SilverHawk

Dock(er) Wok

2026-02-12T14:57:00.000-08:00

Docker is a simple way to create virtualization without having full systems running. I took a moment to figure out how to run docker in relation to Postgres.

Note you need at least three servers for this for it to have any sort of redundancy. One is the "master". You can probably get away with two (run the manager on one of the nodes), but I do not recommend this. Really, you could get away with one server that runs the manager and two instances (I did), but know that if you don't run the docker instances on separate hosts, you've now lost all high availability, because if that server fails, the whole stack will cease to exist.

First, install Docker :

sudo apt-get install docker.io
sudo systemctl enable docker
sudo systemctl start docker

Then, create your swarm (anything recent should have swarm built in, you just need to set it up).


sudo docker swarm init --advertise-addr 192.168.x.x

Next, join a docker worker to the swarm

username@server1:~$ sudo docker swarm init --advertise-addr 192.168.0.3
Swarm initialized: current node (NODE IDENTIFIER) is now a manager.

To add a worker to this swarm, run the following command:

    docker swarm join --token TOKEN_STRING 192.168.0.3:2377

To add a manager to this swarm, run 'docker swarm join-token manager' and follow the instructions.
username@server1:~$

Then, on each worker node that isn't the manager, run the command you were provided (use sudo) :

sudo docker swarm join --token TOKEN_STRING 192.168.1.237:2377

Next is your basic configuration. Each host is relatively identical, with just a few modifications. Each host has 3 different configuration files, pg_hba.conf, postgresql.conf, and pg_ident.conf. Let's start by creating our directory structure.

I created a directory to house all of this so I can clear it out quickly after my learning curve.

mkdir cluster-postgres
cd cluster-postgres
mkdir -p {master,slave-1,slave-2}/config

Next, create our three files for each host :

touch {master,slave-1,slave-2}/config/{pg_hba.conf,postgresql.conf,pg_ident.conf}

With the files created, let's populate them. Here's the gist for each host, but at the end, I'll identify differences as needed.

postgresql.conf

This file contains :

# -----------------------------
# PostgreSQL configuration file
# -----------------------------
#

data_directory = '/data'
hba_file = '/config/pg_hba.conf'
ident_file = '/config/pg_ident.conf'

port = 5432
listen_addresses = '*'
max_connections = 100
shared_buffers = 128MB
dynamic_shared_memory_type = posix
max_wal_size = 1GB
min_wal_size = 80MB
log_timezone = 'Etc/UTC'
datestyle = 'iso, mdy'
timezone = 'Etc/UTC'

#locale settings
lc_messages = 'en_US.utf8'   # locale for system error message
lc_monetary = 'en_US.utf8'   # locale for monetary formatting
lc_numeric = 'en_US.utf8'    # locale for number formatting
lc_time = 'en_US.utf8'       # locale for time formatting

default_text_search_config = 'pg_catalog.english'

#replication
wal_level = replica
wal_keep_size = 512MB  # Adjust this value as needed
archive_mode = on
archive_command = 'test ! -f /mnt/server/archive/%f && cp %p /mnt/server/archive/%f'
max_wal_senders = 3

However, on the slave nodes, lines 28-34, which contain :

#replication
wal_level = replica
wal_keep_size = 512MB  # Adjust this value as needed
archive_mode = on
archive_command = 'test ! -f /mnt/server/archive/%f && cp %p /mnt/server/archive/%f'

are removed.

pg_hba.conf

This file contains :

# TYPE  DATABASE        USER            ADDRESS                 METHOD

host     replication     replicationUser         0.0.0.0/0        md5

# "local" is for Unix domain socket connections only
local   all             all                                     trust
# IPv4 local connections:
host    all             all             127.0.0.1/32            trust
# IPv6 local connections:
host    all             all             ::1/128                 trust
# Allow replication connections from localhost, by a user with the
# replication privilege.
local   replication     all                                     trust
host    replication     all             127.0.0.1/32            trust
host    replication     all             ::1/128                 trust

host all all all scram-sha-256

However, on the slave nodes, line 3 :

host     replication     replicationUser         0.0.0.0/0        md5

is removed.

pg_ident.conf

This file contains :

# PostgreSQL User Name Maps
# =========================
#
# Refer to the PostgreSQL documentation, chapter "Client
# Authentication" for a complete description.  A short synopsis
# follows.
#
# This file controls PostgreSQL user name mapping.  It maps external
# user names to their corresponding PostgreSQL user names.  Records
# are of the form:
#
# MAPNAME  SYSTEM-USERNAME  PG-USERNAME
#
# (The uppercase quantities must be replaced by actual values.)
#
# MAPNAME is the (otherwise freely chosen) map name that was used in
# pg_hba.conf.  SYSTEM-USERNAME is the detected user name of the
# client.  PG-USERNAME is the requested PostgreSQL user name.  The
# existence of a record specifies that SYSTEM-USERNAME may connect as
# PG-USERNAME.
#
# If SYSTEM-USERNAME starts with a slash (/), it will be treated as a
# regular expression.  Optionally this can contain a capture (a
# parenthesized subexpression).  The substring matching the capture
# will be substituted for \1 (backslash-one) if present in
# PG-USERNAME.
#
# Multiple maps may be specified in this file and used by pg_hba.conf.
#
# No map names are defined in the default configuration.  If all
# system user names and PostgreSQL user names are the same, you don't
# need anything in this file.
#
# This file is read on server startup and when the postmaster receives
# a SIGHUP signal.  If you edit the file on a running system, you have
# to SIGHUP the postmaster for the changes to take effect.  You can
# use "pg_ctl reload" to do that.

# Put your actual configuration here
# ----------------------------------

# MAPNAME       SYSTEM-USERNAME         PG-USERNAME

There are no differences here for each node.

Last Configurations

Create the docker "network" :

sudo docker network create postgres-cluster-network

This will print a fairly large alphanumeric ID.

Starting the "master"

At this point in time, start the master node using the following command :

sudo docker run -d --name postgres-master  --net postgres-cluster-network \
-e POSTGRES_USER=postgresadmin -e POSTGRES_PASSWORD=admin123 \
-e POSTGRES_DB=postgresdb -e PGDATA="/data" -v ${PWD}/master/pgdata:/data \
-v ${PWD}/master/config:/config -v ${PWD}/master/archive:/mnt/server/archive \
-p 5000:5432 postgres:latest -c 'config_file=/config/postgresql.conf'

If you haven't downloaded the postgres docker image yet, this will actually cause it to try and install. If you get an error about it already existing because you tried to start it one and it gave you an error :

docker: Error response from daemon: Conflict. The container name "/postgres-master" is already in use by container "6362ca473d3f29b7fe1bf02558f5a871fb6cf6827eff23fe01714f34cb386951". You have to remove (or rename) that container to be able to reuse that name.

Then list the dockers, and delete it :

username@server1:~/postgres-cluster$ sudo docker ps -a
CONTAINER ID   IMAGE             COMMAND                  CREATED          STATUS    PORTS     NAMES
6362ca473d3f   postgres:latest   "docker-entrypoint.s…"   10 minutes ago   Created             postgres-master
username@server1:~/postgres-cluster$ sudo docker rm 6362ca473d3f
6362ca473d3f
username@server1:~/postgres-cluster$ sudo docker ps -a
CONTAINER ID   IMAGE     COMMAND   CREATED   STATUS    PORTS     NAMES
username@server1:~/postgres-cluster$

Now try to start it if it had previous failed again, and it should simply give you a large alphanumeric ID again.

Next, create a replication user :

username@server1:~$ sudo docker exec -it postgres-master bash
[sudo] password for username:          
root@63919c16e356:/# createuser -U postgresadmin -P -c 5 --replication replicationUser
Enter password for new role: 
Enter it again: 
root@63919c16e356:/# exit
exit
username@server1:~$

Just type exit and get back to the prompt.

Now, we can move on to starting the slaves.

Starting the slaves

For each node, run :

username@server1:~$ sudo docker run -it --name postgres-slave1 --rm \
> --net postgres-cluster-network \
> -v ${PWD}/slave-1/pgdata:/data \
> --entrypoint /bin/bash postgres:latest
root@447571e3e11e:/#

This will put you in a bash prompt as "interactive", where you then run :

pg_basebackup -h postgres-master -p 5432 -U replicationUser -D /data/ -Fp -Xs -R

The command initiates replication, and will use the password you specified on the master node when you started that up with the "createuser" command.

Repeat for slave-2 (of course, replacing postgres-slave1 with postgres-slave2, and the slave-1 folder names with slave-2).

Create standby instances

Run the following :

sudo docker run -d --name postgres-slave1 --net postgres-cluster-network \
-e POSTGRES_USER=postgresadmin -e POSTGRES_PASSWORD=admin123 \
-e POSTGRES_DB=postgresdb -e PGDATA="/data"  \
-v ${PWD}/slave-2/pgdata:/data -v ${PWD}/slave-2/config:/config \
-v ${PWD}/slave-2/archive:/mnt/server/archive -p 5002:5432 \
postgres:latest -c 'config_file=/config/postgresql.conf'

Repeat for slave-2 (of course, replacing postgres-slave1 with postgres-slave2, and the slave-1 folder names with slave-2).

Test It

Connect to the master node using :

sudo docker exec -it postgres-master bash

From in here, you can run your psql commands to create databases and manipulate whatever you need.

psql --username=postgresadmin postgresdb

Exit, and then check the other slave nodes by connecting to them (sudo docker exec) and running psql to query any tables you've created and populated with data.

Just Pulley-ing My Leg

2026-01-19T09:21:00.000-08:00

A year or so back, I had heavily modified am early 1940's tiny lathe with the intent of using it to turn pens. There are a number of small lathes that are likely cheaper than the cost put into that one, but for some odd reason, I love the old iron, and did it anyway.

The results were, less than stellar. Indeed, it did work, but I had to take my time because the stepper motor did not have enough torque. And that is with a small pen.

So, enter the next chapter. Someone over on the hobby machinist forum sent an old sewing machine motor to me (think Pfaff 130, and you have the footprint of it), just for the cost of shipping. With the pulley that was on it, I found myself in a bit of a quandary - I needed more modifications to see if this would work. So, I set about creating a new belt tensioning system. I patterned this variation off of a cars alternator - where it is installed and then tightened over an arc-bracket.

I made the bar using my drill press, because, I didn't have a rotary table large enough for the arc (mine is 4", the arc is 8" - but that is due to only having a scrap bar with an 8" arc that was an off cut for the 127-tooth compound gear), and also because I was too lazy to drag out the milling machine onto a cold back porch and cut it. Really, it was because I was way too lazy.

But, it was close enough for what I needed.

So, I drilled a series of holes, then used an endill in the drill press to smooth out the arcs. I drilled a few extra holes in some bar stock, and used a 1/4"-20 bolt to put it all together.

Next, I needed a pulley. So I grabbed another off-cut/drop (2" diameter aluminum), chucked it up, and made a 2L pulley that was as large as this lathe could accept, but small enough to fit (1.950" with the belt). This was faced, drilled, then reamed. It was then turned on the boss, then the flat part of the pulley, then the angles.

The result is a usable 2L pulley to match the one on the sewing machine motor.

I wired the motor to test it, and once that checked out okay, it was time to put it all together.

This immediately feels like more torque than the stepper. But, the proof is always in the pudding. What is the end result? Let's give it a try. I ran through the normal steps preparing a pen blank (this time, I chose a gun-metal "mini" bolt action pen kit from Penn State Industries - I have a lot of family that love the regular ones, so I thought I'd give this more "unique" variant a whack.)

Usually, mark the blank lengthwise to keep grain lined up. In this case, it's small, and I was fixing another pen while I was at it, so this step didn't matter.
Cut off two chunks to fit your pen tubes, about 1/4" longer than the tubes themselves. Again, this was small enough, and I was using a scrap blank I had from another project.
Drill the centering hole for the blank(s).
Glue in the tubes. Many people just use CA glue. I have always loved 5-minute epoxy for this.
Barrel-trim the ends to get it flush with the brass tubes (don't take off brass or it might not fit).
Put the tubes on the appropriate bushings, then on the mandrel, and turn.
Once to slightly larger than size, use the lathe to sand the blanks. Get as high a grit as possible for the best finishes.
Remove from the lathe and assemble according to the instructions.

So, once I got to the lathe part, I grabbed a photo.

They do look fantastic (I had two on there - I was repairing another pen while I was at it). Then, assembly prep :

Then, the final assembly.

While the motor was slightly "under-powered" (It IS a 1/5 HP sewing machine motor), it still did the trick. I do have to take things a little slower with this, but it does mean I don't have to drag out the larger twin for this little thing. I can simply pick it up, clamp it to a work bench, plug it in, and turn a pen.

Finding God in Failure

2026-01-11T12:04:00.000-08:00

Many religions believe we are being punished for Adam's transgression. We do not.

"And now, behold, if Adam had not transgressed he would not have fallen, but he would have remained in the garden of Eden. And all things which were created must have remained in the same state in which they were after they were created; and they must have remained forever, and had no end.

"And they would have had no children; wherefore they would have remained in a state of innocence, having no joy, for they knew no misery; doing no good, for they knew no sin.

"But behold, all things have been done in the wisdom of him who knoweth all things.

"Adam fell that men might be; and men are, that they might have joy.

"And the Messiah cometh in the fulness of time, that he may redeem the children of men from the fall. And because that they are redeemed from the fall they have become free forever, knowing good from evil; to act for themselves and not to be acted upon, save it be by the punishment of the law at the great and last day, according to the commandments which God hath given." - 2 Nephi 2:22-26

Man is fallen. The earth is Telestial - imperfect.

It is this imperfect world that allows us to be imperfect.

Were it not for the imperfection of the world in which we live, we could not know how to hope for a better one. Two verses before the famous "faith" scripture in Ether, we have this little gem.

"Wherefore, whoso believeth in God might with surety hope for a better world, yea, even a place at the right hand of God, which hope cometh of faith, maketh an anchor to the souls of men, which would make them sure and steadfast, always abounding in good works, being led to glorify God." - Ether 12:4

All of us find themselves in times of "hoping for a better world".

Sermon on the mount - after stating that we should pray for those who curse us, states :

"for he maketh his sun to rise on the evil and on the good, and sendeth rain on the just and on the unjust." - Matthew 5:45

Do I mention cancer? Heartbreak? Heart Failure?

By Elder Gong tells the following story.

A Chinese story begins as a man’s son finds a beautiful horse.

“How fortunate,” the neighbors say.

“We’ll see,” says the man.

Then the son falls off the horse and is permanently injured.

“How unfortunate,” the neighbors say.

“We’ll see,” says the man.

A conscripting army comes but doesn’t take the injured son.

“How fortunate,” the neighbors say.

“We’ll see,” says the man.

This fickle world often feels tempest tossed, uncertain, sometimes fortunate, and—too often—unfortunate. Yet, in this world of tribulation, “we know that all things work together for good to them that love God.” Indeed, as we walk uprightly and remember our covenants, “all things shall work together for your good.” - Elder Gerrit W. Gong, Apr 2024, "All Things for Our Good"

Do we want to be "tempest tossed", or to wonder "why me?" How do we avoid this?

"We become our truest, best selves only as we put off the natural man or woman and become a child before God." - Elder Gerrit W. Gong, Apr 2024, "All Things for Our Good"

How do we put off the natural man?

For the natural man is an enemy to God, and has been from the fall of Adam, and will be, forever and ever, unless he yields to the enticings of the Holy Spirit, and putteth off the natural man and becometh a saint through the atonement of Christ the Lord, and becometh as a child, submissive, meek, humble, patient, full of love, willing to submit to all things which the Lord seeth fit to inflict upon him, even as a child doth submit to his father. - Mosiah 3:19

Yield to the enticings of the Holy Spirit.

Create an environment where the world can be quiet. Temple, at home in your personal room, or sitting in your car while waiting.
Commit to follow the direction you are about to receive.
Pray for the enticing of the Spirit of God to show how you can serve someone.
Remain in the quiet environment until you feel that you have something to do.
Obey immediately.
Repeat.

As you put this effort in to interact with the Spirit of God and to follow through, you are becoming a saint through the atonement of Christ the Lord. You become meek, humble, patient, full of love, and willing to submit to all things the Lord seeth fit to inflict.

You are building a trusting relationship with Christ.

"When we trust God and His love for us, even our greatest heartbreaks can, in the end, work together for our good." - Elder Gerrit W. Gong, Apr 2024, "All Things for Our Good"

This trusting relationship does not obliterate our enemies, or push obstacles out of our way.

Shadrach, Meshach, and Abed-nego, would not cede their covenants to a King, and were to be thrown into a furnace of fire. When the King questioned the power of their God, their response was :

"If it be so, our God whom we serve is able to deliver us from the burning fiery furnace, and he will deliver us out of thine hand, O king.

"But if not, be it known unto thee, O king, that we will not serve thy gods, nor worship the golden image which thou hast set up." - Daniel 3:17-18

The second half of that, "But if not", is the crucial component. Whether something happens or not does not cause a break in what our commitment should be.

Remember, we are mortal, living in a mortal world. But when we get lost in that "forest for the trees", pause, to also remember this.

"Even if you cannot always see that silver lining on your clouds, God can, for He is the very source of the light you seek. He does love you, and He knows your fears. He hears your prayers. He is your Heavenly Father, and surely He matches with His own the tears His children shed." - Jeff Holland, Oct 1999, "An High Priest of Good Things to Come"

God knows you, deeply and fully. If he is "endless", pause and realize that his love for you is also endless.

"He is the light and the life of the world; yea, a light that is endless, that can never be darkened; yea, and also a life which is endless, that there can be no more death." - Mosiah 16:9

We must recognize that our timing is never God's timing.

"Some misunderstand the promises of God to mean that obedience to Him yields specific outcomes on a fixed schedule. They might think, “If I diligently serve a full-time mission, God will bless me with a happy marriage and children” or “If I refrain from doing schoolwork on the Sabbath, God will bless me with good grades” or “If I pay tithing, God will bless me with that job I’ve been wanting.” If life doesn’t fall out precisely this way or according to an expected timetable, they may feel betrayed by God. But things are not so mechanical in the divine economy. We ought not to think of God’s plan as a cosmic vending machine where we (1) select a desired blessing, (2) insert the required sum of good works, and (3) the order is promptly delivered.

"God will indeed honor His covenants and promises to each of us. We need not worry about that. The atoning power of Jesus Christ—who descended below all things and then ascended on high and who possesses all power in heaven and in earth—ensures that God can and will fulfill His promises. It is essential that we honor and obey His laws, but not every blessing predicated on obedience to law is shaped, designed, and timed according to our expectations. We do our best but must leave to Him the management of blessings, both temporal and spiritual." - Elder D. Todd Christofferson, April 2022, "Our Relationship with God"

And once you recognize this, you will also recognize that, while we are imperfect and mortal, there is a full regalia of blessings waiting.

"Some blessings come soon, some come late, and some don’t come until heaven; but for those who embrace the gospel of Jesus Christ, they come." - Jeffrey Holland - Oct 1999, "An High Priest of Good Things to Come"

Recommend Jan 2009, Thomas Monson's speech to graduating students called "Great Expectations".

Rudyard Kipling, "If" :

If you can keep your head when all about you
Are losing theirs and blaming it on you,
If you can trust yourself when all men doubt you,
But make allowance for their doubting too;
If you can wait and not be tired by waiting,
Or being lied about, don’t deal in lies,
Or being hated, don’t give way to hating,
And yet don’t look too good, nor talk too wise:

If you can dream—and not make dreams your master;
If you can think—and not make thoughts your aim;
If you can meet with Triumph and Disaster
And treat those two impostors just the same;
If you can bear to hear the truth you’ve spoken
Twisted by knaves to make a trap for fools,
Or watch the things you gave your life to, broken,
And stoop and build ’em up with worn-out tools:

If you can make one heap of all your winnings
And risk it on one turn of pitch-and-toss,
And lose, and start again at your beginnings
And never breathe a word about your loss;
If you can force your heart and nerve and sinew
To serve your turn long after they are gone,
And so hold on when there is nothing in you
Except the Will which says to them: ‘Hold on!’

If you can talk with crowds and keep your virtue,
Or walk with Kings—nor lose the common touch,
If neither foes nor loving friends can hurt you,
If all men count with you, but none too much;
If you can fill the unforgiving minute
With sixty seconds’ worth of distance run,
Yours is the Earth and everything that’s in it,
And—which is more—you’ll be a Man, my son!

Elder Holland spoke about one of his own "travails" - he and his young family had car troubles. When he drove past that spot later in life, how he wished he could tell himself :

"Don't you quit. You keep walking, you keep trying, there is help and happiness ahead. Some blessings come soon. Some come late. Some don't come until heaven. But for those who embrace the gospel of Jesus Christ, they come. It will be alright in the end. Trust God and believe in Good Things to Come." - Jeff Holland, Oct 1999, "An High Priest of Good Things to Come"

And in the same talk, he continued :

To any who may be struggling to see that light and find that hope, I say: Hold on. Keep trying. God loves you. Things will improve. Christ comes to you in His “more excellent ministry” with a future of “better promises.” He is your “high priest of good things to come.” - Jeffrey Holland - Oct 1999, "An High Priest of Good Things to Come"

The Spindle Depth Stop

2025-12-18T14:51:00.000-08:00

I had a small run of parts on the lathe that needed to be the same length. There are a couple of ways to do this, such as :

After creating the part (and making sure both ends are faced, but a wee bit long), measure the length of the part.
Calculate the amount you need to remove by subtracting the final length of the part from your previous measurement.
Re-insert the part into the chuck (or collet), and measure the length of the stick-out from a known reference point.
Subtract the distance you need to remove from this stick out measurement.
Face the part until you reach that distance calculated in step 4.

However, I'm lazy, and that's a little too much effort. I'd rather do that only once, and on a single-run part, I'd have to do that, anyway. The micrometer stop does a bang up job of getting it right on once you've gone through that. But if I have four parts? Uh-huh. I'm too lazy, since I could insert the parts into a depth, while leaving the carriage/cross-slide in the one position.

It's time to make a spindle depth stop - something that locks into the spindle so you have a consistent reference point on all the parts.

Mind you, my spindle has either a collet closer attachment installed, or it has the big fat gear used by the collet, so I can't just throw a standard one in place. But, just in case, my South Bend Heavy 10L (large spindle bore) regular, non-gear depth-stop was also designed. This spindle depth stop can fit all three, just by making the locking spider for all three. That's only two extra parts, allowing this to fit three different configurations. I also wanted to use as many parts that I didn't have to make. Because, again, I'm lazy.

Okay, okay, enough blab. Let's get to it. Grab the materials (I'm going to use CR1018/cold rolled steel).

First, let's get the stainless-steel 3/8"-16 all thread cut to length. I used a 36" bar, and cut it at the 24" mark. No precision necessary here, I just have two variations in case I don't like too much sticking out the back for an extra long part.

Next, I already had the nuts in hand (that's not what I meant, and you know it) for the 3/8"-16 all thread, and a 5/8"-11 nut for the locking lug.

First part to "really" make was the locking lug. Really, it was a simple turning job for the most part. Turning to 0.625" for the threaded end, threading to 11 TPI to match the nut, and then getting the main body set, the internals bored for clearance for the threaded rod (and one end threaded to 3/8"-16 of that internal bore. The hardest part was the "cone" on the end. What I had drawn up in the plans had a 17.615 degree angle, and to be honest, you just need it "close". Using the compound, you can set the angle, then lock that in. Do not change that angle until you have completed your locking spider(s)!

This was my first taper turning, and I went with a standard 17 degree. I don't care how close I am to 17 - it's not moving until I've made the entire thing.

Next up was the washer. It was coming from the same large stock as the spiders, so it's easy to slap in there and turn, face, and bore.

You just might notice that the proportions aren't lining up with the drawing. I actually opted to make this 1/2" thick instead of 1/4" thick, and I bored out 1/4" of the internal bore to match the larger diameter on the locking lug I've finished. This will give me a little more expansion room, and allow me to use less material on the spiders.

Now I just need at least one spider. I'll do the gear-based spider first :

Then, I'll do the collet attachment spider :

Now, I'm not making this next spider, because I have that fat gear that sticks out for the spindle attachment. It means I can't really use it. However, I drew the plans up anyway, just in case someone needed them.

So, I started out creating my two spider blanks. I chucked up the material, faced one end and the diameter in order to get the base set up. I also rough-turned one to get it not quite to dimension (I'm 0.050" oversized on the outside). I did not go all the way to dimension yet because I needed the bore set up so that as stresses are relieved, the outside will then be truly concentric to the bore.

Once I had a close-but-no-cigar dimension on the outside, I turned (pun is not intended) to the inside for the boring process. First, I used drill bits to get it to 0.5" (needed to get to 0.810"), and then I could use the boring bar.

Once the inside bore was complete, it was time to also turn the taper. This is the exact reason why the compound was not moved - it already has the matching taper from the internal part, so I knew it would be a match made in heaven (or at least on the lathe).

With the spider's internal bore/taper and the locking lug fitting as they should, I went back and turned the outside down to the final dimension, and then turned the 0.125" circular relief groove using a rounded insert (could only find metric, and it was close enough). Yes, in the following photo, the lathe is OFF. I refuse to lose a hand. But, I did it to provide a visual on the groove.

After the groove, I removed the chuck from the lathe and ensured I had a snug slip fit into the targeted location. At that point, I could pull it out of the chuck and do a test fit up. It definitely locks in there, but without the expansion slots, it won't truly lock.

I do know that I need to remove more on the outer end (largest diameter face), because I didn't get it short enough to fit on the locking lug. So, into the chuck it will go with some brass shim stock to protect the finish, and then I can face it down to size until I have a perfect arrangement.

It's not often you hit the target width so closely. I thought it was close enough, and assembled everything (not usable until the slots are cut).

