|
[00:11] <TheMuso> imbrandon: I seem to be having a problem with apt-mirror. The error I get is: http://www.pastebin.ca/841988 My mirror.list file is: http://www.pastebin.ca/841990 |
|
[00:16] <Fujitsu> TheMuso: s/pub2/pub/? |
|
=== Spec is now known as x-spec-t |
|
[00:26] <TheMuso> Fujitsu: Its the way tinternode have their setup. |
|
[00:26] <TheMuso> Fujitsu: if you go to the site, ftp://mirror.internode.on.net/pub2/ubuntu/ubuntu/dists/gutsy/multiverse/sources you will see the files do exist. |
|
[00:26] <TheMuso> well, not entirely right but you get the idea |
|
[00:27] <TheMuso> Fujitsu: As it is, either pub or pub2 works. |
|
[00:28] <Fujitsu> TheMuso: pub2 doesn't seem to be visible from the outside. |
|
[00:28] <TheMuso> Hrm. |
|
[00:28] <TheMuso> Well it works for me. |
|
[00:29] <bddebian> Heya gang |
|
[00:30] <ion_> Hi |
|
[00:30] <bddebian> Hello ion_ |
|
[00:37] <zul> gday |
|
[00:38] <TheMuso> Hey zul. |
|
[00:38] <zul> hey TheMuso how goes it? |
|
[00:39] <TheMuso> zul: Well thanks. Yourself? |
|
[00:39] <zul> good a bit tired though |
|
[00:39] <brandonperry> so, if I wanted to deb up bkhive and samdump2, where is a good place to look |
|
[00:39] <brandonperry> ? |
|
[00:40] <brandonperry> look to learn how |
|
[00:40] <Fujitsu> !packagingguide | brandonperry |
|
[00:40] <ubotu> brandonperry: packagingguide is The packaging guide is at http://wiki.ubuntu.com/PackagingGuide - See https://wiki.ubuntu.com/UbuntuDevelopment/NewPackages for information on getting a package integrated into Ubuntu - Other developer resources are at https://wiki.ubuntu.com/UbuntuDevelopment - See also !backports |
|
[00:40] <brandonperry> danke |
|
[00:42] <LaserJock> hi bddebian |
|
[00:43] <bddebian> Heya LaserJock |
|
[00:43] <LaserJock> bddebian: how's it going? |
|
[00:44] <bddebian> Same as always. You? |
|
[00:44] <LaserJock> similar |
|
[00:44] <zul> hey laserjock |
|
[00:44] <LaserJock> trying to get my new laptop going well |
|
[00:47] <nenolod> cool |
|
[00:47] <nenolod> apparently my ram is ok as nothing else crashed :D |
|
[00:48] <nenolod> brb. restart. |
|
[00:49] <nenolod> firefox somehow screwed over memory management. i guess i'll need to file a bug about that. |
|
[01:18] <nenolod> anyway |
|
[01:18] <nenolod> back to making libflashsupport useful on amd64. |
|
[01:19] <crimsun> make sure you're basing off Debian's flashplugin-nonfree-pulse tree. |
|
[01:24] <LaserJock> has there been any more discussion about flash SRUs? |
|
[01:27] <nenolod> crimsun, i'm just fixing libflashsupport in universe |
|
[01:27] <nenolod> also |
|
[01:28] <nenolod> i think the gnome-panel crash has to do with gtk |
|
[01:29] <crimsun> ah crap |
|
[01:30] <crimsun> we really should gut libflashsupport in favor of Debian's flashplugin-nonfree-pulse |
|
[01:31] <ion_> Are the based on the same code, or something completely different? |
|
[01:31] <crimsun> same code, essentially |
|
[01:31] <TheMuso> But for syncing reasons, it makes sense to use Debian as a base. |
|
[01:32] <crimsun> I think CJ ripped out everything not related to flash+pulse, though |
|
[01:32] <nenolod> crimsun, i agree |
|
[01:32] <TheMuso> And maintenance reasons |
|
[01:32] <nenolod> but right now, this gnome-panel crash is annoying me |
|
[01:32] <crimsun> I'm trying to chase down that rmap.c foo'er |
|
[01:32] <nenolod> ah. it's happened to someone else? :) |
|
[01:33] <crimsun> it's been happening for some time |
|
[01:33] <crimsun> (according to google) |
|
[01:33] <nenolod> it's the first time it's happened on this machine |
|
[01:33] <nenolod> i was worried that i bought defective ram :/ |
|
[01:33] <crimsun> are you using kvm? |
|
[01:33] <nenolod> no |
|
[01:33] <crimsun> ok, that's a start |
|
[01:33] <nenolod> but the memory passes memtest86+ |
|
[01:33] <nenolod> (i tested it before putting the memory into production ;)) |
|
[01:34] <nenolod> crimsun, i'll look into flashplugin-nonfree-pulse then |
|
[01:34] <nenolod> persia, gnome-panel crashes for me when i run pbuilder |
|
[01:35] <nenolod> persia, so i think it's I/O related, as a lot of files are opened and closed during pbuilder setup |
|
[01:35] <nenolod> since gtk+ has this retarded "recent files" thing now |
|
[01:35] <ScottK> crimsun: Would you be available in a bit to upload a source backport for clamav in Feisty? I did a "bad" thing today and I need to fix an i386 FTBFS. |
|
[01:36] <persia> nenolod: That matches my experience. I can replicate with sbuild, but never encountered with working with datasets in lots of little files with total size of twice my available RAM. |
|
[01:36] <crimsun> ScottK: if it's not huge, yes. |
|
[01:36] <persia> nenolod: Shouldn't that not track inside a chroot? |
|
[01:37] <nenolod> persia, it shouldn't track anything outside $HOME in my opinion |
|
[01:37] * Fujitsu has noticed gnome-settings-daemon crashing a bit lately, but not gnome-panel. |
|
[01:37] <ScottK> crimsun: Debdiff should be a new debian/changelog entry and one build-dep change in debian/control |
|
[01:37] <persia> nenolod: I can't agree with that. I touch lots of files in /data and in /usr/share/doc that would be handy in "recent files" if I used "recent files". |
|
[01:37] <nenolod> persia, i don't understand why gtk+ needs to track what files are opened and closed ;) |
|
[01:38] <nenolod> maybe it's an FD exhaustion thing |
|
[01:38] <persia> nenolod: "Recent Files" has several uses. Whether it belongs in gtk+ is a different issue. |
|
[01:38] <Ng> can a debian/watch file reasonably track the "Download project files" section of a Launchpad project? |
|
[01:39] <nenolod> persia, see https://bugs.launchpad.net/ubuntu/+source/gnome-panel/+bug/180463 |
|
[01:39] <crimsun> nenolod: just noticed that's with -2-. When -3- is bumped (or if you want to try it locally by grabbing and installing manually), can you update the bug report? |
|
[01:39] <persia> Ng: Should be able to do so. You might need to do it in screen-scraper mode. |
|
[01:39] <Ng> persia: hmm |
|
[01:39] <nenolod> crimsun, gnome-panel? |
|
[01:39] <nenolod> crimsun, or linux-source |
|
[01:39] <Fujitsu> Or convince the LP people to give us a nice way to do it ;) |
|
[01:39] <persia> nenolod: I can't for some reason. Subscribe me? |
|
[01:39] <nenolod> ok |
|
[01:39] * Fujitsu blinks. Why would that be private? |
|
[01:40] <crimsun> nenolod: 180461 (linux) |
|
[01:40] <nenolod> crimsun, ok. will do. |
|
[01:40] <nenolod> Fujitsu, i don't know. i didn't set it private. |
|
[01:40] <Ng> I'll have a look tomorrow. I've fixed up enough stuff tonight :) |
|
[01:40] <nenolod> persia, idle_populate_func() is a function which populates the recent files chooser |
|
[01:41] <nenolod> persia, i think it's an FD exhaustion issue |
|
[01:42] <bddebian> persia: Hey! |
|
[01:42] <persia> nenolod: That might make sense. I was only working with hundreds of files in my dataset activities, which is significantly less than sbuild. On the other hand, I only encounter it with some builds: about 40% of the time there is no issue. I wonder if it might be related to the installation of some of the gtk libraries in the chroot. |
|
[01:42] <persia> bddebian: Hi! Noticed zynaddsubfx: are you chasing sound now too? |
|
[01:43] <bddebian> I'm just chasing my tail, as always :) |
|
[01:43] <nenolod> crimsun, 180461 didn't happen for me until i increased my memory size in both amd64 machines |
|
[01:43] <bddebian> persia: Did you see that slangasek fixed jugglemaster? |
|
[01:43] <nenolod> crimsun, so < 1.5GB seems to not trigger the bug |
|
[01:43] <persia> bddebian: Yes. Nice fix too :) |
|
[01:44] <persia> nenolod: data set size or RAM size? |
|
[01:44] <nenolod> crimsun, it might also have to do with 32UL apps (flashplugin-nonfree) being used at the same time as 64UL apps. |
|
[01:44] <LaserJock> persia: you know anything about ubuntu-main-sponsors? |
|
[01:45] <persia> LaserJock: A little. What information do you seek? |
|
[01:46] <nenolod> crimsun, because, the last time, i had a 32UL app going + apport running on gnome-panel and firefox hosed the slab allocator mapping table when apport opened a new tab to start the bug reporting process |
|
[01:46] <ScottK> crimsun: Would you rather have a debdiff to the current feisty-backports version of clamav or a .dsc link to the entire source package? |
|
[01:47] <crimsun> ScottK: both, please. |
|
[01:47] <ScottK> Will do. |
|
[01:47] <nenolod> crimsun, wish i had more information, but i'm not much of a kernel monkey ;) |
|
[01:47] <ScottK> I'm test building right now. |
|
[01:47] <bddebian> persia: Hey, I tried :-( |
|
[01:47] <persia> bddebian: You did all the heavy lifting. The final patch is fairly small. |
|
[01:47] <bddebian> persia: BTW, newpki-client iw freakin' worse than jugglemaster was... :( |
|
[01:48] <bddebian> s/iw/is/ |
|
[01:48] <LaserJock> persia: well, I just noticed that there are quite a few bugs and wondered if there was somebody looking after it |
|
[01:48] <LaserJock> I'm guessing dholbach |
|
[01:49] <crimsun> I can look at u-m-s later tonight |
|
[01:49] <crimsun> look -> help look* |
|
[01:49] <persia> LaserJock: Essentially. Some bugs in queue get hit because the people who watch those packages see them and incorporate the changes. When things sit for a while, dholbach usually subscribes someone specific, and pokes them. |
|
[01:50] <nenolod> persia, i have an idea on how to verify that it's a gtk+ |
|
[01:50] <nenolod> persia, one moment ;) |
|
[01:50] <persia> LaserJock: For packages that aren't "owned", I believe processing is somewhat like UUS, except the lack of unsubscription. For "owned" packages, best to poke the interested party. |
|
=== kitterma is now known as ScottK2 |
|
[01:50] <pochu> Ng: I have this in one of my packages: https://launchpad.net/andvare/+download https://launchpad.net/andvare/(?:.*)/(?:.*)/\+download/andvare-(.*)\.tar\.gz |
|
[01:51] <nenolod> persia, now i have two apps which should crash |
|
[01:51] <pochu> Ng: should work for you changing andvare with your project name |
|
[01:52] <pochu> Ng: (those two links should be in the same line) |
|
[01:52] <pochu> Night all |
|
[01:53] <LaserJock> persia: seems a bit scattered |
|
[01:53] <pochu> stgraber: interesting Minutes btw ;) |
|
[01:53] <nenolod> hmm. can't trigger it now. |
|
[01:53] <persia> LaserJock: That it is. Nobody has ever sorted the culture mix for main. |
|
[01:53] * nenolod tries updating his pbuilder |
|
[01:54] <persia> pochu: There wasn't a lot of activity in the meeting. See the raw log :) |
|
[01:54] * persia should have slept in |
|
[01:55] <bddebian> heh |
|
[01:57] <LaserJock> persia: yeah, I've still haven't figured out "ownership" in Main |
|
[01:58] <crimsun> what ever became of keybuk's suggestion regarding main+universe and seeds? |
|
[01:58] <nenolod> crimsun, so, you think i should scrap libflashsupport (it's not my package), and instead issue a sync request for flashplugin-nonfree-pulse, and then recommends: on that? |
|
[01:58] <bddebian> crimsun: We killed him :-) |
|
[01:59] <ScottK> crimsun: Near as I can tell sabdfl didn't like it and nothing more was said. |
|
[01:59] <persia> LaserJock: Well, there's several classes of things in main: flavour-specific stuff, kernel stuff, toolchain stuff, startup/shutdown stuff, core libraries, and miscellany. As I understand it, miscellany is not "owned", and core libraries are typically coordinated with all the rdepends, the others generally have teams that coordinate stuff. |
|
[01:59] <crimsun> nenolod: I'd ping ogra regarding the availability of http://git.debian.org/?p=pkg-pulseaudio/flashplugin-nonfree-pulse;a=summary |
|
[02:00] <persia> crimsun: https://lists.ubuntu.com/archives/ubuntu-devel/2007-December/024885.html was the last result (as my further response was ignored) |
|
[02:00] <LaserJock> persia: right which makes u-m-s a bit pointless |
|
[02:01] <persia> LaserJock: Well, maybe. It's hard for people not sure of the details (and some of us who ought be) to determine who gets a package. U-M-S is a nice way to coordinate the requests for sponsorship, and have the (few) main sponsors ensure the right people are notified about the candidates. |
|
[02:01] <ScottK2> crimsun: Please see Bug #180466 |
|
[02:01] <ubotu> Launchpad bug 180466 in feisty-backports "Latest clamav backport (0.91.2-3ubuntu2.1~feisty1) stuck in dependency wait" [High,Confirmed] https://launchpad.net/bugs/180466 |
|
[02:01] <ScottK2> That has both the debdiff and a link to the .dsc |
|
[02:02] <nenolod> crimsun, well, that package is i386-only too |
|
[02:02] * ScottK2 runs off to get the 4 year old to bed. Be back in a bit. |
|
[02:02] <nenolod> crimsun, we need pulse to work on amd64 as well |
|
[02:03] <crimsun> ok, so we'd just need an addition to the arch line. |
|
[02:03] <crimsun> flashplugin-nonfree works with the wrapper on amd64, no? |
|
[02:03] <TheMuso> crimsun: But will that work on amd64? |
|
[02:03] <nenolod> crimsun, right, but ia32-libs-dev is still all screwed up |
|
[02:03] <crimsun> still? even after pitti's upload recently? |
|
[02:03] <crimsun> granted I didn't check it, just noted it |
|
[02:03] <nenolod> crimsun, -dev is not in the archive yet AFAIK |
|
[02:04] <nenolod> nenolod@petrie:~/upkg/libflashsupport/libflashsupport-1.9$ apt-cache search ia32-libs-dev |
|
[02:04] <nenolod> nenolod@petrie:~/upkg/libflashsupport/libflashsupport-1.9$ |
|
[02:04] <nenolod> synced 1 hour ago |
|
[02:04] <nenolod> so... can't build-depend on it |
|
[02:05] <nenolod> so... have to do a bunch of ugly hacks to get it to build with ia32-libs |
|
[02:05] <nenolod> (which i have a way of doing, but like i said, it's ugly, and i'd rather see a proper solution be realised for building 32UL code in 64UL multilib environment) |
|
[02:07] <nenolod> the bug requesting ia32-libs-dev is here --> https://bugs.launchpad.net/ubuntu/+source/ia32-libs/+bug/117131 |
|
