nostalgix.org

Jekyll rebuild for DSGVO

madhatter@nostalgix.org (Arvid Warnecke) — Tue, 15 May 2018 00:00:00 +0200

Jekyll rebuild for DSGVO

The DSGVO (GDPR) is coming. The internet will stop to exist. At least the internet we knew here in Germany. All other european countries were able to read the paper in time to take care of those changes. Germany didn’t and now we have to make sure that no lawyer will sue us for storing cookies on your computer, to track your clicks on our sites and try to figure out what might be of most interest.

To make sure that I don’t have to spent the money I save for my kids’ education in their future, I removed everything that stores anything. This website/domain does not store any data at all from you. Nor from anybody else. I used to use etracker to know every week I have 2 visitors and 4 pages views. For this I had to send cookies to make sure I might remember you if you would ever return one day. As I don’t write something new here, I don’t see the point in doing so any longer anyway.

The webserver running this site had an access log where I stored your ip addresses. I don’t do this anymore. What for?

But to make sure that my static sites are not changed by a man-in-the-middle attack I renew the ssl certificate every now and then. If you would send some information/data to my websites it would be encrypted, too. But you can’t, because I don’t care. And I don’t save any of it.

So, let’s make this clear one last time: this is a private and personal website that does not make any profit at all out of it and which does not store any data of it’s visitors and readers.

Das obligatorische Neujahrsposting

madhatter@nostalgix.org (Arvid Warnecke) — Fri, 17 Feb 2017 00:00:00 +0100

Das obligatorische Neujahrsposting

Na, das hat doch hervorragend geklappt mit den vielen Vorsätzen, die ich mir letztes Jahr so vorgenommen habe. Allein schon, wenn ich mir anschaue, wieviel ich hier zu posten geschafft habe. Dennoch bin ich mir nicht zu fein, mal zu prüfen, was noch alles nicht so gut geklappt hat, wie ich es mir vorgenommen habe:

mehr Bücher lesen
Hier hat sich leider nicht so viel getan im letzten Jahr. Ich hatte mir wieder 10 Bücher zu lesen vorgenommen und habe ungefähr 3 oder 4 davon geschafft. Je nachdem ob man auch angelesene Bücher zulässt oder nicht. Für 2017 habe ich wieder 10 Bücher als Ziel. Mal sehen, ob ich die in diesem Jahr ansatzweise erreiche.
Comics regelmäßiger lesen
Das habe ich komplett gelassen. Nachdem ich ziemlich schnell bemerkt habe, dass ich da so gar nicht vorwärts komme, habe ich alle meine Comic-Abos gekündigt. So zahlt es wenigstens auf das Ziel weiter unten ein, weniger Geld auszugeben.
wieder mehr bloggen und auf deutsch bloggen Gerade auf deutsch habe ich genau einen Post geschafft. Und auch sonst war da nicht viel dabei. Allerdings muss ich zu meiner Verteidigung sagen, dass ich ja doch viel technischen Kram blogge, wenn denn überhaupt. Und genau dafür war im letzten Jahr dann fast keine Zeit, sich mit viel zu befassen. Das ist natürlich auch nicht gut, aber viel Zeit habe ich schon damit vertan, dass ich mir den Ellenbogen gebrochen hatte und somit nicht viel am Rechner gesessen habe in der Zeit.
weniger Geld für Scheiß ausgeben
Das hat generell gut geklappt. Kaffee koche ich jetzt immer morgens zuhause und nehme ihn mit. Der schmeckt eh viel besser und ist um einiges günstiger. Und auch sonst war ich im letzten Jahr schon eher sparsam.
mehr Geduld mit den Kindern haben
Klappt. Oft jedenfalls. Aber leider so oft, wie ich es mir wünschen würde.
kein Facebook mehr nutzen
Noch bin ich angemeldet. Aber meine Nutzung geht aktuell schon eher gegen Null. Insofern: halb das Ziel erreicht oder so.

Neues nehme ich mir nicht vor. Ich denke ich habe mit meinen alten Zielen noch genug zu tun.

Howto fix RVM path issues

madhatter@nostalgix.org (Arvid Warnecke) — Tue, 29 Mar 2016 00:00:00 +0200

Howto fix RVM path issues

I had some trouble with rvm lately on the Odroid setup which took me some time to figure it out. I had no idea what might be wrong because I have been using rvm on all my setups for the last 5 years without any issues. And as I have all my dotfiles stored in a GitHub repository I was quite sure, that I had a working zshrc.

But when I logged in I noticed that I was always falling back to system’s ruby even though I had set the rvm ruby version as default. And when I tried to re-set the ruby as default again I got the following warning from rvm:

Warning! PATH is not properly set up, ‘$HOME/.rvm/gems/ruby-2.1.5/bin’ is not at first place, usually this is caused by shell initialization files - check them for ‘PATH=…’ entries, it might also help to re-add RVM to your dotfiles: ‘rvm get stable –auto-dotfiles’, to fix temporarily in this shell session run: ‘rvm use ruby-2.1.5’.

It took quite some while and lots of useless googling to find the reason: I made the mistake to install rvm for the root user, too. I never did that before (and I probably never will ever again) and did not know that rvm installs /etc/profile.d/rvm.sh if it has root rights.
Because this zlogin was evaluated first and I later moved parts of the PATH in my local ~/.zshrc becore finally calling the rvm scripts again, rvm got messed up and did not work correctly. After deleting /etc/profile.d/rvm.sh this was fixed.

Howto Setup a Tor Access Point (on an Odroid C2)

madhatter@nostalgix.org (Arvid Warnecke) — Sat, 19 Mar 2016 00:00:00 +0100

Howto Setup a Tor Access Point (on an Odroid C2)

I got an Odroid C2 for testing from a work-mate. Because he told me that his initial tests with Ubuntu’s alpha version went not too well, I gave ArchLinux a try. As I use Arch as my daily driver on non-arm architectures I appreciated that option.
Installation went smooth, but I had no fancy idea to take the Odroid for a ride. So I set up a Tor Access Point for the living room. I already had that in mind for my dated Raspberry Pi, but never was in the mood.
I came accross a quick howto in german a few months ago and started from there. Because I had some small pitfalls I thought it might be a good idea to write it down.

