HomeCategoriesChoose a templateRecent EntriesEmdebian Grip updated
Monday, September 6 2010 screen, irssi and page control Sunday, September 5 2010 FreedomBox Wednesday, August 4 2010 check-deps.sh and xapt Thursday, July 8 2010 Switching from iceweasel to chromium Sunday, June 27 2010 World Cup QA Sunday, June 20 2010 multistrap 2.1.5 Monday, May 31 2010 HP laptop battery recall Monday, May 24 2010 DebConf10 Monday, April 26 2010 pdebuild-cross Saturday, April 24 2010 |
Wednesday, March 28. 2007Comments (0) Trackbacks (0) Defined tags for this entry: Emdebian
Cross-building a rootfs
With updated cross-building toolchains and a much improved emdebian-tools package, work has finally started on actually building cross-built packages for Emdebian. Wookey has identified the first set - a usable rootfs. http://wiki.debian.org/EmdebianRootfs
Nineteen packages have been cross-built so far, from base-files and popt to util-linux and ncurses so I thought I'd make a note of how it has gone at this early stage. Packages, patches and build logs can be viewed at: http://www.emdebian.org/packages/search.php Results: ====== 1. CDBS cross-builds like a dream. No failures, no missing files, no misplaced files, no files Emdebian doesn't want (like changelogs and manpages). em_make replaces one makefile with a customised one and everything else falls into place. Fantastic. 2. debhelper packages generally cross-build quite well - as long as debian/rules sticks to actually using debhelper rather than creating bespoke rules that just do what dh_foo would have done anyway! (I then have to go behind them and remove the custom rules to get rid of manpages etc. when if the task had been left to dh_foo, em_make would have removed dh_foo automatically. Grrr.) 3. Packages built without CDBS or debhelper are an absolute PITA. I know it is a good exercise to build without "high-level" tools, but PLEASE, only put non-debhelper packages in the main archive if you really cannot have debhelper involved. It'd make my life SO much easier right now. 4. Splitting out the translation files into individual, dedicated, packages one per language is having two related effects: there are a plethora of packages in the repository with only 19 source packages built; the size of the main binary packages has dropped substantially. Some translation packages are 30-60kb in size but in Debian, 12 maybe 30 of these are packaged inside the main binary or a similar combined _common package. Users get no choice, it's no translations or all translations. I cribbed the idea from OpenEmbedded and it's nice to see it paying off, even if it means that ' apt-cache search' listings will be at least an order of magnitude longer in Emdebian. (Sounds like a wrapper is needed.) Need to start testing 'langupdate' soon to make sure it can cope with keeping users up to date with translations of installed programs.So far, haven't needed too many cache files to be used with ./configure --cache-file but I'm bound to meet those soon.Overall, things are going well (famous last words). Saturday, March 24. 2007Comments (0) Trackbacks (0) Defined tags for this entry: Emdebian
Emdebian repository layout
OK, after a long time and a few reshuffles, emdebian has a 'logical' SVN repository layout with a clear division between current and superceded code, clear module separation and updated documentation.
[ wiki docs ] [ website docs ] [ repository browser ] To check out the current code: svn co http://buildd.emdebian.org/repos/current/ A full check out of even just the current code can easily be over 20Mb and is likely to increase as more packages are emdebianised. See the list of modules available. Friday, March 23. 2007Comment (1) Defined tags for this entry: Debian
sid or unstable in /etc/apt/sources.list
I've just come across a tricky little bug in using non-standard directory options to apt when /etc/apt/sources.list contains references to 'sid' rather than 'unstable'.
Update: guess what I'll be working to fix tomorrow . . . Friday, March 23. 2007Comments (4) Defined tags for this entry: Debian
gecko, nvidia, fonts and editing forms - Bug 345700
Update: Fixed. The workaround described in bug 345700 works perfectly.
Add Option "XaaNoSolidFillRect" to the Screen Section of /etc/X11/xorg.conf, as described in the bug report. It's hard to identify initially - first thoughts are always to blame Gecko-based browsers but gradually the problem migrates into the nvidia driver. When editing text in a HTML form entry field, moving the cursor with the cursor keys results in horrid vertical lines: a|b||c|4|e which disappear when clicking elsewhere in the control but make editing VERY difficult. The vertical lines do not always appear between characters, they often overlay the characters, something I cannot represent here. (The bug report provides a good description.) It is definitely something to do with Gecko because Gtk does not suffer this problem (which is one reason why I use drivel for blog entries). Galeon, Iceweasel and epiphany-browser are all affected. One simple example is the Debian wiki - normal, uncomplicated text inputs, no special Javascript, but moving around the text in a control is almost impossible. Monday, March 19. 2007Comments (3) Trackbacks (0) Defined tags for this entry: Debian
network-manager failures
In response to a couple of comments about my post about pesky binary firmware, I can only say that after removing the last entrails of network-manager and network-manager-gnome, everything appears fine. Command-line configuration wins over GUI eye candy, again.