Now, the remaining task is to cut the slots into the surface. There are six slots, but using a band saw (my slitting saws are not large enough), it's only three cuts. So, next I need to build a jig to keep it perpendicular to the saw blade, mark the three cuts, and then start slicing. This had my anxiety running high, because there is no way I'm willing to lose a finger. And, with two separate spiders for different dimensions, I need something adjustable for that smaller diameter. I need it to be at least close to parallel to the axis. So, after thinking on it for a month, one morning, I needed to get out of the house, and I ran out and grabbed some 2"x1/2" aluminum flat bar, and simply marked out a few parts. Cut, drilled, tapped for M3 screws (because I didn't want to wake my wife early in the morning), and I had this little thing.

I can thread those two screws in or out, depending on the diameter of the spider end, so there is the adjustability that I needed for the two spiders. It allows me to adjust those two screws until I have a parallel setup, and I can then feed that into the bandsaw blade without fear of something catastrophic. It does fit like I hope.

The next task was the actual cutting of the slots. The jig held up quite nicely, though I should have adjusted the screws part way through. While there, I also drilled relief holes where those slots ended so that there wasn't a hard corner or edge to start cracking on.

I used a triangular (bastard) needle file to deburr the slots (outside and inside), and a round needle file to remove the burr on the inside where the drill bit broke through. But, once that was done, I had a relatively decent part. A quick test fit of everything together....

Now, it's time to try it out.

It locked in there without much of a tightening on that large nut (it could still slip, so I'd likely use a wrench when I really need it). Functioned pretty well!

The second one can now be chucked up and replicated to the dimensions of the first (with the smaller diameter for the collet attachment tube).

Finally Got My Pulse

2025-11-26T20:12:00.000-08:00

Well, we've been dealing with side effects of the cancer for a while. The symptoms were dizzy spells, and not just when changing positions (like standing up). These came while sitting down and staying there.

Officially, we finally got it half-way pinned down to a heart failure diagnosis with a "reduced ejection fraction", meaning the heart wasn't pumping enough blood. When we had tests ran, the numbers came in at 41% of what it should be.

[sigh].

We next did an ablation on the heart to knock off some arrhythmia that had crept into the picture while the chemo was happening. Things were not perfect, but much better. After watching things a bit, we found that the procedure had a definite impact :

It went from about 10-20 spells per day down to about 2-5, with some days not having any at all.

Gotta get the ol' ticker running properly, and this was a good step!

I Drove a Porch

2025-10-24T14:42:00.000-07:00

No, it's not misspelled.

As you age, so do things around you. I knew of a couple who was getting long in the years, and their back porch was becoming rickety. The "porch" was simply a set of welded metal steps. And if I stood on them without moving, they'd rock back and forth. Seriously, it was the only way for me to look like I had any dance moves. So, their daughter volunteered her and her siblings to build them something new. Somehow, I ended up becoming a de facto decision maker.

I started out by making "dad" mark out where it was going to go. While they headed off to take care of some legal things, one of the siblings hubby and I (and their kids) ran over to get started.

We fired up an auger to drill footings - but by the time we got it running, the young lad had already dug three of the holes. Ah, youth and vigor.

We dropped some "Sonotubes" in place, shoved some rocks in the bottom to stabilize it a bit more, and filled them with concrete followed by some 4"x4" post base brackets, and let that set up overnight.

The next morning, we started installing vertical posts. I made the young lady of the group do the bolt hole drilling - she did fantastic. Truth is she did a lot more than that. She worked on the actual decking, the stairs, etc. She ran the drills, the impact drivers, etc.

Part way through, another siblings kids showed up. They just cranked through the project, and did fantastic. Apparently, I was simply there to be beautiful.

But, I was the idiot who climbed onto the roof panels before they were fully anchored in place. Those sheets had to be screwed down from above, so someone needed to, and I've had a good life. I figured I had less life to lose, eh?

Anyway, we put up the railing, and built a "roof line" that was close to the house roof line. Unfortunately, it needed to be "separate" so that we didn't require a permit. We kept it close, but still separate. In one weekend, we got the foundation, the frame, the deck, and the base of the roof built out.

We still need to install the moisture barrier and the roofing itself, but the basic shape is there, and it is much more stable.

Rebuilding my Roller Saw Horse Stand

2025-09-26T13:34:00.000-07:00

My workshop is fairly limited in size. While that can be changing, in the immediate time, I don't have a lot of space. It means I have to be able to set up in a mobile fashion, e.g. I take over the back patio when I need to.

I've used saw horses to set up a large work bench before. And about 6 years ago, I did what I shouldn't do, and I used roller stands (the kind you'd use to allow long boards to roll off of a table saw, jointer/joiner, or a planer). Now, they "roll", which they are supposed to, so the idea 6 years ago was tested to keep them from rolling when needed. I used some 2" PVC pipe, which (when cut lengthwise to allow the arms supporting the rollers to slide into the "pipe") seemed to fit absolutely perfectly on the rollers.

The problem is that over time, the wood split, warped, and encountered some heavy water damage. They became unusable (not quickly, but just over time). I needed something that would last a little longer. So, I made a few purchases, and set to work on "revision 2".

Here's what you'll need :

Two HDPE sheets measuring 3/4" thick x 12" wide x 48" long
Four aluminum angle iron 1-1/2" x 1-1/2", with legs that are 1/4" thick (thick enough to act as a "nut")
48" of 2" PVC pipe, cut into 12" lengths
Eight 2" wide x 1-3/4" thick x 3" long HDPE bars
Sixteen heat-set threaded inserts for assembly screws of your choice
Two thread inserts for "locking pins"
At least two transfer punch screws (4 is preferable)
One countersink screw bit
One drill bit for tapping screws of your choice
One drill bit for clearance of the same screws
One drill bit for the thread "locking pin" inserts
One drill bit for the heat-set threaded inserts
Eight spacers (I used drops from other projects) that need to be the same length - about 1" in length, maybe 1" in diameter
One (or more if you break them) tap
A soldering iron with a heat-set insert tip
A screwdriver

How to :

Identify (and mark) which side of the angle iron is to be the "surface" side.
Drill the tapping-size holes for the screws of your choice into the aluminum angle iron. I used 5 on each surface-side, with spacing at uneven (check that, my CDO/OCD had issues, but I wanted to decrease linear distance the surfaces would be screwed down with). Two pieces of angle-iron are used per "saw horse", and the downward legs will face each other. Orientation is key here.
Tap the holes.
Drill four holes along the downward legs of ONE of those angle iron bars per horse (if building two horses, you'll do two).
Tap these holes.
Use the transfer screws in the downward legs and bring the two pieces of angle together to mark the other set.
Drill the clearance size holes in the second piece of angle aluminum.
Put your spacers in between, and screw the two pieces of angle together into a structural arm.
Center the arm on one piece of HDPE sheet, and mark ONE hole on the HDPE sheet.
Drill that hole for the clearance size.
Bolt the HDPE sheet and the structural arm together with that one bolt.
Mark the opposite end of the structural arm to the HDPE sheet.
Drill with the clearance size, and connect the two together. The two parts should not move at this point from each other.
Use the remaining holes and either use transfer screws, or use transfer punches.
Drill the remaining HDPE holes for structural arm attachment.
Countersink the holes from the top-side of the saw horse table top.
Attach the two together (note, they might come off the table later in this process, see step 27).
Use a hole-saw in a jig (I just built a square "box" I could clamp them to on the drill press, then clamped that box to the drill press table) to cut an arc on one side lengthwise. This should be about 2-3/8" in diameter, but I just used a 2" hole saw and made two different runs so that they matched.
Drill two holes (threaded insert sized)that will be perpendicular to the arc all the way through the bars. These will allow threaded inserts to be set in from each end, and allow the pipe to be screwed to these blocks.
Use threaded inserts on the back side of the HDPE blocks as well (allowing the blocks to be attached to the HDPE table top surface).
Heat-set inserts into the blocks.
Once cooled, use transfer screws on the non-arc side of the blocks to mark them onto the HDPE table top.
Drill (clearance sized) holes for each block.
Countersink the remaining holes.
Attach the blocks (permanently - we shouldn't have to remove them again) to the table tops.
Use the PVC pipe as a guide and mark the arc pattern into the angle aluminum structural bar. If the pipe clears, you don't need to do anything.
If the pipes do not clear the structural arms, remove one piece of aluminum angle at a time, and grind or cut the clearance so that the pipe will not touch the aluminum. I gave between 1/16" to 1/8" of clearance.
Once clearanced, re-install the piece of angle aluminum and tighten the screws.
Repeat for the other one as necessary.
Use a table saw to cut the pipe lengthwise. A word of caution - PVC pipe is under tension, and when it is fully cut through, will snap together and pinch your saw blade. This is extremely dangerous. There are other options, but if you go this direction, find a good way to hold the pipe securely without having your fingers close, go slowly, and if it pinches the saw blade and brings it to a halt, do NOT panic. Just turn off the saw, and work things free.
Use the table saw to cut the pipe to the width to clear the roller arms.
Use the transfer screws to mark the pipe in conjunction with the HDPE bars.
Drill the holes in the pipe and countersink so that you can attach the pipe to the HDPE bars.
Drill the CENTER of the table top to take the threaded inserts.
Countersink the remaining clearance holes (PVC pipe, table top, etc).
Screw everything together.
Heat the inserts up and screw them into the table. These allow you to put in a pin and prevent the workbench from sliding on the HDPE.

The screws themselves won't be holding the weight; they just keep things from moving around and shifting. It means that you don't need shear strength in picking your hardware.

The end result can just clip together when you need a "saw horse" mode from your roller stands. When you need roller stands, just take them back apart.

They are quite nifty for the "mobile" or "temporary" workshop.

Hammo Can - Radiated Nerdism

2025-09-11T11:41:00.000-07:00

I'm a HAM. Not that I'm funny, seriously, I'm a HAM. An amateur radio operator. I function as an emergency contact for neighborhood issues when infrastructure is down, as well as a point of contact and support for the city (I work with both data networks as well as voice networks in the radio environment). I also work for a local healthcare company on their radio connection network in my spare time. It's not what I do for a career, this is what I do for kicks and giggles. My current portable radios are poor, my wife doesn't want an awful antenna sticking up from the house, and that leaves either a permanently mounted in the vehicle (it's been broken into once or twice, so things are not left in there) or a portable base station with some power.

Old ammunition storage boxes are often referred to as "ammo cans", so taking it into the world of radio and adding an "h" on the front is not a long stretch. I am NOT the coiner of the phrase "hammo can". I refuse to take credit on coining the terminology because, frankly, I'm not that smart.

Under the circumstances outlined above, a "hammo can" is the perfect option. It will be light weight, portable, and not tied to a vehicle or property. I can easily set one up under a canopy in the yard if it's needed. So, I finally decided (11 years after first obtaining my radio license) to venture into the "hammo can" world. So, that proverbial bullet was bitten, and I started ordering parts :

The radio is a TYT's "TH9800D". It's tried and true, and a lot of HAM operators swear by it. Theoretically, it will fit a Yaesu FT-8900, but I cannot confirm that without a radio. This one also requires three m4-0.7 flat head machine screws, one being 16mm long and the other two being 12mm long.
The container is a 50 caliber ammo can off of Amazon. I ordered it in purple (they have other colors, and this one visually separated itself from my grab-and-go case of hard drives).
The battery is a 12v LiFePo battery to be able to run this station for an extended period of time.
A power manager (AtemPower AP100BM) is used in this build (and some switches so I can completely disconnect the battery for storage).
Battery terminals (2 negative [black], one positive [red])
Random switches and battery connectors for easier charging.
Heat-set inserts for m3 screws - while it started out with some m2 screws threading directly into the plastic, it was eventually changed to all-m3 (except for the radio itself) for consistency.
27 m3x10mm machine screws (add four if you print this with the two-part top plate).
2 m3x10mm pan head screws (for the radio face-plate-to-battery box).
Heat shrink tubing
1/4"-hole circular connectors
1/4" female spade connectors for the battery
A PL-259 90-degree elbow
4" Cable with PL-259 ends (male and female)
Mounting connector for a PL-259 for the top plate
Wiring (I used a radio adapter that had long cables, and I just used the extra cable for all my power needs)

So, I have the hardware to have a "hammo can", but it is kinda ugly piled into a chaotic mess, and I don't want to re-organize everything before I can use it in an emergency. I need this thing to be ready to run with only setting up an antenna. The simpler it is in an emergency, the better.

Design

That lead me on a course to print an insert for the "hammo can". I mapped out a lattice, and started gluing things together in FreeCAD. Please note - by the end of this write-up, the structure will change. There is no lattice for rigidity, and it is all handled completely by brackets. This will give more parts to print, but reduce the over all time it takes to print it. But, let's skip this digression and just move to the original design.

There are compartments to slide the battery into, the radio into (faceplate goes on top), and the power management adapter. So, I also loaded the "thing" to thingiverse if anyone else wants to go down this course of action.

I think this will work sufficiently, but it looks like it's just about 2 weeks to print (17 days). Plus, without a brim, it looks like it just barely fits on a Creality CR-10s build plate (with a brim, it did not fit). I do want to make one change - I'd like to make the radio area a bit more "modular" so you could potentially use other radios, so this could become a more viable option to build one-off hammo cans. We'll see if I do.

I got into gear by printing the power module bezel (I was using it as the test platform for interrupted filament prints because the chassis is going to be 2.5 reels of filament (yes, that much). I don't want to be so attentive for two weeks while this prints, so an empty filament reel should result in a pause. That's two prints that I can use to make sure it is all running properly. The power bezel print did abort, while I figured out the issues with the filament sensor housing, and I printed a new housing and had it working okay, so I could run some more tests.

I managed to get the bezel printed with interrupted prints, so I think I can test this once I get the time to do it (mid-to-late August at this point).

The bezel plate came out great.

The bezel plate will just screw in to the chassis. There is one section I'm going to have to modify for a power shunt, so I might as well get the radio caddy a little more modular, too. I spent a few hours getting things set up for a modular setup.

I then built out the shunt plane, as well as a modular radio, and also printed out the switch plane while I was at it.

The switch panel looks good.

There are two "generics" in the thingiverse file - one radio module and one shunt plane. But I also have the shunt plane for the AtemPower monitor and the TH-9800D radio loaded in there. The "generics" are in case you want to try something a wee bit different (e.g. fit in a different radio, for example).

I then started to print the shunt plane for the AtemPower (not the generic one - I don't need one of those). Of course, the shunt plane print failed four times. I re-leveled the bed, changed the nozzle, and then finally gave up and ordered a new PEI printing bed plate (I ordered a few, just because this could use up more consumables like the print bed). Then I started printing it yet again. I'm getting close to running out of filament on one spool, so this is perfect - it's going to give me a real test of the filament sensor break.

The print was good - it finished right on time. The shunt plane finally finished, and did not warp. Now, if this should look funny, it's because I wanted to finish using up my silver filament. I knew during one of these parts I would have an opportunity to prove out the filament sensor interruption, and it came about 2/3's of the second to last layer. So, the base is "black", but the last layer is "silver". It's going to be buried in the can, so I'm not at all worried about a discrepancy. This print took about 4.5 hours with the filament reload.

Now for the radio module. Since I had the silver installed, I simply just went for that. It will all be hidden on the back end, so I'm not terribly worried about presentation here.

Again, I created a "generic" in the thingiverse file for this, but also a specific TH-9800D variation for my implementation. If someone wants me to do a non-generic module for a different radio (if they fit in the same basic size), they can always donate a radio and I'll try to design it to make it happen. Again, in theory, this may fit a Yaesu FT-8900 or 8800 radio (since that is what the TYT TH-9800 is based on), but no guarantees. Send me one, and I'll make sure it does work. I probably won't send the radio back when I'm done.

I only printed the TH-9800D radio bracket, not the generic, so I cannot vouch for the generic one's functionality, but the specific one came out well. I assume the generic one is fine, since the TYT-specific one was based on it. This print took about 16 hours.

I realized that my measurements were 0.039" off on the screw spacing, so I adjusted the module, but rather than take 16 hours to reprint, I just modified it using a drill. The radio body fits in using one m4-0.7-16mm screw on the top and two m4-0.7-12mm screws on the bottom.

Next came the print that would take a few weeks, the one I've been testing the filament change in anticipation of - the actual chassis. Seeing that my previous prints have come out slightly undersized, I did a quick google search, only to find out that an STL doesn't have real dimensions in it, so your slicer can only guess so much. This means that you have to adjust the size in your slicer before you slice it. That leads me to the following. The chassis itself is 6.102" (154.9908mm) wide, 11.217" (284.9118mm) long, and 6.315" (160.4010mm) tall. Okay, on to the printing. My slicer seemed to have it pegged on dimensions well, so, off to the printer. I know I should adjust for shrinkage, but meh. We'll see.

This should start out in silver filament (because, again, that's what is on there), but should switch to black before the top layers when the silver filament runs out (it takes 2.5 rolls of filament, so it will definitely run out) - and that means it should look just fine on the outside. My CDO (OCD in the right order?) will know the bottom is silver, but it will still function.

Yeah, that's a few layers of silver below the black. Sure, it's going to look whacky, but I don't care. (The fact that I am yet again referencing this "issue" probably means a second print later.)

This is going to take a while. The slicer said 13.4 days, but the print isn't coming across at the same rate. After 33 hours of printing time (started on the 11th at 8:54 AM), I've calculated each percentage point as taking 249 minutes, or 4.15 hours. The stats claim that it will take a grand total of 17.3 days (on the dot, which puts it at finishing on the 28th at 2:54 PM) - which is what 415 hours equates to.

At the 20% mark (8:10 AM this morning on the 14th, 72 hours and 9 minutes in, I revisited the stats. The target date shifted from the 28th at 2:54 PM to the 26th at 8:46 AM. The per-percentage time went from 249 up to 277, which makes me question my math in one of these two check points. We'll see what happens. It does look good, and taking random photographs will help me post-process the supports, clearing this up. As I work through it, the photos will definitely come in handy.

The next check point, 7:37 AM on the 15th, was 27% complete, with about 1.888" of height, 94 hours and 36 minutes into this (5,676 total minutes so far). On this one, I took the time to also measure the chain guard distance to the stabilizer, because I know the stabilizing rods get in the way of the x chain. That came out as 6.146" of height remaining, and the calculations are to 5.664" of height left (which falls a lot longer than the 6.315" of total height).

This 21,022.2222 minutes in total, or 14 days, 14 hours, and 22 minutes. That brings the new estimate of completion to the 25th at 11:16 PM. Of course, there are some filament changes that will have to happen in there somehow.

My next check point came in a few times on the 18th, while I started paying closer attention to the filament running out for a second time. I clocked it at 171 hours and 51 minutes at 12:51 PM, and had the filament changed at 172 hours and 15 minutes at 1:39 PM, approximately 24 minutes in print time, but 48 minutes apart. That means that the print timer itself is total time, and does not include time between filament being out to the change time - or print pauses. That's good to know.

So, the estimate for the next filament change is the 24th at 8:46 AM. At that point, we should be about 97% to 98% done with this, and we'd only have 1 day and 7 hours, and 53 minutes remaining (25th at 4:39 PM is the new completion target, depending on how much time I waste forgetting I need to change the filament).

It is now 62% done, but at 59% done, it was 203 hours and 45 minutes into the print, and 4.374" tall. If this was fully linear, it would mean that the print would be 7.414" tall, but we know that the print should only be 6.315", so we have nearly a full inch shorter. This is a solid indicator that the print percentage is NOT in layer count, it is either in over-all time, or in total distance of filament printed. At this rate, it should be 5 days, 21 hours, and 35 minutes remaining, a target completion of the 25th at around 9:50 PM.

Now, if only I could wait patiently (and hope we don't have a power outage in the next 5 days, 13 hours, 22 minutes, and 46 seconds).

Bugger.

9 days in, at 66.3% complete, I walk into the room (I'm not living in the same room as this thing for a week) and hear odd noises.

The nozzle is being scraped across the top of the print!

Apparently, I printed a 6.35"x6.1"x4.6" pile of a roll of filament. I aborted the print job, and grudgingly went back to the drawing board. [sigh].

No fancy lattice work, no huge set of prints, and all of the modular parts will still fit. This also includes other improvements that were re-developed int6o the new structure (I thought I would see interference on the corners with the curves in the can itself, so I added some fillets to work around that). So, off to print a few parts for the printer itself, then to start with the four new parts :

The bracket that ties the shunt plane to the radio, the battery, and to the top plate
The extra corner bracket (where I have some space that could be used in the future if I wanted to add networking components or Raspberry Pi's)
The new top plate with mounting tabs
Battery box

All said and done, these four new parts should take approximately 5 days to print - a massive difference in time down from the 15 days it was going to take. Additionally, these new brackets can also be adjusted to make this even more customizable. The top did not fit with a raft - so I created a 2-part top plate. However, I really don't like the visual, but it does still function, so if you are making one of these without a large printer, it will still work.

The top parts did get plastic-welded together after being bolted with screws. Here it is without the plastic-weld, which is still functional.

And the visual... really isn't so bad! The top panel pieces took 40 hours and 40 minutes on the CR-10s with 100% infill and a raft, which is why the surface is a bit sub-par. Here it is with the top panels all combined, and with no wiring.

Next was the battery box. Started at 9:44 Aug 31, followed by the brackets. And, after all that, it finally started to come together!

Assembly

Install the heat-set inserts in ALL the parts. If you used the two-part top plate, you will use an extra four of the inserts, and an extra four of the m3x10mm machine screws.
Next, install the radio itself to the radio module.
If you printed the two-part top plate, combine the two top plates together with four screws. Plastic-weld the two parts together (*this is purely optional*).
Fit the switch plate to the top plate, and screw together.
Install wiring on the radio body side of the switch (you won't be able to get to that after you install the radio).
Temporarily install the radio face plate (the part that comes with your radio) to the adapter block, but do not tighten.
Install the radio face plate to the top plate.
Measure where the radio adapter bracket (what came with your radio, not the block itself) sits, and remove the block and radio face plate.
Re-install the radio adapter bracket without the radio face plate itself to the adapter block.
Install the radio face plate adapter block and bracket to the top chassis plate again.
Shift the radio face plate adapter until it matches what was measured above in step 7.
Tighten the radio face plate adapter bracket (what came with the radio) to the adapter block.
Remove the adapter block, and add the cables required for the radio face plate, and install the face plate to the adapter block/bracket combination.
Install the battery box to the top plate through the radio face plate connector.
Install the radio face plate to the top chassis, running cables into the cavity underneath the box where they can be connected to the radio.
Fit the power module into the bezel, then attach it to the top plate.
Install the battery connectors for the top plate.
Install the radio module itself to the top chassis.
Install wiring to the battery, and to the power module, the remaining switch terminals, the battery terminals, the battery box (add the connectors for the battery, too) and the radio itself. In my case, the negative battery was wired to the switch, then to one battery terminal. I essentially used the battery posts on the top chassis plate for the interconnection of wires.
Install the extra bracket.
Install the shunt plane bracket (not the adapter).
Insert the battery to the battery box, and strap it in using a velcro strap (not really required, but recommended to prevent the battery from bouncing around) to secure it in position.
Route wires around the shunt plane bracket.
Install the radio module shunt plane adapter and shunt plane.
Program the radio using your computer (again, while optional, I highly recommend it).
Drop the block into the hammo can.
Weather strip around the top edge to lock the panel into position and secure the chassis.

I did use a strap to hold the battery when I started putting it together. It rattled around and went flying through the air and landed on toes when pulling the chassis in and out of the can while assemling.

I finally got the last few parts to finish putting this together (the one remaining battery terminal, and the antenna connector).

It does fire up, and things appear to be operational!

Your radio should now be ready for an antenna and an emergency. A good source of information is the "Radio Reference" database (https://www.radioreference.com/db/aid/7771) where you can look up frequencies specific to your location. I would recommend (if your radio supports these frequencies) :

27.065 (CB Channel 9) - Emergency
52.525 (6-meter) - FM emergency band for ham radio operators
138.225 - FEMA disaster relief (primary)
146.52 (2-meter) - non-repeater, HAM frequency
Local Fire, Police, and Search and Rescue (frequencies will vary by location)
156.75 - maritime weather alerts
156.80 - maritime distress
163.4875 - national guard emergency channel
163.5125 - armed forces disaster preparedness
168.55 - federal civilian agency for emergencies and disasters
243.00 - military aviation emergency
462.675 (70cm) - GMRS emergency channel 20

Once set up, connect your antenna and give it a test run.

Summary

Now, while I don't like the finish of the top surface, nor the fit, I think this will be operational for what I need, and I secretly think this is awesome. I'm hoping to surround the top plate with weather stripping, and put some custom gaskets beneath the switch and power bezels for a bit more moisture-sealing, but with the gaps under the radio faceplate itself, this is definitely NOT water tight. I'd like to revisit an antenna (if I can find a purple pipe wide enough to protect the antenna as well as act as a bit of a stand) to see if I can get that a bit more appealing.

With all of these prints together, it was time to assemble the whole thing.

And with that, we have a "hammo can" (portable base station) for emergency use!

Old Hard Drives

2025-08-22T10:40:00.000-07:00

When you visit family, and someone hands you an old hard drive from 22 years ago and asks for a data recovery because it could have your moms letters on it, you don't question it. Grab that hard drive and take it.

So I started a little journey. I popped it into my PATA USB case, and plugged it in. It did not register with the system (I'm using Linux so I should at least see a USB interface get logged, even if the hardware has failed).

It spun up. There were no clicks (so the hardware had not failed). I dug into the logs - nothing, indicating the interface board was likely the problem.

So, I took the interface board off. It was a Western Digital WD400 from 2003 (I told you it was 22 years old). I grabbed my Western Digital WD400 from my case and started to swap the interface boards temporarily to commence a recovery.

One screw changed between them. There are associated with the WD400 product line, but actually have different model numbers. I'd wager that if it wasn't for that screw being in the wrong place and not being able to be anchored down properly, it would have been perfect, yet this would not be.

