auf.kante

Wed, 31 Aug 2005

goCluster updates

Not that I would care about publications anymore but the paper on goCluster has been published in Bioinformatics and drew some additional users to the package which is nice.

So I added a project page for the software and this will hold some additional information on the package. For now I just added answers to some of the questions I got, a link to the newest version and the developer versions of the vignettes.

Since I wanted to check that the package is up-to-date for Bioconductor 1.7 (to be released soon) I rebuilt an R-development environment. This required the developer version of R so I adapted the Gentoo ebuild to the developer version. You can download it here. I'll provide no guarantee that this thing works. It installed without problems on my desktop but failed on my laptop (though I believe this was not related to the ebuild but some other problem on the machine).

Posted at: 22:58 | Permalink | category: /english/technical/projects

Tue, 30 Aug 2005

This is "boring" ...

... a nice piece of music from Carouge. You can download some of their songs at music.download.com. They classify it as "Euro-Pop" but I think it's not really dance music. Little bit too moody for that, but I like it.

Posted at: 22:04 | Permalink | category: /english/personal/media

Added svnreplicate, updated delicious API

Since John added some fixes to the delicous-API for Emacs I updated my ebuild. I am also trying to get an automatic backup system in place for my subversion repository so I created another ebuild for SVNreplicate. I did not get it working yet but that's not a problem with the code itself. Right now Apache just does not like me...

Posted at: 21:57 | Permalink | category: /english/technical/gentoo

Sun, 28 Aug 2005

Muse mode update

I updated my muse ebuild to the newest version 3.01.91. The directive for configuring the pyblosxom blog has changed and it looks like recursive directories work out-of-the-box now. Thanks Micheal!

My blog configuration as a reference:

	("Blog"
	 (,@(muse-project-alist-dirs "~/usr/devel/website/home"))

	 ,@(muse-project-alist-styles "~/usr/devel/website/home"
				      "~/usr/devel/website/test/home"
				      "blosxom-xhtml"))

Looks like the font-highlighting has also improved. Nice :)

Posted at: 11:05 | Permalink | category: /english/technical

Thu, 25 Aug 2005

py-webapp-config

webapp-config is the application Gentoo provides for the management of web applications. It provides the capabilities to safely install, upgrade or uninstall web tools into different virtual hosts. I can definitely recommend it to anyone hosting applications.

The first version of the tool has been written in bash which has some drawbacks. The main problem is the time it takes to install a larger web application into a virtual host. In addition bash scripts are hard to handle once the application grows in size. So one outstanding task has been porting the application to python.

Since I did nearly all my coding in python the last three months it wasn't too much of an effort to produce a first draft version of a pythonized webapp-config.

It is able to install and clean web applications but I expect that this version still has a number of serious bugs. I recommend using the --pretend flag when running it. That should avoid most damage to your system.

Download the current version here.

Posted at: 13:48 | Permalink | category: /english/technical/gentoo

Sun, 14 Aug 2005

Updated the list of experimental ebuilds

The ebuild page has been updated to include some more ebuilds. Most of them can also be found in the gentoo bugzilla. The easiest way to access them is using gensync to access my repository (instructions at the end of this page).

Two of my ebuilds have been added to the portage tree (knowledgeTree and OpenWebStats). Hope they will be useful...

Posted at: 18:32 | Permalink | category: /english/technical/gentoo

A new accounting tool

During the last few months I have used GnuCash to do track where my money went. Recently I discovered legder which can be used for the same kind of task. Is it any better? I'd say without a doubt. It has no graphical frontend and uses plain text files for the management of the data but it's capabilities are clearly superior to GnuCash. One thing I was lacking in GnuCash was the ability to plan ahead. This is absolutely no problem with ledger and the generated plans can later be used to easily compare the budgets with the amounts of money actual spent.

Some parts of the application are still beta (didn't get python support to work) but the basic functionality is already impressive.

There is an emacs mode available for the application but this is still very basic. I hope I'll get the time to add some functions I would like to have.

If you want to run the thing under Gentoo then you can download the experimental ebuild. The python use flag of this ebuild does not work yet.

Posted at: 18:16 | Permalink | category: /english/technical/gentoo

Thu, 11 Aug 2005

Jammin Inc

I think a full album ("Mit Anlauf") released under the Creative Commons License deserves a note. Jammin Inc is a german band which crosses Hip Hop, Reggae and some rock to an interesting new mix. They gave a really good interview (in german) to the Tonspion and explained why they chose to use the free license.

Though I am not much of a HipHop fan I can definitely recommend the songs "Stop" and "System overload".

Posted at: 07:53 | Permalink | category: /english/personal/media

Wed, 10 Aug 2005

Scientific fraud

This article from Wired News has some info on the recent rise in scientific fraud. This topic started to fascinate me when I read somewhere that the phrase scientific fraud did not exist before the seventies. This is probably not quite correct because there are documented cases of scientific fraud even in the twenties and thirties. There is for example the case of Paul Kammerer who tried to prove the theories of Lamarck. Another famous case is the research in heritability of intelligence conducted by Cyril Burt.

Nevertheless the number of incidents of scientific fraud have been clearly increasing since the eighties. A reasonable explanation is given by David Goodstein:

In the past few decades, however, that state of affairs has changed dramatically. Science is now held back mainly by the number of research posts and the amount of research funds available. What had been a purely intellectual competition has become an intense struggle for scarce resources. In the long run, this change, which is permanent and irreversible, will probably have an undesirable effect on ethical behavior among scientists.

Currently the number of junior scientists is constantly rising while the number of positions in higher ranks remains constant. The impact factor of journals gets more and more important since it allows to compare the "quality" of scientists. There are a number of such indications of increasing competition in science.

What if competition does not provide the most productive basis for science? True, this type of rivalry apparently works quite well for our economy. But there is more to human interaction than competition and science might benefit far more from the collaborative part of our nature. Projects like WikiPedia present fascinating examples of the human ability to aggregate knowledge at an extremely fast pace.

While I'd like to see competition in science actually dwindling I don't believe it to happen anytime in the near future. Hell, I would not have left the field otherwise...

Posted at: 23:24 | Permalink | category: /english/personal/news/science