[00:02] <slangasek> ok
[00:02] <slangasek> I think I'll hold it until seeing what else blows up that would require a rebuild :)
[00:04] <heno> slangasek: server rebuild is ok by me - the server team seems on top of the testing
[00:25]  * heno starts an ubuntu 64 expert install (the other installs are I/O limited ATM)
[00:35] <stgraber> LTSP failed
[00:35] <stgraber> same reason, I'm making sure we have the fixed package in
[00:36] <slangasek> remind me which package that was?  It's been hours now
[00:36] <slangasek> I've entirely forgotten :
[00:36] <stgraber> the ltsp udeb I'd think
[00:36] <stgraber> ltsp-client-builder
[00:37] <stgraber> it's not the latest on the cd ...
[00:37] <stgraber> don't you need to do something in order to have d-i pick up a new udeb ?
[00:37] <stgraber> hmm, in fact none of the ltsp packages are the latest
[00:37] <stgraber> http://cdimage.ubuntu.com/daily/current/intrepid-alternate-i386.list
[00:37] <stgraber> says ubuntu2
[00:38] <stgraber> and rmadison says ubuntu3
[00:38] <stgraber> so looks like the alternate was rebuilt before ubuntu3 reached the archive
[00:38] <slangasek> heno: last time, there was a web page for tracking all the post-freeze packages accepted, for regression-testing; do we have something like that this time?
[00:39] <slangasek> stgraber: er, that's very strange; especially considering it lists gnome-terminal as out-of-date too, and that's what I used to trigger the build
[00:39] <slangasek> (i.e., http://paste.ubuntu.com/60638/)
[00:39] <stgraber> slangasek: yeah but I did a quick check and ubuntu3 is what's now on the archive and the file listing of the Cd shows ubuntu2
[00:39] <heno> slangasek: https://wiki.ubuntu.com/Testing/ISO/FixValidation
[00:39] <stgraber> so something went wrong
[00:40] <slangasek> stgraber: I absolutely agree that something went wrong, I just don't understand how or why
[00:40] <stgraber> should we mark all alternates as broken and needing rebuild ?
[00:40] <slangasek> no
[00:41] <slangasek> http://people.ubuntu.com/~ogasawara/weatherreport.html tells you which ones need a rebuild
[00:42] <slangasek> how do you guys feel about me accepting a kernel then, and you can start clean in the morning?
[00:42] <heno> stgraber: could you test ltsp on kub i386 atl which seems to be current?
[00:43] <stgraber> heno: does kubuntu has ltsp now ?
[00:43] <heno> slangasek: just for the CD-drawer bounce bug or is there more?
[00:44] <slangasek> heno: there are packages in unapproved right now to fix bugs #285572 and #285779
[00:44] <stgraber> looks like it has (at least the packages and the udeb), I'll need 2 hours to download the ISO though so I'd prefer someone else trying it
[00:44] <stgraber> I'll try to manually apply the patch from ubuntu3 on my current install
[00:44] <heno> ok
[00:45] <stgraber> slangasek: how many ISO would that be, just Ubuntu alt or do you want to rebuild more (for the kernel thing) ?
[00:46] <slangasek> heno: so the CD-drawer fix is also not included here yet
[00:47] <slangasek> stgraber: if we do the kernel, it's a full-spectrum rebuild
[00:47] <slangasek> I only found out about bug #285572 after we'd started the last round of rebuilds, otherwise I probably would've delayed for it
[00:48] <heno> slangasek: right, those bugs look more deserving
[00:48] <stgraber> and the kernel isn't built yet at all so that's : kernel build time + publishing + building new images ?
[00:48] <heno> I assume we're getting the xorg part of the fix too then
[00:48] <slangasek> heno: correct
[00:48] <slangasek> stgraber: yes - but you guys can sleep through that part :)
[00:49] <slangasek> if we decide before 00:03, I can guaranteeably control the timing such that everything will be ready for you guys in the morning
[00:49] <sbeattie> slangasek: there's a fix for the cd-drawer thing?
[00:49] <slangasek> sbeattie: not that I see
[00:49] <slangasek> hence, we don't have it ;)
[00:49] <sbeattie> hwh
[00:49] <sbeattie> heh, even
[00:50] <stgraber> slangasek: define "morning": what time zone ?
[00:50] <heno> ok, I'm not sure we can guarantee a Thursday morning release at this rate ;) but those bugs look grave enough
[00:50] <slangasek> stgraber: morning UTC... oh, did you say you're in Canada now?
[00:50] <stgraber> slangasek: yeah
[00:50] <slangasek> ok, well
[00:50] <heno> I would also rather minimise post RC changes
[00:50] <slangasek> exactly
[00:51] <stgraber> I guess if RC is late thursday, that should be fine
[00:51] <heno> sbeattie: no, it's just me getting groggy :)
[00:51] <stgraber> if you plan to release thursday morning western time, there may be a problem
[00:52] <slangasek> personally, I'm confident that we can rally the resources to get the RC out on Thursday European time; doesn't have to be in the morning, though preferably while London is still in the office
[00:52] <heno> btw, I'm in London all Friday so won't be able to test much
[00:52] <slangasek> but since a lot of that "rallying" depends on you guys, I'd like to know if you think I'm wrong :)
[00:52] <heno> my test box is not very portable :)
[00:53] <slangasek> this schedule also allows me to get some testing in this evening, before turning things over to Europe again
[00:53] <stgraber> well, I will be stuck in meetings all tomorrow morning but can likely free some time in the afternoon for ISO testing, then I still have the evening to finish with the remaining tests
[00:53] <stgraber> heno: use kvm -vnc and a VPN access to your box :)
[00:54] <heno> slangasek: WFM - but the kernel team have now burned all their testing slack ;)
[00:54] <slangasek> heno: absolutely :)
[00:54] <slangasek> ok, here goes everything
[00:55] <stgraber> ok, so let's rebuild the world :) And hope nothing breaks because I doubt we can afford one more rebuild after that :)
[00:55] <heno> slangasek: and 3g connection in the car, sure will do :)
[00:55] <heno> ok, I'll wander off to bed then
[00:55] <charlie-tca> You're rebuilding xubuntu too?
[00:55] <slangasek> stgraber: that's the other point, to be sure - are you guys comfortable with what's tested so far, that we shouldn't have any more break-the-world showstoppers?
[00:55] <slangasek> charlie-tca: yes
[00:56] <stgraber> slangasek: Ubuntu server looks good (I did two installs), so does Alternate (except LTSP that will be fixed with ubuntu3), I also had a friend install Desktop without any problem. So yes it looks good
[01:00] <heno> agreed, AFAICT it looks good
[01:01] <slangasek> ok, accepting packages
[01:03] <stgraber> I'm testing the ltsp fix by hand at the moment, just to make sure we didn't forget something. But the fix looks good and should fix the issue.
[01:05] <slangasek> ok, great
[01:06] <stgraber> would be something like the first time we have an Intrepid with LTSP installing from the CD :)
[01:09] <slangasek> stgraber: you said that you're bandwidth-starved today; are there images it would be useful to you to get at yet this evening?
[01:09] <slangasek> (where in Canada, BTW? Montreal?)
[01:10] <stgraber> I'm 90mi from Montreal
[01:10] <stgraber> the internet at work is fast so I have a copy of Ubuntu alternate, Ubuntu Desktop and Ubuntu server and I just rsync the diff
[01:10] <stgraber> so testing anything else will take hours to download
[01:11] <slangasek> ok
[01:11] <slangasek> so is it useful to you if I can deliver the Ubuntu images first tonight, or are you looking at tomorrow for all of your testing regardless?
[01:12] <stgraber> if you can get something relatively soon (let's say 3-4 hours) I can test them
[01:12] <slangasek> 4h is in the realm of possibility
[01:14] <slangasek> heno: have a good night :)
[01:14] <stgraber> seems to work, LTSP didn't crash at the usual place :)
[01:17] <heno> thanks. have a nice evening in the western hemisphere :)
[01:23] <stgraber> ok, install finished correctly with the LTSP fix
[02:56] <slangasek> stgraber: I see that the linux kernel builds average over 2 hours; so this won't be done before that 4-hour mark, after all
[03:00] <stgraber> and it's on the builder now ?
