TL;DR:

To delete all old versions of a document, open the document, navigate to File > Revert To > Browse All Versions… (opens Time Machine type versions view), select the most recent non-current version on the left, move the mouse into the menubar region to reveal the menubar (it’s hidden in full screen view by default), navigate to File > Revert To > and notice that your choice is now Delete This Version…. Hold down the option key and it becomes Delete Old Versions…. And that is the option you’re looking for.

After leanring and I didn’t find an answer easily so after I stumbled upon the answer via too much trial and error I thought I’d document it for next time I forget

In Numbers (as an example), if you’ve saved several versions of a document and use the File > Revert To > Browse All Versions… menu item you will be shown a full screen interface that looks like Time Machine backups. It has the Current Version of your frontmost document on the left, a stack of document versions on the right going back in time, and a timeline on the very right edge of the screen for quick navigation.

There is a “Done” button at the bottom of the Current Version that exits this mode or view. There is a “Restore” button under the document stack view on the right. If you you hold the option key down that “Restore” button becomes “Restore a Copy”.

What is missing here is a way to delete versions, or, better yet, delete all old versions.

To get to this functionality you have to do some magic and that magic is…

…move the mouse up to where the menu bar should be but is hidden in full screen mode by default and it will appear (!).

Once the menu bar is visible, navigate to the File menu and notice some things:

• If you have selected a revision by selecting it before opening the File menu (by clicking on the frontmost non-current version or using the navigation arrows to go back in time) then File > Duplicate has become File > Restore A Copy.
• Likewise, File > Revert To > Browse All Versions has become File > Revert To > Delete This Version…. If you hold the option key down this becomes Delete Old Versions… and that is the option you want to delete all old versions of the current document.

Cheers.

P.S. Sorry about the crappy WordPress.com ads. Migrating off of wordpress.com is on the To Do list, but it’s a long list…

I want to use some C++ code from a Swift Package, so I:

1) Added a target with the C++ code
2) Added an umbrella header
3) Added this new target as a dependency to the client target
4) Import the new module in the client target

But now when I try to call a C++ function, which I can autocomplete to, the linker throws “Undefined symbol”.

I’d never done this before so I thought I’d take a look. The definitive documentation I could find seemed to be “Setting up Mixed-Language Swift and C++ Projects” on Swift.org.

The poster on Mastodon wanted to use a couple of C++ files from https://github.com/aappleby/smhasher/tree/master, specifically MurmurHash3.h / MurmurHash3.cpp. Not really sure what this repo is all about, but I grabbed those files and played around some.

Got it working and you can see a test repo here (ignore the hacky Swift code trying to make some test parameters to call the C++ MurmurHash3 function with; not sure what this function is really expecting but I was just trying to solve the linking issue the poster was running into.

The key to getting this to work is the addition of:

swiftSettings: [.interoperabilityMode(.Cxx)])

on the target that is going to use the C++ containing target.

Take a look at the repo.

UPDATE: Added a second branch “three layer” which increases the complexity to match what the person asking was doing. It now has three layers of Swift modules dependencies before the base C++ module. So:

Swift Executable target (“testCmd”) -> Swift Library target (“SwiftMumurLib” -> Swift Library target (“MurmurTest”) -> C++ Library target (“Murmur”)

The key to getting this deeper level of nesting to work was to include the swiftSettings: [.interoperabilityMode(.Cxx)] in each target that was using the C++ target or a target that was using it even indirectly.

let package = Package(
    name: "testCmd",
	platforms: [.macOS(.v14)],
	products: [
		.executable(name: "testCmd", targets: ["testCmd"]),
		.library(name: "SwiftMurmurLib", targets: ["SwiftMurmurLib"]),
		.library(name: "MurmurTest", targets: ["MurmurTest"]),
		.library(name: "Murmur", targets: ["Murmur"])
	],
    dependencies: [
        .package(url: "https://github.com/apple/swift-argument-parser.git", from: "1.2.0")
    ],
    targets: [
        .executableTarget(
            name: "testCmd",
            dependencies: [
                .product(name: "ArgumentParser", package: "swift-argument-parser")
				, .target(name: "SwiftMurmurLib")
            ]
			, swiftSettings: [.interoperabilityMode(.Cxx)]
        )
		, .target(name: "SwiftMurmurLib",
			dependencies: [
				.target(name: "MurmurTest")
			]
			, swiftSettings: [.interoperabilityMode(.Cxx)]
		)
		, .target(name: "MurmurTest",
			  dependencies: [
				.target(name: "Murmur")
			  ]
			  , swiftSettings: [.interoperabilityMode(.Cxx)]
		)
		, .target(name: "Murmur",
				dependencies: []
		)
    ]
)

(Repo was tested with Xcode 16.2, macOS 14.7.5).

