[00:00] uds-gb-h: This session has ended. === ChanServ changed the topic of #ubuntu-uds-grand-ballroom-h to: Track: | What exactly is juju anyway? | Url: http://summit.ubuntu.com/uds-q/meeting/20754/what-exactly-is-juju-anyway/ | Audio: http://icecast.ubuntu.com:8000/grand-ballroom-h.ogg.m3u [00:08] o/ [00:10] everybody knows Al Gore invented the cloud [00:21] _________________________________________ [00:21] / I typed juju deploy wordpress and all I \ [00:21] \ got was this burning sheep / [00:21] ----------------------------------------- [00:21] \ . . . [00:22] \ . . . ` , [00:22] \ .; . : .' : : : . [00:22] \ i..`: i` i.i.,i i . [00:22] \ `,--.|i |i|ii|ii|i: [00:22] UooU\.'@@@@@@`.||' [00:22] \__/(@@@@@@@@@@)' [00:22] (@@@@@@@@) [00:22] `YY~~~~YY' [00:22] || || [00:24] chroot + network and pid namespacing [00:30] o/ [00:30] juju set mysql dataset-size=xxxG tunes it for your dataset [00:31] juju set mysql tuning-level=fast <--- removes the brakes from the car [00:33] the charm *could* lock to a single version [00:34] IMO, charms *should* lock to a version in most cases, but this is not a universally held belief. :) [00:35] remove-unit always removes an explicit unit [00:37] common feature request: juju remove-unit x/4 --terminate-machine [00:42] 8 charms for core openstack [00:43] mysql, rabbit, nova-cc, nova-compute, glance, swift, swift proxy, keystone [00:43] Suggests: apt-cacher-ng, lxc, libvirt-bin, zookeeper [00:43] Can you use juju on trystack.org? [00:43] zookeeper == java [00:44] txwikinger: that should work if trystack has S3 exposed [00:44] I have not figured that out yet [00:44] how do I configure juju for it? [00:45] There is but its haacky [00:46] yes, but you won't be able to use juju after that [00:46] no you can't shut it down because its address will change [00:46] each node gets the address of node 0 stored as a configuration [00:47] thats just an implementation detail .. we could make the bootstrap node discoverable by other means. [00:47] \o/ subway good! [00:52] I am! [00:53] unless I dreamed it ;) [00:53] yes, we'd have to put that charm in a different charm store though [00:54] uds-gb-h: 5 minutes left in this session! [00:54] If you guys come up with a charm that can only run on Solaris or BSD or something.. come talk to us, I want to see that and will help get that integrated w/ the charm store. :) [00:55] s3 is not properly abstracted for charms yet [00:55] planned feature for the future [00:55] uds-gb-h: 4 minutes left in this session! [00:55] it does have the abstraction in the provider [00:55] but there's no charm tool to access it [00:56] official images only please! [00:56] uds-gb-h: 3 minutes left in this session! [00:57] uds-gb-h: 2 minutes left in this session! [00:57] .sssssssss. [00:57] .sssssssssssssssssss [00:57] sssssssssssssssssssssssss [00:57] ssssssssssssssssssssssssssss [00:57] @@sssssssssssssssssssssss@ss [00:57] |s@@@@sssssssssssssss@@@@s|s [00:58] _______|sssss@@@@@sssss@@@@@sssss|s [00:58] / sssssssss@sssss@sssssssss|s [00:58] / .------+.ssssssss@sssss@ssssssss.| [00:58] / / |...sssssss@sss@sssssss...| [00:58] | | |.......sss@sss@ssss......| [00:58] | | |..........s@ss@sss.......| [00:58] | | |...........@ss@..........| [00:58] \ \ |............ss@..........| [00:58] \ '------+...........ss@...........| [00:58] \________ .........................| [00:58] |.........................| [00:58] uds-gb-h: 1 minute left in this session! [00:58] /...........................\ [00:58] |.............................| [00:58] |.......................| [00:58] |...............| [00:58] Cheers folks! [00:59] uds-gb-h: This session has ended. === ChanServ changed the topic of #ubuntu-uds-grand-ballroom-h to: Currently no events are active in this room - http://summit.ubuntu.com/uds-q/grand-ballroom-h/ - http://ubottu.com/uds-logs/%23ubuntu-uds-grand-ballroom-h.log === daker is now known as daker_ === ChanServ changed the topic of #ubuntu-uds-grand-ballroom-h to: Track: Foundations | Detail and begin the arm64/aarch64 port in Ubuntu | Url: http://summit.ubuntu.com/uds-q/meeting/20330/foundations-q-aarch64-porting/ | Audio: http://icecast.ubuntu.com:8000/grand-ballroom-h.ogg.m3u [16:02] hello everyone :) [16:02] hey folks [16:02] 你好 [16:02] come closer to microphones, folk :) [16:02] yes [16:03] better [16:03] better yes [16:03] still not