[00:06] Keybuk: what are the dotted lines? === emma_ is now known as emma [00:14] sladen: useful points [00:14] where we do budgeting [00:14] end of kernel/initramfs [00:14] start of X [00:14] start of gdm session [00:14] end of boot [00:17] Keybuk: they do look useful; could you try adding a vertical label for which each one is [00:24] does google use android to track traffic patterns? i've noticed the traffic data for the US seems much better than it used to be, and probably better than is actually being tracked by highway departments [00:24] and there are weird gaps that would lead me to believe phones might be being used as a source of the data [00:24] ccheney`: yes [00:25] ccheney`: they blogged about it. [00:25] http://googleblog.blogspot.com/2009/08/bright-side-of-sitting-in-traffic.html [00:26] "It takes almost zero effort on your part — just turn on Google Maps for mobile before starting your car — and the more people that participate, the better the resulting traffic reports get for everybody." [00:27] lifeless: ah cool, i couldn't find the article [00:28] lifeless: now with google maps navigation lots more people will be using it :) [00:28] regular google maps app wasn't all that useful [00:28] sladen: I don't see the point [00:29] the lines already meet the appropriate process [00:30] Keybuk: the output becomes significantly more useful to people other-than-yourself if it's annoted [00:31] Keybuk: the more information processing you can farm out to other people, the less explaining and work you need to do for yourself [00:31] sladen: and the more time I spend adding annotations ;) === micahg1 is now known as micahg [00:32] if only it were open source, then you could add your own features [00:32] OH, WAIT [00:34] Keybuk: upload/publish it and I will [00:36] sladen: I already have [00:36] bzr branch lp:ubuntu/pybootchartgui === Jon is now known as Guest23770 === dendrobates is now known as dendro-afk [01:33] cjwatson: http://people.canonical.com/~scott/daily-installer/ === jelmer_ is now known as jelmer === dendro-afk is now known as dendrobates === dendrobates is now known as dendro-afk [03:47] *sigh* [03:47] Stupid KDE transition breaking compiz builds :/ [03:52] Amaranth: Qt upstream that does a post-beta binary incompatible change. [03:52] Not actually even KDE's fault. [03:53] ScottK: hrm, I don't think we do all that when GTK+ changes the ABI during development [03:55] Generally Qt doesn't break ABI except on major version updates, so I gather this was some kind of mistake as the 'fixed' ABI is compatible with 4.5 [05:26] I am just curious about creating IPv6 readiness blueprint for Lucid Lynx I am a Debian Developer experienced in networking, having developed OSPF for zebra/quagga, and extensive IPSEC administration and configuration experience under linux. [05:40] matgeek: You'll probably get more of an answer during regular European or US working hours. [05:53] ScottK : Thanks, I will try later. One of the problems with living in NZ.... [06:02] matgeek: You might also try #ubuntu-server. Perhaps more interest there. [06:37] morning [06:37] component mismatches smells fishy: a lot of multimedia packages (libfaac, mplayer, xvidcore) try to get promoted to main for no understandable reason. at least not to me. [06:41] who should i bug about firefox apparmor profiles, -security or -mozilla team [06:41] ? [06:44] 102923.396229] type=1503 audit(1259217826.151:207): operation="exec" pid=25141 parent=25079 profile="/usr/lib/firefox-3.5.*/firefox" requested_mask="::x" denied_mask="::x" fsuid=1000 ouid=0 name="/usr/bin/liferea-add-feed" [07:33] Good morning [07:34] morning pitti! [07:35] siretart: kdemultimedia Recommends: mplayerthumbs [07:38] hey Hobbsee, how are you? [07:39] pitti: studying :( [07:39] pitti: last exam tomorrow [07:39] \o/ [07:39] Hobbsee: oh, all the best then! [07:39] thanks :) [07:39] Hobbsee: which is why you're on irc? :-) and yes, Good Luck! [07:40] spm: well.. ;) [07:40] spm: trying to put a bit of a break in between doing one, and studying for the next one ;) [07:40] sounds fair to me! [07:43] superm1, ScottK: package set questions> I'm afraid I need to discuss that with cjwatson