[03:00] <slangasek> yes
[03:03] <charlie-tca> Whoo, Xubuntu LiveCd passed manual install
[05:18] <calc> anyone tried a kubuntu i386 oem install?
[05:18] <calc> it seems to hang at a black xorg screen when i try to reboot into the oem bit after the install
[05:18] <calc> i am reinstalling again to see if it was just a fluke
[05:46] <calc> grr
[05:46] <calc> it did it again
[05:46] <calc> i'll just mark as failed and riddell can examine it
[05:49] <slangasek> calc: fwiw, I'm in the process of rerolling all images; sorry, I see now that you just joined the channel, and I haven't marked them invalid on the tracker, so you probably didn't know
[05:50] <calc> oh ok
[05:50] <slangasek> calc: kubuntu alternate will be among the first images to post in the re-run, but they're still at least a half hour out
[05:50] <calc> ok
[05:50] <calc> so far it seems to work fine except for the oem install part
[05:52] <slangasek> yes, there are no major failures that we know of with kubuntu alternate (except for ltsp), the rebuild is to pull in some more release-critical fixes while there's still time
[05:52] <calc> ok
[05:52] <calc> well it seems oem install is completely broken, or something weird is happening in my vmware
[05:53] <calc> X comes up but its just a black screen
[05:56] <slangasek> well, kubuntu desktop got successful reports with OEM mode
[05:56] <slangasek> so, I dunno
[05:58] <calc> ok, yea this was for kubuntu alternate i386
[05:58] <calc> in vmware 6.5
[05:58] <calc> vmware 6.5 is a bit weird wrt keyboard anyway
[05:59] <calc> i had to use xkeymap.nokeycodeMap = "TRUE"
[05:59] <calc> to get it to work right for me
[06:25] <slangasek> calc: fresh kubuntu alternates posted, if you care to resync
[06:25] <calc> resyncing already :)
[06:26] <slangasek> cool
[06:26] <calc> made a script to pull all the kubuntu/ubuntu images via rsync
[06:26] <stgraber> slangasek: ubuntu too ?
[06:26] <calc> so i wouldn't have to do it manually anymore :)
[06:26] <slangasek> stgraber: yes
[06:26] <persia> slangasek, During your reroll-everything effort, are you also rerolling MID and Mobile, or shall that be organised in parallel?
[06:26] <stgraber> ok, rsyncing then
[06:28] <slangasek> persia: I haven't scheduled those; I'm a bit fuzzy on what my role is supposed to be as far as managing Mobile/MID builds, but if you need me to push buttons to trigger a rebuild I'm happy to do so
[06:28] <slangasek> (though it will be several hours before the livefs buildds have quiesced)
[06:30] <persia> slangasek, I can bug someone else to push the trigger.  I'm a bit fuzzy on what needs to happen when adding new official flavours, and this will be the first release for both of these.
[06:31] <persia> slangasek, Do you know of any special registration that needs doing to get them to follow the same rough procedues as Xubuntu, Mythbuntu, or Ubuntu Studio?
[06:32] <slangasek> well, to let you know where I am, I think that's the first semi-official statement I've heard that they're releasing as part of 8.10, rather than based on it
[06:33] <slangasek> in the past, the process seems to be "let the RM know you're releasing together, integrate yourself into the alpha series, and go"
[06:33] <persia> Ah.  I seem to have missed the "let the RM know" part.  My apologies.
[06:34] <calc> looks like 48MB updates
[06:34] <slangasek> and since we didn't integrate these into the alpha series, that leaves a lot of unanswered questions we'll need to answer quickly, like where we're going to host the images, how we link to them in the announcements, ...
[06:34] <persia> hosting is at cdimage.ubuntu.com  announcements would probably best link to a wiki page (we're drafting release notes this week).
[06:35] <slangasek> ok
[06:35] <persia> I'll go poke people about releases.ubuntu.com, and see what I can find.  You make me thing several steps have been missed, of which I was unaware.
[06:35] <persia> s/thing/think/
[06:36] <slangasek> yeah... I knew there was a statement of intent to be better based on the archive for releases, but I didn't have a clear concept of where you were in that process :)
[06:36] <stgraber> though all the new images aren't correctly integrated in the tracker yet as that would have required a code update
[06:36] <stgraber> so no rsync/http/md5sum for these
[06:37]  * slangasek nods
[06:37] <persia> Yeah.  Unfortunately, the images didn't work reliably until just around beta, which was my main motivation for not trying to add them to the tracker.
[06:38] <persia> Is there a document somewhere that outlines the many things that need to be done?
[06:38] <slangasek> no
[06:39] <slangasek> it's not something that's been done often enough yet to get a document :)
[06:39] <persia> That makes sense :)
[06:40] <persia> I've heard plans for Kubuntu Mobile (and maybe also MID) for Jaunty, so I've certainly an interest in putting a checklist together to reduce confusion next cycle on that side.
[06:42] <slangasek> is it at all likely that some of these would converge with the CD livefs down the line?
[06:42] <slangasek> that would save build time, as we start to add more flavors...
[06:43] <persia> It is extremely likely that all of them would converge with the same image building tools.  The existence of usb-creator makes it less important to have a USB image.
[06:43] <persia> That just didn't get released early enough to make the change for intrepid.
[06:44] <calc> hold on did ubuntu desktop get respun also?
[06:44] <calc> i don't see updated tests for it
[06:45] <persia> I've just received confirmation that images are planned to be hosted at releases.ubuntu.com for Mobile and MID.
[06:45] <slangasek> calc: everything's being respun; everything with a date stamp < 20081022 is being superseded
[06:45] <slangasek> persia: ok.  has anyone worked out the directory structure for that?  i.e., does publish-release know what to do with them?
[06:46] <persia> slangasek, Maybe, and very unlikely.
[06:47] <calc> slangasek: ok
[06:47] <calc> so is RC still happening tomorrow?
[06:47] <slangasek> Thursday
[06:47] <calc> ah yea its wed here already :)
[06:47] <slangasek> depending on whether that's "tomorrow" to you :)
[06:48] <slangasek> the rest of the alternates have posted now
[06:48] <slangasek> all the alternates can post in the time it takes for one livefs build :P
[06:49] <calc> ah
[07:43] <stgraber> yeah, alternate is good
[08:00]  * stgraber starts Ubuntu alt 64bit and goes to sleep
[08:02] <slangasek> 'night!
[08:36] <davmor2> Good Morning Everybody
[08:39] <mvo_> hi davmor2!
[08:47]  * davmor2 starts with Xubuntu Alt
[09:39] <slangasek> davmor2: ubuntu desktop posted
[09:40] <davmor2> slangasek: Cool :)
[09:53] <heno> morning
[09:53] <davmor2> heno: morning
[09:54] <davmor2> desktops should be trickling through now
[09:54] <davmor2> I got xubuntu Alt running now
[09:55] <heno> ok, good. all alternates are ready?
[09:56] <heno> was there some additional reason to rebuild desktops than the kernel?
[09:56] <davmor2> no the livefs screwed up so it got delayed :(
[09:57] <davmor2> it's the rebuilds from yesterday
[09:57] <davmor2> as far as I know anyway
[09:59] <heno> ok, cool
[10:02] <slangasek> correct - the screw-up was a hung livefs buildd that took a bit to sort out
[10:03] <slangasek> but kubuntu desktop should be along in the next half hour; xubuntu desktop a bit after that; then the DVDs last
[10:22] <slangasek> kubuntu desktop up
[11:00] <davmor2> slangasek: so that's all the cd images back up now yes?
[11:00] <slangasek> CDs, yes
[11:00] <slangasek> DVDs are still in progress
[11:00] <davmor2> cool
[11:01]  * wgrant rsyncs.
[11:01] <slangasek> someone needs to invent a technology to make DVDs take no longer than CDs to build
[11:01] <wgrant> slangasek: Use The Rails Solution.
[11:02] <slangasek> wgrant: lie down on them and wait for the train?
[11:02] <wgrant> slangasek: I was thinking of throwing more hardware at it, but that works too.