[02:07] <ubotu> Launchpad bug 117131 in ia32-libs "32bit development libraries missing for amd64 systems systems" [Wishlist,Confirmed] |
|
[02:08] <nenolod> but it's not a big deal |
|
[02:13] <nenolod> crimsun, i think it'd be better to stay with libflashsupport for now |
|
[02:13] <nenolod> crimsun, cause people are using OSS4 and such |
|
[02:20] <SirBob1701> wheres better to join motu launchpad or elsewhere? |
|
[02:20] <ScottK> Here is good. |
|
[02:21] <SirBob1701> ah |
|
[02:21] <SirBob1701> ok how does one go about joining? |
|
[02:21] <persia> SirBob1701: Jump in, do some stuff, and after a while we'll tell you to request upload access. |
|
[02:21] <crimsun> nenolod: hmm, interesting point WRT OSS. |
|
[02:21] <SirBob1701> persia: ok |
|
[02:22] <persia> SirBob1701: I like to make sure people have reviewed https://wiki.ubuntu.com/MOTU/Contributing, and would be happy to share tasks if you are looking for some. |
|
[02:26] <nenolod> fghgds |
|
[02:26] <nenolod> libflashsupport reruns configure without CFLAGS="-m32" :| |
|
[02:27] <nenolod> hmm. exporting it fixes things. |
|
[02:27] <SirBob1701> persia: i'll look that site over thoughly and then talk to you :) |
|
[02:27] <persia> SirBob1701: When you're looking for something, just ask here. Others may also have suggestions if I'm not around. |
|
[02:28] <SirBob1701> ok |
|
[02:28] <crimsun> ScottK: 180466 processed. |
|
[02:28] <SirBob1701> persia: if i want the source for a project to work on do i use the synaptic source or bazaar? |
|
[02:29] <persia> SirBob1701: Depends on where you want to apply the fix. I usually start with the Ubuntu package, (which might be repo source or bzr, depending on the package), and then push back up to Debian or upstream. |
|
[02:30] <SirBob1701> ok |
|
[02:30] <LaserJock> SirBob1701: generally apt-get source |
|
[02:30] <SirBob1701> LaserJock: thanks |
|
[02:31] <LaserJock> often times if the package is maintained in bzr it'll tell you, I think |
|
[02:32] <nenolod> stupid gtk+ bug |
|
[02:32] <nenolod> it's definitely some FD exhaustion thing |
|
[02:37] <nenolod> TRY TO CRASH IRSSI, GTK+ |
|
[02:37] <nenolod> I DARE YOU. |
|
[02:37] <nenolod> yeah. that's what i thought. |
|
[02:37] <nenolod> watch gnome-terminal crash now. |
|
[02:40] <nenolod> ia32-libs needs new libraries |
|
[02:40] <nenolod> :D |
|
[02:40] <Fujitsu> Again? |
|
[02:40] <nenolod> yeah |
|
[02:40] <nenolod> pulse is not there |
|
[02:40] <Fujitsu> Wasn't it uploaded less than 24 hours ago? |
|
[02:40] <Fujitsu> Ah. |
|
[02:41] <nenolod> sdl is fixed, libao is added (zsnes) |
|
[02:41] <nenolod> pulse is missing |
|
[02:41] <persia> Can't we just leave it for a week or so, and push a new upload then? It's truly huge. |
|
[02:41] <nenolod> persia: yes |
|
[02:41] <nenolod> persia: i'm more interested in fixing this GTK+ bug NOW. |
|
[02:41] * nenolod nods |
|
[02:41] <nenolod> because it's really starting to tick me off |
|
=== nenolod is now known as nenolod` |
|
[02:42] <nenolod`> well, pbuilder isn't going, so... |
|
[02:42] <nenolod`> lets see if conspire can run without gtk going insane |
|
[02:42] <nenolod`> ok. |
|
[02:43] <nenolod`> i'm running memtest86+ again. |
|
[02:43] <nenolod`> back later |
|
[02:49] * bddebian starts build 150 of newpki-client :-( |
|
[02:50] <ScottK> crimsun: Thanks. |
|
[02:51] <Fujitsu> bddebian: What are you trying to do to the poor creature? |
|
[02:52] <ScottK> slangasek: Would you have a moment to push clamav out the door in feisty-backports? It fixes two CVEs currently open in the backports version. |
|
[02:52] <Fujitsu> Who needs viruses when you have vulnerable virus scanners? |
|
[02:52] <bddebian> Fujitsu: Build it with wxgtk2.6 |
|
[02:52] <Fujitsu> bddebian: Ah. Ew. |
|
[02:52] <ScottK> Better than 2.4. |
|
[02:52] <Fujitsu> Marginally. |
|
[02:52] <bddebian> Trying to get rid of 2.4 :-) |
|
[02:53] <ScottK> bddebian: Why not reach for the stars and do it with 2.8? |
|
[02:53] <bddebian> Debian doesn't have it yet :-( |
|
[02:53] <ScottK> Ah. You're doing Debian first. That makes sense. |
|
[02:54] <Fujitsu> Are we trying to prevent wx from turning into db4.x? |
|
[02:54] <bddebian> heh |
|
[02:54] <ScottK2> Yes. |
|
[02:55] <Fujitsu> We have 3 versions of that in universe now, I think. |
|
[02:55] <ScottK2> More as of the last day or two. |
|
[02:56] <persia> Fujitsu: Getting rid of wx2.4 has benefits beyond just duplication. It renders mojibake in all non-latin locales. |
|
[02:56] <Fujitsu> I thought we had 5 in total, and only 2 in main now. |
|
[02:56] <Fujitsu> Ah. |
|
[02:58] <Fujitsu> Are there only 4 packages left on 2.4, or am I looking at the wrong list? |
|
[02:59] <Fujitsu> nenolod: That wasn't a very long test. |
|
[03:00] <nenolod> Fujitsu, turns out my old memory was a different CAS timing, and i think that's what the problem was |
|
[03:00] <nenolod> so i just pulled out the remaining 512M dimm, and things seem to be stable |
|
[03:00] <nenolod> we'll see what happens ;) |
|
[03:00] <tuxmaniac> Hi. When is the last date until which I can propose/upload new packages for Hardy? |
|
[03:00] <tuxmaniac> Is it feature freeze date? |
|
[03:01] <Hobbsee> yes |
|
[03:01] <Fujitsu> Isn't it some time before FF, to give archive admins time to NEW them? |
|
[03:01] <tuxmaniac> good! thanks Hobbsee and wish you a happy and a great new year 2008 |
|
[03:01] <Hobbsee> Fujitsu: i thought it was all the same now |
|
[03:01] <Hobbsee> tuxmaniac: you too :) |
|
[03:01] <nenolod> at any rate |
|
[03:01] <nenolod> pbuilder fun |
|
[03:02] <tuxmaniac> setting up pbuilder is not fun for me. Darn! I wanna a faster connection. 256 Kbps is just not for pbuilder stuff! :-( |
|
[03:03] <tuxmaniac> Fujitsu, is that so? Can you please let me know the dates. I couldnt find that info in the releaseschedule wiki page |
|
[03:04] <tuxmaniac> FF seems to be in Feb |
|
[03:04] <LaserJock> there is no set date |
|
[03:04] <tuxmaniac> LaserJock, ok |
|
[03:04] <LaserJock> FF is when archive admins stop processing without an exception |
|
[03:04] <LaserJock> so you'll want to have them in a bit before FF to allow for that time |
|
[03:05] <ScottK> Fujitsu: Any idea why we still have libdb1-compat? |
|
[03:06] <Fujitsu> ScottK: No idea. |
|
[03:06] <ScottK> Apt-cache says no rdepends. |
|
[03:06] <Fujitsu> Could be some proprietary things that want it. |
|
[03:06] <Fujitsu> cjwatson might now, being the maintainer and all. |
|
[03:06] <ScottK> Sounds like an excellent rationale for removal. |
|
[03:06] <Fujitsu> *know |
|
[03:07] <nenolod> aha. |
|
[03:07] * ScottK makes a list. |
|
[03:07] <nenolod> i found the source of my memory problem, i think. |
|
[03:07] <nenolod> i had two different brands of DIMM being used in a dual-channel configuration |
|
[03:07] <nenolod> :D |
|
[03:07] <ScottK> Hobbsee: Are you able to publsh backports? |
|
[03:08] <Fujitsu> We might have a chance of getting db4.2 to go away if the OpenLDAP issue can be resolved. |
|
[03:08] <persia> nenolod: That surely wasn't the cause of your gnome-panel crash, was it? |
|
[03:08] <Hobbsee> ScottK: no |
|
[03:08] <ScottK> Hobbsee: OK. Thanks. |
|
[03:09] <Fujitsu> ScottK: You mean a backport with source changes? |
|
[03:10] <ScottK2> Fujitsu: Yes. It's been uploaded already. I need an archive admin to publish it. |
|
[03:10] <Fujitsu> Any core-dev can upload those, and Hobbsee should be able to approve them |
|
[03:10] <Hobbsee> oh. |
|
[03:10] <ScottK2> I've lost track of what the web interface can and can't do. |
|
[03:11] <Hobbsee> Fujitsu: i thought that needed special publishing scripts to accept |
|
[03:11] <Fujitsu> It can approve stuff from NEW and UNAPPROVED, and even do it properly. That's about it. |
|
[03:11] <ScottK2> Ah. |
|
[03:11] <ScottK2> Right. Should have said. |
|
[03:11] <Hobbsee> archive admin runes public page may well be incorrect / out of date though |
|
[03:11] <Fujitsu> Hobbsee: As far as I know there's nothing special about backports, except that there's a script on drescher to generate them. |
|
[03:12] <ScottK2> Hobbsee: If you wouldn't mind having a look and seeing if there's a clamav backport in feisty-backports waiting approval? |
|
[03:12] <Fujitsu> If you're making source changes, there should be no need for the script. |
|
[03:12] <ScottK2> Right. |
|
[03:13] <ScottK2> Makes sense. |
|
[03:13] <nenolod> persia, no |
|
[03:13] * ScottK2 waits to see. |
|
[03:13] <nenolod> persia, my X server started doing whacko stuff |
|
[03:14] <nenolod> persia, like, it forgot it had a mouse cursor, so i could hover the cursor over any app, and it would kill it with a BadCursor X11 exception |
|
[03:14] <Fujitsu> Nice. |
|
[03:14] <ion_> That’s a feature. No more need for running xkill separately. |
|
[03:15] <persia> nenolod: Ah. I can't reproduce that one :) |
|
[03:16] <nenolod> ion_, yeah, except it kept killing my irc client whenever i went to go click on something in it |
|
[03:16] <nenolod> ion_, which started to annoy the living shit out of me |
|
[03:17] <nenolod> oh well, i just replaced the other 2 banks with more dimms out of that same batch |
|
[03:17] <nenolod> so now i have 4GB of server ram in here ;) |
|
[03:17] <Hobbsee> ScottK2: i'm heading out, sorry |
|
[03:17] <nenolod> persia, going to look into gtk+ FD exhaustion issue now |
|
[03:17] <ScottK2> Hobbsee: OK. Maybe slangasek will show up. |
|
[03:18] <nenolod> oh great |
|
[03:18] <nenolod> ubuntu renumbered my devices again |
|
[03:18] <nenolod> er, not ubuntu, alsa |
|
[03:18] <nenolod> just more reason to get libflashsupport fixed ;) |
|
[03:32] <SirBob1701> hmm guess i'll request something to work on tommorow |
|
[03:34] <nenolod> persia, i added an appointment to evolution to remind me to file a bug about adding libpulse to ia32-libs in 5 days. ;) |
|
[03:35] <ScottK2> nenolod: No reason not to file the bug now. |
|
[03:35] <SirBob1701> what ide's do you guys use? |
|
[03:36] <ScottK2> kate and vim are my favorites ;-) |
|
[03:36] * Fujitsu uses vim. |
|
[03:36] <persia> nenolod: Please do file the bug now, and include the patch, just don't subscribe the sponsors until it's ready to be uploaded. |
|
[03:36] <SirBob1701> i like vim |
|
[03:36] <ion_> gvim without the tool/menu/scrollbars. |
|
[03:36] <nenolod> persia, i'd like to restructure the ia32-libs package from the ground up someday |
|
[03:37] <persia> SirBob1701: Depending on how much you like vim, you might want to use it in non-IDE mode. I find it's easier to just keep two terminals open. |
|
[03:37] <SirBob1701> persia: i usually do do it all terminal based. i don't use gvim or any of that |
|
[03:38] <persia> nenolod: If you're up for that, now's your chance. It'll take long enough to do that most people won't complain about another 500MB download. Please fix gnome-panel first :) |
|
[03:51] <nenolod> persia, yes. gnome-panel is my current project, but it's not crashing atm when i run pbuilder |
|
[03:51] <nenolod> persia, it's being picky about when it wants to do it |
|
[03:52] <nenolod> persia, i do know where the crash is, though, and knowing is half the battle. |
|
[03:52] <nenolod> but at least my X server is presently sane. ;) |
|
[03:52] <nenolod> so hopefully i can get some work done |
|
[03:54] <nenolod> persia, nvidia drivers also need to be bumped, but that's main, and it's a no-change-upload |
|
[03:55] <persia> nenolod: Why do they need a bump? For the new X? |
|
[03:56] <nenolod> persia, nod |
|
[03:56] <nenolod> i'm using nv right now which is unaccelerated blitting |
|
[03:57] <nenolod> SirBob1701, i use nano |
|
[03:57] <nenolod> SirBob1701, don't listen to these emacs and vi fans. nano is the best. |
|
[03:58] <SirBob1701> nenolod: bit biased are we? haha |
|
[03:58] <ion_> Yeah, every editor that requires you to use more keypresses for everything is good. |
|
[03:58] <LaserJock> gedit FTW! |
|
[03:58] <ScottK> SirBob1701: You'll love the intuitive interface were ctrl o means save. |
|
[03:58] <SirBob1701> i like :w |
|
[03:59] <LaserJock> ScottK: obviously ;-) |
|
[03:59] <nenolod> ScottK, nano is intuitive to me |
|
[03:59] * ScottK too |
|
[03:59] <nenolod> but i'm a little bit nuts apparently |
|
[03:59] <SirBob1701> you guys european? |
|
[03:59] <nenolod> so :P |
|
[03:59] <ScottK> Not me. |
|
[03:59] <LaserJock> not I |
|
[03:59] <SirBob1701> hmm |
|
[03:59] <ScottK> nenolod: You also use Gnome, so you're doubly suspect in my book. |
|
[03:59] <nenolod> ScottK, :D |
|
[04:00] <nenolod> ScottK, i only use gnome because right now i find xfce's direction to be too vague ;) |
|
[04:00] <SirBob1701> I like xfce network manager doesn't work the way i'd like all the time in it tho |
|
[04:00] <SirBob1701> actually never |
|
[04:00] <nenolod> networkmanager sucks |
|
[04:00] * nenolod needs to remove it |
|
[04:01] <SirBob1701> lol what do you use? |
|
[04:01] <SirBob1701> or want to use? |
|
[04:01] <nenolod> i already killed off tracker for wasting my disk I/O resources building an index i won't ever use |
|
[04:02] <SirBob1701> ya i should prolly yank that out. i only use locate |
|
[04:02] <nenolod> SirBob1701, /etc/network/interfaces |
|
[04:02] <SirBob1701> nenolod: i jump too many wireless networks |
|
[04:02] <nenolod> SirBob1701, this is a desktop |
|
[04:02] <SirBob1701> there you go |
|
[04:02] <nenolod> SirBob1701, networkmanager is fine for my laptop |
|
[04:03] <SirBob1701> honestly its only on my desktop back at school cause i'm lazy |
|
[04:03] * nenolod wishes there was a ubuntu-desktop-lite |
|
[04:03] <SirBob1701> haha |
|
[04:03] <nenolod> you know, gnome without networkmanager, tracker, and other crap i don't want |
|
[04:03] <SirBob1701> gobuntu? |
|
[04:03] <persia> nenolod: What would it contain? Maybe you seek ubuntu-standard? |
|
[04:04] <nenolod> SirBob1701, networkmanager, tracker, and other crap i don't want, are all free software |
|
[04:04] <SirBob1701> ah |
|