First, install the required software

# pacman -Sy dnsmasq hostapd tor iptables

Now configure the dnsmasq service in /etc/dnsmasq.conf

interface=wlan0
dhcp-range=10.10.0.25,10.10.0.255,6h

Next, configure the access point via /etc/hostapd/hostapd.conf

interface=wlan0
driver=nl80211
ssid=Doors of Passion
country_code=DE
ieee80211d=1
hw_mode=g
ieee80211n=1
channel=9
macaddr_acl=0

Btw, you probably want to make sure, that the wifi adapter you are using is able to enter master mode. (iwconfig wlan0 mode master)

We now need a unit file to assign a static ip to the wifi adapter. Create
/etc/systemd/system/openwifi.service

[Unit]
Description=Assign static ip to wlan0
Wants=network.target
Before=network.target
BindsTo=sys-subsystem-net-devices-wlan0.device
After=sys-subsystem-net-devices-wlan0.device

[Service]
Type=oneshot
RemainAfterExit=yes
ExecStart=/usr/bin/ip link set dev wlan0 up
ExecStart=/usr/bin/ip addr add 10.10.0.1/24 dev wlan0
ExecStop=/usr/bin/ip addr del 10.10.0.1/24 dev wlan0
ExecStop=/usr/bin/ip link set dev wlan0 down

[Install]
WantedBy=multi-user.target

And the tor configuration itself in /etc/tor/torrc

AutomapHostsSuffixes .onion,.exit
AutomapHostsOnResolve 1
TransPort 9040
TransListenAddress 10.10.0.1
DNSPort 53
DNSListenAddress 10.10.0.1
User tor
Log notice syslog
DataDirectory /var/lib/tor
BridgeRelay 0
PublishServerDescriptor 0

To bind Tor to privileged ports the service must be started as root. We can modify the tor service by adding /etc/systemd/system/tor.service.d/start-as-root.conf with following content:

[Service]
User=root

And to make all this work together we need to do a few more changes to the system. We have to enable ip forwarding via sysctl. Create a file like /etc/sysctl.d/99-sysctl.conf with this line

net.ipv4.ip_forward=1

And load this configuration by

sysctl --system

At last we need some iptables rules to redirect tcp and udp traffic and to route it from the wlan0 device to eth0:

# iptables -t nat -A PREROUTING -i wlan0 -p udp --dport 53 -j REDIRECT --to-ports 53
# iptables -t nat -A PREROUTING -i wlan0 -p tcp --syn -j REDIRECT --to-ports 9040
# iptables --table nat --append POSTROUTING --out-interface eth0 -j MASQUERADE
# iptables --append FORWARD --in-interface wlan0 -j ACCEPT
# iptables-save > /etc/iptables/iptables.rules

Let’s give it a try and start all services:

# systemctl start iptables
# systemctl start openwifi
# systemctl start hostapd
# systemctl start dnsmasq
# systemctl start tor

And when everything is working you can enable all the services to get them started after reboot:

# systemctl enable dnsmasq hostapd tor openwifi iptables

If it is not you have to take a further look into it.

On the Odroid C2 it performs very well and if it was any different they did something very wrong. So I am looking for some other task for the board.

Frohes Neues Jahr

madhatter@nostalgix.org (Arvid Warnecke) — Wed, 13 Jan 2016 00:00:00 +0100

Frohes Neues Jahr

Frohes Neues! Das darf, nein muss man ja eigentlich den ganzen Januar über wünschen. Und insofern ist es sicherlich auch nicht verwerflich, wenn man mit seinen Vorsätzen für das neue Jahr erst Mitte Januar fertig wird.
Normalerweise halte ich ja nicht viel von solchen Dingen; überhaupt kann ich dem ganzen Silvester/Neujahr Gefeier nicht wirklich etwas abgewinnen. Aber dieses Jahr habe ich komischerweise unheimlich viele Punkte auf der Liste mit Vorsätzen. Darum dachte ich mir, ich schreibe sie einfach mal auf. Denn so kann ich dann zum Ende des Jahres prüfen, wieviele Punkte ich erfolgreich umgesetzt habe:

