[00:00] RainCT: well, things happen *g*... maybe I now reported a bug twice, in case exim was smart enough to still retry the original message (the log is a little bit vague, saying completed, which I guess is valid for the bounce only though) [00:01] gouki: I usually use debsign -k... as for me DEBSIGN_KEYID never seemed to work *shrug* [00:02] well, good night [00:02] gn8 RainCT [00:03] sistpoty, indeed! That worked like a charm. Weird, as DEBSIGN_KEYID doesn't work. Any ideas why? [00:05] gouki: not really... maybe a bug in debsign? [00:07] gouki: from a glimpse, maybe also setting DEBSIGN_MAINT ? [00:08] Random Quote: One of the most striking differences between a cat and a lie is that a cat has [00:09] sistpoty, hmm, didn't work either :S [00:12] gouki: hm... no idea then, even with staring at debsign's source for some time now [00:22] Hmmm ... That's weird. Are those variables saved and persistent after reboots? It keeps forgetting the name and email [00:23] did you write them to your .profile or .bash_profile or some other similar file? [00:41] cyphermox, I thought they were written when using export. [00:42] ah, no. you need to add those to the profile file used by you're shell. maybe there's some other way elsewhere, but i'm not aware of it [00:46] how do i figure out who the maintainer(s) for a particular package is [00:48] CTho: "apt-cache showsrc $pkg | grep Maintainer" [00:49] most packages in Ubuntu don't have maintainers, only some packages in main have maintainers (but any coredev can still upload) [00:52] cool [00:56] hey stgraber :) [00:57] hi pochu [00:57] how's it going ? [01:00] stgraber: pretty well! you? [01:00] * pochu just applied to NM [01:00] yeah, getting used to living in Canada :) [01:00] pochu: coming to UDS ? [01:00] stgraber: not this time :( [01:00] stgraber: oh, are you in Canada? [01:01] yep, moved there 3 months ago [01:01] argh, too bad we'll never meet if you continue to go to UDS only when I don't :) [01:03] I'm now working on LTSP and educational softwares at full time and I'm paid for that :) [01:04] I'll probably apply for MOTU membership soon though as it seems to be a requirement to get limited upload rights to main now ... (I maintain iTalc/ltsp/ltspfs/ldm and soon ltsp-cluster there but need to get everything sponsored ...) [01:05] not that I don't like to speak to my sponsors :) but it'd make things a lot faster to directly upload that to Ubuntu. [01:05] stgraber: that's cool! yeah, hopefully next UDS... [01:06] pochu: yeah, now that I'm working on Ubuntu full-time I should be able to come to every UDS either sponsored by Canonical or by my company (if Canonical sponsors me we can then send someone else) [01:07] so now all the pressure is on my side :) [01:07] yeah :) [01:08] * pochu thinks of a cool proposal for the Kind Koala's UDS :D [01:10] topics aren't the problem at UDS, time usually is :) [01:11] I hope next UDS is not in June ;) [01:11] * pochu is off to bed [01:11] pochu: why ? exams ? [01:12] yeah [01:12] yeah, that'd be a problem :) [01:12] pochu: good night [01:12] good night stgraber! see you [01:14] GREAT SUCCESS! Intrepid boots to ext4 :D [01:15] jdong: cool [01:15] yeah, pretty neat :) [01:16] jdong: can we install on ext4 or do you need to install on ext3, then debootstrap on ext4 ? [01:16] stgraber: I just tune2fs -E test_fs -O extents my / [01:16] stgraber: but BE CAREFUL: (1) you need to add ext4dev to /etc/initramfs-tools/modules (2) you need a separate ext2/ext3 /boot because grub can't read ext4 [01:17] stgraber: right now I'm just going to mangle all my files in-place to take advantage of extents (i.e. rsync usr usr.new, mv usr.new usr) [01:18] ok, that'll take a while :) (Copying all your files) [01:18] stgraber: yeah, maybe 15 minutes of mangling around; this is a fairly new install [01:19] stgraber: I'm not sure how much you'd have to hack ubiquity to make it install to ext4dev, probably it'll be a simple hack [01:19] stgraber: though you will have to manually make sure ext4dev.ko is in initramfs or the next boot will be interesting [01:20] yeah :) and force it to create a /boot (as it does with xfs/jfs) [01:20] correct [01:22] uh oh [01:22] I think I found a FTBFS [01:23] can someone verify or should I just file? [01:27] hello? Did everyone just run away as soon as I mentioned FTBFS? [01:27] I'm just being curious, but what fails? [01:28] geordi [01:28] missing a haskell library in the build-dep I think === csilk_ is now known as csilk [01:49] coppro: hm? iirc geordi built last fine a few weeks ago? could you pastebin the error message eventually? [01:50] sure [01:55] sistpoty: http://codepad.org/HOYP77NO [01:56] coppro: did you try that in pbuilder? [01:56] sistpoty: I can't get pbuilder to do it; how do I make the dsc for it to use [01:57] coppro: sudo pbuilder build [01:57] which dsc? [01:57] coppro: well, if you take a pacakge from the archive, you've already got a dsc [01:58] coppro: otherwise you could build a source package first (dpkg-buildpackage -rfakeroot -S -us -uc) [01:58] oh there we go [01:58] coppro: or just use pdebuild in the top source dir [01:58] I forgot the build [01:58] coppro: however in this case, are you sure you have all the build-dependencies of geordi installed? [01:58] yes [01:59] pbuilder fails at satisfying them though [02:00] that would be bad [02:00] probably [02:02] * sistpoty test-builds geordi [02:11] jdong: I was planning to reinstall my laptop later today... I think it might be hack-Ubiquity-on-the-live-CD time. [02:11] wgrant: it'll be fun! :) [02:11] FWIW an hour later nothing exploded yet [02:11] Not surprising. [02:12] yeah, I wouldn't expect ext4 to be all that risky [02:12] performance improvements are modest and incremental [02:12] I'd have to tweak /etc/modules and update-initramfs, wouldn't I? [02:12] for ordinary workloads, little difference, but for huge files, it helps significantly [02:12] yeah you need to tweak /etc/initramfs-tools/modules [02:12] Ah, that's the one. [02:12] and of course, separate /boot [02:13] Yep. [02:13] how do I get ext4? I already have a separate /boot, so that's not an issue [02:13] * wgrant hits clueless forum users. [02:14] wgrant: That's redundant. [02:14] They're now complaining that you can delete root-owned files. [02:14] Withou being root. [02:14] If you own the parent. [02:15] bug, clearly [02:16] ubuntuforums actually looks a bit better without CSS. [02:17] you mean everything looks better when it isn't in a crap brown and orange theme? [02:17] [02:17] * ScottK likes L better that way. [02:17] jdong: I'd agree with that. [02:17] * ScottK likes blue. [02:17] :) [02:17] Native widgets help. [02:17] coppro: well depends on how daring you are [02:18] coppro: to convert the filesystem, simple tune2fs -E test_fs -O extents /dev/root_block_dev [02:18] is ext4 installed in the kernel now? [02:18] coppro: then make sure to add ext4 to initramfs as mentioned [02:18] coppro: yup [02:18] err s/ext4/ext4dev/ [02:18] coppro: then edit fstab change ext3 -> ext4dev [02:18] then reboot [02:18] * coppro blindly follows jdong's instructions [02:19] afterwards for large files you may have, it helps to make a copy of them and mv the copy over the original [02:19] extents are only used for newly created files [02:19] jdong: You saw my comment about a bzr backport? [02:19] ScottK: yeah [02:19] can ext3 still mount the ext4 ones if it doesn't work? [02:19] coppro: *NO* [02:19] coppro: once you enable extents there's no going back! [02:19] coppro: Make a backup of anything you actually care about. [02:19] even if they aren't actually used? [02:19] coppro: it may be prudent to -E test_fs *WITHOUT* -O extents first [02:19] coppro: then it's backwards compatible [02:20] ok [02:20] coppro: Make a backup of anything you actually care about. [02:20] coppro: until you're sure you can mount ext4 properly [02:20] coppro: then add extents and do the copy-move mangling [02:20] coppro: oh yeah ScottK is absolutely right :). [02:20] ScottK: yeah, all the stuff I care about is in source code repos :) [02:21] actually, I know [02:21] OK. If you're going to play with new file systems, consider the data totally expendable. [02:21] OK. [02:21] I'll try it on just /usr for now, and see how that goes [02:21] * wgrant stab stab stab. [02:21] 99% of that is all package files, so it's recoverable [02:21] Don't fricking file bugs after a dev tells you you are *wrong*. [02:22] wgrant: Unless it's Launchpad and a Launchpad Dev. [02:23] Haha. [02:23] True, true. [02:23] okay, so I've enabled ext4 but no extents on /usr... time to reboot and hope I don't end up having to dig out a boot disk [02:23] * ScottK should probably point out to cprov that you said that. [02:24] coppro: oh you'll be fine. [02:24] kk