**** BEGIN LOGGING AT Mon Jun 07 02:59:56 2010 Jun 07 04:54:38 Android has unique developer signatures on applications .... does openmoko based software stacks use a similar mechanism? Jun 07 04:58:07 mickeyl can u help me with the above question? Jun 07 04:58:32 Also put on at the #oe Jun 07 04:59:23 Kensan and lindi: Can you explain the signature based scheme of Android vs. openmoko Jun 07 04:59:32 lindi- ... Jun 07 05:05:28 shazkhan: i have not compiled android Jun 07 05:05:35 shazkhan: and openmoko distro is not alive anymore Jun 07 05:07:31 shr .... Jun 07 05:08:04 lindi- do we have any signature scheme on packages that pertain to the developers identity? Jun 07 05:09:12 lindi-: What about debian sw stack and packages :) Jun 07 05:54:17 shazkhan: i don't know about shr sorry Jun 07 05:54:43 shazkhan: in debian each source package is signed by the maintainer Jun 07 05:55:59 yeah .... I got some clues from #oe Jun 07 05:56:31 lindi- are u available on #oe Jun 07 05:56:45 plz have a look at it if u r. Jun 07 05:57:35 wait Jun 07 05:59:59 lindi- Debian provides a signature (gpg) on the feeds of a release and same is the case for opkg and ipkg based repos as far as I understood Jun 07 06:04:38 zub: quite common is to send that patch with [PATCHv2] prefix as reply to previous patch (git send-email will ask for "In-reply-to" message id and prefix is defined with --subject-prefix), in case you can send e-mails from your dev box now Jun 07 06:08:40 shazkhan: so? Jun 07 06:09:05 not sure Jun 07 06:22:38 lindi- Why need to sign the package when the feed can be signed too? Jun 07 06:31:10 lindi- this should mean that if a package is from a trusted repository than still we should check that it is from a trusted developer/maintainer. Two levels of trust don't make sense to me. There can be usecases but it is not suitable generically. It would make sense to sign the package if the repo is not signed, which will be in case the package is not from a standard repository. Jun 07 06:31:47 standard .... trusted repository Jun 07 06:33:14 JaMa|Wrk: thx, will try - when I find the time Jun 07 06:46:56 shazkhan: the package? are you talking about source or binary package? Jun 07 07:07:25 lindi- can be both ... depends but most of the time it is binary when we consider a distro at runtime installing packages on cell phones. Although have never come across a situation where a ditro at run will install from source package. ... Jun 07 07:12:34 shazkhan: I do not understand what you are trying to say Jun 07 07:13:25 lindi- Lets just consider binary packages. Jun 07 07:13:51 lindi- if feeds are signed and the packages are signed .... so is this sensible? Jun 07 07:13:56 shazkhan: in debian source packages (dsc) are signed by debian developers (DD). the binary packages are built automatically by a buildd and added to the archive after the developer signs the build log Jun 07 07:14:04 shazkhan: as far as I understand Jun 07 07:15:02 lindi- Is the repository signed too after the package is added to it? Jun 07 07:15:16 Got a guest .... Jun 07 07:18:23 shazkhan: lindi- is right, the source packages are signed by the developers, the package-list of the repository by the debian archive key Jun 07 07:18:57 got it Jun 07 07:19:01 iirc DDs also build and sign a binary package but it is thrown away and not included to the archive Jun 07 07:19:27 lindi- Why? Jun 07 07:20:45 shazkhan: 'Given that other distros experience has shown that allowing source only uploads results in a huge loss of quality checks and an increased load on the buildds from packages FTBFSing everywhere, some members of the team believe that source+binary uploads should happen as currently, but that the maintainer built binaries should be rebuilt by the buildds (i.e. be thrown away at accept time).' Jun 07 07:20:48 because the ftpmasters can only read code, not binary Jun 07 07:20:51 shazkhan: -- http://lists.debian.org/debian-devel-announce/2009/11/msg00001.html Jun 07 10:12:22 morning Jun 07 10:13:22 moin Jun 07 11:43:57 interlaced graphics to get round slow bus.. interesting idea Jun 07 12:22:26 mickeyl: hi, did you heard about Linaro? Jun 07 12:26:02 mickeyl: perhaps, they could be interested in FSO framework Jun 07 13:02:04 Martix_: just heard about that on weekend Jun 07 13:02:11 Martix_: i'll drop them a mail Jun 07 13:02:22 ok Jun 07 13:03:08 mickey|office: you can ping hrw too, he is part of Foundations team Jun 07 13:03:27 hm... just got no resume after suspend... no reaction on power button at all Jun 07 13:03:43 Q-Master: which kernel? Jun 07 13:03:50 2.6.29 Jun 07 13:04:09 JaMa|Wrk: ah, so that's