mehr Bücher lesen
So wenig habe ich wahrscheinlich gar nicht gelesen letztes Jahr. Aber meine Goodreads Challenge von 10 Büchern in 2015 habe ich, mit nur einem dort als gelesen vermerkten Buch, ganz knapp nicht geschafft. Zu meiner Verteidigung muss ich sagen, dass ich das ein oder andere Buch dort nicht eingetragen habe. Der Kindle tut das nicht automatisch und ohne Lokalisierung auf amazon.com erst recht nicht. Aber ich habe eben auch etliche Bücher lediglich angefangen und nie zu Ende gelesen, da ich schnell das Interesse verliere.
Comics regelmäßiger lesen
Bücher lesen ist gut und schön, aber wenn man regelmäßig Comics per Post bekommt, weil man sie bei seinem Comicshop abonniert hat, dann wäre es sicher sinnvoll sie auch zu lesen. Leider mache ich das viel zu selten und teilweise stapeln sich dann Serien mehrere Monate lang, bis ich mich endlich aufraffe und sie mit zur Arbeit schleppe, um sie in der Bahn zu lesen. Einige Serien abonniere ich dadurch viel zu lange, weil ich viel zu spät bemerke, dass mir die Story gar nicht mehr gefällt.
wieder mehr bloggen und auf deutsch bloggen
Der Anfang wäre hiermit getan. Jetzt muss ich es nur noch schaffen, dass es nicht wieder bei einem kümmerlichen Post pro Jahr - wenn überhaupt - bleibt, sondern es vielmehr wieder zur Regelmäßigkeit wird.
Und warum plötzlich auf deutsch? Weil ich die Befürchtung habe, dass einfach alltägliches nicht den Weg in einen Blogpost findet, weil es mir zu umständlich ist, meine Gedanken in englisch zu formulieren. Bei technischen Themen ist das kein großes Problem, aber so bei Alltagsthemen finde ich es schon anstrengender.
weniger Geld für Scheiß ausgeben
Das ist ja eigentlich immer eine gute Idee. Das Problem ist ja aber, dass man nicht immer merkt, wieviel Geld man im Monat für unnötiges ausgibt.
Allein für Kaffee unterwegs gebe ich im Monat ungefähr 30 - 40€ aus. Und dabei gehe ich nicht mal zu den großen teuren Kaffeeketten wie Starbucks oder Balzac. Hinzu kommt dann aber oft noch anderer Kram wie Gebäck oder Zeitschriften. So genau weiss ich das aber eben leider immer gar nicht mehr hinterher. Und bislang habe ich auch noch keine App gefunden, mit der ich derlei Kram komfortabel tracken könnte.
Für viele mögen solche Beträge kein Grund sein, sich damit gezielt auseinander zu setzen, aber mich ärgern sie. Ich erfreue mich einfach nicht nachhaltig an diesen Dingen und würde das Geld lieber sinnvoller investieren.
mehr Geduld mit den Kindern haben
Ihr kennt das.
kein Facebook mehr nutzen
Da bin ich noch unschlüssig, ob ich das wirklich für eine gute Idee halte. Generell geht mir der ganze Facebook-Hass auf den Nerv. Und eigentlich möchte ich keiner Community angehören, wo solche Dinge passieren, auch wenn meine eigene Timeline gar nicht (oder nur minimal) davon betroffen ist. Auf der anderen Seite frage ich mich, ob man dem Pöbel das Feld kampflos überlassen sollte.¹ Und man wird es wohl auch diesmal nicht schaffen, seine Freunde und Bekannte in ein anderes/besseres Netzwerk zu migrieren.

Obwohl man den Pöbel ja eh lieber ignorieren sollte, als sich auf deren Niveau zu begeben. ↩

Seriously, when shall I ...?

madhatter@nostalgix.org (Arvid Warnecke) — Fri, 10 Apr 2015 00:00:00 +0200

Seriously, when shall I …?

Welcome back. I wasn’t sure if you were coming. It’s been quite a long time.
But seriously, I really don’t know how all people manage to keep up with all these things in life. When I read other’s people blogs I am always stunned what they seem to be able to handle in there spare time.
Maybe I am wrong, but I reckon that most have full-time jobs about 8 hours/day. And they have to get there and have to take a break at lunch. So alreay 9-10 hours per day are gone. Further I guess that they eat in the morning and sometime before they go to sleep. Let’s calculate an hour for these meals. 11 hours gone.
Doctors will tell you that you should sleep at least 7 hours per day …err night, better 8. What is left then are about 5 hours per day not assuming that you have to work overtime very much or do any sports after work. But maybe you have a friend or a family you would like to spent time with? No? Fine. But all the others only have an hour or two per day left where they can

code on pet projects
blog about something interesting
listen to music
read books
scroll on tumblr

If you are like me, you want to do all these things. And this is where I get into trouble. I only have time for one of these things, because switching activities within the week leads to distraction and no real result. Aside from listening to music maybe.

This post was written on 2015/10/04 and never finished. I guess because I ran out of free time. I only found it now under drafts, because I was planning to revamp my blog and start blogging more often than in the last few years. That is why it will be released with all the other changes made.

vim - 12 years and counting

madhatter@nostalgix.org (Arvid Warnecke) — Fri, 14 Jun 2013 00:00:00 +0200

vim - 12 years and counting

Everybody is telling you right now, how they configured vim and which plugins they are using. At the moment you can get the idea, that vim becomes one of the hippest editors and people want to show off by hanging their .vimrc out. Maybe the girls like that.
So, here is the repository with my vim configuration, knock yourself out… Just kidding. I don’t know if I have build a .vimrc which makes a grown man cry and it also is pretty well documented, but I think this is not what it is about.

I came to vim about 12 years ago in 2001 and Debian Potato was the distribution of my choice back then. If I remember correctly it was vim version 5.6 or 5.7. Code folding wasn’t available and came with vim 6.0. Can you imagine that?
I remember trying emacs before, but I had terrible issues with the keybindings (still have) and used gEdit or something like that since then. Until a room-mate told me about vim and got me hooked.¹ But things were different in these days: the internet was made from wood and there weren’t too many websites available that told you how to configure vim best. You mostly had to figure out that by yourself and it wasn’t about adding lots of plugins, but change one or two settings here and there. You were busy enough to configure vim’s core functionalities but adding lots of plugins. And let’s be realistic, there hasn’t been any git back then that you might wanted to call from your editor.²

Of course, it took me some time to get used to vim’s commands, but it got much better when I stopped using different editors from time to time and instead used vim for everything: writing emails, taking notes, writing news posts, coding, … you name it. And that is what bothers me about lots of articles I read in the last few months: telling people how hard it is to get used to vim! If you use it regularly you will get used to it very fast. The only thing you should keep in mind is that you might have to take some extra time to figure out how you can do things faster in vim when you get the feeling that you are using to many keystrokes to do something. There is a better way.
I think it’s the same problem when people switch from Windows or OS X to Linux or Unix or whatever: you have to use it for everything and don’t ignore when something does not seem to work as well as it should.
And it might not be the best idea to copy whole vim configurations from others for a starting point. Just read them and look up those things that sound interesting in vim’s help and configure them to your needs then.

I think the best tip I can give you from my perspective is to make yourself comfortable. Vim is an editor that can do lots for you, but you have to know what you want first.
And maybe vim isn’t the right editor for you if you don’t feel comfortable with it. Don’t try to use it just because so many people are writing about it at the moment. There are lots of other editors available that might suit you better. I have a colleague who uses emacs. And he likes using it. Nobody forces him to do so. I can’t understand that, but people are different.