Today I noodled around and got the same thing working in Alfred. I’m just learning how to do Alfred Workflows (this is only my second one), so if there’s an easier way that doesn’t require using AppleScript, I’d love to hear about it either here or on Mastodon.

Here’s a screenshot of the Alfred Workflow editor for a “Show Series Menu” workflow that has two nodes. A Hotkey node which is set to the control+command+S sequence (^⌘S) and a Run NSAppleScript node which is opened to the AppleScript editor and shows:

on alfred_script(q)
tell application "System Events"
tell item 1 of (the processes whose frontmost is true)
tell menu bar item 2 of menu bar 1
click
click menu item "Services" of menu 1
end tell
end tell
end tell
end alfred_script

That seems to work.

The top of the line new Mac Studio M3 Ultra (32-core CPU, 80-core GPU) with 512G RAM (!) and a 4TB SSD is an impressive and tempting machine, but ~$11K (!!). Is that “ridiculously expensive”?

A more “reasonable” configuration (for me) would be Mac Studio M3 Ultra (28-core CPU, 60-core GPU) with 96GB RAM and 2TB of SSD and that’s only $4400.

Future proof it now that I’m not replacing machines very often and go 256G RAM and 4TB SSD? $6K. Bump the M3 Ultra to 32-core CPU, 80-core GPU, $7500.

I can’t really justify any machine upgrade because I’m not working again (yet), but it’s fun to price them out and think about it

What’s amazing to me is that this ‘around $5K’ price for a solid Mac is very close to to the “standard developer mac” I’ve been buying for quite a few years despite inflation. That means that basically the performance per $ has increased to keep up with inflation.

Decided I’d compare prices then to now and, since these have been business expenses, they’re in my quicken file going back quite a ways. Some are missing (not sure why!) and some are uninteresting (test machines so lower spec, etc).

Just for fun I put in the equivalent price in today’s dollars (am I trying to justify the maxed out Studio’s 11K as “reasonable?” Maybe! ).

(The prices below are apple developer program prices until the hardware purchase benefit went away (not sure what year that was). Had a large discount early on in the 1990s but that went down over time. The developer program cost $500 a year for years there, but the hardware discount made up for it).

• (1986-1988: 512Ke, MacPlus, Mac II, missing data)
• 1989 Macintosh IIcx in 1990 (dev discount was big then!) was $4200; $9700 today (and base retail was $5300 but I got more RAM and larger HD options).
• (1990 IIci, missing)
• 1991 PowerBook 170 in 1991 at $3600 + $1400 in RAM & HD upgrades in the first year; $5K $11,600 today.
• (IIfx, missing)
• 1993/1994 PowerBook 180 was $3,335 + $1700 for extra batteries, charger, RAM, modem, hard drive upgrade) so about $5K total; $10,900 today.
• (Quadra 700, missing)
• 1997 PowerBook 3400c was $4786 plus $700 of upgrades; $10,700 today.
• 2002 Titanium PowerBook $1750 + $450 upgrades; only 2.2K – great deal for a powerful laptop! $4100 today.
• 2003 17″ Powerbook G4 $2800 with RAM upgrade; $4800 today.
• 2005 Dual dual-core G5 Mac (QuadG4) for $2997 (had to get a bigger UPS to handle its startup surge for $250 though! ); $4900 today.
• (First Intel PowerMac, missing)
• 2007 MacBook Pro $1700 + $275 RAM & HD upgrades so right around $2K; $3100 today.
• 2010 27″ iMac $2200 + $100 RAM upgrade; $3400 today.
• 2011 13″ MacBook Air + case cables adaptors and accessories $2250; $3100 today. (Great portable mac!)
• 2017 MacBook Pro $3300; $4300 today.
• 2020 12″ M1 MacBook Pro $2,250 + $500 dongles ( USB-C switch and dongles); $3400 today.