[11:02] <davmor2> or just hit them with rails :)
[11:15]  * davmor2 readies vista and xp for wubi and m-a testing
[11:49]  * davmor2 starts wubi testing
[11:58] <wgrant> Hmmm. Are the two added applications menu items (Orca and System Cleaner) really meant to be in their own categories?
[11:58] <wgrant> And should System Cleaner lack its icons?
[12:06] <liw> system cleaner does not have an icon (at least not yet), and it's moving to system/administration
[12:07] <wgrant> liw: It also has a rather generic image in its top left corner, looking like it's missing some other image.
[12:10] <liw> wgrant, that's the image gnome uses when the app does not specify anything else; unlike everyone else in the world, I don't think that's a problem, but I'll add an icon if someone makes one
[12:10] <liw> (Me, I think the abundance of icons for every little thing is a problem, not the lack of icons :)
[12:11] <wgrant> liw: I don't mean the icon in the decorations. I mean the icon in the top left corner of the window.
[12:14] <liw> wgrant, yes.
[12:15] <wgrant> liw: Why is it there if there's no icon?
[12:16] <liw> wgrant, as a placeholder until someone makes one
[12:16]  * heno takes ubuntu 64 bit alternate
[12:17]  * slangasek starts syncing xubuntu amd64 desktop
[12:18] <slangasek> but I won't test it 'til the morning
[12:18] <stgraber> morning
[12:18] <slangasek> stgraber: hi again :)
[12:19] <stgraber> even managed to sleep 4 hours, not so bad :)
[12:36] <liw> I want to test something. Anything in particular need of attention?
[12:37] <davmor2> liw: everything :) all new cds :)
[12:39] <Hobbsee> [22:25] * Hobbsee expects to do entire disk, auto-resize, & live sesion, and maybe an oem session if i can figure out how to do that.
[12:39] <Hobbsee> (for i386 ubuntu)
[12:40] <heno> liw: studio or xubuntu perhaps?
[12:40] <stgraber> liw: I'll do Ubuntu server both arch in an hour or so
[12:41] <heno> I'll do kub 64bit alt next
[12:41] <liw> I'll start on Xubuntu amd64 then
[12:41] <davmor2> liw: live alt is done :)
[12:43] <wgrant> i386 live looks good for me.
[12:43] <wgrant> davmor2: live alt? Aren't they mutually exclusive?
[12:43] <liw> davmor2, "live alt"?
[12:43] <davmor2> test desktop.  Alt is finished
[13:13] <ara> morning all :-)
[13:14] <Hobbsee> morning ara!
[13:14]  * Hobbsee starts testing
[13:14] <davmor2> morning
[13:14] <ara> morning Hobbsee :-)
[13:14] <ara> morning davmor2
[13:15] <Hobbsee> hmmm.  is the RC *supposed* to work on virtualbox?
[13:16]  * ara starts testing ubuntu desktop i386
[13:16] <Hobbsee> ara: oh darn, tha'ts what I was doing.  Which tests are you running?
[13:16] <ara> Hobbsee: no worries, I'll go for alternate
[13:16] <ara> :)
[13:16] <Hobbsee> ara: cool.  I don't have that one downloaded :)
[13:18] <Hobbsee> oh wow, i like the new partition thing!
[13:18]  * heno starts kub alt 64 expert install
[13:23] <heno> *** DVD images are up ***
[13:23]  * heno presses the rsync button
[13:33] <ara> davmor2: that was a quick lunch indeed!
[13:33] <davmor2> too much to test just went and eat :)
[13:44]  * stgraber is done with Ubuntu server (amd64 and i386), going to some remaining tests on Alternate now
[13:46] <Hobbsee> wow...so that's waht orca does.
[13:46] <BUGabundo_work> what is Hobbsee?
[13:47] <Hobbsee> BUGabundo_work: screen reader and such
[13:47] <BUGabundo_work> yep
[13:48] <davmor2> don't test auto resize on ubuntu I need to do it for migration-assistant
[13:49] <Hobbsee> davmor2: for which?
[13:49] <davmor2> desktop
[13:49] <davmor2> both 32 and 64 bit
[13:49] <Hobbsee> oh
[13:51] <heno> well, having two people test the same case is not a disaster
[13:51] <liw> having a working X would also not be a disaster *sigh*
[13:51] <Hobbsee> that's true, i haven't started the test yet
[13:52] <heno> doubling up on a few images can be a good idea
[13:52] <heno> esp on different hw
[13:52]  * heno takes a short break
[13:52] <Hobbsee> yeah, that's true.  I think i'll do a manual partitioning on real HW, assuming the machine doesn't catch fire.
[13:52] <Hobbsee> i'd be interestedto see if it picks up the atheros card.
[13:54]  * Hobbsee scratches head
[13:54] <davmor2> Hobbsee: it should pick it up
[13:54] <Hobbsee> so, if i've gone from http://iso.qa.ubuntu.com/qatracker/result/2093/5, how do I get back to the whole list of tests for that iso?
[13:54] <Hobbsee> clicking on Ubuntu Desktop i386 seems to just give me info on how to download ig
[13:54] <liw> Hobbsee, I use tabs...
[13:56] <davmor2> Hobbsee: click on the tab under it for testcase
[13:56] <Hobbsee> davmor2: which one?
[13:56] <Hobbsee> liw: i usually do, but i forgot :(
[13:57] <davmor2> Hobbsee: Do you want the testcase or do you want to know what other tests are available?
[13:57] <ara> Hobbsee: use the breadcrumbs on top
[13:57] <Hobbsee> davmor2: the latter.  I
[13:58] <ara> Hobbsee: QA Tracker -> Test list -> Result list
[13:58] <Hobbsee> oh.
[13:58] <ara> Hobbsee: click on Test list
[13:58] <davmor2> click on result list
[13:58] <Hobbsee> right.   got it, thanks.
[13:58] <davmor2> or test list
[13:58] <Hobbsee> how do i do an oem setup?
[13:59] <davmor2> Hobbsee: hit f4 on the welcome screen and select oem
[14:00] <Hobbsee> davmor2: oh, i see, thanks
[14:02]  * ara takes xubuntu i386 desktop
[14:06] <liw> I'm having trouble with Xubuntu's auto-resize (from live cd): the disk had a previous xubuntu installation on it, and the auto-resize complains it can't make enough free space on the system for a new installation; should that work?
[14:07] <Hobbsee> liw: if you don't change the value?
[14:07] <Hobbsee> oh, wait, different bug, nvm.
[14:07] <liw> Hobbsee, what value? size of new partition? how do I change it?
[14:08] <Hobbsee> liw: drag the slider?
[14:08] <liw> oh, there's a widget hiding in there
[14:08] <Hobbsee> yeah
[14:08] <liw> that's not entirely hard to miss...
[14:08] <Hobbsee> i know...
[14:09] <liw> ubiquity would be the responsible package, I guess (filing bug)
[14:12] <Hobbsee> liw: actually, i think i'm hitting your bug now
[14:12] <liw> which one of my many bugs?
[14:13] <Hobbsee> [00:06] <liw> I'm having trouble with Xubuntu's auto-resize (from live cd): the disk had a previous xubuntu installation on it, and the auto-resize complains it can't make enough free space on the system for a new installation; should that work?
[14:13] <liw> I got past that after using the slider
[14:13] <Hobbsee> oh.  Then you hit the bug that i'm just reporting, i think.
[14:13] <liw> https://bugs.launchpad.net/ubuntu/+source/ubiquity/+bug/287546
[14:13] <ubot5`> Launchpad bug 287546 in ubiquity "ubiquity's partition resize option has a near-invisible slider" [Undecided,New]
[14:14] <Hobbsee> although that seems to refer to thecurrent system, not theone about to be installed.
[14:14] <Hobbsee> liw: right
[14:15] <Hobbsee> liw: https://bugs.edge.launchpad.net/ubuntu/+source/ubiquity/+bug/287546 is what i reported.
[14:15] <ubot5`> Launchpad bug 287546 in ubiquity "ubiquity's partition resize option has a near-invisible slider" [Undecided,New]
[14:15] <liw> Hobbsee, does one of us have an identity crisis?
[14:16] <Hobbsee> liw: sorry, 54*7*, not 6.
