16:06:36 #startmeeting 16:06:36 Meeting started Fri Jun 28 16:06:36 2013 UTC. The chair is dgilmore. Information about MeetBot at http://wiki.debian.org/MeetBot. 16:06:36 Useful Commands: #action #agreed #halp #info #idea #link #topic. 16:06:43 .fas jonmasters 16:06:43 jonmasters: jcm 'Jon Masters' 16:06:59 .fas blc@ 16:07:00 bconoboy: blc '' 16:07:02 #meetingname Fedora ARM 19 GO/NO-GO 16:07:02 The meeting name has been set to 'fedora_arm_19_go/no-go' 16:07:29 .fas dmarlin 16:07:29 dmarlin: dmarlin 'David A. Marlin' 16:07:59 #chairs pwhalen 16:08:04 #chair pwhalen 16:08:04 Current chairs: dgilmore pwhalen 16:08:32 lets wait a little for everyone to turn up 16:11:26 #link https://fedoraproject.org/wiki/Category:Fedora_19_ARM_RC3 link to test overview 16:12:48 ok 16:12:51 lets get moving 16:13:10 #topic QA status 16:13:21 and no paul 16:14:26 might be out sick- he's been ill all week. shouldn't really be working. 16:14:29 he was having network issues earlier... will probably be back on shortly 16:14:49 he has been bouncing in and out all week 16:14:55 all day 16:15:09 *nod* 16:15:34 so we have an issue with the kernel on cortex-a8 systems 16:15:39 none of them work 16:15:47 sorry, apparently dropped 16:15:52 pwhalen: :) 16:15:57 this has been around for a while, no one testing it 16:16:05 pwhalen: want to give us the run down of the QA status 16:16:23 what was the last message you got here? 16:16:30 ive been testing but could have done better about trying to get it fixed 16:16:31 pwhalen> https://fedoraproject.org/wiki/Category:Fedora_19_ARM_RC3 16:16:33 writing out a minimal vfat card to see how omap fairs, initial-setup-text ran on the armhfp minimal image 16:16:38 highbank install succeeded, with abc, going to try extlinux next 16:16:40 fossjons tool for writing vfat cards seems to work great, used the kde image, just finished soas 16:17:01 pwhalen: i just did a minimal image test on pandaboard 16:17:05 and? 16:17:07 pwhalen: initial-setup ran 16:17:11 w00t 16:17:20 user created and all that jazz 16:17:24 amazing, that was my biggest concern 16:18:16 a couple of things. has audio ever worked? I have never got it working myself, but someone mentioned it did 16:18:26 ive never tested audio 16:18:47 ok, been trying, hdmi and plugging in my headphone, nothing worked 16:18:55 huh 16:18:59 on which system? 16:19:07 trimslice and pandas 16:19:14 ok, ill test 16:19:59 thanks. 16:20:41 pwhalen: what issues do we have? 16:21:31 i think we've now covered our biggest issues 16:21:52 So...can we ship it yet? :) 16:22:37 need to file a bug on konqueror, it does not take text input on fas2, google 16:22:42 segfaults 16:23:00 ok, but that can be fixed later 16:23:13 right, mentioning things I found 16:23:19 thanks Paul 16:23:24 :) 16:23:30 pwhalen: and you're a trooper for even being here today...don't push it! 16:23:37 dgilmore, have you discovered anything? 16:23:48 jonmasters, I can breathe today, that feels good :) 16:24:26 pwhalen: it's the little things... :) 16:24:44 some of the desktops I dont really know the default behavior , soas is a pita to test 16:25:33 pwhalen: audio on trimslice works 16:25:37 with sugar, downloads seem to be an issue, but I have never used it before, other then to get out of it on an xo 16:25:45 dgilmore, what are you using ? 16:25:57 pwhalen: firefox 16:26:08 right, hdmi? 16:26:12 did a seach for ogg sample file 16:26:21 there is one in the testing matrix 16:26:23 pwhalen: sorry headphones 16:26:37 awesome, i'll take it 16:26:50 pwhalen: i had to plug them into the port labeled videoin to get sound 16:27:00 that's self explanatory! 16:27:02 :) 16:27:24 lol, hey if it works 16:27:24 at least it wasn't the port labeled "DC input" 16:27:32 the one labeled s/pdif gave nothing 16:27:52 thats the one i used, didnt even think to try video in 16:28:04 in fairness, I would never have tried that either 16:28:19 looking bellow the words s/pdif is microphone and video in is output the pictures seem to be right 16:28:40 pwhalen: there is little pictures below the text 16:29:15 see it now, weird though 16:29:25 pwhalen: ive tested highbank kickstart installs using extlinus and abc as boot loaders 16:29:37 both worked okay 16:30:02 i do feel that our installation coverage is a bit weak, but its better than f18 was 16:30:08 so its a step forward 16:30:39 dgilmore, sure, could use your or some help with increasing the different tests we do 16:30:59 so...the only thing you and I discussed Dennis was Cortex-A8. Now, I can promise to help work on a patch that allows that errata to be turned on and boots on A8 but I don't want to block for that 16:31:13 the pa testing matrix is very much the same as what we did previously. the tests are based on release criteria, so end up being almost the same 16:31:13 pwhalen: well we really have done very little to next to nothing interactively 16:31:17 I propose we ship now and consider 3.10 updated images soon that add OMAP3 devices 16:31:45 * handsome_pirate swigs from the rum bottle 16:31:47 I am +1 on shipping. 