**** BEGIN LOGGING AT Wed Aug 13 02:59:57 2008 Aug 13 07:56:51 bonjour Aug 13 09:42:06 * * OE Bug 4495 has been created by ronan(AT)aimao.org Aug 13 09:42:08 * * recipe for Fusil, a Python library used to write fuzzing programs. Aug 13 09:42:10 * * http://bugs.openembedded.net/show_bug.cgi?id=4495 Aug 13 09:43:00 !oebug 4495 Aug 13 09:43:00 * * Bug 4495, Status: NEW, Created: 2008-08-13 11:42 Aug 13 09:43:03 * * ronan(AT)aimao.org: recipe for Fusil, a Python library used to write fuzzing programs. Aug 13 09:43:04 * * http://bugs.openembedded.net/show_bug.cgi?id=4495 Aug 13 09:43:05 oops :) Aug 13 11:16:08 hi Aug 13 11:16:51 i am buildinf fso and i experiment unusal issue when compiling illume-theme-freesmartphone Aug 13 11:17:04 http://rafb.net/p/Wc2H1Z86.html Aug 13 11:18:59 any idea for fixing this please ? Aug 13 11:27:16 ahem Aug 13 11:27:38 can it be that all of OE's monotone repositories are down? Aug 13 11:27:49 you are late ;) Aug 13 11:28:33 josch|nsn: yeah, wanted to start working with it after lunch ;) Aug 13 11:29:27 there are mirrors (one was in the topic once) Aug 13 11:29:29 mtn pull is workig for me Aug 13 11:29:45 okay then it's only the web if being down Aug 13 11:29:55 Crofton|work: which server? monotone.openembedded.org Aug 13 11:30:00 yeah Aug 13 11:30:09 * thebohemian tries again Aug 13 11:31:19 Crofton|work: oh no. its probably the companies firewall .... :( Aug 13 11:31:25 arg Aug 13 11:31:35 that is actually a really annoying mtn issue Aug 13 11:31:43 which is not mtn's fault Aug 13 11:33:19 Well, there's a good solid reason why most VCS software has a HTTP transport mode. Aug 13 11:35:46 nobody get a  illume-theme-freesmartphone_git.bb failed to compil: error ? Aug 13 11:36:03 someone could point me to a way of fixing this please ? Aug 13 11:36:07 well, it was fixed now by opening the port :) Aug 13 11:36:42 they set up a new gateway here today Aug 13 11:38:58 lambda: did you inspect the log to find out what the actual error was? Aug 13 11:39:11 yes Aug 13 11:39:12 http://rafb.net/p/Wc2H1Z86.html Aug 13 11:39:23 but logs are rather light Aug 13 11:39:44 ah, the "no such file or directory" thing? Aug 13 11:39:55 so, did you verify whether those directories do in fact exist? Aug 13 12:23:31 mickey|zzZZzz: It looks like mtn is not running on amethyst Aug 13 12:23:42 verified with "pidof mtn" Aug 13 13:09:30 How is debian doing with pushing patches upstream? not so good, right? Aug 13 13:10:04 I was wondering whether we could exempt trying to push patches upstream when we have borrowed it from debian. Aug 13 13:10:48 Laibsch: Depends on the package and how the maintainer feels but Debian wants patches upstream normally. Aug 13 13:10:56 My thinking was "debian is a big and widely adopted distro, debian will push it or the upstream authors will likely recognize themselves" Aug 13 13:11:43 Laibsch: Debian kernel patches have been going upstream nicely ... Aug 13 13:12:02 OK Aug 13 13:12:05 Encouraging Aug 13 13:12:20 Most of Debian will be upstreaming patches. Aug 13 13:12:25 but of course the kernel is central and Debian has a huge interest for this stuff to go upstream Aug 13 13:12:34 good Aug 13 13:12:56 Let's exempt debian patches then and hope they make it upstream in most cases Aug 13 13:13:09 http://bugs.openembedded.net/show_bug.cgi?id=4491 Aug 13 13:13:32 4490, rather Aug 13 13:14:48 * Laibsch wishes debian had their packaging info web browsable