[14:16]  * Hobbsee smacks her copy key
[14:17] <liw> Hobbsee, right, different bug, but same place
[14:17] <Hobbsee> liw: yes
[14:22] <Hobbsee> davmor2: i've marked autoresize on i386 ubuntu desktop as a fail - would be interestedin seeing what your results are
[14:28] <davmor2> Hobbsee: Pass on both counts I'm just check the file transfers
[14:30] <Hobbsee> davmor2: interesting.  I'll retry tomorrow
[14:31] <Hobbsee> the quit button doesn't seem to work on ubiquity either.
[14:31]  * stgraber takes Ubuntu desktop i386 OEM + manual
[14:32] <Hobbsee> oh, darn, i'm on the wrong bit of it
[14:36] <davmor2> Whose doing what now?
[14:36] <Hobbsee> i've added mine, i'm out for the night
[14:37] <davmor2> anyone doing xubuntu desktops?
[14:37] <liw> davmor2, I am
[14:37] <liw> (amd64)
[14:38] <davmor2> and ara is on i386
[14:38]  * davmor2 takes studio
[14:40] <davmor2> anyone got any tv cards? if so can you have a look at mythbuntu?
[14:42] <liw> Xubuntu Desktop amd64 *done*
[14:47] <liw> hmm, mythbuntu and studio will want to test audio, right?
[14:48] <davmor2> liw: I would think so :)
[14:48] <liw> I can't touch those, then.
[14:51] <stgraber> Anyone on Kubuntu alt ? or should I take them
[14:51] <Hobbsee> i'd be interested to konw ifquit doesn't, on other ISO's.
[14:51] <Hobbsee> (for ubiquity)
[14:52]  * liw starts on "Upgrade Ubuntu amd64" and "Upgrade Server amd64"
[14:53]  * ara is installing xubuntu in Georgian to test ubiquity UTF-8 support
[14:53]  * ara thanks she knows by heart ubiquity screens :-)
[14:55] <stgraber> anyone saw: http://www.stgraber.org/download/langpack-notify.png ?
[14:56] <liw> stgraber, I think mvo is fixing that (based on #ubuntu-devel discussions), can you check with him?
[14:56] <stgraber> that's when enabling the french langpack on top of an english install (OEM install in english, then french in the user creation assistant, then enabling the missing packages using language support)
[14:56] <davmor2> stgraber: I did yesterday but then was told the cd's were being respun haven't seen it today
[14:56] <stgraber> davmor2: that's with today's
[15:01] <davmor2> stgraber: mine was on a standard english install
[15:01] <stgraber> ok, so not linked to OEM, good
[15:05]  * stgraber takes Kubuntu alternate i386: export + entire disk + entire disk with encryption
[15:09] <ara> liw: did you succeed with the kubuntu resize installation? it is failing for me on virtualbox. I will try on hw now
[15:09] <liw> ara, I was doing xubuntu, not kubuntu, but yes, it worked
[15:10] <ara> liw: I meant xubuntu
[15:10] <ara> liw: ok, I'll try on hw
[15:17] <davmor2> ara: worked for me on netboot too
[15:19] <ara> davmor2: ok, thanks :)
[15:20] <davmor2> ara: and alternate thinking about it :)
[15:20]  * liw goes out to have a bit of food, back in a bit
[15:26] <davmor2> who works on ubuntustudio?
[15:33] <stgraber> ogra: LTSP works !!! (first time since Intrepid development started IIRC)
[15:33] <ogra> YAY !°!!
[15:33] <davmor2> :) \o/
[15:33] <ogra> finally
[15:34] <ogra> stgraber, thats not ture btw, at least one apckage iteration worked when i was at the ahckfest
[15:34] <ogra> *package ... hackfest
[15:34] <ogra> for a day or two ... before i broke it again :)
[15:34] <heno> so mythubuntu installs and runs,though I cannot confirm that it's doing the right thing
[15:35] <heno> autoresize has been working for me in kvm
[15:36] <stgraber> anyone on Ubuntu studio ?
[15:36] <davmor2> yes just started but I need some info on it so I need to know who works on it
[15:37] <heno> TheMuso works on studio
[15:41] <davmor2> heno: I think ubuntustudio may need a respin
[15:42] <heno> davmor2: what's up? I can have a look
[15:43] <davmor2> no backdrop, gl screensavers listed under system tools, no office stuff (not sure if that is correct)
[15:45] <heno> davmor2: what options did you install, just the desktop?
[15:45] <davmor2> all of them
[15:45] <heno> not sure it should have desktop tools
[15:45] <heno> ok
[15:45] <heno> did it have gimp, blender and such?
[15:46] <davmor2> everything seems to of installed it just seems broken
[15:48] <davmor2> no email client either
[15:49] <ogra> you might have to sing your mails :)
[15:49] <heno> ok, trying a basic install now
[15:50] <davmor2> heno: when I did this install with beta everything looked fine.
[15:55] <heno> davmor2: ok, I'll have a quick look and if that's equally broken we'll request a respin
[15:55] <heno> of course the issues will need fixing first :)
[15:56] <davmor2> you're right of course :)
[15:56]  * heno kicks off a kub alt 64 oem install
[15:57]  * davmor2 quits on ubuntustudio for the time being
[16:09] <stgraber> uhm, OEM doesn't work with Kubuntu (from Alternate)
[16:09] <stgraber> first boot is ok, then it reboots and I don't get the oem assistant
[16:10] <davmor2> stgraber: that's an old bug that got fixed once.  Looks like it got broke again :(
[16:10] <stgraber> heno: tell me your result, just to make sure it's not something specifc to my setup
[16:11] <davmor2> stgraber: you on hardware?
[16:11] <stgraber> nope, kvm (that was kubuntu alt i386)
[16:13] <davmor2> burns cd to try it on hw
[16:18]  * davmor2 rsyncing for dvd's
[16:23] <heno> stgraber: in progress. Is the OEM assistant in the menu somewhere?
[16:23] <stgraber> you should have an icon on the desktop, then it'll tell you you need to reboot
[16:24] <stgraber> the assistant is supposed to start after reboot
[16:27] <heno> stgraber: right. I'm asking because some time ago the install icon itself was missing from the kub desktop but it worked fine from an entry in the menu - though this might be similar
[16:27] <heno> has anyone tested kub desktop oem?
[16:27] <heno> actually I can look that up :)
[16:28] <heno> tracker says 'no'
[16:30] <heno> my ubuntu oem tests were fine, fwiw
[16:30] <stgraber> mine too
[16:31] <davmor2> stgraber: about halt way through the install now
[16:32] <calc> is here still resyncing though
[16:32] <calc> heno: so did the kubuntu oem install work for you?
[16:32] <calc> heno: for me it just showed a black screen after rebooting into the oem-config-prepare thing
[16:32] <heno> calc: mine isn't complete yet
[16:32] <stgraber> calc: didn't for me (using alternate), same thing that you describe
[16:32] <calc> heno: i tested alternate install though
[16:32] <calc> stgraber: ok good, i thought i was breaking it somehow :)
[16:33] <calc> stgraber: i ran through it twice before reporting it to verify i didn't do something wrong
[16:33] <heno> I'll try kub desktop oem on a different box
[16:34] <heno> running short on IO juice here ...
[16:34] <calc> i am using vmware 6.5 for test installs here, so i don't know if it has any effect on it
[16:34] <stgraber> calc: mine was in KVM
[16:34] <calc> stgraber: ok
[16:34] <calc> stgraber: is there a bug open about this issue already? i can add my info to it if so
[16:39] <stgraber> calc: nope, I was waiting for someone to confirm it first (to be sure it wasn't a chair<->keyboard interface problem)
[16:39] <davmor2> desktop up now stgraber
[16:41] <calc> stgraber: ok
[16:41] <calc> yea it would be good to get confirmation on real hardware since both failures so far were on VMs
[16:41] <stgraber> calc: feel free to report one, I'm in a meeting at the moment
[16:42] <davmor2> reboot
[16:42] <calc> what do we report the bugs against kubuntu-meta ?
[16:43] <calc> ah i found the reporting tab
[16:43] <calc> hmm it doesn't say there either just to file them in general
[16:44] <calc> so what package should i file a oem install fails bug to since its in the oem bit, ubiquity or kubuntu-meta, or something else?
