Dull

The Last Post

ipenburg — Tue, 10 Feb 2009 22:36:00 GMT

This is my last post on this platform. I've migrated my content from this platform to a Bricolage implementation which I use to publish some kind of blog to my XS4ALL webspace: http://www.xs4all.nl/~ipenburg/blog/.

Spectrographing Windowlicker

ipenburg — Tue, 27 Jan 2009 19:03:00 GMT

I've all my CDs ripped to oggs and served by Mediatomb running on my OS X PPC to my Playstation 3 (just to avoid AirTunes). Ok, that was running for a while, but now I looked at the import script and found out the whole chain supports a bit of Unicode. The issue is that the Playstation's user interface is a bit annoying when you need to scroll through lists of hundreds of artists or albums, so that's why Mediatomb supports customizing the generation of these lists through a JavaScript import script. Using that script, I now have my music first divided into an index of first letters, and the next level is a combination of albums and artists starting with that letter. But to distinguish between an artist and an album I wanted a nice indicator. Which is possible because I can edit the import.js script as utf-8 and then Mediatomb can treat it as utf-8 and the Playstation also does utf-8. So I put in the Unicode symbol for sun ☉ to indicate the entry is an album, and renamed my server from New York to I♥NY just to check that out. Then I wanted to get the notorious second track of Windowlicker right, but that would need MathML because it's a bit to complicated for just Unicode. And just because you can't trust a Wikipedia article, I checked the Spectrograph myself by opening the ripped wav in Sonic Visualizer. Would be nice if the Playstation had a realtime spectograph visualizer instead of the just a bit off useless ones. But the Wikipedia article is right about the spectrograph:

CPAN author

ipenburg — Fri, 23 Jan 2009 09:21:00 GMT

I finally released a thingy on CPAN. It doesn't do anything complicated, but what it does it does well, I think. It includes lots of tests, has Kwalitee, showers under Perl Critic, is signed, has my Gravatar, has a bug and more...

REWE 20+h

ipenburg — Wed, 31 Dec 2008 07:10:00 GMT

I just found out why I prefer to go to the REWE in Blankenese after 20:00.

p3nfsd for mounting Sony-Ericsson P800 filesystems on Mac OS X PPC

ipenburg — Sun, 30 Nov 2008 03:36:00 GMT

Finally managed to get around to getting p3nfs working with my P800 and iBook. It involved compiling the source for the server, installing the provided binary .sis on the P800, adding a Bluetooth Serial port to the P800 in OS X, using plutil -convert xml1 to be able to edit /var/root/Library/Preferences/blued.plist by hand to change the channel of the just created SerialPort to 13, setting up some nice StartupItem to have the service allways running and after a reboot it all works.

FileVault

ipenburg — Sat, 04 Oct 2008 01:48:00 GMT

It's Tag der Deutschen Einheit in Germany, so a good reason to encrypts some disks. But I hit a snag:

Turning on FileVault requires an additional 4053.9 GB of free disk space to create an encrypted copy of the home folder. Try emptying the Trash or deleting files you don't need.

Why do I need 4TB of free space to encrypt my home folder which weighs currently in at 68GB?

WordPress?

ipenburg — Sat, 30 Aug 2008 12:07:00 GMT

Because this .Text blogging application is sucking harder and harder I looked at some other applications, WordPress being the most prominent. WordPress is a PHP+MySQL rig, so we don't expect too much from it. It's nice that it's apt-get installable on Debian, but if we then try to upload an image:

Unable to create directory /usr/share/wordpress/wp-content/uploads/2008/08. Is its parent directory writable by the server?

Huh? We've given WordPress write access to our MySQL database already, why would it need write access on the filesystem to store images? Because somehow the database is too crappy to store blobs without a performance hit, the programmers don't know how to do it, or it somehow is a feature to make scaling and migration more complicated than just moving a database around?

Misunderstanding Social Networks

ipenburg — Mon, 18 Aug 2008 01:09:00 GMT

Spending some time using social networking sites like LinkedIn, Xing and Hyves I noticed some users trying to get to the core of a network. Like there is some hierarchy involved and social networking is about getting to the one person at the top, or becoming the one person at the top. But there is no hierarchy and there is no top (and no visualization tool in LinkedIn that shows that). It's easy to mistake LinkedIn's 1st, 2nd and 3rd grade connections for a layered organization with you on top, but you are the only one seeing it that way. Influence doesn't flow that way either, there is not one source that provides knowledge and trends to other nodes and thus there is no point in being close to the source to have an advantage. The social network is like a traffic flow in which you can't understand the congestion of the flow by just looking at the individual vehicles, especially if you're one of the vehicles yourself. You can't find the cause of a traffic jam by asking all individual drivers involved if they caused the jam, because it takes complex interaction between a group of several drivers to cause it. In the same way there is no single node in a social network that causes waves of influence in the network. It will be a group of nodes, or the network as a whole that generates waves of influences. Social networking isn't about cutting out the middleman to get closer to a source, the middleman is part of the source.

Blankenese

ipenburg — Thu, 14 Aug 2008 14:15:00 GMT

I moved to Blankenese, by the way. A lot of big old trees with big new cars parked under them, as I noticed during the storm the last few days.

View Larger Map

Compiling KDE 4.1

ipenburg — Wed, 13 Aug 2008 00:42:00 GMT

Since I didn't have a lot of stuff on my mind for a couple of days I decided to screw up my Debian system some more by going for a compiled from source KDE 4.1. I've enough experience with open source stuff to be not very optimistic about getting it all working in the end. Before you know it you are up to your waist in dependencies and if one of them fails it's over. While I do admire the community for releasing open source packages like KDE, building from source shows that it's a lot easier to just buy a Mac. The main issue is that the current state of development means for every little piece of functionality there is a library, and they all have an API, and the chance that the whole cascade of used libraries collapses due to one incompatibility between versions of a library is way to big. Libraries get more specialized, easier to maintain due to the use of other specialized libraries, but beyond one library depending on another a big project like KDE is just to vulnerable to incompatibilities between hundreds of used libraries. In the end - after a lot of tar xvzf && ./configure && make && make install - I got something working which looked like a nice KDE, but along the way I got a lot of errors that showed some libraries lied about their dependencies, and obviously the install was never tested on a system that started from scratch. Of course we can wait for the official distributed packages, but how open source is that?