good [02:24] and when it breaks I blame it on jdong [02:24] ScottK: The bugs I file are normally fairly sane, I suspect. Sometimes they object violently but then a few months later the politics sort themselves out and they become high priority... [02:24] :P [02:25] wgrant: but because you complain, they don't get listened to. duh. [02:26] Hobbsee: kiko has a good point, but I generally complain more relevantly now. [02:26] wgrant: doesn't seem to matter. [02:26] * Hobbsee shrugs [02:42] hm.. geordi builds fine as expected, but now coppro is gone [02:42] oh god, he's still not back? [02:42] crap.... [02:42] * jdong runs away and hides [02:42] heh [02:43] Haha. [02:43] * Hobbsee throws rotten tomatoes at jdong [02:44] * wgrant throws in some broken volumes. [02:44] hey, these are still edable Hobbsee :P [02:44] sistpoty: feel free to eat them... [02:44] *yum* === nellery_ is now known as nellery [02:44] if you really wish :P [02:44] erm... maybe I was wrong *g* [02:45] they're all squishy and rotten, so... [02:56] * ScottK figures he got what he deserved for listening to a known crackhead. [02:56] haha [02:57] I see it's business as usual in here. [02:59] hello, i have a question, how do i change display resolutions in 8.10? it seems the X11 doesnt care about xorg.conf and the "graphical" program only gives me 800x600 (i can get way more). Xrand doesnt seem to work either. I was used to get 1024x768 with VESA driver, now i can't get more than 800x600 (in 8.10). [03:00] This isn't a support channel. Try #ubuntu [03:00] jfc [03:00] people in #ubuntu tells me to come here [03:00] well, they're wrong. [03:00] and people here tells me to go to #ubuntu! [03:00] this still isn't a support channel. [03:00] i notice it. [03:00] Go to #ubuntu+1 then [03:00] orly_owl: is closed. [03:00] KOSHrf: might be a bug, though output of your xorg.log might shade some light [03:01] Ah. fair enough. [03:01] s/shade/shed/ [03:02] sistpoty, well it is pretty empty... i only put the modes in it, thats pretty much all. [03:02] xorg.log doesnt says anything. [03:02] /var/log/Xorg.0.log? [03:02] KOSHrf: /var/log/Xorg.0.log is completely empty? [03:02] * ScottK hands KOSHrf https://wiki.ubuntu.com/X/Troubleshooting/Resolution [03:03] It's not your exact problem there, but I suspect the solution will work. [03:03] by anything i meant "anything wrong" :) [03:03] ScottK, let me try that, thank you. [03:03] * ScottK then mutters about #ubuntu-x [03:05] ScottK, nop, it doesnt do anything, xrandr says there is only "800x600" and "640x480" available. [03:06] Then I'd ask in #ubuntu-x, but it's generally VERY quiet on the weekends. [03:06] hehehe, okie, thanks tho [03:06] * jdong points towards the last part of the page [03:06] * wgrant points people towards unbuggy hardware. [03:07] wgrant: That page exists because of a monitor that worked fine in Hardy and Intrepid falls over on. [04:20] jdong: I couldn't hack the extra FS into partman+ubiquity properly, so I just s/ext3/ext4dev/ in the mkfs and mount commands... it seems to be working. [04:20] wgrant: wait, what's the proper way? [04:20] :) [04:28] It seems to now be happily installing with extents enabled. [04:31] wgrant: awesomeness [04:31] I have to keep a swapfile and extents are just wonderful for that [04:39] Note to self: The changelog entry isn't enough. You have to actually make the change for it to count. [04:40] ScottK: lol there needs to be a lintian added-dpatch-but-series-unchanged alarm :) [04:40] I've caught myself doing that several times when inspecting a debdiff pre-upload [04:41] Yeah. I had one of those last week. I even uploaded it. Fortunately it was an SRU, so I could just get it rejected. [04:43] One of my main uploads towards the end of Intrepid had one of the patches removed by the sponsor... but he didn't add either of them to series, and nobody noticed for a couple of days. [04:43] It's too easy to do. [04:44] yeah, indeed [04:48] hmm I'm going to give the Kubuntu CD a try. [04:48] I admittedly haven't touched KDE for almost a year now and am curious [04:52] jdong: IMO KDE4 makes Gnome look a bit tired and old fashioned. [04:53] ScottK: I've heard enough good things that I'm at least curious [05:09] ScottK: you know if anything will blow up if I install kubuntu-desktop on Ubuntu? :) [05:10] the livecd was definitely intriguing [05:14] jdong: I don't know because I haven't done it, but it should be fine (assuming you mean install the metapacakge from the archives and not some silly use the CD thing) [05:14] I tried the 8.10 livecd and had some issues but overall it blows the gnome look out of the water. I just don't like some of the functionality. I couldn't get the twitter widget to work without it screwing up pretty bad. [05:14] jdong, In 8.04 I installed it and it replaced my Ubuntu screen with Kubuntu and it threw all of the kubuntu stuff into my gnome menus. It was pretty annoying. [05:15] Well KDE4 is young and so it's definitely got some rough edges. [05:15] yep [05:15] It looks great though. [05:15] well it's going to add kubuntu stuff all over the menu for sure [05:15] and a splash screen isn't going to kill me [05:16] and the autoremove thing should give me a quick uninstall path if I don't like it :) [05:16] jdong, Yeah I rarely rebooted and when I did that was the only time the other IT guys hassled me for having Kubuntu on it :P [05:16] (at work anyhow) [05:16] rrittenhouse: update-alternatives on the usplash-artwork.so file should do the trick [05:16] ah [05:16] I went to Ibex there today anyhow =/ [05:17] I have to go back in tomorrow and keep working through all of my issues and get a running system again. [05:17] might just be me, but ext4 seems to unpack these debs a lot faster [05:17] hmm [05:17] Does Ibex do ext? [05:17] er [05:17] ext4* by default [05:17] No [05:17] no, it's experimental in 2.6.27 [05:18] you've got to be a daring idiotic fool like me to try to hack it working :) [05:18] Oh ok. I swore it was Ext3. Ok [05:18] i see [05:18] haha :-D [05:18] I normally am. I'm actually trying to involve myself into the project more to see what I can (hopefully not) screw up :P [05:18] (e2fsprogs in Intrepid along with the kernel support ext4 but neither advertises its abilities by default) [05:19] interesting. [05:20] it's been marked stable in 2.6.28 (i.e. recommended for adoption) but AFAICT there's no significant code changes to get to that marking [05:20] but the improvements for the average user IMO are not worth it to make your system experimental and unsupported :) [05:20] yeah [05:20] I should just VM that then? :P [05:21] that is if VMWare wants to work with the newer kernels. I was having a horrid time with vmware server 2 in 8.04 with the last 2-3 kernels. [05:24] testing from KDE4 :) [05:24] :) [05:24] I [05:24] I'm just not sure if I could convert over to KDE.. I've tried. :P [05:25] jdong: Feel free to stop in on #kubuntu-devel if you want to discuss the experience. [05:25] ScottK: absolutely if I find something productive to say, I'll do so :) [05:25] jdong: Also we'll have KDE 4.1.3 hitting intrepid-proposed next week, so the bug fix goodness level should go up. [05:26] ScottK: fantastic [05:26] Good night. Hope you enjoy it. [05:26] thanks :) [07:57] what files should I submit to the ubuntu source if I wants to maintain a deb package? [08:33] Hi Aron. You can find more info here: https://wiki.ubuntu.com/MOTU/Contributing. Is it for a new package or an existing one? [08:36] Aron_: anyway, all begins with a bug in launchpad [08:36] ... [08:37] I am just reading the page you gave me [09:13] Good morning. [09:21] morning [09:22] it there any special SRU versioning scheme? [09:26] didrocks: you can use the same as for security updates [09:27] RainCT: that was what I reckoned! Thanks a lot ;) === RainCT_ is now known as RainCT [10:54] * RainCT released https://wiki.ubuntu.com/MOTU/Packages/REVU/CheckList into the wild for people to improve it (this includes coming up with a better name for the page :P) [11:04] RainCT, what about asking archive-admins for common reject cases (e.g. for copyright issues)? It could be an added value to that page [11:08] * Package is empty: reject [11:10] * Package is empty, or only contains changelog and copyright file for a non-meta package: reject [11:11] * Package matches /matix$/: reject [11:12] wgrant: ^5 [11:13] wgrant: uhm? [11:13] wgrant: +1 [11:13] RainCT: automatix [11:13] :) [11:13] or ultamatix [11:14] Hobbsee: so, empty packages are common? o.O [11:15] I would also like to propose automatic rejection of NEW packages with 'ubuntu' in the version. [11:15] i once had an empty package approved. i forgot to bzr add stuff :/ [11:15] RainCT: Um, sometimes. I've seen a few [11:15] i don't really do much of it [11:15] wgrant: what about ubuntu-specific packages? [11:15] wgrant: why? [11:15] RainCT: re: src and bin package names, should they be exactly the same? [11:15] karooga: no, they are fine [11:16] karooga: No they shouldn't. They should be *sane* [11:16] karooga: the name of the binary package should follow the Python policy [11:16] karooga: that one of the source package name can be the same as upstream gave it [11:16] laga: That is my point. We don't want them. [11:16] Hobbsee: We don't want to maintain many hundreds of packages on our own. [11:17] wgrant: no, i mean if they're only going to work on ubuntu. [11:17] stuff like ubuntu studio meta packages etc [11:17] Ah. [11:18] wgrant: well, i'd say a more effective way of doing that woudl be to close REVU, adn give people other things to work on [11:18] But that would make sense... [11:18] wgrant: they only end up in the new queue after being reviewed and uploaded, so that seems slightly silly. [11:18] RainCT: great. When I did the packaging, I did it for hardy and intrepid - I presume I should still do it for these too, in addition to jaunty? [11:18] The Ubuntu Studio meta packages I daresay didn't hit REVU, and they don't contain ubuntu in their version, since they're native [11:18] wgrant: well, that's true. It would actually make sense. But I think people are *finally* starting to see exactly why it doens't, and that they needto do something about it. I live in hope, anyway! [11:19] Hobbsee: But we're the final line of defense! [11:19] StevenK: oh? [11:19] StevenK: why? [11:19] Hobbsee: -archive is [11:19] karooga: Jaunty is enough but yes, if you aim to get them backported later then ensuring that they work on previous releases is a good thing :) [11:20] StevenK: we were not talking about REVU, we were talking about possible reasons why an archive admin would reject a package :) [11:20] StevenK: that's true. But i'm not seeing the relevance, as such :) [11:20] Hobbsee: Off the cuff comment [11:20] StevenK: i was under the impression that non-native packages with an upstream tarball are preferred. like the mythbuntu stuff is handled [11:20] StevenK: ah, right. [11:21] laga: The seeds and meta packages ought to be native [11:21] yeah, they are. other stuff isn't. like default settings packages [11:22] Seeds are special [11:22] default settings and such, I can go either way [11:22] -m{id,obile}-default-settings are native [11:24] laga: actually, there's a really good debian link about this. [11:26] laga: http://ftp-master.debian.org/REJECT-FAQ.html iirc [11:27] then let's link that on the REVU page [11:27] not all are relevant, though [11:29] fails to build from source in a sane way <- got to love that one [11:29] yup [11:33] laga: what might be useful is if you document which of those are relevant for that page, and link it. [11:35] i have no clue which ones are relevant and i dont have my wiki login handy ;) [11:36] laga: oh, sorry, it was RainCT who was adding them, not you. [11:36] * Hobbsee got the purple people mixed up [11:36] RainCT: ^ [11:37] purple? [11:37] too much fluor in your drinking water? [11:37] purple people ? but he's geen ! [11:37] heh [11:37] *green even [11:37] you show as purple in my client. as does ogra. [11:38] can i get another color? [11:38] you should really rework your color scheme, you are yellow btw ... [11:38] heh === calc_ is now known as calc [11:44] All of you are black [11:44] i don't have a nick list [11:44] Until you say my name, and then for that line you're red [11:44] All of you are white. [11:44] Unless you're yellow. [11:45] wgrant: irssi? [11:46] laga: Affirmative. [11:46] \o/ [11:50] wgrant: good choice \o/ [11:59] Hi all [12:00] Can somebody help with a packagin issue, pls? [12:00] hi marrow [12:00] I am stuck [12:00] !ask [12:00] Please don't ask to ask a question, simply ask the question (all on ONE line, so others can read and follow it easily). If anyone knows the answer they will most likely reply. :-) [12:00] Are people ready for hte open day? [12:00] ubottu: noted [12:00] Sorry, I don't know anything about noted [12:02] I am new in pachaging. I was trying to upgrade small packages for practice, but at pbuilder I always get an error. Always when it tries to apply the patches of the previous version. [12:03] marrow: well, the source changed then, so you'll have to see if the patches are still necessary and if they are redo the patch against the new source [12:04] I see. And do I need to note whether the patches are necessary or not in the changelog? [12:05] marrow: yep, that's usually a good thing to do [12:05] Ok. [12:05] One more thing which is not clear to me. A patch is a debdiff, right? [12:06] So if I make a debdiff from a modification, then I simply rename that debdiff to xyz.patch? [12:07] marrow: no, a patch is a patch (and a debdiff is basically also a patch, but is unrelated to this) [12:07] marrow: which patch system is the package using? (you can check by running the command what-patch, available in ubuntu-dev-tools) [12:09] Ok. I install the ubuntu-dev-tools first, and check [12:10] any other way to check the patch system, btw? [12:10] marrow: yep, look at debian/rules [12:11] RainCT: it says quilt [12:12] marrow: OK. I still haven't got to understand quilt but there are instructions on how to use it on the wiki :P [12:12] quilt is great IMHO [12:12] marrow: https://wiki.ubuntu.com/PackagingGuide/Howtos/Quilt [12:12] laga: simple-patchsys is great :P :P [12:13] RainCT: if a patch is a patch, and not a debdiff than how do you generate the patch [12:13] yeah. [12:13] RainCT: i only ever used quilt outside of packages.. but that was for 150 patches. it was great for that [12:13] marrow: Using quilt :). That wiki page explains how to do it. [12:14] Any other wiki pages? For simple-patchsys for example? :) [12:15] marrow: not sure if there's one, but you shouldn't change the patch system used by existing packages [12:15] 'lo all [12:15] marrow: and hey, I don't like it but this doesn't mean that you can't use it :) [12:15] hi Laney [12:16] Anyone fancy some reviewing? http://revu.ubuntuwire.com/details.py?package=goocanvasmm [12:16] for a dist-upgrade, does update-notifier show you the release notes? [12:16] Hi RainCT. How are you? [12:16] I am not considering to change the patch system. I take what is given [12:16] marrow: there are two ways how you may update this patch: quilt refresh .diff - this will probably not work, as you said that the patches don't apply [12:17] marrow: or deleting the old patch and adding a new one which does the same (instructions on the wiki) [12:17] RainCT: for simple patchsys? I thought that was the cdbs'ey one. [12:18] Hobbsee: that's for quilt. and yes, simple-patchsys = cdbs [12:19] oh, right, it actually is quilt. [12:19] Laney: fine, fine. I should be learning maths but well, I'm doing some Ubuntu stuff as it's funnier :P [12:19] hm, maths is fun too [12:19] ...not as fun as reviewing packages........ [12:19] * Laney runs away fast [12:19] * RainCT added a section about the copyright file to https://wiki.ubuntu.com/MOTU/Packages/REVU/CheckList [12:20] Laney: not if you have to learn the stuff from the past three months :P [12:20] (erm, it's 2 months actually, but anyway :P) [12:22] RainCT. Thank you for your help. I will dive further into the patch systems. [12:23] One more question: what is pushd and popd? [12:23] No manual entries for them [12:23] I hope this is the right place to report this, I just installed libvisual-dev to play around with it. But when I include libvisual-0.4/libvisual.libvisual.h and I compile these errors are given: /usr/include/libvisual-0.4/libvisual/libvisual.h:27:32: error: libvisual/lvconfig.h: No such file or directory. [12:23] I haven't seen these words since I learnt assembly [12:23] I think the files should be in /usr/include/libvisual/*.h [12:24] marrow: that's bash stuff.. pushd x = cd to directory x, popd = cd back to where you were before the pushd [12:24] I see [12:24] Thank you again RainCT [12:25] you're welcome :) [12:26] Laney: looking at it [12:26] \o/ [12:26] Don't forget your maths though [12:26] (If you want to see it working, look at glom in my PPA) [12:29] can anyone confirm whether I can use hardy script in /usr/share/debootstrap/scripts/ for jaunty? (I'm running hardy still...) [12:30] karooga: I guess so [12:32] Laney: why don't you use version 3 in debian/watch? (the only change is that you can remove the "debian uupdate") :P [12:32] RainCT: No reason. I copied it from elsewhere [12:33] Laney: oh, btw, I think I already looked at your package yesterday (the "Most of the docs are junk" made me remember) :) [12:33] heh [12:33] You didn't leave any comments [12:34] yeh, debian/* looked good at a first glance and I