I could have 3D printed a bracket that fit across and held the interface circuit board down, but I just put the original back in, and looked at the jumpers. Old PATA/IDE hard drives had a jumper on them for "Master", "Slave", and sometimes one or two other options (such as "CS" or "Chip Select"). There were no more than two IDE drives on the same IDE cable because of the electronic design. One would be designated the "master", and the other a "slave". In later years, they added a "chip select" option to try and make it a little more plug-and-play.

I had the thought I should move the jumper from Master to CS, and suddenly the old interface loaded when it was reconnected. Ah hah!

It loaded no files (an empty filesystem). The next step is always to run a recovery, just in case a new filesystem was put over the top of an old one. So, I ran :

photorec /d Documents/recovery /dev/sdc

This dumped a few web files (CSS, XML, and HTML), a few executables (EXE and DLL), a few TXT files (that turned out to be nothing more than configuration files, registry files and some text extracted from executables). Two files, a WAB and an EDB, both stood out to me, because a WAB is supposedly a Web Address Book, and EDB is an Exchange Database. Turns out, they were simply mislabeled data files for the executables.

No, the drives did not have any letters from my mother. But it's better to know than not to.

Doh! A Delicious Dish of Dumb Dumb!

2025-08-05T12:21:00.000-07:00

Alright. I'm only posting this because of the blatant error in my own reverse engineering. I had a Coleman pop-up canopy, with fancy "push button" locks and unlocks. It's really not bad - it's survived for 10 years with only minimal issues. (Two years ago, one of the locking mechanisms on one leg wouldn't release.) Here it is in all it's glory (still in it's bag).

This year, that same locking mechanism released (perhaps a better term is it fell apart). Since I had silver filament and these were gray, I thought, what could I lose? I'll see if I can draft and model up a replacement part, and print it.

I started out by taking the few remaining parts off to see what was broken. This is a pretty ingenious device. I'm impressed with the design (but not so much the implementation).

The broken part is circled in red above. A few more photos, just to make sure :

It is time to create the model. I whipped this up in FreeCAD :

About 50% done (I'm doing 100% in fill so it is solid), I picked up one of the parts, and realized - I didn't look deep enough. There are two locking tabs on each end (I got both of those), and... recesses in the mating part where there were supposed to be four more locking tabs on the sides.

Really? I play the part of an idiot VERY well, sometimes. I'm missing the ones in green here. Notice, I had them listed as an alignment pin. Looking closer to the original, and these locking tabs had broken off (three of the four were down to the plate, one was half height, and I built all four off of that one).

So, back to the drawing board. I simply copied one of the two tabs and positioned the four new instances.

Okay, we can try printing this once more. Here we go. 8 hours and 38 minutes.

Alright. The overall length should be 2.904", but the final print came out at 2.890. In Cura, it shows 73.6, which is about 2.898", so we're losing about 0.008" in the print, and Cura is also not getting sizes from the STL file with much accuracy from FreeCAD. That should be just fine, really, as long as I remember to adjust the size by 100.48% before slicing.

It looks good, the clean up was good (the aborted print gave me an idea of how this was going to be cleaned up), and with some fine adjustments (hobby knife and a Dremel for some inside work), it works exactly as I'd hoped.

I went to install it, and it simply didn't lock together. So, that led me on a re-do, and I added the toggle cover with a new base that screwed together (no tabs). This particular variation uses screws to bring the cover and the main housing together, but uses the same toggle switch itself, the same spring, and the same spring assembly.

Note, be careful how you print this - if you print side-to-side instead of top-to-bottom, the locking tab that holds the housing to the actual metal upper leg will break, and then you have to superglue it into place (ask me how I know).

I have my canopy back!

Moving Badge Access to your Finger

2025-07-24T10:00:00.000-07:00

I work for an entity that implements physical security for pretty much everything. Workstations in the office can even have their own badge reader (through "Imprivata") to "login". One group of employees we cater to absolutely uses these systems to keep notes, and to be able to move from one office to another quickly and effectively. The badge access systems make "authentication" quicker and slicker.

I wanted to make it even easier than reaching for my badge to tap in and out of the computer, so I bought a T5577 RFID writable ring off of Amazon, and an RFID reader/writer. At first, I got fairly frustrated that neither of those tools did what they were supposed to do.

Then I made one adjustment, and things were suddenly magic.

So, what did I do?

Use the mode key to select "HID 125KHz".

Once the HID 125KHz is selected, insert the badge, and press "read".

The device will then say to put the T5577 card in and press "Write". This will take some trial and error to find the right position to hold the ring for writing.

Once you have a successful position and press write, it will tell you it was successful.

With that, it was time to test it. At a workstation, I used my primary badge, and locked the session. Then I tried with my hand (ring) to unlock, and the light indicator went to "green" from "red", showing success.

The workstation unlocked, angels sang from the skies above, and I was suddenly being served pina coladas on a beach in Tahiti. (Well, one of the three was true.)

Now, it did NOT work at the office doors - those apparently require encryption on the badges that this simply cannot do. I'll have to think around that.

Folding Desk

2025-05-27T07:21:00.000-07:00

There have been a few times where I've had to temporarily set up "shop" (my workstation) in different places around the house. For example, when we have a visitor and the bedroom immediately next to the workstation is occupied and the time they wake up doesn't match.

These situations have usually had me splitting the expanding table and attaching a dual monitor stand in between the two haves, then pushing back together and setting up in the kitchen.

I have a folding drafting table, and I wanted something similar I could use for a workstation. I was severely tempted to just purchase a butcher block top along with a folding stand, but if you know me, if I'm going to put one together, I might as well go all the way. I had some aluminum extrusion laying about for a different project that ended up not being used, so my brain said.... why not?

I set about designing the desk in my head, and then just threw it together. Here's what you'll need (with modifications from what I did because you likely will make changes). Since I was doing this with 2020 series extrusion, my bolts and nuts were all M5.

3 fir boards 2x6 that are 10 feet long
Biscuits
Wood glue
2x 48" horizontal extrusion rails for the main body (mine were shorter, and you'll see overlap on the folding ends later, which messes with my CDO, er, OCD)
2x 43" slanted extrusion rail to stabilize the main body (these form the "x" in the main body and stabilize the whole thing - one will need to be cut into two pieces, but do that after the first one is in place)
4x 23.5" horizontal extrusion rails for the folding ends
2x 27" slanted extrusion rails to stabilize the folding ends
6x 26" vertical extrusion rails for the height of the desk
8x 45-degree brackets for the extrusion rail
8x 135-degree brackets for the extrusion rail
16x corner hardware brackets
76 bolts (I had 40 for the piano hinges, so your mileage may vary depending on your hinges)
76 washers
76 extrusion nuts
2x Piano hinges about 20" long
12x extrusion end caps

Tools you'll need :

A square
Allen keys for the bolts
A biscuit cutter
A chop saw for the boards
A band saw to cut the extrusion to length (and some angled cuts)
Panel clamps (I used quick clamps along with off-cut boards to keep flat and Saranwrap to keep those off cuts from sticking to the glued up surface)
Sanding pads
Planes (optional)

First, cut the extrusion to the lengths you need.

Second, for the slanted pieces, cut the folding ends to 45 degree angles (angles are NOT parallel for these, but opposite of each other). The main body angles ARE parallel.

Third, start bolting each frame it all (except for the second cross bar in the main body) together.

Once it is all together, you can mark where to cut the second cross bar in the back, cut it, and install that one half. Then you can line up the other half and mark where that one needs to be cut as well, cut it, and then install it.

With that done, you can then mark where the top bar of the main body connects to the cross bars. Then, remove the top bar on the main body, and cut those two pieces off, then reinstall the two pieces. This should keep the whole thing somewhat stable. That middle section shouldn't be discarded yet - in fact, you will cut a little more off and re-install it about 6" below where it originally was. This should give you clearance. (Pictures don't show it offset until you get to the end of this post, and they don't show the second cross bar installed, so be aware).

Fourth, install the hinges (but don't tighten quite yet). With the folding area in place, open one end perpendicular to the main body and get it square and vertical to the floor. While in that position, tighten the open ends' hinges to secure it in place. Repeat for the other folding end.

You should now have a folding rack and desk base.

Yes, notice the folded ends overlapping. This is because my desk was less than the 48" I've now recommended above. But it's still "functional". Anyway, back to the build.

Next, cut the fir 2x6 boards in half. This should fit the span of the desk. Line them up in position on the desk surface, and mark and cut for biscuits.

Then, glue it all up and wait.

With the surface glued up, it's time to finish it. Granted, you don't have to, but you won't be using the desk to write with the way 2x6's have curved corners. I'd recommend you use the hand planer to get it flat, and then sand and finish. However, you may want to simply pour epoxy in (sealing the ends, of course) until you build it up far enough to have a flat surface. My unfinished, but usable workstation :

I used lacquer spray to somewhat seal the bottom (I did not spray the top), and then taped the bottom and edges so I didn't have epoxy run all over. While there, I also hammered some small gears, screws, etc into the surface. All of that was filled with a metallic bronze epoxy mixture, then sanded. I went to 220 grit top, bottom, and edges, then went 800 grit on the top and edges. Then I lacquered the whole thing (yes, even the bottom again). Once done, I hit it briefly with 2000 grit sand paper to knock off the little spikes that form (this spray can lacquer isn't the greatest and smoothest finish after spraying).

That gave me a phenomenal surface that won't poke holes in paper, and won't snag on clothes. It also has a great look to it. The epoxy will look black or dark unless the light hits it at a specific angle, so everything blends together in a fantastic way.

I will have to build the nipples for the bottom of the table, the threaded tubes to make it a standing desk, and the fixtures for the top of the stand itself. I whipped up the specs for machining the 12 remaining parts required (for what I'm doing at least - you COULD adjust with a top hinge and make it only a fully folding sitting desk. I want the option of standing, so I'm manufacturing a few pieces yet.

I built a fixture plate that bolted to my lathe faceplate (so that I could screw the next part onto it and just make it match), and then did some bandsaw work so I could reduce the amount of turning work I'd have to do, and started cranking through.

I needed four nipples for the bottom of the desktop, four threaded mounts for the frame, and four pipes with one end threaded to the same thing as the frame mount. Plus, I had to shorten some M5 screws.

Once these were created, I could install the frame-to-table parts.

Then it to flip the desktop over, line it up, then install the nipples to the bottom.

A final test before I install the standing spacers :

And the final spacers :

I did slap the laptop on it and work in standing position for an afternoon. This will work well for me when I need to be on the move.

Next, I didn't want to lose the aluminum "standing" rods, so I printed a storage case that simply connects in and locks them when not in use.

Now, I went to use this for a little bit, and there have been some issues. When it was in "sitting" mode, the height just wasn't quite right. So, I trimmed off lengths 3" (I already updated the measurements above). When I tried to use a dual monitor stand that snugs onto the desktop itself, the rear, top bar was in the way. I added a second cross bar in the main body (pic below), and I cut out the top cross bar and moved that down 6" so I could have clearance for the monitors. When I threaded the standing-mode bars on, some wouldn't go on all the way. I ran a threaded die onto them and cleaned up the threads (I had to use an adjustable die so that I could get it perfect, and that meant running it over the threads a few times while adjusting it tighter in between until it fit just right.

I'm loving this thing! It is now perfectly level, it has excellent stability (though it is a bit heavier), and it all just works as it was designed.

Now, if I have a mobile Internet connection, we could go just about anywhere and work, and my family can wander off to see sights while I work. This would be good, as long as the family doesn't get eaten by bears without me. That would be bad. Perhaps if I'm eaten by bears along with them, but not without them.

Chess - Ah, Nuts!

2025-05-27T06:33:00.000-07:00

A group of chess enthusiasts returned to their hotel room after a tournament, and were still standing in the lobby while bragging about the defeats of their opponents. After an hour of loud nerds, the manager told them they needed to disperse.

"But, why?"

"Because," he replied, "I can't stand chess nuts boasting in an open foyer."

Perhaps, I had better apologize for the pun. I'm not sorry for it, but I will apologize.

I'm one of "those" kinds of people who loves chess, cars, and I do spend my fair share of time on Pinterest. I spend little precious time playing video games, and a lot of time just exploring. And, while in Pinterest, I have seen numerous instances of chess sets made out of fasteners (you know, nuts-and-bolts kinds of things).

Now, a little insight into me in general - I have chess sets. I have chess sets my dad gave to me when I was 8 or 9 years old. I have marble chess sets, glass chess sets, wood chess sets (a folding one, and a wall hanging one), and I had a friend make a set for my dad in a machining class out of aluminum (it's anodized blue and clear - it is a beauty). I'd even made my own by using an existing set and creating molds for the pieces, and casing my own out of green and gold resin (I made a folding chess board back in 8th grade wood shop class - it is definitely rough, but I'm keeping it as is). I have sets for Lord of the Rings, and I'd like some other resin-cast variations. Simply put, I LOVE chess sets!

But I do not have my own metal set, and these looked intriguing. I wanted one. The materials are not a cheap way to get into the game - trust me, you can find MUCH better options out there (that are even regulation sizes and a bit more portable for things like chess in the park) as cheap as $21.

So, I decided to try my hand at it, just because they look pretty darn cool. Here's what I ordered :

10 brass M10-1.5 20mm bolts (for 8 pawns and 2 knights)
2 brass M10-1.5 25mm bolts (for 2 rooks)
4 brass M10-1.5 40mm bolts (for 2 bishops, a queen and a king)
10 brass acorn nuts for the pawns and bishops
4 brass M10-1.5 nuts (simply to lock some other parts to the bolts)
3 "castle" M10-1.5 nuts - if you find these in brass, I'd be surprised, so you are looking for actual castle nuts with a gold tint - it's close enough for the boys my wife dates
2 brass lamp rings (these were M10-1.0, because you will not find these with a 1.5 thread - but that's okay, because you're just going to tighten them on anyway (without the nuts to lock them) as far as they'll go.
1 brass flange nut (for the queen)

Repeat the purchase for the other color (I went with black to give it a good contrast - for machining, you basically have brass/gold, black oxide, polished stainless, or just a steel). I've thrown it all into an amazon list if you are extra lazy. Just know this is not exactly cheap.

Now, it is time for assembly.

For a pawn, just take 8 of the 20mm bolts, and tighten the acorn nut on as tight as you can by hand. Maybe a little tighter, but this isn't structural, so you don't need to worry about it being really tight.

For a rook, spin onto the 30mm bolt a regular nut, followed by one of the "castle" nuts (they are officially called "castellated" nuts, and they are used with cotter pins to keep a nut from spinning off of a thread for shafts in things like cars). Put the peaks upward, and when you have all rooks top nuts in the right position, bring the other nut against it to lock it in place.

For the knights, just thread the lamp rings on. The improper thread size will lock them in place fairly quickly.

Bishops, treat JUST like the pawns, but with the 40mm bolts.

For the king, just like the rook but with a 40mm bolt.

And for the queen, use a flange nut upside down and a lock nut.

The set doesn't look bad at all in the above photo, (it is incomplete, e.g. the queen is not done). Additionally, I assembled some black pieces, because contrast between black and gold is just... beautiful.

If you look closely enough, there are still some parts missing (black rooks and king are both missing the "crown", brass queen is missing the flange nut). Once those arrived, it was relatively easy to complete the set.

Still, I'm debating in my head around making a chess board using gold and black without machining one from stock and going through the black oxide process, etc. Still working on that thought, but it might ultimately come to that.

A Dimensional Awakening - An Old 3D Printer (Creality CR-10s)

2025-04-10T15:49:00.000-07:00

Since the cancer scare of 2024, a host of old projects are just itching to be scratched [off the list]. One of those projects was my old alarm system needing new hardware around the garage door (the wood it was made out of seems to disintegrate over time). So, it was time to break out my old Folgertech Prusa i3.

I couldn't get it upgraded to handle ABS, which is where I wanted to go for these parts. I ran through updates, I adjusted beds, I even bought an enclosure for my Prusa. ABS was no bueno.

So, I hit the local classifieds and bought an old Creality CR-10 for $75 with hopes of getting it to be my ABS-dedicated machine. It was extremely dirty, gross, and misaligned. Plus it was a CR-10s, not a CR-10 (note, not a CR-10 4S - that extra digit and the space make the "s" mean something totally different). This "s" means it's supposed to be "smart", and has a filament sensor on it which allows it to detect when filament is running out. There were a few broken parts on it, and a few customizations (e.g. the fan shroud was a dual 5015 fan that was a simple, but really elegant design).

First, filament sensor housing had snapped and was loose. There was a spare piece of filament jammed into it to bypass this, which was okay to get this thing back online.
Second, the extruder cooling fan shroud had cracked. The previous owner had glued it together using superglue glue, but this had failed under temperature.

The housings for both failed parts were after custom, and printed with PLA, which doesn't hold up to heat very well and considering sits around the extruder with high temperatures, is not an ideal material for a fan housing. But, I think I can deal with a fan that is slightly loose while I use the printer to replace some of it's own parts.

So, I cleaned it up, bought an enclosure, bought a new build surface (the gold PEI plates work best - the black one stuck a little too well and shredded after a few uses because I couldn't get the parts off), and leveled it out. I started playing with settings and ended up at 255°C on the nozzle and 95°C on the bed. That worked very well for the first few prints (I used real, usable and needed parts as my test prints, such as x-chain links for small parts).

I then backed down to 250°C on the nozzle and 90°C on the bed. I'm using a brim to keep things as glued to the PEI as possible (again, the gold magnetic PEI build plates seems to be the best as they release when the temperature cools enough, but don't hold onto it like a snake locking jaws in a bite).

I have successful prints. Now that I know it's ready to repair itself and it functions for what I need, I bought some 6mm extrusion cover in black, and changed this thing to start blacking this thing out.

I have a full x-axis cable support chain ready for a fan shroud housing. It's time to do the filament detector housing, as well as the fan shroud to connect the x-axis support chain up to.

Bugger. I can't find the original custom parts online. I was going to use those because I KNEW they worked.

So, I had to design and build my own. Out came the calipers, micrometers, and gauge pins to map out and chart the holes. This was a few weeks of printing and trial and error while I got everything mapped. First was the filament detector housing. I started with someone elses STL, loaded into FreeCAD, and converted it to a solid to start playing with it. I angled it up a bit to reduce tension on the filament inside the housing, and ended up with something I liked.

Since I stole someone else's design, I could re-use the same lid/cap they had designed. I printed it out, and... it warped. Another bugger! But, it was still usable enough to get that part online. So, onto the machine it went. And it was good... for a little bit. When I started installing the z-chain cable management, not so much. So, back to the drawing board. I came up with this thing posted to thingiverse.

Next, I designed the replacement fan shroud. I used the same tools (FreeCAD, calipers, micrometers, gauge pins) to map out the existing, known-to-have-been-working design, plus I wanted to add the x-axis chain cable support. I got to where I liked it.

I tried printing the fan shroud, and... it warped so bad it was unusable. [sigh].

So, I turned off the cooling fans, and reprinted.

There was MUCH less warping (but there was still warping around). I cleaned it up, and tried to get it installed....

Fantastic! The third time was the charm! To finish my x-axis chain, I just need to print the vertical bracket and install it! This printer is essentially better than when I got it. The broken parts have been repaired, and additional parts are being completed. The final fan shroud is available on thingiverse under the name of "Creality CR-10s Dual 5015 Fan with X-Chain Interface" :

https://www.thingiverse.com/thing:7006624

Started working on the cable management (I'm doing all three axis). One thing I don't like about the x-chain is that if you are using tie-rods to stiffen it up, you DO lost z-axis. I'll have to address that later.

So, once the cable management was all complete, it is a pretty solid printer. I really do like this thing.

Now, about that filament sensor.... the next itch that needs scratching is the upcoming "hammo can". It's an insert for a .50 caliber ammo can that sets up a mobile HAM radio station. It doesn't fit on this Creality CR-10s with a brim, but the bottom plate is large enough that it shouldn't need one, and without the brim, it just barely fits. That's going to take 2 weeks to print on this server (and 2.5 roles of 1kg filament), but it's now doable.

Pfaff! A New Machine for Me!

2025-04-08T15:36:00.000-07:00

305.076.08

I have an amazing aunt (well, my fathers aunt, but I claim her). Every once in a while, she contacts me about a family heirloom. Well, she knows exactly what I do for kicks and giggles, and when she realized her family had quite a few of these sitting around, she immediately contacted me and asked if I wanted one.

My first thought was "it's the wrong family", but it didn't take much to twist my arm. Truthfully, I was twisting my own arm.

I ran and grabbed it.

It is a Pfaff 130. The machine itself is in surprisingly excellent shape. It doesn't rotate very well right now completely around, which indicates that something is likely bent inside, or it just needs a good cleaning. The belt has disintegrated, but other than that, it appears to be well taken care of as there's not a spot of rust on it. Kudos to that side of the family for taking such good care of their machines.

First things first, I need to try to identify this a little more than just using the model number off the front, because they were made for quite a while. I found the serial number on the inside, "5 123 825" :

A not-as-easy-as-expected search (took a few manipulations of key words, but that is all), revealed a page out there where Brenda Dean of ISMACS' locale released a basic chart

https://ismacs.net/pfaff/pfaff_manufacture_dates.html

And that tells me precisely that my new toy was manufactured in 1954.

I went through this thing, and I ordered a belt (some websites linked to this 14-3/4" belt) #305.076.08. It did NOT fit. Instead, I opted for a 15-1/2" belt, and with some adjustment, it works beautifully. The motor and electrical are good to go, and everything has been oiled!

This will let me sew mild leathers for car seats, canvas for carrying cases, etc. I'm excited for this new toy!

Port-A-Lign - Misappropriated For Another Use

2025-03-01T11:34:00.000-08:00

I have a need. There were a number of un-identified reamers that I need to officially know their sizes. My milling machine doesn't have the reach, and I don't want to chuck up long bars of stock sideways to be drilled. The best option I came up with was the old Craftsman Port-A-Lign. It's led to some newer variants, but those were mostly plastic, and I struggled with those thoughts - the Craftsman has the drill chuck centered between the rods, and that makes me think it will be a better option to abuse.

Starting out, I need more length (some of these reamers are much longer than jobber drill bits). I grabbed the diameter of the rods for this as 0.496" :

Next up was to identify the spindle that the chuck sits on. It's a 3/8-24 thread, but I'm missing part, and also the chuck is a 3/8" chuck. I needed to be able to handle a 3/4" diameter reamer, so a new chuck was in order, and none of those have a 3/8-24" thread mount. That means I need to make a new spindle, with a JT3 mount on the end for the 3/4" chuck.

I pulled the spindle out. Two clips hold that spindle in place :

I found a 9mm socket fit perfectly on the end to drive that spindle out. I got my measurements :

(Not written down is the JT3 taper end, which is 1.2188" long, and the two ends are 0.7461" and 0.8110", or a taper-per-foot of 0.639", and an angle off of center of 1.5251 degrees.)

I chucked the new drill attachment into the lathe (1/2" straight shank, with a JT33 end for a 3/4" drill chuck), and cut a groove down to 0.467" (inside diameter for the new e-clip). Next the shaft went into the freezer, and the chuck went onto a heater vent to get a good temperature differential for assembly.

So, I have a fully functional Craftsman Port-A-Lign with two interchangeable chucks, the original chuck, and a 3/4" banger for use in first drilling, then reaming a hole. It's only to test reamer sizes, remember? Sheesh, my memory has really taken a hit with this chemo thing.

Re-Visiting an Old Friend - ESP32-WROOM-32

2025-01-07T17:02:00.000-08:00

There are three sections :

ESP32 Code
GPIO message translator
Nagios Interface

I've had an ESP32-WROOM-32 sitting around for a while, and I thought I'd throw it back into action as an MQTT data ingress point for my home system. It's been a LONG time (about 8 years, I think) since I was implementing one of these things. I needed the following :

Base it on MQTT (message brokering) to reduce complexity of the tools required
DHCP compatible on-network presence for a drop-in, rapid security network
Ease of configuring or adapting a new sensor into the system
Robust enough that it could actually run and not encounter a memory leak

Add into the mix that I don't use the Mac any longer, and suddenly, I'm starting from scratch. I started out following Espressif's guide, but ran into a snag straight out of the gate, and that was I couldn't get the stupid thing to compile in the Espressif environment. Bugger. I have to go back to the Arduino IDE.

So, back into the Arduino IDE, slap some code together (usual copy-and-paste from things like Stack Overflow, the forums, etc).