[04:04] <nenolod> persia, gnome, ubuntu-artwork, other useful tools like gimp, OO.o and such |
|
[04:04] <ScottK> Fortunately they are easliy removed |
|
[04:05] <SirBob1701> i mike switch over to xubuntu. gotta see how vmware workstation works in it |
|
[04:06] <persia> SirBob1701: Should be no change from how it works in any other flavour, due to the nature of dependencies management. |
|
[04:06] <SirBob1701> hmm |
|
[04:09] <nenolod> persia, i think ia32-libs should be a metapackage for various lib32* |
|
[04:09] <SirBob1701> do you guys all have day jobs and do all this in your free time? |
|
[04:09] * Fujitsu doesn't see any Canonical employees lurking, so assumes so. |
|
[04:10] <ScottK> Almost all. |
|
[04:10] * ScottK sees at least one. |
|
[04:10] <nenolod> SirBob1701, well i make money working with linux, but i don't make money packaging up software for debian/ubuntu ;p |
|
[04:10] * TheMuso sees at least 3. |
|
[04:11] <Fujitsu> Sorry, I meant lurking in the `likely to be around' sense, not in the IRC `always present but not speaking' sesnse. |
|
[04:11] * ScottK2 uses linux to make money, but doesn't get directly paid for working on Ubuntu. |
|
[04:11] * nenolod declares memory issue fixed since his box hasn't gone into automatic xkill feature mode |
|
[04:12] <nenolod> ;) |
|
[04:13] <SirBob1701> i took a job for after graduation devloping windows applications. guess i'm a traitor lol |
|
[04:14] <nenolod> no, just foolish |
|
[04:15] <nenolod> people pay big bucks for linux work |
|
[04:15] <nenolod> microsoft pays their employees fairly poorly |
|
[04:15] <TheMuso> nenolod: You know of this how? |
|
[04:16] <nenolod> TheMuso, there's a fairly large microsoft installation where i live |
|
[04:16] <nenolod> i hear of these things :) |
|
[04:16] <SirBob1701> i'm not working for microsoft i'm working for a dutch company |
|
[04:16] <SirBob1701> only in america 2 years |
|
[04:16] <LucidFox> SirBob1701> Nobody can forbid you to develop for a dying platform. The only question is whether you want that yourself. |
|
[04:16] <nenolod> TheMuso, most of the microsoft developers i know are below the poverty line |
|
[04:17] <SirBob1701> LucidFox: point taken |
|
[04:17] <TheMuso> nenolod: Ouch. |
|
[04:17] <nenolod> TheMuso, well it's why windows is crap |
|
[04:17] * persia champions platform agnosticism for employment, so long as it provides time for Ubuntu |
|
[04:17] <LaserJock> I think I might be below the poverty line ... |
|
[04:18] <nenolod> TheMuso, nobody is motivated to do a good job because they are paid so poorly |
|
[04:18] <TheMuso> nenolod: I'm not so sure I believe that. |
|
[04:18] <nenolod> TheMuso, well i'm sure it's one of the reasons |
|
[04:18] <nenolod> windows is also needlessly complex |
|
[04:18] <nenolod> that's probably another ;) |
|
[04:18] <persia> nenolod: I'm not. I'm paid very poorly to work on Ubuntu, and don't believe it impacts my motivation. |
|
[04:19] * minghua doesn't count "work for free" as "be paid very poorly". :-) |
|
[04:19] * persia thinks minghua has never held a "working visa" |
|
[04:20] <nenolod> persia, yes, but, you're motivated because you're making a /difference/ |
|
[04:20] * persia loses the set |
|
[04:20] <ScottK> LaserJock: But you're a graduate student. You've volunteered to be poor. |
|
[04:20] <minghua> persia: No. |
|
[04:20] <nenolod> persia, your average microsoft employee is doing whatever he is told to do. he has no latitude for making his own choices about defining the direction in which a specific component of a product should go |
|
[04:20] <LaserJock> ScottK: true ... although I don't feel much like a volunteer |
|
[04:21] <nenolod> when working on ubuntu, you are in control of your own tasks |
|
[04:21] <persia> nenolod: To me that would be more of an issue than the remuneration. |
|
[04:22] <nenolod> persia, yes, but if you have no choice in how to go with something, and you are being paid poorly to do this work, than odds are you're not going to be very non-apathetic about it |
|
[04:23] <persia> nenolod: I've been apathetic and bitter while being paid very highly to do things I had no choice about. It's not about remuneration. |
|
[04:23] <nenolod> persia, true |
|
[04:23] <nenolod> persia, but at least you can be apathetic and bitter in style with proper remuneration. ;) |
|
[04:25] <nenolod> but the backbone of the linux economy is small privately owned businesses anyway |
|
[04:25] <persia> s/linux/any/ |
|
[04:25] <nenolod> companies like redhat, canonical, while being a large component, are not the majority share of how linux (and FOSS in general) gets developed |
|
[04:26] <nenolod> whereas in windows, you're mostly doing business with giant corporations which are inhumane |
|
[04:26] <nenolod> and don't really care if you like their product or not, as long as you didn't download it from a torrent |
|
[04:30] <SirBob1701> i'll go for inhumane if it pays well :) |
|
[04:31] <nenolod> i wouldn't |
|
[04:31] <LaserJock> I think I'd just go for a job at this point |
|
[04:31] <LaserJock> I've been in school too long |
|
[04:31] <nenolod> i'd rather make barely enough to live on and do the right thing, than make lots of money doing the wrong thing |
|
[04:32] <SirBob1701> uh oh somebody has morals ;) |
|
[04:32] <LaserJock> nenolod: for a great many people "right thing" can mean a lot of thing I think |
|
[04:32] <LaserJock> for instance providing for their family, etc. |
|
[04:32] <nenolod> LaserJock, indeed |
|
[04:33] <SirBob1701> providing for oneself... |
|
[04:33] <persia> nenolod: Just for the record, large inhumane corporations also use linux and FOSS (frequently wrapped by Value-Added-Resellers who charge lots of money for training sessions and GPL software). |
|
[04:33] <nenolod> persia, AKA: redhat enterprise linux ? :) |
|
[04:35] <persia> nenolod: No. More like Niksun |
|
[04:35] <nenolod> at least canonical's support contract offerings i have heard are fair |
|
[04:35] <nenolod> redhat support contracts are like hundreds of thousands of dollars for enterprises |
|
[04:35] <nenolod> from what i have heard |
|
[04:35] <nenolod> i could be wrong |
|
[04:36] <nenolod> persia, oh dear |
|
[04:36] <persia> nenolod: Yep. You've never paid so much for snort and wiretap before :) |
|
[04:36] <LucidFox> Does anyone here run hardy? (Not in pbuilder, but with an actual shell) |
|
[04:36] <persia> LucidFox: Several people. |
|
[04:37] <nenolod> LucidFox, i'm running it |
|
[04:37] <Vorian> LucidFox: me too |
|
[04:37] <nenolod> LucidFox, i thought it went insane, but it turned out that i had a memory DIMM being overclocked to all hell |
|
[04:37] * Fujitsu has been running it for more than 2 months. |
|
[04:37] <nenolod> :D |
|
[04:37] <LucidFox> I'd like someone to execute "apt-get build-dep avidemux" and report if anything is missing |
|
[04:37] <LucidFox> to kill bug #71802 |
|
[04:37] <ubotu> Launchpad bug 71802 in avidemux "apt-get build-dep fails " [Low,Incomplete] https://launchpad.net/bugs/71802 |
|
[04:37] <persia> LucidFox: How would a live system differ from a chroot for that test? |
|
[04:38] <LucidFox> It wouldn't. |
|
[04:38] <LucidFox> You can run it in a chroot. |
|
[04:38] <persia> LucidFox: You know about pbuilder --login, right? |
|
[04:38] <Fujitsu> debcheck says it's OK on all archs. |
|
[04:38] <LucidFox> I didn't. Sorry. :) |
|
[04:38] <Vorian> LucidFox: i can confirm an error if that's what you are looking for |
|
[04:39] <Fujitsu> But indeed it fails here. |
|
[04:39] <persia> LucidFox: No need to apologise: that's a large part of why we have this channel :) |
|
[04:40] <Vorian> persia: !!! that's awesome! |
|
[04:40] <nenolod> pbuilder --login is indeed very useful. |
|
[04:40] * persia still prefers schroot on LVM |
|
[04:40] <Vorian> i've been wondering all day, "where the heck is ~/.pbuilderrc" |
|
[04:41] <persia> LucidFox: Works for me |
|
[04:41] <nenolod> Vorian, it's not in your pbuilder |
|
[04:41] * Fujitsu finds schroot on LVM to be very good. |
|
[04:41] <nenolod> Vorian, it's a file you put in your home directory. |
|
[04:41] <nenolod> Vorian, it allows you to override default behaviours of pbuilder |
|
[04:41] <Vorian> nenolod: aye, and when I just did the pbuilder --login it created it |
|
[04:42] <nenolod> i should see if this dimm i accidentally overclocked to all hell because it didn't have an SPD chip on it is still usable |
|
[04:43] <nenolod> probably not |
|
[04:43] <crimsun> ah, nice, so it was a hardware issue? (likely?) |
|
[04:43] <persia> LucidFox: I can reproduce the bug in gutsy (and don't have a feisty chroot handy to test more). |
|
[04:43] <nenolod> crimsun, the kpanic is not. |
|
[04:43] <nenolod> crimsun, cause it happens on my other amd64 box |
|
[04:44] <crimsun> nenolod: right, was referring to the gtk+ |
|
[04:44] <LucidFox> It does occur in gutsy. |
|
[04:44] <LucidFox> Because avidemux used to depend on a library removed there. |
|
[04:44] <nenolod> crimsun, well, yes and no. there's a bug in the recent files chooser code relating to FD exhaustion. |
|
[04:44] <crimsun> and ya gotta love pci device enumeration being nondeterministic. |
|
[04:45] <nenolod> crimsun, but my gtk going insane is indeed fixed after pulling that dimm |
|
[04:45] <nenolod> it's been up for ~90 minutes now, which is much longer than before ;) |
|
[04:46] <persia> Does anyone happen to know why the fortran support libraries still depend on gcc 3.4? |
|
[04:46] <nenolod> persia, yes |
|
[04:47] <nenolod> cool |
|
[04:47] <nenolod> ia32-libs is broken in a new way now |
|
[04:48] <Vorian> do i copy /etc/pbuilderrc to ~/.pbuilderrc? |
|
[04:49] <nenolod> oh. |
|
[04:49] <nenolod> nevermind. |
|
[04:49] <nenolod> it's because of a thirdparty repo. |
|
[04:50] <nenolod> :D |
|
=== cassidy_ is now known as cassidy |
|
[04:57] <nenolod> persia, going to work on gtk+2.0 bug |
|
[04:57] <persia> nenolod: Thanks. |
|
[04:57] <nenolod> if i can wake up, that is |
|
[04:57] <ScottK> Vorian: Why would you want to do that? |
|
[04:57] * persia proferrs purple pep pills |
|
[04:58] <Vorian> ScottK: I don't have a ~/.pbuilderrc |
|
[04:58] <nenolod> Vorian, you don't need one |
|
[04:58] <ScottK> Vorian: Is this causing a problem? |
|
[04:58] <Vorian> no, not at all |
|
[04:59] <ScottK> Then you don't need one. |
|
[04:59] <Vorian> alrighty then |
|
[04:59] <ScottK> The only reason you might is to set per user pbuilder prefs differently |
|
[04:59] * minghua gets back to his Ubuntu machine and installs firefox-3.0. |
|
[05:00] <nenolod> http://launchpadlibrarian.net/11160730/ThreadStacktrace.txt |
|
[05:00] <Vorian> ScottK: thanks for the tip :) |
|
[05:00] <nenolod> this isn't very useful. |
|
[05:00] <persia> ScottK: You might be interested in following http://wiki.debian.org/GfortranTransition as part of your longer-term goal to reduce the gcc version count |
|
[05:01] <ScottK> That or I might get a severe headache just thinking about it. |
|
[05:02] <persia> ScottK: Looks to me like progress is well underway, and we've a good chance of being able to mostly sync, with only 5-10 patch applications (and some rebuilds). |
|
[05:02] <ScottK> I'm currently wrapping my head around doing a libclamav2 to libclamav3 transition in backports. |
|
[05:04] <persia> Don't let me distract you then: that's needful and painful enough :) |
|
[05:05] <nenolod> i don't see how it could crash there |
|
[05:05] <nenolod> bleah |
|
[05:06] <nenolod> oh. |
|
[05:06] <nenolod> it's a GSlice |
|
[05:06] <nenolod> no wonder the address is weird |
|
[05:10] <nenolod> idle_populate_func() isn't thread safe either. |
|
=== bigon is now known as bigon` |
|
[05:14] <nenolod> persia, ok. i figured it out. |
|
[05:14] <nenolod> persia, what happens is that occasionally, idle_populate_clean_up() gets called BEFORE idle_populate_func() |
|
[05:15] <nenolod> persia, i'm trying to figure out how to recover from that safely |
|
[05:25] <nenolod> persia, i have a debdiff which fixes it for me. would you like to try it? |
|
[05:27] <persia> nenolod: Sure. I'll give it a shot. You've attached it to which bug again? |
|
[05:28] <nenolod> i'll do that in a bit. i'm still testing it out ;) |
|
[05:31] <nenolod> persia, what should i change a core package to for Maintainer: ? |
|
[05:33] <persia> nenolod: For the first time today, you've asked me specifically a non-specific question. Please don't. The answer is "Ubuntu Core Developers <ubuntu-devel-discuss@lists.ubuntu.com>" according to https://wiki.ubuntu.com/DebianMaintainerField |
|
[05:34] <nenolod> sorry about that |
|
[05:34] <nenolod> i should have just looked at gcc ;) |
|
[05:35] <nenolod> oh. i hate them. they have a control.in. |
|
[05:40] <ion_> Hey, it could be worse. At least they don’t have rules.in |
|
[05:40] <persia> ion_: Is that even permitted? |
|
[05:40] <ion_> I sincerely hope not. :-D |
|
[05:41] <minghua> How do you generate rules if you have rules.in? |
|
[05:42] <ion_> You could ”bootstrap” it by initially writing a small debian/rules that just generates debian/rules from debian/rules.in. ;-) |
|
[05:42] * persia discovers it is permitted, as long as rules is generated at packaging time, as part of the packaging tool that shall not be named. |
|
[05:42] <minghua> Ah right. Generated at packaging time. |
|
[05:43] <nenolod> persia, it's attached to bug 180463 |
|
[05:43] <ubotu> Launchpad bug 180463 in gtk+2.0 "gnome-panel crashed with SIGSEGV in idle_populate_func()" [Undecided,New] https://launchpad.net/bugs/180463 |
|
[05:44] <StevenK> persia: Twitch |
|
[05:44] <persia> StevenK: Yes. This is why the packaging tool shall not be named, and any REVU packages using it are rejected summarily. |
|
[05:47] * nenolod goes back to enjoying pbuilder without gnome-panel crapping out |
|
[05:48] <persia> nenolod: I'm looking at the patch, but I don't understand why it's good. Could you explain a little? |
|