hearing any 64-bit long words :) [16:03] SteveMcIntyre: are you able to do a hard ping on wookey? [16:03] :-) [16:03] rsalveti: I'll go look for him [16:04] hello [16:05] schedule moved :-) [16:06] I can't hear yet [16:06] wookey_: they want to know if the port is done yet? [16:07] Not quite :-) [16:07] wookey_ had headphones on, dunno if he was actually connected to the audio [16:07] wookey_: So, Friday then? [16:07] OK. I'm here [16:07] wookey_: see the audio link in the topic! [16:08] can anyone share the link? [16:08] http://people.linaro.org/~wookey/buildd/precise/sbuild-ma/status.html [16:08] people.linaro.org/~wookey/buildd/quantal/sbuild-ma/status.html [16:08] thanks [16:08] status is very broken right now [16:08] as libc doesn't seem to want to install [16:09] yes we do [16:09] sorry arm does, linaro doesn't [16:09] And that toolchain is hacked from a bninary build [16:09] sorry what ricardo? [16:09] it builds, but a lot of the patches need fixing for newer gcc [16:10] wookey_: the results of the cross buildd based on the new toolchain from arm [16:10] can we get agreement about the multiarch triplet now? [16:10] doko: heh :-) [16:12] correct, no qemu for a while [16:12] there will be "issues" with adding support there [16:12] who's providing the QEMU? [16:12] no qemu means we need different cross-fixes for some things than on a qemu-having arch [16:12] yes [16:13] not at the moment, no [16:13] is there some plan to work on qemu? [16:13] main blockers are multiarchy deps for perl, python [16:14] suihkulokki: not that I know of, yet [16:14] suihkulokki: we'll have at linaro, but I believe it's probably something to start at Q4 [16:14] arm64 is favourite, I think [16:14] * SteveMcIntyre nods infinity [16:14] wookey nods too :-) [16:15] wookey_, multiarchy deps? [16:15] arm is using multiarch paths for linker and libs on internal toolchain, they are changing to linker in /lib and libs in /lib64 [16:15] SteveMcIntyre, favourite for what? gnu triplet, multiarch triplet? [16:15] doko: for the Debian/Ubuntu arch name [16:16] cjwatson: it's still an open question [16:16] colin we need to argue upstream about that if we care [16:16] SteveMcIntyre, and the gnu triplet is already defined? [16:16] the upstream glibc folks will whinge *again* if we try to push m-a [16:16] gnu triplet is already decided and upstream (aarch64-linux-gnu ) [16:16] doko: yes [16:16] doko: aarch64-linux-gnu [16:17] and this should be used foe the multiarch triplet too? looks a bit "generic" [16:17] what happens on an amd64/aarch64 machine? what is /lib64 symlinked to? [16:17] the clear thing we have at least from the armhf mess is that we can have ld.so in /lib [16:17] as /lib/ld-linux-aarch64.so.X IIRC? [16:17] ramana: ? [16:18] SteveMcIntyre, the GNU triplet is already upstream isn't it ? [16:18] ah. OK [16:18] ramana: yes, the triplet is [16:18] ramana: have we specced the linker name? [16:18] we have a lot of cross-build fixes already [16:18] but there's issues with getting them pushed upstream from within ARM :-( [16:19] I have a big list of fixes and effort estimates [16:19] 10% is aarch64. 50% is multiarch deps, 40% is assorted crossbuild [16:19] SteveMcIntyre, probably worth talking to marcus about. I haven't followed all those developments recently. [16:19] cjwatson: yes, agreed [16:20] this is a reason why wookey_ has got involved [16:20] infinity yes, the prblem is that the best set of patches is ARMs and they won't let it out directly yet [16:20] but I can still give you a summary [16:20] Peter was very good at coming up with patches, but not pushing them upstream [16:20] so a number of them rotted :-/ [16:21] And what happened then was that the packaging moved on so they became irrelevant or unmergeable [16:21] yes [16:21] There were several cases near the end of precise where there was a patch somewhere but it didn't make sense any more so I just did it again from scratch [16:21] ack [16:21] Where is the right venue to disucss crossfixes? [16:21] doko: you're quite quiet? [16:22] SteveMcIntyre, as always ;) [16:22] I sometimes need to ask and it's not obvious where... [16:22] any regular Ubuntu development channels are fine [16:22] #ubuntu-devel etc. [16:22] many of the old cross-building patches are not quite valid in the multiarch world anymore [16:22] SteveMcIntyre, just want to