20:00:46 #startmeeting 20:00:46 Meeting started Wed Jun 13 20:00:46 2012 UTC. The chair is pwhalen. Information about MeetBot at http://wiki.debian.org/MeetBot. 20:00:46 Useful Commands: #action #agreed #halp #info #idea #link #topic. 20:00:46 #chair pwhalen jonmasters bconoboy ctyler pbrobinson 20:00:46 Current chairs: bconoboy ctyler jonmasters pbrobinson pwhalen 20:00:46 Anyone? 20:00:52 .fas jcapik 20:00:52 * pwhalen is here 20:00:52 jcapik: jcapik 'Jaromír Cápík' 20:00:59 .fas ctyler 20:00:59 ctyler: ctyler2621 'Christopher Tyler' - ctyler 'Chris Tyler' 20:01:01 .fas pwhalen 20:01:01 pwhalen: pwhalen 'Paul Whalen' 20:01:09 * maxam is here 20:01:12 buenas dias amigos 20:01:15 * bconoboy is here 20:01:19 * agreene is here 20:01:21 .fas blc@redhat.com 20:01:21 bconoboy: blc '' 20:01:22 * Frojoe is here 20:01:31 * satellit__ listening 20:01:33 .fas maxamaxim 20:01:33 maxam: maxamaxim '' 20:02:11 export LANG=en_US.UTF-8 20:02:13 * tinker-f595_ tinker-f595_ 20:02:31 ctyler: but i use es_US.UTF8 20:02:32 let's get started then 20:02:40 #topic 1) F17 GA - remaining blockers? 20:03:10 pwhalen: whats the blocker tracker bug number? 20:03:38 Last week we decided to make the grubby patch and resizefs scripts integrated into proper fedora packages. That is now the case, but the packages haven't been built on arm.koji 20:04:21 * jonmasters is in, sorry for slight delay 20:04:36 We need the following in f17 arm: 20:04:38 grubby-8.12-2.fc17 20:04:38 rootfs-resize-0.3-2.fc17 20:04:46 bconoboy: i think the resizer is built but grubby is not 20:04:57 grubby-8.12-2.fc17 includes my patch 20:05:06 (or says it does in the changelog) 20:05:32 The resizer is only built for f18 20:06:17 ok, need it built for f17 as well then 20:06:39 I couldn't submit for f17, it told me it wasn't in "the list" 20:06:52 But I did do a scratch build. 20:07:00 ctyler: it woulndt have been in the f18 list either 20:07:45 any other blockers? 20:07:55 ctyler: that's strange 20:08:06 ctyler: you asked for f17 .... you should be able to submit for f17 20:08:34 jcapik: its not strange when you know the processes used to add it 20:08:40 the f17 branch was created https://admin.fedoraproject.org/pkgdb/acls/name/rootfs-resize 20:08:43 If there aren't any other blockers I'd like to propose a test day at the end of this week. 20:09:02 =Friday, I presume 20:09:04 sounds good 20:09:10 jcapik: the syncing of oweners is done at the lowest tag supported, so dist-f15 right now. 20:09:14 Basically, get the 2 packages built, integrate their use into the rootfs resizer, and do a testathon friday. 20:09:20 If it's good, ship it. 20:09:32 jcapik: its not fully automated for secondary arches 20:09:52 bconoboy: we need to ensure that no updates are included in the images 20:09:56 dgilmore: the package is noarch 20:10:09 dgilmore: Ack. 20:10:09 bconoboy: because of GPL compliance 20:10:14 jcapik: doesnt matter 20:10:16 (GPL!?) 20:10:31 bconoboy: yes, we distribute the distro under GPL 20:10:50 bconoboy: and we say the sources can be found in the source tree 20:10:57 its not true if there is updates 20:10:59 dgilmore: Okay. That's interesting. I will make the images without reference to fedora-updates. I will rely on you to make the right packages part of the base fedora package set. 20:11:01 +1 to test day on Friday 20:11:06 dgilmore: I really don't understand .... we're submitting that for primary arches 20:11:28 dgilmore: and I know the process .... I've done that many times 20:11:35 bconoboy: we need updates enabled for use after installation. we can not have it in use during composition 20:11:57 dgilmore: Sure, this will just be a yum --disablerepo sort of change when installnig stuff. 20:12:04 cool 20:12:04 I won't touch the yum.repos.d dir. 20:12:14 that's about right 20:12:47 #idea VFAD on Friday June 15th at 12pm EST for testing F17 20:13:02 +1 20:13:14 can we make it a little later? 20:13:21 join when free? 20:13:26 sure 20:13:31 +1 20:14:22 #agreed GA Final Candidate testing to take place at VFAD on Friday June 15th at 12pm EDT 20:14:52 between now and then we'll need to have the grubby and resize issue resolved. Let's handle that in #fedora-arm rather than today's meeting. 20:15:52 pwhalen: What's next? 20:15:55 any other foreseen blockers? 