once he's back from vacation; the overlaps are intended, but I'm not sure why some key packages are missing [07:43] so I'll answer next week === mac_v is now known as mac_v`afk [07:44] pitti: I was back from vacation yesterday [07:45] pitti: but I do not see the questions you're responding to in my scrollback [07:45] cjwatson: oh, oh u-devel@, wrt. "ubuntu-desktop team members can now upload" [07:48] cjwatson: I actually noticed the same: gnome-panel can't be uploaded by -desktop, only by -core; but it is in http://people.canonical.com/~ubuntu-archive/germinate-output/ubuntu.lucid/desktop [07:48] (I understood that the package sets are derived from seeds?) [07:52] good morning [07:52] hey dholbach! [07:52] hi mvo === mac_v`afk is now known as `afk === mac_v is now known as mac_v`afk === mac_v`afk is now known as mac_v [07:57] pitti: indeed, but gnome-panel also builds a library that's used all over the place, so *shrug* [07:58] ah, ok [07:58] so it's due to the overlap with xubuntu, etc. [07:58] no [07:58] overlaps don't cause something to go to core [07:58] it's because the reverse (build-)dependency chain from libpanel-applet leads into core, somewhere along the line [07:59] the test is "can you break core by uploading this source package?" [07:59] since gnome-panel really is a fairly fundamental platformish package loosely disguised as desktop, I don't feel too bad about it [08:00] ok, thanks for the heads-up [08:00] cjwatson: what defines "core"? [08:00] I think superm1 has found at least one genuine bug [08:01] core> minimal and standard seeds? [08:01] hm, no, that wouldn't grab panel [08:01] platform.lucid/(required,minimal,boot,standard,d-i-requirements,installer,build-essential,language-packs,supported-(development,kernel,(hardware,installer,network,sysadmin)-common)) right now [08:02] we can change some of that, but I think reducing down to standard would be too far [08:02] agreed [08:02] besides, build-dependencies do still tend to suck in a lot of the world surprisingly quickly [08:02] look at the build-dep graph for just the kernel sometime :) [08:03] I just wondered where this is defined [08:03] in a branch I need to push somewhere more sensible than cocoplum:~cjwatson/packagesets/ [08:03] it's a miracle that it doesn't pull in gtk -- yet [08:04] ah, seed_sets in packageset-report [08:39] hi [08:41] I'd like to learn C, have you any suggestion? [08:43] rossera, a good textbook. and somewhat offtopic for this channel [08:46] I see. thanks [08:56] Hi! I am a DD that wants to help out with IPv6 support in Ubuntu, testing Lucid for functionality. Is there a blueprint for any of this ples? May I write one? [08:59] hey matgeek; I don't think that we specifically discussed IPv6 at UDS; my gut feeling is that problems with it should be handled through bug reports, unless there's some large architectural change that you plan? [09:01] pitti: No,Just want to vet functionality. Create a how to. Pull a few IPv6 auto address functionality packages from universe to supported. [09:02] pitti: IPv6 is coming quickly here in NZ. My ISP is running trials. I have the background and expertise to do this sort of work. [09:03] great to hear that it's getting some real-life adoption! [09:03] pitti: Have developed and maintained OSPF for a router and zebra/quagga. Ran IPSEC for NZ EEC using openswan. [09:04] pitti: I have what s needed to do this properly. [09:04] that's great, thanks for looking into that [09:05] pitti: Who would I talk to to organise this? [09:06] matgeek: until there are concrete bug reports, I'd say that discussing it on ubuntu-devel@ would be best [09:07] to make some more IPv6 literate people aware, and coordinate testing === micahg1 is now known as micahg [09:08] pitti: Is there a place in the Ubuntu community wiki to start organising some documentation? I want to document how to set up lucid for this, as well as create patches for any show stoppers that I find. [09:09] matgeek: https://wiki.ubuntu.com/IPv6 is a starting point, so feel free to create https://wiki.ubuntu.com/IPv6/Testing or something such [09:13] pitti: Cool! Thanks! I will start from there. Testing basic IPv6 on lucid, and how it