was too lazy to do a complete review :P [12:34] I just noticed a typo in the rules file [12:34] 0.1-2 -> 0.1-3 [12:36] Laney: right, consider using a variable for that [12:37] good idea [12:37] oh, why is everything I'm reviewing these days released under the LGPL? :P [12:39] Laney: debian/copyright is incomplete, some files are under the GPL [12:39] RainCT: The ones in examples/? [12:40] Laney: yep, and scripts/ltmain.sh [12:40] I got the examples one [12:40] but that file is from the FSF, not sure if it needs to be listed [12:44] RainCT: http://www.nabble.com/Should-copyright-information-for-automake-generated-files-be-added-to-debian-copyright--td17191924.html [12:45] RainCT: enjoyed your packaging chklist. thanks. [12:46] karooga: great :) [12:47] Laney: Alright, I'm fine with not listing it. [12:47] RainCT: Also, I have those GPL2 files listed, no? [12:47] http://revu.ubuntuwire.com/revu1-incoming/goocanvasmm-0811021315/goocanvasmm-0.12.0/debian/copyright [12:47] Or is the way I did it unclear? [12:48] Laney: oh, sorry. I did an "apt-get source" but haven't updated the package list today :) [12:48] heh [12:49] Laney: uhm.. does the LGPL2.1 need the GPL file or does it include all terms? (I know that LGPL3 requires the GPL3 to be shiped together with it, but about 2.1 I'm not sure, and the file is 4 times as long as 3) [12:50] I don't know what you mean [12:51] Laney: nevermind, I see now that LGPL2.1 seems to be a license by itself :) [12:51] RainCT: for more stuff you shouldn't od, you could also add: http://mjg59.livejournal.com/99905.html [12:52] but maybe it's a bit too upstream-based [12:52] Laney: I asked because the LGPL3 file just includes "Additional clauses" to be applied upon GPL3, so I wasn't sure if it was the same for previous versions [12:52] I see [12:52] "First, let me make one thing clear. This isn't constructive criticism. This is just criticism" LOL [12:53] RainCT: wait till you read the multiple points of fail... [12:53] would it be wrong to format the guy's computer to vista, and lock the bootloader? [12:56] Hobbsee: good to know that our archive admins have a *matix filter for the archive :P [12:56] sebner: yeah, looking at that... [12:56] * Laney remembers jdong's april fool [12:56] although i note they haven't been stupid enough to submit that. [12:56] Laney: Indeed. [12:56] Hobbsee: yet ... :P [12:57] sebner: well, if they do, i'm sure they won't live long. That will become a job for someone like mneptok [12:57] oh, mjg59 rage [12:57] RainCT: Check the latest upload. I added a variable to rules, fixed watch and relicensed in copyright [12:57] \o/ [12:58] hrh === bureflux is now known as afflux [13:01] Hobbsee: that's scaring.. and I'm going to love mjg59 :P [13:02] RainCT: oh yes...The last one was also pretty good, too [13:03] RainCT: you should read his blog. it's awesome [13:04] * laga is a huge mjg59 fanboy. embarrassing. [13:05] laga: if he's on planet.ubuntu.com, I'm already reading it :P [13:05] RainCT: not all of his postings go there [13:10] He is certainly rather high in the awesomeness ranks. [13:11] Link? [13:12] http://mjg59.livejournal.com/99905.html [13:12] orly_owl: ^^ [13:27] is there any preferred for patch names using cdbs (cdbs-edit-patch some_name)? [13:36] hi [13:37] short question - is there any guide that can describe how to create package if upstream is available only as cvs or svn checkout? [13:37] or maybe I should just checkout and tar? [13:37] am I right? [13:41] marcin_ant: yes, but create a get-orig-source target in debian/rules to generate the tarball [13:48] anyone familiar with cdbs-edit-patch? === fta_ is now known as fta [13:49] karooga: what's the question? [13:49] RainCT: is there any preference naming patches? [13:50] karooga: just use something that makes what the function of the patch as clear as possible [13:50] karooga: cdbs will apply patches alphabetically, so a common option is to append a number to the patch name [13:51] like 01_fix_foo.patch [13:51] RainCT: Ok cool. And if there are multiple changes, is it good practice to create different patches for each change? [13:51] karooga: yep [13:52] pochu: thanks [13:56] Laney: looks good, but I have no pbuilder environment around right now [13:57] persia: use case for neutral comments ^ :P [13:57] RainCT: I've just uploaded it to my PPA, should be built soon [13:57] ;) [14:01] ok, I'll go study maths in the meantime :P [14:09] RainCT: Built [14:17] RainCT, Except you're asking for another revision when you do this, which ought be a rejection. [14:20] persia: EMENOTUNDERSTAND [14:23] any suggestions on a good UML tool? [14:24] RainCT, Your use case for neutral comments is that you want patches to be named sanely. You have asked for a new revision with sane names. This is a rejection. If you liked it, you'd advocate it without comment. [14:24] rawler, You might try gaphor, but it's not had a history of working well. [14:25] persia: that's not my use case :). the use case is "looks good but I have no pbuilder environment around right now" :P [14:25] RainCT, Umm... That's not a useful comment. [14:26] (and whether this is a good use case or not is arguable, but it is one :P) [14:26] You're just saying "I didn't actually review this", which doesn't add value. [14:27] Yeah. I'll give it to you as a use case, although I'm of the opinion it's more of a useless case. [14:27] persia: it is if I forget which package it was that I had reviewed, as it can then find it easily looking at my last comments, or if I look at it then just by seeing the comments I remember "ah, that's the one I reviewed" :P [14:28] yeh, I think those which siretart had were better [14:28] I suppose. I've always been a fan of doing proper reviews in FIFO order, and I try to avoid reviewing the same package twice in a row as I don't believe that packages benefit from having the same reviewer for multiple revisions. [14:29] So, I don't see any value in keeping track of which packages were reviewed, especially if the review wasn't actually performed completely. [14:30] Anyway, late here. We can resume the debate another day :) [14:30] persia: oki, will try.. thanks.. :) [15:31] * RainCT feels the silence [15:35] heh [15:36] I'm trying to install Intrepid on my new macbook [15:40] Laney: how new is that ? [15:40] 5 days [15:40] is that one with the dual-graphic cards thing or the latest of the previous serie ? [15:41] I think only the pros have that [15:41] ok [15:42] dual cards for better performance or multi monitor? [15:43] orly_owl: One is low-powered for battery life and one is for performance [15:43] you can switch which one it uses [15:43] ah [15:43] i've heard of this. was there another company offering it? [15:43] Dell maybe [15:49] wow tiny mobo http://cache.gawker.com/assets/images/gizmodo/2008/10/SHdpS4vY1kdqB4TU_01.jpg [16:03] if creating an initial package, which requires some patches - do I add this information to the changelog? [16:04] karooga: I don't believe so. Just document the changes (in text) in the patch (you should use a patch system) [16:24] did I miss any replies? my connection got dropped (again)! [16:25] 11:04 < Elbrus> karooga: I don't believe so. Just document the changes (in text) in the patch (you should use a patch system) [16:26] jdong: ta. [16:27] RainCT: I think "I didn't do a full review, but what I looked at looks good" is a useful comment. [16:31] ScottK: wow, today you're supporting all my thoughts :) [16:31] RainCT: Keep having good ones then. [16:31] ;-) [16:34] karooga: I usually write a line for each patch in debian/changelog (like "* debian/patches/: fix foo"), but if you write a description in the patch itself then it's fine too just write the name of the patch in debian/changelog (imho the name is important so that you can easily see when each patch was introduced) [16:36] RainCT: didn't realise you where notified of my last upload automatically. Was just checking for new lintian errors... still working out how I'm going to do the licensing. [16:36] RainCT: I think karooga was asking about an initial packaging. Of course I agree with later changes getting mentioned. [16:36] karooga: (note that if you'd add a patch to an existing package from Debian then it's good to add as much information as possible (links to bug reports, whether the patch is Ubuntu specific or not, etc.), so that the Debian maintainer can decide without too much work if he wants to adopt it or not) [16:37] karooga: heh yes, I subscribed to mail notifications :) [16:37] Elbrus: yes, but I think it's also good to list at least the names in the initial changelog entry so that people the package in the future see "aha, this patch has been here since the initial packaging, and it was added by " [16:38] (other may have a