[05:48] <nenolod> persia, ah, yes. the part which makes it work is not shown. |
|
[05:49] <nenolod> in gtk_recent_chooser_menu_populate(), it checks to see if populate_id is set to non-zero. |
|
[05:49] <nenolod> if so, it returns to avoid recursion. |
|
[05:49] <persia> nenolod: Yes. That's the part I'm having trouble with. It looks like you are delaying the variable assignment, and checking it later, but I don't understand why the problem then goes away. |
|
[05:49] <persia> Ah. That makes sense. Thanks. |
|
=== brandonperry is now known as bperry |
|
[05:50] <nenolod> in vanilla gtk, idle_populate_func() always sets populate_id to zero regardless of whether or not it should be. |
|
[05:51] <nenolod> at any rate, that's enough to fix it for me, but the underlying problem is still there: gtk_recent_chooser_menu_populate() should never be called more than once. |
|
=== emgent_ is now known as emgent |
|
[05:52] <nenolod> so, my patch sets populate_id to 0 when idle_populate_func() will never be called again. |
|
[05:53] <nenolod> i also check populate_id to in idle_populate_clean_up() to make sure that it hasn't been called preemptively (due to possible thread race condition issues) |
|
[05:53] <persia> nenolod: OK. So just moving it before the return FALSE; makes it clean. Got it. Amusingly enough, I'm not triggering it building your candidate. |
|
[05:54] <nenolod> what's strange is: anyone who knows GTK should know that the way that code worked was entirely bogus |
|
[05:54] * nenolod wonders if they audit gtk2 anymore ;) |
|
[05:55] <nenolod> the part in idle_populate_clean_up() where populate_id is just a sanity check to ensure that populate_id is zero before it frees the context data that idle_populate_func() uses |
|
[05:58] <nenolod> persia, what is curious to me though is: why is gtk_recent_chooser_menu_populate() being called in gnome-panel |
|
[05:58] <nenolod> sadly, we probably won't ever know the answer to that |
|
[05:59] <nenolod> because gobject wraps everything up in a hard to debug mess |
|
[05:59] <persia> nenolod: FTBFS in my chroot :( http://paste.ubuntu.com/3297/ |
|
[05:59] <nenolod> bleah |
|
[06:00] <nenolod> persia, i copied in the wrong patch :( |
|
[06:00] * persia cheers the power of peer review |
|
[06:03] <nenolod> persia, try that. it's new and improved. |
|
[06:06] <persia> With the new patch, the build triggers the bug :) |
|
[06:08] <nenolod> persia, :D |
|
[06:18] <persia> nenolod: I'm off for a bit: I'll update the bug with my results after a dozen builds or so. |
|
[06:20] <slangasek> ScottK2: there's a clamav package in feisty-backports/unapproved, yes |
|
[06:20] <ScottK2> slangasek: Would you please approve it. |
|
[06:20] <ScottK2> slangasek: It fixes a couple of CVEs and I'll sleep better knowing we're all caught up. |
|
[06:21] <ScottK2> Or more precisely fixes the package so that it will actually build on Feisty. |
|
[06:21] <LucidFox> Hmm, apport now reports upgrade problems. Very useful. |
|
[06:22] <slangasek> ScottK2: are there guidelines on accepting backports that I should be following? |
|
[06:23] <ScottK2> It should be approved by a member of ubuntu-backporters (I'm one of those). |
|
[06:24] <ScottK2> Source backports should be rare and must be uploaded by a core-dev. Crimsun took care of that for me. |
|
[06:24] <ScottK2> Other than that, not that I know of. |
|
[06:24] <ScottK2> slangasek: ^^^ |
|
[06:24] <slangasek> right, good enough. :) |
|
[06:25] <ScottK2> Version numbering should be sensible would be another rule I guess. This should be fine on those grounds. |
|
[06:25] <slangasek> acceptzored |
|
[06:25] <ScottK2> slangasek: Thanks. |
|
[06:26] <ScottK2> And LP is mighty quick with the mail. |
|
[06:26] <nenolod> persia, great. thanks for testing my fix. |
|
[06:26] * slangasek wonders why content is split between wiki.ubuntu.com and help.ubuntu.com |
|
[06:26] <nenolod> persia, if it doesn't whack it, then we'll have to look for other causes. |
|
[06:35] <LucidFox> persia, I have prepared a debdiff for avidemux 2.4 final, if you still care about that :) |
|
[06:36] <LucidFox> I've also been approved into ubuntu-bugcontrol |
|
=== asac_ is now known as asac |
|
[08:33] <slytherin> how come a package gets built in Debian if a build dependency is missing? |
|
[08:34] <Fujitsu> slytherin: Because Debian allows binaryful uploads. |
|
[08:35] <Fujitsu> So arch-indep packages will only ever get built on the maintainer's system. |
|
[08:35] <Fujitsu> Most don't even use pbuilder, it seems. |
|
[08:35] <imbrandon> slangasek: that was thanks to our wonderfull community marketing |
|
[08:35] <Fujitsu> That leaves the Debian archive in a very questionable state, but they don't seem to like the idea of binary-only uploads. |
|
[08:36] <Fujitsu> imbrandon: What was? |
|
[08:36] <Fujitsu> Er, source-only uploads. |
|
[08:36] <imbrandon> Fujitsu: help. and wiki. split |
|
[08:36] <Fujitsu> Ah, yes. |
|
[08:38] <slytherin> Fujitsu: libcommons-lang-java has ftbfs in ubuntu and guess what the build dependency 'ant' is missing. I have just fixed it and will upload debdiff soon. |
|
[08:42] <Fujitsu> slytherin: Right, that's because the Debian maintainer is probably braindead. |
|
[08:43] <Fujitsu> Make sure there's a bug in Debian - Lucas has probably already filed one. |
|
[08:43] <slytherin> let me check |
|
[08:44] <slytherin> no bugs found in debian. I will file one. |
|
[08:45] <Fujitsu> Check that it's a problem in Debian first, please. |
|
[08:46] <slytherin> Fujitsu: problem must be there since ubuntu version was a sync form debian. Still I will make sure |
|
[08:47] <slytherin> yes, problme is present in debian too |
|
[08:48] <warp10> Please review http://revu.ubuntuwire.com/details.py?package=tennix a cute little tennis game. It had 1 advocacy yet, and is waiting for the second one. |
|
[09:08] <slytherin> Fujitsu: I am getting 'Relay access denied' error when trying to use reportbug. Is there anyway I can submit report from gmail? |
|
[09:09] <LucidFox> slytherin> Submit a mail to submit@bugs.debian.org |
|
[09:09] <LucidFox> with the subject being the bug name |
|
[09:09] <LucidFox> and the first two lines being: |
|
[09:09] <LucidFox> Package: (package) |
|
[09:10] <LucidFox> Version: (Debian version) |
|
[09:10] <slytherin> LucidFox: I have the report created by reportbug. Just want to know if I need to sign the message |
|
[09:12] <Fujitsu> No signature is necessary. |
|
[09:14] <slytherin> Ok, done. |
|
[09:14] <slytherin> By the way, debdiff for ubuntu is available at bug 180502 |
|
[09:14] <ubotu> Launchpad bug 180502 in libcommons-lang-java "[patch] Fix for FTBFS - 'ant' is missing from Build-Depends" [Wishlist,Triaged] https://launchpad.net/bugs/180502 |
|
[09:16] * Fujitsu attacks the triager of that bug. |
|
[09:16] <Fujitsu> Wishlit for an FTBFS. Nice one. |
|
[09:16] <LucidFox> Whoops. |
|
[09:17] <Fujitsu> Heh. |
|
[09:17] <LucidFox> Sorry, I was only approved yesterday. What should the importance for FTBFS be? Critical? |
|
[09:17] <Fujitsu> High. |
|
[09:17] <LucidFox> Ack. |
|
[09:17] <Fujitsu> Critical is more for bugs that kill things. |
|
[09:18] <Fujitsu> Like the one that dropped people from the admin group, for example. |
|
[09:18] <minghua> Or data loss, I suppose. |
|
[09:19] <Fujitsu> I think that comes under `kill things' |
|
[09:19] * minghua wonders how that bug wasn't caught by Lucas's mass rebuild. |
|
[09:19] <minghua> Fujitsu: Right. |
|
[09:20] <Fujitsu> minghua: That's what I was thinking above. |
|
[09:20] <Fujitsu> Has it only needed that dependency recently? |
|
[09:24] <minghua> Fujitsu: Digging in LP, I found 2.3-2 built fine on October last year. |
|
[09:25] <Fujitsu> The changelog for the Debian upload yesterday seems to indicate some move to ant. |
|
[09:25] <Fujitsu> * debian/rules: |
|
[09:25] <Fujitsu> - Rewritten to use CDBS' ant task. |
|
[09:26] <minghua> Yeah, so likely a new bug. |
|
=== Lure is now known as Lure_ |
|
=== Lure_ is now known as Lure |
|
[09:59] <persia> warp10: I couldn't find any evidence that the "Royalty Free" loops from http://www.melodyloops.com/ carried a license anywhere. Could you expand on your sources as to why they are public domain? |
|
[10:24] <nenolod> persia, how's the gtk patch going? |
|
[10:24] <nenolod> persia, you might have to restart to make the patched version kick in. i'm not certain. |
|
[10:25] <nenolod> (since gtk2 is already memory resident) |
|
[10:25] <persia> nenolod: Not had very much testing yet (build completed whilst I was away) |
|
[10:56] <AnAnt> man-di: hello, I made a package for JOSM that builds against icedtea, but I found out a couple of hours ago, that Debian GIS team are making a JOSM package (but based on sun java6) |
|
[10:56] <AnAnt> man-di: their package is even better than mine btw |
|
[10:57] <AnAnt> man-di: http://svn.debian.org/wsvn/pkg-grass/packages/josm/trunk/debian/?rev=0&sc=0 |
|
[10:58] <man-di> AnAnt: you can file a merge request and change the (build-)depends for Ubuntu |
|
[10:58] <AnAnt> man-di: how? |
|
[10:59] <AnAnt> man-di: how file a merge request I mean |
|
[10:59] <man-di> !merging |
|
[10:59] <ubotu> Merging is the process of including changes from other distributions (most commonly Debian) into Ubuntu packages, and is typically a major focus at the beginning of each Ubuntu development cycle. Please see https://wiki.ubuntu.com/UbuntuDevelopment/Merging for more information. |
|
[10:59] <man-di> !merging | AnAnt |
|
[10:59] <ubotu> AnAnt: please see above |
|
[11:00] <LucidFox> Wow. Bug #180511 is the first time I see a bug not filed in English. |
|
[11:00] <ubotu> Launchpad bug 180511 in evolution "keine Termine mehr im Kalender" [Undecided,Incomplete] https://launchpad.net/bugs/180511 |
|
[11:01] <persia> AnAnt: If you process as a merge, please strive for minimal variation from Debian. |
|
[11:01] <AnAnt> man-di: but it isn't in Debian yet |
|
[11:01] <man-di> AnAnt: oh |
|
[11:01] <AnAnt> man-di: it's just in alioth.d.o |
|
[11:02] <Fujitsu> LucidFox: I've seen quite a number. |
|
[11:02] <man-di> AnAnt: thats a problem, I dont know how Ubuntu handles this. I'm a Debian developer, sorry |
|
[11:02] <man-di> persia: any idea? |
|
[11:02] * jussi01 pokes persia |
|
[11:03] <AnAnt> well, I was thinking of sending my work to the Debian GIS team to merge my work into theirs |
|
[11:03] <persia> AnAnt: We've not yet organised exactly how to pull from alioth cleanly. Best current practice is to work with the alioth team, and apply any patches as "Ubuntu variations". This is less ideal for NEW packages, in which cases the REVU submitter typically works closely with the Debian packagers to ensure -0ubuntu1 matches SVN closely. |
|
[11:03] <AnAnt> btw, there is an RFP on Debian for icedtea |
|
[11:04] <AnAnt> isn't it possible that Matthias sends his IcedTea package to Debian? |
|
[11:04] <man-di> AnAnt: I know, I#m working on it this weekend |
|
[11:04] <AnAnt> man-di: ah, cool ! |
|
[11:05] <LucidFox> I'd _love_ to see icedtea in Debian. |
|
[11:05] <AnAnt> man-di: well if IcedTea gets into debian, then there won't be a need for Ubuntu variations, right ? |
|
[11:05] <man-di> LucidFox: me tee, but I fear too many people will depend on this and forget to think |
|
[11:05] <man-di> AnAnt: not necesarrily |
|
[11:06] <AnAnt> man-di: depend on what ? and think about about ? |
|
[11:07] <jonnymind> Hello, MOTU people. |
|
[11:07] <slytherin> AnAnt: depend on icedtea and forget think if the package can be compiled with GCJ |
|
[11:07] <man-di> AnAnt: depend on only icedtea and forget about the alternatives and then people on e.g. sparc file a lot of bugs because icedtea doesnt exist there |
|
[11:08] <AnAnt> oh |
|
[11:08] <AnAnt> man-di: well, what are the alternatives ? |
|
[11:09] <man-di> e.g. java-gcj-compat(-dev) |
|
[11:09] <AnAnt> man-di: that's part of gcc, right ? |
|
[11:09] <man-di> mostly, yes |
|
[11:09] <man-di> but there are more |
|
[11:10] <jussi01> persia: Jaako did the work, I only copied the patches/files in and compiled |
|
[11:11] <persia> jussi01: If Jaako made the patches, and you integrated the patches with the package, your name goes in the changelog, and you should mention that Jaako is the patch author in the specific entry for the patch. |
|
=== Lure_ is now known as Lure |
|
[11:13] <nenolod> is it a known issue that pornlib is utterly fucked in firefox-3.0? |
|
[11:13] <nenolod> or should i file a bug. |
|
[11:13] <nenolod> ;) |
|
[11:13] <Fujitsu> !ohmy | nenolod |
|
[11:13] <ubotu> nenolod: Please watch your language and topic to help keep this channel family friendly. |
|
[11:13] <nenolod> oh right. |
|
[11:13] <nenolod> sorry. :( |
|
[11:13] <jussi01> persia: no, you misubderstand. Jaako crated the patch, and gave me an edited control file/changelog. |
|
[11:13] * Hobbsee can imagine that certain people would be very happy to look into that... |
|
[11:13] <persia> jussi01: Did you need to make any further changes? |
|
[11:13] <jussi01> persia: nothing |
|
[11:14] <persia> jussi01: Then Jaako's name belongs in the changelog |
|
[11:14] <nenolod> let me rephrase is --> http://nenolod.net/pornlibisgreat.png <-- a known issue with pornlib in firefox3? |
|
[11:14] <jussi01> persia: are you able to test the deb I sent you? i haven a hardy system as yet |
|
[11:15] <persia> jussi01: No, firstly because I don't like binaries, and secondly because that deb isn't compatible with my system. |
|
[11:15] <nenolod> while i'm sure people enjoy images that are scaled looking like abstract art, i don't think it's very useful ;) |
|
[11:15] <jussi01> persia: ahh |
|
[11:17] <white> Hobbsee: so the final plan is you text me, when i am in .au? :) |
|
[11:17] <Hobbsee> white: do you prefer lunch or dinner? |
|
[11:17] <Hobbsee> but yeah, that sounds OK |
|
[11:18] <white> don't mind |
|
[11:18] <Hobbsee> ok |
|
