[00:00] feel free, it's a wiki [00:01] oh, I hadn't realised the main release pages were [00:01] I meant http://www.ubuntu.com/testing/maverick/alpha3 [00:02] I can't change that either. Just work on MaverickMeerkat/TechnicalOverview [00:04] cjwatson: btw big thanks for the gparted update. While not having used the new version yet, I've experience amazing improvements after aligning my partitions to match my (cheap) hdd's 4096 sector size [00:05] *experienced* even [00:06] sistpoty: Is that an internal drive? [00:07] sistpoty: yeah, parted was updated last release, I hadn't realised until I saw the release notes that gparted needed a separate update [00:08] penguin42: internal as in I mounted it inside my case... WD15EARS-00Z [00:09] sistpoty: interesting that the 4k drives are starting to appears [00:09] they've been on their way for some time [00:09] nod [00:09] it'll be interesting to see how stuff works with >2T drives as well [00:12] penguin42: actually I got that drive since january, and have only now found out that the sector size was causing huge problems (I first blamed my other hdd, being confused about my partitions, then my new mainboard) [00:12] sistpoty: I guess lots of people have the same problem but don't realise it [00:13] penguin42: might be... I never thought that misalignment would have such negative net effects (e.g. firefox used to be unresponsive for about 20-30 seconds while updating its cache, I guess. That's completely gone now) [00:14] youch [00:18] fresh lucid installs should generally be fine, but of course not if you tried to be clever and did the partitioning with gparted beforehand :) [00:23] of course I have a developer system (wich is always weirds) and an installation that dates back to dapper, I guess (it'd be even earlier, but I reinstalled once I got amd64 hw) === fta_ is now known as fta [00:29] yeah, when I restored onto my SSD I took some care to create the partition table afresh rather than restoring it directly [00:30] does LVM do anything sensible on those drives? [00:31] it's supposed to have alignment handling; no idea how well it works [00:31] man pvcreate /alignment [00:32] (hm, one of these days I should make that actually work when typed literally) [00:32] penguin42: I've read that it automatically aligns to 4096 bytes, but I don't have a clue if it works (it's static data there in *my case*, problem with the 4096 sector size are writes FWIW) [00:33] sistpoty: Yeh - getting this all right in partition+LVM+fileystem+who knows what can probably be a nightmare; even with 512 byte drives making sure it all ligns up [00:33] lines even === yofel_ is now known as yofel [03:57] Hey guys, what should I do about a bug that I have provided a patch towards that has had no movement since May (when I posted the patch) (libvirt LP Bug #455832) [03:57] Launchpad bug 455832 in libvirt (Ubuntu Lucid) "segfault when attaching disk with same physical device" [High,Confirmed] https://launchpad.net/bugs/455832 [03:59] g: well, the ubuntu-reviews team is susbscribed, they'll get to it eventually, but you can ask someone in #ubuntu-reviews if they can look at it sooner [04:00] micahg: righty'o thanks [04:02] micahg: and I'm guessing the next thing to say is "We'd love help with reviewing other patches as well" right? :) [04:02] g: that's up to you :) [04:07] micahg: heh :) [04:36] G: have you tested the bug with lucid? [04:37] I have no clue how to test it :/ [04:39] micahg: I'm writing that long pending email to you now :D [04:51] nigelb: I'll work on it at our jam next weekend :) [04:51] * nigelb hugs micahg :) [04:51] nigelb: lucid yep [04:51] G: ok, and the patch fixes it? [04:51] nigelb: all you need is a VM that you can't care about the data on [04:51] In that case our normal action is to forward it upstream [04:52] nigelb: I'm yet to confirm if it's still present in latest upstream [04:52] well, libvirt is much newer in maverick [04:52] hmmm thats a good point [04:52] haven't tried libvirt in Mav yet [04:53] you can forwarding upstream... [04:53] * nigelb cannot make sense of it much. [04:53] please see if maverick is still affected [04:53] G: aha, so you're the nigel jones :) [04:53] micahg: just doing so now [04:54] nigelb: yeah [05:08] yeah, it doesn't apply to Maverick, it's fixed upstream, just lucid where it still happens [05:08] micahg: nigelb: ^, can't see anything in the Debian changelog, just going to get upstream git to see if they did what I did [05:09] G: if you can isolate a patch, maybe we can backport it for lucid. [05:11] upstream did the exact same thing, except used if (cgroup) instead of if (cgroup != NULL) which is basically