different opinion here, though) [16:38] RainCT: I do think you have a point there. [16:38] :) [16:38] * Elbrus luckely didn't have to patch my first package... [16:39] s/my/his ;) [16:39] * RainCT maintains at least one package with over 5 patches since the first release :) [16:40] (most of those Debian specific, though) [16:41] RainCT: cool - i mention the patches in changelog. How do I handle adding licenses? You mentioned repackaging - I think patches should sort this out (elegantly)? [16:43] RainCT: oh and this isn't in debian (although I would like to get it into debian too). :-) [16:45] karooga: if you say that because of point 6, Ubuntu is a Bebian system too. (and getting the package into Debian would be awesome! :)) [16:46] karooga: about the first question, I'm not sure but I think the licenses have to be in the .orig.tar.gz [16:46] ScottK: does the GPL need to be in the orig tarball or is enough if it's in debian/ (ie, .diff.gz)? [16:48] A full copy of the license needs to be in the orig tarball [16:49] ScottK: right, thanks for confirming that :) [16:49] karooga: so yes, ask upstream to release a new tarball including the LGPL [16:50] karooga: (and nevermind about mentioning the GPL in debian/copyright, that's only for LGPL3) [16:51] RainCT: yup, already asked him. In your review you mentioned repackaging myself to include LGPL... [16:51] karooga: yeah, but only if upstream doesn't answer [16:51] RainCT: GPL referring to my packaging, right? [16:53] karooga: no, about point 6 (forget that) and "ask the author to add both files to the source tarball" (only the LGPL is enough) [16:53] I got confused about this because LGPL3 does require to be shipped together with the GPL, because it just contains some "Additional clauses" to add to the GPL, but LGPL2.1 seems to be a complete license on its own [16:54] Laney: advocated :) [16:54] RainCT: \o/ [16:56] Anyone up to give http://revu.ubuntuwire.com/details.py?package=goocanvasmm a second advocate (or to hit Laney with complains about something being wrong *g*)? :) [16:58] When's the repo opening for jaunty? [16:59] And I have a package that I've made but havn't submitted to debian/ubuntu. Any clue on whether I should persue debian first, or will that take too long for my package to get in before the import freeze? [17:00] lfaraone: I'd say try Debian first [17:01] for me the response time there has been pretty good (although I'm not sure how it's now with lenny being freezed) [17:01] RainCT: Ok, I'll ask at OFTC. [17:01] RainCT: Is work in debian considered when applying for MOTU? (a long-term goal of mine) [17:02] RainCT: ok so only LGPL is needed. Does your comment about licensing info in the .c and setup.py still stand? [17:02] Oh, and is there a WYSIWYG way to write manpages? [17:02] karooga: yes [17:03] lfaraone: (to get a package into Debian, submit it to mentors.debian.net and then mail debian-mentors@ asking for review -there's a template for that on the webpage) [17:04] RainCT: Thanks. [17:05] lfaraone: There may be some year old (ugly) stuff, but it isn't difficult at all to do them yourself. One easy way, for instance, would be https://wiki.ubuntu.com/PackagingGuide/Howtos/PODManpage, but just doing them in roff directly isn't very difficult neither (I may write a howto for that too some day) [17:06] RainCT: ah, kk. [17:06] RainCT: http://paste.ubuntu.com/66332/ [17:06] RainCT: is it upstream's task to write a manpage, or is it acceptable to write one for packaging? [17:06] (for the package itself) [17:06] RainCT: is that right? [17:07] lfaraone: You can write one, but submit it back upstream once you do. [17:07] Laney: Ok. [17:07] lfaraone: it's good to write it yourself if there isn't one (and send it upstream for them to include it into the tarball) [17:07] (fortunately, I'm familiar with the upstream, so I'll try to get that accepted) [17:08] karooga: that isn't the most usual formatting but yes, if dch can parse it it's fine [17:09] RainCT: what is the program that does proper formatting of changelog? [17:10] ah, dch [17:10] :-) [17:10] karooga: the human mind :) [17:10] and dch can help too ,) [17:10] Hey, where are the app descriptions that show up in gnome-session-preferences derived from? [17:11] what is gnome-session-preferences? [17:11] Laney: the app that lets you choose what starts up when you login [17:11] Oh, -properties :P [17:11] Laney: ie pidgin, gnome settings daemon, etc [17:11] (I tried to run it) [17:11] Laney: lol, yeah [17:12] karooga: http://paste.ubuntu.com/66336/ that'd be a more normal format [17:12] RainCT: do you have an example pkg? with initial release + patches? [17:12] RainCT: ta [17:12] karooga: the explanation for patch 02 is unclear, btw (why is that needed for hardy?) [17:13] karooga: Did you see https://wiki.ubuntu.com/UbuntuDevelopment/PatchTaggingGuidelines ? [17:15] RainCT: there is a move away from g2c apparently, gfortran is the future. [17:16] karooga: then that isn't only necessary for Hardy, is it? [17:16] Laney: didn't see that - was just looking at the patch section in cdbs documentation. [17:16] karooga: write something like "Change from g2c to gfortran because the former is not available in Ubuntu." then [17:16] RainCT: nope, already changed to ">= hardy" [17:17] k [17:17] So, when does jaunty unfreeze? [17:17] lfaraone: when there's a mail on ubuntu-devel announcing that it's unfrozen :) [17:17] *or open, rather [17:18] RainCT: lol. [17:22] If you are packaging in a PPA, should you put yourself as the maintainer instead of motu? [17:23] I want my package in ubuntu after some build blocking bugs are solved [17:23] yep [17:24] hmm, than I need to fix all the packages in my PPA :( [17:25] but in retrospect it is quite trivial... [17:36] Elbrus: if you aren't doing so, also use ~ppaX or something in the version number [17:36] (so that the package will be replaced once it's in Jaunty) [17:37] that's exactly what I am doing [17:37] laney: those patch tagging guidelines are awesome! [17:37] so I wouldn't need to change the maintainer? [17:38] Elbrus: Alright then. You do, the Maintainer shouldn't be MOTU if we the MOTU Team isn't responsible for it, but it isn't something urgent to change unless people start mailing ubuntu-motu@ to complain about them [17:39] I completely understand. But I guess the idea is that it could be pulled from the ppa. [17:39] and IIRC it prevents lintian warnings [17:44] Where should the environment variables bet set? DEBFULLNAME, etc. [17:47] karoog1: Yeah, thery're a good idea imo [17:47] gouki: I have them in .profile [17:47] Er, no that's not right. I have them in .bashrc [17:48] usually they are set in ~/.devscripts I believe [17:50] Laney: do you know where they tie in later? [17:50] karoog1: What do you mean? === karoog1 is now known as karooga === bddebian2 is now known as bddebian [17:52] Laney: i.e. does it integrate into launchpad later, or something like that? [17:52] karooga: It's used by dch and some other apps to fill in your name and email [17:53] Elbrus: the warning about the maintainer not being @ubuntu.com? you can ignore that for PPAs [17:53] Laney: cool. [17:53] karooga: Launchpad checks the e-mail in the debian/changelog file to know who to credit for an upload [17:54] (but to authenticate the package when it's upload it looks at the GPG signature in the changes file) [17:55] RainCT: did the launchpad src code get opened? [17:56] karooga: not yet === ScottK is now known as ScottK2 === ScottK2 is now known as ScottK [17:59] * RainCT notices that he has reverted the work the had done on REVU yesterday and headdesks XD [18:02] * ScottK hands RainCT an aspirin. [18:03] ScottK: thx ^^ [18:03] well, it isn't that bad as I didn't like the way I had done the changes anyway :P [18:07] btw, do you think it would be of any use if ubotu could fetch info about packages on REVU? [18:09] I just got a big spew of stuff, and need a hint as to what the problem is [18:10] sudo apt-get build-dep gqcam filled my ssh term buffer with stuff like var/lib/scrollkeeper/C/scrollkeeper_extended_cl.xml:1726: parser error : attributes construct error To Add Files to an Archive [18:10] here is the whole buffer http://dpaste.com/88322/ including an attempt to do it again [18:11] hmm... myabe this is an #xubuntu thing... [18:13] CarlFK: here it works (Ubuntu Intrepid and Ubuntu Server Hardy) [18:13] yeah, worked on a normal ibex box too [18:14] hmm, your -server doesn't have gnome either. this made me think it was because I was on xubuntu: Extra content at the end of the document [18:47] If the INSTALL file doesn't mention any dependencies, the other option is trial and error, right? [18:48] gouki: look at configure.ac or configure.in [18:48] INSTALL is usally generic for autotool projects [18:48] azeem, thank you. [18:49] gouki: especially check the PKG_CONFIG macros in there [18:49] README might have a list of dependencies as well [18:49] azeem, README also didn't mention anything. I'll look into the file you mentioned. [18:51] gouki: apt-get build-dep [18:52] CarlFK, it's a new package that is not in the repositories. [18:52] oh... never mind :) [18:52] :) [18:52] I'm trying to understand how to fill the dependencies field in control. [18:52] you mean Build-Depends, right? [18:53] azeem, yes, I'm sorry. [18:55] azeem, BTW, in configure I already found out it requres libssl :) Thanks. [18:56] Does mention version, so it's safe to assume the latest one will work? [18:57] it's not safe [18:57] but it's obviously the only thing you can assume [18:57] OK [18:58] Also in Build-Depends, should I leave debhelper and autotools-dev and add the new ones? [19:03] gouki: yep [19:04] RainCT, thank you. [19:07] RainCT: looking at another pkg now - which has LGPL in orig tar ball. Is it necessary to reference it in copyright? [19:10] karooga: you've to point to /usr/share/common-licenses/LGPL2-2.1, as the LGPL file won't be included in the .deb [19:13] RainCT: really? so debuild actually excludes it? [19:13] It's LGPL-2.1. [19:14] karooga: debuild includes what debian/rules gives it [19:14] right, typo [19:15] * iulian is going to pick up a package from REVU and reviewing it. [19:26] Hello. [19:26] What is wrong with using checkinstall to create Debian packages? [19:27] * Elbrus would like REVU to have the desciption of the packages on the front page... [19:27] Tetracomm: you can do that for private use ;) but such packages won't ever be in the archive [19:28] !checkinstall | Tetracomm [19:28] Tetracomm: checkinstall is a wrapper to "make install", useful for installing programs you compiled. It will create a .deb package, which will be listed in the APT database and can be uninstalled like other packages. See https://help.ubuntu.com/community/CheckInstall - Read the warnings at the top and bottom of that web page, and DO NOT interrupt CheckInstall while it's running! [19:28] sebner: Why? [19:29] Tetracomm: because they aren't valid packages. it's more like a quick and dirty dirty dirty solution for normal users [19:29] Ok. [19:30] sebner: Is there a method which isn't too complicated for the creation of Debian packages? [19:30] !packaging | Tetracomm [19:30] Tetracomm: 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 [19:31] !backports | karooga [19:31] karooga, please see my private message [19:31] sebner: Isn't that very complicated? [19:32] Tetracomm: with a little bit of motivation, asking here if you have problems .. no. this is the only way to create valid packages [19:33] Ok. [20:20] So, is anyone working on Gnome bugs in 8.10? [20:32] pangloss: given that gnome is in main, it's not our area [20:32] pangloss: #ubuntu-desktop is what you want [20:32] or gnome upstream, ofc [20:32] ajmitch =/ im having window redraw issues, makes the top of the window unreadable.. [20:33] time to go kde? [20:33] hah [20:33] * ajmitch wouldn't expect too many people to be around at this hour [20:33] you've filed a bug about it? [20:34] tried with & without desktop effects? [20:34] is it appropriate to change the title of an old bug (LP: 242668) so that it has '[needs packaging] package'? [20:34] pangloss: have a nvidia card? [20:34] ajmitch im actually searching launchpad now to see if someone's files a bug [20:35] I see it has already been triaged (but it was not changed). [20:35] karooga: if you're going to do that, make sure it has the right tag as well [20:35] afaik tags are still being used to find those bugs [20:36] jdong yes nvidea [20:36] pangloss: consider if it's bug 186382 [20:36] Launchpad bug 186382 in nvidia-graphics-drivers-177 "[nvidia 169.xx] gtk window decorator somtimes draws decoration ugly or not at all" [Undecided,Confirmed] https://launchpad.net/bugs/186382 [20:36] ajmitch: what would the tag be? [20:36] karooga: needs-packaging [20:36] pangloss: not a GNOME bug, but an nvidia proprietary driver bug [20:36] >.<\ [20:37] didn't do it in 8.04 [20:37] pangloss: yes, new revision of nvidia driver in 8.10 [20:37] ah [20:37] pangloss: the old one is incompatible with the new xorg-server-1.5.0 [20:37] pangloss: nvidia engineers said at nvnews.net that they are aware of the problem and are investigating internally [20:37] pangloss: I'm afraid there's not much we can do but keep an eye for developments [20:38] jdong thanks for the info =) [20:38] sure thing [20:38] ls [20:38] jdong, think I should use free driver until then? [20:39] ooopss, sorry. [20:39] ajmitch: apparently the tag "has now been used before"? is this normal LP behaviour? [20:39] pangloss: AFAIK the free driver doesn't do Compiz desktop effects to begin with [20:39] ajmitch: s/now/not/ [20:39] karooga: you've misstyped it then [20:39] pangloss: as long as you disable desktop effects this problem shouldn't present itself [20:40] jdong: AFAIK is the default when you install right? [20:40] pangloss: that's correct [20:40] jdong, ok then, disabled effects till then [20:40] RainCT: I have? "needs packaging"? [20:41] karooga: it's needs-packaging [20:42] RainCT: weird, that's what I thought i said...? [20:43] missing hyphen [20:43] in case your irc client has a funny font that doesn't display it :) [20:45] ajmitch: yup, i did have have the hyphen. If the bug title has "[needs packaging] some_pkg", does the tag also have "needs packaging"? (or needs-packaging in this case?) [20:46] not sure, I haven't looked at many of those bugs [20:46] Where on control can I add suggested packages? [20:46] I think in this specific case, the bug hasn't been filed against ubuntu [20:47] so it needs to be marked as being against ubuntu, not just the upstream project [20:47] * ajmitch didn't notice that when looking at it before [20:47] * karooga thinks he needs packaging time with his pillow, lights out and computer off. [20:48] ajmitch: think i'll file a new bug then [20:53] If one needs the package libssl0.9.8, it would be libssl (>=0.9.8) on the control file, right? [20:56] having a build-dependency on libssl-dev & debian/rules doing the right thing would put that in there [20:56] unless there is a really really really good reason why you need a particular version :) [20:58] yay, mjg59 blogging again :) [20:58] ajmitch: Yeah. Great post that one. [20:59] sad to see how things just haven't improved [21:00] ajmitch and jdong, thank you. [21:01] what on earth is it with everyone starting new forums this month? [21:01] RainCT / Laney / ajmitch: shots for the help today. Lights out time for me. cheers [21:01] * jdong takes an aggression inhibitor pill [21:02] jdong: it's the cool thing to do [21:02] * ajmitch hands jdong a copy of ultamatix to play with [21:03] ajmitch: do... I even want to know what that is? [21:04] oh you haven't read that wonderful dissection of it? [21:04] http://mjg59.livejournal.com/99905.html [21:04] how can i public an ubuntu package in a repository? [21:06] ajmitch: my eyes are bleeding already on the 3rd bullet. [21:06] I mean once that i have made a package... where do I upload it?, so that the other people can use apt-get to install it [21:07] one of the easiest ways for that is to upload to a PPA on launchpad, or to some other hosted service [21:08] ajmitch: (/me was away.. it's "[needs-packaging]" in the title and the tag is "needs-packaging" too) [21:08] unless we switch to wnpp :) [21:09] RainCT: right, in this case I didn't notice that the bug was only against an upstream project on LP === thunderstruck is now known as gnomefreak [21:15] time to try & hunt for a new laptop that won't suck with ubuntu === TheMuso_ is now known as TheMuso [21:16] * ajmitch spots a TheMuso [21:20] ajmitch: Tell me if you find one on this side of the world, please. [21:20] wgrant: the problem is that it's awfully hard to tell just from the manufacturer specs provided [21:22] Indeed. [21:23] And getting a reasonable one without the Windows tax is difficult. [21:27] on top of that, I need to find a decent ISP [21:27] * ajmitch is about 500MB from being capped for 2 weeks at home [21:29] ajmitch: Hey, I lived reasonably badly on 1GB per month until 18 months ago. [21:29] You should know that there are no decent ISPs down here. [21:29] I have no idea how you could manage that [21:30] Neither. [21:30] But parents wouldn't have it any other way. [21:30] of course not :) [21:30] my parents are still on dialup at home [21:31] Hi, I was thinking of packaging revoco, but I see a debian ITP. However, after reading http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=406377 , should I bother trying to package it for ubuntu? Are the copyright requirements less