[11:18] <white> just want to make sure we know how to actually fix a place and time, before i go offline here and we miss each other :) |
|
[11:18] <Hobbsee> cool, yup |
|
[11:19] <white> i'll be crying, if you don't call me during tuesday :) |
|
[11:19] <jussi01> ok, can someone remind me where the documentationon how to make a debdiff is? |
|
[11:19] <white> Hobbsee: become a DD and put your mobile phone number in LDAP |
|
[11:19] <white> at once :) |
|
[11:19] <Hobbsee> white: why do i want to become a dd? |
|
[11:20] <Hobbsee> jussi01: debdiff foo1.dsc foo2.dsc > foo.debdiff |
|
[11:20] <white> jussi01: debdiff $old_version*.dsc $new_version.dsc |
|
[11:20] <StevenK> white: My mobile number is in LDAP, feel free to call me |
|
[11:20] <raof> white: Shouldn't you have suggested that a couple of years ago if you wanted it done soon? |
|
[11:20] <StevenK> Hrm. Lowercaseroaf |
|
[11:20] <white> StevenK: you are in sydney? |
|
[11:20] <StevenK> white: Aye |
|
[11:20] <white> oh great :) |
|
[11:20] <Hobbsee> raof: you coming? |
|
[11:20] * raof is without scrollback |
|
[11:21] <StevenK> What are you using, telnet? |
|
[11:21] * white gets to see the whole sydney cabal, great :) |
|
[11:21] <raof> Hobbsee: If I'm in Sydney, then yes :) |
|
[11:21] <Hobbsee> irc over crackpipe. |
|
[11:21] <Hobbsee> raof: will you be in sydney on wednesday? |
|
[11:21] <raof> I'm currently in Perth, and my home box is down :( |
|
[11:21] <raof> No. Curses! |
|
[11:21] <Hobbsee> drat |
|
[11:21] <StevenK> white, Hobbsee: I'd prefer dinner, but lunch is do-able |
|
[11:21] <Hobbsee> fly home earlier then |
|
[11:22] <white> raof: will you be in alice springs at the end of the week? |
|
[11:22] <raof> I'm back on Friday. Probably at some godawful time in the morning if previous form is followed. |
|
[11:22] <nenolod> hi DarkMageZ |
|
[11:22] <white> StevenK Hobbsee: i am fine with whatever |
|
[11:22] <Hobbsee> ok |
|
[11:22] <DarkMageZ> hey nenolod |
|
[11:22] <raof> white: No. Perth is *quite* hot enough for me, thanks :) |
|
[11:22] <nenolod> DarkMageZ, you should see this cool bug in firefox3-pornlib |
|
[11:22] * white can't wait to leave German winter |
|
[11:22] * Fujitsu sends raof to Darwin. |
|
[11:23] * jussi01 would love to be in .au , but cant |
|
[11:23] <DarkMageZ> nenolod, lol. i remember they called the lib libporn. what's the bug? |
|
[11:23] <nenolod> DarkMageZ, https://bugs.launchpad.net/ubuntu/+source/firefox-3.0/+bug/180525 |
|
[11:23] <ubotu> Launchpad bug 180525 in firefox-3.0 "firefox-3.0 pornlib resizes images into works of abstract art" [Undecided,New] |
|
[11:23] <raof> Hot *and* humid, urgh! |
|
[11:23] <jussi01> #¤%#¤%¤ -20 degrees |
|
=== \sh_away is now known as \sh |
|
[11:23] <raof> jussi01: is that -20 degrees in some sensible scale, like Celcius? |
|
[11:23] <jussi01> raof: yes |
|
[11:23] <nenolod> firefox3 is a much nicer browser than firefox2. |
|
[11:23] <raof> jussi01: Oh, dear. |
|
[11:23] <nenolod> it doesn't memory leak. |
|
[11:23] <nenolod> i'm .. stunned. |
|
[11:24] <jussi01> raof: I am after all an aussie... |
|
[11:24] <nenolod> it just has a bunch of other bugs instead :D |
|
[11:24] <Fujitsu> nenolod: That's what I thought. It's also much nicer in other ways. |
|
[11:24] <Fujitsu> And it's fast. |
|
[11:24] <raof> nenolod: And Epiphany now builds against xulrunner, so it looks ggood. |
|
[11:24] <DarkMageZ> nenolod, lol. nice |
|
[11:25] * raof decides that telepathy-idle isn't for him, and gets a real local irc client |
|
[11:25] <nenolod> Fujitsu, the download manager in fx3 is a significant improvement |
|
[11:25] <\sh> moins |
|
[11:25] * slytherin waits for epiphany-webkit |
|
[11:25] <Fujitsu> nenolod: Is it? I haven't used Firefox much for many months. |
|
[11:25] <Fujitsu> slytherin: It's deliberately disabled in Ubuntu :( |
|
[11:26] <slytherin> Fujitsu: I know |
|
[11:26] <nenolod> slytherin, webkit isn't ready for primetime yet |
|
[11:26] <raof> slytherin: You can get something pretty close with epiphany right now. Yay native form widgets! |
|
[11:26] <Fujitsu> raof: Apart from the grey rectangles around them. |
|
[11:26] <persia> slytherin: If you really want webkit & GTK, try midori |
|
[11:27] <nenolod> midori is a neat little browser |
|
[11:27] <slytherin> persia: I will when I upgrade my machine |
|
[11:27] <nenolod> i hope that it someday gets complete enough to replace firefox in ubuntu entirely |
|
[11:27] <raof> fujitsu: That seems to have gone away for me? |
|
[11:27] * nenolod is "not a fan" of firefox |
|
[11:27] <persia> slytherin: You could backport it :) |
|
[11:27] <Fujitsu> raof: Hm, not for me. WHen did you last upgrade? |
|
[11:28] <slytherin> persia: too many dependencies will need to be backported |
|
[11:28] <Fujitsu> Has raof caught persia disease? |
|
[11:28] <slytherin> nenolod: webkit is ready, epiphany-webkit combo is no5t |
|
[11:28] <nenolod> slytherin, not true. webkit-gtk and webkit-qt have incomplete spots. |
|
[11:29] <RAOF> Fujitsu: I upgraded this morning sometime. Ish. |
|
[11:29] <DarkMageZ> nenolod, midori is Very fast! |
|
[11:29] <Fujitsu> RAOF: Hmmm. I upgraded a few hours ago, and I still have the rectangles. |
|
[11:29] <RAOF> Fujitsu: Hm. Some things have borders (gtk.entry, for example), but bettons don't. |
|
[11:29] <nenolod> slytherin, try to download something in a webkit-gtk browser. it just shows up as crap on the screen |
|
[11:29] <slytherin> nenolod: Right. How is that webkit's fault. The rendering engine is perfect. That is why Safari uses it |
|
[11:29] <RAOF> Or even buttons. |
|
[11:29] <persia> Fujitsu: I like to think of it as ease ;P |
|
[11:30] <StevenK> RAOF: You fixed your home box? |
|
[11:30] <nenolod> slytherin, when i say "webkit is not ready for primetime" i mean in the scope of ubuntu, which is not MacOS, and therefore incapable of running safari ;) |
|
[11:30] <RAOF> StevenK: No, I installed an actual irc client. |
|
[11:30] <StevenK> Heh |
|
[11:30] <RAOF> nenolod: Wine :P |
|
[11:30] <nenolod> RAOF, safari runs on wine now? it bailed a while back. |
|
[11:31] <slytherin> nenolod: Ok. Perhaps I was unclear. I was talking about webkit as rendering engine. I want to test webapps in two different rendering engines. |
|
[11:31] <RAOF> nenolod: According to some blog on my reading list, yes, it works. As long as you've got the mscorefonts in the right place. |
|
[11:32] <Ubulette> nenolod: i can't reproduce your ff3 bug |
|
[11:32] <nenolod> Ubulette, weird. |
|
[11:33] <nenolod> Ubulette, are you using the one in hardy? |
|
[11:33] <Ubulette> could you try in safe mode ? |
|
[11:33] <DarkMageZ> i can replicate nenolod's firefox bug. |
|
[11:33] <nenolod> Ubulette, yeah. one sec. |
|
[11:34] <RAOF> Can anyone see this? |
|
[11:34] <nenolod> Ubulette, same garbage |
|
[11:34] <StevenK> RAOF: No |
|
[11:35] * StevenK smirks |
|
[11:35] <Ubulette> i've tried with http://nenolod.net/canhasdesktop.png it looks fine, i can zoom in and out, no problem |
|
[11:35] <nenolod> Ubulette, i can zoom in, and it's fine |
|
[11:35] <nenolod> Ubulette, zoom out, it's crap. |
|
[11:35] <nenolod> Ubulette, what version are you using? |
|
[11:36] <nenolod> DarkMageZ is running the same one as I am and he gets the same crap |
|
[11:36] <nenolod> DarkMageZ, are you running x86 64UL or 32UL? |
|
[11:36] <nenolod> it might be a 64UL bug. mozilla guys have done it before. |
|
[11:36] <DarkMageZ> 32-bit hardy. the ff3 build in the repositories. |
|
[11:37] <nenolod> so it's not 32UL vs 64UL |
|
[11:37] <Ubulette> i'm with b3pre like hardy but not the one from hardy, i'm running the one from my ppa |
|
[11:37] <Ubulette> should not matter much |
|
[11:38] <nenolod> Ubulette, what *version* |
|
[11:38] <nenolod> Ubulette, i didn't ask where you got it, i asked what *version* |
|
[11:38] <nenolod> Ubulette, as in, `dpkg -s firefox-3.0 | grep Version` |
|
[11:38] * Hobbsee notes that Ubulette answered that |
|
[11:39] <nenolod> Hobbsee, no, Ubulette answered "like hardy but not the one from hardy" |
|
[11:39] <nenolod> ;p |
|
[11:39] <Hobbsee> b3pre |
|
[11:39] * DarkMageZ thinks this is offtopic for -motu and more appropriate for -mozillateam or -bugs. |
|
[11:39] <nenolod> yeah |
|
[11:40] <nenolod> -> -mozillateam it is |
|
[11:40] <Ubulette> b3pre from yesterday. providing i've packaged most of ff3 and xul1.9 myself, i'm pretty confident it doesn't matter |
|
[11:57] <slytherin> One question, can i install lintian and linda from hardy repository in gutsy? Or do I need to get them from gutsy-backports? |
|
[12:01] <geser> slytherin: check the depends |
|
[12:01] <slytherin> geser: other than that there shouldn't be any problem, right? |
|
[12:02] <geser> correct |
|
[12:02] <persia> slytherin: Best to get from backports if they are available. That indicates the backport was tested. You ought be able to run them directly, but this may be unsafe, and is definitely not supported. |
|
[12:07] <slytherin> persia: Ok. Downloading them from backports |
|
[12:09] <slytherin> persia: linda from backports gives warning about standards version 3.7.3 |
|
[12:09] <Fujitsu> linda is yet to be updated for 3.7.3 anywhere. |
|
[12:10] <persia> slytherin: Yes. That's unfortunately currently blocked by an effort to come up with a better way to handle the package analysis (with significantly reduced disk I/O). In the meantime, ignore her complaining about the version. |
|
[12:12] * slytherin leaves for now. More java package fixing later |
|
[12:18] <white> !info kradio hardy |
|
[12:18] <ubotu> kradio: Comfortable Radio Application for KDE. In component universe, is optional. Version 0.1.1.1~20061112-3 (hardy), package size 1986 kB, installed size 9384 kB |
|
[12:18] <POX_> Kmos: ping |
|
[12:19] <white> bah, can't you guys fix my bugs? :) |
|
[12:19] <persia> white: We try to be generous, to also give you a chance :) |
|
[12:19] <white> pfff |
|
[12:20] <IntuitiveNipple> Is there a channel for PPA tips and help? |
|
[12:21] <persia> IntuitiveNipple: #launchpad is the closest (but not really) |
|
[12:22] <IntuitiveNipple> I thought that was the case ... I'll go see :) |
|
[12:26] <DktrKranz> Vorian, around? |
|
[12:29] <POX_> Kmos: can I touch startupmanager's debian/changelog with your name before uploading? (kmap is no longer in Uploaders) |
|
[12:30] <Kmos> POX_: sure |
|
[12:52] * persia has just noticed a FTBFS due to a call to dh_iconcache, and notes that anyone interested in a few quick patches might want to look for packages not updated since early gutsy to see if they need s/dh_iconcache/dh_icons/ to avoid FTBFS in hardy. |
|
=== sbucat is now known as Hulman |
|
[13:00] * minghua wishes he has access to a mirror so that he can grep dh_iconcache. |
|
[13:01] <dennda> Hi |
|
[13:01] <persia> minghua: You could always generate a mirror :) |
|
[13:01] <dennda> I am just wondering: Isn't there a particular deadline for uploading packages to universe? https://wiki.ubuntu.com/HardyReleaseSchedule |
|
[13:01] <Hobbsee> yes - feature freeze on there |
|
[13:02] <persia> dennda: No, but the deadline for accepting them into hardy is Feature Freeze. Best to get it uploaded at least two weeks beforehand,. |
|
[13:03] <dennda> ok |
|
[13:03] <dennda> Thank you guys. We will hurry then :) |
|
[13:04] <dennda> And ladies |
|
[13:04] <dennda> of course |
|
[13:04] <minghua> persia: Yeah, But I don't have enough disk space, or bandwidth, or use of a mirror to justify it... :-) |
|
[13:05] <persia> minghua: You have usage caps? |
|
[13:05] <Hobbsee> a lot of places do |
|
[13:05] <minghua> persia: No. It's not common in US. I'm on a DSL line though, 300 KB/s tops, I think. |
|
[13:09] <persia> minghua: Just run while you are asleep for a few days, and you'll have a mirror soon enough. |
|
[13:10] <minghua> persia: Good to know. But I am lazy. So probably not going to happen. :-P |
|
[13:11] <minghua> And you really need an unpacked archive to do really efficient grepping. |
|
[13:12] <cyberix> What would be the most straight forward way for a _user_ to make a custom version of a package? |
|
[13:13] <cyberix> for personal use only |
|
[13:13] <cyberix> Is there a guide about it somewhere? |
|
[13:13] <cyberix> I'm talking about changing a precompilation option and such. |
|
[13:15] <persia> cyberix: I generally used apt-get build-dep; debuild when I did that. If people don't want the build-deps installed, they need to use pbuilder/sbuild. changelog doesn't really matter for purely private use. |
|
[13:16] <minghua> cyberix: debian/changelog actually matters, because of the version number. |
|
[13:16] <geser> pochu: Hi, are you working on merging mono-tools? |
|
[13:16] <cyberix> I recall doing apt-get build-dep; apt-get source; dpkg-buildpackage; But that didn't work because it erased my modifications to the source. |
|
[13:18] <minghua> cyberix: The correct sequence is: apt-get source <foo>; apt-get build-dep <foo>; <make your modifications>; dpkg-buildpackage |
|
[13:19] <cyberix> That is actually what I did |
|
[13:19] <cyberix> And it erased my modifications before building |
|
[13:19] <Hobbsee> sounds like yada |
|
[13:20] <Hobbsee> cyberix: which package? |
|
[13:20] * Fujitsu looks dangerously at Hobbsee. |
|
[13:20] <Hobbsee> Fujitsu: it does, though. sounds yada-based. |
|
[13:20] <cyberix> Can't recall. |
|
[13:20] <Fujitsu> Perhaps. |
|
[13:20] <Fujitsu> cyberix: If you're making changes in debian/control, look for a debian/control.in. |
|
[13:21] <StevenK> Sob |
|
[13:22] <StevenK> If it's yada, the details are in debian/package or some nonsense |
|
[13:22] <persia> Yes. debian/package. Complete repackaging patches accepted. |
|
[13:23] <Fujitsu> s/accepted/greatly encouraged/? |
|
[13:24] <cyberix> Ok. Got it to work. |
|
[13:24] <Fujitsu> What did you change |
|
[13:24] <Fujitsu> *? |
|
[13:25] <cyberix> Just made a test with hello package. Changed one letter. |
|
[13:25] <cyberix> Now sauna, then hacking++. |
|