would run on home lan etc. [09:14] pitti: We need to get Ipv6 smooth as to maintain Ubuntu's reputation. [09:14] * pitti wishes that German T-Online would finally offer IPv6, too [09:15] IPV6 still hasnt got around enough so its kinda hard to test [09:18] fagan: Yes I know, but I think I can make sure that lucid works well enough out of the box. Within a year we will be seeing a lot more IPv6. I just want to make sure that all the basica packages needed are supported etc, and file bugs where needed, and fix show stoppers. [09:19] true but do a lot of testers have IPv6? [09:23] fagan: Not yet. I am a Debian Developer with a background in networking, IPSEC, and zebra/quagga OSPF development. === azeem_ is now known as azeem [09:23] fagan: should be able to test things using few virtual machines, and map things out. [09:24] I dont think virtual machines can emulate them [09:24] * fagan checks [09:24] hm, host <-> VM should work with IPv6 [09:25] but kvm doesn't set up interfaces on the host side by default [09:25] I think there's a way to do that, though; I'm sure that kirkland would know :) [09:29] pitti: Already doing it with kvm!!. I will keep quiet for a few days, and get back after exploring more, there looks like there may be quite a community using IPv6 with Ubuntu. [09:31] ogra: still remember the modprobe -b change that you did to pm-utils ages ago? I forwarded that to https://bugs.freedesktop.org/show_bug.cgi?id=25254, but there's no explanation why the change was done [09:31] Freedesktop bug 25254 in General "respect module blacklists on resume" [Normal,New] [09:31] ogra: what was the actual problem there? (rationale is that a module which was loaded before should be loaded afterwards) === TeTeT_ is now known as TeTet === TeTet is now known as TeTeT === mac_v is now known as mac_v_ [10:22] MacSlow, yo ... i hear you know about OSD [10:23] apw, sadly ;) [10:23] MacSlow, does OSD take any X grabs in its life cycle showing popups etc? [10:23] apw, no [10:24] MacSlow, just had a permenant hang of my OSD (and got a core off it), trying to understand how it can trigger unresponsivness in applications which don't use it [10:24] *shrugg* [10:24] i've added the core to bug #347444 [10:25] Launchpad bug 347444 in notify-osd "OSD hangs up applications leading to them greying out" [Medium,Confirmed] https://launchpad.net/bugs/347444 [10:25] its getting more and more frequent over time [10:25] apw, the only bug I know of (and have not yet been able to fix) is it stalling something on DBus [10:26] it seems supprising gnome-terminal would use dbus, but anything is possible [10:26] apw, I even asked DBus-experts about possible causes for this... no clear answer as of yet [10:27] so do we know what osd is trying to do when it goes non-responsive? [10:27] * apw is suspicious that its going non-responsive when its queue gets large [10:27] apw: I take it I should add https://wiki.ubuntu.com/KernelTeam/ReleaseStatus/LucidTasks to kernel work item scanning? [10:28] pitti, yep that is the roll up ... i meant to tell you about it [10:28] pitti, -b makes modprobe respect the blacklist files [10:28] apw, nope [10:28] ogra: right, but why? [10:28] MacSlow, well there is a core file on the bug now, of it mid hang, so we should be able to tell from that right [10:28] pitti, well, if a module is blacklisted i dont want it to be loaded [10:28] ogra: if a module was loaded before suspend, shouldn't it be loaded after resume, regardless of blacklisting? [10:29] apw, an isolated test-case triggering this bug in notify-osd would be a good start... sofar nobody was able to provide such [10:29] pitti, but it shouldnt be loaded before suspend, if it was, thats wrong ... [10:29] MacSlow, i have tried hard to get it, and i get it about once a week on average, and cannot find a simple trigger for sure [10:29] unless i misunderstand what a blacklist is for [10:29] ogra: right, but if it is, you apparently loaded it manually [10:30] blacklist should be for suppressing auto-loading [10:30] right [10:30] but pm-utils just restores the situation as it was before suspend [10:30] it didnt when i supplied that patch [10:30] ogra: I thought it was for