the same http://libvirt.org/git/?p=libvirt.git;a=blob;f=src/qemu/qemu_driver.c;h=656a1e4a02c00ff6fb2ebc52ec3788d6d654f7f9;hb=92af69abad8b6531ea5d2dc16bfc36654c8f531f#l8313 [05:11] g: k, that's a good idea [05:12] G: aweesome, you can now go for sru since its fixed in maverick [05:12] !sru [05:12] Stable Release Update information is at http://wiki.ubuntu.com/StableReleaseUpdates [05:12] G: great, if you can grab the commit and make a debdiff, that would be the next step [05:12] micahg: so I should remake my patch to match what the libvirt guys did (assuming I can find the line where they did it) [05:13] g: no, see if you can find the specific commit and an upstream bug that's mentioned [05:13] ahhh right, I get ya [05:14] will do so after coffee [05:14] G; great :) [05:40] micahg: found the original diff & ML Post reg bug [05:42] \o/ [05:43] * G hugs the '$VCS blame' command :) [05:45] so with regard to SRU, I guess it qualifies based on the fact that it's a regression? [05:45] G: well, it was already accepted for Lucid, so you're good to post the debdiff [05:45] G: Make sure there's a good test case as well [05:46] G: and follow the rest of the steps of the SRU [05:47] micahg: if I was to review other libvirt bugs w/ patches would it be acceptable to bundle the patches (if I can verify them) into the same SRU? [05:49] G: idk, we should ask the SRU team since now they approve in the queue [05:49] jdong: around? [05:52] G, micahg: https://wiki.ubuntu.com/StableReleaseUpdates#Fixing%20several%20bugs%20in%20one%20upload [05:53] kklimonda: I know, but Idk if they thought about that with the new "review in upload queue" policy [05:53] kklimonda: thanks [05:53] in actual fact, there is just this bug that I'd be prepared to put forward [05:54] G, micahg: I'm pretty sure you can do that. If they NACK one of the fixes, you can just reupload, I'd assume [05:55] micahg: I'd say that we should follow policy and if sru-team want to change it they should update documentation. at worst they will just reject it and we'll know that documentation is out of date ;) [05:55] * micahg prefers to think tihngs out [06:10] G, micahg: as ebroder says, it's generally fine to bundle multiple fixes in a single SRU and upload to the queue; provided you're following the guidelines to begin with, the rate of SRU rejects out of queue is really going to be quite low [06:10] okay [06:10] slangasek: great thanks [06:10] jdong: unping [06:10] I think I'll just push this one, the others with patches I've not familiar with [06:29] So I'm starting to think about things like flights for UDS-N. I know robbiew said 10/25-10/29, but does that actually include weekends on either side? [06:33] ebroder: the developer summit is always M-F [06:34] Right, but should I plan on there being events on either end? [06:34] wrt UDS, no :) [06:35] you might not want a flight too early on Saturday morning since Friday night tends to be big for socializing [06:35] however there may happen to be other developers present with whom you want to spend time at a bar the sunday before [06:35] or that [06:35] or...ya know... DISNEY! [06:36] maco: I have absolutely no intentions of being in Orlando *without* going to Disneyworld. I don't understand what the point would be [06:36] * maco pouts at the U1-4-KDE client [07:11] yes, what better epilogue to a free software conference than to pay tribute at the altar of perpetual copyright ;) [07:14] :-) [07:15] Ubuntu doesn't have awesome roller coasters :-P [07:16] slangasek: i was thinking go to disney first that way by being at uds you make amends [07:16] like penance, but fun [07:16] heh [07:16] well... funner than 10 hail mary's [07:17] I've been to Disney World, don't really need to go again :) [07:17] “Write 100 lines of free code as a penance” [07:17] and I had the excuse that I was a child ;) === Lutin is now known as Guest87666 [07:17] Eh. I don't care how old I am. I'm still a kid when it comes to theme parks [07:18] heh i was 8 when i went i think... but that was before they got all the themed areas and stuff... is the CA or FL one the one that has the new Hogwarts castle? [07:18] eh, haven't heard of that one [07:18] maco: That's Universal Studios Orlando [07:18] oh thats close! [07:20] the next uds is in orlando? [07:20] yes [07:21] sweet [07:26] any pointers for debugging a partly working usb wifi dongle? === lucidfox_ is now known as lucidfox [07:27] I'm not 100% certain if I'm using the correct driver, but it's certainly a realtek chip [07:40] can I update a package while I've got a submitrequest pending? === Jordan_U_ is now known as Jordan_U [08:46] !regression-alert latest