stringent in ubuntu, or is a revoco package likely to get rejected for the same reasons as in debian? [21:31] Debian bug 406377 in wnpp "ITP: revoco -- control Logitech's MX-Revolution mouse" [Wishlist,Open] [21:32] * ajmitch looks around for an ftp master [21:33] Depends on exactly why it got rejected. [21:33] copyright policies are generally pretty similar, though that doesn't say anything about the reason for rejection [21:33] "Written November 2006 by E. Toernig's bonobo - no copyrights. [21:34] " that's the copyright in the source and the author is apparently unwilling to change it. He states in a personal email, that it could be considered public domain [21:35] my question is whether that will be an issue. From the last comment in the bug report, I see the copyright as the reason it was rejected [21:37] I guess I can just try and see what happens :) [21:37] we're not going to be able to give a definitive answer either way [21:37] jrib: Does the source indicate any kind of license? [21:38] ScottK: nope, that quote is the only non-programming line in the tarball (and the author's email) [21:39] I suppose, if I contact him and it's public domain, I could release it and just add GPL right? [21:40] I will be rejected. [21:40] *It [21:41] wgrant: referring to the original or with the addition of GPL? [21:42] I would suggest rejection unless the upstream tarball has the GPL licenses. [21:42] It doesn't seem to make sense for the license to be added downstream. [21:43] jrib: In some countries it's essentially impossible to for something to be created in the public domain. [21:43] Germany is one, isn't it? [21:44] * ajmitch detests copyright issues [21:44] jrib: ask him to release it under the ISC License or something [21:44] (him = upstream) [21:45] jrib: Maybe you could convince him to license it under http://sam.zoy.org/wtfpl/ [21:45] That will, in fact, be enough. [21:46] ScottK: heh, are there any packages in ubuntu with that license? [21:46] I'm pretty sure I remember hitting one at one point. I don't recall which. [21:47] okay, so my first step will be to contact the author. Thanks ScottK, RainCT, wgrant, and ajmitch [21:47] jrib: http://www.codinghorror.com/blog/archives/000833.html seems like a decent overview. [22:23] wgrant: Oh, I didn't know that there's an ubuntu-directory team.. Prepare to be annoyed with question :D [22:25] RainCT: There isn't, really. We do approximately nothing, other than get spammed with Samba bugs. [22:26] it's terribly exciting [22:26] lol [22:26] what questions do you plan to ask anyway? [22:27] RainCT: I think you should go ahead and do your proposed REVU cleaning, just be sure to include a nice note about resetting the queue and update your packages for Jaunty. [22:29] ScottK: sure, but I'll wait some more time for people to complain to ignore them all then :) [22:29] ajmitch: well.. dunno xD [22:30] RainCT: I think a nice comment is the only real resolution to the whining. [22:30] RainCT: Otherwise do nothing and then I'll look at REVU, see the list, and throw up my hands and find something else to spend my time on. [22:30] cleaning up while it's being discussed on the mailing list isn't too helpful though [22:31] ajmitch: I don't feel the discussion is particularly productive. [22:31] It seems a lot of 'don't do that' but no decent alternatives [22:31] Agreed that there's no huge rush though. [22:32] ScottK: And then, I really don't undersand why there's people against the proposal but then say "archive everything every cycle", which is even harder o.O [22:32] you've always been one who wants due process with making changes [22:32] ajmitch: For MOTU policy sure. I don't think that applies here. RainCT is the REVU coordinator and I'm glad to let him do it. [22:32] but I'm going to wait a bit more before I do anything, let's see if the discussion leads anywhere [22:34] * NCommander wakes up [22:34] RainCT: do you really believe you will see a usable result? [22:34] What about REVU? [22:35] geser: not really, but well.. who knows? :) [22:35] NCommander: check your mails [22:35] * Hobbsee wonders about a week allocated to reviewing revu, starting now (or last week) [22:35] seeing as people have nothing to upload atm anyway [22:35] and should probably wait for important main stuff to come thru. [22:35] * wgrant wonders about a week running TRUNCATE upload; [22:36] Hobbsee: I hope you have some spare whips handy already [22:36] geser: to get people to do it? [22:36] Hobbsee always has whips at hand [22:36] geser: i could try - but i don't do it myself much, so.... [22:37] I personally think anything sitll targetting intrepid should be nuked [22:37] we have a huge signal/noise rational on REVU [22:37] NCommander: the bigger problem is that the MOTU can't/doesn't process the signal regardless - perhaps due to the noise [22:37] NCommander: but a lot of the signal becomes noise, due to the lack of reviews being done by the MOTU. [22:38] Agreed [22:38] But there are times packages in the Needs Work category have been there for awhile [22:38] We only zap packages after they've been there for a few months [22:38] I do think that new contributors posting on REVU is a complete waste of time. [22:38] Hobbsee, ? [22:38] NCommander: I'll add automatic archiving for stuff without comments since 6 months or something like that (most of the code is already there) [22:39] so noone really has any ideas for encouraging MOTUs to review stuff, since it's not exactly a fun job by any means [22:39] We could modify REVU to only support uploads from people who are UUC or greater, but there are people start off w/ packaging stuff [22:39] * NCommander did that w/ codeblocks [22:41] Can we restrict to one new package per 10 other uploads or something? [22:41] Its tricky [22:43] I prefer if people start with fixing the current packages instead of creating new ones, so I'd be for limiting REVU for UUC (and greater) as I hope those people know how much work maintaining is and don't package everything they find on the internet [22:43] I can make the necessary changes to launchpad_login.py [22:44] geser: +infinity [22:44] But unless we have consensus on this from the MOTU community [22:44] Based on how people seem to be having a lot of trouble with PPAs, and getting packages to build (even substandard ones, which probably don't have correct fields, etc), it's clearly a very difficult task. [22:44] I'm unconfortable from the change [22:44] Well, we have a working PPA importer [22:44] (its actually on production, RainCT still hasn't finished cleaning up my daemon though) [22:44] it's far too high a barrier to entry, and hardly anyone will review a new packager's work anyway. [22:44] maybe we should restrict direct REVU uploads to UUC and greater [22:45] NCommander: now that might be a good idea.... [22:45] But allow imports from anyone [22:45] bah. why would that make any difference? [22:45] because you at least have a package that builds cleanly [22:45] oh, right, because it would actually have had to build somewhere. [22:45] NCommander: actually, it's finished [22:45] * NCommander hugs RainCT [22:45] it's just missing a check to only accept packages that have build correctly [22:46] NCommander: that's probably one of the best ideas i've heard all day, in conjunction with RainCT's documentation stuff last night. [22:46] which should be quite easy to add.. [22:46] RainCT, there was one ... [22:47] NCommander: the text on the pages mentions it, but it wasn't actually implemented, or? [22:47] SO launchpad_login.py would only sync keys for UUC or greater [22:47] I thought I implemented it [22:47] It might not have actually gotten committed because my last commit never got pushed [22:47] So we need to modify launchpad_login to import a GPG key to the REVU keyring if your a UUC (or how about just Ubuntu member) [22:48] * NCommander notes that means resetting the keyring again ... [22:48] NCommander: because ubuntu members aren't necessarily clueful, w.r.t. packaging. [22:48] Fair enough [22:49] someone coming from e.g. LoCo has still no clue about packaging and how much work is it to keep a package up-to-date [22:49] So UUC for direct uploads? [22:49] UUC makes sense - they've been known to put good packaging stuff forwards. [22:49] I've been wanting to propose changes for UUC [22:49] I think they should have the right to request syncs [22:49] * ajmitch wonders if he's indirectly in UUC [22:49] NCommander: CC the list on your suggestion about it, but i suspect you've found a winner, at least. [22:49] Hobbsee, which suggestion [22:49] NCommander: htat's been discussed a few times before, and the answer has been a resounding "no". [22:49] NCommander: a sync is potentially as harmful as a normal upload, if not more [22:49] NCommander: the one about REVU and UUC. [22:49] RainCT: it is more - it's often gettign rid of existing changes. [22:50] yeh [22:50] and we've seen what happens when people request many, many untested syncs [22:50] * Hobbsee twitches and shudders [22:50] Heh heh heh. [22:50] Ok, fair enough [22:50] * NCommander would think a UUC would be experienced enough though to say that [22:50] ajmitch: i think he's on brainstorm, you know. Just in case you wanted to go say hello. [22:51] yes, well [22:51] * ajmitch wasn't naming any names, of course [22:51] no, of course not [22:52] NCommander: not necessarily. u-u-c applications don't look at technical skills [22:52] NCommander: you'd have to make it that MOTU could upload as well, there's no team association that I can see at the moment [22:52] ajmitch: ubuntu-dev rather [22:52] either one [22:52] whatever works properly [22:52] * NCommander grumbles [22:52] Figuring out if someone in a team is already a huge headache [22:53] but I thought ubuntu-dev is in u-u-c [22:53] RainCT, nope [22:53] NCommander: api doesn't support that yet? [22:53] LauncpadLogin doesn't use it [22:53] Yet [22:53] I was thinking RDF parsing [22:53] RainCT: it could be changed to do that, if it doesn't cause any bugmail problems [22:53] Erm. [22:53] motu is a member of u-u-c, not ubuntu-dev. [22:53] That looks wrong. [22:53] quite probably [22:54] the teams are a bit of a mess [22:54] indeed [22:54] * ajmitch is not really a motu :) [22:54] NCommander: are you sure? I thought that the launchpad api already allows to get a list of people in a team [22:54] and the teams a person is in [22:55] The RDF API we used when I wrote the inital login code didn't support it [22:55] NCommander: I mean the Python API [22:56] uhm.. Kasper's mail has just remembered me about some stuff I was thinking about a while before [22:56] which is, getting Debian into REVU :P [22:56] o_o? [22:56] RainCT: what, and why? [22:57] that'd be, have a "I'm interested in maintaining this package in Debian." option on REVU, and have a team of "Friendly DD's" who'd care about assisting such people with getting the package into Debian [22:58] given how difficult it is to get MOTUs to review, how do you plan to coerce DDs? [22:59] (the idea is still rather unclear.. I though about it before while I was showering :P) [22:59] ah, one of *those* ideas :) [23:00] How about telling people that there are much better things to do than make new packages? [23:00] intrepid was released with a bug-free universe, wasn't it? [23:00] the packages on REVU will be likely to have already been reviewed by MOTUs so should be in rather good shape. and perhaps we could get the utnubu guys to give their thoughts about this [23:00] ajmitch: Definitely. [23:01] What Utnubu guys? [23:01] ajmitch: of course. what are bugs even? [23:01] well, there was an utnubu team.. not sure if it's still alive [23:01] (probably not) [23:01] RainCT: It is alive in the same sense as ubuntu-directory. [23:02] wgrant: I think a different sense and I think it largely did what it was meant to do. [23:03] NCommander: will you raise the "u-c-d only" topic? [23:03] that reminds me of what I was going to write - get an overview list of stuff that should get back to debian [23:03] hm.. perhaps we should discuss it this Friday [23:04] on the ML it would just result in a flame :P [23:04] * ajmitch hasn't exactly seen any flames on the list [23:04] (on IRC probably too, but well.. wasn't IRC created to have flames on it? :D) [23:04] no [23:04] ajmitch: ;) [23:04] we try & keep things flame-free, rather than calling people retarded idiots for even dreaming up the idea [23:05] it's much better to call ideas stupid, rather than people :) [23:05] when did I/we/anyone say that someone is a retarded idiot o.O [23:05] * ajmitch will be back to flame^Wcorrect after lunch [23:06] (beside the creator of anything ending with *matix :D) [23:06] ajmitch: enjoy your meal? (where do you live, btw?) [23:08] now that's called an efficient ML http://lists.alioth.debian.org/pipermail/utnubu-discuss/ [23:09] * NCommander reinstalls REVU on 127.0.0.1 [23:09] Utnubu isn't dead? [23:09] NCommander: of course not!! there's one thread per month on the ML! [23:09] :P [23:11] Second question [23:12] s/meal?/meal/ [23:12] Should we give UUC the ability to +1 on Ubuntu? (aka, one MOTU and one UUC would be good enough to qualify for upload of a package) [23:12] s/Ubuntu/REVU/g [23:13] -1 [23:13] I hope not. [23:13] I agree, but I figure it was worth asking [23:17] i'm wondering, what would you think about having a "package of a day" (having a wiki page where contributors could add themselves to say "I'll be available day X to be fix any problems with my package" and then during that day we aim to get the package into Ubuntu)? [23:17] (yeah.. also a shower idea :P) [23:18] RainCT: It sounds similar to the hug days that we hold [23:18] but those have tended to be kinda effective. [23:19] well, as effective as stuff like five-a-day and such is [23:19] yeah.. revu integration into five-a-day is something I considered yesterday XD [23:21] 5-rejections-a-day? [23:21] it would be 5 comments a day, surely :P [23:22] Hobbsee: right [23:23] but I still have to reconsider this at the next shower [23:23] as well as revu integration into ubottu :P [23:24] (well, I'm unlikely to pursue this last idea unless someone wants it) [23:25] well, I'm off for today. good night [23:30] NCommander: I also think the 'only UCD can upload to REVU' idea is a policy change thought ought to be agreed to my MOTU in general. [23:31] ScottK: which would be why it's now on the list, no? [23:31] Hobbsee: OK, but that's not a "OK, go do that". That's something we need to seriously discuss (unlike the rest of the discussion). [23:32] I think if RainCT had wanted to just clear the queue without discussion, as REVU coordinator that'd have been fine. [23:32] ScottK, well, I proposed the idea [23:32] ScottK, I think that's a given, but we get a tremendous amounts of uploads to REVU ... [23:34] I can see it has merit, "No cookie for you until you prove yourself." [23:34] It also means UDC also actually means something w.r.t to privelleges give [23:34] *given [23:38] * ScottK loves the new wiki.ubuntu.com [23:39] ScottK: Haha, indeed. [23:40] oh that person deserves talking to :) [23:40] * wgrant reverts. [23:40] spoilsport [23:41] too late :P [23:41] Bah. [23:41] I was trying to revert, and it was complaining that I hadn't changed the page content. That was rather confusing. [23:41] * ajmitch wonders why it says immutable page [23:41] ajmitch: Not when logged in [23:42] that's one page that probably shouldn't be changed even by random logged-in people [23:42] I thought it was restricted. [23:43] Anyone got a link to that nice graphic of all releases and how long they will be supported for? [23:43] it was a Canonical marketing one, AFAIK [23:45] Laney: http://blog.canonical.com/?p=15 [23:46] james_w: Winner [23:49] * ajmitch spots a suggestion on brainstorm for jaunty to be feature complete by alpha 1 [23:49] If we decided to have no new features, sure. :) [23:49] and if GNOME decided to sit out a cycle. :) [23:49] oh no, there's a list of required features by alpha 1 ;) [23:50] including linux 2.6.28 [23:50] But we can't stop it. [23:50] Didn't you see it? It forbids us from demoting it. [23:51] The thing is, there's no good way to close that sort of thing. [23:51] Perhaps Won't Implement. [23:51] Anyone have a second to help me with a problem I am having in one of the packaging guide tutorials? [23:51] But 'WTF - go away' is more appropriate. [23:51] pangloss: Ask away [23:51] Laney: in here https://wiki.ubuntu.com/PackagingGuide/Recipes/PackageUpdate === nhandler_ is now known as nhandler [23:52] laney @ step 4 [23:52] wgrant: it's just a misunderstanding that feature development & integration does actually require some work [23:52] Laney:I do not have a brasero-0.5.2/ folder [23:53] pangloss: What messages did you get after the dpkg -x line? [23:53] pangloss: Do 'dpkg-source -x brasero_0.5.2-0ubuntu1.dsc' [23:54] Laney: there is no dpkg line, but there is dget line [23:54] I was getting to that ;) [23:54] dget, yes [23:54] I guess it just failed to verify the key and then died extracting there [23:54] it should probably be dget -xu [23:54] Laney nhandler yes I never did a dplg [23:54] dpkg* [23:54] I meant dget :( [23:54] =( [23:54] so dpkg-source -x brasero_0.5.2-0ubuntu1.dsc [23:54] ? [23:55] yep [23:55] Yes pangloss [23:55] That should give you the brasero-0.5.2 directory [23:55] Thanks guys =) [23:55] You're welcome pangloss [23:57] there, I updated the options [23:57] Thanks for the pointer pangloss [23:58] Laney, updated the options? [23:59] on the wiki