On Sunday 18 March 2007, Michael Schwendt wrote: >On 18/03/07, Sam Varshavchik wrote: >> At this point, what I would ordinarily do now is fire up gdb, and step >> through the stupid thing to see exactly what it's doing. But, if you >> knew how to do this, I'd think you would've done it already, I >> presume, so that's not going to work. Time to try a different >> approach. > >First would come simple tests with "fc-list" and "fc-cache -v > /usr/share/fonts". > We've done the latter many times now, with the same error output. The former spits out a list of fonts, and its pretty eclectic, but one thing sticks out, none of the msttcorefonts seem to be listed there. >> It looks to me like that there's a corrupted file somewhere that the >> stupid thing is trying to read, and when it can't make heads or tails >> of it, it blows up. > >/etc/fonts/fonts.conf could turn out to be a hot tip, Is this an xml file in disguise? Both it and local.conf appear to be xml code, although that's not anything I've ever walked around in, so I wouldn't know a bad file if it hit me over the head. >as nothing has >written to /var/cache/fontconfig since Nov 9th. Reinstalling the >fontconfig package or just the config file might be enough. I went to rpmfind and found it, dl'd it and did this, which seems self-explanatory: [root@coyote dlds-rpms]# ls fontconfig* fontconfig-2.4.1-3.fc6.i386.rpm [root@coyote dlds-rpms]# rpm -q fontconfig fontconfig-2.4.1-3.fc6 [root@coyote dlds-rpms]# rpm -ivh --force --nodeps fontconfig-2.4.1-3.fc6.i386.rpm Preparing... ########################################### [100%] 1:fontconfig ########################################### [100%] /usr/share/fonts: failed to write cache /usr/share/fonts/bitmap-fonts: failed to write cache /usr/share/fonts/bitstream-vera: failed to write cache /usr/share/fonts/default: failed to write cache /usr/share/fonts/default/Type1: failed to write cache /usr/share/fonts/default/ghostscript: failed to write cache /usr/share/fonts/dejavu-fonts: failed to write cache /usr/share/fonts/dejavu-lgc: failed to write cache /usr/share/fonts/japanese: failed to write cache /usr/share/fonts/japanese/TrueType: failed to write cache /usr/share/fonts/japanese/misc: failed to write cache /usr/share/fonts/linux-libertine: failed to write cache /usr/share/fonts/msttcorefonts: failed to write cache /usr/share/fonts/perl: failed to write cache /usr/share/fonts/terminus-font: failed to write cache /usr/share/fonts/tv-fonts: failed to write cache /root/.fonts: failed to write cache /root/.fonts/kde-override: failed to write cache /usr/local/share/fonts: failed to write cache error: %post(fontconfig-2.4.1-3.fc6.i386) scriptlet failed, exit status 19 Sob, moan, sniff, same story. What else could this be? Humm, using mc, the swiss army knife of file managers, I find that the font.conf in my /etc/fonts directory is a hair over 11k, but the one in the rpm is only 5239 in length. So I forcibly overwrote the one in /etc/fonts with the one from the rpm, something I would have expected the install command line I gave rpm above to do, no questions asked. But it DIDN'T! But guess what folks: [root@coyote yum]# fc-cache -v /usr/share/fonts: skipping, 0 fonts, 4 dirs /usr/share/fonts/bitmap-fonts: skipping, 31 fonts, 0 dirs /usr/share/fonts/bitstream-vera: skipping, 10 fonts, 0 dirs /usr/share/fonts/default: skipping, 0 fonts, 2 dirs /usr/share/fonts/default/Type1: caching, 35 fonts, 0 dirs /usr/share/fonts/default/ghostscript: skipping, 8 fonts, 0 dirs /usr/share/fonts/dejavu-lgc: skipping, 21 fonts, 0 dirs /usr/share/X11/fonts/Type1: skipping, 29 fonts, 0 dirs /usr/share/X11/fonts/OTF: skipping, no such directory /root/.fonts: caching, 75 fonts, 1 dirs /root/.fonts/kde-override: caching, 0 fonts, 0 dirs /usr/local/share/fonts: caching, 0 fonts, 0 dirs /usr/X11R6/lib/X11/fonts: skipping, no such directory /var/cache/fontconfig: cleaning cache directory /var/cache/fontconfig: invalid cache file: test-by-gene fc-cache: succeeded Now my question is "why was it 'skipping' many of those?"? However, if I run it without any arguments, presuming it would only output errors, its 5 seconds of silence and the prompt is back. Yay! Whoopeee (and several other best left out of mixed company expletives) I *think* the SOB is fixed. But, whyinhell didn't rpm fix it when I _distinctly_ told it too? That I believe is something the rpm maintainers need to check out, ASAP!!!! It should have at least made an .rpmsave out of the old one it found and written the archive copy in its place, but it didn't touch it! IMNSHO, that failure=rpm IS busted. And this definitely isn't rocket science to come to that conclusion. ggrRRROOOOOWWWWLLLLLLLLllllll....... Somebody with more patience for bz than I have right now should fwd this whole damned tread to bugzilla. Underlined, capitalized, complete with all the sound effects of a frustrated user. I hope the neighbors don't call the law to have them investigate all the screaming coming from our little gully in the West Virginia hills. -- Cheers, Gene "There are four boxes to be used in defense of liberty: soap, ballot, jury, and ammo. Please use in that order." -Ed Howdershelt (Author) electromagnetic radiation from satellite debris