At no point did network-manager[-gnome] cope with my setup, principally due to WPA difficulties, it did connect once when WPA was disabled at my Access Point at home but that simply isn't an option in the real world. (And it's not a problem with my home setup because network-manager also failed (comprehensively) with a completely different wifi setup - the one I'm using here on hols.) Oh and by the by, if anyone has a problem with the template I've chosen for my blog, there is always the option to select a different template via the drop-down box. I happen to like the current one and it's my blog so, umm, that's how it will stay. Sunday, March 18. 2007Comments (2) Trackbacks (0) Defined tags for this entry: Debian
pesky binary firmware
Aaarrgh! Trying to post the last blog entry revealed that my wireless connection had simply died. Various restart options failed (ifdown/ifup, dbus restart etc) so the realisation dawned that the binary blob that gets the Airport Extreme card working is not always coming out of suspend correctly. This time, I did a full restart (shutdown -r now) but I hope this isn't necessary every time. Next time, I'm going to see if a quick suspend / unsuspend is sufficient - watch this space....
Sunday, March 18. 2007roaming SMTP in sylpheed
After finally getting wireless connectivity whilst on hols, the only thing left was email. I'd already configured fetchmail by copying the fetchmailrc from my email server before I left but I'd forgotten to do procmail|maildrop. Some reason, can't seem to get maildrop to work with maildir (which is what sylpheed uses for 'local' email accounts) so I'm filtering messages manually for now. Adding a custom action to sylpheed to run 'fetchmail -s' is nice but it would be even better is sylpheed could call that action when checking for email. (I feel a wishlist bug coming on.)
SMTP was initially more of a problem - SMTP AUTH was available via the excellent ukfsn so creating a new Roaming account in sylpheed allows the configuration of SMTP AUTH just for that account. Problems with forwarded email addresses (like codehelp at d.o) and mailing list subscriptions meant that I'm now using a private server but ukfsn were perfectly usable for non-mailing list usage of 'normal' email accounts. It's only my own, overly complicated, forwarding and greylisting systems that made mailing list submissions into a problem. The remaining problem is power consumption - the wireless connection on this iBook (Airport Extreme internal) seems very power hungry. All the testing and configuring that was needed for the above cut my battery life in half. I can normally get 4hrs plus of 'average workload'. This will have implications for using wireless at DebConf etc. Even if sylpheed could call a custom Action (like fetchmail) automatically, it wouldn't be wise to keep the original 10minute checks without AC power. Wednesday, March 14. 2007Defined tags for this entry: Emdebian
g++-3.4 and g++-4.1 cross-built side-by-side
edos-debcheck reports that the emdebian toolchain for alpha built from gcc-3.4 cannot be installed on i386 (and possibly amd64) when the repository also contains the toolchain for gcc-4.1:
libstdc++6-dev-alpha-cross (= 3.4.6-5): FAILED The following constraints cannot be satisfied: libstdc++6-dev-alpha-cross (= 3.4.6-5) depends on libstdc++6-alpha-cross (>= 3.4.6-5) {libstdc++6-alpha-cross (= 4.1.1-21)} Yet there is no problem with these for arm or other architectures (except possibly s390). In the main Debian archive (native build), g++-3.4 depends on the gcc-4.1 version of libstdc++6. http://packages.debian.org/unstable/devel/g++-3.4. $ edos-debcheck -explain -failures -check libstdc++6-dev-arm-cross < combined reports no errors. ('combined' here is a result of appending the Packages file from Debian to the Packages file from Emdebian although the Emdebian repository is slightly behind my own data. I'll sort out the update tomorrow.). The repository can only contain one version of libstdc++6, just like any one system can only install one version (at least in /usr/lib): $ reprepro list unstable libstdc++6-alpha-cross unstable|main|i386: libstdc++6-alpha-cross 4.1.1-21 unstable|main|amd64: libstdc++6-alpha-cross 4.1.1-21 unstable|main|powerpc: libstdc++6-alpha-cross 4.1.1-21 By building separately, the emdebian cross-build has a dependency on 3.4.6-5 and despite the >=, edos-debcheck reports that the dependency cannot be satisfied. $ dpkg -I /var/emdebian/debian/pool/main/g/gcc-3.4/libstdc++6-dev-alpha-cross_3.4.6-5_all.deb | grep Depends Depends: gcc-3.4-base (= 3.4.6-5), g++-3.4-alpha-linux-gnu (= 3.4.6-5), libstdc++6-alpha-cross (>= 3.4.6-5), libc6.1-dev-alpha-cross (>= 2.3.5-10) The next cross-build of gcc-3.4 on i386 for alpha needs to use the version of libstdc++6 from the repository instead of making an older version via dpkg-cross. Not sure why this only affects alpha cross-builds. |
ArchivesSyndicate This BlogQuicksearch |