[13:27] <tuxmaniac> Hi. I have done some changes inorder to make the "NEW" package build with pbuilder hardy base. The original package was already uploaded by another friend of mine and I would like to give a update for this. Can I do that? Will there be two uplaods in REVU then for the same package? |
|
[13:28] <StevenK> persia: I've got to stop repackaging yada-using packages as something else - dexter things I have a vendetta against yada and keeps mailing me about it |
|
[13:28] <persia> StevenK: You get a pass then. Recruit some others :) |
|
[13:29] <persia> tuxmaniac: You can upload to REVU, but your friend won't be able to comment (only the last uploader can comment). |
|
[13:33] <tuxmaniac> hmm persia but he has done most of the hard work and if we would like to comment on a few "feedacks" from MOTU guys on my upload? Its not possible? |
|
[13:33] <tuxmaniac> s/we/he |
|
[13:34] <persia> tuxmaniac: Better to send him the package then, and have him upload. |
|
[13:34] <tuxmaniac> persia, ok thanks |
|
[14:06] <minghua> "RPM has many more features, more of an industry standard, etc and yum has just as many features as apt including some apt doesn't have. There is a yum is faster and uses cache just like apt and even has plugins like fast mirror." |
|
[14:06] <minghua> Seen in a slashdot comment. Are these all true? |
|
[14:08] <geser> minghua: there are more rpm from 3rd-party-repos outside than deb but that doesn't say that the rpm works in your rpm-based distro |
|
[14:11] <minghua> geser: I know. But the quote was not talking about quantities of packages, it's talking about the features of RPM format and yum. |
|
=== bigon` is now known as bigon |
|
[14:15] <tuxmaniac> Hi. If a package name from upstream is alliance-5.0-20070718, I give 'alliance' as the package name in changelog, version is 5.0-20070718-0ubuntu1 and orig is alliance_5.0_20070718.orig.tar.gz |
|
[14:15] <tuxmaniac> is that right? |
|
[14:15] <tuxmaniac> i meant the version in changelog there |
|
[14:18] <persia> tuxmaniac: alliance_5.0-20070718.orig.tar.gz |
|
[14:18] <tuxmaniac> aah ok |
|
[14:20] <tuxmaniac> persia, but I get this error while doing a pdebuild |
|
[14:20] <tuxmaniac> <error> dpkg-source: error: source package has two conflicting values - alliance-5.0-20070718 and alliance </error> |
|
[14:21] <tuxmaniac> sorry. I am totally new to packaging. So asking such questions. I have seen through the wiki but couldnt find why this happens. |
|
[14:21] * minghua thinks persia got it right. |
|
[14:22] <minghua> The error message should be some mistake in the package. |
|
[14:22] * tuxmaniac checks |
|
[14:22] <minghua> a blind guess is the source package name in debian/control and debian/changelog don't match. |
|
[14:23] * man-di suspects the same as minghua |
|
[14:23] <minghua> another blind guess is the name of the directory is wrong. |
|
[14:23] <tuxmaniac> minghua, heheheh. you hit the bull's eye |
|
[14:23] * tuxmaniac bangs his head and makes the change |
|
[14:48] <\sh> ScottK: ping wine...libgif-dev fixed...it's on the usual place : |
|
[14:59] <n3xu|laptop> looking for original Quake font, plz pm me </amsg> |
|
[15:13] * persia finds it very odd when upstream releases an updated tarball including "dfsg" in the new version name |
|
[15:15] <Kmos> iceowl should be removed from hardy ? we have sunbird package |
|
[15:15] <persia> Kmos: sunbird isn't free |
|
[15:16] <persia> ls |
|
[15:17] <Kmos> we have it at universe |
|
[15:17] <Kmos> it's the mozilla calender |
|
[15:19] <persia> Kmos: Yes. I know. However, not all ubuntu derivatives can ship it, as it may not be patched without approval of the Mozilla Foundation. That is why iceowl exists. We need both for now. |
|
[15:20] <Kmos> persia: ah ok =) 0.5-2 doesn't build.. and 0.7-1 at debian builds. so i'll request a sync to fix the ftbfs |
|
[15:20] <Kmos> the iceowl |
|
[15:20] <persia> Kmos: Does 0.7-1 build on hardy? |
|
=== thekorn__ is now known as thekorn |
|
[15:21] <Kmos> persia: yes |
|
[15:21] <Kmos> 0.7-2 |
|
[15:21] <Kmos> :) |
|
[15:21] <zul> then request 0.7-2 |
|
[15:22] <Kmos> yep |
|
[15:22] <persia> Kmos: If you're sure, and you've tested things, and you can provide sufficient justification for the DIF exception, you may be able to request a sync, but definitely request 0.7-2 rather than 0.7-1. |
|
[15:22] <Kmos> done =) |
|
[15:22] <Kmos> i build the 0.7-2 |
|
[15:22] <Kmos> i made a typo |
|
[15:22] <Kmos> in hardy pbuilder |
|
[15:24] <Kmos> thanks for the help |
|
[15:26] <geser> persia: if we keep iceowl for other derivatives why not also the other ice* packages? |
|
[15:26] <persia> geser: I'm firmly of the opinion we should, if we really mean to release gobuntu, but maybe I'm opinionated in that regard. |
|
[15:27] <persia> There's some documentation about possible plans at https://wiki.ubuntu.com/Gobuntu/pkg-non-free |
|
[15:28] * minghua thinks Gobutu is going to be vaporware. |
|
[15:28] <minghua> But I'm usually pessimistic. |
|
[15:28] * persia offers minghua http://cdimage.ubuntu.com/gobuntu/daily/current/ |
|
[15:29] <persia> Looks like it's just firefox and thunderbird at this point, although I don't know the state of the kernel |
|
[15:30] <geser> I remember discussions about firmware without source in the kernel in the debian-devel ML but I don't know if it's resolved now |
|
[15:30] <minghua> persia: Maybe vaporware is the wrong word. I meant "distributions that have more developers than users". |
|
[15:31] <persia> minghua: Well, maybe. Personally, I think most gobuntu users might be happier with Debian, but it may be that the more frequent release cycle is considered interesting. |
|
[15:31] <minghua> Is RMS fine with kernel modules that don't work unless you have a binary blob? |
|
[15:32] <persia> Certainly not. If there's not a stripped free kernel for gobuntu, certain parties won't be pleased. |
|
[15:32] <persia> (stripped in the sense of missing blobs, not in the sense of missing symbols) |
|
[15:33] <minghua> persia: I mean, the kernel won't ship the binary blobs, but will ship modules (with source) that won't work unless the blobs are installed. |
|
[15:34] <minghua> I am asking because of the recent "mention, therefore recommend non-free software" argument RMS made. |
|
[15:35] <minghua> I don't think it's worth it to do so much work just to please RMS's ideology (or earn his recommendation). But it's just me. |
|
[15:35] <minghua> If somebody else wants to, more power to him. |
|
[15:36] <persia> minghua: That's the nature of free software: if someone else wants to, they can. That's the idea behind gobuntu, because for some parts of Ubuntu, we can't (which can be frustrating for both developers and bug submitters). On the other hand, I use Ubuntu, so take my arguments with seasoning of choice. |
|
[15:40] <mok0> As much as I sympathize with RMS' view and goals, one of the reasons I chose Ubuntu over Debian is the existence of packages with the restricted nVidia drivers. |
|
[15:40] <persia> mok0: Those exist in Debian as well (or at least did when I was a Debian user). |
|
[15:41] <mok0> Oh? My mistake ;-) |
|
[15:41] <mok0> An other reason is the friendly atmosphere in the ubuntu community... |
|
[15:42] <mok0> :) |
|
[15:43] <minghua> mok0: http://packages.debian.org/sid/nvidia-glx |
|
[15:44] <minghua> Admittedly they are not as easy to install as in Ubuntu. |
|
[15:44] <mok0> Yeah, the restricted software gui is awesome |
|
[15:44] <minghua> And the chance that you'll get it out-of-box is rather slim. But it's there. |
|
[15:49] <pochu> geser: not now, and I won't for some time, so feel free to do it yourself |
|
[15:54] <jonnymind> PPl I had not any other news about bug #174470 after the last fix |
|
[15:54] <ubotu> Launchpad bug 174470 in ubuntu "Package for the Falcon Programming Language" [Wishlist,In progress] https://launchpad.net/bugs/174470 |
|
[15:54] <jonnymind> A couple of weeks have passed... |
|
[15:56] <AnAnt_> Hello, I am making a package for themes that are based on Human themes , how should I mention the attribution in copyright file ? |
|
[15:57] <joejaxx> Good Morning All |
|
[15:57] <joejaxx> :) |
|
[15:59] <persia> jonnymind: I'm suspicious of that. Every REVU submission pending review was addressed over New Years, so either you have pending work from that, or your package was awaiting a previous update prior to the clear. |
|
[16:00] <jonnymind> Eh? |
|
[16:00] <jonnymind> Emilio Pozuelo Monfort wrote on 2007-12-20: (permalink) |
|
[16:00] <jonnymind> I fixed it and |
|
[16:00] <jonnymind> gniccolai wrote on 2007-12-27: (permalink) |
|
[16:01] <jonnymind> And then there's no new notification... |
|
[16:01] <AnAnt_> btw, can someone review this upload: http://revu.tauware.de/details.py?package=usplash-theme-ubuntume ? |
|
[16:01] <jonnymind> Afaik, there was no other issue open, but again afaik the fix has not been checked. |
|
[16:03] <persia> jonnymind: I commented with 15 action items at 00:16 UTC 1 January 2008 on falconpl. There has since been no update. Please check again. |
|
[16:06] * minghua hates software that can only be build once (i.e., make clean doesn't restore the source tree status). |
|
[16:06] <jonnymind> ? |
|
[16:06] <AnAnt_> I forgot to mention that Human themes are licensed under Creative Commons ShareAlike license |
|
[16:06] <jonnymind> where do I have to look? |
|
[16:06] <persia> minghua: It's bug-worthy. File away. |
|
[16:06] <minghua> jonnymind: I think persia means http://revu.tauware.de/details.py?package=falcon |
|
[16:06] * persia thanks minghua for the fast URL lookup |
|
[16:06] <minghua> persia: Not in archive yet. Just a piece of my pet software. |
|
[16:07] <minghua> persia: No problem, had it open in firefox. |
|
[16:09] <minghua> persia: BTW, I think Lucas should have caught most of one-build-only packages in his FTBFS-if-built-twice-in-a-row mass bug filing against Debian. |
|
[16:10] <jonnymind> persia: I am a bit confused about the copyright thing. |
|
[16:10] <persia> minghua: most, yes. There's quite a few Ubuntu-local packages that have that issue, and the Debian QA team is likely still processing QA uploads to fix some of the orphaned packages. |
|
[16:11] <jonnymind> persia: also, I checked the watch file, and it was correctly finding everything. |
|
[16:11] <jonnymind> persia: where do I reply to those comments? |
|
[16:11] <persia> jonnymind: uscan --report-status and uscan --force-download didn't work for me. What confuses you about debian/copyright? |
|
[16:12] <jonnymind> In the file, there are the copyright owner, the license and the years. |
|
[16:12] <persia> jonnymind: for reply, best to ask in this channel. You can leave a rebuttal as a comment on REVU, but clarification, etc. is best done here, as few people look for updated comments on REVU. |
|
[16:12] <jonnymind> IC |
|
[16:15] <persia> jonnymind: Your debian/copyright is very oddly ordered. It does not contain the download location. It does not specify the authors (although in this case they happen to match the copyright holder and packager), has an inappropriate plural for "Copyright owners", doesn't follow standard formatting, and fails to include a license for the packaging. |
|
[16:15] <jonnymind> So, the package license is different than the software license? |
|
[16:16] <persia> jonnymind: Possibly. Generally they are the same, but this should be made clear. Further, if you used other packages as sources when creating your packaging, you may be constrained by the terms of the licensing of the packaging for those packages. |
|
[16:17] <jonnymind> This is free software and free for everyone to use. |
|
[16:17] <jonnymind> Sorry, |
|
[16:18] <jonnymind> I meant: I just need a thing that is ok for you. I thought you wanted the software copyright in the copyright but it seems it's not the case. |
|
[16:18] <jonnymind> I'll copy another copyright then. |
|
[16:19] <persia> jonnymind: debian/copyright contains lots of different information. Best to read the email which URL is in the comment, and follow that. |
|
[16:19] <jonnymind> I see a loophole here. |
|
[16:20] <jonnymind> However, never mind. |
|
[16:20] <jonnymind> I am absolutely uninterested to the contents of that file as long as it complies with your specifications. |
|
[16:20] <jonnymind> 4) This appears to be a new license which may cause delays in NEW processing |
|
[16:21] <jonnymind> Yes, there is a new license. |
|
[16:21] <jonnymind> 5) manpages are typically installed with dh_installman |
|
[16:21] <jonnymind> Typically |
|
[16:21] <persia> That's just a warning, really. The archive admins tend to take longer when the license is not one with which they are familiar. |
|
[16:21] <jonnymind> 6) the binary packages seem odd: no-md5sums-control-file is very rare << how is that created? |
|
[16:22] <persia> jonnymind: Typically with dh_md5sums |
|
[16:23] <jonnymind> this has the same informative content as saying that the copyrhight file is tipically written with emacs or vi |
|
[16:23] <persia> jonnymind: Except that you can read the dh_md5sums source to see what it does. |
|
[16:24] <jonnymind> 9) Don’t run strip on shared libraries: use dh_strip << strip is part of my build process. |
|
[16:24] <persia> jonnymind: Yes, which means we can't generate -dbgsym packages as part of the build, which means we can't effectively trace issues that occur when your program crashes. |
|
=== Lure_ is now known as Lure |
|
[16:26] <ScottK> \sh: Looking |
|
[16:26] <jonnymind> 7) You probably want #DEBHELPER# sections in the maintainer scripts << what does this mean? -- I don't want to use debhelper. |
|
[16:28] <jonnymind> persia: Let me understand. A package debian/ubuntu compliant should not use strip in the build process, and the packaging process should use dh_strip. Correct? |
|
[16:29] <persia> jonnymind: That's not absolutely required, if you don't want to use debhelper, you can alternately provide a symbols package for debugging as -dbg |