differentiate between cross build failure and package-needs-porting [16:22] yeah [16:22] Right [16:23] the list is natty-based yes, but modified to include all that are fixed upstream [16:23] and some pat bugs [16:23] apt [16:23] some are still OK [16:24] but yes some are mostly useless [16:24] then there's still some *nasty* ones [16:24] perl \o/ [16:24] cross-building perl isn't that hard, but making it multiarch-useful is [16:24] ditto python [16:24] ack [16:24] because a lot of that is defining rules for extensions [16:25] (where by "isn't that hard" I mean relative to the usual difficulty of building perl and python :-) ) [16:25] I am maintaining public status of issues here: https://wiki.linaro.org/Platform/DevPlatform/CrossCompile/MultiarchCrossBuildStatus [16:25] working on python3.3 cross buildability. one more reason to switch to 3.3 ;) [16:25] cjwatson: *grin* [16:25] who's the perl maint in Ubuntu? [16:25] we keep it synced from Debian [16:25] I had a 5.12.x patch that worked, but would need m-a tweaks [16:25] SteveMcIntyre, we only have a last upoader ... [16:25] so we don't have one [16:25] (by hand, so often a few days out of date) [16:25] then 5.14.x happened before it was accepted [16:25] I'm happy to act as a stunt perl maintainer [16:26] we'd need to work on 5.14.x and 5.16.x and get upstream to play too, ideally [16:26] I guess this cross-build perl/python modules is a thing to do at linaro connect [16:26] cjwatson: well volunteered :-) [16:26] * doko turns on the camera ... [16:26] agreed [16:26] so we can move back to aarch64 specific issues here [16:26] The automatic build will be definitive [16:27] but it's good for things like 'this apt bug breaksa thes 8 builds' [16:28] We need agreement on the staged build/bootstrap stuff too [16:28] to get it into dpkg/apt [16:28] I'm not sure we do [16:28] That's good for automatic bootstrapping, but it wouldn't be the end of the world to have some manual work involved in a handful of packages [16:29] it's still massively better than anything we had before [16:29] true - depends if you want to bootstrap more than once [16:29] which you do until there is some hardware... [16:29] compiler bugs is one thing yes [16:29] wookey_, well, the final bootstrap will be a native one anyway (the one where packages hit the archive) [16:30] true [16:30] they're different, but the staged bootstrap helps both ways [16:30] arm want to keep bootstapping due to toolchain issues [16:30] cjwatson: yes [16:31] we just need it to be pushed and working [16:31] cjwatson: yes, but dpkg people want some evidence we are agreed [16:31] and then fix the issues when they arrive [16:31] before putting it in [16:31] wookey_: who? [16:31] There are bugs filed: [16:32] http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=661538 [16:32] Debian bug 661538 in dpkg-dev "Add support for Build-Depends-Stage1 in order to allow breaking of cyclic Build-depends loops" [Wishlist,Open] [16:32] http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=661537 [16:32] Debian bug 661537 in apt "Support for staged/bootstrap builds to break cyclic build-dependencies" [Wishlist,Open] [16:32] so eveyone pile in and say 'yes it's a good idea' [16:33] * SteveMcIntyre nods infinity again [16:33] it's up to the people *using* this to say what's needed [16:33] OK [16:33] wookey_: any other patches? [16:34] e.g. sbuild and friends? [16:34] OK I have a pyhton-debian patch [16:34] and there is an xdeb [16:34] just to recognise extra fields IIRC [16:34] and python-apt? [16:35] well yes xdeb is mostly supeceded [16:35] but there is a patch for people still using it [16:35] didn't catch that...? [16:35] On toolchains we need to do gcc/glibc/libffi/binutils/gdb packaging [16:35] That's #669250? [16:36] cjwatson: yes [16:36] we have a kernel for aarch64 inside ARM, yes [16:36] review is ongoing, will go public soon [16:37] pass :-) [16:37] I did it once for maverick-vintage toolchain [16:37] rsalveti: yes [16:37] arm lawyers are huge impediment to getting anything done [16:38] an 'IP based business' finds contributing very difficult [16:39] I have a half-arsed patch for adding aarch64 to toolchain packaging [16:39] it's all gone quiet... [16:39] should I just send it to doko to update for current? [16:39] Is someone going to add arm64/aarch64 line to dpkg? [16:40] multi-assed? [16:40] hrw: OK [16:40] it's a one-line patch :-) [16:41] cjwatson: it's alkmim's first patch so may not be perfect. It looked approx OK to me. [16:41] yep, I shall ponder it [16:41] SteveMcIntyre, could you add the multiarch triplet to http://wiki.debian.org/Multiarch/Tuples ? [16:41] it obviously requires patched dpkg [16:41] doko: yes, will do [16:41] please add an action for that for me [16:42] So are we bootstrapping in debian or ubuntu first? [16:42] wookey_: yes [16:42] [16:42] infinity: I'm doing it *now* already [16:43] planning to get it working this cycle? [16:43] we'll end up bootstrapping a few times anyway, I expect [16:44] wookey_: yup :-) [16:44] releasing before hw will be hard [16:44] infinity: compared to older models, they *are* fast [16:44] believe me! [16:45] imagine a few tens of MHz type of level [16:45] fast enough to run things, but not something you'd want to build on it... [16:45] Oh yes - There is an alternate reduced packge list: http://people.linaro.org/~wookey/buildd/quantal/sbuild-ma/status-bootstrap.html [16:45] at least, that's my impression from others [16:46] not played directly yet [16:46] which is basically the current targeteed package set [16:46] can someone take a look and see why _no_ build-deps work currently [16:46] that's just what arm used [16:46] infinity: the list we started with was *more* than bootstrap to start with [16:46] becaasue they wanted apche to demo [16:46] people wanted to demo a simple LAMP stack [16:47] click on any build-dep link there [16:47] then go down to the bottom [16:47] we can fix about 100 builds if that is unbunged :-) [16:48] libc6-dev is m-a:same [16:49] you need a cross-build-essential [16:50] it makes sense, yes [16:50] yes, someone needs to do that [16:50] I think ThibG was going to [16:50] sbuild has it's own internal list for now [16:50] uds-gb-h: 5 minutes left in this session! [16:50] but it has to chage from debian to ubuntu [16:51] hell yes [16:51] infinity: you at Connect too? [16:51] yay [16:51] uds-gb-h: 4 minutes left in this session! [16:52] :-) [16:52] anything else? [16:52] all done? [16:52] uds-gb-h: 3 minutes left in this session! [16:52] agreed [16:52] doko: did you already fix M-A python? [16:52] so we just have to try it? [16:53] or is there more to do? [16:53] wookey_, it should be, but I really didn't test anything with it [16:53] uds-gb-h: 2 minutes left in this session! [16:53] well quantal if armhf [16:54] precise was all armel [16:54] right, time to bail [16:54] have fun folks [16:54] uds-gb-h: 1 minute left in this session! [16:54] Well, glad to hear you are all going to fix cross stuff :-) [16:54] thanks! [16:55] I look forward to it [16:55] uds-gb-h: This session has ended. [16:56] OK, cheers colin === ChanServ changed the topic of #ubuntu-uds-grand-ballroom-h to: Track: Community | Continuing Packaging Guide Improvements | Url: http://summit.ubuntu.com/uds-q/meeting/20682/community-q-packaging-guide/ | Audio: http://icecast.ubuntu.com:8000/grand-ballroom-h.ogg.m3u [16:57] good - not just me then :-) [17:11] I knew of the packaging guide on the wiki, but not this new one [17:15] I will more then willing help anywhere that I can just send a email.But I dont know if I am the best person but I def can help [17:15] josephjamesmills [17:16] https://launchpad.net/~josephjamesmills [17:16] np thank you :) [17:23] new packaging videos ? [17:25] Update the old packaging videos (which were done with Hardy?), so they're more current for Precise/Quantal [17:28] is there anyway that you can turn up the mic ? [17:31] How about this idea, get Kahn Academy to do a session on creating and building a package? [17:31] call for translators could be a milestone [17:32] what did you ask ? [17:32] I can here [17:32] hear [17:33] I forgot [17:39] uds-gb-h: 5 minutes left in this session! [17:40] Promise them cake [17:40] (WIKI)I would like too bring up that there is a team for ubuntu-wiki that is new I will tell them about this ? [17:40] uds-gb-h: 4 minutes left in this session! [17:41] uds-gb-h: 3 minutes left in this session! [17:42] uds-gb-h: 2 minutes left in this session! [17:43] I wrote code to make html too moinmoin [17:43] uds-gb-h: 1 minute left in this session! [17:44] uds-gb-h: This session has ended. [17:45] quiet UDSbot === ChanServ changed the topic of #ubuntu-uds-grand-ballroom-h