the ubuntu thing? Jun 07 13:04:22 battery eject/insert helped Jun 07 13:05:06 damn Jun 07 13:05:15 something drained the accums to 0 Jun 07 13:05:37 wtf??? Jun 07 13:06:15 mickey|office: cannoniccal is not listed in first announcement (IIRC), but seems like lots of ubuntu arm platform people are involved in linaro now and 2nd meeting is about linaro+ubuntu.. Jun 07 13:07:01 mickey|office: and that 2nd meeting is here in Prague.. so I hope they will invite you and then we can go for a beer :) Jun 07 13:13:33 JaMa|Wrk: iirc, the idea of linaro comes from shuttleworth. I may be wrong but I believe linaro is more or less a way to impose ubuntu on arm Jun 07 13:17:10 linaro aims at kernel and middleware for all linux based platforms including MeeGo Jun 07 13:18:01 maybe Android as well ... I think I saw it on their website. Jun 07 13:19:15 shazkhan: when reading such thing : http://wiki.linaro.org/Events/LinaroPlatformSprint072010 it's hard to believe it's not ubuntu based Jun 07 13:19:37 shazkhan: and as JaMa|Wrk said, a lot of ubuntu arm people are in linaro's teams Jun 07 13:19:52 rtp: then I would expect Canonical listed as one of founding corporations.. Jun 07 13:21:43 JaMa|Wrk: they may be part of but not pay anything, thus not being listed I gues Jun 07 13:21:46 s Jun 07 13:24:24 rtp: as long as they pay those devs.. :) Jun 07 13:24:56 :) Jun 07 16:29:49 hi mickeyl Jun 07 16:30:17 do you think it's a good idea to make a recipe for compiling the kernel out of tree with oe so I can develop it faster? Jun 07 16:30:39 so it does mkbootimg etc...with just a bitbake Jun 07 16:40:50 GNUtoo|laptop: i compiled my kernel with OE toolchain without any issues Jun 07 16:41:02 GNUtoo|laptop: just export CROSS_COMPILE and you're done. Jun 07 16:41:36 GNUtoo|laptop: you can borrow the script that does mkimage and tars a special modules archive from andy-tracking branch. Jun 07 16:42:22 GNUtoo|laptop: going through bitbake autorev process every time you attempt to compile a kernel is obviously a pita. Jun 07 16:53:03 JaMa: do you have a moment? Jun 07 16:53:14 sure Jun 07 16:53:21 do you have some patch for test Jun 07 16:53:23 ? Jun 07 16:53:58 well Jun 07 16:54:03 not yet Jun 07 17:04:34 larsc: should I try something? Jun 07 17:05:53 * JaMa didn't want to frighten away larsc by patch request... Jun 07 17:09:06 JaMa: http://pastebin.com/zvxAGCXm Jun 07 17:10:34 larsc: + that debug from last time, right? Jun 07 17:12:25 JaMa: no. just om-gta02-2.6.34 and that patch Jun 07 17:12:44 ok **** BEGIN LOGGING AT Mon Jun 07 18:28:14 2010 Jun 07 21:45:59 larsc: resume seems fast again, dmesg: http://pastebin.ca/1878703, patch used: http://paste.pocoo.org/show/223035/ Jun 07 21:48:42 JaMa: hm. it looks as if pcf50633_irq_{suspend,resume} are not called Jun 07 21:49:51 yes.. thats probably also the reason I've never seen those printks from previous debug patch Jun 07 21:54:44 i wonder why Jun 07 21:59:48 could 54067ee20645a4ee12a9546aeb3b048b4c44cf60 be the cause? Jun 07 22:01:21 CD L Jun 07 22:01:26 whoops :) Jun 07 22:02:22 and 2f60ba706bd9af84c4eab704243b262e69556f2e is the fix Jun 07 22:03:34 the fix is only local? Jun 07 22:03:40 cannot see it in git.om repo Jun 07 22:03:50 or upstream git? Jun 07 22:03:57 upstream Jun 07 22:04:00 its in 2.6.35 Jun 07 22:08:40 good, looking forward for backport :) Jun 07 22:09:21 btw is there any way to get ts filtering back to kernel, or really only possible way is in userspace? Jun 07 22:09:53 JaMa: http://pastebin.com/EjqavHsJ Jun 07 22:09:56 should work aswell Jun 07 22:10:04 so you don't need to packport that patch Jun 07 22:10:26 larsc: I've already reported false bug in EFL when some components previously responding to single-click needed now double-click (single was usually handled as really small drag) Jun 07 22:11:28 well, the upstream kernel folks don't want filtering in kernel. Jun 07 22:11:49 and i don't want to maintain a local patchset either Jun 07 22:13:02 ok, I've seen some evfilter project.. so we just need to adapt it somehow and integrate it in our images Jun 07 22:16:21 or tslib Jun 07 22:40:42 larsc: http://paste.pocoo.org/show/223057/ Jun 07 22:42:16 larsc: s/client/dev? Jun 07 22:44:03 will test tomorrow morning.. Jun 07 22:44:18 larsc: thanks! Jun 07 22:48:08 larsc: also s/PC50633_PM_OPS/PCF50633_PM_OPS/g (missing F) Jun 07 23:03:19 freesmartphone.org: 03Frederik.Sdun 07vala-dbus-binding-tool * r4630ec41d7cb 10/Makefile.am: Jun 07 23:03:19 freesmartphone.org: Add config.vapi to EXTRA_DIST Jun 07 23:03:19 freesmartphone.org: Based on a patch by David Wagner **** ENDING LOGGING AT Tue Jun 08 02:59:56 2010