Anyway. An amusing way to procrastinate doing a few chores

The take away though is that while they seem expensive, Apple computers are a heck of a value for the money considering how much more capable they are from the early ones and how the prices are basically the same despite this and inflation.

Variations: Chocolate chunks + hot red chili pepper (ground up)

(Weight is Double / Single batch)
250g / 125g | 4.5oz Sugar
450g / 225g | 8oz Unsalted butter, softened (room temp)
600g / 300g | 10.5oz plain flour (whole wheat pastry flour)
100g / 50g | 1.75oz brown rice flour (ground brown rice)

• Sugar + Butter in KitchenAid mixer with beater blade
  start slow but speed up to about speed 5 and
  cream butter & sugar.
  Scrape down sides and blade and run a bit more.
• Sift flours & add to mix
  Also add lavender or chocolate+chili pepper
• Mix AS LITTLE AS POSSIBLE, slow speed (lowest)
  The idea is to just mix the ingredients and then test to see if it’ll squish together and hold shape (indicating that the butter is well mixed in).
• Pull out and knead gently until smooth & roll into a log with diameter the diameter of the cookies you want.
• Chill for 2 hours+ in fridge in waxed paper (rolled up and ends twisted)
• Preheat oven to 350F
• Slice “1/2 inch” thick disks off the roll (I’ve always done 1/4 inch) and place on cookie sheet with space between them.
  (We’ve also done it on parchment paper on cookie sheet).
• Cook 18-20 minutes (until golden brown on edges and centers are firm).
  -> 28-30 minutes for double batch (two cookie sheets), swapping cookie sheets on shelves top/bottom half way through.

Chocolate + chili peppers variant:

1/2 tsp ground chili peppers (I used the Mexican market red chillies I grind for burritos).
1/2 to 3/4 cup chocolate chips (not really chips, little round chunks from bulk place. 63% cocoa with sugar (and milk?)).

Notes:

• Both the dough and the cooked cookies freeze quite well so you can make a big batch and freeze them for later eating.

A good product idea:

• solves a real problem that real people have
• these people feel the pain of this problem regularly
• they have the money to solve it, and will spend it
• there are enough of such people
• you can solve it well with software
• [ideally] there is little competition or only significantly inferior solutions to yours

You have to spend time speaking with people to learn these answers.

If your idea doesn’t pass the first 5, find a different problem to solve because this one isn’t likely to succeed.

There are lots of other things required to develop and market a successful software product, but working on the right product idea is a very important initial decision.

Background (feel free to skip)

FreeCAD is super powerful, but confusing, and, more importantly, buggy.  Spent a long time drawing the part and then STL export failed to export some holes.  Redrew the part a different way and some holes worked but others didn’t. GRRRRR!

Gave up on that project in frustration and moved on to more fun activities.

After that experience I did some research on alternatives.  

Fusion360 has a hobby tier that would be appropriate but I seem to recall them cancelling that at one point (and, apparently, bringing it back) and the paid subscription is way too expensive for the very infrequent use I make of 3d printing and CAD software (largely because of the CAD software hurdle).  

I want to love Shapr 3D, but again the every single month subscription cost vs how rarely I’ll actually use it is problematic, and I did run into some difficulties using it when I paid for a month to try it out.  I loved working on the iPad though.

I had a friend who worked at OnShape and had taken a look at it some time ago, and another friend on Mastodon has recently commented that the’s liking it, so that’s something I need to look into but it goes from “Free but share all your cad files publicly” to $1500 per year (way way too much), so that’s pretty useless to me.

Hello OpenSCAD

I have a friend who’s long sung praise of OpenSCAD and a talented maker I follow on Mastodon has made it clear he’ll use nothing else.  I tried it out briefly some time ago but the idea of typing in visual things seemed off to me; draw visual things, type textual things just makes the most sense to me.  But I’m to cheap to pay for the visual tools, so OpenSCAD is what I’ll be trying out for my next phase of experiments but this time I’m going to take the time to learn it more thoroughly.

