=== jalcine_ is now known as jalcine === jalcine is now known as JackyAlcine === JackyAlcine is now known as jalcine [04:14] Good morning [04:15] bdmurray: ubuntu-bug does accept executable paths, but only the full one; it does not search $PATH [04:16] slangasek, Sweetshark: LibO did use it, but it seems not any more? [04:43] @pilot in === udevbot changed the topic of #ubuntu-devel to: Precise Beta-1 Released. Archive: open | Development of Ubuntu (not support, not app development) | build failures -> http://bit.ly/xmGdCW | #ubuntu for support and general discussion for hardy -> oneiric | #ubuntu-app-devel for app development on Ubuntu | http://wiki.ubuntu.com/UbuntuDevelopment | See #ubuntu-bugs for http://bit.ly/lv8soi | Patch Pilots: pitti [04:43] Morning [04:52] hi pitti :), how long are you around for? [04:52] micahg: today? for about 30 mins [04:52] pitti: oh, ok [04:52] couldn't sleep any more, my brain told me all the stuff I still need to do today before I get sedated :) [04:53] * micahg will need some copies in a few hours, but will find another AA [06:21] stgraber: I filed LP bug #959037 about not starting the local resolver if network-manager detects that a dns server (bind9, dnsmasq) is already installed [06:21] Launchpad bug 959037 in network-manager (Ubuntu) "Don't start local resolver if a DNS server is installed" [Undecided,New] https://launchpad.net/bugs/959037 [06:35] pitti, beer and all green in the smoke tests this morning [06:35] ! [06:37] rickspencer3: yep, sorted out the remaining bits on Friday evening :) [06:38] pitti, nice work, will put us in a good position for beta freeze [06:38] pitti, was it hard to get the archive problems sorted? [06:40] rickspencer3: no, not particularly [06:40] just some discussion involved, and then pressing a few buttons [06:40] thanks pitti [06:41] http://people.canonical.com/~ubuntu-archive/component-mismatches.txt needs some work still [06:41] apw, janimo`: are linux-{ac100,n900} really supposed to get into main? [06:44] infinity: d-i FTBFS on powerpc, presumably those kernel images went away in favor of -smp? [06:46] pitti: Yeah, we're holding off on resolving that while we decide if we need to resurrect powerpc. [06:46] pitti: (ppc-smp doesn't boot on my UP G3, at least, so we may need to revert the mess) [06:46] pitti: And linux-ac100 and n900 are *not* main kernels, no. [06:48] ^ that's what I thought [06:48] http://people.canonical.com/~ubuntu-archive/component-mismatches.txt says open-iscsi-udeb and others now want them [06:49] pitti: Seems like a virtual package oops, maybe? [06:49] pitti: Confused resolver... [06:49] it's fairly new on c-m [06:49] pitti: Yeah, but neither of those kernels are new, so the fault lies elsewhere. [07:12] pitti: AIUI, we don't want keywords for random universe packages just like we don't want the quicklists [07:13] or rather don't want ATM [07:18] quicklists need translations, but keywords not necessarily? [07:18] but it was forwarded upstream, so it seemed harmless enough [07:18] but I see that it's debatable [07:19] ok, http://reports.qa.ubuntu.com/reports/sponsoring/ is down to 13 now, and that's about as much as I'm comfortable with sponsoring [07:19] and I need to disappear anyway [07:20] pitti: I'll take a look at the rest in ~20 hours when I pilot [07:21] pitti: BTW, that was something else I saw the conversation about, not quicklists, so I might be confused, I'll check with seb128 when he gets online [07:23] micahg: there's two or three SRUs which can still be sponsored, and the grub2 stuff if you feel up to it [07:23] but there's a couple which still need some baking/upstreaming/fixing in precise first [07:24] @pilot ot [07:24] (pilot (in|out)) -- Set yourself an in or out of patch pilot. [07:24] @pilot out === udevbot changed the topic of #ubuntu-devel to: Precise Beta-1 Released. Archive: open | Development of Ubuntu (not support, not app development) | build failures -> http://bit.ly/xmGdCW | #ubuntu for support and general discussion for hardy -> oneiric | #ubuntu-app-devel for app development on Ubuntu | http://wiki.ubuntu.com/UbuntuDevelopment | See #ubuntu-bugs for http://bit.ly/lv8soi | Patch Pilots: [07:24] pitti: ok, I'm sure there will be a few more