to: Currently no events are active in this room - http://summit.ubuntu.com/uds-q/grand-ballroom-h/ - http://ubottu.com/uds-logs/%23ubuntu-uds-grand-ballroom-h.log === ChanServ changed the topic of #ubuntu-uds-grand-ballroom-h to: Track: | Lubuntu work items for Q | Url: http://summit.ubuntu.com/uds-q/meeting/20575/lubuntu-q-work-items/ | Audio: http://icecast.ubuntu.com:8000/grand-ballroom-h.ogg.m3u [17:52] Thanks !! [18:03] I'm waiting 5 min to close the door and to start the session :) [18:08] hum, as usual, no many people in the room, but the chan is also quiet :-/ [18:08] \o [18:08] hey Unit193 :) [18:09] I feel less alone :) [18:10] hey gilir [18:10] hi IAmNotThatGuy :) [18:11] hey! [18:12] Hello valdur55 [18:12] do you want me to start talking, or could we cancel the session ? not sure it's really useful with the low number of people :( [18:13] please keep it :) [18:14] gilir, Important action items can be discussed at this time. People can look at the logs later I believe [18:15] Yep. [18:15] the etherpad : http://summit.ubuntu.com/uds-q/meeting/20575/lubuntu-q-work-items/ [18:23] That's the one I switched out since I needed a timed option. [18:31] if this is the right time: i think that "Menu" should get more futures. i mean that it should get two looks : one "simple" (as it is now) and other more complet but light (like in winXP) [18:31] i mean "Menu" the applet in lxpanel [18:31] Along the lines of menu, there was some sort of "LXmed" menu editor, but not in the repo. [18:38] I use Synapse semantic launcher for launching programs :) [18:41] i think "audacious" should have a more avanced applet for lxpanel if it will not use more Mo [18:45] Translations of the site/wiki? [18:47] Xfce 4.10 upgraded Application Finder. : http://xfce.org/about/tour . Maybe should have lxde desktop same runner? [18:50] uds-gb-h: 5 minutes left in this session! [18:51] uds-gb-h: 4 minutes left in this session! [18:52] uds-gb-h: 3 minutes left in this session! [18:53] i'm not sure, but installing an other session like gnome-shell or unity with lubuntu will make lubuntu-session more slow (use more memory because some daemons of others sessions open in the lubuntu-session) [18:53] uds-gb-h: 2 minutes left in this session! [18:53] or not? [18:54] uds-gb-h: 1 minute left in this session! [18:55] uds-gb-h: This session has ended. === ChanServ changed the topic of #ubuntu-uds-grand-ballroom-h to: Track: Desktop | Supporting USB video devices | Url: http://summit.ubuntu.com/uds-q/meeting/20574/desktop-q-xorg-usb-video-support/ | Audio: http://icecast.ubuntu.com:8000/grand-ballroom-h.ogg.m3u [19:46] i have come across some usb mimo monitors with integrated touchpad. the device is basically a usb hub that connects to a usb hid input device for the touch and a display link screen.. the hard part are how to associate the touchscreen orientation with the usb screen [19:46] can we support these kind of usb screens? and make the touch work? [19:46] hello [19:46] !!! [19:47] have a nice lunch [19:54] uds-gb-h: 5 minutes left in this session! [19:55] uds-gb-h: 4 minutes left in this session! [19:56] uds-gb-h: 3 minutes left in this session! [19:57] uds-gb-h: 2 minutes left in this session! [19:58] uds-gb-h: 1 minute left in this session! [19:59] uds-gb-h: This session has ended. === ChanServ changed the topic of #ubuntu-uds-grand-ballroom-h to: Currently no events are active in this room - http://summit.ubuntu.com/uds-q/grand-ballroom-h/ - http://ubottu.com/uds-logs/%23ubuntu-uds-grand-ballroom-h.log === ChanServ changed the topic of #ubuntu-uds-grand-ballroom-h to: Track: Hardware | Ubuntu Kernel Delta Review | Url: http://summit.ubuntu.com/uds-q/meeting/20341/hardware-q-kernel-delta-review/ | Audio: http://icecast.ubuntu.com:8000/grand-ballroom-h.ogg.m3u [22:54] uds-gb-h: 5 minutes left in this session! [22:55] uds-gb-h: 4 minutes left in this session! [22:56] uds-gb-h: 3 minutes left in this session! [22:57] uds-gb-h: 2 minutes left in this session! [22:58] uds-gb-h: 1 minute left in this session! [22:59] uds-gb-h: This session has ended. === ChanServ changed the topic of #ubuntu-uds-grand-ballroom-h to: Currently no events are active in this room - http://summit.ubuntu.com/uds-q/grand-ballroom-h/ - http://ubottu.com/uds-logs/%23ubuntu-uds-grand-ballroom-h.log === bobweaver is now known as josephmills