[01:13] Anybody free to look at http://revu.ubuntuwire.com/details.py?package=cvc3 ? [01:20] postalchris1: You've got a typo in debian/copyright. There's an incorrect GPL reference at the end. [01:21] postalchris1: I haven't managed to do a full re-review, but I still disagree with the weird version clause that I comment on [01:21] *commented [01:21] I don't think it accomplishes what it claims to [01:21] maxb: That's is possible [01:23] maxb: I failed to Google the source of the, because "(<< ${binary:Version}.1~)" looks like it has search operators in it, and I don't know how to escape them. [01:24] postalchris1: Are you in contact with upstream for this package? [01:24] ScottK: Good catch. Looking up the proper boilerplate now... [01:24] ScottK: Upstream is my academic advisor. So yes. [01:25] postalchris1: The license is technically Free by our terms, but I would not recommend this go in the Ubuntu archive with the current license. Others may have a different view. [01:25] The requirement to get permission to modify or rename is technically acceptable, but very difficult to deal with. [01:26] Consider what happens if someone discovers a security issue and he is on vacation? [01:26] The packaging of a CVS snapshot from a CVS repository that is not publically accessible also is.... remarkable [01:26] postalchris1: I would ask you to encourage him to use a less restrictive license. [01:27] That's another interesting issue. [01:27] It's not a legal problem though. [01:28] ScottK: We're working on relicensing, but who knows when it will go through... [01:28] With the current license I don't think the package is maintainable. [01:29] maxb: You know what? I'm dumb. We publish daily snapshots at http://www.cs.nyu.edu/~barrett/cvc3-latest/ [01:29] correction: With the current license and name it's not maintainable. [01:33] postalchris1: Basically if you're Firefox you can, just barely, get away with must have permission to modify without renaming as it's technically DFSG Free, but it's sufficiently onerous that I don't think we want to deal with it. [01:36] ScottK: That makes sense. It's unfortunate. We have the intention to relicense *and* the firm assurance of all parties involved that Clause 3 wouldn't be invoked in the case of routine maintenance. [01:38] postalchris1: Could they publish some kind of addendum or exception and add it to the source tree? [01:38] ScottK: Not sure what you mean. [01:39] If there was something in the upstream source that said 'normal maintenance is ok', perhaps we could call that good enough. [01:40] ScottK: I'm sure I could swing that. Seems pretty weak, legally, though. [01:41] postalchris1: My thought is if that 'firm assurance' could be reduced to writing in the tarball, then maybe it's good enough. [01:41] The thing is it's DFSG Free and legal for the archive as it is, just very uncomfortable. [01:41] So I'm trying to reduce that uncomfort. [01:44] ScottK: Cool. I'll look into it. === d-b_ is now known as d-b [07:13] good morning [07:17] Morning dholbach. [07:18] hi iulian [07:39] hi guys [07:49] good morning o/ [07:49] hey di [07:49] didrocks :) [07:49] Hi dholbach :) [08:49] good morning [08:52] morning geser [09:17] would a kde3->kde4 port be considered bugfix or new feature (in terms of FF)? [09:17] package is minirok [09:28] I'd consider it as both: bugfix and new feature (but I'm not from ~motu-release so you can ignore me :) [09:53] I have a question about encoding [09:54] I've got a tar.gz with every file in it in iso-8859-15 encoding ... With french accent éèàâ, etc [09:55] does it cause a problem ? May I "patch" the source to get everything in utf8 with iconv ? [09:55] it shouldn't cause a problem, but you could still talk upstream into converting to utf8 [09:56] okay [09:56] because i have added some manpages totally in utf8... the fact that some files are in utf8 and others in iso-8859-15 doesn't cause any problem too ? [09:57] as long as your editor can deal with it, it should be fine [10:38] liw: btw, is there