ESP32 Code

I did have an issue where the ESP32 dropped off the network after a few hours. That almost killed my DHCP requirement out of the gate, and then one night, I had a dream about lines of code. A quick modification in the code to actually check if the ethernet had disconnected, and I had something that stayed with me a bit longer. The code turned out to be :


/**
 * A simple firmware for sending MQTT queue messages based on GPIO changes
 * Written (well, hack and slashed from examples) by Joe Lewis, December 18, 2024
 */

/*
 * Some defaults
 */
//#define DEBUG_WIFI
#define USE_EXTERNAL_RES_PINS_34_39   1
#define DEFAULT_MQTT_CLIENT_PREFIX    "mqttgpio-"
#define DEFAULT_MQTT_BROKER           "mqttserverhostname"
#define DEFAULT_MQTT_TOPIC            "GPIO"
#define DEFAULT_INIT_MESSAGE          "init-"
#define DEFAULT_MQTT_MSG_PREFIX       "pin-"
#define DEFAULT_MQTT_PIN_SEPARATOR    "-"
#define DEFAULT_MQTT_PORT             1883
#define DEFAULT_WIRELESS_NETWORK_NAME ""
#define DEFAULT_WIRELESS_NETWORK_KEY  ""
#define HALF_SECONDS_BEFORE_ETHDOWN   60

typedef struct {
  int             p;
  int             v;
  int             lastSteadyState;
  int             lastFlickerableState;
  unsigned long   lastDebounceTime;
} pin_t;

#include "driver/gpio.h"
#include <soc/gpio_struct.h>

// the trailing 0 is for the end of the list
#ifdef USE_EXTERNAL_RES_PINS_34_39
#define NUMBER_OF_PINS  16
pin_t pins[NUMBER_OF_PINS] = {
   {  GPIO_NUM_0, 0, LOW, LOW, 0 }, // possible boot time issues
   {  GPIO_NUM_1, 0, LOW, LOW, 0 }, // possible boot time issues
   {  GPIO_NUM_2, 0, LOW, LOW, 0 }, // possible boot time issues
   {  GPIO_NUM_3, 0, LOW, LOW, 0 }, // possible boot time issues
   {  GPIO_NUM_4, 0, LOW, LOW, 0 }, // possible boot time issues
   {  GPIO_NUM_5, 0, LOW, LOW, 0 }, // possible boot time issues
   { GPIO_NUM_13, 0, LOW, LOW, 0 },
   { GPIO_NUM_14, 0, LOW, LOW, 0 },
   { GPIO_NUM_15, 0, LOW, LOW, 0 },
   { GPIO_NUM_16, 0, LOW, LOW, 0 },
   { GPIO_NUM_32, 0, LOW, LOW, 0 },
   { GPIO_NUM_33, 0, LOW, LOW, 0 },
   { GPIO_NUM_34, 0, LOW, LOW, 0 }, // has a pre-built pull up resistor
   { GPIO_NUM_35, 0, LOW, LOW, 0 }, // has to have external pull up resistor soldered in
   { GPIO_NUM_36, 0, LOW, LOW, 0 }, // has to have external pull up resistor soldered in
   { GPIO_NUM_39, 0, LOW, LOW, 0 }  // has to have external pull up resistor soldered in  //USED
};
#else
#define NUMBER_OF_PINS  13
pin_t pins[NUMBER_OF_PINS] = {
   {  GPIO_NUM_0, 0, LOW, LOW, 0 }, // possible boot time issues
   {  GPIO_NUM_1, 0, LOW, LOW, 0 }, // possible boot time issues
   {  GPIO_NUM_2, 0, LOW, LOW, 0 }, // possible boot time issues
   {  GPIO_NUM_3, 0, LOW, LOW, 0 }, // possible boot time issues
   {  GPIO_NUM_4, 0, LOW, LOW, 0 }, // possible boot time issues
   {  GPIO_NUM_5, 0, LOW, LOW, 0 }, // possible boot time issues
   { GPIO_NUM_13, 0, LOW, LOW, 0 },
   { GPIO_NUM_14, 0, LOW, LOW, 0 },
   { GPIO_NUM_15, 0, LOW, LOW, 0 },
   { GPIO_NUM_16, 0, LOW, LOW, 0 },
   { GPIO_NUM_32, 0, LOW, LOW, 0 },
   { GPIO_NUM_33, 0, LOW, LOW, 0 },
   { GPIO_NUM_34, 0, LOW, LOW, 0 } // has a pre-built pull up resistor
};
#endif

// Important to be defined BEFORE including ETH.h for ETH.begin() to work.
// Example RMII LAN8720 (Olimex, etc.)
#ifndef ETH_PHY_TYPE
#define ETH_PHY_TYPE  ETH_PHY_LAN8720
#define ETH_PHY_ADDR  0
#define ETH_PHY_MDC   23
#define ETH_PHY_MDIO  18
#define ETH_PHY_POWER -1
#define ETH_CLK_MODE  ETH_CLOCK_GPIO0_IN
#endif

#define ESP_INTR_FLAG_DEFAULT 0
#define DEBOUNCE_TIME         50

#include <ETH.h>
#include <WiFi.h>
#include <PubSubClient.h>
#include "esp_mac.h"


char  mqtt_topic[255];
char  mqtt_broker[255];
char  mqtt_msg_prefix[32];
char  mqtt_client_id[32];
int   mqtt_port;
int   delayCountSinceLastSend;
int   sendCount;
char  wifi_ssid[128];
char  wifi_password[128];
bool  use_wifi;
bool  eth_connected;

WiFiClient espClient;
PubSubClient mqttClient(espClient);

typedef struct GPIOEvent_t {
  pin_t       *pin;
  GPIOEvent_t *next;
} GPIOEvent_t;
GPIOEvent_t *events;

/*
 * GPIO Functions
 */

void sendMessage(char *msg) {
  if (!mqttClient.connected()) {
    reconnect();
  }
  mqttClient.publish(mqtt_topic,msg);
}

void processEvent(GPIOEvent_t *e) {
  char  msg[255];
  /*
  int   new_value;
  new_value = gpio_get_level((gpio_num_t)e->pin->p);
  if (new_value != e->pin->v) {
    e->pin->v = new_value;
    */
  e->pin->v = gpio_get_level((gpio_num_t)e->pin->p);;  
    sprintf(msg,"%s%s%d=%d",mqtt_msg_prefix,DEFAULT_MQTT_PIN_SEPARATOR,e->pin->p,e->pin->v);
    sendMessage(msg);
//  }
}

static void IRAM_ATTR gpio_isr_handler(void* arg) {
  pin_t *p = (pin_t*)arg;
  GPIOEvent_t *e = (GPIOEvent_t*)malloc(sizeof(GPIOEvent_t));
  e->next = events;
  e->pin = p;
  e->pin->v = gpio_get_level((gpio_num_t)e->pin->p);
  events = e;
}

void reconnect() {
  // Loop until we're reconnected
  int retries;
  retries = 3;
  while ((retries > 0) && (!mqttClient.connected())) {
//  while (!mqttClient.connected()) {
    retries--;
    Serial.print("Attempting MQTT connection...");
    // Attempt to connect
    if (mqttClient.connect(mqtt_client_id)) {
      Serial.println("connected");
    } else {
      Serial.print("failed, rc=");
      Serial.print(mqttClient.state());
      Serial.println(" try again in 5 seconds");
      // Wait 5 seconds before retrying
      delay(5000);
    }
  }
}

/*
 * Ethernet-specific stuff
 */
// WARNING: onEvent is called from a separate FreeRTOS task (thread)!
void ethernetOnEvent(arduino_event_id_t event) {
  switch (event) {
    case ARDUINO_EVENT_ETH_START:
      Serial.println("ETH Started");
      // The hostname must be set after the interface is started, but needs
      // to be set before DHCP, so set it from the event handler thread.
      ETH.setHostname(mqtt_client_id);
      break;
    case ARDUINO_EVENT_ETH_CONNECTED: Serial.println("ETH Connected"); break;
    case ARDUINO_EVENT_ETH_GOT_IP:
      Serial.println("ETH Got IP");
      Serial.println(ETH);
      eth_connected = true;
      break;
    case ARDUINO_EVENT_ETH_LOST_IP:
      Serial.println("ETH Lost IP");
      eth_connected = false;
      break;
    case ARDUINO_EVENT_ETH_DISCONNECTED:
      Serial.println("ETH Disconnected");
      eth_connected = false;
      break;
    case ARDUINO_EVENT_ETH_STOP:
      Serial.println("ETH Stopped");
      eth_connected = false;
      break;
    default: break;
  }
}

void readConfiguration() {
  Serial.println("readConfiguration() not implemented, skipping (we use defaults)");
}

void setup() {
  pin_t   *pin;
  uint8_t mac_addr[6];
  char  *mqtt_pin_message,*tmp_message;
  Serial.begin(115200);
  events = NULL;
  strcpy(mqtt_topic,DEFAULT_MQTT_TOPIC);
  strcpy(mqtt_broker,DEFAULT_MQTT_BROKER);
  strcpy(mqtt_msg_prefix,DEFAULT_MQTT_MSG_PREFIX);
  delayCountSinceLastSend = 0;
  sendCount = 0;
  mqtt_port = DEFAULT_MQTT_PORT;
  use_wifi = false;
  // read the configuration file if we can
  readConfiguration();
  if (strlen(wifi_ssid) > 0) {
    // wifi network is defined, use it
    use_wifi = true;
    strcpy(wifi_ssid,DEFAULT_WIRELESS_NETWORK_NAME);
    strcpy(wifi_password,DEFAULT_WIRELESS_NETWORK_KEY);
    Serial.print("Starting connecting WiFi, connecting to ");
    Serial.print(wifi_ssid);
#ifdef DEBUG_WIFI
    Serial.print(" / ");
    Serial.print(wifi_password);
#endif
    Serial.println("...");
    delay(10);
    WiFi.begin(wifi_ssid, wifi_password);
    while (WiFi.status() != WL_CONNECTED) {
      delay(500);
      Serial.print(".");
    }
    Serial.println("WiFi connected");
    Serial.println("IP address: ");
    Serial.println(WiFi.localIP());
    esp_read_mac(mac_addr,ESP_MAC_WIFI_STA);
    sprintf(mqtt_client_id,"%02X%02X%02X%02X%02X%02X",
            mac_addr[0],mac_addr[1],mac_addr[2],
            mac_addr[3],mac_addr[4],mac_addr[5]);
    strcpy(mqtt_msg_prefix,mqtt_client_id);
  } else {
    use_wifi = false;
    Network.onEvent(ethernetOnEvent);
    ETH.begin();
    while (eth_connected == false) {
      delay(50);
    }
    Serial.print("Using Ethernet, IP address : ");
    Serial.println(ETH);
    esp_read_mac(mac_addr,ESP_MAC_ETH);
    sprintf(mqtt_client_id,"%02X%02X%02X%02X%02X%02X",
            mac_addr[0],mac_addr[1],mac_addr[2],
            mac_addr[3],mac_addr[4],mac_addr[5]);
    strcpy(mqtt_msg_prefix,mqtt_client_id);
  }
  mqttClient.setServer(mqtt_broker,mqtt_port);

  gpio_install_isr_service(ESP_INTR_FLAG_DEFAULT);
  for (int x = 0;x < NUMBER_OF_PINS;x++) {
    pin = &(pins[x]);

    // set up the pin
    gpio_reset_pin((gpio_num_t)pin->p);
    
    //set the direction
    gpio_set_direction((gpio_num_t)pin->p,GPIO_MODE_INPUT);
    
    if ((pin->p != GPIO_NUM_34) &&
        (pin->p != GPIO_NUM_35) &&
        (pin->p != GPIO_NUM_36) &&
        (pin->p != GPIO_NUM_39)) {
      gpio_set_pull_mode((gpio_num_t)pin->p,GPIO_PULLUP_ONLY);
    }
    
    // run on any edge
    gpio_set_intr_type((gpio_num_t)pin->p,GPIO_INTR_ANYEDGE);
    
    // add the handler
    gpio_isr_handler_add((gpio_num_t)pin->p,gpio_isr_handler,pin);
  }
}

void loop() {
  Serial.println("start sending events.");

  if (events == NULL) {
//    if (sendCount > 0) {
//      delayCountSinceLastSend = 0;
//      sendCount = 0;
//    }
//    if (delayCountSinceLastSend > HALF_SECONDS_BEFORE_ETHDOWN) {
//      if ((use_wifi == false) && (eth_connected == true)) {
//        ETH.end();
//      }
//    } else {
//      delayCountSinceLastSend++;
//    }
    delay(500);         // no events, let's just pause
  } else {
    if ((use_wifi == false) && (eth_connected == false)) {
      ETH.begin();
      while (eth_connected == false) {
        delay(50);
      }
    }
    while (events != NULL) {
      GPIOEvent_t *event = events;
      events = events->next;
      processEvent(event);
      free(event);
//      sendCount++;
    }
  }
  mqttClient.loop();
  // at some point, perhaps an esp_restart() might be needed.
}

I have NOT modified the code from the network. I did not add WIRELESS information into those, but if I wanted to use wireless, I could. I'm running this on a POE network, so I might as well use that.

Essentially, the basic code starts up the ethernet (it would use wireless if I defined those fields), and then uses the MAC address (I'm refusing to override it) as it's client name.

Additionally, there is no configuration for a GPIO pin other than "listening on this pin". With these Olimex POE devices on the ESP32-WROOM-POE, pins 34, 36, and 39 are all needing to use an external pin grounding (e.g. add a resistor to the 3v pin) if you are going to use them. Aside from that, it's plug-and-play. Change the hostname, compile the code, and upload it to your devices.

Then, deploy a Raspberry Pi with mosquitto (the Linux MQTT server, and install the development libraries, too) with the same hostname you threw into your ESP32 code. That likely needs a static IP address (either configure the device with a static, or configure your DHCP server to always assign the same IP address).

With that, you simply need the two following services.

GPIO translator - this is a tool that subscribes to a channel of your choosing, and if something matches, it converts that into a message for another channel. For example, A3C1FFD901-PIN33=1 would get translated to "gate=OPEN". There was a very deep fear that I had to keep track of states for more complex things like a garage door where you had an OPENING, OPEN, CLOSING, and CLOSED states with two sensors, before I realized I didn't care, mapping the states would already take care of that. I DID need to swallow duplicate events because the ESP32 would send the same message a few times (you COULD bypass that using a debounce function, but why when I could potentially swallow those duplicates here?). Turns out to function pretty great. Code is below.
Nagios-to-CMD file - this was fairly simple in comparison to the last one. All it did was listen on an MQTT channel, and parse a message, then dump it into the Nagios CMD file used for passive or external checks.

So, you have to make sure Nagios is set up to process external checks, define checks with no actual check (set passive_checks_enabled to 1 in the service definition for your sensor), set command_file (it might already be set) and check_external_commands to a "1" in your global Nagios configuration file, and restart Nagios. Take note of your host name inside of your Nagios config for these services - you'll need it in one of the configs later. Also, make sure the directory and file specified in command_file is writeable by the user you will be running these things as.

GPIO message translator

First, let's look at our MQTT GPIO message translator. This is yet another hack-and-slash, I mean, copy-and-paste, chunk of code. The code :


/**
  * @file sentinel_translator.c
  *
  * @brief Mqtt client used to translate ESP32 MQTT messages to other messages
  * 
  * @date 19-Dec-2024
  * @copyright GNU General Public License v3
  *
  */

#include <stdio.h>
#include <stdlib.h>
#include <unistd.h>
#include <stdarg.h>
#include <string.h>
#include <semaphore.h>
#include <stdint.h>
#include <time.h>
#include <pthread.h>
#include <errno.h>
#include <stdbool.h>
#include <syslog.h>

#include <mosquitto.h>

#define	MAXIMUM_WORKING_QUEUE_LENGTH	128

#define	DEFAULT_MQTT_PIN_SEPARATOR	"-"
#define	DEFAULT_MQTT_INPUT_TOPIC	"GPIO"
#define	DEFAULT_MQTT_OUTPUT_TOPIC	"SENSORS"
#define	DEFAULT_MQTT_SERVER		"localhost"
#define	DEFAULT_MQTT_PORT		1883
#define	DEFAULT_CONFIG_FILE		"/etc/sentinel/translator.conf"
#define	DEFAULT_SYSLOG_IDENTITY		""
#define	DEFAULT_SYSLOG_FACILITY		1
#define	DEFAULT_SYSLOG_OPTIONS		LOG_NOWAIT|LOG_PID
#define	MAX_LENGTH			255
#define	__SHOW_MOSQUITTO_INFO__		1
#define	DEFAULT_MQTT_CLIENT_NAME	"sentinel-translator"

char	config_file[MAX_LENGTH];
char	mqtt_input_topic[MAX_LENGTH];
char	mqtt_output_topic[MAX_LENGTH];
char	mqtt_server[MAX_LENGTH];
int	mqtt_port;
char	mqtt_pin_separator[MAX_LENGTH];
int	log_level;
char	*program_name;
char	syslog_identity[MAX_LENGTH];
int	syslog_facility;
bool	syslog_started;
int	syslog_options;
bool	daemonize;
char	*last_message_sent;
struct mosquitto	*mosq = NULL;

typedef struct gpio_rule_t gpio_rule_t;
struct gpio_rule_t {
	char			*mac;
	char			*state;
	char			*translation;
	gpio_rule_t		*next;
};
typedef struct working_queue working_queue_t;
struct working_queue {
	// Maximum length of the queue in number of entries
	unsigned int max_queue_size;
	// Pthread mutex for controlling the queue access
	pthread_mutex_t access;
	// Conditional variable queue is empty. Informs the waiting thread
	// that the queue is empty.
	pthread_cond_t empty;
	// Conditional variable queue is not full. Informs the waiting thread
	// that new entries can be written in the queue
	pthread_cond_t not_full;
	// Conditional variable queue is not empty. Informs the waiting
	// thread that there are some entries in the queue
	pthread_cond_t not_empty;
	// Head of the FIFO queue. Worker thread process the head entry first
	int head;
	// Tail of the FIFO queue. New entry is appended at the tail of the
	// queue.
	int tail;
	// Current number of entries in the queue
	int number_of_entries;
	// Pointer to memory reserved for the queue
	void *entry;
	// Size of one queue entry in bites
	int entry_size;
};
typedef int (*do_work_f)(void *work_entry);
typedef struct worker worker_t;
struct worker {
	// FIFO queue for storing the received payloads from MQTT messages
	working_queue_t working_queue;
	// Worker will run in this thread
	pthread_t working_thread;
	// Stop the processing of the working queue items
	bool stop_working;
	// Function that will be called for each queue entry. MQTT payload
	// processor.
	do_work_f do_work;
};
void add_work_entry(working_queue_t *working_queue, void *working_entry);
void logMessage(int logLevel,char *msg,...);
static void *worker_thread(void *worker_thread_arguments);
void worker_clean_up(worker_t **worker);
working_queue_t *mqtt_message_queue;

gpio_rule_t		*rules;

/**
 * @brief This function will be called by the worker thread for
 * processing each entry from the queue of MQTT message payloads.
 * 
 * @param[in] process this entry from the queue
 */

void send_translation(char *msg) {
	logMessage(LOG_INFO,"sending translation to MQTT topic %s: %s",mqtt_output_topic,msg);
	mosquitto_publish(mosq, NULL, mqtt_output_topic, strlen(msg), msg, 2, false);
};

int isLastMessage(char*msg) {
	// if it's the same last message, do not send
	if ((last_message_sent != NULL) && (strcmp(last_message_sent,msg) == 0)) {
		logMessage(LOG_INFO,"swallowing a duplicate message (%s) as we've had multiple",msg);
		return 0;
	}

	// if we've already done a check, clean up
	if (last_message_sent != NULL) free(last_message_sent);

	// store this one
	last_message_sent = (char*)malloc(strlen(msg)+1);
	strcpy(last_message_sent,msg);
	return -1;
};

int process_message(void *message) {
	time_t		local_time;
	int		rules_found;
	gpio_rule_t	*rule;
//	struct tm	tm_result;
//	char		time_stamp[32];

	//Build the timestamp header
//	local_time = time(NULL);
//	localtime_r(&local_time, &tm_result);
//	strftime(time_stamp, sizeof(time_stamp), "%d.%h.%Y %H:%M:%S", &tm_result);

	char *e = (char *) message;
	rule = rules;
	rules_found = 0;
	while (rule != NULL) {
		if ((strcmp(e,rule->state) == 0) && (isLastMessage(message) != 0)) {
			logMessage(LOG_DEBUG,"process_message(): %s matched to '%s'",message,rule->translation);
			send_translation(rule->translation);
			rules_found++;
		};
		rule = rule->next;
	};
	if (rules_found == 0) {
		logMessage(LOG_ERR,"ERROR: Unknown message payload (%s), missing configuration item?",message);
	};

	fflush(stdout);
	
	return 0;
}


/**
 * @brief Call back function for received MQTT message.
 * 
 * Libmosquitto thread will call this function for every received MQTT message.
 * It extracts the payload info from the MQTT message and writes it into the
 * working FIFO queue.
 * 
 * @param[in] pointer to libmoquitto MQTT client instance
 * @param[in,out] pointer to the data defined by the Libmosquitto user/caller
 * @param[in] points to the received MQTT message 
 */
void my_message_callback(struct mosquitto *mosq, void *userdata, const struct mosquitto_message *message)
{
	char		*e;
	char		msg[255];
	e = (char*)malloc(strlen((char*)message->payload)+1);
	strcpy(e,message->payload);
	add_work_entry(mqtt_message_queue, message->payload);
}

static void clean_up_libmosquitto(struct mosquitto *mosq) {
	mosquitto_destroy(mosq);
	mosquitto_lib_cleanup();
}

int create_worker(worker_t **worker, unsigned int working_queue_size, unsigned int working_queue_entry_size, do_work_f do_work) {
	if (*worker) {
		//Worker already created/exists.
		return -1;
	}
	pthread_attr_t attr_thread;
	int rc;
	*worker = malloc(sizeof(worker_t));
	(*worker)->working_queue.head = 0;
	(*worker)->working_queue.number_of_entries = 0;
	(*worker)->working_queue.tail = 0;
	(*worker)->working_queue.entry_size = working_queue_entry_size;
	(*worker)->working_queue.max_queue_size = working_queue_size > MAXIMUM_WORKING_QUEUE_LENGTH ? MAXIMUM_WORKING_QUEUE_LENGTH : working_queue_size;
	(*worker)->working_queue.entry = calloc((*worker)->working_queue.max_queue_size, working_queue_entry_size);
	(*worker)->stop_working = false;
	(*worker)->do_work = do_work;
	//Init the objects for synchronisation of the threds
	pthread_mutex_init(&((*worker)->working_queue.access), NULL);
	pthread_cond_init(&((*worker)->working_queue.not_full), NULL);
	pthread_cond_init(&((*worker)->working_queue.not_empty), NULL);
	pthread_cond_init(&((*worker)->working_queue.empty), NULL);
	pthread_attr_init(&attr_thread);
	pthread_attr_setdetachstate(&attr_thread, PTHREAD_CREATE_JOINABLE);
	//Inherit all of the thread scheduling properties from the calling thread.
	pthread_attr_setinheritsched(&attr_thread, PTHREAD_EXPLICIT_SCHED);
	//Start the worker thread
	rc = pthread_create(&((*worker)->working_thread), &attr_thread, worker_thread, *worker);
	if (rc != 0) {
		strerror(errno);
		worker_clean_up(worker);
		rc = -1;
	}
	return 0;
}


void add_work_entry(working_queue_t *working_queue, void *working_entry) {
	pthread_mutex_lock(&(working_queue->access));
	while((working_queue->number_of_entries) == working_queue->max_queue_size) {
		pthread_cond_wait(&(working_queue->not_full), &(working_queue->access));
	}
	// Place the new work entry on the working queue
	memcpy(working_queue->entry + working_queue->entry_size * working_queue->tail, working_entry, working_queue->entry_size);
	working_queue->tail = (working_queue->tail + 1) % working_queue->max_queue_size;
	working_queue->number_of_entries++;
	pthread_cond_signal(&(working_queue->not_empty));
	pthread_mutex_unlock(&working_queue->access);
}


void *worker_thread(void *worker_thread_arguments) {
	worker_t *worker = (worker_t *) worker_thread_arguments;
	void *working_entry = NULL;
	while(1) {
		pthread_mutex_lock(&(worker->working_queue.access));
		/* Check if the queue is empty. */
		while ((worker->working_queue.number_of_entries == 0) && (!worker->stop_working)) {
			/* Wait till the other side signals that there a new item on the queue. */
			pthread_cond_wait(&(worker->working_queue.not_empty), &(worker->working_queue.access));
		}
		if (worker->stop_working) {
			pthread_mutex_unlock(&(worker->working_queue.access));
			if (working_entry) {
				free(working_entry);
			}
			pthread_exit(NULL);
		}
		if (worker->working_queue.number_of_entries != 0) {
			working_entry = malloc(worker->working_queue.entry_size);
			memcpy(working_entry, worker->working_queue.entry + (worker->working_queue.head * worker->working_queue.entry_size), worker->working_queue.entry_size);
			worker->working_queue.head = (worker->working_queue.head + 1) % worker->working_queue.max_queue_size;
			worker->working_queue.number_of_entries--;
		}
		if (worker->working_queue.number_of_entries < worker->working_queue.max_queue_size) {
			pthread_cond_signal(&(worker->working_queue.not_full));
		}
		if (worker->working_queue.number_of_entries == 0) {
			pthread_cond_signal(&(worker->working_queue.empty));
		}
		pthread_mutex_unlock(&(worker->working_queue.access));
		//Process the entry from the working queue
		if (working_entry) {
			worker->do_work(working_entry);
			free(working_entry);
			working_entry = NULL;
		}
	}
	pthread_exit((void *) 0);
}


int stop_worker(worker_t *worker) {
	pthread_mutex_lock(&worker->working_queue.access);
	if (worker->stop_working) {
		pthread_mutex_unlock(&worker->working_queue.access);
		return 0;
	}
	while(worker->working_queue.number_of_entries != 0) {
		pthread_cond_wait(&worker->working_queue.empty, &worker->working_queue.access);
	}
	//The queue is empty. Set the stop_working flag and release the mutex.
	worker->stop_working = true;
	pthread_mutex_unlock(&worker->working_queue.access);
	//Wake up the worker thread that waits for non-empty queue so it can check the stop_working parameter.
	pthread_cond_broadcast(&worker->working_queue.not_empty);
	//Wait for the worker thread to make an exit
	pthread_join(worker->working_thread, NULL);
	return 0;
}


void worker_clean_up(worker_t **worker) {
	if (*worker) {
		pthread_cond_destroy(&((*worker)->working_queue.not_full));
		pthread_cond_destroy(&((*worker)->working_queue.not_empty));
		pthread_mutex_destroy(&((*worker)->working_queue.access));
		free((*worker)->working_queue.entry);
		free(*worker);
		*worker = NULL;
	}
}

void startLog() {
	syslog_started = true;
	openlog(syslog_identity,syslog_options,syslog_facility);
};
void logMessage(int logLevel,char *msg,...) {
	va_list		ap;
	char		complete_message[MAX_LENGTH];

	if (logLevel > log_level) return;
	va_start(ap, msg);
	vsnprintf(complete_message, sizeof(complete_message)-2, msg, ap);
	va_end(ap);

	if ((daemonize == false) || (syslog_started == false)) {
		fprintf(stdout,"%s: %s\n",program_name,complete_message);
	} else {
		syslog(logLevel,complete_message);
	}
};
void endLog() {
	syslog_started = false;
	closelog();
};