Some initial cool/useful things I learned

So far I’m really liking OpenSCAD’s capabilities, though the macOS app’s text editor is very annoyingly different than the native macOS editing key commands and shortcuts.  I finally switched to using BBEdit (https://barebones.com) and an outdated OpenSCAD language module for it I found in the OpenSCAD repo (https://github.com/openscad/openscad/tree/master/contrib).  I plan on spending the time to update that language module and submitting a PR, but not until after this project is done (no Yak Shaving allowed! ).

A couple of really helpful things I’ve run across so far:

Modifier Characters for debugging and visualization aid.  

Super helpful.  I’m using % and # so far.

Example use:  If you are using one shape to cut out a portion of another, # lets you visualize the shape doing the cutting.

Without #, the following statements:

c_h = 20;
c_r = 10;
pipe_r = 16;
difference() {
    cylinder(c_h, r= c_r);
    rotate([90,0,0])
    translate([0,0, -(pipe_r/2+2)])
    cylinder(h=c_r*2, r = pipe_r);
}

Gives you:

If you add the # sign in front of the last cylinder call you can see the shape doing the cutout of the bottom of this vertical cylinder:

c_h = 20;
c_r = 10;
pipe_r = 16;
difference() {
    cylinder(c_h, r= c_r);
    rotate([90,0,0])
    translate([0,0, -(pipe_r/2+2)])
    #cylinder(h=c_r*2, r = pipe_r);
}

Which makes it a lot easier to see what’s going on.

Nested module definitions

This isn’t mentioned in the manual, but I think it’s going to be very helpful for statement modularization. Turns out you can nest module definitions inside each other.  

Here’s a made-up example:

module test_outer() {
    inner_test();
    inner_test2();
        module inner_test() {
        translate([100, 150, 0])
        sphere(10);
    }
    module inner_test2() {
        translate([100, 100, 0])
        sphere(20);
    }
}

test_outer();

Which gives you:

This is handy for complex part assemblies where you want a single line test_outer(); to represent a complex part with multiple sub-parts but don’t want those sub-parts to be constructible outside of the assembly.  Using inner_test(); outside of the body of test_outer(); raises an error.  Excellent.

Anyway, that’s it for this short initial OpenSCAD post.  IFF I find OpenSCAD working out for me I hope to post more as I figure more stuff out

Pretty neat poem.

The Calf-Path by Sam Foss – Poems | Academy of American Poets

The Calf-Path

Sam Foss

I.

One day through the primeval wood
A calf walked home as good calves should;

But made a trail all bent askew,
A crooked trail as all calves do.

Since then three hundred years have fled,
And I infer the calf is dead.

II.

But still he left behind his trail,
And thereby hangs my moral tale.

The trail was taken up next day,
By a lone dog that passed that way;

And then a wise bell-wether sheep
Pursued the trail o’er vale and steep,

And drew the flock behind him, too,
As good bell-wethers always do.

And from that day, o’er hill and glade.
Through those old woods a path was made.


III.

And many men wound in and out,
And dodged, and turned, and bent about,

And uttered words of righteous wrath,
Because ’twas such a crooked path;

But still they followed—do not laugh—
The first migrations of that calf,

And through this winding wood-way stalked
Because he wobbled when he walked.


IV.

This forest path became a lane,
that bent and turned and turned again;

This crooked lane became a road,
Where many a poor horse with his load

Toiled on beneath the burning sun,
And traveled some three miles in one.

And thus a century and a half
They trod the footsteps of that calf.


V.

The years passed on in swiftness fleet,
The road became a village street;

And this, before men were aware,
A city’s crowded thoroughfare.

And soon the central street was this
Of a renowned metropolis;

And men two centuries and a half,
Trod in the footsteps of that calf.


VI.

Each day a hundred thousand rout
Followed the zigzag calf about

And o’er his crooked journey went
The traffic of a continent.

A Hundred thousand men were led,
By one calf near three centuries dead.

They followed still his crooked way,
And lost one hundred years a day;

For thus such reverence is lent,
To well established precedent.
 

VII.