And at last I have to admit, that I don’t use vim for everyting anymore today. I spent months by trying to code Java with it, but I never made vim such easy and helpful as a full-featured Java-IDE like Netbeans.³ And it makes pair-programming so much easier with workmates.

You should not be too dogmatic about these things.

He also got me hooked on mutt and slrn. And it’s a shame that he does not use anything of these anymore… ↩
I know people you use IDEs like eclipse or Netbeans for Java, but execute git/mercurial and maven from the command line next to it, because they don’t see the benefits in using a plugin for their IDE that might not be able to give them all the functionality the original tools do. ↩
If anybody has the solution for that, I would be happy to hear about it though. ↩

See It In The Dunst

madhatter@nostalgix.org (Arvid Warnecke) — Wed, 20 Feb 2013 00:00:00 +0100

See It In The Dunst

As I mentioned before I am happy using dwm as my windowmanager for every day. But dwm does not come with a systray¹ by default and I felt not like patching one in, because I was wondering why I should really have one. The only things sitting in the systray back in my awesome days were the nm-applet, dropbox and sometimes the pidgin notificiation icon.

Only I few days later I noticed that that pidgin icon was the only thing I really was missing because I ignored all jabber messages from my co-workers the whole day.
But would it be a good idea to add another patch just for one icon?

No. So I have been looking for alternatives and ended up with installing dunst, a dmenu-ish notification daemon.² The only problem was, that pidgin did not seem to like dunst at all and I haven’t been able to get any notifications from it. This was the last reason necessary to get rid of pidgin at all and move to some other client instead. For my private instant messaging pleasures I have been using bitlbee for a while now to combine it with my IRC client³.
I couldn’t use that environment though, because it’s located on a remote host and it would be a lot more difficult to get local notifications from there.

As I already know irssi I installed weechat on my workstation to check it out for this project. But it seemed that there were no scripts available that worked with libnotify⁴, so I wrote one myself. In Ruby, of course.

def weechat_init Weechat.register("notify.rb", "madhatter", "0.1", "GPL3", "Notifications via libnotify", "", "") Weechat.hook_print("", "", "", 1, "notify_msg", "") return Weechat::WEECHAT_RC_OK end def notify_msg(data, buffer, date, tags, visible, highlight, nick, message) # Get the channel's metadata. data = {} %w[ away type channel server ].each do |meta| data[meta.to_sym] = Weechat.buffer_get_string(buffer, "localvar_#{meta}"); end data[:away] = !data[:away].empty? # Remember the own nick name to ignore it in private messages my_nick = Weechat.info_get('irc_nick', data[:server]) # If it is a highlight, a private message and not myself notify me. return Weechat::WEECHAT_RC_OK if highlight.to_i.zero? && data[:type] != 'private' or nick == my_nick `notify-send -i /usr/share/pixmaps/weechat.xpm "#{nick}" "#{message}"` return Weechat::WEECHAT_RC_OK endThis Gist hosted on GitHub.

The script does nothing special, but notifies me about mentions of my nickname in a channel or a query. And it ignores everything from my nick, because first I got even noticed about everything I wrote, too.

My dunstrc hasn’t been too much changed compared to the default configuration file that comes with dunst, but I went with the dwm way for dunst, too, and build my own package with a slightly modified config.h.⁵
Now I am aware of what is happening in the office chat all the time and the nice thing is, that the dunst notifications follow the focussed monitor. That, and that the notification window is larger than the icon I used to have with pidgin makes it much more difficult to miss now.
I noticed that dropbox makes use of the dunst daemon by default when it is running, so I somehow replaced the systray icon for dropbox, too.

The only thing I haven’t been able to accomplish yet is to make any kind of application icons appear in dunst. I don’t really need this feature, but I have been wondering what it might look like. But maybe it’s just that libnotify does support that and dunst does not.

http://en.wikipedia.org/wiki/Taskbar ↩
To be honest I don’t see what is so dmenu-ish about it. If you do, please let me know. ↩
irssi: http://www.irssi.org ↩
https://wiki.archlinux.org/index.php/Libnotify ↩
I haven’t uploaded it anywhere, because I did not patch dunst anyhow, but only changed some colors and the font. Maybe I will later, when I am aware of what I might tweak in the header file, to make much better use of dunst. ↩

Jekyll with Pygments on Arch Linux

madhatter@nostalgix.org (Arvid Warnecke) — Fri, 08 Feb 2013 00:00:00 +0100

Jekyll with Pygments on Arch Linux

As I am trying to blog more I came accross Pygments again and wanted to give it a try. But I had to notice that Jekyll failed to highlight anything, because of Arch Linux using Python3 as default (/usr/bin/python is linked to /usr/bin/python3).
Lots of googling later I found out, that it isn’t exactly Jekyll’s fault, but a problem with the pygments.rb gem. There already is a pull request for this issue which has been opened 3 months ago.
It really is easy to fix though, but I guess it might cause issues on other Linux distributions which have no Python 3 at all and therefore no python2 command available.
For now I went to the gem’s installation path and changed mentos.py manually:

#!/usr/bin/env python2

Quick and very dirty. A better solution probably would be to fork the gem’s repository, patch it to your needs and install that one.

DWM is Awesome (for now)

madhatter@nostalgix.org (Arvid Warnecke) — Wed, 06 Feb 2013 00:00:00 +0100

DWM is Awesome (for now)

Since I moved from OS X back to Linux I have been using the tiling window manager awesome and most of the time it has been what it’s name let you think it would be.
A few days ago I moved to dwm. Why? And how can it make such a big difference when awesome is based on a dwm fork?

