[00:05] http://cdimage.ubuntu.com/kubuntu/ports/daily-live/current/ still has Lucid images too. [00:07] fixed for next sync [00:08] must fix that cdimage publication bug some day [00:09] kubuntu-netbook and xubuntu on tracker [00:10] mythbuntu should be done within half an hour or so, but I'm not going to wait up for it [00:12] also, have queued up Ubuntu, Kubuntu, and Edubuntu DVD builds [00:12] though I don't think alpha-2 should block on those getting tested [00:19] cjwatson: while you are updating iso.qa, could you kindly bump Ubuntu Arm Daily Preinstalled to 20100630.6? Thanks. [00:21] GrueMaster: done [00:21] thanks. === bjf[afk] is now known as bjf === bjf is now known as bjf[afk] [02:08] cjwatson: (or someone with admin access to iso.qa). I am unable to add more bugs beyond the default 3 fields in http://iso.qa.ubuntu.com/qatracker/result/4290/695 === jjohansen is now known as jjohanse-afk [04:35] I'm afraid I'm travelling to akademy today so won't be able to test CDs, I've asked for volunteers so hopefully it'll happen === bjf[afk] is now known as bjf === bjf is now known as bjf[afk] [07:01] Good morning [07:01] cjwatson: Didier mentioned that he expected netbook to be heavily oversized [07:02] cjwatson: it's actually not that bad any more with the recent package fixes, but we thought it'd be okay for alpha-2, since most folks are installing netbooks from USB sticks anyway (no CD drive) [07:04] Good morning [07:04] bonjour ttx [07:12] pitti: amd64 server is oversized [07:13] ttx: did that change over night? [07:13] the last respin made it pass the limit, yes [07:13] I have a script to compare two images [07:14] pitti: could you have a look ? [07:14] sure [07:14] so, 30.1 was 694 MB, 30.2 is 701 [07:15] which sounds like a good inflation for just a build1 [07:15] le huh [07:15] ttx: http://paste.ubuntu.com/457691/ [07:16] beeeh [07:16] seems this suddendly started pulling in half the desktop [07:16] please note that those are just differences > 200 kB [07:16] sorry, no [07:16] added packages are complete [07:16] changed packages are just > 200 kB [07:17] ttx: I blame yesterday's hplip [07:17] do you ship that? [07:17] yes [07:17] it changed things quite heavily [07:17] * ttx grumbles [07:17] there's obviously some dependency problem here [07:17] we should just blacklist it [07:17] * ttx looks [07:18] ttx: I think it might be the new python-notify dependency [07:19] ttx: would you mind filing a hplip bug about that, milestone it, and assign to tkamppeter? [07:19] I can do that. Though I always wondered why hplip was in server-ship in the first place [07:20] you ship cups, and HP printers are rather popular [07:20] but it should just ship the drivers, the desktopy bits should be separate [07:21] do we need a quick workaround ? Or is tkamppeter around to fix it ? [07:21] not sure whether he's awake yet [07:21] ttx: so you could unseed hplip, or we just declare it a known bug and leave it oversized [07:22] if you guys can test a respin again, I'm happy to respin, but it'd need to happen rather quickly [07:22] how much time do we have to test ? [07:23] this afternoon? [07:23] we can prepare the notes in the meantime and finish testing the other images [07:23] but we need some time to publish all the images and the announcement, etc. [07:23] well, it's not as if so much testing was done overnight [07:24] we're at 0/17 on amd64 [07:24] 6/17 on i386 [07:24] ttx: is there something in ship which we could drop? [07:24] you mean, besides hplip ? :) [07:24] i. e. that wouldn't require two publisher runs? [07:25] ttx: like, we could drop postgresql-doc for alpha-2 [07:25] which should bring us below 700 [07:25] oh, hmm [07:26] ttx: hplip is in the print-server task, and updating tasks takes two publisher [07:26] hm, so is postgresql-doc [07:26] why are