A moral lesson this might teach
Were I ordained and called to preach;

For men are prone to go it blind
Along the calf-paths of the mind,

And work away from sun to sun,
To do what other men have done.

They follow in the beaten track,
And out and in, and forth and back,

And still their devious course pursue,
To keep the path that others do.

They keep the path a sacred groove,
Along which all their lives they move.

But how the wise old wood gods laugh,
Who saw the first primeval calf.

Ah, many things this tale might teach—
But I am not ordained to preach.

This poem is in the public domain.

During this period Dad worked remotely for a client programming a Mac application. But a bit of research will show you that while WiFi – which we all take for granted now – had been invented some 10 years earlier Lucent hadn’t been able to get anyone to adopt it and it wasn’t actually available on a consumer computer until Apple launched the iBook at NY MacWorld in 1999.

Oh, but WiFi is so passé now in the 2020s, we have 5G digital wireless on the cellular network! Oh the dream of digital transmission on the cellular network was alive in 1997, but it was mostly just that a dream. Digital cell service was just starting to roll out and I think AT&T had 12 towers in big cities. I had to buy a special Nokia phone that was analog but also supported digital cellular connection. And a booster that had to be connected to the car to get enough power to push the signal. And the highest speed I ever got was 1200 baud despite the advertised 2400 maximum (pretty sure I’m remembering that right) – and that was parked directly under one of the few digital towers in the country

Here’s an excerpt I just found on my computer (which prompted this post) of a Bulletin Board System post I found on the “Bus Conversion News Board” from January 1999 which makes it clear that even nearly two years later it was still a challenge to get online from the road:

            Re: Wireless Internet, Cheap (relatively), nationwide

       [ Follow Ups ] [ Post Followup ] [ Bus Conversions News Board ]

Posted by Ed Carroll on January 17, 1999 at 18:22:25 PST:

In Reply to: Wireless Internet, Cheap (relatively), nationwide posted by Joe Solbrig on January 17, 1999 at 15:32:17 PST:

Thanks for the info Joe. Sounds like something a lot of us would be
interested in. A couple of questions though: 1. Do I understand you
correctly that if we connect up with this service then AT&T becomes our
regular ISP for unlimited connect time via wireless connection for only $60
per month regardless of our location? (Does that mean worldwide or just the
lower 48?) And 2. You mentioned a $100 "Pocketnet phone". Does that $60
monthly fee allow any voice communication from this "Pocketnet phone" to
friends on a regular land line phone?

: Hello again folks,

: I'm back, thanks to CDPD service from ATT wireless.

: There now is a way to connect to the internet wirelessly, nationwide and
for less than an arm and a leg.

: It takes a bit of looking around to actually get the information on this,

: "CDPD" and "Wireless" are data-broadcasting standards that uses the
cellular radio bands. ATT has used it for things like meter-reading and
stuff for a while. The problem is they traditionally charged per data
transmission and this made the service too expensive. However, recently,
they started a "nationwide unlimited" service which gives unlimited data at
a set price of $60/month. You have to buy a special connection equipment
also (request a "pocketnet phone" for $100 plus the connection kit for
another $100. The special CDPD moderm ).

: This is a business service, so it's best to say you're in business. Aren't we all ;). Also, you have to deal with the ATT office via phone and fax. But they're quite patient.
: They ran no credit check and started my service without my having paid for the equipment (which is good, given the delay my mail gets sometime).

: So all you full-timers, check it out!!

: Joe

          ---------------------------------------------------------

Follow Ups:

          ---------------------------------------------------------

Post a Followup

 Name
 Email Address
 Subject
 Comments:



          ---------------------------------------------------------
       [ Follow Ups ] [ Post Followup ] [ Bus Conversions News Board ]

Makes the kind of speed and service we can get now pretty amazing in contrast, eh?

So, how did I upload my source code, pull down shared source, get my email, and otherwise work remotely from a 40′ transit (city) bus as we drove around the country for 8 months?

Two techniques I figured out:

1. Most businesses have a fax machine and that’s often the only analog line in the office (the rest were part of the business PBX digital phone system). So I’d ask to borrow that line and since most didn’t need their fax machine online all the time, they’d often let me. Some hotels had a business communications area off the lobby that you could access without having a room if you acted confident and carried a laptop briefcase (and didn’t look too shaggy :)).