linux -security update in hardy breaks -xen flavour [08:46] Error: I am only a bot, please don't think I'm intelligent :) [08:46] !regression-alert [08:46] cjwatson, jdong, pitti, slangasek, ScottK, mdz, kees, ttx, marjo, seb128: reporting regression in a stable release update; investigate severity, start an incident report, perhaps have the package blacklisted from the archive [08:46] lool: you're looking for the | key [08:47] maco: Thanks, will know for next time [08:51] LP #620994 [08:51] Launchpad bug 620994 in linux (Ubuntu) "xen kernel bug: 'kernel BUG at /build/buildd/linux-2.6.24/debian/build/custom-source-xen/mm/memory.c:2704'" [Critical,Confirmed] https://launchpad.net/bugs/620994 === _martin is now known as m4rtin [09:24] apw, lag, ^^^ [09:30] mdz: I've escalated to security@u.c, Cc:ing Stefan and Pete; do you think this requires further escalation (I think security@u.c is read quickly)? [09:30] mdz: I've pinged, but not phoned elmo === bilalakhtar is now known as bilalakhtar_ [10:00] lool, given it's a universe package, no, I don't think so [10:01] lool, You always have the alternative of uploading a reversion, if you feel strongly enough about it. [10:03] persia: Since it's a bunch of security updates, and all hardy users would get the updates, I'm not tempted to decide by myself, I'll leave it to security team who needs to approve the uploads anyway [10:04] OK. Sometimes I find that on the weekend, if something bad happens, it can be better to just fix it (and let it get fixed right later), because some people aren't about. [10:04] But stable security updates end up being all sorts of special :) === Guest87666 is now known as Lutin [10:24] james_w, is it just me, or is lp:ubuntu/pm-utils out of date (it seems to contain 1.3.0, while maverick has 1.4.1-1) [10:29] hmm, lp:ubuntu/flashplugin-nonfree is also out of date [10:30] mdz and micahg: See http://package-import.ubuntu.com/status/ [10:31] Laney: does that include uploaded packages, I don't see anything for flashplugin-nonfree except for the Debian import failure [10:32] I expect it's all part of the same thing [10:33] micahg: it might be because the Debian package didn't get imported the Ubuntu uploads don't get either anymore as this would break the history later [10:34] geser: hmm...why? aren't they different branches? [10:35] Ubuntu uploads are branched from the corresponding Debian revision [10:35] That's how merges work. [10:35] hmm...well, this package is on the blacklist list [10:35] And trying to fix history later would be exceedingly painful, given how bzr tracks things. [10:36] doe that mean I shoudl just use a debdiff [10:36] instead of UDD [10:36] It's not really because of how bzr tracks things. But yes, all existing branches after that point would be broken. [10:37] micahg: the package importer probably doesn't know that we forked flashplugin-nonfree [10:37] Hrm? I thought that recovering history was bad because of bzr's (current) inability to keep track of files by name and contents, rather than identity (and that this was a target to be fixed). I agree it would be painful anyway, but this makes it more so. [10:38] * micahg is just going to use a debdiff for branches that won't update [10:39] persia: Regardless of how files are tracked, you'd still have to rewrite every later revision (with a new rev ID) if you wanted to add a merge into the history. [10:39] Which then makes existing branches incompatible. [10:39] True [10:39] Well, incompatibility has to do with lack of understanding contents, but needing to update everything is annoying. [10:39] Somewhat. [10:48] hmmm...weird, UDD just overwrote my 2 revisions (merge/release) that I used to update a package [10:48] had you done a mark-uploaded when you uploaded, and pushed that ? [10:49] lifeless: no, I must have missed that step on the instructions [10:51] ah, it's in the upload document, not the merge document... [10:52] I'll remember for next time [11:09] ne one trued ubuntu lts??? [11:10] nisiop, that's not english [11:11] lol [11:11] has anyone tryed ubuntu LTS eddition [11:12] i'd guess quite a few people, given it's the current stable ubuntu release. but i'd also guess that your question isn't really an #ubuntu-devel question [11:12] Heaps and heaps of folk. You may find some of them happy to answer questions about it in the support channel (#ubuntu). At this point, most of the developers are focused on the next release. [11:20] man copers are cranky [11:20] coders* [11:21] ne one help me set up a domain useing bind?? does that count [11:22] * nisiop finds another room === unimatrix_ is now known as unimatrix [11:24] i doubt i would have