**** BEGIN LOGGING AT Sat Apr 02 02:59:58 2011 Apr 02 10:36:39 hmm.. Apr 02 10:37:06 linux-kexecboot fails to compile under testbuilder but compiled fine by direct bitbake invocation.. Apr 02 10:37:51 /var/tmp/tb-clean-build/work/collie-angstrom-linux-gnueabi/linux-kexecboot-2.6.38-r35.1/linux-2.6.38/scripts/gen_initramfs_list.sh: Cannot open 'initramfs.cpio.lzma' Apr 02 12:36:14 /na Apr 02 12:36:17 sorry Apr 02 12:36:18 XD Apr 02 13:01:52 03Andrea Adami  07org.openembedded.dev * r3e4c19efcd 10openembedded.git/recipes/linux/ (7 files in 7 dirs): Apr 02 13:01:52 linux-kexecboot: add new kernel task do_savedefconfig Apr 02 13:01:52 * shrink and recommit Zaurus defconfigs Apr 02 13:01:52 Signed-off-by: Andrea Adami Apr 02 13:22:06 hi mickeyl Apr 02 13:22:53 hi woglinde Apr 02 14:40:28 Can someone look at the patches I sent to mailing list? I sent them 2 days ago and got basically no feedback neither those were pushed. Apr 02 14:43:00 otavio, the patchwork is broken and is beintg fixed Apr 02 14:48:12 ka6sox-away: sure but it has been sent to ml as well Apr 02 14:48:23 ka6sox-away: so I see no problem on tracking it there too Apr 02 14:49:45 in the last 4 days I counted over 200+ patches that have been sent to the ML...takes time to review... Apr 02 14:50:26 ka6sox-away: and if it is the case, I can make them available ... hold ... done, check at https://projetos.ossystems.com.br/git/?p=users/otavio/org.openembedded.dev.git;a=summary Apr 02 14:51:00 ka6sox-away: I see ... but some patches were sent a week before without an ack/nack (for example the udev one) Apr 02 14:51:57 otavio, I am just a poor sysadmin who is struggling to keep up with the volume of work to keep things running. I can imagine that the folks who are reviewing are just as swamped. Apr 02 14:52:16 ka6sox-away: heh Apr 02 14:58:05 heh.. overriding IMAGE_FSTYPES affects linux-kexecboot at least Apr 02 14:58:23 because of initramfs image is not generaged then [in right format] Apr 02 14:58:32 *generated Apr 02 14:58:44 otavio, part of the reason things are so backed up is that somebody pushed a HUGE TS7500 patch and it choked the patchwork server. Apr 02 14:58:51 base64 encoded. Apr 02 14:59:08 ynezz: be careful with IMAGE_FSTYPES in testbuilder Apr 02 15:01:14 heh Apr 02 15:01:29 heh or not...that is what broke it. Apr 02 15:02:25 so everything AFTER that patch has to be Hand fed into the system...again... Apr 02 15:04:14 * mwester rather suspected that patch-nuclear-bomb was going to cause trouble... Apr 02 15:04:55 A simple fix might be to simply bounce incoming patch emails that exceed a certain size... Apr 02 15:05:31 mwester, if we had a python-django programme it would probably be simple Apr 02 15:05:37 programmer. Apr 02 15:07:12 Yep. I see it all the time in my line of work -- someone finds a "new" language, and writes some wizz-bang code with it, and never once figures the ongoing cost of maintaining that... So I get asked to provide APIs for my employer's product for some of the most ridiculous languages that have no place in build systems... Apr 02 15:07:42 Too many programmers think only of the "artistic" or "creative" aspect, and ignore the practical. Apr 02 15:08:52 yes, and worse even modern languages like ruby don't seem to care about backwards compatibility and routinely break because the gemset isn't supported on the *next* version. Apr 02 15:09:28 even point releases are guaranteed to not work. Apr 02 15:11:47 anyways...so now someone has to manually cull thru the ML for the last 5days, find the *original* patch submittal, cut it out of the ML and hand feed those to the patchwork portal... Apr 02 15:13:59 and of course, any patches that are based off that original nuclear bomb patch have to be discarded or they will choke the system to as having no reference. Apr 02 15:14:13 s/to/too Apr 02 15:14:22 is patchwork working as hook to smtp or fetching mail via imap/pop3? Apr 02 15:14:32 the later Apr 02 15:14:36 latter Apr 02 15:15:08 fetchmail to a parser Apr 02 15:18:42 :-/ Apr 02 15:20:16 so instead of processing and reviewing patches for the last number of days the group has been trying to find the bomb that went off and the group has been dealing with the PC LOAD LETTER issues and not working on reviewing and processing patches Apr 02 15:45:20 ka6sox-away: would the limit of the email size help? http://pastebin.com/55LQit82 Apr 02 15:45:31 ka6sox-away: or it's that decoding causing problems? Apr 02 15:46:14 Jay7: noted Apr 02 15:48:21 ynezz, okay I like that kind of approach for limiting size, but somewhere that needs to be documented in the submission guidelines... Apr 02 15:48:34 