not loading blacklisted module dependencies, etc. [10:31] if thats different today then the -b isnt needed [10:31] ogra: ok, thank you [10:31] apw: so it is on http://piware.de/workitems/kernel/lucid/report.html now [10:32] pitti, most excellent [10:32] apw: right now, wiki page WIs can't have assignees (not sure where to fit them in), so they appear as "nobody" [10:33] if you have an idea/request how to add an assignee to it, I'm fine to add it [10:33] could e. g. be "assignee:apw" or so [10:33] but it needs a prefix or other special marking, otherwise any word in a table row could be considered an assignee [10:33] pitti, well for my purposes i exposed INPROGRESS and added allowed it to be INPROGRESS apw [10:34] pitti, though the other option would be to simply follow the current plan [apw] at the beginning of the descriptio [10:34] that would work too, right [10:34] pitti, does it work now? [10:35] not with current workitems.py [10:35] but I can hack it in [10:35] (TODO|INPROGRESS|POSTPONED) (.*) || [10:35] where the middle .* would be the assignee [10:36] yeah that is what i am doing as things stand [10:36] my version of your workitems doesn't flatten INPROGRESS to TODO as i care which are being worked [10:37] I'll fix that as well [10:37] pitti, i am also parsing out the wiki pages in a slightly differnt way [10:37] one which lets me see the titles of the sections so our page shows up as four sections on my report [10:38] apw: oh, hang on, it just looks for DONE/POSTPONED, and treats everything else as TODO [10:38] so it works with INPROGRESS [10:38] for my purpose i need to tell them appart at the consumer of that API so i've frigged it to keep that too [10:38] i use the same data to produce this: [10:38] https://wiki.ubuntu.com/KernelTeam/ReleaseStatus/Lucid [10:39] note the two tables below your burn down chart [10:39] (TODO|INPROGRESS|POSTPONED) (.*) || <-- note that this is probabally needs to be [10:40] (TODO|INPROGRESS|POSTPONED)( (.*))? || [10:43] ok, added [10:44] apw: http://piware.de/workitems/kernel/lucid/report.html -> one less for "nobody", one more for you now [10:44] pitti, 'yay' [10:45] pitti, how do you determine the owner of the blueprnt, doesn't seem to be a workitems.py call for it :) [10:46] apw: it's done in get_blueprint_workitems() [10:46] assignee_re = re.compile('') [10:46] hrrm [10:46] m = assignee_re.search(l) [10:46] if m: [10:46] default_assignee = m.group(1) [10:46] (yay screenscaping) [10:47] pitti, how does that hit the assignee and not any of the other people in there [10:47] there are what 4 names on that page? [10:48] apw: the check above it [10:48] if '

Assignee:' in l: [10:48] found_assignee = True [10:48] it's really a quick'n'dirty hack [10:48] ahh ... [10:48] ok i don't have that in my version [10:48] but I didn't feel like setting up proper XML parsing and xpath [10:48] could you zap me a new one :) [10:48] "zap"? [10:48] heh no, i am sure its close enough [10:49] point me at a new copy, can't remember how i got the previous one even if it was 2 days ago [10:50] apw: http://piware.de/bzr/bin/workitems.py [10:50] (it's also on WorkItemsHowto) [10:50] it's a bzr branch, too (although that has all my other ~/bin/ crap) [10:50] sorry i am soo lazy [10:50] apw: np :) [10:50] you interested in being able to get status and title for each group in the wiki page? [10:50] i've got a fn which looks at the wiki page as a collection of groups not one big group [10:51] group -> any heading? [10:51] yes, that could be useful [10:51] in the report, the wiki page could appear as one line per wiki page section instead of just one line for the entire page [10:51] yeah i am using the == xxx == as the title of the blueprint, the text before the table as the blueprint status: text, and the table as you do [10:52] yeah thats exactly what i do in my lucid status page [10:52] https://wiki.ubuntu.com/KernelTeam/ReleaseStatus/Lucid [10:52] if the Kubuntu guys tweak their page a bit, it'd work for them, too [10:52] the last four groups in the two tables are the sub sections in the wiki page [10:53] yeah [10:54] i'll clean up what i ahve here and send it back to you so you can see what i am