2. This gizmo:

What is this? How is it used? It’s an Acoustic Coupler and it’s used to connect to a POTS telephone, like this:

Yah, that’s an “antique” rotary phone and you can see the acoustic coupler held against the handset with the integrated velcro strap. This is literally a physical connection between the speaker on the acoustic coupler and the microphone on the phone handset and the microphone on the acoustic coupler and the speaker on the handset. So primitive!

And yet, it worked. Well, assuming I could find a payphone that hadn’t had beer poured on it and wasn’t trashed from being pounded against the side of the phone booth in frustration, etc.

My strategy was to find the largest bank of payphones – yah, they made them in rows, something like this:

but I’d look for at least 5 and 8 or 10 was a better bet. I’d take my laptop and a folding camp chair and start at one end and work my way down the bank of pay phones until I found one that could connect at 9600 baud or better (14.4 was rarely possible, and 28.8 was a challenge) and hold the connection long enough to upload/download etc.

You’d use a calling card in case you had to call long-distance to get an access point (phone number that would answer and connect you to the internet) and you had to program all that stuff into modem dial string so it would call the calling card billing number, put in your calling card number and pin, and then dial the closest access point number.

Fun times!

Anyway, it was either this or the Iridium satellite phone service which was $25K for the antenna/phone unit and $4/minute for pretty slow data (I don’t remember but it wasn’t 9600 baud even, IIRC). My clients weren’t paying me that kind of money

So, next time you open your smartphone and surf the web, or work remotely off of a wifi or you cellular hotspot – be thankful at the freedom we have now and laugh at how primitive working remotely was back in 1997.

I’m pretty sure there’s a way to do this with diskutil and whatnot instead of booting from the clone and turning it on there, but I didn’t feel up to digging through how to do that (it’s complicated by the need to add users manually before turning on encryption, is my vague memory of the steps I had to do last time I dug into this).

However, David – maker of SuperDuper! – was kind enough to mention that once you turn it on you can then boot back from your main drive and connect the backup clone and let it finish the encryption in the background (this is good because it said 10 days to complete – mine is a very slow USB external drive, apparently!).

The drive kept blinking for two days and I was trying to figure out how to tell how much progress it had made and if it was done or if spotlight was indexing the drive or what. Here are some commands that I found to help with that.

diskutil cs list

will produce a listing of Core Storage volumes and their “Encryption Status” and “Conversion Status” which will indicate complete or not.

Unfortunately, the bootable combined volume + “Data” doesn’t come up in that list (it’s something different, apparently).

So for that I was able to find what I wanted using:

diskutil apfs list

which outputs an entry like this for the in-progress file vault encryption volume (emphasis mine):

|   +-> Volume disk7s2 8C8FD74F-ACE0-4E1D-B67B-2F021A3C5DD7
|   |   ---------------------------------------------------
|   |   APFS Volume Disk (Role):   disk7s2 (Data)
|   |   Name:                      Data (Case-insensitive)
|   |   Mount Point:               /Volumes/Data
|   |   Capacity Consumed:         1459954380800 B (1.5 TB)
|   |   Sealed:                    No
|   |   Encryption Progress:    75.0% (Unlocked)

You can specify the container disk (which is disk 7 in this case) and see just the volumes in that container: diskutil apfs list disk7

The other funny thing is the sketches are on the back of ImageWriter II (dot-matrix) printout of Pascal code from 1988 for an iChing app I did for a client

The Dissolve(…) routine call in it is calling out to a 68k asm routine I wrote to do a fast dissolve from an offscreen grafport to the screen – likely the first assembly language code I shipped in a commercial product

Anyway, likely only amusing to me as a reminder of a time in the past, but they were some good times so posting it here to save the smile they brought to my face.