void splitStringInTwo(char *input,char **output_one,char **output_two) {
	char	*tmp,*secondInput;
	*output_one = input;
	*output_two = NULL;
	if ((input[0] == '\'') || (input[0] == '\"')) (*output_one)++;
	// find the space (unless quoted)
	tmp = *output_one;
	if (*output_one == input) {
		// go until we have a space;
		while ((tmp[0] != ' ') && (tmp[0] != '\t') && (tmp[0] != '\0')) tmp++;
		*output_two = tmp;
	} else {
		while ((tmp[0] != input[0]) && (tmp[0] != '\0'))  tmp++;
		*output_two = tmp;
		// okay, we did find a closing quote of some sort, process it
		if (*output_two[0] != '\0') {
			if ((*output_two)[0] == input[0]) (*output_two)++;
		};
	}
	if (*output_two[0] != '\0') (*output_two)++;
	tmp[0] = '\0';
	// eat the space in between
	while ((*output_two[0] == '\t') || (*output_two[0] == ' ')) (*output_two)++;
	secondInput = *output_two;
	if ((secondInput[0] == '\'') || (secondInput[0] == '\"')) {
		(*output_two)++;
		tmp = *output_two;
		while (tmp[0] != secondInput[0]) tmp++;
		if (tmp[0] != '\0') tmp[0] = '\0'; // null terminate this thing
	} else {
		tmp = *output_two;
		while ((tmp[0] != '\t') && (tmp[0] != '\0') && (tmp[0] != ' ')) {
			tmp++;
		}
		if (tmp[0] != '\0') tmp[0] = '\0';
	};
};
void parseRule(char *txt_rule) {
	gpio_rule_t	*rule;
	char		*state;
	char		*translated_state;

	splitStringInTwo(txt_rule,&state,&translated_state);
	rule = (gpio_rule_t *)malloc(sizeof(gpio_rule_t));
	rule->next = rules;
	rule->state = (char*)malloc(strlen(state)+1);
	rule->translation = (char*)malloc(strlen(translated_state)+1);

	strcpy(rule->state,state);
	strcpy(rule->translation,translated_state);

	rules = rule;
};

void print_help(char *program_name,int die_when_done) {
	if (die_when_done > 0) {
		endLog();
		exit(-1);
	}
};

void makeLowerCase(char *str) {
	for (int x = 0;x < strlen(str);x++) {
                if(str[x] >= 'A') && (str[x] <= 'Z')) {
			str[x] = str[x] - 'A' + 'a';
		}
	}
}
void setConfigItem(char *option,char *value) {
	if (strlen(option) < 1) return;
	// let's make this easier on ourselves and lowercase this string
	makeLowerCase(option);
	if (strcmp(option,"log_level") == 0) {
		makeLowerCase(value);
		if (strcmp(value,"critical")) {
			log_level = LOG_CRIT;
		} else if (strcmp(value,"error")) {
			log_level = LOG_ERR;
		} else if (strcmp(value,"warning")) {
			log_level = LOG_WARNING;
		} else if (strcmp(value,"info")) {
			log_level = LOG_INFO;
		} else if (strcmp(value,"debug")) {
			log_level = LOG_DEBUG;
		}
	} else if (strcmp(option,"input_topic") == 0) {
		strcpy(mqtt_input_topic,value);
	} else if (strcmp(option,"#") == 0) {
		// this is a comment, don't do anything
	} else if (strcmp(option,"output_topic") == 0) {
		strcpy(mqtt_output_topic,value);
	} else if (strcmp(option,"server") == 0) {
		strcpy(mqtt_server,value);
	} else if (strcmp(option,"hostname") == 0) {
		strcpy(mqtt_server,value);
	} else if (strcmp(option,"port") == 0) {
		mqtt_port = atoi(value);
	} else if (strcmp(option,"rule") == 0) {
		parseRule(value);
	} else if (strcmp(option,"mqtt_pin_separator") == 0) {
		strcpy(mqtt_pin_separator,optarg);
	} else if (strcmp(option,"daemonize") == 0) {
		makeLowerCase(value);
		if ((strcmp(value,"1") == 0) || (strcmp(value,"true") == 0)) {
			daemonize = true;
		} else {
			logMessage(LOG_INFO,"daemonize set to %s, not equal to \"1\" or \"true\", assuming false",value);
			daemonize = false;
		};
	} else if (strcmp(option,"syslog_facility") == 0) {
		syslog_options = atoi(value);
	} else if (strcmp(option,"syslog_identity") == 0) {
		strcpy(syslog_identity,value);
	} else if (strcmp(option,"syslog_options") == 0) {
		syslog_options = atoi(value);
	} else {
		logMessage(LOG_CRIT,"ERROR: unknown configuration option \"%s\" encountered (%s)",option,value);
		endLog();
		exit(-1);
	}
};
void parseConfigLine(char *line) {
	char	*cmd;
	char	*linecopy;
	char	*param,*t1,*t2;
	int	pos;
	linecopy = (char*)malloc(strlen(line)+1);
	strcpy(linecopy,line);
	cmd = linecopy;
	while ((cmd[0] == ' ') || (cmd[0] == '\t')) cmd++;
	param = cmd;
	pos = 0;
	while ((param[0] != ' ') && (param[0] != '\t') && (param[0] != 0)) param++;
	if (param[0] == 0) {
		setConfigItem(cmd,param);
		free(linecopy);
		return;
	}
	param[0] = 0;
	param++;
	while ((param[0] == ' ') || (param[0] == '\t')) param++;
	if ((param != NULL) && (strlen(param) > 0)) {
		setConfigItem(cmd,param);
	} else {
		setConfigItem(cmd,NULL);
	};
	free(linecopy);
};
void read_configuration(char *filename) {
	FILE	*fp;
	char	*line;
	size_t	len;
	ssize_t	read;

	line = NULL;
	len = 0;
	logMessage(LOG_DEBUG,"reading config file %s",filename);
	fp = fopen(filename,"r");
	if (fp == NULL) {
		logMessage(LOG_CRIT,"Failed to read configuration file : %s (%s)",filename,strerror(errno));
		endLog();
		exit(-1);
	};
	while ((read = getline(&line,&len,fp)) != -1) {
		// length of line = read, actual line = line
		while ((line[strlen(line)-1] == '\n') || (line[strlen(line)-1] == '\r')) line[strlen(line)-1] = 0;
		parseConfigLine(line);
	}
	fclose(fp);
	logMessage(LOG_DEBUG,"config file %s successful",filename);
	if (line) free(line);
};

int main(int argc, char *argv[]) {
	char		ch;
	char		subscription_string[MAX_LENGTH];
	bool		config_read,config_error;

	program_name = argv[0];
	config_read = false;
	config_error = false;

	// set defaults
	strcpy(mqtt_pin_separator,DEFAULT_MQTT_PIN_SEPARATOR);
	strcpy(mqtt_input_topic,DEFAULT_MQTT_INPUT_TOPIC);
	strcpy(mqtt_output_topic,DEFAULT_MQTT_OUTPUT_TOPIC);
	strcpy(mqtt_server,DEFAULT_MQTT_SERVER);
	strcpy(config_file,DEFAULT_CONFIG_FILE);
	strcpy(syslog_identity,DEFAULT_SYSLOG_IDENTITY);
	syslog_options = DEFAULT_SYSLOG_OPTIONS;
	syslog_facility = DEFAULT_SYSLOG_FACILITY;
	mqtt_port = DEFAULT_MQTT_PORT;
	log_level = LOG_WARNING;
	daemonize = false;
	syslog_started = false;
	last_message_sent = NULL;

	logMessage(LOG_DEBUG,"starting to parse command line");
	while ((ch = getopt(argc,argv,"a:c:df:h:i:l:o:p:qr:s:t:v")) != 255) {
	switch (ch) {
	case 'a':
		setConfigItem("output_topic",optarg);
		break;
	case 'c':
		read_configuration(optarg);
		config_read = true;
		break;
	case 'd':
		setConfigItem("daemonize","1");
		break;
	case 'f':
		setConfigItem("syslog_facility",optarg);
		break;
	case 'h':
		setConfigItem("server",optarg);
		break;
	case 'i':
		setConfigItem("syslog_identity",optarg);
		break;
	case 'l':
		setConfigItem("log_level",optarg);
		break;
	case 'o':
		setConfigItem("syslog_options",optarg);
		break;
	case 'p':
		setConfigItem("port",optarg);
		break;
	case 'q':
		log_level--;
		break;
	case 'r':
		setConfigItem("rule",optarg);
		break;
	case 's':
		setConfigItem("mqtt_pin_separator",optarg);
		break;
	case 't':
		setConfigItem("input_topic",optarg);
		break;
	case 'v':
		log_level++;
		break;
	case '?':
		logMessage(LOG_CRIT,"unknown option: %c",optopt);
		print_help(argv[0],1);
	};
	};
	for (; optind < argc; optind++) {
		logMessage(LOG_WARNING,"extra arguments on the command line that were ignored: %s",argv[optind]);
		config_error = true;
	};
	logMessage(LOG_DEBUG,"finished command line parsing");
	if (config_error == true) {
		exit(-1);
	};

	if (config_read == false) {
		read_configuration(config_file);
	}
	if (daemonize == true) {
		pid_t pID = fork();
		if (pID < 0) {
			logMessage(LOG_CRIT,"ERROR: Failed to fork()");
			exit(-1);
		} else if (pID != 0) {
			// parent
			exit(0);
		}
		if (setsid() < 0) exit(-1);
		pID = fork();
		if (pID < 0) {
			logMessage(LOG_CRIT,"ERROR: Failed to fork()");
			exit(-1);
		} else if (pID != 0) {
			// parent
			exit(0);
		}

	};
	startLog();

	// libmosquitto MQTT client instance pointer
	mosq = NULL;
	// Semaphore for blocking the main thread execution.
    	sem_t blocking_sem;

	// thread for processing received payload from all publishers
	worker_t *mqtt_message_processor = NULL;
	// FIFO queue for payloads
	mqtt_message_queue = NULL;

#ifdef __SHOW_MOSQUITTO_INFO__
	int major, minor, revision;
    
	// Get libmosquitto version info
	mosquitto_lib_version(&major, &minor, &revision);
	logMessage(LOG_DEBUG,"using libmosquitto version: %d.%d.%d\n", major, minor, revision);
#endif

	// Create a working thread used by the subscriber for processing
	// the received MQTT messages. 
	if (create_worker(&mqtt_message_processor, 32, MAX_LENGTH, process_message)) {
	//if (create_worker(&mqtt_message_processor, 32, sizeof(char*), process_message)) {
		logMessage(LOG_CRIT,"failed to create mosquitto message thread");
		endLog();
		exit(-1);
	}

	// FIFO queue for the payload from MQTT messages received by the
	// subscriber. Worker thread will process the items as they land
	// in the queue.
	mqtt_message_queue = &mqtt_message_processor->working_queue;

	// libmosquitto initialization
	mosquitto_lib_init();

	// Init the semaphore
	sem_init(&blocking_sem, 0, 0);
	
	// Create new libmosquitto client instance
	mosq = mosquitto_new(DEFAULT_MQTT_CLIENT_NAME, true, mqtt_message_queue);

	if (!mosq) {
		logMessage(LOG_CRIT,"failed to create mosquitto client");
		endLog();
		exit(-1);
	}

	// Define a function which will be called by libmosquitto client
	// every time when there is a new MQTT message
	mosquitto_message_callback_set(mosq, my_message_callback);
	
	// Connect to MQTT broker
	if (mosquitto_connect(mosq, mqtt_server,
			mqtt_port, 60) != MOSQ_ERR_SUCCESS) {
		logMessage(LOG_CRIT,"Error: connecting to MQTT broker (%s:%d) failed",mqtt_server,mqtt_port);
		endLog();
		exit(-1);

		stop_worker(mqtt_message_processor);
		worker_clean_up(&mqtt_message_processor);

		clean_up_libmosquitto(mosq);

		exit(-1);
	}

	// Subscribe to any GPIO channel
	sprintf(subscription_string,"%s/+",mqtt_input_topic);
	logMessage(LOG_INFO,"subscribing to %s",mqtt_input_topic);
	mosquitto_subscribe(mosq, NULL, mqtt_input_topic, 0);

	// Run libmosquitto client in a separate thread
	mosquitto_loop_start(mosq);
	
	while(1) {
		// Block the execution of the main thread
		sem_wait(&blocking_sem);
		break;
	}

	// Stop the worker thread
	stop_worker(mqtt_message_processor);

	// Stop libmosquitto cliet thread
	mosquitto_loop_stop(mosq, true);

	// Worker clean up
	worker_clean_up(&mqtt_message_processor);

	// Clean up/destroy objects created by libmosquitto
	clean_up_libmosquitto(mosq);
	endLog();
}

The code can be compiled using :


gcc  -o sentinel_translator.o -c sentinel_translator.c
gcc -lmosquitto -lpthread -o sentinel_translator sentinel_translator.o

uses a configuration file called /etc/sentinel/translator.conf, and you will need to make sure there are "rules" defined inside. My configuration :


input_topic	GPIO
output_topic	SENSORS
hostname	localhost
port		1883

daemonize	true

log_level	INFO

Rule A3C1FFD901-34=1 'Attic Door=OPEN'
Rule A3C1FFD901-34=0 "Attic Door=CLOSED"
Rule A3C1FFD901-33=1 'Bedroom Door=OPEN'
Rule "A3C1FFD901-33=0" "Bedroom Door=CLOSED"

# I always thought I needed something special, but the
# more I thought about it, the less we actually needed
# anything special for a four-state-sensor like garage
# doors
Rule A3C1FFD901-13=1 'Garage Door=OPENING'
Rule A3C1FFD901-13=0 'Garage Door=CLOSED'
Rule A3C1FFD901-14=1 'Garage Door=OPEN'
Rule A3C1FFD901-14=0 'Garage Door=CLOSING'

The rules will be unknown when you start. Simply run "mosquitto_sub -t \# -v" and then physically work the sensors tied to your GPIO pins (e.g. open the door). It will print the topic, and the "A3C1FFD901-13=1" message. Since you know it was slapped in there when your garage door started opening, you can map the rule appropriately. Run through each part until you know how each one behaves, and then you can launch the program (no parameters required).

This will see the pin message from the MQTT on the GPIO channel, and write it back as the second parameter, e.g. "Garage Door=OPENING". Note that the left side of the equal sign ("Garage Door") MUST match the service check name in Nagios for each one you configure.

Now, all we need to do is get that second parameter into Nagios.

Nagios Interface

Since Nagios can be configured to notify via e-mails, etc, I'll not worry about that here. You can play with that on your own. But, we DO need an MQTT subscription tool that writes these sensors into the external command file for Nagios. The code (yes, another hack-and-slash) :


/**
  * @file sentinel_nagios_fifo.c
  *
  * @brief Mqtt client used to take MQTT messages and write them to Nagios
  * 
  * @date 19-Dec-2024
  * @copyright GNU General Public License v3
  *
  */

#include <stdio.h>
#include <stdlib.h>
#include <unistd.h>
#include <stdarg.h>
#include <string.h>
#include <semaphore.h>
#include <stdint.h>
#include <time.h>
#include <pthread.h>
#include <errno.h>
#include <stdbool.h>
#include <syslog.h>

#include <mosquitto.h>

#define	MAXIMUM_WORKING_QUEUE_LENGTH	128

#define	DEFAULT_MQTT_PIN_SEPARATOR	"-"
#define	DEFAULT_MQTT_TOPIC		"SENSORS"
#define	DEFAULT_MQTT_SERVER		"localhost"
#define	DEFAULT_MQTT_PORT		1883
#define	DEFAULT_NAGIOS_FIFO_FILE	"/var/lib/nagios3/rw/nagios.cmd"
#define	DEFAULT_NAGIOS_CHECK_HOST	"localhost"
#define	DEFAULT_CONFIG_FILE		"/etc/sentinel/nagiosfifo.conf"
#define	DEFAULT_NAGIOS_STATE		"UNKNOWN"
#define	DEFAULT_SYSLOG_IDENTITY		""
#define	DEFAULT_SYSLOG_FACILITY		1
#define	DEFAULT_SYSLOG_OPTIONS		LOG_NOWAIT|LOG_PID
#define	MAX_LENGTH			255
//#define	__SHOW_MOSQUITTO_INFO__		1
#define	DEFAULT_MQTT_CLIENT_NAME	"sentinel-nagiosprocessor"

char	config_file[MAX_LENGTH];
char	mqtt_topic[MAX_LENGTH];
char	mqtt_server[MAX_LENGTH];
int	mqtt_port;
int	log_level;
char	*program_name;
char	nagios_service_host[MAX_LENGTH];
char	nagios_fifo_file[MAX_LENGTH];
char	syslog_identity[MAX_LENGTH];
int	syslog_facility;
bool	syslog_started;
int	syslog_options;
bool	daemonize;
char	*last_message_sent;
struct mosquitto	*mosq = NULL;

typedef struct state_t state_t;
struct state_t {
	char		*name;
	char		*nagios_state;
	state_t		*next;
};
typedef struct working_queue working_queue_t;
struct working_queue {
	// Maximum length of the queue in number of entries
	unsigned int max_queue_size;
	// Pthread mutex for controlling the queue access
	pthread_mutex_t access;
	// Conditional variable queue is empty. Informs the waiting thread
	// that the queue is empty.
	pthread_cond_t empty;
	// Conditional variable queue is not full. Informs the waiting thread
	// that new entries can be written in the queue
	pthread_cond_t not_full;
	// Conditional variable queue is not empty. Informs the waiting
	// thread that there are some entries in the queue
	pthread_cond_t not_empty;
	// Head of the FIFO queue. Worker thread process the head entry first
	int head;
	// Tail of the FIFO queue. New entry is appended at the tail of the
	// queue.
	int tail;
	// Current number of entries in the queue
	int number_of_entries;
	// Pointer to memory reserved for the queue
	void *entry;
	// Size of one queue entry in bites
	int entry_size;
};
typedef int (*do_work_f)(void *work_entry);
typedef struct worker worker_t;
struct worker {
	// FIFO queue for storing the received payloads from MQTT messages
	working_queue_t working_queue;
	// Worker will run in this thread
	pthread_t working_thread;
	// Stop the processing of the working queue items
	bool stop_working;
	// Function that will be called for each queue entry. MQTT payload
	// processor.
	do_work_f do_work;
};
void add_work_entry(working_queue_t *working_queue, void *working_entry);
void logMessage(int logLevel,char *msg,...);
static void *worker_thread(void *worker_thread_arguments);
void worker_clean_up(worker_t **worker);
working_queue_t *mqtt_message_queue;

state_t		*states;

/**
 * @brief This function will be called by the worker thread for
 * processing each entry from the queue of MQTT message payloads.
 * 
 * @param[in] process this entry from the queue
 */

void sendUpdateToNagios(char *sensor,char *stateTxt,char *nagiosState) {
	time_t		local_time;
	struct tm	tm_result;
	char		time_stamp[32];
	char		nagios_cmd[MAX_LENGTH*2];
	int		stateNum;
	FILE		*nagiosCmdFile;

	logMessage(LOG_INFO,"sensor '%s' state set to %s (Nagios: %s)",sensor,stateTxt,nagiosState);
	if (strcmp(nagiosState,"OK") == 0) {
		stateNum = 0;
	} else if (strcmp(nagiosState,"WARNING") == 0) {
		stateNum = 1;
	} else if (strcmp(nagiosState,"CRITICAL") == 0) {
		stateNum = 2;
	} else {
		stateNum = 3;
	};

	local_time = time(NULL);
	localtime_r(&local_time, &tm_result);
	strftime(time_stamp, sizeof(time_stamp), "%s", &tm_result);
	sprintf(nagios_cmd,"[%s] PROCESS_SERVICE_CHECK_RESULT;%s;%s;%d;%s - %s %s",time_stamp,nagios_service_host,sensor,stateNum,nagiosState,sensor,stateTxt);
	logMessage(LOG_DEBUG,"sending NAGIOS message \"%s\" to cmd file %s",nagios_cmd,nagios_fifo_file);
	nagiosCmdFile = fopen(nagios_fifo_file,"a");
	if (nagiosCmdFile == NULL) {
		logMessage(LOG_ERR,"ERROR: Cannot open %s: %s",nagios_fifo_file,strerror(errno));
	} else {
		fprintf(nagiosCmdFile,"%s\n",nagios_cmd);
		fclose(nagiosCmdFile);
	}
};
int process_message(void *message) {
	time_t		local_time;
	char		*msgState;
	char		*sensor;
	char		*nagiosState;
	state_t		*state;

	sensor = (char*)malloc(strlen(message)+1);
	strcpy(sensor,message);
	logMessage(LOG_DEBUG,"process_message(): got message: %s",sensor);
	msgState = sensor;
	nagiosState = NULL;
	while ((msgState[0] != '=') && (msgState[0] != '\0')) msgState++;
	if (msgState[0] == '=') {
		msgState[0] = '\0';
		msgState++;
		// now find the corresponding state string
		state = states;
		while (state != NULL) {
			if (strcmp(msgState,state->name) == 0) {
				logMessage(LOG_DEBUG,"process_message(): mapping state %s to '%s'",message,state->name);
				nagiosState = state->nagios_state;
			};
			state = state->next;
		};
		if (nagiosState == NULL) {
			logMessage(LOG_WARNING,"ERROR: state %s is NOT mapped anywhere, defaulting to UNKNOWN",msgState);
			nagiosState = DEFAULT_NAGIOS_STATE;
		};
	} else {
		nagiosState = DEFAULT_NAGIOS_STATE;
	}
	sendUpdateToNagios(sensor,msgState,nagiosState);
	free(sensor);

	fflush(stdout);
	
	return 0;
}


/**
 * @brief Call back function for received MQTT message.
 * 
 * Libmosquitto thread will call this function for every received MQTT message.
 * It extracts the payload info from the MQTT message and writes it into the
 * working FIFO queue.
 * 
 * @param[in] pointer to libmoquitto MQTT client instance
 * @param[in,out] pointer to the data defined by the Libmosquitto user/caller
 * @param[in] points to the received MQTT message 
 */
void my_message_callback(struct mosquitto *mosq, void *userdata, const struct mosquitto_message *message)
{
	char		*e;
	char		msg[255];
	e = (char*)malloc(strlen((char*)message->payload)+1);
	strcpy(e,message->payload);
	add_work_entry(mqtt_message_queue, message->payload);
}

static void clean_up_libmosquitto(struct mosquitto *mosq) {
	mosquitto_destroy(mosq);
	mosquitto_lib_cleanup();
}

int create_worker(worker_t **worker, unsigned int working_queue_size, unsigned int working_queue_entry_size, do_work_f do_work) {
	if (*worker) {
		//Worker already created/exists.
		return -1;
	}
	pthread_attr_t attr_thread;
	int rc;
	*worker = malloc(sizeof(worker_t));
	(*worker)->working_queue.head = 0;
	(*worker)->working_queue.number_of_entries = 0;
	(*worker)->working_queue.tail = 0;
	(*worker)->working_queue.entry_size = working_queue_entry_size;
	(*worker)->working_queue.max_queue_size = working_queue_size > MAXIMUM_WORKING_QUEUE_LENGTH ? MAXIMUM_WORKING_QUEUE_LENGTH : working_queue_size;
	(*worker)->working_queue.entry = calloc((*worker)->working_queue.max_queue_size, working_queue_entry_size);
	(*worker)->stop_working = false;
	(*worker)->do_work = do_work;
	//Init the objects for synchronisation of the threds
	pthread_mutex_init(&((*worker)->working_queue.access), NULL);
	pthread_cond_init(&((*worker)->working_queue.not_full), NULL);
	pthread_cond_init(&((*worker)->working_queue.not_empty), NULL);
	pthread_cond_init(&((*worker)->working_queue.empty), NULL);
	pthread_attr_init(&attr_thread);
	pthread_attr_setdetachstate(&attr_thread, PTHREAD_CREATE_JOINABLE);
	//Inherit all of the thread scheduling properties from the calling thread.
	pthread_attr_setinheritsched(&attr_thread, PTHREAD_EXPLICIT_SCHED);
	//Start the worker thread
	rc = pthread_create(&((*worker)->working_thread), &attr_thread, worker_thread, *worker);
	if (rc != 0) {
		strerror(errno);
		worker_clean_up(worker);
		rc = -1;
	}
	return 0;
}


void add_work_entry(working_queue_t *working_queue, void *working_entry) {
	pthread_mutex_lock(&(working_queue->access));
	while((working_queue->number_of_entries) == working_queue->max_queue_size) {
		pthread_cond_wait(&(working_queue->not_full), &(working_queue->access));
	}
	// Place the new work entry on the working queue
	memcpy(working_queue->entry + working_queue->entry_size * working_queue->tail, working_entry, working_queue->entry_size);
	working_queue->tail = (working_queue->tail + 1) % working_queue->max_queue_size;
	working_queue->number_of_entries++;
	pthread_cond_signal(&(working_queue->not_empty));
	pthread_mutex_unlock(&working_queue->access);
}