helped him even if it were #ubuntu, the inability to spell would send me into a flying rage within minutes [11:24] srsly??? [11:24] YA RLY [11:27] hehe :P [11:31] could someone please retry https://edge.launchpad.net/ubuntu/+source/junit4/4.8.2-1 , it builds fine now [11:37] * shadeslayer wants a sponsor too :P [11:37] shadeslayer: bug #? I know I cannot sponsor, but will just have a loo [11:37] s/loo/look [11:37] no bug :P [11:37] shadeslayer: branch? [11:37] koffice 2.2 [11:38] from : https://edge.launchpad.net/~rohangarg/+archive/experimental/+packages [11:38] they had a bug fix release [11:38] ah, this is outside my scope [11:39] :) [11:39] i asked aplg but he doesnt have the bandwidth :P [11:41] if anyone has the time .. please :) [11:42] shadeslayer: try asking on #kubuntu-devel [11:42] bilalakhtar: already did.. everyone is sleeping [11:42] im there most of the time anyway === almaisan-away is now known as al-maisan === al-maisan is now known as almaisan-away === jelmer is now known as Guest99497 === Guest99497 is now known as jelmer === yofel_ is now known as yofel [15:32] anybody knows where does sudo source it's envronment vars from? [15:38] cjwatson, jdong, pitti, slangasek, ScottK, mdz, kees, ttx, marjo, seb128, lool: I've commented in bug #620994 [15:38] Launchpad bug 620994 in linux (Ubuntu Hardy) "linux 2.6.24-28.75 breaks xen flavours (xen kernel bug: 'kernel BUG at /build/buildd/linux-2.6.24/debian/build/custom-source-xen/mm/memory.c:2704')" [High,Confirmed] https://launchpad.net/bugs/620994 [15:39] in essence it is my opinion that we should not yank the patch as it would reintroduce a serious vulnerability with a known exploit (there is a paper on it) for all hardy users [15:40] the kernel team should be online to look at this completely in 15 hours or so [15:40] while unfortunate, I think pulling the patch would affect more people than leaving it in [15:41] and xen users can always boot into their previous kernel === glick is now known as Guest8937 [16:10] Sigh, I wonder if my internal TI card reader will ever work properly [16:12] MattJ: What does it do? [16:12] Aug 22 16:01:40 silver kernel: [ 7079.424137] tifm_sd0:1 : card failed to respond for a long period of time (12, 1) [16:12] * penguin42 thinks the TI one in here generally works [16:12] and then gazillions of other syslog messages about IO errors [16:13] It's been a problem for me since, hmm, Dapper I think... but it's intermittent [16:13] Though it's puzzling, I think it started working (at least more reliably) in one release [16:13] and when searching for bug reports, it seems that's consistent [16:14] yeh PCIxx12 working on Maverick here - just tried [16:14] But now I'm having it again, so perhaps it's a regression... I can't find any recent reports of it [16:14] MattJ: My experience is they can be very touchy with regard to the particular card [16:14] I have: 07:06.2 Mass storage controller: Texas Instruments 5-in-1 Multimedia Card Reader (SD/MMC/MS/MS PRO/xD) [16:14] I've used this card in here before [16:14] Hmm, actually maybe I didn't [16:15] MattJ: Same here, lspci -n shows it as 07:06.2 0180: 104c:803b [16:15] I could have given up and used my (slow) camera via USB [16:15] Oh well [16:18] yeh I've been tending to do that, my new machine doesn't have an SD card slot, and the external USB card reader doesn't seem to like the current 2GB card I have [16:20] Actually I think I'll borrow another laptop and network it, that'll save flattening my camera battery :) [16:20] Now the cards are bigger I just don't empty them until they're full :/ [16:43] I do tend to like to copy any particularly nice stuff off before I have the chance to screw it up [16:44] jdstrand: thanks a lot for commenting [16:54] can someone take a look at bug 74065? [16:54] Launchpad bug 74065 in vmware-player (Ubuntu) "apt-get remove vmware-player doesn't remove vmware-player service" [Undecided,Confirmed] https://launchpad.net/bugs/74065 [16:54] OLD bug [16:54] Yes, but it has a patch. [16:55] I'd like to know what they think about symlinks is the right. [16:55] nigelb: That's the wrong way to fix the bug, though. The right thing to do is for the init script to check if the package is still installed at the beginning [16:55] the normal solution to that is to check for the existance of the binary on startup [16:56] nigeld: Look at /etc/init.d/ssh; at the beginning it does "test -x /usr/sbin/sshd || exit 0" [16:56] Ah. but symlinks getting removed only at --purge is normal? [16:57] Yes [16:57] Ok, thank you :) [17:09] cjwatson: do you have comments thoughts to add on bug 110144? [17:09] Launchpad bug 110144 in debootstrap (Ubuntu) "No lo interface after debootstrap install" [Wishlist,Triaged] https://launchpad.net/bugs/110144 === Guest8937 is now known as glick [18:47] hi, just a question: will the 10.10 have onscreen keyboard preinstalled by default [18:48] just wanted to point this out, since i ran into troubles one day using 10.04, i wasn't able to find it [18:50] you can actually login to system with mouse, but can't use it after that, or atleast i didn't find onscreen keyboard anywhere [18:55] not sure, the accessibility stuff is normally in by default isn't it? [18:56] well, i double checked this.. as the usercase with me was with a server that i installed a while ago [18:56] i just installed 10.04 from scratch to virtualbox and i can't find onscreen keyboard in menus [18:57] gdm has own onscreen keyboard, but gnome doesnt [18:57] that's a bit weird [18:57] yes [18:58] Interestingly I've just checked, on here the only one I have is the KDE one under kvkbd [18:59] and character map isn't useable for that [19:00] onscreen keyboard used to be included by default [19:00] i wonder when it got removed? [19:00] not unless you primarily want weird symbols [19:00] visual system bell for deaf users was removed from gui in karmic, but i have a patch to add it back to mav [19:01] * maco goes to say something in #ubuntu-accessibility [19:01] well, i had to make compromize with the server machine i was working with, either mouse or keyboard with memoristick, apparently i wasn't able to get into system menus without mouse.. [19:01] yeh that sucks, should be able to get to most stuff via keyboard [19:02] reaby: cant get into system menu without mouse? [19:02] what menus cant you open? [19:02] well i didn't find a keyboard maping to applications menu [19:02] alt+f1 [19:02] ach [19:03] lol [19:03] :) [19:03] same in both gnome and kde [19:03] ok, good to know [19:03] thats set in the "keyboard shortcuts" window as "show the panel's main menu" [19:08] hmm.. you guys prolly don't like the idea for using now-a-days standard keyboard start-button for opening applications menu? [19:09] "standard?" the keyboard i am using does not have such a button. [19:09] in terms of usability [19:09] :/ [19:10] not everyone has a pc105 [19:10] also, the shortcuts dialog wont let you set without another key [19:10] * penguin42 hugs his Model M [19:10] modifier keys have to have a normal key with them [19:10] penguin42: ticka ticka ticka [19:10] although I have some sympathy for using super for some things, as long as I can switch away [19:10] mneptok: No, much much louder :-) [19:10] TICKA TICKA TICKA [19:12] penguin42: http://pckeyboards.stores.yahoo.net/linux101.html [19:12] mneptok: Nod, I've got 3 standard ModelM's (UK 102 key) - one here, one at work and a spare [19:13] penguin42: Unicomp is still making them. and labeling keys for Linux users. :) [19:13] well sorry for mentioning :( [19:13] mneptok: Yeh [19:13] reaby: as far as i know, that key is only standard on windows anyway [19:13] reaby: when you hit the apple key on a mac, it does not open the menus [19:13] reaby: no need to apologize. but it's not safe to make assumptions about the hardware "most people" have. [19:14] reaby: No I have some sympathy with the use of the extra key; at some point I do have to acknowledge I'm a weird grey beard :-) [19:14] yuck i wouldnt use that [19:14] ctrl's in the same spot as those awful sun keyboards [19:15] now, if they made a *vi* keyboard... you know, with esc next to A...i'd be happy [19:15] What prevents you from mapping it there? [19:15] i do map it there [19:15] the key's just labeled wrong when anyone else tries to use my computer [19:16] maco: they tried. but the vi(m) community responded with "we make our own keyboards by specifying Chinese plastic fab factories in .vimrc. kthxbai." [19:16] they hit caps lock, and the window closes :P [19:16] I use it as ctrl, though. And yeah, i’m a vim user. [19:16] maco: Oh, HP used to do one like that, I used one at college - it had it just next to left shift [19:16] pity the rest of the layout was so weird [19:17] but thanks for listening a random users note from this issue [19:18] chrisccoulson: had some time yet for the gnome-power-manager bug ? :) [19:20] dupondje, i might look at it later, but it's still the weekend ;) [19:20] i generally try to do as little work as possible at the weekend ;) [19:20] bye [19:40] chrisccoulson: no problem :) you saw its also an issue with other drivers then nouveau? Or is it not nouvea-related ? [19:41] dupondje, it affects any driver that doesn't support XBACKLIGHT [19:42] ok :) [19:42] seems like almost all drivers ? [21:47] Sarvatt: around?