[16:44] <davmor2> stgraber: black background with a white outlined cross on reboot
[16:44] <stgraber> davmor2: on hw ?
[16:44] <davmor2> yeap
[16:45] <stgraber> ok, so we are sure it's not a VM issue now. OEM is broken in Kubuntu
[16:45] <calc> davmor2: same as our issue then
[16:46] <davmor2> https://bugs.launchpad.net/ubuntu/+bug/218144
[16:46] <ubot5`> Launchpad bug 218144 in ubuntu "Kubuntu KDE4 Alternative amd64 OEM installation failure" [Undecided,New]
[16:46] <calc> fails on i386 also so i'll update the bug accordingly
[16:46] <davmor2> it's been hit and miss with this for a while
[16:47] <calc> oh wow that is an old bug
[16:47]  * davmor2 tries to track down Riddell 
[16:47] <stgraber> why isn't he in the channel ?
[16:48] <davmor2> nope
[16:50] <davmor2> well I think that's every channel informed :)
[16:52] <calc> i marked the bug as ubuntu-8.10 (hopefully that is ok?)
[16:52] <heno> just trying this on kubuntu desktop oem
[16:52] <heno> I'd hate to have to respin all kubuntu discs at this point
[16:52]  * calc is now scping his images to his test machine and will start beating on images :)
[16:53] <davmor2> heno: and dvd
[16:53]  * ara is leaving now
[16:54] <heno> davmor2: indeed
[16:54] <stgraber> heno: Is that something that needs to be fixed in RC (I'd say yes as it's a whole testcase that's broken) ?
[16:54] <davmor2> worse is it was working in beta
[16:55] <heno> stgraber: it probably does need fixing, yes
[16:56] <calc> i don't know if this issue is the error messages that running oem-prepare-config spit out or not, but if you run it in a terminal you will see some messages
[16:56] <calc> i forgot to save them off last night :\
[16:57] <davmor2> persia: ping
[16:57] <persia> davmor2, yes?
[16:57] <davmor2> are you involved much with ubuntustudio?
[16:57] <persia> Yes.
[16:58] <davmor2> should it be missing email,office,background, and have system tools full of gl based screensavers?
[16:59] <liw> hmm, grub-install failed when installing _hardy_ (in preparation for an upgrade test), under kvm. I vaguely remember seeing this before
[16:59] <davmor2> persia: ^
[16:59] <persia> davmor2, checking now.
[17:02] <heno> studio looks ok to me (install just completed) appart from the odd screensaver menu
[17:02] <heno> studio-desktop does not depend on openoffice and friends
[17:03] <heno> the gray background may be a design choice (?)
[17:03] <persia> davmor2, ubunstudio-desktop does not include evolution or openoffice.  There should be a background.  It does include extra screensavers.
[17:03] <davmor2> heno: what backdrop do you have
[17:03] <heno> it's quite common in studio-type software
[17:03] <heno> davmor2: just a gradient
[17:04] <davmor2> that's no backdrop
[17:04] <davmor2> the default is the camera lens saying maro
[17:04] <heno> the screensaver package should be fixed to remove the menu entry
[17:04] <davmor2> macro even
[17:04] <heno> ok
[17:05]  * liw re-tries ubuntu amd64 upgrade and takes on ubuntu dvd amd64
[17:06] <persia> There *was* a background.  Looking at package histories now.
[17:06] <stgraber> calc: can you update the tracker with your test results for Kubuntu alternate amd64 ?
[17:07] <calc> stgraber: i tested kubuntu alternate i386 and already added it last night
[17:07] <heno> the camera lens image works fine when selected in the backgrounds selector, fwiw
[17:08] <calc> stgraber: i updated the lp bug report to reflect that it affects both i386 and amd64
[17:11] <davmor2> heno: to be honest I thinking it's just a settings misconfig
[17:11] <heno> yep
[17:11] <persia> davmor2, Seems ubuntustudio-wallpapers isn't on the image for some reason this time.  I'll go hunt image build logs.  Thanks.
[17:11]  * persia finds tuesday - thursday to be the least available time for testing, and is paying the price
[17:12] <liw> has anyone gotten an oosplash.bin crash when testing the examples on a live cd (or dvd)?
[17:13]  * calc hopes not :)
[17:13] <persia> No.  It is there.  Hrm.
[17:13]  * persia looks harder
[17:13] <liw> calc, three times now...
[17:13] <calc> hmm :\
[17:13] <liw> calc, after closing the program
[17:14] <persia> davmor2, I'm still rsyncing from the last update.  Do you have a /usr/share/backgrounds/ubuntustudio-lense.jpg ?
[17:14] <heno> drat. kubuntu desktop oem is broken too
[17:15] <liw> calc, but now I can't reproduce it anymore
[17:15] <calc> liw: ok
[17:16] <liw> calc, so if this was an isolated instance, let's not worry about it
[17:17] <calc> ok
[17:17] <liw> ooh! the trash icon changes when something is moved to trash
[17:17] <calc> liw: hasn't it done that forever?
[17:17] <liw> calc, nope
[17:17] <liw> calc, there was a bug when I did these tests for gutsy and, I think, hardy
[17:17] <calc> hmm its always shown a bit of paper above the trash can for me
[17:17] <calc> oh
[17:19] <charlie-tca> Is there a known issue with live cd and auto resize? I ran Xubuntu for today and it errors with "too small size".
[17:19] <charlie-tca> The partition is 20 GB with less than 500MB used
[17:20] <persia> charlie-tca, partman error 141, or something different?
[17:20] <charlie-tca> different
[17:20] <charlie-tca> This is the first time I tried auto-resize. The only thing I got is "size too small".
[17:21] <charlie-tca> I'll run Ubuntu next and see it if errors if there is no reports on it
[17:22] <charlie-tca> The /var/log/installer/debug log gave the same message. There is no error number
[17:25] <liw> charlie-tca, I had that kind of a problem, but it was because I did not realize I needed to adjust the partition sizes using a slider -- did you do that?
[17:25] <charlie-tca> What slider?
[17:26] <charlie-tca> It's a check box for auto resize under guided partitioning
[17:27] <liw> at the stage when it shows you the sizes of the partitions it will make (something made smaller, and the new one), there is a small, near-invisible slider between the partitions
[17:27] <charlie-tca> I'll try it again, and see if I can find a slider
[17:27] <charlie-tca> Thanks for the info
[17:33] <calc> i think i will fully reinstall my laptop with amd64 after RC releases to see if it suspend/resumes properly now
[17:33] <calc> it has some sort of weird issue when running off the cd
[17:33]  * davmor2 takes ubuntu i386 dvd
[17:43] <liw> calc, I keep getting the oosplash.bin crash, so I filed a bug, 287642
[17:44] <calc> ok
[17:44] <davmor2> liw: what happens if you open writer direct?
[17:45] <liw> well, this is not so nice: [13982.782919] kvm[10821]: segfault at 284 ip 0000000000435355 sp 00007ffff6c63500 error 4 in kvm[400000+19e000]
[17:46] <liw> one of my VMs just died (not the one having oosplash trouble)
[17:46] <calc> liw: did you mark that bug as security or something? i can't even see it
[17:46] <calc> oh no its automatically marked like that since its apport
[17:47] <calc> who do i talk to get added to the bug asap
[17:47] <liw> davmor2, nothing unusual, in half a dozen tries
[17:47] <davmor2> liw: which item triggers it?
[17:48] <liw> davmor2, randomly any of the oo files I've tried
[17:48] <liw> calc, what's your lp username?
[17:48] <calc> liw: ccheney
[17:48] <liw> calc, you're now subscribed
[17:48] <davmor2> liw: works fine on my main machine :-/
[17:48] <calc> liw: thanks
[17:49] <liw> davmor2, it doesn't happen every time for me, either
[17:49] <davmor2> okay thats 10
[17:50] <davmor2> heno: is there going to be a respin of all thing Kubuntu?
[17:51] <liw> hm, ubuntuvideo.com does not exist according to my dns
[17:52] <liw> and going to youtube doesn't make firefox offer to install a flash plugin... shouldn't it?