First of all I never liked the whole lua configuration. I can’t tell exactly why, but to me lua is not very appealing. It starts with the “–” for comments and leads to the odd habit of counting from 1 instead of 0. But the main reason is that awesome is not awesome when it comes to error handling and logging. If you have written something wrong in your rc.lua (which is the main configuration file) awesome will not tell you what exactly is wrong but start with some kind of default configuration. But there may be times when not even Mod4+Enter will work to start a terminal.
The last reason came a couple of weeks ago when I have been forced to update my configuration to work with the update for version 3.5. The changes have been not too many, but because I had lots of widgets configured I had to change a lot to match the new syntax. It took so long to have everything updated that I downgraded to 3.4.x for some time, because I haven’t been in the mood to do the work.
And when I was done… what did I get? Right. Nothing. There were no benefits noticeable at all. Others were even noticing performance issues and crashes.

So it was about time for me to take a look for a new windowmanager. A tiling one of course.
I have been using Xmonad a couple of years ago when I wanted to use FreeBSD on an old IBM Thinkpad T60. I did like it back then, but I had some problems with fullscreen mode for watching videos. And a huge problem is the configuration file written in Haskell. I remember hacking a bit Haskell back then, but even though reading and changing the config was a PITA.

I remembered some other windowmanagers I read or heard about or even gave a try yet some time ago: monsterwm, herbstluftwm, ratpoison¹, wmfs2² and dwm³.
I went with dwm. I can’t exactly say why, but maybe it was because I kinda liked the idea that you have to compile dwm when you want to change something. Arch makes it very easy to compile your own versions with ABS. I just had to tweak the PKGBUILD file a bit, configure dwm to my personal needs in the config.h header file and call makepkg.⁴
That’s the basic version. If you want to have things that dwm does not offer to you by default your can patch the source code yourself or try to find somebody who already did that. The only problem might be that you can’t apply lots of patches you find here and there above each other. But as you already have the lines of code that have to be added or changed you can patch it manually and then create your very own patch file(s). Then you are able to patch dwm everytime from vanilla source again when you change the header file.
Sounds difficult? You’ll get used to it.

I don’t know if I am already done with my dwm configuration yet, but at the moment I don’t miss something special. And when I thought that wmfs2 might be a bit too spartan, I now think that you don’t need much more than that what dwm or wmfs2 have to offer. Most of the time I only used two tiling modes in awesome anyway.
I haven’t found something like widgets, but at the moment I am very happy with conky and I already had something ready from my days with wmfs2…

For now I am very happy with dwm. But I am pretty sure it won’t be the last windowmanager I install. What is it about this Unity everbody is talking about lately…?

Update
I almost forgot to list the patches I have applied at the moment:

statuscolors
pertag
scratchpad-stay
cflags
single_window_no_border
centered-floating
save_floats

More information on those patches can be found on dwm’s homepage.

I don’t know for sure, but I think I even had ratpoison running for a few days back in 2002. But I think I haven’t been ready for a tiling windowmanager back then. ↩
I used wmfs2 a few months ago for a week or so. But compared to awesome (at that time) I missed a few things and found it too minimalistic. ↩
The first time I came across dwm must have been when I was looking for dwb some time ago. A nice browser you should take a look at if you haven’t already. ↩
For more information on how to install dwm via the Arch Build System take a sharp look at the wiki. ↩

More bad news from systemd

madhatter@nostalgix.org (Arvid Warnecke) — Mon, 21 Jan 2013 00:00:00 +0100

More bad news from systemd

The last weeks have been horrible. Every now and then (but within 2 days for sure) my Macbook did not recover from suspending very well. I wasn’t able to figure out what went wrong exactly, but something seemed to fail when setting the harddisk asleep. The terrible result was that the root partition was mounted read-only when the book woke up again. Because I only have /boot and /home on own partitions and /var is part of the / partition I wasn’t able to remount it read-write again. So I had to reboot often and do manual filesystem checks every two days at last.

A few days ago I sat down and tried to figure out when those problems might have happened for the first time. And then I realized that it probably began with the upgrade to systemd. I had issues back then with the supsend because acpid and systemd wanted to take care of it. The result was that acpid called pm-suspend exactly when systemd woke the Macbook up again and it never woke up at all. So I edited /etc/systemd/login.d to set HandleLidSwitch to ignore and made the handler.sh in /etc/acpi handle the supsend by pm-suspend again.

Why I hate systemd even though migration was easy

madhatter@nostalgix.org (Arvid Warnecke) — Wed, 31 Oct 2012 00:00:00 +0100

Why I hate systemd even though migration was easy

I finally moved my Arch to Systemd because in the last few days I had the feeling that if I won’t do it now I will run into serious trouble.

I have to admit, that I have never been a big fan of the movement to Systemd. Initscripts was simple. It worked well for me and never had complaints about it. So why move to something else? And I had the feeling that Systemd is just the new sow that gets driven through the village (german saying).

Okay, but I don’t make the decisions and there is no opt-out from Systemd. So I sat down last night and went through the three steps of migration as described in the wiki. I rebooted every now and then and made sure that I did not break anything. And today my MacBook Pro is running a initscripts free Arch Linux. The only problem was the pommed daemon, which had no real systemd unit and was relying on Systemd’s initscript compatibility. But I found a light version in the AUR that also had the missing units. Nice. Took only about an hour and I am done.

But today I noticed that pm-suspend triggered by acpid did not work too well anymore. Or better: suspending worked as usual, but then waking up again it seemed that now Systemd’s own powermanagement tools kicked in and send the book back to sleep again. I now had to wake my laptop twice (now more akin to me, but not what I had in mind). I disabled the acpid unit again with no effect, because I missed the acpid.socket, which I never explicitly enabled. Until I noticed this I just removed the lines in the handler.sh file.

And now what? All done? I am afraid not. There probably will turn up other issues I would not have when I would be using initscripts again. And that is what bothers me most. I have to take care of problems I did not get, because I wanted to change my system. I don’t like spending my free time on all that.

The Apple Days are over

madhatter@nostalgix.org (Arvid Warnecke) — Thu, 25 Oct 2012 00:00:00 +0200

The Apple Days are over

I originally began this post on March, 19th and never finished it.