by the time I get around to piloting [07:24] heh, or that :) [07:24] pitti: you have time for a quick copy? [07:25] micahg: really need to run now, sorry [07:25] doctor appointment [07:25] pitti: np, I'll find someone else, thanks [07:25] off for today, sick day [07:30] bye pitti, hope you feel better [07:30] wow, down to 13 items in the patch pilot queue? [07:30] what a day [07:32] StevenK: are you still around to do a copy from a PPA? === tkamppeter_ is now known as tkamppeter [07:34] micahg: If you ask nicely, I might be able to do something for you. [07:35] rickspencer3_: dholbach's shaming has had an effect. [07:35] infinity, good ;) [07:35] No one likes Daniel upset with them. [07:35] infinity, combined with 0 archive problems and all smoke tests passing ... we're in good shape [07:35] It's like accidentally kicking a puppy. [07:36] I feel chuffed [07:36] * infinity shoves a few archive problems under the rug. [07:37] rickspencer3_: And yeah, +1 maint really does seem to be doing the trick. I'm curious to see how well we can do during one of the "insane" cycles, rather than during a conservative LTS. [07:37] * infinity has fond memories of edgy never actually being usable. Like, ever. === rickspencer3_ is now known as rickspencer3 [07:38] micahg: Last chance before I go and get some sleep (yes, I really do sleep sometimes, honest). [07:39] infinity: ah, sorry, I stepped away for a minute, still available?/ [07:39] micahg: No, I left in the 15 second gap there. [07:39] micahg: What's up? [07:39] infinity: xulrunner-1.9.2/lucid from ubuntu-mozilla-security to lucid-security [07:40] micahg: Just that? [07:40] yeah, I still have a wee bit more testing on the other stuff [07:40] Err, wait. [07:40] I copied maverick and natty already, lucid timed out [07:40] To security, not to proposed? [07:40] infinity: yes [07:41] Isn't there some USN magic that should happen with that? [07:41] infinity: we have tools to unembargo, but they timeout when there are a lot of binaries [07:41] Oh, or you just need manual intervention due to lp API lolz. Check. [07:43] 2012-03-19 07:43:06 INFO 49 packages successfully copied. [07:43] Confirm this transaction? [yes, no] yes [07:43] 2012-03-19 07:43:10 INFO Transaction committed. [07:43] 2012-03-19 07:43:10 INFO Done. [07:43] micahg: Enjoy. [07:43] infinity: thanks [07:53] good morning === smb` is now known as smb [08:32] slangasek, hi [08:41] infinity, if you are still concious there are powerpc kernels here: http://people.canonical.com/~apw/master-next-precise/ [08:53] pitti, linux ac100 doe snot need to be in main as far as I know, we have been making images for ac100 just fine so far. Unless tooling or policies have changed it can stay a simple community supported image seeded from universe [09:19] ev, when woopsie is at the 'you had a problem' stage and has yet not collected details, does it have the binary name of the crash? i would quite like to be able to see that before details is collected if so, and am suggesting perhaps making it a tooltip on the detail button or something subtle === fenris is now known as Guest97336 [10:39] uf. according to my script we have 4717 packages without -dbgsym ones [10:39] how many of those have a -debug package? [10:39] http://paste.ubuntu.com/890445/ is list and http://paste.ubuntu.com/890445/ is script if someone is interested [10:40] geser: -dbg check is my next step. but amount of packages may be lower as I have some external ppas [10:40] hrw, how many of those are not candidate to having a debug? [10:40] hrw, like a shell script [10:41] seb128: many probably - have to go though list and improve script [10:41] hrw: and you linked the same paste twice [10:41] hrw, there is no reason we would miss dbgsym out of publishing bugs (which happen) [10:41] ops [10:42] geser: 890444 [10:42] hrw, basically anything calling dh_strip (i.e stripping binaries) should have a dbgsym [10:42] I know [10:42] or ones which call pkg-create-dbgsym by hand (like binutils) [10:43] hrw, what are you trying to get at? [10:44] seb128: I am working on creating ubuntu based sysroots for Linaro's binary toolchain. Making sysroot imge with -dev packages is quite easy. problem starts when someone also wants debug symbols [10:44] seb128: so based on one of my scripts I wrote extra one which checks fo packages without dbgsym support [10:44] hrw, your script is buggy [10:44] hrw, like gnome-session is listed but it has ddeb support, http://ddebs.ubuntu.com/pool/main/g/gnome-session/ [10:44] seb128: yes, it is - fast written [10:45] why there is an symbol link "ubuntu" in archive.ubuntu.com/ubuntu to .? [10:45] hrw, so what's the point of making boggus number claims on that channel? ;-) [10:46] hrw, I'm still not sure what feedback you were aiming for claiming that 4700 "packages" don't have a dbgsym there [10:46] seb128: sorry, just wanted to share yet-another-rather-useless-for-most thing [10:47] hrw, well it would be useful if you didn't come with boggus numbers ;-) it seems like a finger pointing on Ubuntu or something [10:47] sorry, does not planned to offend anyone [10:47] hrw, it would be useful if you came with a list of packages than should have a dbgsym and don't rather [10:48] seb128: will post such one during this/next week [10:48] hrw, rather than coming stating "css themes and shell scripts don't have a dbgsym"... sure what would be a dbgsym for a css ;-) [10:48] ;) [10:49] seb128: thank you for it - added notes [10:49] hrw, thanks, looking forward having a reasonable list of what is buggy and we should fix ;-) [10:52] seb128: css themes and shell scripts are usually 'architecture:all' so I skip them. but there are some false negatives on my list still [10:55] hrw, well you have weird stuff like hundred of language-support packages [10:55] hrw, and you have bugs, i.e gnome-session is listed but it has a dbgsym [10:56] hrw, oh, i.e gnome-user-docs is listed but it builds only gnome-user-guide which is arch all and documentation [10:56] hrw, so your arch all filtering is somewhat not working I guess [10:57] seb128: I started next run with some fixes [11:00] is it normal to have 65 "/usr/lib/cups/notifier/dbus dbus://" processes running? [11:00] tkamppeter, ^ :) [11:04] dropping oneiric from apt config helps ;D [11:09] mvo, poing [11:14] dholbach, no. [11:14] tkamppeter, what could be the cause for having that many? [11:16] dholbach, never heard about that. Perhaps you should report a bug and I will forward that straight upstream. [11:16] dholbach, who is dbus expert here? [11:16] will do [11:18] tkamppeter, a lot of people know about dbus in here [11:19] dholbach: i see this on two of my systems (52 instances on one, 100 on the other - each ~1.7mb) [11:19] https://bugs.launchpad.net/ubuntu/+source/cups/+bug/959195 [11:19] Launchpad bug 959195 in cups (Ubuntu) "65 cups notifier processes running" [Undecided,New] [11:19] tkamppeter, htorque: ^ [11:19] you might want to subscribe to it or at least "affects me too" [11:20] done and done [11:22] cool [11:47] ok, 2568 packages to check [11:47] but later [11:51] htorque: could I pm you? [11:51] Can* [11:51] vibhav: sure === MacSlow is now known as MacSlow|lunch [12:03] hi all! to complete an answer on ask ubuntu, i'd like to know if using one's real name is mandatory for code contributions in general and for signing the 'Canonical CLA' in particular. [12:04] apw: bug 938707 [12:04] Launchpad bug 938707 in apport (Ubuntu) "Hard to see which program crashed for "Internal error" reports" [Medium,Fix committed] https://launchpad.net/bugs/938707 [12:05] ev, ta [12:11] htorque, i would imagine that as long as the email that the 'signed' CLA was sent by and the contributors LP email are the same it should be ok [12:11] htorque, i'm not an authority though === _salem is now known as salem_ [12:12] htorque: do you have an exact example? [12:14] sladen: for example, "Duffy Duck " contributes to one of the projects that require signing the CLA - is that possible? [12:15] brendand: thanks, do you know who to contact for getting a more official answer? [12:15] htorque: ultimately you can put anything you want on the Commit ID [12:15] htorque: the key comes to getting somebody to accept that [12:16] htorque: which means the person-accepting-it needs to be sure that they are allowed to do so [12:16] htorque: eg. that acceptor needs to be happy that a CLA exists that covers that contribution [12:17] htorque: and that the CLA is valid [12:17] htorque, i think in our environment the only secure way to confirm someone is who they say they are is that they can access the same email account for two purposes [12:19] sladen: but are there any legal problems when signing the CLA with anything else than your real name? [12:19] *than/but [12:19] htorque: what's actually going to happen in practice, so that you'll probably push a proposed branch (using Bzr-over-SSH) to launchpad.net/~htorque/+junk/proposed-patch-for-bug-1234 [12:19] is that [12:20] htorque: and somewhere separate, that account will be on the ticklist for accounts that have signed the CLA [12:20] htorque: and when filling out the CLA you will have put in sufficient details to allow you to be contacted [12:21] htorque: (I perhaps understand your problem to some extant, as having a _boat_, that doesn't have a fixed address of its own) [12:21] bryceh: right, no sense in uploading it if we know it's going to FTBFS - I just thought you were uploading it already before that. no matter :) [12:22] pitti: as I understand it, the libO that uses it has not yet been uploaded [12:22] tkamppeter: hi there [12:22] htorque: if you can convince somebody else that you're called Duffy Duck, you can give it a shot [12:23] htorque: though since there are several humans in the chain, one of them would likely spot and double check it [12:23] htorque: in which case you'd probably have to pursuade a lawyer that you're called Duffy Duck [12:24] htorque: OTOH, if it's tied to your Launchpad and there a GPG key on there signed by 20 Debian Developers confirming you _are_ known as Duffy Duck then that would probably solve it :) [12:25] /win 6 [12:26] htorque: as those people are lending a small part of their own credibility into reenforcing that identity [12:26] sladen: yeah, i guess it's hard to get it signed by even one person. ;-) so, to summarize: if there's no CLA, than it's basically up to the one accepting the contribution. if there is a CLA, it depends on the requirements of that CLA? [12:28] htorque: there are a number of Canonical back projects that require CLA to contributions, in order to allow them to be relicensed in a flexibly and legally-sound way. Getting people to sign CLAs is _not_ generally hard [12:28] Canonical-backed projects [12:29] htorque: (this does not stop you from contributing to Ubuntu) [12:29] htorque: and is somethere for some of the upstream projects that Canonical steers [12:30] htorque: and is only there for some of the upstream projects that Canonical steers [12:30] sladen: but there are people in the ubuntu community that prefer to not use their real name for any activity. they are definitely ruled out by the canonical CLA? [12:30] as far as i am aware, any real names that are on signed contributor agreements are not made public [12:31] htorque: I don't think anyone stops that [12:31] vibhav: but the signed contribution itself is [12:31] oh yes [12:31] hmm [12:32] htorque: yes, Canonical care about having a legally-sound basis for the projects they steer. If you're wanting to contribute to a Canonical Upstream project then you need to do that (with Canonical) in a legally sound way [12:33] htorque: the options are generally: (1) an employment contract (2) a contracting agreement (3) a legally-sound CLA [12:33] i see. the Canonical CLA seems to require: first name, last name, email, phone, address, country [12:35] htorque: if you have some specific situation (eg. war-torn country, stalking issues, living a boat without a fixed address), then we could look into it. But both yourself and Canonical would have to be happy that the result was a legally-sound understanding entered into in good faith by both parties [12:36] htorque: (and not that this is only about contributing code upstream, _to_ Canonical upstream projects. It doesn't apply to eg. Ubuntu, or to forked your own versions of Canonical upstream projects) [12:36] and note [12:37] sladen: thanks, that pretty much answers the question (not "my" question, it's on ask ubuntu → http://askubuntu.com/questions/112434/is-there-a-real-name-policy-in-the-ubuntu-community). [12:37] htorque: (and if you'd prefer to discuss this in private, I'm happy to do that aswell) [12:37] htorque: ah