20:16:07 #topic 2) Images to be shipped with F17 20:16:29 pwhalen: only other potential issue is the older packages 20:17:13 dgilmore: Is this a GPL issue or a policy issue? 20:17:26 bconoboy: policy 20:17:40 in that we should ship the same or newer than primary 20:17:59 I'd like to bend that policy for F17. Just F17. 20:18:02 where newer are for blocker bugfixes 20:18:20 bconoboy: it is less than 100 packages 20:18:38 dgilmore: There is a reason koji-shadow hasn't built them. It could take weeks to sort out. 20:19:31 http://142.204.133.82/jon/koji/kc.17.diff.html according to that page it is 80 srpms 20:19:38 bconoboy: sure 20:19:53 just making sure we still keep it in mind 20:20:27 Okay. Let's make this a priority for F18. 20:20:29 we're close in F17, we'll get it nailed in F18 20:20:47 dgilmore: Can you take ownership of tracking this down? 20:21:13 (IE, post -F17 release) 20:21:29 bconoboy: sure 20:21:56 #task dgilmore to track down older packages 20:22:07 actually a meeting owner will need to do that 20:22:11 #topic 2) Images to be shipped with F17 20:22:31 #task dgilmore to track down older packages 20:22:32 I have an opinion about this. 20:23:04 We will ship those images that pass validation testing on Friday. It needs to minimally work with Panda and Vexpress, our original requirements. 20:23:19 If other images also pass testing, we'll include them. 20:23:54 (I expect this means we will also work with trimslice, imx serial, kirkwood and possibly highbank) 20:24:20 dgilmore: Do we need the 3.4.2 kernel? 20:24:21 bconoboy: so we have vexpress, imx, pandaboard, beagleboard, kirkwood, tegra, highbank? 20:24:41 bconoboy: probably, as it builds in rtc for arm boards 20:24:41 and highbank will target only the model 20:24:43 dgilmore: Yes, that's the widest net I would propose casting 20:24:49 (for now) 20:24:57 bconoboy: so if we want time set on boot right we really need it 20:25:05 bconoboy: but it could be an update 20:25:16 bconoboy: it doesnt prevent working installs 20:25:22 dgilmore: Okay, I'm happy either way if our minimum kernel version is 3.4.1-2 20:25:22 and can be fixed via an update 20:25:46 bconoboy: i think we should go with 3.4.1-2 as that works on phyical highbank hardware 20:26:10 bconoboy: what do the validation tests involve 20:26:11 ok 20:26:28 tinker-f595_: There is a wiki. Pwhalen, do you have a link handy? 20:26:47 (or ctyler) 20:26:54 * jonmasters is on the phone 20:27:23 http://fedoraproject.org/wiki/Architectures/ARM/Quality_Assurance/Beta_Release_Criteria 20:27:26 I can do smoe tests after 7pm PDT friday 20:27:28 bconoboy: when will kernel-3.4.2-3 be available for use (in images)? 20:27:43 #info http://fedoraproject.org/wiki/Architectures/ARM/Quality_Assurance/Beta_Release_Criteria 20:27:49 dmarlin_: it's already built, just needs to be tagged and pushed by dgilmore 20:28:05 if it goes into updates-testing today I'll make an image that uses it. 20:28:19 #chair dgilmore 20:28:19 Current chairs: bconoboy ctyler dgilmore jonmasters pbrobinson pwhalen 20:28:49 I'm not hearing any objections to shooting for all 6 platforms 20:29:27 #info http://fedoraproject.org/wiki/Architectures/ARM/Quality_Assurance/Beta_Release_Criteria beta release criteria 20:29:28 the tests are simple to perform 20:29:30 bconoboy: but Panda and Vexpress are the only "required" platforms/blockers 20:29:31 #agreed F17 ARM GA will minimally include Vexpress and Panda, but may also include Trimslice, IMX, Kirkwood and Highbank subject to testing. 20:29:37 bconoboy: id prefer all of them 20:29:52 dgilmore: #disagreed? 20:30:01 bconoboy: no. 20:30:08 bconoboy: +1 20:30:17 bconoboy: i would rather ship images for all supported platforms 20:30:45 #topic 3) Documentation drive - wiki gardening, increase visibility on fedoraproject.org, update for F17 20:31:19 bconoboy: +0.5 20:31:22 I asked Paul to add this because the F17 release is going to get a lot of attention 20:31:23 dgilmore: +0.5 20:31:36 ... but the wiki is still largely F14 oriented 20:32:09 Would like to know if anybody has time to spruce things up 20:32:38 we need to really work hard on the wiki 20:32:42 Perhaps a second VFAD would be helpful 20:32:53 We could make that part of friday's affair 20:32:54 the docs are pretty horribly out of date and largely not relevant 20:33:13 dgilmore: +1 20:33:20 #info wiki for arm installation and configuration largely out of date, needs love 20:33:23 while testing maybe a great time to add things 20:33:37 bconoboy: +1 to second VFAD 20:33:38 Once we've done this we could then ask for better placement on the main fedora pages 20:33:39 dgilmore: I got some complaints about the wiki 20:33:47 #task ALL: spend 5 minutes cleaning up bits that you work with 20:33:55 jcapik: its a wiki fix it :) 20:34:08 maxam: As opposed to doing it during relesae validation? 