The reason was that I got lost in a long story about when I started with Linux, moved to Unix, ended up with Apple and then finally found my way back to Linux… Today I finished that post to get it out of my mind. It still is that long story though. But who knows if there isn’t somebody who is interested in it.

tl;dr

How did I end up with Mac OS X anyways?

I came a long way: Microsoft MS-DOS, Microsoft Windows, SuSE Linux, Debian Linux, FreeBSD and finally Mac OS X. Not included the other Linux distributions I tried once in a while and the Commodore Basic V2.

At college I used most of my free time (and the other time, too) to learn about and try things on Linux. Starting with full desktop environments like KDE and Gnome¹ I more and more learned to love the power of the Shell and the tools available for it. So when I opted out of college in 2003 I read my mails with mutt, read the news with tin, did some serious chatting with irssi and hacked code in vim. My favorite windowmanager at that time was Openbox, because it was incredibly fast. And it did what I needed it for: managing my program windows. I had Openbox configured to be fully controlable with the keyboard. I only touched the mouse when I surfed the interwebs².

I started my new job as system administrator and software developer and still was hoping to be able to stick to my tools. But they gave me an Apple PowerBook G3 (Pismo) and told me that everybody should use Mac OS X³ because of ease of use and compatibility.

Compatibility ment, that I had to use Apple Mail for office mails and BBedit for my code. There was almost a new tool for everything. That was the reason I installed Openbox on the Mac and used it within the X11 environment for my private stuff. I now had two parallel systems on the Mac: the Mac OS X world and a world made up from tarballs and lots of compilation cycles. It was exhausting, but I tried as hard as possible to keep the tools I mastered over the years.

But because I had to switch between OS X and X11 it did not feel right. And I did not like the compile-almost-everything strategy I had chosen. I tried Macports and fink to install the necessary libraries for other things I wanted to install. It helped to speed up installations, but did not make it less complex to know later, what has been added to the system by which method or tool. So I erased the whole disk on the Pismo and installed NetBSD⁴.

I should have known the reaction of the management, when I came back to office showing around the result of the last weekend’s work: nyet!

I installed Mac OS X (“Panther” by that time) again and build vim from the sources. I resigned and from that day on I primarily used what OS X had to offer.

And I got used to it over the years.

Why did I spent my own money on Apple Hardware later?

While working in my frist job I got some hardware upgrades over the years. The Powerbook Pismo was very slow with newer versions of Mac OS X, so I got the latest and fastest iBook. Which I really enjoyed: it was really fast (at that time) and the 12” version was very handy but still useable.⁵ When the iBook died I got a MacBook Pro. One of those that still looked like the late Powerbooks. I had left the Power-PC behind me finally. And besides the benefits of being able to run virtual machines on my new hardware, I now was able to watch videos on YouTube, too.

When I left the job I had to give the MacBook back. In the last 5 years I really got familiar with the operating system and even though there were not all the great console tools available out of the box, most of the applications available for the Mac had evolved to be not so dramatically feature-less like a few years ago. And I did not really work on the MacBook at that time anymore. The last year in that job I worked at customer sides, so I often had to use their hardware or I just needed the Terminal application of OS X to get to work.

So when I quit the job I had to buy a new MacBook Pro. Because it was faster, lighter and shinier than the other laptops I knew. And I had the idea, that I might do something useful with it, when driving to my new job.

But I didn’t.

Time to become productive again!

At the new job I had to use Windows XP at the office. So this did not feel like home at all. But I already had been able to lower my expectations on productive tools once…

A couple of months later I bought a Lenovo T60 Thinkpad to install FreeBSD on it. I can’t exactly remember what has been the thought behind all this. Maybe I tried to get back the good old system I used to have before I started to work and had lots of time to play around with tools and languages…. I really don’t know. But problem was that FreeBSD did not work too well on the Thinkpad. I was able to install everything I needed and play a bit with the Haskell based Xmonad windowmanager. The main reason to get rid of the Thinkpad again was, that I wasn’t able to get the powermanagement work properly.

Another reason was the fact that I tried to make a laptop work for every day usage that was less powerfull, heavier and with an awfull screen resolution while having the shiny MacBook Pro lying around at home. It turned out that running FreeBSD on old hardware wasn’t the solution to get the feeling of being productive and creative again. I had the chance to take a new job. And I did. Even though the old job was pretty damn safe.

And lots of things changed:

when I arrived I had a Linux workstation waiting for me
I began using the awesome tiling windowmanager⁶
vim, my favorite editor was available again

Now I had a job where I had the base for a hacker’s life again. Now what? - It turned out that the job really had lots of cool projects waiting and lots of new stuff to learn and do. But this is not the story about that.

Good bye, OS X.

Because I now used Linux again the whole day it began to feel weird when I went home and had to use OS X again. I started hacking on a private Ruby project and had to admit that developing with up to date versions is not ment to be on a Mac. You always have to install the version you really need besides the version Apple had in mind for you. And the problem is, that Apple only seems to update Java once in a while.

Then you begin to use tools like Homebrew. A few years ago you used fink or macports, now Homebrew is the standard to easily install standard unix tools in newer versions or at all. Except when it does not work and that happened a lot…

Then I stumbled upon “Why OS X just doesn’t cut it” and started to look for a Linux distribution that might be installed on my meanwhile old Macbook Pro (late 2008). First I took the easy way out and installed Linux Mint 12. I had Ubuntu at work, so it felt almost the same and I was able to install into a complete Gnome 3 desktop environment which was able to recognize all hardware. The good thing was that I was able to install everything I had at the office, too.

But that changed soon and I moved from Mint to Arch. Because it’s cooler, you know? Just kidding. The more time I spent in the community again, I came across more and more stuff I wanted to test myself and I had to admit that Ubuntu/Mint still is limited when it comes to up to date software⁷ and I don’t want to install to much manually. Except for vim.

Today I removed all of my TimeMachine backups on the NAS. Good bye, Apple, I won’t come back.