[17:53] <davmor2> liw: doesn't here either are you sure it's not screencast.ubuntu.com
[17:53] <davmor2> liw: go to google.videos instead :)
[17:54] <liw> davmor2, it was ubuntuvideos before, and https://wiki.ubuntu.com/Testing/Cases/UbuntuDesktop says that still
[17:54] <heno> davmor2: not my call, but it's very possible
[17:54] <heno> first we need the fix though
[17:54] <davmor2> I shall not be tempted to do kub dvd then :)
[17:54] <Daviey> liw:    Domain Name: UBUNTUVIDEO.COM
[17:54] <Daviey>       Created on: 03-Aug-06
[17:54] <Daviey>       Expires on: 03-Aug-09
[17:54] <Daviey>       Last Updated on: 03-Aug-08
[17:55] <liw> Daviey, can you actually access it?
[17:55] <Daviey> no :)
[17:55]  * calc fights the urge to file a bug about the big logout button on the top right now, pushes the clock way out of the corner
[17:57]  * davmor2 hit calc with big hammer to help him forget he ever saw it
[17:57] <calc> on the whole i don't like the theme changes in 8.10
[17:57] <calc> the right arrows on the menus stand out too much, the separate logout/shutdown/lock buttons, etc :(
[17:57] <liw> installing flash via add/remove makes it work in firefox, too. good enough.
[17:58] <calc> maybe i should switch to XFCE ;-)
[17:58] <liw> the logout/etc buttons aren't a theme change, though
[17:58] <calc> liw: unfortunately true, so i can't fix that
[17:58] <calc> well not easily anyway
[17:59] <heno> QA team meeting in 2 min.
[17:59] <tomjon> yay
[18:02]  * calc hopes the LP change for change sake doesn't take over ubuntu desktop ;-)
[18:02] <calc> wrt LP's gui changes ;-)
[18:10] <calc> Hey shouldn't System Cleaner be under System->Administration instead of Applications?
[18:10] <calc> its the only thing in Applications->System Tools
[18:10] <liw> calc, yes, and moving there
[18:10] <calc> ok
[18:11] <liw> what's npviewer.bin and why is it crashing?
[18:11] <calc> also is orca supposed to be appearing in the menu now? iirc it used to not show up there either
[18:11] <heno> liw: it's something firefox uses to run flash and such
[18:11] <calc> liw: probably flash
[18:11] <schwuk> liw: no idea, but it crashes a lot on my laptop as well
[18:12] <schwuk> flash would make sense, as that's normally what is used just before it crashes
[18:16] <calc> uh is VMware doing something weird or does the i386 ubuntu alt cd not display at all if Grub is on the hard drive?
[18:16] <calc> i've tried telling it to boot off cd first and tried telling it to directly boot of cd and it still just shows me the grub menu
[18:17] <calc> oh it somehow got disconnected
[18:17] <calc> i wonder how that happened
[18:18] <davmor2> calc: I thinks it's vm I've done a dozen or so installs when something eles was on the disc
[18:20] <heno> davmor2: was your DVD failure Live or alt install?
[18:21] <davmor2> heno:  it's live
[18:21] <davmor2> still running cd check at the minute
[18:21] <calc> davmor2: ok
[18:21] <calc> davmor2: well i know Ubuntu couldn't tell vmware to disable the drive, heh
[18:22] <calc> davmor2: just a very weird vmware bug (or feature?) i guess
[18:22] <davmor2> calc: it's a feature it's so you don't overwrite your favourite setup :)
[18:23] <calc> davmor2: heh, well you still have to tell your CD OS to install
[18:29] <calc> wow i'm killing the IO on my laptop :\
[18:30] <davmor2> heno: I need to sod off at 19:00 lug meeting
[18:30] <heno> davmor2: I'm getting the same hang on the dvd live
[18:30] <heno> davmor2: ok
[18:33] <calc> wow my whole system is non-responsive due to oversaturation of disk io
[18:33] <calc> i can't even get firefox to repaint
[18:34] <persia> calc, Are you swapping heavily, or just from file accesses?
[18:34] <calc> just large amount of file accesses
[18:34] <persia> You have free RAM and little in swap?
[18:34] <calc> i'm doing a resize in vmware and a verify local data in transmission
[18:34] <calc> yep 4MB in swap and 2374 free on the -/+ line
[18:35] <calc> er 4MB used in swap and 2374MB free in ram
[18:35] <calc> all the 'free' ram is being used for cached
[18:35] <persia> OK.  That's a less worrisome number :)
[18:35] <liw> oh dear, I repeatedly cannot install hardy
[18:36] <calc> its so loaded X isn't even repainting terminal drags
[18:36] <liw> grub-install fails... anyone remember this?
[18:36] <persia> Oh, so it's not actually free then.  Strange that you only have 4MB of swap, and are still having painting issues.  I'd consider that a bug, as disk i/o *shouldn't* block memory i/o.
[18:36] <calc> grub-install can fail for lots of reasons
[18:36] <persia> liw, How is it failing?
[18:37] <liw> persia, I didn't look for details (and killed the vm)
[18:37] <davmor2> heno: no errors :)
[18:37] <calc> persia: well if its doing disk io its also doing memory io of some sort, but apparently isn't being very fair about dividing that up
[18:37] <persia> liw, OK.  It's just that there are several ways it can fail, and which is interesting to solve the problem.
[18:37] <heno> davmor2: same here - the alt install is going fine too
[18:37] <liw> persia, sure
[18:38] <calc> some of the grub install failures are due to crappy hardware
[18:38]  * calc ran into that type with HPT fake raid
[18:38] <davmor2> just speaking to evand he wants some extra info so I'm going to try again
[18:38] <persia> calc, Right.  That's why I call it a bug.  There are tunables in the kernel, but by default we ought provide something that avoids perception of lockup, even if something is doing *lots* of torrenting and VM stuff.
[18:38] <liw> calc, this is under kvm... I remember doing this a few dozen times this year already
[18:38] <calc> liw: ah
[18:38] <calc> persia: yea this definitely not optimal, heh
[18:40] <persia> calc, You could try increasing /proc/sys/vm/swappiness : that might page out some more of what you aren't using.  Also, is your video card sharing main memory?
[18:40] <calc> it appears the paint issue is one of the apps is unresponsive
[18:41] <calc> when i minimize everthing then i can drag terminal around on the desktop without paint issues
[18:41] <persia> OK.  That's not as bad as it sounded then.
[18:41] <calc> persia: yea i think so, its an intel 945
[18:41] <heno> davmor2: oooooo after 10-15 min the dvd install continued :)
[18:42] <calc> oh hmm the app is responsive again, not sure what was going on
[18:42] <persia> calc, But you can paint, so that's not the issue.  Probably just over-aggressive caching.
[18:42] <heno> I'm trying an XFS install too just for giggles, and that's waiting at 5% too
[18:42] <calc> well it seems it was evolution
[18:42] <calc> which seems to not deal well with low IO ability
[18:43] <calc> i think it was evolution that wasn't repainting when damaged
[18:43] <sbeattie> Hah, that's awesome. Had a power glitch that rebooted my VM server while a VM was doing an upgrade test. Now that VM just cycles starting gdm over and over again.
[18:47] <calc> a 28gb verify takes a long time on a laptop hd :\
[18:48] <heno> the xfs install has resumed too
[18:49] <charlie-tca> liw: thank you, thank you, thank you. The auto resize works
[18:49] <liw> charlie-tca, cool
[18:50] <heno> did anyone file a bug about that? it seems it should pick more sensible defaults
[18:50] <liw> heno, I did, and Hobbsee filed a related bug
[18:50] <heno> I've seen it goof once now, but it's mostly ok
[18:51] <heno> liw: ok, thanks
[19:03] <calc> wow linux is really really bad about disk io
[19:03] <liw> hm, I did a hardy install on my kvm host running hardy, and that worked like a charm, so it might be that the intrepid kvm has some trouble
[19:03] <calc> i can't even use pidgin due to the disk io from transmission
[19:03] <calc> and evolution is pretty much useless as well
[19:03] <liw> calc, that's especially noticeable on laptops, yes
[19:04] <calc> liw: probably wouldn't be much better on a desktop when doing a torrent check
[19:04] <liw> calc, it is, at least for me, but it's still a problem
[19:04] <calc> ok
[19:05] <persia> calc, The top of http://www.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.7/2.6.7-mm1/broken-out/vfs-shrinkage-tuning.patch might be interesting for you to read.  I think that's the tunable that you would want to tweak.