ha. Well in that case, I can just answer there then : [12:38] sladen: vibhav won't like that (due to the bounty going on) :P [12:38] he basically answered all the other points, just the code contribution was left. [12:39] htorque: :( [12:40] htorque: vibhav: well, I _hope_ the aim is to get a high-quality answer up there, not who gets the prize [12:40] sladen: sure [12:41] htorque: vibhav: PS. next time you've got a similar query, could you try and post the link giving the high-level summary _first_, not _last_! [12:43] sladen: right, sorry. i just mentioned ask ubuntu - forgot the link. [12:43] me too [12:57] pitti, we hold ddebs 14 days from when the master binary packag [12:57] package is reaped right ? [13:04] sladen: I think it's a mis-statement to claim the CLA is in any way required for Canonical to be "having a legally-sound basis for the projects they steer". If that were true, Ubuntu would be legally unsound. It's not. [13:05] htorque: There's no real name policy in Ubuntu. [13:07] ScottK: but then, why require a CLA at all? [13:07] htorque: Ubuntu doesn't require one. It's other Canonical projects that do. [13:07] (if not for legal reasons) [13:07] htorque: It's so that they can also ship code in their other projects under a proprietary license if they want to. [13:08] So it's not needed for any legal reasons related to Ubuntu. [13:08] It may be needed for legal reasons for other proprietary stuff. [13:08] I guess that's a more correct way to put it. [13:09] ScottK: however, it's needed for the contribution to be accepted (in ubuntu) in the first place? [13:09] htorque: Not in Ubuntu, just those other projects. [13:09] ScottK: Jumping immediately to proprietary is a bit unfair, it's for ease of relicensing, period. [13:11] ScottK: but then the code cannot be merged into the upstream project and ubuntu would need to carry a patch around, right? [13:11] infinity: True, but they could solve a whole world of trouble by just saying "sign the CLA or submit your changes with a BSD like license". [13:12] htorque: For those projects that's true, but there's a huge amount of Ubuntu that's nothing to do with them. Also Ubuntu has carried patches for such projects in the past. [13:12] ScottK: I suppose the FSF could do that too, but they don't. [13:12] ScottK: that is basically what the CLA (now) says [13:13] (CLA != CCA) [13:13] Yes, but the FSF isn't a for profit company. Even if I trust Canonical, I can't assume they won't eventually be bought (cf Sun/Oracle), so I think the comparison with the FSF is irrelevant. [13:14] ScottK: charities can make a profit, and companies can make a loss. Why would a random label of "for profit" have anything to do with it [13:15] sladen: OK. That's shorthand for a corporation. At least in the US, non-profit entities can't be bought and sold. [13:16] There are not for profit corporations, so the for profit isn't 100% correct, but they are rare. [13:16] ScottK: occasionally like-minded people with a cause in common club together to form an organisation of some sort. Be it familes, or charities, or foundations, or companies, PLCs, LLPs, amateur radio clubs, canoe clubs etc === bladernr_afk is now known as bladernr_ [13:19] Charities can be usurped, such as companies can [13:19] In the US, these are formed under section 501(c)3 of the tax code. They have a governing board and a charter to do certain things to which they are limited. [13:20] I haven't read the FSF charter, but I expect it includes some reference to free software, so I believe there would be some legal reprocussions if somehow it's board were packed and it's policies flipped to be proprietary friendly. [13:20] I'm not saying it's impossible, but it's an entirely different thing than a corporation being sold. [13:20] The latter is a normal, expected thing. [13:25] janimo`: ok; not really sure why it's on c-m right now [13:25] apw: yes === MacSlow|lunch is now known as MacSlow [13:28] sladen, ScottK: who would you need to talk to if there were any problems with signing the CLA? the project's contact listed at http://www.canonical.com/contributors? just want to add that to the existing AU answer (unless sladen is going to add his own answer). [13:28] * ScottK has no