and not only diff.gz Aug 13 13:14:59 would save me a step or two ;-) Aug 13 13:15:06 being so selfish today Aug 13 13:18:06 * * OE Bug 4494 has been RESOLVED (FIXED) by Aug 13 13:18:08 * * lzma-native_4.17.bb fails to fetch source tarball Aug 13 13:18:10 * * http://bugs.openembedded.net/show_bug.cgi?id=4494 Aug 13 14:13:19 zecke: ping Aug 13 14:13:24 zecke: you in berlin? Aug 13 14:19:06 * * OE Bug 4489 has been marked as DUPLICATE of bug 4490 by Aug 13 14:19:08 thebohemian: yes, I will see you tonight Aug 13 14:19:08 * * recode-native fails do_compile Aug 13 14:19:11 * * http://bugs.openembedded.net/show_bug.cgi?id=4489 Aug 13 14:20:50 zecke: Do you have a minute? I am thinking about the necessary changes for aptgetable-OE when going from mtn to git. I'll take the opportunity to release openembedded-openmoko while I'm at it (thanks, again!). In fact, I will use openmoko as my git testbed internally. Aug 13 14:21:27 Laibsch: maybe Aug 13 14:22:02 So far, I did have a single instance of the repo that everyone used. Aug 13 14:22:09 One central repo per branch Aug 13 14:22:38 I think I will go to one directory for OE metadata and switch to the different branches with git Aug 13 14:22:47 on a single-user system that would be all Aug 13 14:23:07 But I want to have it so that my packages can be used on a multi-user system as well Aug 13 14:23:26 zecke: at c-base I suppose :) Aug 13 14:23:47 Now, imagine user A switching to org.openembedded.dev.git branch and the next guy switching to org.openmoko.asusomething Aug 13 14:23:54 They'd step on each other's toes Aug 13 14:23:57 solution: Aug 13 14:24:08 clone to home directory Aug 13 14:24:12 problem: Aug 13 14:24:28 waste of bandwidth and HD space Aug 13 14:24:38 solution: Aug 13 14:25:00 users use git clone from local, centrally managed copy of upstream Aug 13 14:25:19 git apparently uses hard links when possible Aug 13 14:25:31 now comes the issue I have no solution for yet: Aug 13 14:26:13 zecke: hail zecke Aug 13 14:26:35 subsequent updates will apparently not use hardlinks but create new files instead Aug 13 14:27:57 zecke: I was told there were "git alternates" that I could use but that it was an advanced technique. I'm being told, you are an advanced git user. I wonder if you can possibly help me manage this alternates stuff. Aug 13 14:28:40 Laibsch: admitingly, I would start over if I run into disk issues :] Aug 13 14:28:48 Laibsch: like git-repack one repo and clone again Aug 13 14:29:03 yes Aug 13 14:29:26 I think you are aware that I try and make these deb packages zero maintenance Aug 13 14:29:54 Take the hassle and need for knowledge out of the process and package it Aug 13 14:30:03 That is sort of what I am trying to do Aug 13 14:30:19 Advanced OE devs would know what to do on their individual computer, no doubt Aug 13 14:30:46 The ones who benefit from my packages most are those that just want to build a package here and there Aug 13 14:32:09 * Laibsch looks at git-repack man page but is not 100% sure how it would fit into the situation he is facing Aug 13 14:32:34 If the alternates is what I think you mean it's about sharing the object/pack store between multiple repos. Aug 13 14:32:37 Laibsch: I think these users do not have a multi user setup? Aug 13 14:32:54 zecke: yes, it is Aug 13 14:33:18 I know at least of one example ;-) This is also my main testing ground Aug 13 14:33:19 Laibsch: are you referring to the --local or --shared arguments to git clone? Aug 13 14:33:35 * broonie suspects so. Aug 13 14:34:06 * rwhitby is also looking at separate checkout directories for separate branches for MokoMakefile, so that I can built fso, asu.testing and asu.dev all at the same time Aug 13 14:34:33 zecke: There is a kind soul that donates CPU cycles. I have aptgetable OE installed on there. Any user with an account on that machine can immediately build for OE with "oe-setup-sonkei dev;bitbake opie-image" Aug 13 14:35:03 Laibsch: I'm tending towards assuming that anyone who is building OE has enough disk space for multiple git clones. Aug 13 14:35:23 Given how well git packs stuff I'd expect that to be the case. Aug 13 14:35:38 broonie: Thank you for the explanation. I am still fairly new to git and its terminology. I tried to read up on git packs and objects, but I am still unclear as to what they are exactly. Aug 13 14:36:08 The database. Aug 13 14:36:19 Laibsch: I would say start with git-clone -l, but then don't care... actually I do that for my local branches Aug 13 14:36:20 An object is the contents of a given revision of a file. Aug 13 14:36:34 A pack is a compressed database containing multiple objects. Aug 13 14:37:19 zecke: I think you are right. That is surely safe and the least trouble. I was wondering if there is "the perfect solution". Too much of a perfectionist here ;-) Aug 13 14:37:38 broonie: Thanks Aug 13 14:37:49 Where are the packs stored? Aug 13 14:37:52 .git? Aug 13 14:37:54 yep Aug 13 14:38:14 * Laibsch takes a peek into .git/ Aug 13 14:39:15 .git/objects/pack Aug 13 14:56:07 * * OE Bug 4496 has been created by ronan(AT)aimao.org Aug 13 14:56:09 * * new recipe for python-ptrace , a debugger using ptrace Aug 13 14:56:10 * * http://bugs.openembedded.net/show_bug.cgi?id=4496 Aug 13 15:02:02 mickeyl: good morning Aug 13 15:03:07 * * OE Bug 4492 has been RESOLVED (FIXED) by Aug 13 15:03:09 * * python-pycurl requires curl 7.16.4 but 7.18.0 is in oe Aug 13 15:03:11 * * http://bugs.openembedded.net/show_bug.cgi?id=4492 Aug 13 15:04:20 fusil seems to be a good fuzzing software , i should try it on opkg :) Aug 13 15:04:43 cf http://fusil.hachoir.org/trac/wiki/CrashList Aug 13 15:07:43 pb___: hey, you still didn't get any feedback for your package change? Aug 13 15:08:09 zecke: no, it seems nobody loves me Aug 13 15:08:16 or, at least, nobody loves my changes ;-} Aug 13 15:08:25 hey pb___ Aug 13 15:08:49 zecke: still, on the bright side, there were no objections! Aug 13 15:09:16 pb___: I will reply today... I have some air for oe this week... but this means I spend creating a map of 363 authors to name and email address for git... Aug 13 15:09:24 * XorA looked, but it was fairly voodoo Aug 13 15:09:40 zecke: thanks :-) Aug 13 15:09:50 zecke: doh, is there no automated tool to do that? Aug 13 15:10:16 pb___: yeah, but it looks like I'm the automated tool :) Aug 13 15:10:22 zecke: ah, that kind of automation Aug 13 15:10:39 pb___: git-log | grep Author | sort | uniq | tr -d ': ' -f 1 > authors brought me far enough Aug 13 15:11:05 pb___: but now I have to make pb, Phillip Bundel, Blundell, Blundel map to one right name and mail address :) Aug 13 15:12:34 s/[pP].*[bB].