doing [10:58] apw: I think I'll register a proper project for it on LP now, and create a branch with workitems.py, the chart generator, etc. === lionel__ is now known as lionel [10:59] gosh, blocker: how to name it [10:59] pitti, hehehe [10:59] something burner [10:59] launchpad.net/manager-sleeping-pill [10:59] it makes them happy right ... so manager-upper [11:00] manager-burner! [11:00] *cough* [11:01] work-burner [11:01] time-burner [11:02] item-burner [11:02] bah hard [11:02] call it off, too difficult [11:02] https://edge.launchpad.net/work-items-tracker [11:02] pyburner [11:02] heheh [11:02] I know, boring [11:04] pitti, there is a standard way to say 'i am a program' isn't there? [11:04] you mean #! ? [11:05] normally main is included in something like 'if package == 'main':' so you can use it as a library [11:05] ah [11:05] if __name__ == '__main__': [11:05] could we get that added :) so i can use it as a library : [11:05] yeah that thing [11:05] sure; let me finish set up the branch, etc., then I'll commit it [11:06] $ bzr commit -m 'workitems: check if called as a program' [11:10] cjwatson: may I just remind you about the change to the Ubuntu Studio iso status mails that we talked about at UDS. [11:12] pitti, is there an appoved syntax for more than one assignee? [foo] [bar] or [foo, bar] or ? [11:12] there's no concept for this right now [11:13] [canonical-kernel-team] perhaps? [11:13] it'd need a DB schema extension [11:13] https://launchpad.net/work-items-tracker set up, with bzr get lp:work-items-tracker [11:14] you can commit [11:14] pitti, ok thank will avoid multiple users for now :) [11:20] pitti, most excellent i now have assignee's across the board [11:23] wiki howto page updated === MacSlow is now known as MacSlow|lunch [11:49] jussi01: you mean making them not be sent when there are no problems? I did that a few days ago [11:51] cjwatson: I received no problems found both today and yesterday [11:52] I installed glibc-doc, but apropos printf doesn't return the right manpage. [11:53] fasta: printf(3) is in manpages-dev not glibc-doc [11:53] jussi01: huh, odd [11:53] cjwatson, ok, thanks. Is there some easy way to know that next time? [11:54] cjwatson: yeah, curious. just for clarification, this is a mail with the title: Ubuntu Studio daily CD health check [11:54] Ive received it every day for a long time ;) [11:54] fasta: packages.ubuntu.com, or you could just use manpages.ubuntu.com [11:54] jussi01: fixed [11:55] cjwatson: excellent. thank you. [11:55] (properly this time, tested) [11:55] :) [11:57] cjwatson, then I am curious what exactly you typed in on packages.ubuntu.com. [11:57] fasta: *I* didn't, because I knew the answer [11:58] cjwatson, ok, ok. One should be able to ask the packages.ubuntu.com system "manpage printf" and get the right answer. [11:58] (that doesn't work, btw) [11:58] "Search the contents of packages" -> Keyword "printf.3" -> "packages that contain files whose names contain the keyword" [11:58] fasta: manpages.ubuntu.com is much, much better at that particular kind of thing [11:58] since, you know, you then get the man page too [11:59] it's a bit too specific for packages.u.c [11:59] manpages.ubuntu.com tells you which package it's in, too [11:59] cjwatson, ah, nice. That was what I needed. === mt-lith is now known as miket === miket is now known as michaelt === michaelt is now known as mt === mt is now known as mj-t [12:17] somewhere there is a report on bugs which have upstream links, can anyone remember where it is? [12:21] apw: https://launchpad.net/ubuntu/+upstreamreport [12:23] cjwatson, thats the summary, somewhere there is a breakout [12:23] by upstream bug # === MacSlow|lunch is now known as MacSlow === ebroder_ is now known as ebroder === Zic_ is now known as Zic [12:45] StevenK: your cdimage changes are OK as long as we never need to rebuild old ubuntu-netbook-remix images (because you've removed the ability to do so). You broke the #! line in make-web-indices, though [12:45] StevenK: please let me know when you do the unr -> netbook seed collection change, as that's going to involve changes in a bunch of places [12:50] hi [12:50] I have a problem with my