idea. [13:28] htorque: My main request is that you not say Ubuntu requires the CLA. It doesn't. [13:29] ScottK: yeah, sladen made that clear already. :-) [13:30] OK. Just making sure. [13:33] htorque: if in doubt, you can post the query to me [13:33] htorque: I think my name is still in the CLA PDF metadata ;-) [13:33] htorque: and wendar will be able to give some detailed insight too [13:35] Hi guys, just wondering was ubuntun accepted on the gsoc 2012? [13:38] didrocks, seb128: the patches from https://bugzilla.gnome.org/show_bug.cgi?id=671780 are not yet part of libgnome-desktop-3-2, right?! [13:38] Gnome bug 671780 in libgnome-desktop "gnome-bg: improve "representative colour" algorithm" [Enhancement,Unconfirmed] [13:38] DiegoTc: It was not. [13:38] MacSlow, didrocks: they are in precise [13:38] (missed the application deadline by 10 minutes) [13:38] didrocks, seb128: sorry... wrong bug URL... one sec [13:38] MacSlow, didrocks: I backported them [13:39] MacSlow, I know what bug you talk about, that's in precise for a week [13:39] seb128, didrocks: ah ok... thanks [13:44] htorque: vibhav: ScottK: is http://askubuntu.com/questions/112434/is-there-a-real-name-policy-in-the-ubuntu-community/114250#114250 any use? [13:45] sladen: thanks. are you going to add an answer to that question (the bounty is still open for three days), or do you want me to edit it into an existing one (or do it yourself)? it's really up to you if you want to go for the reputation. :-) [13:45] sladen: ah [13:45] htorque: There's no requirement for a key signed by other people. [13:46] I have, for example, generated a new key, added it to my LP account, and used it immediately to upload packages when I was traveling somewhere I wasn't comfortable taking my existing key. [13:46] Debian has such a requirement, but Ubuntu doesn't. [13:47] What's needed to upload directly is to be accepted as an Ubuntu developer. [13:48] ScottK: okay, so the crucial thing is actually that the DMB sets membership of ~ubuntu-dev in Launchpad [13:48] ScottK: and everything ease hangs off that [13:48] It's not just the DMB. === dendrobates is now known as dendro-afk [13:49] For example, for kubuntu-dev, it's done by Kubuntu Developers (which gets you in ubuntu-dev). [13:49] But generally, yes. [13:49] Which is why I say there's no real name requirement in Ubuntu. [13:50] It does have to look like a name, because (IIRC) LP people get grumpy about things that don't appear to be names. [13:53] not sure it does https://launchpad.net/~wookey [13:57] Laney: that is his name (it makes getting through customs in a hurry hassleful) [13:58] Riddell: My point was that it only has one part [13:58] Laney: he only has one part to his name [13:58] yes, I am aware [14:03] Laney: that name matches his passport :) [14:04] cjohnston: pong [14:04] I wasn't trying to say that it's not a real name (how could Launchpad know that?), just that it only has one component. [14:04] so Launchpad isn't so prescriptive about the form names have to take. [14:04] slangasek, I only want to tell you that Ghostscript upstream guys tell that freetype 2.4.9 has a bug which breaks Ghostscript, so please avoid an FFE to update from our 2.4.8 to 2.4.9. [14:05] tkamppeter: 2.4.9 also includes security fixes (as nearly all of them do). Can you give me a pointer to the bug? [14:05] slangasek, please join #ghostscript. They can explain the problems to you. [14:07] tkamppeter: can't look at it right now, sorry; will someone be around later if I drop in there? [14:08] slangasek, I will ask. [14:09] slangasek, see https://savannah.nongnu.org/bugs/index.php?35847 and https://savannah.nongnu.org/bugs/index.php?35833 [14:11] slangasek, Chris Liddell (IRC: chrisl) has made patches for Freetype which are committed to the Freetype repos but are not in 2.4.9. You will have to backport these patches. [14:12] slangasek, Chris will be available on #ghostscript until 5pm UK time today. === dantti|2 is now known as dantti === tomreyn_ is now known as tomreyn === dendro-afk is now known as dendrobates [14:51] cjwatson, i am perlexed, i just added one module to the d-i configuration for the virtio udeb, and the alternate image no longer have any of the modules in it, whereas