*/Number Six/ Aug 13 15:13:09 zecke: mm, right, I see. do people really check in with that many permutations? Aug 13 15:13:26 zecke: presumably for any commits done via mtn, you can get the mapping from mtn key id to (email, name) from mickeyl's inbox. Aug 13 15:13:31 zecke: You are aware of the mapping table in the wiki? Aug 13 15:15:11 Laibsch: no Aug 13 15:15:24 Laibsch: can I git-clone it? :) Aug 13 15:15:26 It is not well populated Aug 13 15:15:41 but apparently some people started on this task Aug 13 15:15:46 look through the history Aug 13 15:16:07 http://wiki.openembedded.net/index.php/Monotone_to_Git_ID_mapping Aug 13 15:16:46 apparently, cbrake started it and IIRC he also announced it on the ml Aug 13 15:16:51 * rwhitby entered all the nslu2-linux.org entries Aug 13 15:17:46 rwhitby: ah thanks Aug 13 15:19:12 I has entered me :-D Aug 13 15:22:24 me too Aug 13 15:23:19 zecke: IIRC, all mtn accounts are real mail accounts Aug 13 15:23:31 This would simplify the task to finding the name Aug 13 15:23:57 one of the nslu2-linux.org ones was not, but I've got his real address in the table Aug 13 15:24:06 You wanna try and get this done real quick? I guess people who haven't made their entry yet won't any time soon Aug 13 15:24:08 mine isn't Aug 13 15:24:16 (he used @nslu2-linux.org but was not part of the core team so wasn't supposed to use it) Aug 13 15:24:26 mine never was Aug 13 15:24:32 I think most of the @openembedded.org addresses are not valid email addresses Aug 13 15:24:43 I think you had to fiddle with you ltg account to make it a real email Aug 13 15:24:44 and even if things were e-mail addresses tend to bitrot. Aug 13 15:24:49 I thought we changed @openembedded.org to be a real mail address some time ago? Aug 13 15:25:06 * Crofton|work does not think so Aug 13 15:25:10 OK Aug 13 15:25:26 * pb___ sends mail to pb@openembedded.org to see what happens Aug 13 15:25:39 * XorA waits for pb___ to get spammed off the net Aug 13 15:26:23 heh, you mean the @openembedded.org addresses are all honeypots? Aug 13 15:26:27 that would be a mean trick Aug 13 15:26:28 I'll enter a couple of mail addresses that I konw Aug 13 15:26:39 my @oe adress should work Aug 13 15:26:43 pb_: no, I get the real spam ;-) Aug 13 15:26:50 Laibsch: our small number of keys is easy, but --author "WhatEver" is more difficult Aug 13 15:27:06 zecke: Make a task list in the wiki Aug 13 15:27:14 I'll help you with that Aug 13 15:27:53 Laibsch: I can send you my python code... I need it in a special structure :) Aug 13 15:28:09 What do you need in a special structure? Aug 13 15:28:15 The table? Aug 13 15:28:29 Let's worry about that once it is populated and semi-complete Aug 13 15:28:39 There's more holes in it that boxes filled now Aug 13 15:28:43 than Aug 13 15:29:09 Do we absolutely need the names of people commited as --author? Aug 13 15:29:26 As a last resort, use $localpart as name Aug 13 15:29:28 Laibsch: no, I need to hook it into code Aug 13 15:29:40 it? Aug 13 15:29:44 it being the table? Aug 13 15:29:45 the table Aug 13 15:30:13 http://openembedded.org/~zecke/all_names is what I started yesterday Aug 13 15:30:14 Well, that should be rather easy to beat into shape with sed, awk and friends Aug 13 15:31:01 zecke: That list is complete? Aug 13 15:31:13 IOW, anything that needs matching? Aug 13 15:32:01 Laibsch: first Realname : ([ListOfUsedNames], EmailAddress), Aug 13 15:49:06 * * OE Bug 4496 has been RESOLVED (FIXED) by Aug 13 15:49:08 * * new recipe for python-ptrace , a debugger using ptrace Aug 13 15:49:10 * * http://bugs.openembedded.net/show_bug.cgi?id=4496 Aug 13 16:33:09 thx mickeyl :) Aug 13 16:35:09 