16:32:01 handsome_pirate: I don't drink, but I might join you after this week ;) 16:32:10 * handsome_pirate wonders what the status of blockers is 16:32:17 ok so qa's vote is to ship 16:32:23 handsome_pirate: there are none, it's perfect ;) 16:32:33 Heh 16:32:37 #topic releng status 16:32:38 being one man, i woudl like some other input too 16:32:45 Well, if no blockers, then by definition, we ship 16:33:00 ok, let's let Dennis run through topics one by one. rel-eng 16:33:05 handsome_pirate: well is the cortex a8 bug a blocker ;) 16:33:25 hi 16:33:26 its not.. 16:33:32 dgilmore: Duno, don't have A8 h/w 16:33:34 * jonmasters thinks it's not on the grounds that until I looked at it this wasn't even factoring into the equation 16:33:35 so the processes for f19 have been better and im much more comfortable with it that previous releases 16:33:53 +1 16:34:11 Next release we need to work on the qa process some 16:34:13 we have never defined a8 as a blocker. Its not been widely tested. 16:34:25 pwhalen: thanks, ok, cool, so I think that's decided 16:34:30 Well, if it's not a blocker, it's not a blocker 16:34:32 my main concern as to shipping or not is cortex-a8, me raised concerns over it but probably not loud enough since beta 16:34:32 handsome_pirate, input always welcome as mentioned before 16:35:00 pwhalen: This release we had sane criteria 16:35:14 For next, we should include the matrices like primary 16:35:33 I do believe that f19 is much better than f18 was. 16:35:34 in that we went with pa? we still need to read between the lines when doing so 16:35:43 handsome_pirate: we kinda started late in the game on the matrixes 16:35:53 dgilmore: Aye 16:35:56 so how about this, we've plenty of time to brainstorm awesome ways to improve things, but let's do that another time? 16:35:59 handsome_pirate, https://fedoraproject.org/wiki/Category:Fedora_19_ARM_RC3 16:36:07 I'm +1 on ship it 16:36:08 Why don't we do a retrospective right after we ship and plan for F20 process? 16:36:18 And +1 on a remix for cortex-a8 16:36:32 anyway, since it seems cortex-a8 is not a blocker im +1 to shipping 16:36:45 #topic Development status 16:36:45 before I vote, what is teh status of the initial-setup? 16:36:56 masta: initial-setup is fixed 16:36:59 works 16:37:02 ok then +1 16:37:07 bconoboy: works on console and graphically 16:37:07 masta: in RC3 initial-setup ran and was succesful on pandaboard 16:37:12 on minimal image 16:37:17 i tested ts, which works as well 16:37:22 bconoboy: its not fixed 16:37:26 its worked around 16:37:27 dgilmore: "fixed" 16:37:30 dgilmore: thanks for testing that, I was/am about to on a PandaBoard, waiting on minimal image download 16:37:44 * handsome_pirate is +1 ship 16:37:51 dgilmore: out of interest, what was the workaround? 16:38:08 jonmasters: /usr/lib/systemd/system/initial-setup-text.service 16:38:18 ah, say no more 16:38:22 Before=display-manager.service getty@tty1.service serial-getty@ttyS0.service serial-getty@ttyO0.service serial-getty@ttyO2.service serial-getty@ttyAMA0.service 16:38:40 before it had the wrong tty 16:38:41 as shipped it had Before=display-manager.service getty@tty1.service serial-getty@ttyS0.service 16:38:44 I see, you added a bunch of deps on the different tty interfaces 16:38:53 yea, I was assuming that's where you were going 16:39:05 jonmasters: so i added deps on the other serial consoles i know aboout 16:39:07 systemcrap has those tty generators that you block on, ok, cool 16:39:27 yea, think the only one you might have missed is one of the weird Samsung ones we don't need right now, so cool 16:39:28 ok, we're shipping.. what needs to be done before tuesday? 16:39:39 has anyone else tested fossjons tool for creating cards? 16:39:44 #topic todo 16:39:48 so the initial-setup was in my mind the big thing to "work around", I would have let it slip if that was not working 16:39:55 pwhalen: we need to make sure docs are right 16:40:01 especially for omap systems 16:40:09 we do, and I will start on it, but could use some help 16:40:10 as they seem to have caused some confusion 16:40:17 * masta will help with omap docs 16:40:31 well, to be fair, there are no docs on pa that explain tc installs 16:40:40 pwhalen: so for OMAP I'll help you 16:40:53 * jonmasters undertakes to retest any OMAP docs 16:40:56 unless i missed it, but also been testing and havent had time to write them 16:41:07 pwhalen: They're at the top of the install matrices 16:41:08 * dgilmore is about to stage primary and will stage arm on the mirrors 16:41:12 pwhalen: you're sick, take it easy, we'll all pitch in 16:41:13 and statis 16:41:16 static 16:41:31 our releases have been changing from tc to tc 16:41:34 from 2 images to 12 16:41:58 pwhalen: sure, we really should have sorted that out sooner 16:41:58 for OMAP question, dgilmore you said you got the UUID for the rootfs fixed, so it's just copying stuff into a fresh vfat and that's it, right? 16:42:04 jonmasters: yes 16:42:06 dgilmore: and of course updating the UUID of the boot 16:42:10 ok 16:42:14 then we agree on that 16:42:16 jonmasters: no need to touch uuid's 16:42:27 oh because the vfat one will end up being the same 16:42:28 and i tried to copy the files using *, rather then MLO first, and it worked fine, but I think MLO is copied first anyways? 16:42:34 jonmasters: for minimal testing before i xzcat'ed the image to teh sdcard 16:42:39 ran partprobe on it 16:42:50 then mounted 3rd and 1st partitions 16:42:56 coppied u-boot.img 16:42:59 then MLO 16:43:01 when did this change? MLO? 16:43:04 the uEnv.txt 16:43:21 unmounted 16:43:28 dgilmore: so you are saying because the VFAT is created during the image compose, the correct UUID is already in place 16:43:31 dgilmore: ok 16:43:32 and plugged card into panda and it booted 16:43:38 jonmasters: yes 16:43:39 great 16:43:43 with you, ack 16:44:05 jonmasters: at beta appliance-creator used a slightly different method setting up bits of the image 16:44:11 and it caused weirdness 16:44:11 right 16:44:18 i fixed appliance-creator for final 16:44:27 ok, and since initial-setup is in place, we'll set the root that way and we're all happy 16:44:29 great 16:44:41 happy =) 16:44:42 dgilmore: if you need a new PandaBoard ES I'll get RH to buy you one 16:44:43 jonmasters: right 16:44:51 dgilmore: I think you mentioned that earlier? 16:44:52 jonmasters: i dont have a ES 16:44:57 dgilmore: ok, we'll fix that 16:45:01 jonmasters: one would be good to have 16:45:10 dgilmore: consider it done (abusing a slogan) 16:45:29 * jonmasters will look at the OMAP3 issue since I promised I would if you shipped it now :) 16:45:38 jonmasters: so all the images shipped with a locked root account 16:45:40 and I'll look at why ES is failing on older revs for academic interest 16:45:43 https://fedoraproject.org/wiki/Fedora_ARM_Creator 16:45:51 initial-setup is needed to set root passwd or create a user 16:45:53 dgilmore: yea, I noticed :) hence I hacked the shadow files manually 16:46:13 dgilmore: they haven't replaced /etc/shadow with /etc/systemd-shadow yet so I still know how that works :) 16:46:18 jonmasters: one option if a user wants 16:46:21 in the doc for install, is everyone okay with using fossjons tool? has anyone tested it? 16:46:26 makes a concious choice 16:46:38 dgilmore: fair enough, might just not have a root, understood 16:46:44 pwhalen: ive tested it and doing it manually 16:47:06 jonmasters: its safer than shipped with a know root passwd 16:47:07 * jonmasters is +1 on ship, so let's stage this sucker and then do docs 16:47:17 now on the final release docs, what's the next step? 16:47:20 * masta just runs the cmds manually 16:47:24 * handsome_pirate wishes he could have been more help on testing 16:47:36 * handsome_pirate doesn't have any *supported* h/w 16:47:51 handsome_pirate: I'll get RH to buy you something 16:48:01 handsome_pirate: not today, but I will 16:48:07 Coolio 16:48:12 jonmasters: get things documented, write up release announcement 16:48:19 make sure that we document common issues 16:48:26 now, for the docs, are we taking the beta stuff and then staging docs in a similar location to F18? 16:48:30 alright, so I guess we ship, and make sure we have a nice wiki 16:48:39 * jonmasters is hazy on where this needs to go, can someone help me with really simple instructions? :) 16:48:53 jonmasters: in the wiki 16:49:07 dgilmore: really? I was going to write it on my living room walls! 16:49:12 I guess the wiki is more useful 16:49:22 * handsome_pirate lols 16:49:29 jonmasters: :) share photos with your google glass 16:49:32 dgilmore: :) 16:49:37 dgilmore: oh lol man, I so should do that 16:49:47 ok so if I go here: https://fedoraproject.org/wiki/Architectures/ARM 16:49:56 and then make sub-pages for e.g. PandaBoard, that's correct? 16:50:08 So I'll do PandaBoard as I've one here I've been poking, Dennis can review 16:50:18 who's doing Trimslice? 16:50:33 bconoboy: can you do TrimSlice? 16:50:38 I am not sure we need large pages like that 16:50:47 jonmasters: I only have serial console for trimslice 16:50:50 should be simple to do on one page, like I did for the beta? 16:50:58 oh sure 16:51:05 asking though 16:51:07 ok, if you prefer, yea, I was thinking F18 16:51:14 well, what's the concensus? 16:51:21 do we do like F18 or like F19 beta? 16:51:24 dgilmore: ? 16:51:38 * dgilmore doesnt care as long as its widely advertised and simple to find 16:52:07 I am okay either way myself 16:52:08 +1 16:52:30 ok so pwhalen you started something good for beta 16:52:35 pwhalen: you decide 16:53:12 415/428MB downloaded on minimal 16:53:25 ok, I can do the install instructions for everything. I will write up using fossjons tool, or manual options 16:53:51 pwhalen: how about you start a page and point me to it, and I'll make sure the Panda stuff is in shape? 16:54:04 for sure 16:54:20 I am not here next week, so will not be able to do our release announcement 16:54:22 pwhalen: a simple copy of the beta page is sufficient - you're sick, start with that, we can polish. Just let us know when we can have at it 16:54:35 pwhalen: i think we need to document both ways 16:54:47 dgilmore, that is what I meant, sorry 16:54:53 bconoboy or dgilmore want to do the release announcement? Perhaps Dennis? 16:54:54 pwhalen: :) 16:55:03 pwhalen: i should be able to send the announcement 16:55:07 i will be in Brazil 16:55:09 ok Dennis will send 16:55:18 while he drinks MATE :) 16:55:21 pwhalen: I'll do the release announcement 16:55:28 I will hopefully be somewhere on the Cabot trail, breathing in healing sea air 16:55:35 nvm 16:55:39 jonmasters: i probably will be drinking some 16:55:43 dgilmore: :) 16:55:53 dgilmore: the real stuff - I've a Gourd here myself 16:56:00 jonmasters: i have one as well 16:56:02 though people often say I'm out of it 16:56:08 ;) 16:56:09 anyway 16:56:18 #info vote is to go 16:56:30 ack 16:56:31 so the website folks also wanted some documentation, to have translated asap 16:56:31 #info stick a fork in it, shes done 16:56:34 #info Dennis to do release announcement 16:56:47 #info Dennis to do release announcement 16:56:52 jonmasters: your not a chair 16:56:57 you're 16:57:02 #info Paul to start page on release instructions, jonmasters to test OMAP 16:57:06 ah yea, ok sorry 16:57:52 #chair jonmasters 16:57:52 Current chairs: dgilmore jonmasters pwhalen 16:57:58 jonmasters: there you go 16:58:10 #info Paul to start page on release instructions, jonmasters to test OMAP 16:58:16 yay, now I can proceed to rant and yell ;) 16:58:36 ruh-ro, shaggy 16:58:44 #info jonmasters to provide an OMAP3 fix (post-release) sufficient for an OMAP Remix or updated 3.10 images later 16:59:27 I've also got a PandaBoard ES on order for Peter, which I will hand deliver next weekend 16:59:41 do we not plan on dropping panda in the near future? 16:59:49 I will arrange to get an additional one for Dennis soon, and hopefully John also 17:00:38 actually, let's followup on this stuff later, sorry, I guess it doesn't related to the release 17:00:49 (only thing that does is the OMAP3 notE) 17:00:52 * note too 17:01:21 jonmasters: yep 17:01:27 ok we are at an hour 17:01:31 we know what to do 17:01:31 #info Release retrospective will be planned for a week or two after release (when Paul and Dennis are back) 17:01:47 jonmasters: we can start on that in wiki now 17:01:49 shall we call it done? 17:01:49 I like that idea 17:01:52 getting notes together 17:02:00 lets call it doen 17:02:01 done 17:02:17 thanks Dennis, Paul, Brendan, Peter, John, Jon, everyone! 17:02:33 thanks all 17:02:36 #endmeeting