well, it's just mockup Apr 02 15:48:38 I"ve not looked for them Apr 02 15:49:01 ynezz, understood. Apr 02 15:49:12 Looking at that code fragment, it appears that the entire email message is read into a string variable - correct? Apr 02 15:49:21 we still need a django person. Apr 02 15:49:24 and is the problem CPU or memory on the host? Apr 02 15:50:22 memory is fine, cpu is fine, parsemail.py just choked on parsing an oversized thing. Apr 02 15:50:32 ah, ok. Apr 02 15:50:33 how you can check for size of stdin anyway Apr 02 15:51:06 dunno, never figured that one out :D Apr 02 15:51:19 You can't -- which is making me wonder if this program is an architetural disaster waitiing for a DOS attack. Apr 02 15:51:34 well, the check could be moved to the while loop Apr 02 15:51:39 and limit the read then Apr 02 15:51:44 good point Apr 02 15:52:25 if not data or len(data) > size_limit: Apr 02 15:52:46 but well, some python guy could handle it better :) Apr 02 15:58:36 anyway, this v2 should limit the RAM usage also http://pastebin.com/j7abCAhp Apr 02 15:59:40 kergoth: we need some way to limit the size of the stdin input in python, do you think that the above patch is ok? Apr 02 16:00:38 one huge patch took the patchwork down... Apr 02 16:01:33 bbl Apr 02 16:34:55 hello Apr 02 16:36:04 can anyone help me with this error: Error could not inherit file classes/rootfs ${IMAGE_PKGTYPE).bbclass Apr 02 17:39:30 hello Apr 02 17:39:44 i have build oe and all ipks are under deploy/ipk Apr 02 17:40:06 can I move this ipk directory to a local disk(no network access) and somehow tell opkg.conf to install from there? Apr 02 17:40:08 how to do it? Apr 02 17:40:55 you need a Packages.gz Apr 02 17:41:09 woglinde: it's all under the deploy/ipk directory Apr 02 17:41:27 i want to move it to local SD instead of putting it on a webserver Apr 02 17:41:33 as my network is not up yet Apr 02 17:42:06 dont know if opkg can use file Apr 02 17:42:16 even better i'd like to add more stuff to the final rootfs, but i don't know a quick way to do that, say, how to add pciutils to the rootfs Apr 02 17:42:34 in the latter case that i don't need mess up with opkg to SD Apr 02 17:43:20 ? Apr 02 17:43:28 look the recipes/images Apr 02 17:43:36 there are plenty of examples Apr 02 17:44:30 * xxiao wishes there is a menuconfig thing to add needed packages quickly Apr 02 17:44:46 * xxiao is checking recipes/images Apr 02 20:02:46 i want to put helloworld.elf to oe's final rootfs, how can I overlay this to the final rootfs? Apr 02 20:03:18 bascially i want to overlay some binaries to oe's final image each time i build oe, how to do that? Apr 02 20:03:32 under openwrt i can put a files/ directory under the topdir Apr 02 20:13:28 again Apr 02 20:13:33 write recipes Apr 02 20:13:38 look into recipes/images Apr 02 20:14:08 there are plenty examples how to include packages into your rootfs Apr 02 20:28:59 woglinde: not package,i mean some pre-compiled binaries to overlay into various subdir in the rootfs Apr 02 20:29:21 already learned and got packages adding-on going, thanks Apr 02 20:33:40 yeah as I said Apr 02 20:33:44 make a recipe Apr 02 20:33:49 which makes a package Apr 02 20:33:52 and on you go Apr 02 20:34:09 sure can use the one hook Apr 02 20:34:12 but this suckz Apr 02 20:34:34 ok Apr 02 20:34:37 and again look at recipes/images Apr 02 20:34:59 its the last time I am repeating this Apr 02 21:49:45 03Andrea Adami  07org.openembedded.dev * r4f4093253c 10openembedded.git/recipes/linux/linux-kexecboot-2.6.38/ (6 files in 6 dirs): Apr 02 21:49:45 linux-kexecboot: update 2.6.38 zaurus defconfigs (add RTC) Apr 02 21:49:45 Signed-off-by: Andrea Adami Apr 02 21:53:52 I need so simple gui, as It can to run only one GUI application (hope, It could be FLTK-based application). what xserver try better? xorg-fbdev? do I need GPE to run my application? Apr 02 21:56:50 ? Apr 02 21:57:07 gpe is only a set of x apps Apr 02 21:57:21 running on matchbox2 windowmanager Apr 02 21:57:37 woglinde, yes, give me a route/idea to think about... Apr 02 21:57:50 ? Apr 02 21:57:52 woglinde, ah, matchbox? Apr 02 21:58:01 get familiar with what a X app is Apr 02 21:58:29 what is receipt? Apr 02 21:58:37 for matchbox I mean Apr 02 21:58:57 why would the recipe help you? Apr 02 21:59:05 find -name is your fried Apr 02 21:59:08 friend Apr 02 21:59:19 I just need a route to see what I need. matchbox-desktop or something else? Apr 02 21:59:50 matchbox-wm ? Apr 02 22:01:25 find -name "matchbox*" ? Apr 02 22:02:35 yes. there are matchbox-wm and matchbox-desktop at least. -wm is smaller? Apr 02 22:09:46 dont know **** ENDING LOGGING AT Sun Apr 03 02:59:58 2011