all those duplicated? [07:27] is postgresql-doc usually a recommends ? [07:27] postgresql-doc is in the postgresql-server tasks, we can't easily drop it either [07:27] i. e. not more cheaply than hplip [07:27] the seed probably dates from a time when recommends were not installed by default [07:28] I mean why is server-ship so crowded? [07:28] it seems to duplicate everything in all the tasks [07:28] which makes it hard to see which bits are _only_ in ship, i. e. not installed by any task [07:29] looks like the duplication has been there for quite some time [07:32] pitti: how much do we need to free ? [07:32] there is a hplip -> sane-utils recommends but not sure that brings up enough [07:34] ttx: if we do an upload, we can just as well drop the python-notify dependency [07:35] which is a "more correct" thing to do [07:35] pitti: I was thinking about blacklisting sane-utils to avoid it on the CD without changing packages -- but I don't know if that's possible [07:35] ttx: some 1.5 MB, I think (how much to free) [07:36] ttx: no, I don't think it is [07:37] ttx: are you fine with a hplip upload and a respin? [07:37] I can do the upload [07:37] but would be nice if you could file the bug [07:37] yes, if we have until 1500 UTC for ISO test coverage [07:37] i'm filing the bug right now [07:38] gets tight, but should be doable [07:38] * pitti prepares hplip [07:39] https://bugs.launchpad.net/ubuntu/+source/cups/+bug/600504 [07:39] Launchpad bug 600504 in cups (Ubuntu) "Dependency on python-notify makes hplip unsuitable for servers (affects: 1) (heat: 6)" [Undecided,New] [07:39] assigning to tkampetter and targeting to a3 [07:39] or to you on a2 ? [07:40] tkamppeter/a3 is fine [07:40] hplip uploaded [07:40] I'll watch the build, and stall the publisher a bit if needed [07:41] takes 12 mins to build, should make it [07:41] oh argh [07:41] amd64 builds linux and kdeedu [07:42] but kdeedu is purging already [07:43] * ttx wtaches the snail race [07:49] kdeedu done now; c'mon crested, grab it [07:49] there; should just about make it :) [07:49] * ttx cheers [07:50] you might have to stall the publisher a couple minutes [07:52] ttx: hm, I could rebuild amd64 only, and keep i386 with the desktopy bits [07:52] ttx: do you want a clean i386 respin as well? or keep the current one? [07:52] hm, that's tricky [07:53] on one hand I'd like to keep the already-performed tests [07:53] on the other I'd like to have minimal difference between amd64 and i386 [07:53] is there a precedent for such arch-oriented respins ? [07:54] we did that in the past, I think [07:55] I prefer we do both -- I can make ISO testing happen by 1500 UTC [07:55] sounds like a plan [07:55] ttx: also, I guess you don't need to re-test each and every task [07:55] hoping that last respin doesn't bring its own set of surprises [07:55] some on i386, some on amd64 shoudl be enough for alpha-2 [07:55] especially not tasks which we didn't change since yesterday and which were arleady tested [07:55] * ttx prepares a large softfreeze-respect-stick just in case [07:56] hehe, please do :) [07:58] hah, that's my first "Orthopedic Dog Beds" spam [07:59] * ttx wonders about the success rate of that one. [08:02] c'mon, purge faster [08:02] publisher on manual [08:05] there we go, publisher running, and back on auto [08:06] this also caught the new kernel build, I hope that will go well [08:06] fyi I won't be in action until a few hours from now - I start late on Thu due to baby sign language class [08:08] cjwatson: I think I can handle it; see you later then [08:10] morning all [08:10] * ara resyncs [08:11] I did a desktop and a netbook install, and they work fine now \o/ [08:12] good, I'd done the install but couldn't wait up for it to reboot [08:13] you guys are talking about the ones that are posted? [08:14] yes [08:15] ttx: I set up a trigger to rebuild server ISOs as soon as the new hplip lands on the cdimage mirror [08:15] back in 45 mins [08:15] pitti: ok thanks [08:17] ara: we should have new 20100701 server ISOs in a few, will need your help in promoting them to the tracker [08:20] ttx, sure, I'll be around :) [08:32] ttx, do you happen to know if kubuntu is going to be rebuilt? [08:32] ara: I don't know [08:36] rebuilt for anything in particular? [08:37] cjwatson, no, I was just wondering about the build number being lower than ubuntu and xubuntu desktop [08:38] .1 rather than .2? [08:38] that just means that there were fewer builds of kubuntu yesterday [08:42] I know, I just wanted to double check ;-) [08:47] I respun everything containing ubiquity in sequence last night [08:49] mythbuntu posted as well now, FTR [08:50] pitti: server ISO still pending ? [08:50] DVDs should I think be ready too [08:51] haven't checked the tracker [08:51] ttx: yes, I'm watching them [08:51] cjwatson: will do [08:52] DVDs added to tracker [08:54] ttx: hm, not sure what's wrong -- but it seems that the publisher run somehow failed to publish it; the publisher finished minutes ago, but cocoplum still didn't update [08:54] ttx: I'll continue to watch it, but might take a bit longer [08:54] arh [08:55] cjwatson: I'll rebuild ports, unless you are already? [08:59] pitti: not doing so [09:02] mmm, Kubuntu Desktop looks like Kubuntu Netbook [09:03] * ttx gets some coffee [09:09] pitti: by "also caught the new kernel build" you mean the new i386 kernel that was just built ? [09:09] I guess I'll wait for Riddell [09:09] ttx: yes [09:09] pitti: that might make a case for a amd64-only rebuild [09:09] at least they woud get the same kernel source [09:09] s/rebuild/respin [09:10] hplip was published in this publisher run [09:10] very weird [09:10] * ttx wonders if keeping the old i386 spin would not actually end up being closer to the new amd64 one [09:11] ttx: the change in 6.9 was very small; do you have reason to believe that it's bad in any way? [09:12] ttx: if you don't mind having all the desktop bits on i386, we could respin amd64 only, yes [09:12] pitti: 6.9 actually fixes UEC [09:12] ttx: so why wouldn't we want to pick that up? [09:12] pitti: so we would have a fixed i386 and a broken amd64 [09:13] (UEC) [09:13] pitti: ok, let's keep the original plan [09:13] ttx: we can also wait another hour when the amd64 one is done (it's in the debhelper stage) [09:13] ah! don't tempt me [09:14] pitti: what would be the difference in ETA ? 15 vs. 75 min ? [09:15] ttx: 40 vs. 100 min [09:15] hm, I don't think we can afford that. [09:16] I can use that hour for smoketesting and making sure they are good [09:16] pitti: let's go for 40 [09:16] ttx: ack [09:16] (I'll document the need to dist-upgrade, on amd64 only [09:16] ttx: folks who install can upgrade to get the fixed kernel [09:16] ) [09:17] ttx: cheers; you'll put it into "known issues" on the tech notes? [09:17] yes [09:17] merci [09:17] pitti: if you have the URL of the maverick tech notes handy [09:18] ttx: https://wiki.ubuntu.com/MaverickMeerkat/TechnicalOverview [09:18] I'll do it now [09:18] thx [09:18] ttx: cheers; please let me know when you are done, then I'll brush up the "what's new" bits for desktop [09:18] ttx: btw, if you have anything worth mentioning for server, please add it [09:18] ok [09:26] pitti: done with the techoverview [09:26] pitti: will think about what I can add for server without keeping the lock [09:26] cheers [09:34] ttx: I'm done with the first round, if you want to add stuff [09:35] ok, ack [09:44] done [09:58] ScottK, around? [10:00] pitti: we are past your ETA, any new trouble ? [10:02] (if yes, let's just stick with the oversized ISO, we are lacking time now) [10:04] ttx: image is currently