ubuntu desktop [12:50] In installation of some package I screwed up my desktop [12:50] by mistake uninstalled the gtk [12:50] My desktop is now not appearing [12:51] how to solve the problem? [12:51] slnr: This isn't a support channel - if you need help using Ubuntu, you should ask in #ubuntu [12:54] oh [12:54] thanks [12:54] Sure I shall login and contact that channel [13:00] Looking at LTS+1 today, I notice an "abrowser" which appears to be the equivalent of Debian's iceweasel. Where can I find an explanation for why Ubuntu chose to unbrand Ubuntu in a different way to Debian? === steveire_ is now known as steveire [13:12] ooh, second ubuntu theme music of the day from upstairs [13:13] twb: isn't "firefox" our equivalent of "iceweasel"? [13:17] Keybuk: We have an unbranded version of Firefox. [13:18] Keybuk: That is not to say that our version of Firefox is unbranded, but rather that we have an unbranded version as well as the branded version. [13:18] ah [13:19] Oh. I assumed firefox-branding was there as part of a transition, because you were tired of MoCo being asses. [13:19] twb: as I understand it, it's because we're based off upstream rather than Debian (because Mozilla wanted us to do a better job of patch forwarding in order to use their trademark), and calling it "iceweasel" implied that we were based on the Debian package [13:19] cjwatson: that makes sense [13:20] and we didn't want to maintain two entirely separate packages, but rather one with tweakable branding [13:20] From what I've seen of the Debian packaging, the bit that isn't xulrunner is pretty small anyway [13:24] this was a while back, I don't know what it would be like if we did it now [13:28] cjwatson: Yeah, I broke it on purpose. And yes, the unr -> netbook change requires LP and tasksel changes too [13:31] pwnguin: re firefox and apparmor> file a bug against firefox-3.5, and tag it with 'apparmor'. See https://wiki.ubuntu.com/DebuggingApparmor for details [13:53] mkfs.ext4dev. what exactly does it do besides creating an ext4 fs? [13:53] as opposed to mkfs.ext4 [13:57] experimental; now that's something [13:57] :/ [14:00] pitti: do I need to poke you for a regeneration when we *add* work items, as opposed to completing them? [14:01] cjwatson: in general, no; however, you should poke me when you are "by and large" done with setting up WIs, so that I can trash the db and start over, or alternatively adjust the start position of the trend line [14:01] ok [14:01] by default it starts at the #done value of the first day [14:01] I expect that I'll remove all the DBs next Monday and from then on we start to really track them [14:06] cody-somerville: could you please change platform.karmic -> platform.lucid in xubuntu.lucid/STRUCTURE? [14:06] superm1: could you please change platform.karmic -> platform.lucid in mythbuntu.lucid/STRUCTURE? [14:06] (it would be great if core-devs could change that ...) [14:07] * cjwatson wonders if it would be feasible for python2.4 to build-dep on emacs23 rather than emacs22 ... [14:07] and indeed python2.5 as well [14:11] cjwatson: Is that on lp:~ubuntu-core-dev/ubuntu-seeds/platform.lucid ? [14:12] Daviey: xubuntu.lucid [14:12] er, specifically lp:~xubuntu-dev/ubuntu-seeds/xubuntu.lucid [14:13] (mythbuntu) [14:13] or lp:~mythbuntu/ubuntu-seeds/mythbuntu.lucid === jamie is now known as Guest19452 [14:18] cjwatson: pushed, lp:~mythbuntu/ubuntu-seeds/mythbuntu.lucid rev 1185 [14:18] superm1: ^^ [14:21] Daviey: thanks [14:27] whois saix7 [14:27] ops === tkamppeter_ is now known as tkamppeter [14:42] Keybuk, from our discussion last night I still don't understand why drm initialising/gdm starting works on intel and not on radeon [14:42] tormod: I heard there's been similar issues with nVidia too [14:43] well, don't know if similar but some issues anyway [14:43] I wonder if there some specific tweaks somewhere, or if its just that intel has KMS by default or is loaded earlier [14:47] tjaalton, tseliot: I upgraded to xorg-edgers today, and it seems to completely have forgotten about transparency; all the icons etc. look weird; known? [14:49] pitti: what driver? [14:49] tjaalton: intel gm945 [14:50] ok. I don't run