void *worker_thread(void *worker_thread_arguments) {
	worker_t *worker = (worker_t *) worker_thread_arguments;
	void *working_entry = NULL;
	while(1) {
		pthread_mutex_lock(&(worker->working_queue.access));
		/* Check if the queue is empty. */
		while ((worker->working_queue.number_of_entries == 0) && (!worker->stop_working)) {
			/* Wait till the other side signals that there a new item on the queue. */
			pthread_cond_wait(&(worker->working_queue.not_empty), &(worker->working_queue.access));
		}
		if (worker->stop_working) {
			pthread_mutex_unlock(&(worker->working_queue.access));
			if (working_entry) {
				free(working_entry);
			}
			pthread_exit(NULL);
		}
		if (worker->working_queue.number_of_entries != 0) {
			working_entry = malloc(worker->working_queue.entry_size);
			memcpy(working_entry, worker->working_queue.entry + (worker->working_queue.head * worker->working_queue.entry_size), worker->working_queue.entry_size);
			worker->working_queue.head = (worker->working_queue.head + 1) % worker->working_queue.max_queue_size;
			worker->working_queue.number_of_entries--;
		}
		if (worker->working_queue.number_of_entries < worker->working_queue.max_queue_size) {
			pthread_cond_signal(&(worker->working_queue.not_full));
		}
		if (worker->working_queue.number_of_entries == 0) {
			pthread_cond_signal(&(worker->working_queue.empty));
		}
		pthread_mutex_unlock(&(worker->working_queue.access));
		//Process the entry from the working queue
		if (working_entry) {
			worker->do_work(working_entry);
			free(working_entry);
			working_entry = NULL;
		}
	}
	pthread_exit((void *) 0);
}


int stop_worker(worker_t *worker) {
	pthread_mutex_lock(&worker->working_queue.access);
	if (worker->stop_working) {
		pthread_mutex_unlock(&worker->working_queue.access);
		return 0;
	}
	while(worker->working_queue.number_of_entries != 0) {
		pthread_cond_wait(&worker->working_queue.empty, &worker->working_queue.access);
	}
	//The queue is empty. Set the stop_working flag and release the mutex.
	worker->stop_working = true;
	pthread_mutex_unlock(&worker->working_queue.access);
	//Wake up the worker thread that waits for non-empty queue so it can check the stop_working parameter.
	pthread_cond_broadcast(&worker->working_queue.not_empty);
	//Wait for the worker thread to make an exit
	pthread_join(worker->working_thread, NULL);
	return 0;
}


void worker_clean_up(worker_t **worker) {
	if (*worker) {
		pthread_cond_destroy(&((*worker)->working_queue.not_full));
		pthread_cond_destroy(&((*worker)->working_queue.not_empty));
		pthread_mutex_destroy(&((*worker)->working_queue.access));
		free((*worker)->working_queue.entry);
		free(*worker);
		*worker = NULL;
	}
}

void startLog() {
	syslog_started = true;
	openlog(syslog_identity,syslog_options,syslog_facility);
};
void logMessage(int logLevel,char *msg,...) {
	va_list		ap;
	char		complete_message[MAX_LENGTH];

	if (logLevel > log_level) return;
	va_start(ap, msg);
	vsnprintf(complete_message, sizeof(complete_message)-2, msg, ap);
	va_end(ap);

	if ((daemonize == false) || (syslog_started == false)) {
		fprintf(stdout,"%s: %s\n",program_name,complete_message);
	} else {
		syslog(logLevel,complete_message);
	}
};
void endLog() {
	syslog_started = false;
	closelog();
};

void splitStringInTwo(char *input,char **output_one,char **output_two) {
	char	*tmp,*secondInput;
	*output_one = input;
	*output_two = NULL;
	if ((input[0] == '\'') || (input[0] == '\"')) (*output_one)++;
	// find the space (unless quoted)
	tmp = *output_one;
	if (*output_one == input) {
		// go until we have a space;
		while ((tmp[0] != ' ') && (tmp[0] != '\t') && (tmp[0] != '\0')) tmp++;
		*output_two = tmp;
	} else {
		while ((tmp[0] != input[0]) && (tmp[0] != '\0'))  tmp++;
		*output_two = tmp;
		// okay, we did find a closing quote of some sort, process it
		if (*output_two[0] != '\0') {
			if ((*output_two)[0] == input[0]) (*output_two)++;
		};
	}
	if (*output_two[0] != '\0') (*output_two)++;
	tmp[0] = '\0';
	// eat the space in between
	while ((*output_two[0] == '\t') || (*output_two[0] == ' ')) (*output_two)++;
	secondInput = *output_two;
	if ((secondInput[0] == '\'') || (secondInput[0] == '\"')) {
		(*output_two)++;
		tmp = *output_two;
		while (tmp[0] != secondInput[0]) tmp++;
		if (tmp[0] != '\0') tmp[0] = '\0'; // null terminate this thing
	} else {
		tmp = *output_two;
		while ((tmp[0] != '\t') && (tmp[0] != '\0') && (tmp[0] != ' ')) {
			tmp++;
		}
		if (tmp[0] != '\0') tmp[0] = '\0';
	};
};
void parseState(char *txt_state) {
	state_t		*state;
	char		*input_state;
	char		*output_state;

	splitStringInTwo(txt_state,&input_state,&output_state);
	state = (state_t *)malloc(sizeof(state_t));
	state->next = states;
	state->name = (char*)malloc(strlen(input_state)+1);
	state->nagios_state = (char*)malloc(strlen(output_state)+1);

	strcpy(state->name,input_state);
	strcpy(state->nagios_state,output_state);

	states = state;
};

void print_help(char *program_name,int die_when_done) {
	if (die_when_done > 0) {
		endLog();
		exit(-1);
	}
};

void makeLowerCase(char *str) {
	for (int x = 0;x<strlen(str)lx++) P
                if ((str[x] >= 'A') && (str[x] <= 'Z')) {
			str[x] = str[x] - 'A' + 'a';
		}
	}
}
void setConfigItem(char *option,char *value) {
	if (strlen(option) < 1) return;
	// let's make this easier on ourselves and lowercase this string
	makeLowerCase(option);
	if (strcmp(option,"log_level") == 0) {
		makeLowerCase(value);
		if (strcmp(value,"critical")) {
			log_level = LOG_CRIT;
		} else if (strcmp(value,"error")) {
			log_level = LOG_ERR;
		} else if (strcmp(value,"warning")) {
			log_level = LOG_WARNING;
		} else if (strcmp(value,"info")) {
			log_level = LOG_INFO;
		} else if (strcmp(value,"debug")) {
			log_level = LOG_DEBUG;
		}
	} else if (strcmp(option,"topic") == 0) {
		strcpy(mqtt_topic,value);
	} else if (strcmp(option,"#") == 0) {
		// this is a comment, don't do anything
	} else if (strcmp(option,"server") == 0) {
		strcpy(mqtt_server,value);
	} else if (strcmp(option,"hostname") == 0) {
		strcpy(mqtt_server,value);
	} else if (strcmp(option,"port") == 0) {
		mqtt_port = atoi(value);
	} else if (strcmp(option,"state") == 0) {
		parseState(value);
	} else if (strcmp(option,"daemonize") == 0) {
		makeLowerCase(value);
		if ((strcmp(value,"1") == 0) || (strcmp(value,"true") == 0)) {
			daemonize = true;
		} else {
			logMessage(LOG_INFO,"daemonize set to %s, not equal to \"1\" or \"true\", assuming false",value);
			daemonize = false;
		};
	} else if (strcmp(option,"syslog_facility") == 0) {
		syslog_options = atoi(value);
	} else if (strcmp(option,"syslog_identity") == 0) {
		strcpy(syslog_identity,value);
	} else if (strcmp(option,"syslog_options") == 0) {
		syslog_options = atoi(value);
	} else if (strcmp(option,"nagios_service_host") == 0) {
		strcpy(nagios_service_host,value);
	} else if (strcmp(option,"nagios_fifo_file") == 0) {
		strcpy(nagios_fifo_file,value);
	} else {
		logMessage(LOG_CRIT,"ERROR: unknown configuration option \"%s\" encountered (%s)",option,value);
		endLog();
		exit(-1);
	}
};
void parseConfigLine(char *line) {
	char	*cmd;
	char	*linecopy;
	char	*param,*t1,*t2;
	int	pos;
	linecopy = (char*)malloc(strlen(line)+1);
	strcpy(linecopy,line);
	cmd = linecopy;
	while ((cmd[0] == ' ') || (cmd[0] == '\t')) cmd++;
	param = cmd;
	pos = 0;
	while ((param[0] != ' ') && (param[0] != '\t') && (param[0] != 0)) param++;
	if (param[0] == 0) {
		setConfigItem(cmd,param);
		free(linecopy);
		return;
	}
	param[0] = 0;
	param++;
	while ((param[0] == ' ') || (param[0] == '\t')) param++;
	if ((param != NULL) && (strlen(param) > 0)) {
		setConfigItem(cmd,param);
	} else {
		setConfigItem(cmd,NULL);
	};
	free(linecopy);
};
void read_configuration(char *filename) {
	FILE	*fp;
	char	*line;
	size_t	len;
	ssize_t	read;

	line = NULL;
	len = 0;
	logMessage(LOG_DEBUG,"reading config file %s",filename);
	fp = fopen(filename,"r");
	if (fp == NULL) {
		logMessage(LOG_CRIT,"Failed to read configuration file : %s (%s)",filename,strerror(errno));
		endLog();
		exit(-1);
	};
	while ((read = getline(&line,&len,fp)) != -1) {
		// length of line = read, actual line = line
		while ((line[strlen(line)-1] == '\n') || (line[strlen(line)-1] == '\r')) line[strlen(line)-1] = 0;
		parseConfigLine(line);
	}
	fclose(fp);
	logMessage(LOG_DEBUG,"config file %s successful",filename);
	if (line) free(line);
};

int main(int argc, char *argv[]) {
	char		ch;
	bool		config_read,config_error;

	program_name = argv[0];
	config_read = false;
	config_error = false;

	// set defaults
	strcpy(nagios_service_host,DEFAULT_NAGIOS_CHECK_HOST);
	strcpy(nagios_fifo_file,DEFAULT_NAGIOS_FIFO_FILE);
	strcpy(mqtt_topic,DEFAULT_MQTT_TOPIC);
	strcpy(mqtt_server,DEFAULT_MQTT_SERVER);
	strcpy(config_file,DEFAULT_CONFIG_FILE);
	strcpy(syslog_identity,DEFAULT_SYSLOG_IDENTITY);
	syslog_options = DEFAULT_SYSLOG_OPTIONS;
	syslog_facility = DEFAULT_SYSLOG_FACILITY;
	mqtt_port = DEFAULT_MQTT_PORT;
	log_level = LOG_WARNING;
	daemonize = false;
	syslog_started = false;
	last_message_sent = NULL;

	logMessage(LOG_DEBUG,"starting to parse command line");
	while ((ch = getopt(argc,argv,"c:df:h:i:l:o:p:qs:t:v")) != 255) {
	switch (ch) {
	case 'c':
		read_configuration(optarg);
		config_read = true;
		break;
	case 'd':
		setConfigItem("daemonize","1");
		break;
	case 'f':
		setConfigItem("syslog_facility",optarg);
		break;
	case 'h':
		setConfigItem("server",optarg);
		break;
	case 'i':
		setConfigItem("syslog_identity",optarg);
		break;
	case 'l':
		setConfigItem("log_level",optarg);
		break;
	case 'n':
		setConfigItem("log_level",optarg);
		break;
	case 'o':
		setConfigItem("syslog_options",optarg);
		break;
	case 'p':
		setConfigItem("port",optarg);
		break;
	case 'q':
		log_level--;
		break;
	case 's':
		setConfigItem("state",optarg);
		break;
	case 't':
		setConfigItem("topic",optarg);
		break;
	case 'v':
		log_level++;
		break;
	case '?':
		logMessage(LOG_CRIT,"unknown option: %c",optopt);
		print_help(argv[0],1);
	};
	};
	for (; optind < argc; optind++) {
		logMessage(LOG_WARNING,"extra arguments on the command line that were ignored: %s",argv[optind]);
		config_error = true;
	};
	logMessage(LOG_DEBUG,"finished command line parsing");
	if (config_error == true) {
		exit(-1);
	};

	if (config_read == false) {
		read_configuration(config_file);
	}
	if (daemonize == true) {
		pid_t pID = fork();
		if (pID < 0) {
			logMessage(LOG_CRIT,"ERROR: Failed to fork()");
			exit(-1);
		} else if (pID != 0) {
			// parent
			exit(0);
		}
		if (setsid() < 0) exit(-1);
		pID = fork();
		if (pID < 0) {
			logMessage(LOG_CRIT,"ERROR: Failed to fork()");
			exit(-1);
		} else if (pID != 0) {
			// parent
			exit(0);
		}

	};
	startLog();

	// libmosquitto MQTT client instance pointer
	mosq = NULL;
	// Semaphore for blocking the main thread execution.
    	sem_t blocking_sem;

	// thread for processing received payload from all publishers
	worker_t *mqtt_message_processor = NULL;
	// FIFO queue for payloads
	mqtt_message_queue = NULL;

#ifdef __SHOW_MOSQUITTO_INFO__
	int major, minor, revision;
    
	// Get libmosquitto version info
	mosquitto_lib_version(&major, &minor, &revision);
	logMessage(LOG_DEBUG,"using libmosquitto version: %d.%d.%d\n", major, minor, revision);
#endif

	// Create a working thread used by the subscriber for processing
	// the received MQTT messages. 
	if (create_worker(&mqtt_message_processor, 32, MAX_LENGTH, process_message)) {
	//if (create_worker(&mqtt_message_processor, 32, sizeof(char*), process_message)) {
		logMessage(LOG_CRIT,"failed to create mosquitto message thread");
		endLog();
		exit(-1);
	}

	// FIFO queue for the payload from MQTT messages received by the
	// subscriber. Worker thread will process the items as they land
	// in the queue.
	mqtt_message_queue = &mqtt_message_processor->working_queue;

	// libmosquitto initialization
	mosquitto_lib_init();

	// Init the semaphore
	sem_init(&blocking_sem, 0, 0);
	
	// Create new libmosquitto client instance
	mosq = mosquitto_new(DEFAULT_MQTT_CLIENT_NAME, true, mqtt_message_queue);

	if (!mosq) {
		logMessage(LOG_CRIT,"failed to create mosquitto client");
		endLog();
		exit(-1);
	}

	// Define a function which will be called by libmosquitto client
	// every time when there is a new MQTT message
	mosquitto_message_callback_set(mosq, my_message_callback);
	
	// Connect to MQTT broker
	if (mosquitto_connect(mosq, mqtt_server,
			mqtt_port, 60) != MOSQ_ERR_SUCCESS) {
		logMessage(LOG_CRIT,"Error: connecting to MQTT broker (%s:%d) failed",mqtt_server,mqtt_port);
		endLog();
		exit(-1);

		stop_worker(mqtt_message_processor);
		worker_clean_up(&mqtt_message_processor);

		clean_up_libmosquitto(mosq);

		exit(-1);
	}

	logMessage(LOG_INFO,"subscribing to %s",mqtt_topic);
	mosquitto_subscribe(mosq, NULL, mqtt_topic, 0);

	// Run libmosquitto client in a separate thread
	mosquitto_loop_start(mosq);
	
	while(1) {
		// Block the execution of the main thread
		sem_wait(&blocking_sem);
		break;
	}

	// Stop the worker thread
	stop_worker(mqtt_message_processor);

	// Stop libmosquitto cliet thread
	mosquitto_loop_stop(mosq, true);

	// Worker clean up
	worker_clean_up(&mqtt_message_processor);

	// Clean up/destroy objects created by libmosquitto
	clean_up_libmosquitto(mosq);
	endLog();
}

Compile the code using :


gcc  -o sentinel_nagios_fifo.o -c sentinel_nagios_fifo.c
gcc -lmosquitto -lpthread -o sentinel_nagios_fifo sentinel_nagios_fifo.o

Now, remember that command_file and other stuff from Nagios? We're going to need that. Edit your configuration file (/etc/sentinel/nagiosfifo.conf) and make the appropriate edits.

Configuration file :


topic			security
daemonize		true
nagios_fifo_file	/var/lib/nagios3/rw/nagios.cmd
log_level		WARNING

# nagios_service_host	localhost
nagios_service_host	house

state			OPEN	CRITICAL
state			OPENING	WARNING
state			CLOSING	WARNING
state			CLOSED	OK

Then, start the processes up and you have your home door sensors tied into Nagios. Congrats!

The Tropics of Cancer

2024-11-14T09:00:00.002-08:00

First off, if you are the least bit squeamish, or do not like photos of disgusting things, I'll warn you now, this post is replete with them, but they'll be things like boogers and wounds. I do not warn when one will pop up, either, so proceed at your own risk!

This is not a normal thing for me to post. In fact, everything about me says I shouldn't. However, this could be a bit therapeutic, so I'm going to venture way out into an area where I likely would not.

In late December, while on a drive and following a lengthy descent into a valley, one ear popped, and the other one has not. Thinking I'd had a sinus infection for a long time, I headed into urgent care because I didn't want to remain deaf in that ear. At the time, I was attempting to become a "temple worker" - meaning added service in sacred areas. No one would ever call me back, but they definitely called others back around me. I'd later find that this was one of God's "tender mercies" - because what would come definitely would have impacted everything.

The ear pop failure led to an urgent care visit. That urgent care lead to an ENT (ear, nose, throat) specialist in Mid February. The specialist performed a biopsy on some lumps in the neck.

It was between the urgent care visit and my ENT visit that my spouse also was visiting for some medical issues, which led to a medical appointment for March 29th. I would need to be alert and attentive to her needs for a while, and at this point, all seemed well enough.

After my spouses appointment had been set up, I was pushed into some scans, and in late February, I was diagnosed with Nasopharyngeal Carcinoma (the back of the nasal pharynx). This was a week later determined to be stage 3 (identified fully on March 8th).

On March 15th, I began chemotherapy. I have a three week cycle. The first week is a two-drug "cocktail" infusion. The two drugs are Gemcitabine and Cisplatin. The second week is a single-drug "cocktail". The third week is an "off" week. This three-week cycle coincided perfectly in ways I'd never have expected. My "off week" was going to be.... March 29th. The ONE and ONLY time that I could stop being "sick" and instead focus on my dear sweet spouse!

The drug that is common through weeks one and two is a painful one (Gemcitabine). The best description of that pain is when you get hyper cold, and then someone touches your skin. The entire area feels like that while receiving the drug. It is not comfortable.

The nausea was brutal the first week. I was so bloated that I could not consume liquids because everything wanted to come right back up. The second week, I consumed a lot of fiber before going in for the drugs, and I have continually taking Miralax (the nausea medications don't lend them selves well to regular bowel movements). In fact, I dropped 11 pounds in the first week. I do not recommend this diet, I really don't. I did manage to add a few pounds, but that was short-lived.

On March 24, 2024, my arm became very red, swollen, and painful. One provider diagnosed it as "cellulitis", and another diagnosed it as a blood clot. So, I'm taking a few extra medications hoping this will diminish before I receive more infusions. There is more on this a bit later (it's not cellulitis, nor is it a blood clot). This did abate and disappear after about four days.

Miracle number three might be an "odd" miracle. I struggle with "receiving" service. I've always found myself to be the one stepping in to aid others when they are in need. It's one of my joys. However, it is flabbergasting to realize that these same people who I've aided (and others who I have not) are willing to step in to assist. You definitely see God's hand in your life when you get into a situation like this. He cares, He definitely does.

But why would God allow bad things to happen to people? One of the tenants of my spirituality is "agency". We're born into a life where we can choose to be assholes or angels. If God prevented bad stuff from happening, we wouldn't be able to choose the jerk side of life. We're here to be mortal. We fought for that in the "war in heaven" (the book of Revelation) specifically for our agency. We fought to be here and be mortal. In this situation, feeling gross as I do, I'd still fight that war. I'd still want others to have agency. I'd still want to be able to have cancer.

I'd want it, simply because I can see God's love, manifest through peace, as well as the concern and love of others. Without it, I doubt anyone would be able to see the love of God.

So, here I sit, nauseous and queasy, happy that God is aware of me, while I listen to neighbors voluntarily step in and weed flower beds, mow lawns, and spray weeds in other areas. Just amazing care and concern.

April 10th

I actually just went through the start of round 2 (second cycle, two drugs). They re-ordered the drugs, and I didn't have the same pain. Perhaps that was purely because the blood work nurse that inserted the IV did a fantastic job. I don't know. It was better, but the same nausea is just as rampant, even if not higher strength.

However, the nausea and bloating did not end like I thought it would - it took more time before a proper bowel movement happened.

April 16, 2024

At 2:30 AM two days ago (April 14th), my arm once again became red, inflamed, swollen, and painful. After some discussion with the doctor over the next few days, it was determined that this was the result of Gemcitabine - it has a nasty habit of killing veins. On that discussion, it was realized that the pain foot print is the exact footprint of pain from when the Gemcitabine was first being administered on the 15th at full volume. Another swelling point has now appeared on the right arm (where the second dose of Gemcitabine was administered at half volume).

And, Sunday night (April 14, 2024), for the first time in decades, I threw up. Violently.

So, on the 16th, it was determined that I had to have a PICC line installed to prevent more failures of the blood veins. Yay for progress! After installing, the following chemo day blood tests were simply done at the time of the PICC installation.

The following morning (week 2 of round 2, Apr 17), it was decided because of the pain and swelling in the flesh of the arm that we needed to delay the gemcitabine infusion for another week (platelets were okay). This is a tender mercy, because I don't think I could have handled it. I've been experiencing some major PTSD this week. Instead, we headed back home and applied some topical anti-inflammatory to see if we can get the arm back down to where we need to be. The oncologist will look into what kind of gemcitabine is being used (freeze-dried had a 12-times-fewer instance of pain in Japan when compared to the liquid) at the facility, and determine if we can make changes there. He also discussed ending the 3-cycles a cycle early.

Well, the following day, I went in for an ultrasound on the arm. I have superficial vein thrombosis (SVT). That accounts for the swelling and pain. Now I'm on blood thinners, and my wife has banned me from tools that cut. [sigh]. That's the vast majority. Sometimes ya gotta be destructive before you can be constructive. Anyway, it took a few days of the topical cream and the blood thinners to reduce the swelling enough to no longer be excruciating. Now I can get back to normal.

So, back to the Gemcitabine. On the 24th, the drugs from Japan were in (freeze-dried Gemcitabine), so we could get that one started. Surprisingly, they ran it at full steam, and it did not have the same pain. This could be because of the drug itself, or it could be because we're using a PICC line. One side effect is that the bowels seem to function a bit better after this one, so I'm calling it a win. Sure, I still have the nausea, but it hasn't been as strong so far, either (this is added on the 25th, so only 24 hours after infusion). It was on the way home from that infusion that my wife kindly suggested a "playlist" for beating cancer.

AC-DC - Thunderstruck
AC-DC - T.N.T
Alannah Myles - Black Velvet
Alice Cooper - 01 - Poison
Backstreet Boys - Larger Than Life
Black Eyed Peas - Let's Get It Started
Bobby Mcferrin - Don't Worry Be Happy
Breaking Benjamin - 02 - I Will Not Bow
Christina Aguilera - 04 - Fighter
Chumbawamba - Tubthumping
Des'ree - 01 - You Gotta Be
Destiny's Child - 02 - Survivor
Diamante - Bulletproof
Diamante - Had Enough
Diamante - Haunted
Elton John - I'm Still Standing
Eminem - (Disc 2) 03 - Lose Yourself [Clean]
Europe - The Final Countdown
Evanescence - Bring Me to Life
Five Finger Death Punch - Wrong Side of Heaven
Fun. - 04 - Carry On
George Thorogood - Bad to the bone
Gloria Gaynor - I Will Survive
Guns N' Roses - Welcome to the Jungle
Hall and Oates - 06 - You Make My Dreams (Come True)
Heart - Alone
Imagine Dragons - 10 - Thunder
Imagine Dragons - Radioactive
Jenn Johnson - 09 - You're Gonna Be Ok
Jessie J - 13 - Who You Are
Journey - Don't Stop Believin'
Kansas - Hold On
Katy Perry - 01 - Rise
Kelly Clarkson - 02 - Stronger (What Doesn't Kill You)
Kristin Chenoweth, Idina Menzel, Ensemble - Defying Gravity
Lawrence - 01 - Don't Lose Sight [Explicit]
Lee Ann Womack - I Hope You Dance
Metallica - Enter Sandman
Metallica & SanFran Symphony - Enter Sandman
Miley Cyrus - 08 - The Climb
Muse - Uprising
Nickleback - 06 - Burn It to the Ground [Explicit]
Ozzy Osbourne - Crazy Train
Papa Roach - 04 - Born For Greatness
Pat Benatar - Hit Me With Your Best Shot
Pink - Get The Party Started
Pink - So What
Queen - Another One Bites The Dust
Queen - I Want It All
Queen - The Show Must Go On (with Elton John)
Queen - We Are The Champions
Queen - We Will Rock You
Rachel Platten - 05 - Fight Song
R.E.M. - 06 - It's The End Of The World As We Know It (And I Feel Fine)
R.E.M. - Everybody Hurts
R.E.M. - Losing My Religion.mp3
Roxette - Listen To Your Heart
Shinedown - 10 - A Symptom Of Being Human
Simon & Garfunkel - Bridge over Troubled Water
Sixx:A.M. - 05 - Rise
Survivor - Eye of the Tiger
Taylor Swift - 06 - Shake It Off
The All-American Rejects - 03 - Move Along
The Beatles - Let It Be
The Cranberries - Dreams
Twisted Sister - Were Not Gonna Take It

And, a Spotify link to the list :

Spotify - The SilverHawk "Beat Cancer" Playlist

Turns out to be an interesting play list. (I did later add Michael Jackson's "Beat It" and the Bee Gee's "Stayin' Alive", M.C. Hammer "U Can't Touch This"/"Too Legit to Quit" and Softcell's "Tainted Love" among others.) There are too many things that can be mentally tied to cancer, and they just crack me up.

Anyhow, this form of Gemcitabine seems to be much easier on the system.

April 26

Friday rolls around, and I get news from HR that they've rejected my FMLA. So, I start digging in. Turns out the HR person thought I submitted the wrong form. A little back and forth, and suddenly, I've been approved for FMLA. But there is a lot of request start and request stop notifications. And, since it's going to come out of paid time-off anyway, I might as well just do FMLA once in a while (go unpaid to preserve some PTO for when I'm healthy to celebrate), and just PTO on the other days. Yay for awkward situations.

May 1

The off week seemed to be business as usual, with neighbors stepping in to take care of the front yard. I must say that it is a miracle that we are surrounded by such amazing people.