[19:07]  * calc sees if he can get firefox working enough to go to the link
[19:07]  * persia pastes the relevant bits in /privmsg
[19:08] <calc> persia: would that tunable actually help though, i'm not out of memory (afaik) its just that its seemingly not roundrobin'ing disk io
[19:09] <calc> or not very effectivly
[19:09] <persia> Mind you, I could be misunderstanding that.  I strongly suspect that the issue is with you not having spare memory, as it reminds me of the behaviour I had with very high values of swappiness.
[19:09] <calc> i assume its safe to change those numbers while under disk io?
[19:10] <persia> Do you only get slowdown on disk i/o?  Not being able to use pidgin made me think it was something else.
[19:10] <persia> I believe you can change things : I know you can adjust swappiness while thrashing.  I've not played with vfs_cache_pressure before.
[19:11] <persia> (when I have i/o storms, I usually did something stupid, and don't even have a working terminal)
[19:11] <calc> persia: yes its always worked before
[19:11] <calc> and both evolution and pidgin are having paint issues from what is most likely due to disk io
[19:12] <calc> the cpu usage is fairly low but i am > 90% iowait
[19:12] <persia> and pidgin is in iowait?
[19:12] <calc> not sure atm, its responding again right now
[19:12] <liw> it does sound like virtual memory thrashing
[19:13] <calc> increasing the vfs_cache_pressure number didn't make slab increase
[19:13] <persia> Mind you, for disk i/o, you don't see that in swap, as the pages are clean.
[19:13] <calc> unless i am tuning it the wrong way?
[19:13] <calc> actually i think it should make slab smaller
[19:13] <calc> but it didn't do anything noticable
[19:13] <persia> You're tuning it thr wrong way.  See /msg
[19:14] <persia> is free growing any, or cached shrinking?
[19:15] <calc> oh now it is shrinking
[19:15] <calc> it took it a while
[19:15] <calc> it just dropped by ~ 700MB for cache
[19:15] <persia> Did that fix the repainting issue?
[19:16] <calc> i think so
[19:16] <calc> it wasn't happening all the time but i'll see if it stays gone
[19:16] <persia> Also, 50-100MB is probably enough.  Something between 100 and the current value is likely a better balance.
[19:16] <calc> still have 30% of the file to check
[19:17] <calc> the cache is going back up a bit but is still several hundred MB smaller than it was
[19:17] <calc> and evolution is responsive now
[19:17] <calc> well about as responsive as it usually is anyway ;-)
[19:18] <persia> Next step : chat with the kernel team and get some feedback on whether 100 is the right value for most users.  Maybe it's something you want to tune for yourself, and maybe it's something that should be tuned generally.
[19:18] <calc> yea
[19:20] <calc> apparently 1000 is not enough for my system
[19:20] <calc> i'm trying to find where it is usable without being too high
[19:22] <persia> Well, you're also doing *lots* of I/O.  You probably need a temporarily extra-high value.  Setting it too high in the normal case would annoy people who expect the browser cache to do something useful, or expect to be able to keep a ISO in RAM while doing an install test.
[19:22] <calc> persia: so there's no way to schedule io between apps like there is for cpu time (or used to be?)?
[19:23] <persia> I don't know.  My knowledge of kernel stuff is spotty, and limited to the few things that I needed to discover because I broke my system in one way or another.
[19:26] <liw> stgraber, I may have asked this before, but... would it be possible for the iso tracker to have a list of all states except completed?
[19:28] <stgraber> liw: not possible yet and not something I can implement now (I can't do any code update for a while)
[19:28] <liw> stgraber, ook
[19:28] <liw> did someone say that kubuntu images might be re-done?
[19:41] <stgraber> liw: OEM is broken so they'll probably need a rebuild
[19:41] <joaopinto> Hello, does anyone know if the current ISO for amd64 desktop is broken ?
[19:43] <joaopinto> I am having the worse install experience I had ever with an Ubuntu install :\
[19:46] <persia> joaopinto, http://iso.qa.ubuntu.com/qatracker/test/2092 shows current test status, and outstanding bugs from testing.
[19:47] <joaopinto> there is 1 successful test, so the issue must be specific to my hw
[19:48] <joaopinto> all my hw was properly supported on hardy, now I don't have, keyboard, mouse and network
[19:50] <persia> Very odd.  USB keyboard/mouse?
[19:51] <joaopinto> nope, a regular ps2 keyboard (which works fine on the console), and a regular wheel usb mouse
[19:51] <joaopinto> regarding the network I get a strange "unregistered" protocol, when manually running the dchlient on eth0, which gives me the idea it could be related to kernel modules not being properly loaded
[19:52] <persia> Especially odd, as I have the same configuration running from daily updates (not CD install) on amd64.
[19:52] <joaopinto> a get an odd error during boot, related to a missing .deps file on the modules directory
[19:52] <liw> joaopinto, have you ran "check this cd"?
[19:53] <joaopinto> the livecd/install worked fine, so it's not really an hw support issue, it must be something related to the install process
[19:53] <joaopinto> no, just checked the md5sum of the iso
[19:53] <persia> It ought run depmod at some point though.  You're not the only one installing today, and it sounds like you've hit a systemic failure of some sort.
[19:54] <liw> joaopinto, sometimes a broken cd results in _really_ weird problems, and you're having those, so that would be one thing you could try
[19:55] <persia> And matching md5sum doesn't confirm that the burn went smoothly.
[19:55] <joaopinto> ok, I am going to test the cd, I would export some checksum error during install for a burning related error :P
[19:55] <joaopinto> export=expect
[19:56] <joaopinto> brb
[19:56] <liw> joaopinto, so would I, but, you know, weirdness happens
[19:56] <joaopinto> hum, let me also lsmod from the livcd and check agains the installed system
[20:17] <joaopinto> back
[20:17] <joaopinto> the cd is fine and meanwhile I found the problem
[20:18] <joaopinto> because I have selected to use an existing partition, mounted at /, the installed decided to keep my hardy kernel
[20:19] <joaopinto> so now I have a messed up install, an hardy kernel with intrepid modules
[20:19] <persia> joaopinto, The trick when reusing partitions is to check the "format this partition" box.
[20:20] <joaopinto> persia, it was the first time I decided not to format it, because from my readings, it would now be safe to keep /home on the same partition, the install would overwrite the system dirs but not /home
[20:20] <persia> If you want to upgrade an existing install, perform an upgrade rather than a fresh install (this preseves configuration and data).  The alternate CD is good for this if your network is bad (and we'd appreciate test feedback)
[20:20] <joaopinto> however, it seems that is broken
[20:20] <persia> Where did you read that?
[20:20] <joaopinto> luckly I decided to keep /home on another partition :P
[20:21] <joaopinto> persia, I believe it was on the ubuntu devs mailing list
[20:21] <joaopinto> and the installer does warn you that is going to "Remove conflicting operative system files"
[20:21] <joaopinto> however, it removes the existing modules, but keeps the existing kernel
[20:23] <joaopinto> persia, https://wiki.ubuntu.com/UbiquityPreserveHome
[20:23] <persia> joaopinto, Did you have a custom kernel installed?
[20:23] <joaopinto> nope
[20:24] <joaopinto> and the funny thing, it did update the grub strings
[20:24] <persia> Has anyone else tested this use case?
[20:25] <joaopinto> title		Ubuntu 8.10, kernel 2.6.24-21-generic
[20:26] <joaopinto> where do I fund the use cases descriptions ?
[20:27] <joaopinto> I just see the "test Install (manual partitioning)"
[20:27] <joaopinto> which is not specific about keeping an existing partition
[20:28] <persia> joaopinto, The use case ithat seems untested s the one from https://wiki.ubuntu.com/UbiquityPreserveHome, which you tested.
[20:29] <joaopinto> ah, blueprint wise
[20:29] <persia> Be nice to get more tests.  In the meantime, could you file a bug with a description of what happened?