edgers myself, so can't tell if it's known or not :) [14:50] probably mesa has changed [14:50] * tseliot doesn't run -edgers either [14:50] ok, nevermind [14:51] I just need to run it to test the udev patches [14:51] pitti, todays -intel has https://bugs.freedesktop.org/show_bug.cgi?id=25031 [14:51] Freedesktop bug 25031 in Driver/intel "rendering and color corruption since 14109abf" [Major,New] [14:52] 'zactly that [14:52] thanks! [14:52] I held back -intel for weeks because of this, but I guess more exposure and noise can help :) [14:52] oh, not a dri driver bug after all [14:56] pitti, you can still d/l the old build from the PPA [14:59] tormod: thanks [14:59] yay, hal-free xorg [15:01] scary === chuck_ is now known as zul [15:08] pitti: what have I missed to get foundations-lucid-ubiquity-partitioner-optimisation to show up on the burndown chart? [15:14] if anyone is around and cares to weigh in on a strategy , I would like/am willing to write a little script that takes the difference between two debdiffs. debdiff ver1 vs ver2 and debdiff ver1-ubuntu# vs ver2. This would leave only ubuntu changes that have not been incorporated into ver2 [15:15] mannyv: interdiff [15:15] (patchutils package) [15:16] cjwatson, I have tried interdiff and I get this error 'interdiff: Error applying patch1 to reconstructed file' [15:17] kirkland: you around to lend a hand with an enrypted disk that I can't access...something blew up last night :( [15:17] of course I cant rule out that i have called it incorrectly but i have tried a few different ways [15:19] mannyv: interdiff requires that you turn thrice widdershins and annoint the computer with the yolk of a furling egg [15:21] Keybuk, i was telling maco just the other day that despite everyones claim computers really are run on magic. glad you agree =) [15:23] I think the problem is that patch1 and patch2 have to have a common base [15:23] you can't take two slightly different based patches and interdiff them [15:23] bzr is easier ;) [15:23] Yeah, definitely use a VCS. [15:24] well i use it to help me see if a package can be synced [15:24] bzr can help there too ;) [15:24] im listening [15:24] bzr diff --old lp:debian/pkg --new lp:ubuntu/pkg [15:25] whoa - that's awesome [15:25] ebroder: you can branch, you can commit, you can push (if you have upload rights) [15:25] james_w is made of awesome [15:26] Keybuk: Wait - uploads via bzr push works now? Shiny [15:26] ebroder: no, you still have to separately dput [15:26] that's one of the last key pieces (dput with bzr) [15:26] kirkland: don't worry about it...I am just gonna reinstall...most of my stuff was under one vcs or another, so it shouldn't be all that devistating [15:26] do what does it mean to push to lp:ubuntu/pkg? or are you saying that i can push to my own branch? [15:26] s/^do/so/ [15:27] ebroder: it publishes the code *with revision control history* [15:27] so if you merged from debian [15:27] and added other patches [15:27] and committed each separately, then the push means the next person to branch gets your history too [15:27] otherwise if you just upload, you get a new revision that contains your entire upload as one diff [15:28] Keybuk: interesting. so if i push to lp:ubuntu/pkg, and then upload, will there be an additional no-op commit from the dput? [15:28] no [15:29] mannyv: then fix interdiff rather than writing a new script, please :) [15:29] i guess it doesn't create the commit if there's no diff or something? [15:29] we don't need multiple programs for the same job that are broken in different ways [15:29] but yes, there's bzr too [15:30] patch1 and patch2 don't *actually* have to have a completely common base for interdiff to work; there's a good deal of heuristics in there [15:30] but sometimes it does run into trouble [15:32] I'd like to change the kernel personality of my workstation to server. What's the best way to do that? [15:33] cjwatson: o_O but it's already there? [15:35] pitti: hmm, I guess the delay was just longer than I expected; I swear it wasn't there when I asked [15:35] cjwatson: it's not running hourly yet, since that would spam everyone with "blabla has no work items" every hour [15:35]