building [10:05] it took until 3 minutes ago until the new hplip found its way to the cdimage mirror [10:05] new ETA ~ 15 min ? [10:05] more like 5 [10:06] "a watched archive never publishes" or so [10:06] ok, let's live with that :) [10:06] image build done; now mirroring to cdimage [10:06] ttx: http://cdimage.ubuntu.com/ubuntu-server/daily/20100701/ [10:06] will hit in a minute or two [10:07] maverick-server-amd64.iso 01-Jul-2010 10:03 694M [10:07] yippie [10:07] ttx: go wild :) [10:07] ara: please promote to ISO tracker [10:07] added to tracker [10:07] ara: too late, I just did [10:07] yay [10:07] pitti, thanks :D [10:08] * ttx warms up his three test laptops [10:08] * pitti sees ttx jumping around with six hands, controlling three laptops at once === Daviey_ is now known as Daviey [10:13] if I had more HW, ya ha deedle deedle, bubba bubba deedle deedle dum [10:14] kvm FTW? [10:16] pitti: parellizing 6 installs on KVm tends to be slow [10:16] erm, yes; I never do more than two [10:20] for me, running just 1 kvm makes my laptop almost useless for the rest of my daily tasks [10:20] sometimes I can't even read email [10:22] erm, yes; I never do more than two [10:22] sorry, -EFOCUS [11:28] Is there a known issue about compcache's ramzswap kernel module failing to laod when booting the live cds? [11:29] known in the sense that I'd noticed it, although I hadn't reported it anywhere [11:30] it wasn't something I could fix straight away when I saw it, because the kernel interface changed - so best report and milestone (alpha-3) a bug about it [11:30] file it on initramfs-tools and assign to me, please? [11:33] ah, looking at dmesg, it's comlaining about an unknown parameter "disksize_kb" [11:33] okay, will file. [11:45] right, what can I most usefully do now? [11:45] aha, I can test the i386 DVD, nobody's doing that [12:20] * cjwatson tests btrfs via ubiquity while waiting [12:27] https://wiki.ubuntu.com/MaverickMeerkat/TechnicalOverview should be reasonably up to date now wrt. features, known issues, and general "alpha 2"ness [12:27] ara: Just passing through right now. I should be around in a couple of hours. [12:29] ubuntustudio tasks should be fixed; I'll check and remove [12:32] poo, btrfs doesn't work in the desktop image [12:34] no mkfs.btrfs [12:37] we might want to add bug 599450 in there [12:37] Launchpad bug 599450 in linux (Ubuntu) "[apparmor] getattr handled incorrectly in 2.6.35-6.7 (affects: 2) (dups: 1) (heat: 18)" [High,New] https://launchpad.net/bugs/599450 [12:38] used to break the LAMP task, I'm in the process of reproducing with latest ISO [12:40] cjwatson: so we need to seed it, and then teach oem-config and ubiquity to remove it (unless used)? [12:40] no teaching required, it could just be handled the same way as all the other filesystems [12:40] I'll file a bug [12:41] just trying to understand what's necessary [12:41] we'll need to relnote it for maverick [12:41] I'll add it; I'll drop a stanza about that apparmor bug anyway [12:41] one-liner in platform.maverick/live-common [12:41] I'll take care of it [12:41] ttx: this looks like it'd break cups in desktop as well [12:43] pitti: sounds worthy of a TechnicalOverview note, then [12:43] ttx: I'm adding it [12:44] it's breaking mysql, according to zul -- I'm in the process of reproducing [12:46] ttx: done (also mentioned MySQL); proofreading appreciated [12:49] looking [12:51] looks good [13:11] I get broken packages in kubuntu alternate i386 [13:13] mmm, wrong iso [13:13] *sigh* [13:17] ok, the last build of kubuntu is not the candidate one [13:21] hmm? [13:21] 20100630 and 20100630.1 both have some broken packages [13:22] I thought we'd dealt with that; did we just forget to respin? [13:22] bah. I'm going to respin kubuntu alt now [14:06] cjwatson, pitti: could you have a look at