The magic incantation to make it visible on release builds of iOS is to enter:

applefeedback://inbox?make_visible=1

in the Safari address bar to make the Feedback Assistant app on iOS always visible (spotlight, app icon, Siri search, process switcher, etc).

or, if you want to leave it hidden and auto-hiding but want to launch Feedback Assistant to file a new feedback enter the following in Safari:

applefeedback://new

Use just:

applefeedback://

alone to switch to Feedback Assistant without changing visibility or starting a new feedback (because Feedback Assistant is hidden from the process switcher, you need this if you are in Feedback Assistant and want/need to switch to another app but want to come back to Feedback Assistant to finish the feedback you’re editing, for example).

This is somewhat documented by Apple here: https://developer.apple.com/bug-reporting/

Note 1: You’ll need an Apple ID to file feedback reports.

Note 2: You can always go to feedbackassistant.apple.com to access your feedback assistant account without the Feedback Assistant iOS app, but you can’t include sysdiagnose files and it doesn’t have quite the same ease of use as the app.

It wasn’t trivial, because I do enjoy sweets, but wasn’t that difficult.

Here’s how to successfully eat half an ice cream sandwich (or anything else your body craves), from easy to difficult:

1) [Easy] Take an ice cream sandwich out of the package and cut it in half. Hand half to your wife/kid/friend/neighbor/etc and then pick up your half and enjoy it.

2) [Not difficult] Store your ice cream sandwiches in a freezer in the basement/garage.  Go get one and bring it to the kitchen.  Open one end of the package and slide the ice cream sandwich out of the package and put it on a plate. Cut it in half and put half back into the package.  Take that half an ice cream sandwich in a package back to the freezer in the basement/garage and put it away.  Only then do you pick up the half you’re going to eat and enjoy it.

3) [Medium] Same as above but don’t bother taking the half you’re saving back to the basement until after you eat the half you’re going to eat.

4 [Difficult] Open the package and start eating the ice cream sandwich. When you’ve eaten half of it, stop eating it, close up the package, and put it back into the freezer.

So, I wouldn’t be impressed that you ate only half an ice cream sandwich unless you did number 4 on this list

* I did (2) on the list, fwtw

Lots of ways you might share that image which don’t copy that data, but copying the file to another Mac through an external hard drive or file sharing, at minimum, seems to transfer this metadata across.  Sending as an email attachment, or as a Messages message attachment doesn’t seem to transfer this metadata, on the other hand.  It would be nice to have a preference to prevent adding this to the file metadata (FB10992657).

Since I’m not 100% clear when this metadata is getting transferred and when it isn’t, I wrote a small macOS Shortcut to strip it.  First Shortcut I’ve put together and the lack of logging for shell commands or of any kind makes Shortcuts significantly less usable than Automator, for example (where I had to go to figure out what was going wrong).  Also the initial text for an Run Script node in a Quick Action are not as helpful as the Automator environment (FB10993044).

Anyway, here’s the result:

I’d attach the actual shortcut file but Apple chooses to make them tightly linked to the creator and doesn’t allow sharing the source code for the shortcut without attaching my AppleID certificate to the file (WTH?!) (FB10993444).

Got tired of not being able to find something I knew I had open so I wrote an AppleScript to search the text content of all Safari tabs in all Safari windows that aren’t minimized (even across Spaces).  Does not currently search tabs in windows that are minimized to the dock.

I’ve now found this useful enough that I thought I’d share it.  Disclaimer: I’m not an AppleScript programmer and don’t really like AppleScript (probably because I don’t know the language and tricks!).  There are likely things that could be done better or smarter.  Contributions welcome.

That said, this seems to generally work and, thanks to me learning about using references, fast enough that I wasn’t sure it was actually working at first

Source is on GitHub here.

Instructions for installation are in the Read Me there.

//
//  ContentView.swift
//  trashme
//

import SwiftUI

struct ContentView : View {
  var cellSize: CGFloat = 50
  var numRows: Int = 30
  var numCols: Int = 10
  var body : some View {