mickeyl: I don't suppose you're aware of any problems with i2s audio on s3c24xx, right? Aug 13 16:35:32 no, sorry Aug 13 16:35:48 oh well, didn't really expect so Aug 13 16:35:54 for some reason, our stupid '2412 has started malfunctioning Aug 13 16:36:05 eeks, nasty Aug 13 16:36:16 pb___: XorA, broonie, lrg might be better candidates Aug 13 16:36:17 when configured for 16 bits per channel it seems to randomly output anything between 17 and 15 bits Aug 13 16:36:20 pb___: is audio important for your product? ;) Aug 13 16:37:02 I guess this averages out to 16 bits but that isn't quite close enough for us :-} Aug 13 16:38:19 pb___: stupid question: any clock is unstable? :) Aug 13 16:38:39 zecke: not that I can see. I have spent most of the last 2 days inspecting them and they seem to be of top quality. Aug 13 16:39:24 I had hoped that our customers didn't care about audio that much, but some of them are complaining. Aug 13 16:44:55 pb_: it might be worth having a hw expert cast an eye over the schematics, also try and reverse clock master Aug 13 16:46:12 lrg: yeah, reversing clock master is difficult due to the quirks of this dac. Aug 13 16:46:23 our best man is working on that at present but it is a bit of a pain. Aug 13 18:02:06 * * OE Bug 4497 has been created by frances.albanese(AT)gmail.com Aug 13 18:02:08 * * Bitbake fails to build libpam_0.79 Aug 13 18:02:10 * * http://bugs.openembedded.net/show_bug.cgi?id=4497 Aug 13 20:07:47 re Aug 13 20:08:06 hi florian Aug 13 20:08:42 florian do you know some online tutorial for progamming with X espcialy creating simple windows Aug 13 20:09:59 hm seems I found something Aug 13 20:10:16 woglinde: you mean X as xlib? Aug 13 20:10:20 yes Aug 13 20:12:56 woglinde: I had to use xlib for thumbpad some time ago. iirc i had a simple example and used the manpages which are quite good Aug 13 20:18:39 woglinde: hi, do you have some time? Aug 13 20:19:48 mr_nice a little bit Aug 13 20:20:18 woglinde: hey Aug 13 20:20:38 hi ant Aug 13 20:20:49 woglinde: can you apply patch from bug no 4376? Aug 13 20:22:02 * ant_ is mumbling: how to parse the cmdline in C within klibc...without rehinventing the wheel Aug 13 20:22:11 urgs Aug 13 20:22:13 --with-openssl=${STAGING_LIBDIR}/.. Aug 13 20:23:25 mr_nice is unstable okay? Aug 13 20:23:55 woglinde: yes devel branch Aug 13 20:25:19 aeh right Aug 13 20:25:32 woglinde: unfortunately I have no windows box to test it Aug 13 20:33:24 mr_nice than? why you want it integrated? Aug 13 20:36:03 woglinde: It is not my patch it is from someone else. and i think it would be nice to have it as an package in the feeds Aug 13 20:37:17 woglinde: at least it is some one who has an simpad and started to use openembedded. if the first contribution is not getting attention he may will stop contributing his simpad stuff to the bugtracker Aug 13 20:58:50 hms Aug 13 20:58:55 monotone server Aug 13 20:59:07 bah, monotone Aug 13 21:04:55 mr_nice: is there a problem? Aug 13 21:06:14 laibsch no Aug 13 21:06:16 just update Aug 13 21:06:26 you can take the stable part Aug 13 21:06:32 just the usual hydra with two heads... Aug 13 21:06:40 ah yes Aug 13 21:06:43 woglinde: I'm not talking about the code, I don't have a simpad Aug 13 21:06:44 I merged it now Aug 13 21:06:56 woglinde: thx Aug 13 21:06:57 its pushed right now Aug 13 21:06:58 I was wondering if mr_nice perceived some kind of procedural problem Aug 13 21:07:16 laibsch hm