bug #600272 ? needed for the openjdk-6 uploads, needed for the firefox updates [14:06] Launchpad bug 600272 in ant (Ubuntu Karmic) (and 3 other projects) "ant fix to correctly build JAX-WS (affects: 1) (heat: 10)" [High,In progress] https://launchpad.net/bugs/600272 [14:06] ok [14:14] doko: done [15:19] cjwatson: I just went through the iso-testing bugs again, and the relevant ones are now documented in the tech notes [15:19] cjwatson: it seems most images are ok, except for the untested kubuntu alternates and DVDs [15:19] cjwatson: want me to start preparing the publishing? [15:19] ara: I'm around now. I imagine from the scrollback your question got answered? [15:20] (i. e. set it up, but not sync mirrors yet) [15:20] ScottK, well, it didn't, I didn't know who else to ask [15:20] ara: OK. Go ahead then. [15:20] ScottK, in Desktop Live i386 Kubuntu, the live session shows as if it was the Netbook edition [15:20] is that expected? [15:21] ara: For small screens, yes. We are trying to combine desktop and netbook into one image and do first run based on screen size. [15:21] ScottK, and once installed? [15:22] Once installed, there's an option in systemsettings to pick which you prefer. [15:22] Live/first run installed should be the same. [15:22] ScottK, ok, then why there are still two images in the tracker? [15:22] ara: Because we didn't finish the consolidation yet. I expect for Alpha 3 there won't be. [15:23] ScottK, OK, thanks for the explanation [15:23] You're welcome. [15:24] pitti: yes please, I have a presentation followed by an interview, which makes it hard :-/ [15:24] ack [15:30] cjwatson: for the record, I a-x'ed sync-mirrors [15:39] thanks [15:45] ttx, smoser: can you please publish the alpha-2 UEC images? (https://wiki.ubuntu.com/UEC/Images/Publishing) [15:46] pitti, yeah. will-do. [15:46] smoser: cheers [15:51] ara: do you know if someone is testing kubuntu alternates, in particular the amd64 one? [15:53] cjwatson: hm, do you know whether we usually publish-release DVDs? [15:53] (we didn't for alpha-1) [15:53] if they work [15:53] cjwatson: ok, thanks [15:56] pitti, not that Im aware of, I will resync now and give it a try in VM [16:01] pitti, I cannot access the iso tracker at this moment, can you? [16:01] no, indeed it's timing out for me as well [16:02] *sigh* [16:02] launchpad doesn't respond either there [16:04] jeesh, the interweb is broken [16:08] ara: does rsync for you from cdimage? [16:08] pitti, uec images are public and amazon's ami pages have been updated. http://uec-images.ubuntu.com/releases/maverick/alpha-2/ [16:08] it seems everythin http-ish might be down, but e. g. ssh works just fine forme [16:08] smoser: nice, thanks [16:09] pitti, no, I was syncing kubuntu amd64 alternate, and now it is stalled [16:09] oh, seems to be back? [16:10] works again for me [16:10] just joined the IS channel, they are on it === bjf[afk] is now known as bjf [16:20] oh hai [16:20] ah, thanks to whoever tested kubuntu alternate [16:20] hey lamont [16:20] I have no idea which or how many builds failed from that... reenabling the farm now [16:29] ok, done with preparing the image publishing [16:30] I'm waiting for kubuntu alternate amd64 confirmation [16:39] bladernr_: hello! how is your kubuntu amd64 install? any problems with it? [16:39] yeah [16:39] fails at select and install software [16:40] also, in VirtualBox at least, if I go from tty1 to tty2 to run a shell command, then go back to tty1, the installer screen doesn't redraw completely, so I get a bunch of vertical lines [16:41] I see that in kvm from time to time; switching to tty2 and tty1 again usually clears it [16:41] it's not an installer bug as such, must be a bug in either bogl-bterm or the kernel [16:42] bladernr_: can I see the installer syslog please? [16:42] bladernr_: uh, that's