before they did [14:52] cjwatson, actually lets say --- whereas they are claimed to have been there before ... i don't have an old enough ISO to confirm [14:59] cjwatson, indeed i even see the d-i build installing virtio-modules-3.2.0-19-generic-di as if it was going to use it ... hrm [14:59] My memory fails me…what package builds the alternate cd's? [15:00] BenC, it might be the debian-installer ... as it knows about kernel versions etc [15:00] BenC, debian-cd [15:01] (and a bunch of scripts called cdimage) [15:01] apw: had already checked that…it just builds the initrd's and boot compnents [15:01] ogra_: thanks [15:01] (check the branches of the ubuntu-cdimage team on LP [15:02] ogra_, you may know, how does one know which udebs are included in a specific iso ? [15:03] seb128: what options do we have to fix icons on packages in partners? [15:03] any? [15:03] * mhall119 is specifically looking at skype [15:03] mhall119, dunno, I'm not the person to ask, I've no clue about partners, ask mvo maybe [15:03] mvo: ^^ ?? [15:03] apw, look at the respectinve .list file for teh iso on cdimage [15:04] mhall119: sure, a new app-install-data-partner is needed [15:04] and grep for .udeb or so [15:04] mvo, that don't include runtime icons for unity does it? [15:04] mvo: what does that mean? (please forgive my ignorance) [15:05] mvo: specific case of skype: we have a 24px and 48px PNG, which look terrible in alt-tab switching. If we were to get an SVG for it, can it be used? Or does any image have to come from Skype? [15:05] ogra_, and if what i want is missing, where are the master copies? [15:05] apw: let me have a look [15:05] Riddell: Which "corner image" are you referring to in bug 942543? [15:05] apw: there's a manifest in the d-i output [15:05] Launchpad bug 942543 in ubiquity (Ubuntu) "[KDE] UI needs graphics to match wallpaper" [Medium,New] https://launchpad.net/bugs/942543 [15:06] http://archive.ubuntu.com/ubuntu/dists/precise/main/installer-i386/current/images/MANIFEST.udebs [15:06] cjwatson, i think virtio is not a udeb used in the images ... if i have done what ogra asked correctly [15:06] seb128: oh, runtime icons? no, just software-center icons [15:06] ogra_: no, the .list file is no good for udebs in the initrd [15:06] oh, indeed [15:06] apw: ogra_'s instructions are not correct here :) [15:06] only for the installable ones [15:06] sorry [15:06] mhall119: we need to talk to skype about this, I don't think we can simply change that [15:07] ok,that's what I figured the answer would be [15:07] cjwatson, ok ... the manifest claims its in there [15:07] thanks mvo [15:07] virtio-modules-3.2.0-19-generic-di 3.2.0-19.30 amd64 [15:07] mvo, the question for mhall119 was "can we add a svg icon for skype to the skype package in partner", ie can we do such changes or do they need approval from the partners who submitted them? [15:07] apw: only for some images [15:07] mvo, ok, I see you replied while I was typing :p [15:07] apw: so which image were you expecting virtio-modules to be in? the build system only puts it in netboot images [15:07] oh oop [15:08] cjwatson, ahh as the main server images include the -virtual kernel i was sort of assuming virtio was in them ... so that implies the 'they were there' is completly erroneous from the reporter [15:08] seb128: sorry, I missed some context [15:08] apw: -virtual isn't used for installation [15:08] it may be *installed* [15:09] mvo, no worry ;-) [15:09] mvo: who would be a good person for me to talk to about getting better icons for those packages? (currently skype and adobe-flashplayer) [15:09] cjwatson, yeah these images are being booted in virtual environments, even though they are non -virtual kernels for booting [15:09] apw: virtio_blk is in block-modules right now, and virtio_net is in nic-modules - those are probably better [15:09] apw: but I'd need to know exactly which module you're talking about [15:09] cjwatson, these are hyper-v modules, modules for disks and network for that [15:10] it sounds like i may have put them in the wrong hole [15:10] mind you, {block,nic}-modules depend on virtio-modules [15:10] apw: Is that a frequent complaint? [15:10] so that actually ought to be dragged in [15:10] mhall119: the foundations