[20:29] <joaopinto> eventually it was tested during implementation, which was some releases ago
[20:29] <joaopinto> I am going to file a bug report, but I need to get a working system, I will not be able to do further analysis :(
[20:29] <persia> Hrm?  It was completed on 14th June, which is this release cycle.  It's a new feature.
[20:30] <persia> Using the system you're using for IRC, download linux-image-2.6.27-7-generic and dpkg -i it on your broken system.
[20:30] <joaopinto> I remember reading it would be available on Hardy
[20:31] <joaopinto> I am going chroot, sudo apt-get install ;)
[20:31] <persia> From the spec it looks like it was started in January, which would be targeted for Hardy, but it looks like it wasn't completed in time.
[20:49] <liw> this is very vexing: on my hardy host, I can create a hardy system, but it is very, very slow because the UI goes over network; on my intrepid host, I can't create a hardy system at all
[20:50] <liw> I guess I need to sit next to the hardy host for a bit
[20:50] <joaopinto> what should be the package for the livecd X dropping me on the shell without any notification on what happened ?
[20:50] <joaopinto> I am not sure it's xorg, since after install xorg does boot fine (using vesa)
[21:05] <joaopinto> uff, back into Intrepid :)
[21:06] <joaopinto> grub keeps adding the hardy kernel entries, any idea how it creates the kernel list ?
[21:09] <joaopinto> does it make sense to report a bug about an ATI newer video card not supported by Hardy ? Knowing it would require a driver upgrade that will not be provided at this time ?
[21:12]  * liw creates a hardy guest image on the hardy host, and moves it to the intrepid host... this should work
[21:23] <heno> liw: I'm running kvm installs on a headless box that I'm logged into via ssh; I use the -vnc option and run vinagre over an ssh -X session - if that helps
[21:24] <liw> heno, hm, that might have made it better, it is true, I should've rememberd that plain kvm doesn't do vnc but sdl
[21:24] <heno> the GUI updates fast enough to be usable
[21:24] <slangasek> is anyone here able to do ubuntustudio testing?  appears to require audio support in the test environment
[21:26]  * liw takes on ubuntu i386 dvd remaining three cases
[21:29] <heno> bdmurray, sbeattie: can one of you give studio a spin on real HW?
[21:30] <heno> if not, I can try it later on virtualbox, which supports sound
[21:31] <heno> should rescue mode allow me to mount an encrypted drive?
[21:31] <heno> It did ask me for the passphrase
[21:34] <sbeattie> heno: hmm, I somehow managed not to be rsyncing the studio images; I'll give it a go.
[21:34] <heno> sbeattie: thanks
[21:41] <cr3> heno: I apparently have an ubuntustudio auto install but it hasn't been showing on /builds
[21:43] <heno> cr3: that doesn't do any audio playing tests though right?
[21:44] <cr3> heno: no, but it enables me to install the whole thing with only 1 minute of my time
[21:45] <cr3> it's not rsync'ed in my lab though, where I could test with laptops. I'll start the download
[21:45] <cr3> a T1 doesn't seem to be optimal for a lab :(
[21:46] <cr3> anyone happen to have just a cdrom drive, not rw?
[21:47] <heno> cr3: where is it stored, on the DC satellite server?
[21:47] <liw> cr3, while cleaning, I found a plastic bagful of them actually. all broken. (why did I decide to keep them I have no idea anymore)
[21:48] <sbeattie> cr3: I do, why?
[21:48] <cr3> heno: yep
[21:48] <sbeattie> grumble, grumble, tightvncserver in intrepid is sadly unreliable.
[21:49] <cr3> sbeattie: if it's conveniently connected to a machine running hardy and above, I would appreciate the output of the commands: wodim -scanbus; wodim --devices
[21:49] <sbeattie> cr3: let me boot it first.
[21:49] <cr3> liw: thanks, but I wouldn't burden you with connecting those to a working machine
[21:50] <sbeattie> heno: which vnc server are you using>
[22:00] <calc> i made another failure, yipee :\
[22:00] <calc> ubuntu alternate i386 rescue mode fails to enter rescue mode for me
[22:00] <calc> the system did have a resized install so it had two copies of ubuntu on it, don't know if that matters
[22:06] <heno> calc: 'a shell was found but an error occurred while running it'?
[22:07] <calc> it popped up with a red message screen saying something to the effect failed to enter rescue mode
[22:07] <calc> i can try it again once the current install is done
[22:08] <heno> sbeattie: vino for displaying a running desktop, else I connect via XDMCP
[22:08] <calc> i got the failure twice though so i'm pretty sure something is wrong there
[22:09] <heno> sbeattie: I've not played much with tightvnc and friends
[22:09] <heno> calc: I got it first with an encrypted disk (see above) and just again with a regular disk
[22:09] <calc> heno: ok
[22:09] <heno> so I would agree
[22:10] <sbeattie> heno: oh, so you're duplicating the display on the headless machine?
[22:10] <heno> where do we file that, agaist d-i?
[22:10] <calc> heno: i haven't filed a bug about it yet, if you would like go ahead, no other failures had been reported yet so i wanted to make sure it wasn't just me
[22:10] <calc> heno: or ubiquity maybe, alert cjwatson to it once filed
[22:11] <calc> hmm actually ubiquity is the gui version right? so it would be d-i
[22:11] <heno> sbeattie: I log in to it with ssh -X and then run vinagre from there to look at the kvm sessions transmitting to its localhost
[22:11]  * heno grabs a screenshot and files
[22:12] <calc> heno: he said 'rescue'
[22:12] <calc> heno: apparently that is a package, heh
[22:12] <heno> oh, ok
[22:13] <sbeattie> cr3: hrm, I thought this machine had just a cdrom; turns out it has a dvd-rom (no writing ability); is that sufficient?
[22:16] <liw> if firefox starts crashing on anything containing flash after installing flash, does that mean the test fails? I wouldn't fail it, but what's the policy on this?
[22:17] <cr3> sbeattie: yeah, that's good too
[22:18] <heno> liw: anything=1 thing or anything=everything ?
[22:22] <liw> heno, at least three places
[22:22] <calc> anyone here have a double installed ubuntu i386 that they can try rescue mode on quickly?
[22:23] <heno> liw: I would flag it as fail
[22:23] <charlie-tca> calc, what are you looking for?
[22:23] <calc> charlie-tca: for me it fails to enter rescue mode, it is supposed to give an option of which install to change into
[22:24] <calc> i'm trying to verify it isn't just my test system that is screwed up
[22:24] <heno> 64-on-64 works fine
[22:24] <charlie-tca> I'll try it here. be right back
[22:25] <calc> ok
[22:25] <sbeattie> cr3: on hardy http://paste.ubuntu.com/61220/
[22:31] <cr3> sbeattie: thanks!
[22:43] <liw> finished the ubuntu amd64 upgrade test (finally), now I think I may need to go to bed, my eyelids keep closing
[22:45] <slangasek> 'night, liw!
[22:46] <liw> there's not a whole lot of untested stuff, too bad I can't keep myself awake until it's all tested (damn this anti-caffeine lark)
[22:47] <persia> liw, When not using caffeine, drink twice as much water (but sleep well this time)
[22:49] <charlie-tca> sorry about the time
[22:49] <charlie-tca> calc, it gives me a choice of all nine installations for root
[22:50] <calc> ok
[22:51] <charlie-tca> that is with i386, not 64-bit
[22:54] <slangasek> is anyone working on kubuntu amd64 dvd? looks like it still needs attention
[22:57] <sbeattie> hrm, the ubuntustudio cd is a tad oversized.
[23:00] <persia> sbeattie, It's not a CD.  Never has been.  I've tried to get it down a couple times, but always without much success.
[23:01] <persia> sbeattie, Should be able to burn a DVD with the iso, if you have the hardware.
[23:02] <sbeattie> I do. Was just amused, is all.
[23:03]  * sbeattie is punchy, amuses easily, and is reminded that he needs to finish making that second pot off coffee.
[23:04] <persia> sbeattie, Ah.  In that case, it is a CD, but you need a CD burner that can do double-sided.
[23:04] <slangasek> heh
[23:06]  * charlie-tca needs one of those new burners
[23:57] <heno> kub 64 dvd also pauses for a while on 5% of the ubiquity install, but otherwise seems ok
[23:58] <heno> (let's see if it actually resumes though)