May 10

It's been a while since my last update. We've done the first week of the third round (May 8, Gemcitabine and Cisplaten), and we're getting ready for the second week's dosage (the 15th). However, those SVT's are back with a vengeance. In fact, I now have another one in the right elbow, immediately at the confluence, and that is kind of panicking me.

I went through the ringer today trying to get appointments squared away, though. We moved the radiology oncologist appointment, delayed the colonoscopy appointment, and also cancelled the heart stress test (and subsequent cardiologist appointment). Sheesh.

May 16

After the second week (Gemcitabine) dosage yesterday, I was pulled in for an ultrasound. The radiologist was concerned about not as much the right elbow as he was the PICC line. There was no compression on the vein for that. The doctor did not even mention it, so I guess I have to be okay.

This round has been a little rough. Cognitive function was much lower (ability to concentrate, in particular). It made a rougher day of work, especially with so many folks yammering for their particular piece of my work day. At least I know I have a job still.

June 1

Forgive me, it's been a few weeks (it's now June 1st, so I've had my off week and a week of scans). The PET scan this past Tuesday morning was labeled "significant partial progress". It sounds really awkward, but yesterday (Friday), we spoke with the oncologist. He put that into terms I could understand.... only if the cancer is fully cleared and gone can it be anything but partial. Significant means that the progress is not done, but the initial treatments were obvious to the radiologist. In fact, those lymph nodes that were highlighted in the first PET scan back in March all seemed normal and clean, except for two of them, and those showed a major decrease in size, too.

It seems that the first three rounds did exactly what they needed to, and had a major impact.

Now, it was guaranteed that I'd be facing another 6 weeks of chemo along with radiation. That can potentially start in the next week or two. But it is only going to be 6 weeks, not more.

The radiation will be each week day, and the chemo will be Cisplaten once per week. The volume of Cisplaten will be decreased as well (that was the drug that was given once every three weeks). The chemo is only to aid the radiology in doing it's job. The oncologist said that the first two weeks will be uncomfortable, the next two weeks miserable, and the final two weeks I will probably want to die. The radiology will focus on the neck (obviously, that's where the cancer is), and will make my throat feel like major sunburns. It's going to make it hard and painful to swallow anything, so he mentioned four keys :

Hydration - I should be getting 64 fluid ounces of water per day.
High calorie meals, in any way I can get them. Some people hurt so much on the throat they require feeding tubes. If body weight decreases by 10% in the first three weeks, they'll put me on feeding tubes, so I MUST make sure to maintain weight.
Exercise - I need to get up and move around.
Bowel movements - I must make sure everything is moving like it should.

This is going to be a "fun" road to walk. I should be done by late July, and that means I'd have 6 weeks to heal, but 3 months before inflamation has died down to allow good PET scan results - meaning in theory, the end of November I might be free of cancer.

June 6

Every week, I have to get the PICC line dressing changed. Over the course of a few months, the bandage to keep things clean and pure (it's a straight conduit to the heart) has been getting terribly itchy. I finally told the nurses changing the dressing, and when we pulled the dressing off, we had reactions.

Why am I so stubborn that I waited this long? They changed away from Chlorhexidine (which was reacting with my skin) to Betadine dressing.

June 12

Sure enough, moving away from the de facto standard of Chlorhexidine solved the itching problem. Betadine was the way to go for me.

It looks MUCH better after one week!

July 7th

Okay, the "schedule" they provided about uncomfortable-miserable-death is right on. 10 days into that, my taste started to scram. Two weeks in, I started to struggle to swallow. Within 4 days of that, I was experiencing dehydration. Bloody throat, mouth tissue splitting, and exposed nerves along the throat are making this a rockin' good time. I find myself with no appetite (who wants to eat when it just hurts, and nothing has taste?).

I know I need to consume calories, so breakfasts become things like Malt O Meal wheat germ cereal with ice cream (to increase the caloric intake). My meals are slow and small.

They prescribed a "magic mouth wash" to numb my mouth. Every time I take that, I throw up. Gag reflexes are at an all time high. Saliva is thicker than normal boogers, and that just messes things up even more. Yes, this is the real deal. In one week, I'll hit that 4 week mark where things supposedly go from "miserable" to "I'd rather be dead". When that happens, it could be time for a feeding tube. Maybe I can get a breathing tube, too. Probably not, because a trache might interfere with the stupid radiation. [sigh].

July 16

While in getting dressing changed, remember all of those SVT (blood clots)? I thought I'd grab a picture of the confluence one from months ago.

It is still around, but the vein was killed. The picture above doesn't really show much, but the vein leading up to the confluence is brown. It's also fairly rigid and hard.

July 22nd

It's been a few weeks since my last update. I gotta tell you, that "schedule" of Uncomfortable-Miserable-Death is real. We're one week in on the "wanna die" phase. There is nothing like coughing up blood, dry heaving more blood, and having so much grit on the back of the throat that you simply do not want to eat.

So, last Thursday, I've received an "NJ" tube (nasojejunal tube that runs from outside of the face through the nose and into the jejunum of the intestines.

The criteria for a feeding tube is 10% of your body weight being lost. I hit that pretty solidly last week.

So, after a weekend of using the tube, my weight loss has stabilized. A few more data points and we'll know if I'm trending in the right direction. Caloric intake has risen from 300 calories/day to about 2000, though the doctors want another 1k calories being fed in. We have to slowly build up on the rate of the feeding tube so we don't incur the wrong side effects (requested a 120ml/hour, last night was 88.2ml/hour).

The plus side is that it bypasses the stomach, alleviating some of the nausea. It doesn't solve it all, though. I can't swallow, which means primarily that I get the gagging done anyway because of the thick saliva and the inability to properly clear my sinus of mucus.

Open sores and ulcers, nothing in the stomach (so that hurts, too), and having a new tube strapped to my face makes this a bit more interesting. We're nearly complete, though - we finish chemo in 8 days, and radiation in 11 days. Granted, it's not a guarantee that we're finished - we have to wait until the swelling and inflammation has dropped enough to do a PET scan (likely in November). The targeted milestones :

July 30 - last day of chemotherapy
August 2 - last day of radiation
August 23 - shifting from wanting to die to just being miserable
August 30 - taste should start returning
September 6 - shifting from miserable to uncomfortable, sores likely healing over, potential tube removal
November - PET scan to know for sure

Two more weeks of misery before I can have a few more weeks of misery. Yay!

July 26

Well, some more neighbors crashed the "help" scene. We've had a higher water bill this year, and they found out why - the release valve went south and filled the sprinklers manifolds to the top with water.

So, an absolutely kind neighbor spent two days rebuilding our front valve and sprinkler setup. He would not let us pay for parts or labor. This is yet another miracle.

July 30

The last few days have been excruciating, and not just because of pain (in fact, the pain has been the least of the problems).

Mucus.

Maybe that should be my new name. Coughing fits to try and clear throats, and immediately bringing dry heaves into action. I must say that I should have a 6-pack of abs when this is done. My poor tummy has been just pulverized.

The mucus is simply gross. It smells like dirty feet, so the ONLY smells I get are dirty feet, poop, and mold. I have progressed to a point that I disgust myself, and it's not pleasant in any way. The mucus is thick, dark brown (and or red from blood), and the feeding tube does nothing to ease the expactoration of it.

However, being the 30th, we've now finished the last day of chemo. My wife made me ring a bell (to me, it really didn't matter). We're not "done" until Friday, the 2nd, so it felt a little early to me. Oh, well, she has much skin in this game as I do. It means that she has as much claim to that bell as I do. So, I will ring it.

And then we'll ring it again on Friday.

July 31

We have the milestones now.

Aug 2nd, last day of radiation.

1st week after will be the worst week of it all as things continue to open up from the radiation.

2nd week after will have the sunburn healing (skin easing on the red)

3rd week after the pain will begin to subside. This week should be the migration of wanting to die to just being miserable. Swelling should also begin to subside.

4th week after, mucus will begin to return to normal, and taste will start back up (but will be interpreted rather oddly because the body is definitely not used to it - e.g. things will taste different, or wrong).

5th week - shifting into uncomfortable. At this point, sores in the throat should be healing up, and pain in swallowing should be minimized, allowing the potential to move to eating real things.

August 2

Today was the last day. Gagging is high. Exhaustion is extreme. This is one experience I will be glad to have done. I did hit the gong and the bell for both chemo and radiation.

My strength is failing. My timetable is what I'm holding on to - I didn't come this far to quit. There is zero appetite for anything.

The nausea has caught up - I ended up dry-heaving, then tossing morphine, stomach acid, mucus, and blood for 5 minutes. What made it worse was the fact that I was sitting on the great white porcelain throne (toilet) when it all hit. I can't clear the sinus or throat, so I'm stuck with the acid burns on top of the radiation ulcers. Joy.

But, I did manage to fix some sprinklers, so I'm calling this day of ground zero a success!

Aug 9

I waited intentionally for a week. Doctors called this one of the worst weeks of the process. This has been validated.

Sores are still opening up
Swelling is not dropping (we knew this ahead of time)
The sunburn has not diminished yet
Mucus production has been through the roof, and the mucus constantly contains both fresh and old blood
Mucus production leads to volatile dry heaving as the body tries to clear the throat so you can breath
The whole mouth aches terribly (even teeth)
Swallowing is at an all time pain high, requiring controlled substances to maintain some sort of sanity

However, there have been some good signs. Two days ago, over night, it felt like I'd suddenly hit a plateau, meaning the swelling stopped increasing, and the mucus production seemed to hit a constant rate (instead of increasing every hour). I fully expect that (now that I'm officially starting week 2) the sunburn will be the first to go. We've been using a combination of Cetaphil and Aquaphor, hoping to alleviate one issue. The more issues we can fix, the better the body should be able to focus on healing the other parts.

Aug 10

I'm adding out of the every-week cycle, simply because of changes. The mucus seems to have slowed drastically, and some of the sores have started to heal over. We had that pegged as week 5, but it is day two of week two, so I think my body is just healing quickly. This is fantastic news.

Mucus has had fresh blood for a while now, and now it is coming with a combination of fresh blood as well as old blood. Some is red, some is dark red, and also nose boogers are becoming hard and crisp, like scabs finally starting to peal off.

The pain of swallowing seems to have only slightly diminished, with more pain on my tongue (where it feels like I bit it - that has been a painful spot for a while). So, even though week two is supposed to be "wanting to die", that glimmer of hope and the faint bit of recovery has shifted my psyche already. Woohoo!

Aug 16

This has been a whirl wind of a week. Yes, things started improving. Around Monday, pain for swallowing had dropped significantly (yet another miracle), and the mucus was drying up. Swallowing had shifted from excruciating to barely bearable, but that was a HUGE shift.

Then, on Wednesday, the oncologist did his examination, and panicked. I don't have Thrush, but he said, "just in case" and threw a prescription at me. Four times a day, for seven days, I have to rinse my mouth out with this. We're 24 hours into the prescription, and I've learned one thing....

... four times a day for seven days, I'm going to have violent reflexes as my body continues to cough this damn medicine out. It's a rinse-and-swallow, OR a rinse-and-spit. I can't even get to the swallow before my body is tossing cookies I haven't even consumed. And this tossing of the said cookies is brutal, and goes on for 5 minutes.

Doc said "try diluting it". No change. I'm glad my wife took a weekend with a friend to go and relax, because this will reduce the amount of time she has to hear me trying to expectorate a lung. I keep coughing up bloody mucus.

However, aside from this every-4-hours reflex, things aren't bad. No taste yet, and swallowing does still hurt. When that pain subsides, I'll try a little ice cream here and there (not even a teaspoon) until (a) my taste returns, at which point I'll start increasing the amount until (b) my stomach can handle it. It has been five weeks since I've had anything substantial, really, so I know the stomach will take time to recuperate.

Aug 24

Yesterday was the three-week mark. The day before I finished that ugly medicine. I've dealt with clogs. I can smell (at random times), I have a constant ringing in my ears, and my mouth has gone from mucosal to very dry. I've started sipping water lately, and we went to a speech/language therapist who made me eat quite a bit of food. My stomach has been upset ever since. I have to blow my nose every time we do a feeding tube line flush just so I can continue to breathe. It's not wanting to die, but it is miserable, just like they said. Next week should start the healing processes, but the last time we looked at my bloodwork, the red and white blood cells were still very low (hadn't recovered). So Friday's "hydration" infusion included vitamins. It seriously looked like I was being given an IV of mountain dew.

Aug 28

We hit an odd snag. The feeding tube clogged. While we've been able to clear this in the past, we tried for 48 hours to clear it. At that point, we spoke with the oncologist about simply removing and trying without, and then with the dietician as well. Both concurred that we might as well pull it, and if we can't get the nutrition, we can get it put back in. We DO know that they'd have had to pull it before putting one back in, anyway, so we're simply starting this process a wee bit early.

As to being able to swallow, the problem we've faced was the pain in the past as well as not having taste. (Seriously, why eat if you can't taste the food or enjoy the texture?) I, in an attempt to "detect" when taste returns, have been putting a drop of maple syrup or honey on my finger tip, and then running a test.

That little test began to show signs of life on Monday, the 26th, so hope has arisen. Granted, that was very fleeting (taste was gone within an hour and didn't return that day), but again, bright and early on Tuesday (when we detected the feeding tube clog), taste was back, albeit again, fleeting and diminished (the first one probably shocked me more than anything). It returned later that night, and has been hit and miss since (not scheduled for mornings only). When we did a hydration, I had a blue raspberry jolly rancher, and that taste was hit and miss (took me a few hours to finish it).

So, arriving at this point, it's pretty clear that the tube was being worthless, and if I can get by without, I'm that much better. So, I pulled it (just like my stitches - on my own and in privacy). I double-bagged the feeding tube - that was disgusting to remove. Just the sight of it triggered nausea, which the nurses did warn me about (they were kind enough to say that I needed to be somewhere close that I could puke).

The tube was approximately 54 inches long (nearly 5 feet), or 136 cm in length.

We've been consuming 2000 calories nightly with the tube just to maintain weight, and the doctor said to get that in half as the first goal (dietician confirmed), then work from there to get back to the 2000 calories (and potentially more).

So, my first objective is to figure out how to consume 1000 calories. I've been given hints by the dietician such as :

Avoid clear liquids (other than water, because of the potential for choking on it while re-learning how to swallow).
that cream of wheat cereals are okay, with lots of fat added (ice cream added is a good thing).
Things with potent flavors will help get past the minimal taste issue (things like onion, garlic, etc). I'll vouch for this - I added garlic salt to some cream of chicken soup, and the before and after were definitely different!
Fruit nectors (not juices) are thicker and good.
Any cream-of soups are okay (cream of chicken, etc - I didn't know they made a cream of bacon or cream of potato, or even a cream of asparagus, I'll try them and see).
Overcooked noodles are a plus. If needed, add butter, but don't hesitate to use any cream-of soups with the noodles to increase calorie count.

We've been doing hydration three days every week. Last Friday, I thought something was wrong because the stuff they put up looked like Mountain Dew (I wish I could taste some right now).

The yellow bag on the IV - forgive the hat, I was using it to block out neighboring patient info

But, it is just hydration with vitamins (B1. etc).

The cream of chicken soup only has 130 calories. Damn. To get to the 2000, I'd have to eat 15 of those things a day. Luckily, I have other options, and I can't eat one (only 1/4th of one, so about 30 calories). I might add a snack pack pudding (90-100 calories), just to add intake, for a nightly meal. I KNOW I'm going to have to schedule things on my work calendar through out the day to remind me. It's been 6 weeks since I've really had much, anyway, up until the swallow therapist almost force-fed me a week ago.

So, this will make for an interesting few days.

Sept 6

We went in to the speech/language therapist, and ran through a barrage of tests while my throat was scoped. No pictures of the sinus (where I still have discomfort swallowing), but the throat looks good. In fact, the therapist said the swallowing was in good shape, and she cleared me to eat anything. I know I still can't, but I'm taking steps. We moved from liquid/puree to soft food (I ate some refried beans). Hopefully next week I can progress to some more solid foods.

Sept 10

Well, it's been a while since I've had a good, detailed post (skipping the last one, it's been what, 12 days?). The caloric intake has been to the 1 kilocal state, so I've at least gotten to the bare minimum. I still have little endurance, and I don't know if that is the calories, or if that is simply because we nearly killed the body (that's what we do through this process).

So, that's a good sign. As to the other tube, I went in for hydration yesterday, and was told we didn't have an order for it. So, I did not get hydration. However, over the last week, I graduated from the tiny toddler cups to the regular glassware, and I think I'm okay. If I can maintain this stance, it means I can probably have the PICC line removed in the short future. I was told by nurses that the PICC line usually heals within 24-48 hours, so depending on how strapped my body is for healing, I could potentially be able to shower without a sleeve as early as next week (I'm guestimating here, there is no medical timeline provided).

So far, keeping myself hydrated is working. It seems like my body is slowly returning to normal (I know I still have a long way to go, but progress is progress). In fact, this morning, I could actually TASTE SWEETNESS, which is one of the last hurdles I have. It is minimal (I could barely taste it, and it didn't last long before I couldn't taste it), but it is a step in the right direction. The next step I have is being able to eat breads without it feeling coarse or abrasive on the back of the throat when swallowing. Hopefully that comes this weekend.

So, my next goals are :

Swallowing breads
Eating chips and crackers
Start regaining strength

I need to keep this one going!

Sept 12

Well, after meeting with the oncologist yesterday, and not having hydration for nearly a week, they ordered blood tests to make sure the kidney function was still good. And, once it turned out to be good, the doc told me I can come in and have the PICC line removed.

Before removal

After removal

I no longer have medical tubes keeping me alive!

My water levels are solid, my caloric intake is not good but not bad, and the doc has cleared me until after the next PET scan is run, so I don't need to see them for a few months. The amount of time to see doctors is spreading out and clearing!

I do still have ulcerations in the mouth and sinus (oncologist acknowledged them), so I have to keep things to soft foods for now, but slowly making progress!

Sept 14

We pulled the bandage covering the PICC line hole (it's been over 24 hours, the nurse said this was okay to do). For the first time in months, I can take a shower without some weird sleeve on my arm that I have to prevent from getting wet.

Granted, this was gross when pulled (similar to the mucus). If you look closely in the photos, there is a "strip" of yellow, pliable-but-gelled gunk. When pulling the tube, this was in the vein, and it stretched, and then snapped out of the arm into the bandage. While gross, it was pretty cool to witness.

The milestone feels like I'm progressing, even if I don't actually feel like it. Still have some swallowing issues (I have homework from the speech/language therapist for swallowing exercises), so I'm working on that. Gotta keep going!

Sept 29

The last spell was... interesting. Went to church two weeks ago, and caught some sort of virus. That set me back, for sure. While I'd normally be back to normal a day after feeling all the things a virus does, this one took a full week before I turned the corner and started feeling better (which happened this past Tuesday). Last week, in the middle of this, the oncologist said "go to the instacare immediately" - I guess I still have an immune system that is scraping the bottom of the barrel.

So, I went to instacare - who ran one single test - only to find out definitively that I do NOT have COVID. No other tests, nothing was done. It's viral (symptoms, which I'd already determined, and we ran a COVID test at home a day before so we knew it wasn't that already), but we don't know what virus it was.

So, since the medical providers cared less about determining what virus an immune-compromised cancer survivor had, I determined that it was a matter of wait it out and hope it doesn't get worse. 3 days later, I stopped getting worse, and 2 days after that things started to finally get better (this past Tuesday).

On Wednesday, we went to the radiation oncologist, who scoped my sinus. Lots of gross mucus still in the sinus, but the r.o. determined that there is no outward (if you can call the inside of the sinus that) sign of cancer. That doc now thought there was no need to visit until after the PET scan, targeted for early November. This is good news as it means that October only has to deal with the Speech/Language Therapist and also the Lymphadema Therapist. The second one has had one appointment, and only one more is likely needed (in two weeks).

Caloric intake has stalled at around 1.5kcal per day. Friday took a beating because of some health issues, but yesterday, it was pushed to 2.3kcal . I knew I needed calories, so a Five Guys burger, fries, and a shake was where I ended up (I tried for simply some waffle fries and chicken from Chick-Fil-A, but lines were too long and they had their inside ordering shut down). I was already nearly at 1.5k before the burger. I scraped everything off of the bun, only ate 1/3rd of the fries, but I did eat that shake. That managed to put me in caloric intake higher than I've had since the feeding tube.

Taste is back, sort of. There is a mingled taste after a few bites that is still causing things to not taste good. The Radiation Oncologist called it "metallic", and I can see how it could be considered a metallic taste, but it's not quite metallic. It's hard to describe. It is the taste buds still waking up. A few "chews" of a good taste, and then this one is back. I hope that if I continue to chow down on things and force volumes, that will go away.

Well, that and the mucus being removed. Yesterday, I nearly threw up as my body was trying to clear a small piece of mucus. Two cough-ups of dark brown mucus (yay! A little more!), and then while working with the laser cutter, a little more in the nose caused me to run for the bathroom to blow my nose.

It was thick enough, that when it came, it brought a lot more with it (and I am celebrating that one). The extricated mucus was about the size of a quarter in every dimension, plus there was a chunk of hard scab with a curved gap in it from the feeding tube of so long ago.

I never thought I'd be posting pictures of boogers.

One more ulcer has been healed! I think I'm nearly there where my body can start to focus completely on regaining strength and endurance!

Oct 5

Okay. This has been a week of progression. I still suffer from something similar to "orthostatic hypotension", which is what typically causes you to feel light headed, as if you are going to pass out when you go from a laying or sitting position to standing. Is that what I have? I have no idea. But, the symptoms match up. Causes are lengthy bed rest (which I don't really have a problem with) or dehydration (which I also don't have a problem with - I experienced the issues when I was getting hydration at the chemotherapy fusion center). It can also be caused by anemia, which I know I'm dealing with a little bit, however, my oxygen/O2 stats have been high 90's for the most part (e.g. I'm averaging a 98% oxygen saturation), so I really don't know what could be causing it.

Aside from that, and lacking in endurance, I'm doing better. I had the flu shot on Thursday (under advisement from my oncologist - make absolutely certain you consult with your specialist in a timely fashion). That caused weird reactions (I know, I'm supposed to wait for 10-15 minutes after the shot to make sure there is no reaction, but I knew I'd feel it later and simply left to get home quickly before other effects hit). For example, when vision flashes, or patches of your eye sight go white blind like you stared at the sun too long, if you have a migraine - yeah, that, but without the pain and aches of a migraine. It was weird.

I ate alphabet soup, so I've had good vowel movements (sorry, I just HAD to work that in). Speaking of vowel movements, since I've added other dietary things like a hamburger or fries, or a pizza (yes! It was divine!), that's improved. For example, I had a chicken burrito last night with some habanero spice to it. That was also delicious! And today, my dad brought me apple crisp. This isn't your normal apple crisp that has oats in it (I'm allergic to oats), it's an amazing substance.

We did find out this week that because I haven't been drinking our Mountain Dew, it's all expired. And this is stuff that is difficult to track down (it's the "real sugar" stuff, and in this area, it is rarer than hens teeth). If it tasted normal, I'd still drink it, but it tastes like it's expired. Also, I found out that Pepsi no longer makes the stuff. [sigh]. I stop drinking it and a few months into that they stop making it. I know for certain that I wasn't keeping that branch of production alive, but it does feel that way.

I still do have that "metallic" taste. It hits at random times while eating, but I'm forcing myself to eat. This past week, I averaged 2,156 calories per day, which is a huge win (the feeding tube was averaging 2,000 calories per day for reference). My weight has not only stabilized, but started climbing (up to 186 lbs from the previously-stable 183 lbs).

And, get this.... the bald patches of my jaw, the top of the back of my neck, legs, and crotch (yes, I did not expect hair loss in the crotch area, that was a bit of a surprise) has just barely started to grow back. My mustache is very salt-and-peppery, my hair has definitely gone gray through all of this, and that is just the stuff that stayed. I've been told that hair is more curly (which could be bad with me because I already had very wavy hair if it got longer), because the pores are very flattened after the chemotherapy. That's another thing we now get to witness the truth of.

Oct 30

It's been a few weeks. The taste issues have come and gone. Nothing is set in stone, so there is still hope. However, a few weeks back, I had a flu shot. I've had a congestion problem ever since that point. Additionally, my feet (tops and bottoms) have been tingling, and the last few days I've had swelling in my feet (painless, but weird). It might be classified as a peripheral edema (a side effect of Gemcitabine). I'll message the doctor to see what he says.

Nov 13

This afternoon, I met with the docs to review a PET scan. The news is official - I have been declared cancer-free! This is a MASSIVE change. I am not even close to 100%. In fact, I'm not even close to where the docs said I'd be (they said around 70%-75% back to normal in December, I'm more around 50% as of right now). I still struggle to eat (it's a chore to me, not something that is pleasurable). I still have the Neuropathy (tingling in my feet) and the swelling. I still have that awful taste that comes and goes and clouds things. I still have balance trouble. I still have the broken baroreceptors in my carotid artery that help the body self-regulate blood pressure (which makes the world spin when I stand up, some of which is pretty bad and takes me to my knees so I don't fall). I still have patches of beard that doesn't grow. I still have the tinitus (ringing in the ears).

All of these issues are side effects, and may or may not be permanent.

But I'm CLEAR. And that's enough.

Some folks who experienced head/neck cancers at the same time haven't had swelling go down far enough to be able to have their PET scan. Some are still struggling to even get as far as I have. I know that there is a power in prayer, and blessings do follow you when you seek God.

My weight dropped from 217 lbs down to 188.4 (low was around 182).

My caloric intake has been hard hit - but I'm keeping it around 2k calories.

And my A1C has settled - even after the sugar/glucose intake trying to meet my calories - back down to 5.2 (was at 7.3 in May).

But I'm GOOD!