|
[16:34] <guest22> persia: Could you please answer some questions on your recent review of photoml (http://revu.tauware.de/details.py?package=photoml)? |
|
[16:35] <jonnymind> persia: from the dh_strip manual it seems I cannot use it at all. I think I will go for adding a dbg binary output. |
|
[16:35] <persia> guest22: Better to ask questions generally of the channel, as I may well have made mistakes. I'll likely still answer some of them. |
|
[16:35] <persia> jonnymind: That would work for me. As long as symbols can be made available for debugging. |
|
[16:37] <guest22> OK, thanks. The first issue you raised was "differentiate hyphen and minus-sign", but I can't find any instances of undifferntiated symbols. Presumably I'm overlooking something - could you please point out specific examples? |
|
[16:37] <Ng> would someone care to look at my latest upload (http://revu.tauware.de/details.py?upid=1151). I've not yet addressed all of the previous comments, but it has switched back to cdbs, so should be a lot more suitable than the previous version ;) |
|
[16:38] <persia> guest22: I got that from lintian. Do you always use \- or \(hy in the manpage? |
|
[16:38] <jonnymind> persia: man dh_md5sums offer the same informative content about making checksums as... man vi offers informative contents about making debian copyright ;-)) |
|
[16:38] <jonnymind> persia: np; |
|
[16:38] <jonnymind> I will scan the net. However, |
|
[16:39] <persia> jonnymind: Read the source. It's a perl script. |
|
[16:39] <jonnymind> Ah, the surce is the manual!!! :-))) |
|
[16:39] <jonnymind> Fantastic. |
|
[16:39] <jonnymind> However, should I use it on the source package or on the binaries? |
|
[16:40] <ChrisGibbs> gday all |
|
[16:40] <jonnymind> i.e. before or after building? |
|
[16:40] <jonnymind> and |
|
[16:40] <guest22> persia: I used to just use '-', but switched to the other forms after complaints about '-' during review of another package. (I don't recall from whom.) |
|
[16:40] <jonnymind> before or after splitting the build result in the directories that will go in the binary packages? |
|
[16:40] <jonnymind> (I suppose this detail is not to be found on the source). |
|
[16:41] <persia> guest22: If you're absolutely sure, ignore that then. |
|
[16:41] <persia> jonnymind: It's typically done as part of the .deb construction, post-install, post-package split, etc. |
|
[16:42] <guest22> persia: That's how the manpages are formatted in two other packages accepted on REVU. |
|
[16:42] <\sh> ScottK: thxalot :) |
|
[16:43] <persia> guest22: That may be. Not all reviewers perform the same tests. I'm rather strict. |
|
[16:44] <guest22> persia: I noticed :) To the best of my knowledge, though, the escaped forms of '-' are preferable. |
|
[16:45] <mok0> sed 's/\w-/\\-/' < manpage |
|
[16:45] <persia> guest22: Differentiation is much preferable, as otherwise one can't reliably copy & paste to the command line and execute things :) |
|
[16:45] <guest22> persia: Clearly I misunderstood what you meant by differentiation. Could you please explain? |
|
[16:48] <Vorian> would someone be so kind as to give me some pointers on this error: http://paste.ubuntu-nl.org/50867/ |
|
[16:48] <jonnymind> persia: I am going to deal with the various suggestion you pointed out and report here when done. About the watch file I suppose it has been a problem with a temporary misconfiguration of the repository on the server. |
|
[16:48] <dfiloni> persia: in a lintian output, W is a warning, E is a error, what is "I"? |
|
[16:49] <mok0> guest22: Anything that is not a hyphen, is a minus |
|
[16:49] <persia> guest22: manpages ideally never use '-', but always either "\-" to indicate a minus sign and "\(hy" to indicate a hyphen. The character used in commands is the minus sign (for arguments, file names, etc.). As nroff formats manpages with '-' as hyphen as default, this can cause issues in a unicode locale. This is made more complicated because some fonts don7t render the characters differently. |
|
[16:49] <persia> dfiloni: Informational (and that's a question best asked of the channel generally). |
|
[16:49] <dfiloni> persia: ok thanks |
|
[16:49] <persia> jonnymind: OK. Thanks. Next revu day is the 7th, so if you've a candidate up then, you'll likely get another review. |
|
[16:50] <jonnymind> Thank you. |
|
[16:50] <ScottK2> \sh: I'm still having dependency problems. |
|
[16:50] <ScottK2> Did it build for you in a Hardy pbuilder? |
|
[16:52] <\sh> ScottK2: yepp...the fontforge problem is normal, but isn't important for it...I hope fontforge will be fixed for next wine release |
|
[16:52] <persia> guest22: Looking over your response comment, I can see the point of NEWS and README, just generally prefer to avoid redundancy where possible to save archive space. |
|
[16:52] <guest22> persia: Now I'm really confused - the manpages in question use "\-" and "\(hy" rather than "-". What, then, is the problem? |
|
[16:53] <ScottK2> \sh: I get fontforge is uninstallable and the build fails. |
|
[16:53] * persia rebuilds and re-runs lintian to find out which manpage has the problem |
|
[16:53] <\sh> ScottK2: hmm...not at my place... |
|
[16:54] <persia> guest22: Problem is in pmldigital.1 |
|
[16:54] <minghua> Fontforge recently went through ungif -> gif transition, if it is relevant... |
|
[16:55] <guest22> persia: Thanks. (For some reason my version of lintian doesn't pick that up.) |
|
[16:55] <persia> guest22: Do you call it with -iIv ? |
|
[16:56] <dfiloni> persia: pbuilder shows me hostname: Unknown host, do you know how I can fix this problem? I can't build |
|
[16:56] <mok0> guest22, persia: there was an error in my sed line about. The following should take care of most problems in manpages: sed 's/\W-/ \\-/' |
|
[16:56] <guest22> No. Perhaps that's the problem. (I found the offending '-' in pmldigital.1.) |
|
[16:56] <persia> dfiloni: I know nearly nothing about pbuilder. I've never used it. |
|
[16:56] <guest22> mok0: Thanks, I'll try that out. |
|
[16:57] <dfiloni> persia: what you use to build? |
|
[16:57] <persia> dfiloni: sbuild on schroot on LVM |
|
[16:57] <dfiloni> ok |
|
[16:57] <ScottK2> \sh: What arch are you building for? |
|
[16:58] <minghua> guest22: What is your "hostname" command's output? |
|
[16:58] * persia suspects fontforge is a victim of libgif/libungif for now |
|
[16:59] <ScottK2> persia: Is that a known general problem? |
|
[16:59] <ScottK2> Because that's where I'm having trouble. |
|
[17:00] <guest22> persia: (Returning to the other issues) I agree with your desire to reduce redundancy, but in this case, I forsee potential complaints about unclear licensing if the README file is omitted, and the NEWS file really does include useful information. Would you consider it acceptable for the next upload to just fix the '-' issue and the README.debian issue? |
|
[17:00] <persia> ScottK: There's a transition underway, but it's only about 80% done in main, and hardy tackled in universe (53 packages outstanding last I heard). |
|
[17:00] <\sh> ScottK2: i386 |
|
[17:00] <ScottK2> persia: From which to which? |
|
[17:00] <\sh> ScottK2: but I can test amd64 as well |
|
[17:00] <ScottK2> \sh: OK. Me too. That's odd then. |
|
[17:00] <guest22> minghua: I think you meant to address that question to dfiloni. |
|
[17:00] <persia> guest22: With your rebuttal, yes, but I'm not likely to be your next reviewer, as I believe the best packages are made by having lots of different viewpoints. |
|
[17:00] <minghua> guest22: Right. sorry. |
|
[17:01] <minghua> dfiloni: What is your "hostname" command's output? |
|
[17:01] <\sh> ScottK2: I set pbuilder-satisfydepends-classic in my pbuilderrc |
|
[17:01] <dfiloni> minghua: root1-laptop |
|
[17:01] <persia> ScottK: Bug #174252 |
|
[17:01] <ubotu> Launchpad bug 174252 in libungif4 "transition to libgif" [Undecided,In progress] https://launchpad.net/bugs/174252 |
|
[17:02] <ScottK2> \sh: I did that too. |
|
[17:02] <ScottK2> persia: Thanks. |
|
[17:02] <minghua> dfiloni: what about "echo $HOSTNAME"? |
|
[17:03] <dfiloni> minghua: root1-laptop |
|
[17:03] <ScottK2> \sh I swapped the build-dep to try libgif-dev before libungif-dev and it worked fine. Given the direction of the current transition, I think that's generally sensible anyway. |
|
[17:03] <minghua> dfiloni: Very strange. How did you run pbuilder? |
|
[17:03] <guest22> persia: One more question about the '-' issue. Since I'm reluctant to roll out a new upstream release for something so trivial, this needs to be fixed in the packaging. The two options seem to be a patch file, or appropriate use of sed in the rules file. The latter is more attractive since the package currently does not have any patches. Would that be acceptable, or is there some policy against doing such things? |
|
[17:04] <\sh> ScottK2: this is really odd |
|
[17:04] <dfiloni> minghua: sudo pbuilder build --basetgz /var/cache/pbuilder/gutsy.tgz --distribution gutsy file.dsc |
|
[17:05] <persia> guest22: No policy against it. Best practice is to make the packaging easily understood by others. Patches are common, as the patches are sent upstream, and then removed when applied. There are a number of cases where sed is used in that manner, for trivial things. |
|
[17:05] <minghua> dfiloni: Try again without "--distribution gutsy". |
|
[17:05] <ScottK2> \sh: Agreed. |
|
[17:05] <ScottK2> \sh: Do you think there's any downside to trying libgif-dev first for dependencies? |
|
[17:05] <\sh> ScottK2: I thought this morning to remove all other libungif stuff and just leave libgif-dev in place |
|
[17:06] <ScottK2> That'd make backporting tougher though. |
|
[17:06] <\sh> ScottK2: yepp...would you like to try 0.9.52 from my ppa on your gutsy ? |
|
[17:06] <dfiloni> minghua: hostname: Unknown host |
|
[17:06] <ScottK2> \sh: I'll probably build it for gutsy in my pbuilder after I get it working for Hardy. |
|
[17:07] <guest22> persia: In other words, you don't have a strong preference for either approach. (Since I'm also the upstream developer, this will be fixed in the next upstream release, independent of the approach for this package.) |
|
[17:07] <minghua> dfiloni: Do you have a .pbuilderrc file? What's in it? |
|
[17:07] <guest22> persia: That should have been "In other words, you don't have a strong preference for either approach?" |
|
[17:07] <dfiloni> minghua: I don't have .pbuilderrrc file |
|
[17:08] <minghua> dfiloni: What does "hostname -f" say? |
|
[17:08] <\sh> ScottK2: hmm my hardy pbuilder just complains about fontforge not going to be installed because of libgif4 and then he's trying libgif-dev as replacement for libungif4-dev which pbuilder can't find |
|
[17:09] <dfiloni> minghua: hostname: Unknown host |
|
[17:09] <\sh> ScottK2: and continues |
|
[17:09] <minghua> dfiloni: Then it's your local problem. See http://lists.alioth.debian.org/pipermail/pbuilder-maint/2006-July/001082.html |
|
[17:11] <ScottK2> \sh: OK. Mine gives up and quits. In any case preferring libgif-dev over libundif-dev solves the problem and seems like generally a better idea. I'm actually into compiling now. |
|
[17:11] <\sh> ScottK2: ok...I'll swap the b-ds for this.. |
|
[17:11] <\sh> ScottK2: there shouldn't be any problems with this.. |
|
[17:12] <mok0> Isn't there a way to specify a dependency for a specific distribution? |
|
[17:12] <\sh> ScottK2: I'll push it then to the ppa for gutsy and for feisty somehow...let's see who complains first ,-) |
|
[17:12] <ScottK2> \sh: I've already done it and so if this works, I'll upload my modification. Just make sure you pick it up for the next time around. |
|
[17:13] <ScottK2> OK. I've also started my gutsy pbuilder running it too. |
|
[17:13] <\sh> ScottK2: cool :) |
|
[17:14] <\sh> ScottK2: I wanted to push the wine debian dir somewhere in a bzr branch.. |
|
[17:20] <AnAnt_> anyone knows french ? |
|
[17:21] <geser> AnAnt_: #ubuntu-fr? |
|
[17:21] <AnAnt_> err, I wasn't going to ask a question in french |
|
[17:22] <AnAnt_> I am trying to find the author of some theme , I found his website www.baqs.net, but it's in french, and I can't figure out his name from there |
|
[17:23] <AnAnt_> the worst thing in making packages is that copyright file, it's a headache ! |
|
[17:24] <Vorian> DktrKranz: yep, I am now |
|
[17:25] <DktrKranz> Vorian, it was related to pysdm upload, I noticed there is bug 79179 which should be addressed, but I'm arrived late :) |
|
[17:25] <ubotu> Launchpad bug 79179 in pysdm "pysdm: doesn't detect partition UUIDs on /etc/fstab." [Undecided,Fix committed] https://launchpad.net/bugs/79179 |
|
[17:26] <Vorian> DktrKranz: I shouldn't have slept in then :) |
|
[17:27] <minghua> mok0: No, and for a good reason, I think. |
|
[17:27] * DktrKranz grumbles at timezones |
|
[17:27] <dfiloni> minghua: http://ubuntuforums.org/showthread.php?t=313576 fix my problem, thanks |
|
[17:27] <Vorian> lol |
|
[17:27] <DktrKranz> Vorian, no problem, next time :) |
|
[17:27] <Vorian> you got it DktrKranz :) |
|
[17:28] <mok0> minghua: ? |
|
[17:28] <minghua> dfiloni: No problem. |
|
[17:28] <mok0> minghua: What reason is that? |
|
[17:28] <minghua> mok0: "Isn't there a way to specify a dependency for a specific distribution?" |
|
[17:29] <mok0> minghua: I just realized I asked thar question some time ago :-) |
|
[17:30] <minghua> mok0: Package should be differentiated by versions, not by releases, IMHO. |
|
[17:30] <ScottK2> \sh: My version isn't going to backport cleanly anyway since fontforge depends on libungif in Gutsy. |
|
[17:30] <minghua> mok0: So how are you going to have different dependencies for different distributions? |
|
[17:31] <mok0> minghua: hmmm |
|
[17:31] <minghua> mok0: Or, by "distributions", you mean "Debian or Ubuntu"? |
|
[17:31] <mok0> minghua: actually, I mean gutsy, hardy... |
|