we have no automerger right? Aug 13 21:07:20 woglinde: is the automerger a cron job? Aug 13 21:07:24 ehhh Aug 13 21:07:31 dont know Aug 13 21:07:44 is / exists? Aug 13 21:07:53 Laibsch: no I have no windows box but I saw at least at the rdesktop page that the new release also fixes a securety hole Aug 13 21:09:21 I will write a mail to the stable branch mailing list for reviewing it. Aug 13 21:09:50 woglinde: I have not seen the automerger in action for a while Aug 13 21:10:08 The automerger is probably running on amethyst only Aug 13 21:10:17 Last I looked mtn was not running on amethyst Aug 13 21:10:48 Laibsch, do a ps ax | grep mtn ...... Aug 13 21:11:54 will ubi/ubifs be used in OE ? Aug 13 21:11:58 Crofton: I did a "pidof mtn earlier" Aug 13 21:12:04 Crofton: I did a "pidof mtn" earlier Aug 13 21:12:19 That was empty and I pinged mickey|sports here about it Aug 13 21:12:26 I'll take another look Aug 13 21:12:32 gitweb last commit's time: 31 hrs ago Aug 13 21:12:37 you are making a lot of assumptions using pidof Aug 13 21:13:03 Yes, you are right Aug 13 21:13:12 And indeed my assumptions where wrong Aug 13 21:13:23 Boy, I hate all this unpackaged stuff Aug 13 21:13:42 Why can we just not stick to what debian offers us? Aug 13 21:13:59 All these binaries all over the place Aug 13 21:14:34 I still can't pull from amethyst, though Aug 13 21:14:58 I've been pulling from monotone.oe.org and not having issues Aug 13 21:15:52 I'll give it another shot Aug 13 21:16:41 Laibsch: gitweb.openembedded.NET is stuck Aug 13 21:17:51 I think it might just be VERY slow today Aug 13 21:17:54 from org is ok (got recent woglinde's merge) Aug 13 21:18:04 git is, as Crofton pointed out, indeed working Aug 13 21:18:14 just VERY slow to start anything here Aug 13 21:18:34 it seems out of sync http://gitweb.openembedded.net/ Aug 13 21:18:35 I don't touch git on amethyst, so I'm sorry I cannot help you Aug 13 21:18:53 open a ticket on the tracker Aug 13 21:18:55 but you're right: amethyst seems possessed... Aug 13 21:19:02 and include RP and mickey in cc Aug 13 21:21:27 Let just pass some hours more...there are yet enough bugs! Aug 13 21:23:37 Laibsch: same 31 hours delay for http://git.linuxtogo.org/ Aug 13 21:27:26 no use telling me Aug 13 21:30:52 sorry, thought you had admin rights on ltg just to check whats going on Aug 13 21:31:03 brb Aug 13 21:38:18 grr, stupid police hovering outside my house again in their helicopter Aug 13 21:38:23 heh Aug 13 21:38:28 looking for ya? ;) Aug 13 21:38:56 let's hope not Aug 13 21:39:03 * pb_ shakes his fist in their general direction Aug 13 21:39:32 ah good, that did the trick. they're going away. Aug 13 21:41:33 That technique works for punk teenagers up to no good too Aug 13 21:46:46 ant_: I do have admin rights. That does not mean I will touch just anything Aug 13 21:46:51 unless there is an emergeny Aug 13 21:46:53 cy Aug 13 21:49:24 quick mtd-utils question ? Aug 13 21:50:19 RP: Is the following safe to do? Aug 13 21:50:55 start bitbake instance A, wait till parsing has finished Aug 13 21:50:59 make some changes that should not affect A in the meta-data Aug 13 21:51:13 start bitbake instance B (where one wants the changes to apply) Aug 13 21:52:02 I wonder if I can check out the git stuff all into a single directory, switch between branches and not worry about possibly affecting an already running build **** ENDING LOGGING AT Thu Aug 14 02:59:57 2008