a utility to give encoding of a text file ? [10:39] try file [10:43] dolanor, not fully reliably; moreutils has isutf8 to see if it is syntactically valid utf8, file makes some guesses, but on the whole, you have to look at the file in various encodings and decide which is most likely [10:44] ok [10:45] I just had to clean up a major mess caused by schroot's use of logical lvm volumes. Anyone else has experiences with that? [10:55] hi mok0, I've followed your suggestions -> http://revu.ubuntuwire.com/p/mandvd [10:57] I've uploaded the package now [11:09] quadrispro: good [11:11] quadrispro: I need to test-drive my sbuilder after messing with the lvm's, might as well use mandvd as an example [11:12] ok ;) [11:12] yay it works [11:13] yeah :) [11:16] mok0: If you want jeuclid in Ubuntu you better upload it yourself - it might not get into Debian in time. (I think it was you that wanted it) [11:16] Laney: we still have a week :-P [11:16] hmm [11:16] I dunno, it's cutting it fine [11:16] Too close for comfort, huh? [11:16] but it might start moving after lenny releases [11:17] up to you [11:17] k [11:21] mok0: I've got the package that's #2 in the Debian New queue among the packages not obviously on long term hold and it's been there for at least a couple of weeks. With a release is two days, I seriously doubt New is going to be on ftp-masters TODO before our FF. [11:21] ScottK, you're probably right [11:21] I'll take a look at jeuclid then. Laney, have you tried building it? [11:22] Worst case is there's an additional sync, which is not a big deal. [11:22] mok0: No, I'm not interested in it. I just thought you were so I let you know [11:22] ScottK, can we ask the archive admins to sync from the NEW queue? [11:22] no, it's opaque [11:22] No. New isn't publically accessible. [11:22] You can't get at the files [11:22] Huh? [11:23] Their theory is they don't know if it's distributable until after it's reviewed. [11:23] So they want to make sure it isn't being distributed. [11:23] d'Oh [11:23] uhm, I seem to remember it was in a ppa somewhere [11:24] It's maintained by debian science team, so they probably have it in svn [11:24] Generally one can fish the packaging out of a VCS somewhere or worst case mail the maintainer and construct and equivalent package. [11:24] mok0 probably even knows where that one is. [11:24] So, I should probably send it to revu [11:24] ;-) [11:24] Yes I think so [11:25] If you upload it in your name it's not strictly required. [11:25] Last I checked anyway .... [11:25] mok0: Do your own review, and take the DD uploading as the other one [11:25] Koon: did you pickup libjcip-annotations-java directly frmo new queue? [11:25] unless you want to seek further input [11:26] Laney: It's java right? [11:26] I don't know much about java [11:26] yeah [11:27] Laney, RAOF went into hiding without sorting sublib :( [11:27] psh [11:27] wait until tomorrow (hopefully) [11:28] or else sebner should be around [11:30] mok0: if you would see the buildlog -> http://home.alessiotreglia.com/jaunty/result/mandvd_2.5-4-0ubuntu1/mandvd_2.5-4-0ubuntu1.buildlog [11:30] quadrispro: thx [11:32] quadrispro: you should document in debian/changelog that you have authored a manpage [11:32] I know there's some confusion about this so I write it here [11:33] mok0: ah ok, I'm workin on it now [11:33] Also .desktop file [11:33] Anything that is user visible in the package, that you add, should be documented in debian/changelog [11:34] ok [11:34] so, manpage and .desktop files... [11:34] ok [11:34] quadrispro: :-) yes [11:35] I am rehashing it in the channel in the hope that the misunderstanding can go away [11:36] It'a policy §4.4 [11:36] mok0: upstream provides a .desktop file, but it needs to be fixed. What solution could be better? Add a new .desktop or fix that shipped with the original tarball? [11:37] quadrispro: err, well, since there's no patch system used in the package, it's perhaps a bit much just to fix a .desktop entry [11:38] quadrispro: When the new source