liksom info

Now it has almost been a year that I started to experiment with frugal browsers. I started with ion which worked very well (even thouh the lua bit takes some time to get used to). Ion's immature developer's open Anti-Open Source stance then drove me away to wmii. Again tweaking (this time a shell script which houses the configuration) until it did what I wanted. I followed the mailing list but had soon to learn that wmii's man developers neither are the kind of people I (as luser) would trust to drive the development into my direction. So I moved on to Awesome. Very active project with developers who do not think that using firefox is your own fault. So far I really like it: Snappy performance and a very decent kind of eye candy.

Found this little movie over at archive.org:

I've just updated this site to drupal 5.0. It comes with some nifty templates, one is called antique_modern, I wonder what's antique and what's modern about it. But it is nice. Internally some improvements, but nothing I would write home about.

Just what the title says. Here is why:

Jamendo features a lot of fine music published under various cc licenses. And since I like cc and music, and since jamendo has a capable API, I wrote a little perl script which takes one or more tags as argument, grabs and displays coverart fullscreen, while playing randomly music for these tags.

What was once an extensible daemon written in Perl submitting metadata to Audioscrobbler is now a complete personal streaming system, including a shout/icecast compatible streamer and a cgi web frontend to your music collection.
For more info refer to the project's home page.

Once more I reinstalled my whole system, just because this is so much fun. Because there is nothing better than a clean new install. And all went smoothly, the new debian installer really does what it is supposed to do: install. After some three hours everything was just dandy - except that I hadn't saved my mythtv configuration - which is a pain in the ass because you have to figure out the channels yourself here in Norway.

So I was looking for a more lightweight alternative, which does not force a complete webserver and a database (mysql) upon me. I stumled upon gshowtv, which also supports xmltv and has a nifty interface to use every possible recording application.

Five more or less easy steps assuming you use elive and you have perl installed (which you have when u use elive):

• Put all the png's of your animation (aka logo) ordered numerically (xyz00001.png xyz00002.png xzz00003.png ...) in a directory created for this purpose. Put the background image there too.
• Rename your background image to bgimage.png or save the attached one there.
• Copy the attached template (template.edc) to this directory.
• Change to this directory (if you haven't yet)
• Execute the attached perl script (perl mk17bg.pl.txt).

Then you should get a file called mk17bg.edj which - if moved to .e/e/backgrounds - should show up in Configuration->Configuration Panel->Background Settings. The location of the animated logo (formerly knnown as xyz00001.png etc) can be tweaked by changing the values in the template, more specifically here:

part {
name: "elogo";
description {
state: "default" 0.0;
rel1 {
relative: 0.255 0.41;
offset: 0 0;
}
rel2 {
relative: 0.305 0.48;
offset: 0 0;
}

Have fun!
I'll probably develop the whole thing further scratching my own itch, stay tuned.
BTW: The template is of course based on the beautiful e17_bg_layered_sky created by the Rasterman himself - including the bgimage (correct me if I'm wrong).

All right! Now I am an officially approved open source maintainer. It is called scrobblerd and what this tiny collection of perl scripts does is explained on the project's homepage. I should mention that I, a sociologist, cannot hide my pride to be allowed to take part in this fine human effort, which is all about sharing knowledge, puny as it may be in my case. My humble thanks go to RMS, Linus,and all the fine engineers who lay the fundament of this great democratic cathedral (I know what I'm saying, ESR is intentionally not included, though) of free hackerdom.

Ever heard of nslu2 (aka the slug)? Fist of all it's a fabulous way to waste time. Technically speaking it is a NAS, a network attached storage. Doesn't sound time-wasting, does it? Well, the good folks from nslu2-linux.org have found ways to make this NAS, which is basically a hard disk with a network address, a full fletched pc running som sorts of linux among others also my favourite distribution debian. And here it gets interesting (for those of us who feel the need to waste time). The point is to get by with a computer which is severly lagging behind in terms of performance. About one year ago I was working with an early pentium as my main server (standing in the kitchen of my flatshare). The feeling was about the same. But: The slug, as it is called affectionately, is a tiny piece of hard-ware, no cooling needed, the only thing which makes noise is its external usb2-disk. So there is progress after all. The result is that one is working with self-imposed restrictions, why shouldn't I leave my mighty Goliath (vulgo: the Dell PC with about 15 times the memory) running instead of this puny David? Lars von Trier, the Danish director has a similar strategy: Invent restrictions where there are none to create art. His dogma95, was discussed up and down, and it was a formidable business strategy indeed, but it also is about these restrictions which are unnecessary in the strict sense. Or to quote Schiller (Fv): "Grace is the beauty of form under the influence of freedom." Maybe the slug is not grace but art is a waste of time anyway.

