Today was a lazy day, especially since I didn’t feel that well last night, especially since I spent the evening trying to debug why xf86-video-ati from GIT didn’t enable MergedFB on my box. The result is that having “TMDS, TMDS” as MonitorLayout on xorg.conf for a setup with a DVI LCD and a VGA LCD is now wrong (while before prevented me from having a blurry image on the VGA port) and you have to use “TMDS, CRT”. The new driver disables all outputs and enable only the ones set up on the configuration, so you need to know exactly which one you’re using.
I didn’t write the TiMidity++ maintainer’s guide (although I started it) and I didn’t have time to look at audacious yet, but I updated the backtraces guide thanks to an user who reported an improvement.
After fixing jfbterm (another piece of CJK software that was left rotting in Portage, with a default configuration pointing to /usr/X11R6/lib/X11/fonts path and missing dependencies), and seen that it does not build on FreeBSD, I’ve decided to return to what I was doing when Defiant broke down: testing all the fonts package to see if there was something broken for them.
The results are, as predictable, quite good as I didn’t find any blocking issue with the fonts, I only had one failing because of a tar commandline incompatible with bsdtar, but I fixed that easily, and keyworded that package. But I used the time I spent already on fonts to make all fonts I maintain directly (DejaVu, for a while still) and through CJK not to run scanelf a couple of times to look for ELF files and QA troubles: the packages all install only fonts, so I’ve restricted binary tests. This saves around 20 seconds per package, that are nothing if you consider them alone, but it’s quite interesting when you’re doing a full merge of the media-fonts category.
I’ve also changed a few CJK fonts to improve them a bit (like baekmuk-fonts was installing TrueType fonts only with X useflag enabled before, and another one wasn’t gzipping the PCF fonts, causing 11MB to be wasted on disk), and considered it time to ask for stabilising mikachan-font derivatives (see this entry for more information about mikachan-font story), so if all goes well in a week or two I’ll be masking mikachan-font for removal.
Off topic, a new version of GwenView is in the tree (1.4.1), libaudioss and daaplib are gone, and my backspace key is acting funny, I’ll have to try cleaning this keyboard tomorrow, as I don’t really feel like buying another keyboard for a while. And still on this topic, Betelgeuse’s setup shows he’s the third Gentoo developer having this kind of problems (the second was Olivier)… and tells me there’s a third application to set up Logitech mice… why nobody wrote one that has the entries for the devices as an external datafile rather than in a list inside the source code? And then maintained an updated datafile? Should I try to do that myself, maybe?
Anyway, this was mostly a summary of today’s crazyness, waiting for the commits to complete 🙂