still with the 20100701 image? i386 was reported to work [16:42] pitti: actually, I have no idea... it is whatever was up as of yesterday, was this one of the respins? [16:42] oh, yes, yesterday's had broken packages [16:42] bladernr_: yes, it was respun today [16:42] that's why we respun [16:44] ahhh... then disregard my angst... it took me 14 hours to rsync the images yesterday so and I just heard of respins, but wasnt sure which ones were done [16:45] thanks, I'll resync that and try again [16:46] cjwatson: any other blocker from your side? I have the wiki page and email announcement prepared, images publish-release'd on antimony, and newz2000 standing by [16:47] (just to re-sync on where we are) [16:49] cjwatson: syncing mirrors now, since that'll take a while; in the unlikely case that kubuntu amd64 alternate is botched, re-syncing will be faster [16:51] pitti: the baton is yours unless you desperately need it to be mine - I've been on the phone most of the afternoon [16:51] cjwatson: ack [16:51] cjwatson: that's fine, I have nothing particular on the evening (except an hour phone call), I'll be around for another 5 h at least [16:57] I am downloading maverick-alternate-amd64.iso if it still needs tested but it says it won't be done for 40 minutes [16:58] Riddell: thanks [16:59] pitti: we're not doing dvds I take it? [16:59] Riddell: if anyone tests them, we can publish them [17:00] but I guess they aren't that interesting/important for a2, what do you think? [17:00] mm, at these download speeds that may not work for me [17:00] it's interesting only to know that it won't be a sudden headache for alpha 3 or beta [17:05] Riddell, I will start an amd64 test in 5 min [17:06] great [17:08] pitti: synced kubuntu alternate amd64 and redid, and still dies at "select and install software" [17:08] trying one more time just to be sure... [17:08] bladernr_: can you please rescue /var/log/syslog? [17:09] bladernr_: on the iso there is a .disk/info which would have the timestamp [17:09] pitti: working on that, I'll let you know when I've got it. FWIW, this is a VBox VM, not bare metal (my work laptop is my only bare metal 64bit machine) [17:09] that's fine [17:11] bladernr_: on 20100630.1 it was still broken: http://cdimage.ubuntu.com/kubuntu/daily/20100630.1/report.html [17:11] bladernr_: but on 20100701 it should work (http://cdimage.ubuntu.com/kubuntu/daily/20100701/report.html) [17:11] Kubuntu 10.10 "Maverick Meerkat" - Alpha amd64 (20100701) [17:13] that looks right [17:13] bladernr_: if you could switch to a terminal and check /var/log/syslog what it's failing on, that'd be great [17:18] pitti: oddly enough, this second run on that iso seems to be working :/ perhaps the first try (which was a vbox soft reset) cached the bad iso, while a power-down and restart re-read the data? [17:19] but so far it is doing the software install... so far so good, at least... [17:38] bladernr_: ah, thanks; seems we got two amd64 kubuntu testers now \o/ [17:38] so, time to continue then, I guess [17:49] pitti: I could have sworn that I posted the result here... anyway, worked fine on the second try. Installation successful with one UI issue that I filed a bug for (600716) [17:49] bladernr_: much appreciated, thanks for testing! [17:49] moving on to the rest of the kubuntu alt 64 tests [18:04] announcement sent -- folks, it's officially out [18:05] thanks to all for helping with testing, bug triaging, fixing, and release engineering! [18:08] fantastic, thank you so much [18:08] I was expecting to have to do more of that than I in fact did, so sorry about that ... [18:10] roll on having a dedicated RM again. :) [18:52] cjwatson: don't worry, it was good teamwork [18:52] cjwatson: thanks for another nice milestone release [19:03] pitti, cjwatson: thanks for your help !