I downloaded the whole sources tarballs of Gnome v1.0 back in 1999, which made my parents think about spending the rent for me living at a students home in college instead of having me around and paying for the internet bills… ↩
I never got used to terminal web browsers like lynx, w3m or elinks though. ↩
Mac OS X 10.2 “Jaguar” ↩
NetBSD was the only Unix/Linux operation system with serious support for Power-PC. ↩
I studied half-time back then and was allowed to use the iBook outside the office, too. It was nice to always carry it around. But I have to admit that using Eclipse on 1024x768 was a pain. More pain than Eclipse is by itself. ↩
I never forgot about the comfort of a tiling windowmanager since I used Xmonad on the Thinkpad for a few weeks. ↩
Not to speak about removing Sun Java (Oracle, even) from the repositories and crap like that ↩

Changes

madhatter@nostalgix.org (Arvid Warnecke) — Sat, 17 Mar 2012 00:00:00 +0100

Changes

Welcome to my new home. I moved.

Why? - I am working on a little tool, written in Ruby, to collect interesting articles on Twitter. To do so, I need to let the tool run on a server which is online 24/7. But when I tried to run my Ruby program at my old webspace hosters server I noticed that not only Ruby but the Rubygems where totally outdated. I chose the hosting because it was a special Ruby and Rails hosting. Yeah, right.

Here I have all my Ruby stuff installed by myself, so there will be everything I need. And while I am at it, I started to move this blog, too. I read Blogging like a hacker a while ago and wanted to try Jekyll since. At the moment these pages look very much like those from Tom Preston-Werner. I used his public repository to start with Jekyll. To migrate my old Wordpress blog to Jekyll I used a nice tool written in Python called exitwp. But I had to do some manual tweaks in the migrated markdown files, because I did not want to keep the old Wordpress categories.

One feature I really liked about Jekyll (or maybe static page generators in general) is the ability to push a git repository to deploy your website on the server. I used the post-receive skript from the wiki pages, but had to source ~/.bash_profile to be able to use my Ruby and Gems installed via RVM.

So, right now, I have the basic functionality here. Now I can write posts (and I really hope, that I will do that more often) and create my own blog layout when I am in the mood (I don’t like design, CSS and stuff like that very much).

Don't look back in anger¹

madhatter@nostalgix.org (Arvid Warnecke) — Tue, 08 Nov 2011 18:58:39 +0100

Don’t look back in anger¹

The Quest is over. A couple of weeks ago I wrote about how I have been looking for an alternative to del.icio.us’ bookmarking service, because they new guys running it wanted to rebuild it without keeping the old version available as well.

I found one. As already mentioned in the previous post, I have been thinking about giving pinboard a try. I have been afraid to spent money on a service I am not able to test anyhow upfront, but because there did not seem any good alternatives around I hadn’t checked already… what are my opportunities?

I signed up for about $10 a few weeks ago as far as I can tell at the moment, pinboard is the best alternative to the not-longer-available service of del.icio.us. (sigh!) If you are looking for a service like delicious used to be (i.e. working), you might want to give pinboard a try, too.²

¹ “Schiller’s Reel: Don’t Look Back In Anger”

² No, I don’t get any money for telling you this.

Why you learn nothing from tutorials

madhatter@nostalgix.org (Arvid Warnecke) — Thu, 27 Oct 2011 09:31:19 +0200

Why you learn nothing from tutorials

Maybe you do that: you want to learn something new and you are wondering where to start. Everything seems so strange and different. You are heading to Google and ask for a good tutorial to learn it by example. Of course you find lots of hits, because most people ask somewhere for a good tutorial to learn something new.

But I am wondering: did you really learn something from a tutorial yet or do you just keep asking, hoping to find any good one day which will really help you get started?

I don’t believe in tutorials anymore. I did a lot. I am pretty sure that everytime I started to learn a new programming language I looked up a tutorial and tried to learn from that. But it always became as hard and unfulfilling as it could get.

It always starts very simple and often you get lots of hints on what you will be able to do, when you will have finished the tutorial one day. Adrenalin is running out of your nose as you can’t wait to learn about all that crazy stuff, so you are heading to the next chapter.

Now there are three types of tutorials:

Simple - You start at Zero and after a few pages you are asking yourself if the author thinks that this might be the first time that you ever try to write some code. Or that you might be totally retarded…
Steep - You have difficulties from the first page on to follow what the author might be up to. Code examples are loaded with lots of stuff new to you and most of it is not discussed at all.
Bo-ho-ring - Lacking examples that might give you an idea what this is all about the author seems to be busy rambling about the history of … the alternatives to … the additional documentations on …

Of course there are those that trick you at the beginning and you get the feeling, that this might be the first or maybe the only tutorial that is different. And then, just a couple of chapters later you realize that it is just one of the three above or a mixture of them.

The worst about a tutorial is, when you kind of enjoy what you are doing, but you get the feeling that it might be good to do something else with your time. Or you feel derailed in your progress, because the author seems to have different priorities than you.

My consequence always has been to find a good reference book or sample code from others. What you need next is an idea what you want to achieve. A small application, solve one simple task, … It should be something small, should not involve forking of processes or metaprogramming in the very beginning.

Then, just start hacking.

Why did you break delicious?

madhatter@nostalgix.org (Arvid Warnecke) — Wed, 05 Oct 2011 22:15:16 +0200

Why did you break delicious?

I have been using delicious since 2003. I have been early adopter and enjoying it from the very beginning. And while I already had some concerns when Yahoo bought delicious in 2005, I stuck to the service as there was nothing else available that simple and fulfilling around.

When I read that Yahoo planned to sell delicious again without putting to much effort in enhancements ever, I have been hoping that there will be changes that will get useful features for the users who stayed while lots of others seemed to have jumped on new services where you are able to store links for offline browsing with previews and categories depending on the content (images, videos, music files, documents, notes, …).