    VStack(alignment: .leading, spacing: 20) {
      ScrollViewReader() { proxy in
        ScrollView([.horizontal,.vertical], showsIndicators: true) {
          HStack( alignment: .top, spacing: 0) {
            VStack ( alignment: .leading, spacing: 0) {
              ForEach(0 ..< numRows, id: \.self) { row in
                Text("row " + row.description)
                  .frame( height: self.cellSize )
              }
            }
            ForEach(0 ..< self.numCols, id: \.self) { col in
              VStack( spacing: 0) {
                ForEach(0 ..< self.numRows, id: \.self) { row in
                  Text("\(row), \(col)")
                    .frame( width: self.cellSize, 
                            height: self.cellSize, 
                            alignment: .center)
                    .border(.blue)
                }
              }
            }
          }
          .id("root")  // since our ForEach views don't really 
          		// have true ids (\.self ~)  
          		// we make one for the outer HStack
          		// in the ScrollView
        }
        .onAppear {
          proxy.scrollTo("root", anchor: .topLeading)
        }
      }
    }
  }
}

struct ContentView_Previews: PreviewProvider {
    static var previews: some View {
        ContentView()
    }
}

And here’s what it looks like in an iOS SwiftUI preview pane:

from:username
to:username       // to username
@username         // mentioning @username
keyword           // can be a #hashtag
-keyword
keyword OR keyword
min_retweets: x
min_faves: x
min_replies: x
filter:links
filter:images     // any image
filter:twimg      // twitter pic.twitter.com image
filter:video
filter:native_video
filter:periscope
filter:vine
filter:media      // image or video
filter:retweets
filter:safe
filter:verified   // verified users
-filter:____      // use with any of the above filter nouns to exclude
since:YYYY-MM-DD
until:YYYY-MM-DD
near:X within:10mi // (where X is a city name)
list:user/listname
url:apple          // has url containing "apple" anywhere in it
lang:languagecode  // en, da, cs, de, es, fr, it, ja, ko, nl, no, pt, ro, uk,…

Thought I’d give MarsEdit a try.  But the Mac App Store glitched

Worked after I quit and restarted the App Store.

So one thing I’m not seeing is how to have it post to Twitter when I post to my blog.  Maybe that’ll happen inside of WordPress?

It DOES!   Ok, that’s pretty cool.  This might work after all.

Hmm.  So maybe I’ve been using Pages & Numbers too much lately, and the WordPress site, but I find myself looking to the right side “Options” pane for formatting commands.  I often want to format some content as code (go figure!) and having that very handy is nice.

OH MY WOWZA!  You can add your own formatting macros to the Format menu with key command shortcuts.  Ok.  That is slick.

let foo = 12

Little bit of trouble getting out of code mode.. but not terrible.

— 

Wow. The feature where MarsEdit downloads my template from WordPress so that it can preview posts is very cool.

––

There are a few things that are tripping me up, but since I want to put more of the content I usually put on twitter onto a platform that I can keep if I bail on twitter, this seems like a good way to do it.  (I think about micro.blog also, but I just don’t ‘get’ it.  ¯\_(ツ)_/¯ ).

I do wish MarsEdit would save out static blog posts and use FTP to upload them to a statically hosted blog/site.  But a WPXML roc adaptor to Jekyll or something has been sketched out by two different friends so maybe that’ll materialize somehow.

Just ran across an app that failed to do this and it was quite annoying. I needed to open the folder containing the document in the Finder so I could duplicate the file and selecting the folder name in this menu from the document title is the easiest way to do that. Bug report filed with the app maker, but please save me the time of filing one against your app by getting this right from the beginning

* Note that the Finder implements this for Finder windows, so I’d follow their lead and interpret “document” broadly as ‘anything with a location in the file system’.

I opined that this was probably a form of shopping therapy and given our mother’s chronic health issues in recent years, probably understandable. My sister then mentioned that she just goes to the dollar store to satisfy this for very little money but also that one of her kids seemed have a bit of a shopping therapy urge also.

I mentioned that I’d been pretty successful some years back when I hacked what felt like my own occasional shopping therapy urge (“Feeling down? Buy something!” – such an American consumer ingrained message for so many years).

What I decided to do was every time I felt that urge and couldn’t just easily talk myself out of it I’d find one of the organizations I like to support, a creative person trying to make their first album/book /comic/etc on Kickstarter, a request for help that a friend or Twitter acquaintance had shared, or similar, and make a donation instead.

I found that I got a better “feel good” feeling this way, someone benefits, and I don’t end up with more STUFF cluttering up my life. Triple Win!

I thought I’d share this hack here in case anyone else might like the idea and want to adopt it.