team took this over, so slangasek I would think [15:11] oh, of course, block-modules doesn't actually go in the cdrom initrd [15:11] cjwatson, i definatly am not seeing them in the /install/initrd.gz on todays ubuntu-server daily [15:11] the cdrom initrd should only have enough to load stuff off the cdrom, not general block modules [15:11] infinity, that they are missing ? [15:11] slangasek: I'm trying to get SVG icons for skype and flash-player-properties, do you happen to know who I can talk to about that? [15:11] are these modules necessary for talking to a virtualised cdrom? [15:12] infinity: in the ubiquity KDE frontend [15:12] cjwatson, i don't believe they are no, so they may be confused about needing them then [15:12] apw: No, putting tihngs in the wrong hole. [15:12] infinity, *slap* [15:12] apw: then they aren't meant to be in the initrd - the installer loads more stuff at run-time [15:12] Riddell: I gathered that, yes, I meant which image. The swirly watermark looking thing in the bottom left? [15:12] infinity, there is no wrong hole [15:12] cjwatson, ok i will go and debug this with them, thanks for your help [15:13] apw: either virtio-modules or {block,nic}-modules as appropriate ought to work, based on what you've described, and it isn't necessary for them to land in the initrd [15:13] feel free to point me at the bug [15:13] cjwatson, yep i think we have a clear case of reporter not being at all clear what the issue is and we may actually have no issue at all [15:14] or the issue is at a different level ... please flush from your mind until i have something that makes more sense [15:14] flushed :) [15:15] apw: Hrm, while you're around, I suppose I should test your new kernel for my G3. [15:15] infinity, eeep [15:16] infinity: that's the one [15:16] Riddell: Kay. That's not being pulled from a theme distributed with kubuntu, then? I'll poke it after I test a kernel on my firewall. [15:16] cjwatson, so i am hearing you say that the d-i boot kernel can use the udeb off the CD to suppliment its modules [15:17] infinity: I don't remember, it'll be the KDE frontend pointing to a .png file but that file is either shipped with Ubiquity or elsewhere in KDE [15:17] infinity: we can just get rid of it so this doesn't reoccur [15:18] Riddell: Seems fair enough to me, the UI will look nice enough without it. [15:18] /ws 20 [15:18] grar [15:23] apw: yes, that's central to how d-i works [15:23] cjwatson, thanks one day i'll learn enough to run away from it [15:23] apw: the initrd only has enough bits to allow d-i to fetch more of itself [15:23] it's like a base system [15:23] yeah very logical when you have to have the rest on there anyhow [15:23] d-i has rough analogues (though much less capable) of dpkg and apt [15:30] apw: New kernel boots fine. [15:30] infinity, oh dear, thanks [15:31] apw: Did we really expect otherwise? [15:31] infinity, i try and not expect testing results one way or the other, tends to lead to madness [15:32] * infinity did, however, run into a weird initrd issue he needs to try to reproduce sometime... [15:32] any fglrx guru's around? bug 865984 is killing me frequently ATM [15:32] Launchpad bug 865984 in Ubuntu "firegl_sig_notifier crash on shutdown or reboot" [Undecided,Confirmed] https://launchpad.net/bugs/865984 [15:39] cjwatson, ok the issue seems to be that the virtio udeb seems to missing en-toto from the images all of a sudden [15:39] didrocks, seb128: what's your acceptance-window in regards to notify-osd (with the avg-bg-color fixes) ... do you want a new tarball-release or do you want to distro-patch it? [15:40] MacSlow, do you plan to fix any of the other bugs? like the multiscreen support to match the new unity design? [15:40] cjwatson, it was in there on the 14th and not now [15:41] seb128, yes [15:42] MacSlow, I would recommend you roll a tarball once you get those ready [15:43] seb128, d'accord [15:43] mhall119: so the idea is you're going to provide an SVG icon and want to know if we can put it in the package? [15:44] slangasek: or who can I talk to about requesting an SVG file from the upstream [15:44] apw: So, where do we go from here? Resurrect powerpc, or try to figure out how to make powerpc-smp work? [15:44]