Now changes came. And everything I liked about delicious are gone:

The simple but efficient design is gone and has been replaced with some crappy modern web2.0-like design (everything is round and new colors have been added).
Tag groups (I can’t remember if that was the real name) have been removed, so I don’t have such categories like “database” (tags like “oracle”, “mysql”, “rac”, “mongodb”, …) anymore.
The bookmarklet now opens a new post page, which is not able to auto-complete my tags, so that I have to remember if I always used the tag “hacking” or “coding”.
I am not able to manage my content that well. At the moment there does not seem to be an easy way to remove or rename already used tags. I have to search for all links tagged with the faulty tag and remove it from it.
There will be more, if I take the effort to think about it.

I am pissed. So I have been looking for delicious alternatives lately. But as most users seem to be like using one tool for everything there seem to be more services around like evernote or Zootool.

Don’t get me wrong, I use evernote, too, but I don’t think that it is the best service for easily bookmarking URLs and looking them up everywhere. I prefer the UNIX way: one tool for one thing, that does it well, instead of one tool for everything, that does everything only a bit.

As I did not know too much about Zootool I tried that, too, but when I tried to bookmark the first page with it, I got a site with all the images from that site where I could choose from. I don’t need that.

I then remembered that I already gave blinklist a try when it appeared and re-activated my account my resetting my password. But this process already took me about 15 minutes because the site seemed to be almost down. Nevertheless I imported my delicious bookmarks and tried to switch. But I did not understand how the service should be used as it did not be possible to search the bookmarks efficiently.

Another service I found while searching for alternatives was diigo, which was also able to import all my 1500+ bookmarks from delicious. After the import finished successfully I had the list with my bookmarks and all my tags in diigo, but when I tried to filter by using a tag from the list, diigo was not able to find any bookmarks by that tag. That does not help too much in my opinion.

First I thought that it might be the import’s fault, but bookmarks added by that unnecessary detailed bookmarklet did not work either. They have been added to my bookmark library but it was not possible to look them up again.

Last service I have been testing today sounds a bit like delicious, but did not work too well yet: favilous. I have to admit that I gave up after the initial import of my delicious bookmarks backup stuck for more than 5 minutes and I haven’t been able to get the site back to behave as it should. I really hope I did not break it just by trying to upload a 300 kb file…

What’s next? Right now I am not sure if I am more pissed about the broken delicious service or by looking for a worthy alternative. I have been reading about pinboard, which seems to work pretty well for all who switched since the service started. Only reason why I did not give it a try myself: it actually costs almost $10 to sign up. But because I saw so many web services lately that claim to do a great job (and they probably may do this for lots of people), but did not work for me, I am not sure if I want to spent money and risking that my search won’t be over.

But there are lots of posts available by Google which offer links to much more delicious alternatives, so maybe I will try them all first. Or somebody can name the ultimate delicious alternative and will be able to end my pain…

Deploy to Heroku

madhatter@nostalgix.org (Arvid Warnecke) — Tue, 20 Sep 2011 22:11:16 +0200

Deploy to Heroku

A team mate told me about Heroku today, when I was bashing my buggy hosting.

At the moment I am playing around with Ruby on Rails and I was looking forward to it for a while. This was the initial reason to change my hosting and move to a host who claimed to offer RoR services. But then I noticed that just deploying a very small project did not work too well.

I created a rails project

    
$ rails new blog
$ cd blog
$ rails generate scaffold post title:string body:text
$ rake db:migrate

Then I initialized git and commited it to github and heroku.

    
$ git init
$ git add .
$ git commit -m "Initial commit"
$ git add origin git@github.com:madhatter/blog.git
$ git push -u origin master

$ heroku create
$ git push -u heroku master

This went well so far. But the final step did not work that well:

    
$ heroku rake db:migrate
rake aborted!
Please install the postgresql adapter: `gem install 
activerecord-postgresql-adapter` (pg is not part of the bundle. Add it to Gemfile.)

Tasks: TOP => db:migrate => db:load_config
(See full trace by running task with --trace)

First problem, there is no gem like ‘activerecord-postgresql-adapter’.

So I installed the regular pg gem which needed postgresql to install without errors (not too remarkable, eh).

    
$ brew install postgresql
$ gem install pg

I found a few hints on stackoverflow where lots of people seemed to have similar issues and so I went on to add the pg gem to the Gemfile and use bundle afterwards.

    
group :production do
    # Only gems for heroku
    gem 'therubyracer-heroku', '0.8.1.pre3'
    gem 'pg'
end

    
$ bundle install --without production

But while moving from stackoverflow to google to heroku to the manpages I missed one very important step. I have been trying to call

    
$ heroku rake db:migrate

after doing some changes, but with no luck. Then tried something else and again tried to migrate heroku.

Of course you have to commit the Gemfile changes and push ‘em to heroku. (Doh!)

    
$ git commit -m "Adding the pg gem to the Gemfile."
$ git push -u heroku master

And then the migrate will work with no errors.

Just a few years...

madhatter@nostalgix.org (Arvid Warnecke) — Fri, 19 Aug 2011 23:22:52 +0200

Just a few years…

It’s been a while since I wrote in my blog… almost four years.

A few days ago I went to blogger.com. I have no idea why, but I also remembered my login data and there it was: my old blog I used to write in. Not too regularly, but almost once per month and sometimes even twice per day.

But then I became a father - twice - and switched the job and moved and married and … that’s it I guess. But fact is, I haven’t been crawling the web that much in that time and only coded for the money and not for the fun of it.

Because I did not use the blog anymore and did not find the time to answer comments on that old “Oracle on an iBook” article anymore (and I did not own the iBook anymore, too), I removed the blog from my webspace and never thought I might read a word I wrote into it ever again.

I don’t want to migrate the data back here and continue. This is a new blog, right now with a basic theme, as there is not too much content at the moment to have it distributed over and over a again in zillons of categories or pages. And I will write in english again, too, as I can’t imagine writing about tech stuff in german.

The tweets below will be in german. Might not be the best idea, but I don’t care.