20:34:15 maybe we need to just identify what we want/need documented on the wiki for f17 and accept volunteers for each 20:34:37 Maybe we divvy up boards and corresponding wiki pages for Friday 20:34:48 bconoboy: i did talk with rbergeron yesterday and we are going to look at talking to websites folks to make secondary arch downloads more visable 20:34:52 dgilmore: yup :] Once my managers release me from the enterprise world :D 20:35:04 dgilmore: great! 20:35:16 bconoboy: likely it will be a f18 thing 20:35:38 jcapik: is the 'enterprise world' near the underworld? 20:35:43 dgilmore: If we make gradual process from now to F18 GA that's fine 20:35:57 orc_fedo: it's under 20:35:58 bconoboy: indeed 20:36:10 secondary arches have made huge strides from f14 to today 20:36:11 Same with installer/qa stuff 20:36:47 Okay, I propose the following: 20:37:19 #idea Make friday's vfad include wiki updating- It goes with installation testing *and* it means more people can participate, even if they don't have hardware. 20:37:30 +1 20:37:43 +1 20:37:48 +1 20:37:58 bconoboy: sounds good, as long as people can run qemu-system-arm they can test :) 20:38:15 Maybe we need a page to track who's working on which board(s) 20:38:24 * jonmasters is back 20:38:48 ctyler: +1 20:38:49 * pwhalen can create the page for Friday 20:39:00 #help spend 5 minutes cleaning up bits that you work with 20:39:07 ctyler: +1 20:39:19 sounds good 20:39:45 #action pwhalen to create a page outlining who is testing which boards for vfad on friday 20:40:26 #agreed Friday VFAD to include wiki updating- sign up for a board or a wiki page 20:41:04 s/or/and/ # :-) 20:41:20 anything else for docs? 20:42:51 #topic 4) Your topic here 20:43:05 #topic raspi remix 20:43:12 w 20:43:22 * dgilmore wonders the status, i have not seen anything on the work promissed last week 20:44:11 would you like an announcement of some sort posted at the plugcomputer forum. 20:44:29 Ran into some snags (not all directly with the Raspi) so I missed Monday. Aiming to have the test image out tomorrow. 20:44:32 tinker-f595_: that would be awesome 20:44:38 ctyler: I pooled at the RPF pages ester4dat and they still did not mention the Fedora Remix as one of the options to use ... 20:44:42 s//looked/ 20:44:56 ctyler: Please sign up for raspi page ;-) 20:45:20 any special wording for the announcement? 20:45:25 ctyler: whats the plans to better engage the community to help here? and the plans to get fedora better visability in the pi community 20:45:42 tinker-f595_: can we wait till we are done talking about thepi please 20:46:35 dgilmore: There's plenty of interest for the test image and beta, we'll have good visibility. 20:47:08 ctyler: ok. because today it looks like we are not a option at all 20:47:38 i understand that we have some hurdles to overcome to have it be official. but i dont think they are insurmountable 20:47:38 We're not, the F14 image had some issues. We'll be back on the download page when F17 is ready, but I want to be sure it's really ready before it lands there. 20:47:52 I also think we should have that as part of the long term plan 20:48:37 I don't think the foundation wants to carry the task of maintaining the distro, but at the same time we don't want to carry the task of maintaining their frankenkernel, so we'll need to meet in the middle. 20:48:40 ctyler: can i ask you to talk up whats going on please. and ask for help where it can be given 20:48:53 I know we dont have many/any devs with pis 20:50:13 I will be talking plenty more soon. We're seeing the Pi hardware start to flow, hopefully we'll have a lot more Fedora folks with it RSN. 20:50:16 once the kernel is upstream we can enable it as a kernel.armv6l 20:50:49 we should look at doing armv6+vfp2 builds of some key libraries, I think 20:51:10 we should take a proposal to fedora legal to have an exception like we do for firmware for the boot code 20:51:17 esp. since the APC seems to have the same hardware spec 20:51:23 s/hardware/cpu/ 20:51:49 ctyler: we would need to do it via he softfp abi variant 20:52:45 dgilmore: to be really useful on the Pi, we also need to get the ARM videocore libraries too though. I think the firmware exception is straightforward, but we need a license change for the vc stuff. 