Stuff we've learned :

What hair that doesn't fall out does get brittle and breaks off easily
Chemotherapy slows down everything (nails grow slower, wounds take longer to heal, etc)
For nasopharyngeal carcinoma (stage 3 in the lymph nodes in the neck), when the doctors say the first two weeks of radiation will be miserable, the subsequent two weeks (three and four) will be miserable, and the next two weeks (five and six) you will want to die, they are dead accurate (no pun intended)
You will lose weight when you can't taste, and swallowing hurts - do NOT hold back on getting a feeding tube to get you through it
When the radiation oncologist says that (after the radiation is done) you will have the six week schedule in reverse, it is absolutely true - first two weeks of wanting to die, two weeks of misery, and then two weeks of discomfort
If you experience a complication (such as a virus), know that your body cannot do what it would normally take a day or two to recover takes two weeks, and your body doesn't heal during this period - it seriously takes you back
If you feel like you are getting sick, go to the emergency room (wearing a mask, of course), and do not settle only for a COVID test - make them determine what you actually have because that could result in medication that can help you through it faster
Do not expect taste to suddenly be back and normal - it took an extra two weeks after the discomfort for it to start, and I couldn't taste sweet things for a long time (it's still diminished, but a few weeks after that, it's starting to show up)
Once the six-week-after periods are done, you are not "normal" - your body still has a very compromised immune system, your stamina is still in the toilet (where your immune system is, surprisingly), and you can't just go back to doing normal things - you need to start from square one and work back into things, and that progression takes time - two months after finishing chemo/radiation, I'm probably about 20% of normal
Expect that moving from sitting to standing will have fainting, dizziness, and lightheaded symptoms for a long time

There is one thing I'd list in the "learned" section, but this is simply too big to go with the lesser stuff above.

Take time to find the small miracles. Trust me, you NEED this. God is a God of miracles, and even if He doesn't make the cancer disappear, He WILL walk with you through this challenge. He will bless you with small things, like neighbors who step up and bring in garbage cans, or weed a small flower bed. He will bless you with people who are going through the same challenges - reach out to them so they know you care, and they'll do the same. You will form strong bonds and new friendships. Families grow closer together.

Just find those little miracles. They are there.

Altering an Old Toy Wood Lathe

2024-09-18T08:11:00.000-07:00

Way back in the day, I picked up an old, tiny wood lathe when I was picking up some other tools (the guy threw it into the lot). Well, it's been sitting around for a few years, doing nothing. Essentially, the bearing surfaces (no real bearings, just spindle directly on headstock) were not concentric, the tailstock was frozen, and the tailstock handle was broken.

When going through that cancer stage, I needed SOMETHING to do, so I elected to make this usable. I finally decided to turn it into a dedicated pen-turning micro lathe. I started out with the headstock, went to the tailstock, then went to the motor to power it.

The Headstock

I picked up a few bearings, measured, and started to work.

I used a boring head to bore out for the radial needle bearings, and the thrust bearings just fit in like expected.

I'm using a 4" router ER16 collet, and put the headstock onto the mill as my spindle, because an ER16 collet will hold the 7mm pen mandrel perfectly.

Yes, I don't have the pulley installed there. It's marked and the set screw divot has been bored on the spindle, but I still need to drill it out length-wise (so I can easily shorten the pen mandrel without adding spacers), and thread the back end (left handed m12x1.0 - the whole spindle is metric, I might as well do metric).

The lengthwise-drilling was where the puckering came in. I had to drill 4" on a stainless steel 1/2", ground and hardened rod shaft, with a drill bit diameter of 1/4". [shudder].

So, I started drilling. About 1.5" in, everything ground to a halt (no pun intended because of the next paragraph, seriously).

I knew I needed a better drill bit as well, and a longer one, too, (that meant special degree points, and it needed to be an "aircraft" length). I went through this whole project of setting up a drill bit sharpener (which wasn't long enough to accommodate an "aircraft" drill bit needed to get the length of the spindle, so I had to modify the device with a longer rod so the shoulder could support it), then ground the drill bit with a 118 degree point instead of the 135 degree point it came with.

I then took everything in to the lathe and started drilling again, and got a little further before it ground to a halt again (seriously, no pun intended). I re-ground the drill bit and drilled again. I repeated grinding a few times with drilling until I had punched all the way through.

Now, it was next ready for threading on the end of the spindle so I could lock that spindle in place in the headstock. Back to the lathe, I set up for a 32 pitch, left hand thread (I didn't want deep in the spindle with such thin walls) and cut the thread.

Time to figure out how big to drill the hole for the nut that will lock it in place, I pulled out the machinery handbook, and started doing math. I needed wire diameters, micrometers, outside diameter of the spindle, and some mathy skills. Then I just got sick of trying to think, and pulled up a thread wire measuring calculator in order to get the root diameter.

I put the stainless plate into the four jaw, and dialed it in. I drilled to a 27/64's drill bit to open the hole to 0.421" (surprise! the hole was slightly smaller than I thought it would be, and that's just fine for this).

The interior threading was a treat. If I had a 0.472-32" left-handed tap, it would have been a quick job. However, my threads in the new spindle were flat at the root, meaning that the crowns of the threads on this needed to be flat, too. It took a lot of cutting a thread, flattening the crowns, and then repeating until I got the fit I wanted. Granted, this is slightly tight still, but it is good enough for me.

Now I can take this "nut" out of the chuck, and drill a cross hole (metric 4mm fits well on the socket head within the walls of this thing, so it's going to be perfect). So, I need to drill a 3.5mm hole (0.138", or a #29) all the way through the nut, followed by a 4mm (0.1575" or a #21 with added slop), I have the clearance for the threading. If I then use an end mill to put the socket head in deeply enough, I'll be ready to chuck the spindle in, add the nut and tighten it down, then turn the outside edge round and knurl it, then face the remaining side. Well, at least that's the theory.

I wish it was that easy in reality. This stainless does a fantastic job of work hardening. [grrrrr]. I drilled through it with the #29 (which is 0.136", not 0.138", and is a bit shy of the 3.5mm hole I need to tap with). Then halfway through with the #21. It was ugly, but it did it. I went with a letter I drill bit (0.272"), and dulled that thing faster than Spaceman Spiff loses focus in school. Went with a "J", not much further. In frustration, I swore a little and grabbed a 7/32" end mill (0.218") and plunged it in. That finally let me get to depth, but it leaves a gaping hole on the side.

Okay. Now I just need to tap it (beyond swearing just thinking about it) followed by a slitting saw to open the nut up a little bit. Then we can make it round on the outside, and flat on the sides.

I needed a break from the shaky gut wrenching that is tapping hardened stainless. So I broke out the 3D printer (Creality CR-10s), created some nuts for this small lathe in ABS (using a PEI build plate/printer floor, a brim, and 255 degree nozzle with a 95 degree bed temperature), and printed them out.

These have a raised lip on the thread side that will slide between the bed rails on the lathe and not turn, allowing me to simply tighten the 1/4"-20 nut on the top side. I did three, because I thought they'd be great for the tool post and tailstock, too. So with the nuts out of the way, it's back to the tapping, sawing, and turning of the "nut".

I tapped while it was still in the lathe as much as I could (up until the hole clogged up with swarf), and then used a slitting saw to slice it apart.

I then filed the burrs off, took it in to the band saw, and cut some corners so that it was closer to round, and I chucked it up.

Now I have to face and turn, then I can reverse and face so that ALL surfaces are parallel or perpendicular to the spindle.

Now to make it really round. This was less of an issue once I got the collet tight enough, and the cross screw tight enough to prevent it from slipping off, but I'm loving this. I tried to leave as much on the outside as I could, and wasn't worried about the inside (where the clearance hole came through the side) as that would be covered by the bearing race. So, the spindle is now ready to be assembled.

I put things together just to make sure, and bolted the headstock down :

That should work for the headstock. At this stage, I didn't know if it wouldn't work until I got it under power. But, next up, is the tailstock.

The Tailstock

For the tailstock spindle, I faced both ends of a ground 1/2" rod, and drilled and tapped one end. I then ran a calculator to see if I could approximate a #0 Morse taper, and came up with the following drill depths :

Next up, I need to create the motor mount and the tension mechanism. That will make this at least "usable", though I won't like it until I have a proper tailstock assembly with a live center (e.g. a morse taper #0).

Drill Bit	Drill Diameter (Inches)	Drill Depth (Inches)
F	0.257	1.9039208276299027
G	0.261	1.827072358300744
17/64	0.2656	1.7386966185722121
H	0.266	1.731011771639296
I	0.272	1.615739067645558
J	0.277	1.5196784809841102
K	0.281	1.4428300116549515
9/32	0.2812	1.4389875881884935
L	0.29	1.2699209556643458
M	0.295	1.173860369002897
19/64	0.2969	1.1373573460715463
N	0.302	1.0393755476768698
5/16	0.3125	0.8376483156878285
O	0.316	0.7704059050248144
P	0.323	0.6359210836987872
21/64	0.3281	0.5379392853041098
Q	0.332	0.46301202770817973
R	0.339	0.3285272063821525
11/32	0.3438	0.23630904318716262
S	0.348	0.1556181503915468

Once this was drilled out, I started to use the roughing taper reamer, followed by the finishing taper to get a perfect fit. Now, before anyone complains, I absolutely made the gut call to only run the taper 1/2 way in. This was to get as much tailstock ram articulation as I could. The deeper the taper, the less tailstock I have to work. I suppose I could simply just chop the dumb thing in half so it sets deeper, but I'm only working on pens. This will be just fine.

With that, I could run the 1/4"-20 threaded rod (or bolt - I didn't have the rod) to calculate the depth required to pop the taper out.

1.314" (or 1.5") of 1/4"-20 thread on the end of the spindle. Now, I can get the bushing. Just to make sure, I measured the old rod.

Just as I recalled, 1/2"-13. I grabbed some hex bar brass for a bushing, and cut the 1/2"-13 threads on the outside, bored it through, and faced it off. I had the bushing :

Okay. Now to put that in place, and I just HAD to see the ram in place, too.

Yeah, that's it's big brother sitting in the background.

That over-hang of the taper center is just crazy to me, but that's what I will need to deal with because any deeper and we'd be having a 1/16" ram movement before we disengage the spindle vs. losing engagement on the threads to the spindle. I can easily see how this size would be reason enough to use a spindle and thread all together as one unit like the original designers did.

Now on to the spindle itself. Grabbing some measurements just so I'm not out of whack :

Two parts to make are the spindle (on top there), followed by the bushing to attach the handwheel to the spindle. Cutting the bushing for the handwheel to spindle :

Now, please forgive me. I swear I grabbed pictures of the cuttings of the tailstock spindle, and the side milling of the tailstock ram (to lock it from rotating). But, I cannot find them anywhere. So, here is a later picture of the tailstock spindle, immediately after putting the detent to hold the handwheel to the tailstock spindle. Yes, it's still sitting on the vee block used to put that detent in.

Anyway, proceeding to the handwheel, I needed an 8-32 set screw to lock it all together. I drilled the handwheel AND the bushing together at the same time to keep them lined up. Then I tapped them at the same time, too.

That means I have all of the tailstock parts. To say, I was a bit excited at this stage, because I was still unsure of the functionality after I hosed everything up, I mean, made the parts. It worked in my head, but I was a wee nervous.

It is time to assemble. In goes the tailstock spindle, on goes the handwheel (locked in place), and then in with the tailstock ram. It's backwards (I kinda knew it would be back when I chose to use normal right handed threads rather than left handed threads that would require purchasing both a tap and a die, and I was already getting a little too deep on expense).

I made sure the locking lever worked, I made sure the spindle would retract and extend out. Then I put the live center in, and made sure that also would eject as close to the tailstock as I could.

Honestly, as many times as I've done things like this, I'm always amazed when it works. And this thing worked! Next up, I need to build the motor brack and the tensioning mechanism.

The Motor Bracket

First was to take a heavy chunk of aluminum angle extrusion (so that it doesn't twist), and mount the motor. I created a housing for the controller and 3D printed it so that it could get bolted to the bracket, then I cut out a chunk for the housing to fit into. Next, I drilled the bottom for hinges.

That leaves the task of creating the mount for the tension spring. I am going to use the headstock mounting bolt itself for one end of the spring, and then just mount the spring using a bolt to the bracket on the other end.

Before that, I need to get the pulley and get it lined up first. This will then tell me where I can mount the spring to, and then it's a matter of firing this little thing up!

I started out with small hinges on that angle bracket (2mm screws), and those failed as soon as I added the springs for belt tension. I replaced with a small piano hinge, and that got the thing up and turning. I had lost one brass pen mandrel nut, so I decided to make two more out of aluminum. I DEFINITELY like these aluminum ones.

Now, this thing does have some issues. The stepper motor and driver combination does bind, and when it stops rotation, it doesn't start. I might have to change this out, but.... this project works!

Just so you know, the stepper motor not having enough torque did finally overpower my "success", and I converted a small 1/5 HP sewing machine motor for use on it, with better results.

Now, for an odd bit of calculator for anyone who might need it...

Calculating drill depth for a Morse Taper

In case someone else needs to create a taper in preparation for reaming, here's the HTML-based calculator I whipped up. Just click the "load taper info" button and it will populate the taper size and through-hole data. The through-hole drop down can simply be ignored. This will tell you what size of drill bit to use, and how far to drill into the bar in order to get "close" to a taper socket from the large diameter end.

Hopefully it can be helpful for anyone else!

I'm Beaming

2024-06-01T18:16:00.000-07:00

I have an old beam compass. This was my first "beam compass", and it came from my wife's uncle (no one in his family wanted it, and probably didn't know what it was), so I inherited it. It was missing the beam, so I ordered a section of 3/16" hex bar, sliced it where the bar was bent, and ended up with three sections.

As this sat in my drafting kit, I wanted to be able to use all three sections. With my current health conditions, I've been banned from using cutting tools for a few months (and will be for a few months more), so I had to sneak this one in.

A quick run to the lathe with some 3/8" stainless bar stock, and I drilled a 3/16" hole through the middle.

The hex bar stock is 0.214" across the peaks, and that was right up the alley of a #12-24 thread (0.216" outside or major diameter). So, I tapped the "tube" half way to #12-24.

I then did an interrupted thread cut the hex bar stock on lathe, ran a thread cutting die to finish it, and turned the other end of the hex bar stock down to 3/16" to fit those tubes.

The beam compass extension tubes needed a set screw. I opted to make it like the other screws the compass had - a #5-44 thread to match. The only problem was that I did not have horizontal/vertical knurling wheels, and I've spent too much on tooling for this already to make it anywhere near worth while. But, I had the bar left over from the tubes, so I made some thumb screws with some diamond knurls.

The screws fit, and locked everything up as I'd expect.

This is done! (Except for maybe one part to hold lead instead of a hard point like a divider - we'll see.)

I Have a Ream

2024-03-23T11:57:00.000-07:00

Everybody knows that drills put holes in things. Most people assume they are perfect holes - but some people who need accuracy know that is not the case. So, what kinds of holes, and what to expect? My brief, inexpertise scale from worst to best :

Hand drill - if you need a quick hole and don't care about vertical, size, or anything else, do this. Even with "drill guide" blocks, you're likely to not get it vertical. If you are going to thread this hole for a bolt, please avoid this one. I will throw in cheap drill presses here - though with some work, they can be a little more acceptable toward the next category.
Drill press - these are good for getting rid of the above "vertical" orientation issues of a hand drill. For this category, I'm throwing in drill presses, magnetic drill braces, and I'll even throw in the old Craftsman "Port-Align" attachments to the above hand drills. They offer better consistency in your vertical alignment.
Drill in a lathe - sometimes you can throw the part into a lathe. These give you MUCH better control over axis alignment, plus the holes become much more - dare I say it - ROUND! Now, you thought the previous holes were round. Sure, they were, kind of. But now we're getting into a much better category and much more accurate concentricity.
Boring bars in lathes (or boring heads in mills). These give you not just a concentric and round hole - these actually let you control the diameters of the said holes. One drawback of these is the machining marks in the holes, as these are typically single-point-cutter tools that have radii on the cutting end, so while the holes are much more accurate, they still are on the rough side.
Reamers. These are specialty tools designed to give you the best finishes possible, while giving you exactly the dimension you wanted. The holes are round, concentric and parallel (provided you don't let them wander off course).

Now, I've started thinking through a low-profile, small, milling vise build, and I know I'm going to need better holes than I've done. So, a "drill" isn't going to cut it, and what I really want are "reamers". I found a hodgepodge of reamers on eBay that might cover the sizes I need, so I bought it, and classified their sizes (not all are stamped).

So, I measured these, not with actual use, but using some metrology equipment, which means these are likely +/-0.002". For example, the 0.373 might be a 0.374, and the 0.374 might be a 0.376. The only way for me to tell is to use each individual reamer, but I'm too tired and fatigued to drill that many holes without good test equipment that I can chuck some of these reamers into without moving drill heads, bases, or anything the like.

For now, suffice to say I have the following sizes in hand :

0.155"
0.183"
0.187"
0.189"
0.218"
0.219"
0.246"
0.280"
0.308" (2x)
0.309"
0.310"
0.311" (5/16") (2x)
0.354"
0.373"
0.374"
0.382"
0.498"
0.499" (2x)
0.500"
0.515"
0.749"
0.751"
Morse Taper #2 (this was a bit of a surprise!)

Next up, trying to feel a bit better. This could be a longer road to travel before I can get back to projects, but I've gotten this far. If I can do so, perhaps I can start doing the mill vise.

Bolt Organization - Over Engineering

2024-03-21T18:17:00.000-07:00

We've all beheld big bolt bins, and many have begrudgingly bought into them. I held out for as long as I could, and finally declared that I need something better.

I don't want bolt bins. In our house, we had some old canning jars left over from eons ago, and I decided to use those. Yup, glass jars to hold my metal bolts. It should be fine. So, this write up is about the journey to create a bolt storage system.

This project has been on my back burner for some time. Unfortunately, my spouse came home with some Pyrex for a wedding gift that we ended up etching names into for personalization. This was exactly what I'd been wanting to do with the glass jars (save the addition of some colorant to the etched glass when done for contrast), and since we now had a bottle of glass etchant (Armour Etch), I figured I had nothing to lose but start the storage.

First, the etching of the Pyrex :

Yes, I am now worried about the end result, since the etching seems to be a bit hidden - we'll see if the colorant additions (Sharpie) help this later.

I generated a list of bolt sizes from an old Machinists' Handbook to get a bit wider range of sizes, so not having only modern stuff, but some of the more common older sizes, too. Note, I only did imperial since that is most of what my restoration work uses). I fed that into the vinyl cutter, and had my sticky stencils setting in storage.

The next task was to simply cut the stencils out and apply them. I used some "transfer tape" designed for transferring vinyl to do this, and set the stencils into place on the jars.

Short jars were for fine threads, and large jars were for coarse threads, because most cases I've seen had the longer bolts with coarse threads. It will make it easier to know that the full quart jars were coarse, and the pint jars were for fine threads. Visual organization makes for fast grabbing of what is needed when you need it.

I applied the Armour Etch, and let it set for 15 minutes (should be fine in 5, and the additional 10 minutes likely didn't help much, which doesn't matter for me). I rinsed them off, and then I let them dry.

I grabbed my red Sharpie, and filled in the stencils. This made it fast work to color in the lines. The Sharpie doesn't help much (less sticking to glass), but the tint might show up better against black, silver, or yellow (grade 8) bolts. That's why I chose the color - we'll see how well it works when we get there.

Now, peal off the labels. I'm getting more nervous that the Sharpie isn't going to help much.

Okay, with that against a white background, the Sharpie shows superbly. I expect it will help with the silver-colored bolts (zinc-plated) just fine. The black ones... still unsure of. I'm going to let these jars set for a while to get as much locked in place as I can.

I ordered some brackets I designed to bold vertically. They should match up to an 8mm through hole (so I can use a skate bearing, or a 608 bearing - I have a number of 608RS's that I can use - with a 7mm thickness, and a 22mm outside diameter, they should fit a 1.5" x 5/16" aluminum barstock - all I have to do is cut and drill them out.)

The SendCutSend brackets actually look really good. There are some powder coating issues around the bearing hole (flashing, or extra coating chunks), but really look fantastic over all. The DXC file for the brackets was loaded to Google Drive if you need it.

These use M8-1.0x30mm bolts for the 1/4" brackets. You need to use an M8 washer for these, or the bearings will be flush with the brackets and fail to operate, so, one washer per bearing. Those washers must be less than the 22mm outside diameter with ah 8mm inner diameter. Thickness doesn't matter, because they don't make washers thin enough to not work. My washers have a 16mm outside diameter and a 1.5mm thickness.

I cut some 1"x4" x8' common board into 4" lengths (about the diameter of the jars, and I have about 37 of those for these). The thickness is 3/4", so don't be alarmed if it doesn't match up. The lids will screw into the boards, the aluminum bar will screw into the ends of these, and then the bearings will cap that off. For these parts, I need clearance for the bolt heads and nuts. I started by boring a single board at a time, offset for the aluminum bar width for these 3/4" boards, and just decided it was easier to do two at a time. I used a 3/4" Forstner bit, a vise bolted to the table, and then simply added a chunk of aluminum that I could use to align the sides of the vise jaws to the boards as I put them in.

Okay - we're getting close. Here's how they will sit (the lids screw underneath, and I still need to make the parts that connect the boards to the bearings).

Next, I need to cut out bars for each bearing. These should have file holes in them, and be about the same length as the boards. The center hole will be exactly centered, and will need to be a 22" hole diameter. The others will need to be centered on the boards, so those will be 3/8" from the bottom side. A countersink on them will allow those screws to fit all the way in and not interfere with anything.

The bars (1.5" wide, 5/16" thick) were sawn on the bandsaw. I set it for a 45 degree angle to get more braces out of each bar, and then I'd just use the bandsaw end stop and flip the bar (you can see one in the picture below). I didn't care to take the time to mill or clean up the saw cuts other than a deburing tool. I only needed about 42 of them, but cut as much into these things as I could, just in case I needed some metric storage later. I wanted to get as close to a production-style as I could (so that things were all lined up the same way). On the drill press, I used my square angle brackets to get the small screw holes into the same places (I just flipped the parts to get the opposite end), then set the depths and countersunk each one.

Next came the larger bearing hole. It's a 22mm hole, and I'm too cheap to go buy a single drill bit just for this job. A 7/8" drill bit is 22.225mm, so I know there will be a little slack in these holes (0.0088" plus whatever the drill bit oversize is, to be exact).

It was just a ,matter of running through them all to drill, debur, and call those parts done.

Installation is simple. I installed one bracket (the end one), and got it vertical using a level. Then, I would slide one jar-assembly on along with the opposite bracket (for spacing), and then screw down the second bracket. Repeat until they were all lined up.

Instant storage, and when I need one, just find the jar, thread it off, and the nuts and bolts are contained. Thread it back on when done. Each jar pivots to make stowing the tray in an upright position and the seat up makes this flight a successful one.

Storing Pipe Clamps

2024-02-03T20:11:00.000-08:00

Ever since that table project, I've had pipe clamps laying on the floor of the garage, and when it rains, everything is covered in water. I needed to get them off of the floor. I used a 2x4 for each end, and I used another one to create the braces to screw that too the wall. I bored them out using a 1-1/4" Forstner bit, and then finished the cuts for one side (I left the complete hole on the second end).

Bolted them to the wall.

Now, my clamps are off the floor!

Ram 1500 Rack and Pinion Loose

2024-02-03T15:13:00.000-08:00

As I've driven my Ram 1500 lately, I've had these weird thoughts that the steering was getting loose. And then, on Thursday, it started "popping" when I was cranking the steering wheel all the way right and then to the left as I've been parking it. I panicked, and parked the truck until the most recent snow storm finished.

I slid underneath and inspected all of the steering knuckles and bots to make sure the linkage was in good shape. Nothing looked off, so I started inspecting everything, and that is when I noticed one of the rack and pinion bolt fasteners (holds it to the frame) had about a 3/8" gap. I could turn it by finger. Probably not supposed to be like that.

I grabbed my tools, and secured the bolt.

Now, I don't have the torque specifications for that bolt - and I couldn't find them anywhere (I don't have a service manual for this thing). So, I slowly incremented my torque wrench and tried the other bolt until I got to the point where the other one would turn.

That is where I put the torque on this loose one. Steering is again tight, and I feel much better about driving this thing.

Bernina 900 Foot Pedal Fix Request

2024-01-11T16:16:00.000-08:00

I had some family hand me an old Bernina sewing machine (a Nova, looks like a 900 but I cannot confirm this). It's not working. The pedal is simply not allowing it to function. Apparently, they had taken the pedal to a repair shop, and the shop turned them away, because it was only used for a short time, and the parts stopped being made a long time ago.

Granted, they have a few extra sewing machines, but they REALLY like this light weight machine, and were hoping I could get the pedal fixed. It's called an "air pedal". It seems to function just like other sewing machine pedals, but it "floats", or at least, it is supposed to. This one just floated without connection. So, I took it home, and promptly took the thing apart.

First thing I noticed is that there is a barbell of sorts used to connect the pedal to the actual hinge hardware, and it is not in position. Problem?

Perhaps. The blue circle is around the barbell. The green circles are two concave holes where the two ends of the barbell are supposed to go (think of a hip joint - one round ball and one socket).

Here are some more views of the two halves.

Let's put it back together in what we'd call "the Right Way" (trademarked?) and see what happens.

Here we can see the ball joint in it's natural habitat, with the actual pedal ready for installation.

So, was it the actual problem? Yes. The machine actually started up this time. It immediately stopped, which sent me home with the whole machine for some electrical tests.

I flipped it over, and a pin fell out (one you might see holding sleeves on a brand new dress shirt that is being opened, not machinists pins or internal pins). I paused, flipped the machine upright, plugged everything in, and hit the pedal.

Nothing.

While holding the pedal, I rotated the manual crank wheel, and it started working. That's it - a pin held the whole thing hostage. It's back online.

Taper Size:
Through Hole Size: