/srv/irclogs.ubuntu.com/2011/03/25/#ubuntu-release.txt

micahghi skaet, still around?01:07
skaethi micahg, yup,  what's up?01:07
micahgskaet: 1st, I have to apologize, I forgot to tell mr_pouit that gmusicbrowser needed an FFe for the sound menu integration, should I file a bug and get a retroactive FFe?01:08
micahgI think he assumed I already got it, but it got lost in the shuffle01:09
skaetmicahg, yes please.01:09
micahgskaet: ok, thanks, second, what to do with cairo-dock, bug 72399401:09
ubot4`Launchpad bug 723994 in cairo-dock (Ubuntu Natty) (and 1 other project) "FFe: Please update Cairo-Dock to 2.3.0~0rc1 version (affects: 2) (heat: 16)" [Wishlist,New] https://launchpad.net/bugs/72399401:09
skaetyou were waiting for some things to land,  where are they?01:10
* skaet hasn't looked at the bug yet, and probably should :/01:10
skaets/yet/today01:10
micahgskaet: so, now he tells me the Debian package isn't good and we should just use his version, so I'm a little confused, maybe he was hoping Debian would take his changes01:10
micahgI was hoping we would just be able to sync, but that's not possible apparently01:11
skaetmicahg,  my preference is to leave it on the bench then for a while, and sort it out after beta1 comes out.01:12
micahgskaet: we can still upload then?  Does UiF factor in?01:12
skaetIt does.01:13
skaetBut right now,  we're at UIF as well.01:13
skaetUIF/BetaFreeze was 2300 UTC01:13
micahgugh, ok01:14
* skaet should send out a note. Was chasing too many loose ends earlier. Will do now01:14
micahgskaet: bug 74218501:16
ubot4`Launchpad bug 742185 in gmusicbrowser (Ubuntu) "FFe: Please Sync gmusicbrowser (universe) 1.1.7-1 from Debian unstable (main) (affects: 1) (heat: 8)" [Wishlist,New] https://launchpad.net/bugs/74218501:16
micahgchanged title to look less actionable :)01:17
* micahg will have to go hunt for whale food now...01:18
micahgskaet: anyways, thanks01:27
skaetmicahg,  gone in and added comments, and subscribed ubuntu archive admin.01:38
micahgskaet: no need, it's already in01:38
skaetheh01:38
skaetokie01:38
micahgthat's why I changed the title and apologized :)01:38
skaet:)01:39
* skaet needs to get that note out now...01:39
=== slangasek changed the topic of #ubuntu-release to: Beta Freeze in effect! | Natty Narwhal Release Coordination. Please don't upload things during freezes where you shouldn't, or be prepared to apologise to the release managers with beer | we accept payment in cash, check or whale food | melior malum quod cognoscis
lamontslangasek: it's because y'all keep stuffing huge builds through arm, and breaking sleeves and stuff02:22
slangasekbreaking wut02:22
slangasekand hey, the biggest thing I stuffed through the arm queue today was php502:23
slangasekso don't look at me ;)02:23
lamontwell, all y'all02:23
lamontby which I mean DOKO02:23
lamontthough to be fair, that's mostly because he got a bunch of huge phat packages dropped on his head a few years back02:24
lamontdoko_: gcj-4.5 is also playing shooting-gallery with the armel targets.02:24
* lamont plays whack-a-mole with the outdated gcj-4.5 (-ubuntu1) on hubbard02:25
micahglamont: I'm about to upload chromium in a couple hours :)02:26
slangasekyeah, and I think the double gcj-4.5 upload is due to a multiarch buglet, so you can blame that one on me anyway02:30
micahglamont: actually if you're short armel builders, I can save chromium for the weekend02:31
micahghmm, is this new, all uploads need manual approval until release or until beta freeze is over?02:36
lamontslangasek: heh02:37
lamontanyway, the obviously mad armels are all stuffed back into the race02:38
lamontthe not-obviously mad ones will eventually be obviously mad02:38
lamontbut I need to go do some admin work on my daughter's laptop before she bursts into pouting02:38
nhandlerj/4802:49
chrisccoulsonskaet, could i get an NSS upload in to natty (it fixes the same issue as bug 741528, and is a pretty safe change)?09:29
ubot4`Launchpad bug 741528 in firefox-3.5 (Ubuntu Karmic) (and 15 other projects) "Compromised Comodo SSL certificates puts users at risk (affects: 1) (heat: 258)" [Medium,Fix released] https://launchpad.net/bugs/74152809:29
cjwatsonright, we're going to be wanting queuebot then09:37
cjwatsonScottK: mx51> linux-linaro-mx51 is going to need to be promoted to main to make this work.  Can you deal with any exception bugs or whatever that you feel are needed for that?11:20
ScottKSure.11:21
ScottKchrisccoulson: I'd suggest go ahead and upload and someone will review it in the queue.11:21
cjwatsonScottK: can I keep on calling it imx51 in d-i, or is mx51 better?11:21
cjwatsonI don't know the meanings of the terms11:21
chrisccoulsonScottK, thanks11:22
ScottKcjwatson: I'm not sure.  I'd say let's try to take the shortest path to success, but I'm not sure.11:22
cjwatsonit's not significantly harder either way; I'm more interested in what's right11:23
cjwatsongrepping for imx51 and changing it to mx51 is easy enough11:23
cjwatsonif that's the right thing to do11:23
cjwatsonbut if they're just different names for the same thing, I'll stick with imx5111:24
ScottKI'll ask and see if I can find out.11:24
cjwatsonthanks11:24
ogra_cjwatson, i think linaro implemented something in flash kernel thats called mx5111:28
ogra_hrm11:30
ogra_we have pleany of imx arches now in the code11:30
ogra_and the efika HW doesnt have a subarch check at all11:32
ogra_(so from flash-kernel perspective the name doesnt matter)11:32
cjwatsonflash-kernel doesn't care about the d-i image name anyway, only about what archdetect says11:34
ogra_(given that the kernel will likely only run on efika i would suggest calling it efikamx or some such though)11:34
cjwatson(and certainly archdetect names should be static)11:34
cjwatsonhm, if that were the right name shouldn't it also be the kernel package name though?11:34
cjwatsonI don't really like *inventing* names for d-i images11:35
cjwatsonI'd rather either stick with current (inertia) or else match the kernel package name11:35
ogra_well, flash-kernel nowadays checks for "imx5"|"mx51" and for "Freescale MX51 Babbage Board", "Freescale MX53 LOCO Board" and for "Genesi Efika MX (Smartbook)" | "Genesi Efika MX (Smarttop)"11:35
ogra_i dont think archdetect has the latter two at all11:35
cjwatsonso is mx51 a more general term than imx51?11:38
ogra_i think they are different subarches11:38
cjwatsonhttp://www.genesi-usa.com/products/efika says "Freescale i.MX515"11:38
cjwatsonso I guess imx51 is still accurate?11:39
cjwatsonso confusing11:39
ogra_well, i havent been involved with mx51 at all, thats a pure linaro thing, what i know is that their kernel partially works on the nettop and should also work on the babbage11:39
* ogra_ checks how their kernel is called11:40
ogra_-mx5111:40
ogra_so call the subarch that then :)11:40
cjwatsonok11:40
cjwatsonis there any reason to leave the imx51 subarch in place in d-i?11:41
ogra_no11:41
cjwatsonor can I rename it to mx51?11:41
cjwatsonok11:41
ogra_only historical ones probably, but i doubt we'll ever to a ludic image rebuild11:41
ogra_and if so, likely not with this d-i :)11:42
ogra_(god, my typing sucks)11:42
cjwatsonScottK: also, is there a bug for this feature in general?  if so, can it have a d-i task?11:43
ScottKcjwatson: I'll be sure to add it once I write it.12:42
ScottKcjwatson: It's 74243012:52
cjwatsonta12:53
zulskaet: can i get a FFE for new cobbler and openstack snapshot they are both in universe12:59
ScottKzul: Please just file the bugs and we'll look at them.13:05
ScottKcjwatson: I was going to subscribe ubuntu-mir to the bug and ask about promotion after someone approved the FFe.  I've discussed this with skaet and she OK'ed it, so I can either wait for her or some other release team member can mark it approved.13:15
ScottKcjwatson: So now I look around and that's not all.  I was under the impression that we had images for Kubuntu Mobile on n900 already.  We don't.  This already has an FFe (724534), but it's kernel, linux-n900 (which is based on Linaro's) is also in Universe.  Sigh.13:27
ScottKThis image is also on the manifest.13:27
ScottKDo pre-installed images need D-I changes?13:28
ScottKRiddell: ^^^13:29
RiddellI didn't know we were planning on having images for n900, would be awesome though13:31
ScottKIt's been on the manifest for the entire cycle.13:32
cjwatsonI don't think preinstalled images need d-i, but perhaps ogra could confirm13:34
ogra_only oem-config13:34
ogra_but you need a relatively complex debian-cd setup13:35
ScottKMy head hurts.13:36
ScottKogra_: Does headless need D-I?13:36
ogra_no, as i said above, only uses oem-config13:37
ScottKI thought that was pre-installed.13:37
ScottKBut great if it's both.13:37
ScottKcjwatson: If we do the Kubuntu image as a preinstalled image then I don't think we need mx51 support in D-I then.  Which also means we don't need to promote the kernels, right?13:38
ogra_you need the kernels for preinstalled too indeed13:39
cjwatsonwell, not on account of d-i at any rate; I don't know what preinstalled images require/prefer in that department13:39
ogra_preinstalled is pretty much identical to live images with its requirements13:39
ScottKogra_: But does the kernel need to be in Main?13:40
ogra_hmm, i dont think so13:40
ogra_cjwatson, we have a universe mirror antimony can access, right ?13:40
ScottKIt does for D-I, so avoiding that for the propsed linaro based images would be a really good thing.13:40
ogra_debian-cd pulls bootloader and kernel from the archive during build so both need to be accessible from the cdimage server (antimony)13:41
ogra_if thats the case it should work without promotion13:42
cjwatsonantimony can use universe for images that are configured that way, yes13:42
cjwatsonit's separated a bit to avoid universe leaking into main-only images by accident13:42
ogra_well, the debian-cd post-boot-armel+mx51 script needs to be created (preferably by someone having the HW), it needs to be able to pull the bootloader from the archive and pulls the kernel and initrd from the livefs builder livefs13:44
ogra_-livefs13:44
ogra_i see that kubuntu-mobile already has universe enabled in livecd-rootfs13:45
ScottKIt does.13:45
ogra_so it should be able to find the kernel there13:45
ScottKSo it's mainly the debian-cd work that needs doing?13:46
ogra_kubuntu hasnt though13:46
ScottKAh.  Right.13:46
ogra_that will likely need more changes13:46
ScottKHow about Ubuntu Headless?  Is that Main only?13:47
ogra_you need to add a switch for COMP i guess13:47
ogra_yes, headless is main only13:47
ogra_        headless)13:47
ogra_            LIST="$LIST minimal^ standard^"13:47
ogra_            LIVELIST="$LIVE_BOOT_SCRIPTS"13:47
ogra_            ;;13:47
ogra_thats all headless sets as defaults13:47
ScottKBecause if we could just do the mx51 headless image, then I wouldn't worry about the Kubuntu one for this cycle.13:48
ScottKn900 should be doable already since Kubuntu Mobile already has Universe enabled.13:48
ogra_headless is completely focused on serial though13:48
ScottKRight, but if one edits boot.scr before booting it to remove the oem-config preseed, then it should work, right?13:49
ogra_theoretically yes, untested though13:50
ScottKFor mx51 I'd be willing to try it.13:50
ScottK(I've got hardware)13:50
ogra_we currently set console= to serial and preseed debian-installer/framebuffer=false13:50
ogra_theoretically removing both should get you framebuffer oem-config in debconf mode13:51
ScottKSo if we could get the image built, I could try it and see.13:51
ScottKIf it works, then we can go from there.  I think "If you don't want to work via serial, do X, Y, Z" is not an unreasonable release note.13:52
=== apachelogger is now known as releaselogger
ogra_do you have a bootloader in the archive ?13:55
ogra_and a working boot.scr13:56
jibel_skaet, bug 70323014:08
ubot4`Launchpad bug 703230 in pango1.0 (Debian) (and 2 other projects) ""rm: cannot remove `/usr/share/doc/libpango1.0-0': Is a directory" when updating to 1.28.3-4 (affects: 3) (dups: 3) (heat: 79)" [Unknown,Fix released] https://launchpad.net/bugs/70323014:08
ogra_ScottK, i see u-boot-linaro-efikamx as well as u-boot-linaro-mx51evk ... with which in turn strikes me that livecd-rootfs wont be able to handle a kernel with linaro in the name OOTB, that will need a change too14:11
ScottKOK.  Sorry.  Got distracted by $WORK.  I'll look at this later.14:12
=== smspillaz|zzz is now known as smspillaz
doko_skaet, cjwatson: please approve openjdk-6 and openjdk-6b18 (multiarch library path)16:04
=== bjf is now known as bjf[afk]
micahgare unseeded package uploads still welcome?16:52
cjwatsonmicahg: generally yes, though they'll still go through the queue for review and avoidance-of-buildd-load16:55
micahgok, also, will the archive be open again after beta or remain moderated, I wasn't sure based on the beta freeze e-mail16:56
cjwatsonhttps://wiki.ubuntu.com/BetaFreeze clarifies I think16:58
cjwatsonwe'll be open for 10 days after beta-1, then hard-freeze for beta-216:59
micahgright, that's what I remember from the last 2 cycles, just wanted to make sure something didn't change since we switched to a beta 217:00
micahgeh, relatively close at least :)17:00
micahgcjwatson: you commented last release about sync with Debian uploads, when should those stop, hard freeze?17:01
cjwatsonI did?17:02
micahgcjwatson: yeah, I uploaded enigmail that just basically merged a Debian changelog17:02
micahgerr, wasn't sync, was a merge17:03
micahgcjwatson: was this upload: https://launchpad.net/ubuntu/+source/enigmail/2:1.1.2-1ubuntu117:04
micahgI'm just wondering where the point is to only sync/merge where we gain something instead of just being in sync or ready for sync17:05
cjwatsonmicahg: that sort of tidy-up should only really run up to Debian import freeze IMO17:11
cjwatsonlater than that there must be better things to do o:)17:11
cjwatsonsubstantive syncs are fine whenever a corresponding direct upload with the same change would be, IMO17:12
micahgcjwatson: ok, will keep in mind for the future, thanks17:13
bdrungmay i upload the ftbfs fix for lintian or should i wait until the beta freeze is lifted?17:20
=== bjf[afk] is now known as bjf
ScottKbdrung: Go ahead and upload.17:29
ScottKWorst case it sits in queue.17:29
bdrungk, done17:30
ScottKbdrung: The tests you dropped: Did Debian drop them too?17:32
bdrungScottK: no, but they are worthless with our two changes17:33
ScottKbdrung: I don't get the no-upstream-changelog change.  IIRC we are shipping at least partial changelog in all packages.17:42
bdrungScottK: some of the test cases produced the no-upstream-changelog warning, but the testcase didn't expected it.17:43
ScottKSounds like something that shouldn't be.17:44
zulrobbiew17:45
robbiewzul: what's up17:45
bdrungScottK: look at the test cases. they do not carry an upstream changelog.17:46
ScottKOK.17:46
* bdrung has to leave now.17:46
ScottKDid you file a bug with Debian on this?17:46
bdrungScottK: not yet, but i will forward it.17:46
ScottKOK.17:46
zulrobbiew: sorry misfire17:46
robbiewzul: lol17:47
ScottKbdrung: Accepted.17:47
cjwatsonparted and dmraid should either go together or not at all17:50
slangasekrobbiew: why the proposal to drop netboot support for !x86? (https://wiki.ubuntu.com/NattyNarwhal/ReleaseImageContacts)18:00
robbiewslangasek: x86?18:02
robbiewuh18:02
slangasekno, !x8618:02
slangasek:)18:02
slangasekthe netboot images are free as long as we're doing an existing d-i build for the arch18:02
cjwatsonagreed, netboot support should be kept even if there are no testers IMO - it's actively painful to remove them18:02
robbiewslangasek: I can't test them...and that was a requirement18:02
skaetslangasek,  no ones signed up to test them,   d-i's will keep building18:02
cjwatsonskaet: I can't make them stop publishing without doing violence to d-i18:03
skaetcjwatson,  don't want you to18:03
robbiew"Drop from that wiki page"18:03
robbiew;)18:03
cjwatsonwhere are you proposing to drop them from, then?18:03
cjwatsonI don't understand18:03
skaetNot publishing them as part of the 11.04 artifacts.18:04
skaetThis is all back to the UDS discussions on not publishing images we don't test.18:04
cjwatsonto where?  you mean http://cdimage.ubuntu.com/netboot/11.04/ or something?18:04
skaetyes,18:05
cjwatson(those are just links, sure, they're easy to drop)18:05
skaetI don't want the dailies to stop.18:05
cjwatsonok18:05
skaet:)18:05
doko_slangasek: while you are at it, please accept openjdk-6 too (and python2.6, python3.1, python3.2)18:06
slangasekdoko_: ack, looking18:06
doko_only openjdk-6 is on the dvd, but not on the cd's18:06
doko_will try to get binutils fixed over the weekend18:09
* lamont does another round of armel buildd restoration18:33
ogasawaracjwatson: I've uploaded linux-2.6.38-7.39, could I get it approved in the queue please?19:17
cjwatsonogasawara: done19:23
ogasawaracjwatson: thanks19:23
=== releaselogger is now known as apachelogger
kirklandFYI ... ^ bikeshed upload I had okay'd with ScottK earlier this week;  I missed a one-liner in the Makefile to get the change to take effect21:38
ScottKI'll review it.21:38
ScottKkirkland: Accepted.21:39
kirklandScottK: thanks21:39
ScottKYou're welcome.21:39
=== bjf is now known as bjf[afk]
slangasekpango1.0 in unapproved is a multiarch upgrade-handling fix we'll want for beta22:50
cjwatsonslangasek: accepted; I think we want the same the other way round too?23:08
slangasekfrom the plymouth side it just needs to be made more robust wrt these files being missing; plymouth itself shouldn't depend on pango, only plymouth-label does23:09
slangasekso a bit of tuning is still needed there23:09
cjwatsonwell, OK; I just noticed that new plymouth and old pango1.0 broke, and had been meaning to say something about it23:10
slangasekyeah, we should fix it that way 'round too - I just think a better fix is in order than Breaks23:11
cjwatsonwe shouldn't simply ignore those files being missing since that would result in no text being displayed (AIUI)23:11
cjwatsonI suppose we could fall back to the text theme or something23:11
slangaseknot that text gets displayed currently /anyway/ of course23:12
cjwatsonI think Breaks might result in a more consistent user experience though23:12
cjwatsonit gets displayed correctly for me - I can't reproduce that bug, so far23:12
cjwatsonat least IIRC23:12
slangasekah - I can reproduce the bug, and will see what I can figure out23:12
cjwatsonthat'd be lovely, thanks - plymouth bugs I can't reproduce are no fun at all23:13
cjwatsonI should probably check again mind you23:13
cjwatsonworks on nouveau, modulo the need to alt-f1/f7 before it shows anything (which I should investigate, but that's separate)23:14
slangasekthat's with plymouth in the initramfs?23:15
slangasekit may be that the hook is missing something else for the initramfs case23:15
=== bjf[afk] is now known as bjf
cjwatsonhm, now, I didn't think to try that - no, plymouth is not in the initramfs23:33
cjwatsonI have some time nowish, I'll wedge together an initramfs test case23:35
cjwatsongood call, I think that's it23:38
cjwatsonI don't see the blanked-out areas here like were in the image in the bug, but I don't see text either, having touched /forcefsck23:39
cjwatsonso that should be amenable to strace debugging23:39
slangasekcool23:44
cjwatsonthe effective diff in the grub2 upload (it's a merge, so maybe not trivial to work out which changelog entries matter) is:23:56
cjwatson- Fix crash when extending menu entry line beyond 79 characters23:57
cjwatson- Switch back to framebuffer page zero before loading the kernel (basically fixes black-screen that occurred on average 50% of the time when switching VTs on fglrx, and possibly broke other things too)23:57
cjwatson- Stop grub-setup being really slow if the first partition isn't near the start of the disk23:57
cjwatsonplymouth> looks like missing libGL symlink, am poking23:58
slangasek\o/23:58

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!