20:53:16 ctyler: ok, we should have a plan of attack out there 20:54:07 ctyler: is there tasks today people can help with? 20:54:50 test the test image as soon as it's out, and help shape the default package set 20:54:59 * jonmasters discussed the upstreamability of the rPi kernel with upstream kernel folks last week...it's not hopeful 20:55:33 jonmasters: maybe s/hopeful/going to be quick/ 20:55:46 so I think the Fedora Remix will end up with a frankenkernel probably for a while 20:55:47 * dgilmore needs to run in 5 minutes 20:55:58 any news about the runtime linker location? 20:56:05 Alan Cox did an informal "need to do this to upstream" sort of review in github 20:56:07 ctyler: well, Greg and Arnd think it needs rewriting for a start :) 20:56:10 ctyler: again please just be vocal. especially if you cant make deadlines you have said you would 20:56:19 dgilmore: will do 20:56:22 even a quick note to the list with a update 20:56:36 #topic Plug computer forum messaging 20:56:44 tinker-f595_: your up 20:58:27 tinker-f595_: i think we want to say something along the lines of here is fedora 17 with lots of great goodies, and its treated as a first class citizen 20:58:47 tinker-f595_: we would need to make sure that people have the right uboot 20:59:18 dgilmore: that only affects one plug (Guru) afaik - the as-shipped uboot is ok for the others 20:59:53 ctyler: i was running a debain built uboot on my plugs and they would not uncompress the kernel 20:59:54 dgilmore: tinker-f595_ is probably AFK at the mo 21:00:34 If sending an announcement it'd be good to include a link to an updated wiki page on how to use the image 21:00:43 ctyler: i belive the uboot that shipped is fine 21:00:53 The kirkwood image is different from others in that you need to do some manual uboot configuration. It won't just work. 21:00:56 ctyler: but i dont think many people actually use that uboot 21:01:09 for each target, there ought to be a wiki page 21:01:11 bconoboy: right 21:01:13 dgilmore: the as-shipped Guru uboot has the wrong machine ID 21:01:27 I'd like each page to have a photo of the target, a link to the download for the image, and notes/useful comments 21:01:44 Shiny :-) 21:01:44 we can start with one or two, e.g. the plugs if you prefer 21:01:47 Coincidentally, there's no need to wait until friday to start tackling wiki updates for boards of choice... the nightly images are going to be substantially similar to Friday's RC 21:02:29 ideally, we could eventually have the front page for Fedora ARM basically say "hey, what do you have?" and then they get guided to the right page 21:02:33 #link http://lists.denx.de/pipermail/u-boot/2012-February/117020.html 21:02:35 bconoboy: I need to send an email with the info on how to change the uboot environment variables 21:02:45 #info fix for uboot on plugs 21:02:53 tinker-f595_: Best is to update the wiki with the information, then send an email with a link to the wiki included. 21:03:02 someone ought to be able to google for our project and then just come across a guide for whatever they have, the image, the known issues, comments about bootloaders, etc. 21:03:05 ctyler: you need to change 21:03:20 can we do *one* target, discuss the format, then use that as a template? 21:03:33 ctyler: most marvell factory uboot ship with something wrongly setup 21:03:36 mainlinelinux t0 21:03:43 true 21:03:45 jonmasters: sure 21:04:00 * dgilmore needs to run to a optomitrist appointment 21:04:04 how about panda or ts as a starting point? 21:04:09 something we all have lying around 21:04:11 we're past the hour now- shall we wrap up? 21:04:19 yep 21:04:22 ok, we can take this to Friday 21:04:25 yep 21:04:48 guys .... the runtime linker location 21:05:16 jcapik: we're punting on it to F18, or an F17 update. It's staying as it was in F17 GA 21:05:27 are we going to test that in rawhide and then back to f17 if ok? 21:05:35 jcapik: not enough time to ensure it's working, but F18 has already switched and we're testing it 21:05:42 jonmasters: oh ok 21:05:59 #endmeeting