[17:31] <mok0> minghua: but it would be nice to be able to do backports transparently |
|
[17:32] <\sh> ScottK2: hmm...when we leave the b-d with libungif4-dev | libgif-dev it should be ok...when fontforge comes back with correct deps |
|
[17:33] <minghua> mok0: I would rather have different source packages if different explicit dependencies (i.e., not from dpkg-shlibdeps and the like) are needed. |
|
[17:34] <mok0> minghua: It works, but it's more difficult to maintain |
|
[17:34] <\sh> ScottK2: I uploaded yesterday to gutsy ppa and it build cleanly...if you want to test :) |
|
[17:34] <ScottK2> \sh: Isn't the current dep on libgif in Hardy the correct one already? |
|
[17:35] <\sh> ScottK2: nope...fontforge is in depwait mode as mentioned yesterday on libfoobarwhatever |
|
[17:37] <minghua> mok0: Why? Any decent VCS should handle that easily. |
|
[17:37] <\sh> ScottK2: anyways...wine 0.9.53 is just waiting to be released.. |
|
=== cassidy_ is now known as cassidy |
|
[17:41] <ScottK2> \sh: It's not in debwait now. |
|
[17:41] <ScottK2> deb/dep |
|
[17:41] <\sh> ScottK2: then the message is strange |
|
[17:42] <ScottK2> https://launchpad.net/ubuntu/hardy/+source/fontforge/0.0.20071110-1build2 |
|
[17:43] <ScottK2> One option would be to remove the build dep on libgif-dev or libungif-dev entirely and depend on fontforge to pull the right one in. |
|
[17:44] <\sh> ScottK2: nope...libgif is a build-dep needed by wine explicitly ... so all explicit b-ds should be mentioned... |
|
[17:44] <ScottK2> I agree it's crackish, but I think it would work. |
|
[17:44] <\sh> ScottK2: if it's not going to be backported cleanly, we can point the users to winehq gutsy packages |
|
[17:44] <ScottK2> True |
|
[17:46] <\sh> ScottK2: sure it works..but think of the last xorg transition ... all packages which had direct source build dependencies were screwed because of the package transition |
|
[17:46] <ScottK2> Right. It'd have risks. The question is is there sufficient benifit to getting it into gutsy-backports to take the risk. |
|
[17:47] <\sh> ScottK2: actually when we got the message about fontforge being broken..this should be addressed |
|
[17:47] <\sh> ScottK2: nope...we have a alternative (winehq) the packages are ok for the wine users of gutsy for the latest crack |
|
[17:48] <ScottK2> OK. |
|
[17:49] <ScottK2> Putting the build-dep for fontforge ahead of libgif-dev | libungif-dev wouldn't help any, would it? |
|
[17:54] <mok0> minghua: you are right |
|
[17:56] <mok0> minghua: I want to start doing that |
|
[17:57] <ScottK2> \sh: I'm uploading it now. It'd be nice to have a generic solution for 0.9.53. Maybe you and Scott can work on that in the meantime. |
|
[17:58] <\sh> ScottK2: will try :) |
|
[17:58] <\sh> ScottK2: thx :) |
|
[17:59] <mok0> minghua: but the tools will complain over the .git directory I think |
|
=== czessi_ is now known as Czessi |
|
[18:05] <guest22> I've just uploaded a new version of photoml (http://revu.tauware.de/details.py?package=photoml) after discussion with persia. Any MOTUs here willing to take a look? (The previous upload had one advocate, so it should be very close to being acceptable.) |
|
[18:07] <minghua> mok0: Use dpkg-buildpackage -i or equivalent. :-) |
|
[18:08] <mok0> minghua: thx will check it out |
|
=== cassidy_ is now known as cassidy |
|
[18:09] <mok0> minghua: actually, there is a git-buildpackage I realize... |
|
[18:10] <zul> afternoon |
|
[18:11] <minghua> mok0: There is also gitpkg. |
|
[18:16] <Vorian> Could I have some pointers on these errors: http://paste.ubuntu-nl.org/50867/ |
|
[18:18] <ScottK> Vorian: Do you build-dep on something that provides scrollkeeper? |
|
[18:19] <Vorian> i get the same error with or without libscrollkeeper-dev as a depends |
|
[18:19] <jpatrick> hmmm, attack of the clones... |
|
=== neversfelde_ is now known as neversfelde |
|
[18:22] <tuxmaniac> Any pointer on these warnings I get. I think the new line things are pretty trivial. |
|
[18:22] <tuxmaniac> http://paste.ubuntu-nl.org/50877/ |
|
[18:27] <stgraber> just a quick packaging question, when you have a binary file which is only a X program and doesn't take any parameter, do we need the manpage ? |
|
[18:27] <jpatrick> stgraber: all binaries must have a manpage.. (so says policy) |
|
[18:28] <geser> Vorian: build-depend on scrollkeeper or rarian-compat |
|
[18:28] <Vorian> scrollkeeper |
|
[18:28] <geser> both ship scrollkeeper-config |
|
[18:29] <Vorian> geser: I'll give it a shot |
|
[18:30] <stgraber> jpatrick: the policy actually says "should have a manpage", that's the reason of my question :) |
|
[18:35] <nictuku> hi. is there any problem splitting a python software in multiple deb packages even if all python modules are placed under a single python-package structure? |
|
[18:42] <Vorian> me hugs geser |
|
[18:42] <Vorian> that did the trick |
|
[18:42] <Vorian> :) |
|
[18:51] <tuxmaniac> http://paste.ubuntu-nl.org/50877/ Any pointers ? I am unable to figure out the reason |
|
[18:51] <tuxmaniac> atleast for the binary package changed error |
|
[18:53] <jpatrick> tuxmaniac: are you building the source package? |
|
[18:54] <tuxmaniac> jpatrick, yes |
|
[18:55] <jpatrick> tuxmaniac: make sure the .tar.gz is package_version.orig.tar.gz |
|
[18:56] <tuxmaniac> yes I have made sure that, jpatrick , the name is alliance_5.0-20070718.orig.tar.gz |
|
[18:58] <jpatrick> tuxmaniac: and in changelog it's? |
|
[18:58] <tuxmaniac> package name is alliance in changelog. |
|
[18:59] <tuxmaniac> version I have given as 5.0-20070718-0ubuntu1 |
|
[18:59] <jpatrick> no the version :) |
|
[19:01] <tuxmaniac> jpatrick, in controls I have given the source name as alliance and package as alliance-5.0-20070718. right? |
|
[19:01] <jpatrick> tuxmaniac: odd, try 5.0~svn20070718-0ubuntu1 in changelog |
|
[19:02] <jpatrick> (if it uses svn) |
|
[19:03] <tuxmaniac> jpatrick, if I remove the folder named desktop from the debian directory, then it goes ahead in building the package but fails while icons |
|
[19:03] <tuxmaniac> creating icons I mean |
|
[19:04] <jpatrick> tuxmaniac: oh, have you set rules to clean the created icons? |
|
[19:04] <tuxmaniac> jpatrick, yeah seems like. Just saw it |
|
[19:07] <jpatrick> tuxmaniac: works now? :) |
|
[19:08] <tuxmaniac> jpatrick, there seems to be no clean up of icons in the clean rule. Adding one now. |
|
[19:15] <tuxmaniac> jpatrick, still doesnt work. I am not sure whether I did the right thing. I guess I will remove the Desktop folder from the debian directory. Comment all related rules and build once and re add things after that |
|
=== \sh is now known as \sh_away |
|
[19:16] <jpatrick> tuxmaniac: why is there a Desktop folder in debian/ ? |
|
[19:17] <tuxmaniac> that was my next question? Can there be one. The original packager seems to install icons from this folder. |
|
[19:17] <tuxmaniac> jpatrick, :-) |
|
[19:17] <jpatrick> very odd |
|
[19:17] <tuxmaniac> jpatrick, there is also a man folder which makes me feel even odd. |
|
[19:18] <jpatrick> tuxmaniac: are there binaries in there? |
|
[19:18] <tuxmaniac> jpatrick, cleaning things up. Its already on revu but fails to build on Hardy pbuilder. Now after a monor src change it builds on hardy. After which I started cleaning stuff to comply with Ubuntu/Debian Packaging guide and things didnt look great after that |
|
[19:19] <tuxmaniac> jpatrick, there are a few man pages, icons and a few shell scripts in there |
|
[19:20] <jpatrick> hmm |
|
[19:20] <tuxmaniac> jpatrick, i want to get this "NEW" package badly into Hardy. Will help a lot of Electronic Students and enable more contribution if available in Ubuntu |
|
[19:21] <jpatrick> tuxmaniac: I'd say, mv debian ../ and redo the tarball |
|
[19:22] <jpatrick> tuxmaniac: see then put debian back in at see if it works |
|
[19:22] <tuxmaniac> jpatrick, hmm ok. I think thats the clean way out. |
|
[19:28] <tuxmaniac> bddebian, hola! |
|
[19:28] <geser> nixternal: Hi, you added libmotif-dev (multiverse) to build-depends for pdfedit (universe). That doesn't work, can it be build with lesstif? |
|
[19:28] <bddebian> Heya gang |
|
[19:28] <bddebian> Hi tuxmaniac |
|
[19:28] <geser> Hi bddebian |
|
[19:28] <bddebian> Hi geser |
|
[19:47] <blueyed> Hi bddebian et al. |
|
[19:47] <bddebian> Hello blue |
|
[19:47] <bddebian> Err blueyed |
|
[19:48] <blueyed> Is there a way to cache the build result when using (p)debuild? - I mean, pdebuild can take a while and if you are testing just if a patch builds cleanly, this is quite a lot of time.. or should I use "debian/rules binary" then instead? |
|
[19:49] <blueyed> bddebian: tab-enter? :) |
|
[19:50] * jonnymind is away: dinner |
|
[19:50] <bddebian> blueyed: Aye :-) |
|
[19:50] <nixternal> geser: sure, build with what works of course :) |
|
[19:51] <bddebian> blueyed: I don't know about pdebuild but pbuilder takes a --logfile option |
|
[19:51] <nixternal> weird that it doesn't work when pdfedit actually deps on it according to their docs |
|
[19:51] <bddebian> Or is that not what you meant? |
|
[19:52] <geser> nixternal: pdfedit is now in depwait on libmotif-dev |
|
[19:52] <blueyed> bddebian: pdebuild is pbuilder (and similar to debuild in that regard). I'd just like to make the build as fast as possible (e.g. no "clean"). |
|
[19:52] <blueyed> bddebian: yes, logfile is not what I'm after.. :) |
|
[19:53] <geser> nixternal: we have 3 options: move pdfedit to multiverse :(, check if building with lesstif2-dev works, drop libmotif-dev from b-d again |
|
[19:54] <nixternal> ahh, I don't want to do that honestly |
|
[19:54] <dmb> nixternal: have it use lesstiff? |
|
[19:54] <nixternal> use whatever will keep it in universe |
|
[19:55] <dmb> nixternal: well, do you know if it builds with lesstiff? |
|
[19:55] <nixternal> haven't tried it, the reason I went with the libmotif-dev was because the developer said to use it |
|
[19:58] <superm1> jpatrick, you present still? |
|
[19:58] <dmb> nixternal: it looks like pdfedit uses qt3, boost and xlib |
|
[19:58] <dmb> are you sure it requires motif? |
|
[20:01] <nixternal> it will build w/o it, that I do know, but there were issues trying to build something w/o it |
|
[20:03] <nixternal> I would drop libmotif-dev and go back to the way it was done previously |
|
[20:04] <dmb> nixternal: i'll test to see if it would build with lesstif, just because i'm curious |
|
[20:05] <nixternal> I am sure it will...if it doesn't need it, then it will ignore it |
|
[20:05] <dmb> nixternal: version 0.3.2? |
|
[20:05] <nixternal> whatever the latest is |
|
[20:06] <dmb> yes |
|
[20:07] <slangasek> imbrandon: "community marketing" likes breaking continuity of wiki history, and wants people to think that only pages under "help" are helpful? :) |
|
[20:19] <dmb> nixternal: is there a way i can check if the motif component was built? |
|
[20:20] <nixternal> dmb: I just went through that entire package and couldn't find any reference to *motif* at all |
|
[20:21] <nixternal> not package, but the source code |
|
[20:21] <dmb> oh |
|
=== wolfger_ is now known as wolfger |
|
[21:00] <jpatrick> superm1: I am now |
|
[21:00] <superm1> jpatrick, ah okay. would you have a moment to look over a bug? |
|
[21:01] <superm1> i'm going to be gone the next week and a half, so i wanted to get a backporter to look at it for troubles before i left |
|
[21:01] <superm1> https://bugs.edge.launchpad.net/feisty-backports/+bug/180633 |
|
[21:01] <ubotu> Launchpad bug 180633 in feisty-backports "Please backport xmltv 0.5.49-1 to Feisty (from gutsy)" [Undecided,New] |
|
[21:02] * jpatrick looks |
|
[21:03] <jpatrick> superm1: and none of the rdeps of libxml-twig-perl are affected? |
|
[21:04] <superm1> well i wouldn't imagine so, but let me take a look at its changelog |
|
[21:04] <superm1> between revisions |
|
[21:06] <superm1> well they are all bug fixes except for one possible compatibility change |
|
[21:07] <superm1> i'm not sure what the easiest way to query the other apps using that change however. hmm |
|
[21:21] <jussi01> OK, its been a while since I have done stuff, can someone remind me of the command to extract source from a .dsc? |
|
[21:23] <man-di> jussi01: dpkg-source -x ....dsc |
|
[21:23] <jussi01> man-di: thanks :) |
|
[21:23] <ScottK> jussi01: or dget -x ....dsc if it's elsewhere |
|
[21:23] <superm1> jpatrick, yeah i guess it will be a bit difficult to evaluate those rdepends considering the compatibility change |
|
[21:37] <guest22> Since I see there's at least one MOTU hanging around, this seems like a good time to make another request for a review of photoml (http://revu.tauware.de/details.py?package=photoml) |
|
=== davro is now known as davromaniak |
|
[21:47] <LaserJock> anybody know if it's possible to turn off RSA host key checking in ssh for a particular ip/host? |
|
[21:51] <ion_> Why do you want to do that? |
|
[21:51] <LaserJock> because |
|
[21:51] <LaserJock> I'm dual booting |
|
[21:52] <LaserJock> and ssh keeps complaining about the key |
|
[21:52] <ion_> Why not install the same host key to both installations? |
|
[21:52] <LaserJock> hmm, that would make entirely too much sense ;-) |
|
=== jpatrick_ is now known as jpatrick |
|
[22:24] <jussi01> TheMuso: whats my next move with that debdiff ? |
|
[22:26] <TheMuso> jussi01: Well once/if it has been tested and known to be working, I'd file a bug, and attach the diff. Then subscribe ubuntu universe sponsors. |
|
[22:27] <jussi01> TheMuso: ok. the bug is filed. Ill attach the diff once i have a tested working copy :) |
|
[22:27] <TheMuso> ok |
|
[22:28] <TheMuso> It often makes sense to only file the bug once you have a debdiff you are happy with, but to each their own. |
|
[22:29] <DarkSun88> Hi all |
|
[22:29] <DarkSun88> persia: Ping |
|
=== nuu is now known as nu[year] |
|
=== nu[year] is now known as nuu |
|
[23:44] <imbrandon> slangasek: unfortunately, yes |
|
[23:44] <imbrandon> heh |
|
|