This little notice just to give some advice to those who ponder whether to use one of the less stable Debian flavours. I am not particularly good in fixing things but I have learned how to to use apt, how to adapt my config files, my fstab, and such things. I have a Debian Testing on my laptop and an Unstable on my toybox at home. I upgraded the latter from Testing to Unstable when some KDE programs which are essential for me (Amarok, K3B) had to be removed after an upgrade to KDE 3.4. This was 3 or 4 months ago, 3.4 had already been in Unstable a while so all my problems were fixed by the Upgrade. Shortly after Unstable moved to KDE 3.5, and the same problems appeared again but were fixed quite quickly within a couple of days. Now the system runs without flaw - well, kicker dies because of Azureus' task bar icon - which is annoying. But then Testing, which is supposed to be more stable did not recover well from the transition to KDE 3.5, which happened a few weeks ago. Until this morning I got a crash whenever I logged out and when logging in KDEinit did die (at least once), which caused lockups later whose logic I still haven't understood. But since today this system is working as it should again as well.

It was not easy to figure out without knowing what exactly you are looking for. But now my new laptop finally knows where it is and connects to the right wireless access point using the right parameters (with/without key,...).
Some scripts/apps where just too advanced in terms of configurability, for instance whereami - which has a simple but interesting concept, or in terms of resource usage, e.g. kwifimanager (I am running a fluxbox-rox-filer combo, so any kde/gnome app, though working, would be overkill). Thomas Hood did his homework and reviews a lot of roaming scripts/apps for Debian, recommending his own solution. I finally settled with corvar's even less advanced script.

One of the few things for which I had to crawl back to MS Office is history. After many days (and nights) of fighting missing dependencies I eventually got Openoffice to play nicely together with Bibus, which now gives me the same functionality as my earlier MS Word-Endnote combination. I know that many applications provide integration between comfortable administration of Bibtex-bibliographies and Lyx (e.g. JabRef), which would be everything I need in terms of usability. But I am still forced to provide MS Word compatibility for my texts - that is just how it is in academia (at least outside the sciences, I guess).

Following this guide growing qemu disks (without losing the data) is possible using a workaround involving a Knoppix ISO image.

Found out how to insert external (static) content to drupal. I guess you will encounter such content here soon.

ob_start();
include_once "http://www.trondheim-bilkollektiv.no/info/index.asp";   
$output = ob_get_contents();
ob_end_clean();
return $output;

I extended Emiliano's excellent moviewreview.module to automagically fill in imdb data. I shamelessly ripped apart and plugged in functions from Splitbrain's IMDB Parser, hopefully without breaking something somewhere. So far it works.

An impressive list of formats can be converted into each other using the OOo API. I am particularly interested in command line conversions which then could be used in scripts like the one which I plan to implement as post-hook for svn commits (having synchronised text/html-versions of the word/powerpoint stuff which I am constantly producing - for compatibility reasons of course).

Solved a little annoyance thanks to Re: how to run X apps as root?
As root:

  cd
  rm .Xauthority
  ln -s /home/user_name/.Xauthority .Xauthority

I was playing around with linuxdoc, docbook, and their relatives. Here are some magic conversions:

linuxdoc -B latex --output=pdf intro.sgml

and

sgml2html --split=1 intro.sgml

Looks nice (see attachment), is easily edited with LyX, now if only someone would write...

Two months ago my hard disk crashed (yawn), so I decided to backup my data (gasp).

Here's the code.

tar -zcvpf /home/berker/res/backup-`date '+%d-%B-%Y'`.tar.gz --directory --exclude=*/home/berker/res/* --exclude=*/home/berker/burn* --exclude=/proc/* --exclude=/tmp/* --exclude=/var/lib/mldonkey/temp/* --exclude=/var/lib/mldonkey/incoming/* . 


tar -zcvpf /home/berker/res-backup-`date '+%d-%B-%Y'`.tar.gz --directory /home/berker/res --exclude=**film/* --exclude=*backup* --exclude=*qcow* --exclude=*musi* --exclude=*imdb* --exclude=*tbdocs* --exclude=*tmp* .