package format comes into use, there's an implicit patch system, that will help a lot [11:39] mok0: yeah, I think it's not good introduce a patch system only for a .desktop file, then I will fix mandvd.desktop [11:40] quadrispro: ok, perhaps send the corrected file upstream [11:40] of course [11:41] ScottK, I'll make you a sweet deal: if you review some of my advocated packages in revu, I'll review/upload the plasma_widget ones B-) [11:42] * ScottK looks [11:43] ScottK, there are two libraries that have dependents stuck in the pipeline [11:43] mok0: OK, but it may be Saturday before I have time. [11:43] ScottK, NP [11:44] Has anyone looked at my mock-up revu site? [11:46] mok0: new package uploaded to REVU [11:46] quadrispro: super [11:46] :) [11:53] mok0: don't hate me :) http://revu.ubuntuwire.com/p/nfoview [11:53] * mok0 is confused... is it REVU day :-P [11:54] Should start in about 6 minutes, right? [11:54] Heh [11:54] yes [11:56] mok0: that package is very simple, and I think it's ready really [11:56] :) [11:56] * mok0 is looking [11:57] Hm, never heard of nfo format before now [11:58] quadrispro: tsk tsk not lintian clean [11:59] * quadrispro nooooo [11:59] it's true :( [11:59] heh [11:59] working on it [12:00] don't upload yet there might be more [12:00] mok0: all the fans of emule/bittorent services know well .NFO files :) [12:00] mok0: ok [12:01] Ah, perhaps my son knows about it [12:01] maybe :) [12:01] mok0: http://en.wikipedia.org/wiki/.nfo [12:02] quadrispro: I'm gonna use that for my next personal homepage :-) [12:03] quadrispro: the "retro" look [12:03] yeah, it's very "cool" :D [12:06] * mok0 wonders why the .desktop "Encoding" statement is deprecated... [12:06] Seems rather sensible /me thinks [12:07] quadrispro: for you next package, use this: http://wiki.debian.org/Proposals/CopyrightFormat [12:08] ah good [12:08] quadrispro: Take a look at http://packages.debian.org/changelogs/pool/main/g/git-cola/current/copyright for an example. [12:11] mok0: I could try to apply now that format [12:12] * template [12:12] quadrispro: just curious, why didn't you choose a cdbs rules file? [12:12] quadrispro: great [12:12] mok0: mmm I don't know why, I prefer to use setup.py... [12:13] quadrispro: so does cdbs [12:13] in fact [12:13] ok, i'll use cdbs and apply that template :) [12:14] quadrispro: http://pastebin.com/f64a2d3ca [12:14] ok [12:15] quadrispro: ... and you don't need to tussle with reviewers about debian/rules :-) [12:15] lol :) [12:15] ok [12:16] quadrispro: your rules file was fine though, no problems [12:21] mok0: http://paste.ubuntu.com/117226/ [12:22] quadrispro: yes, but remember the format is like debian/control. You need " ." to specify a line-break [12:22] quadrispro: (at lines 13 and 18) [12:23] quadrispro: that makes it RFC822 format, and it can be read by the standard tools [12:23] ah ok [12:23] quadrispro: so, perhaps someone will write a nice "copyright-viewer" [12:24] mok0: http://paste.ubuntu.com/117228/ [12:25] quadrispro: TBH, I am unsure if it is correct [12:25] quadrispro: AFAIC, the text will only be wrapped correctly if it starts in column 2 [12:26] you're right [12:26] quadrispro: ... and if it is indented 2 spaces or more, it will not be wrapped [12:26] I am? [12:26] :) [12:26] eh [12:26] :D [12:26] I'm doing a lot things at the same time [12:27] OK, I am going to lunch, see you later [12:27] ;) bye [12:28] slytherin: I had to merge it to fix a default-jdk/gcj mismatch. It's published now. libjboss-cache2-java can now build too. [12:39] mok0: I've uploaded nfoview to REVU, I didn't change copyright/rules because a lot of lintian warnings [12:39] see you later, I'm going to lunch [12:45] Koon: I had seen that problem. I was planning to log bug in Debian. [12:45] slytherin: I'll push it back there. Though I'm not sure they are affected by it [12:46] Koon: They are not. But the problem in my opinion is that when you specify default-jdk as build dep you should use default-java as java_home [12:46] slytherin: definitely. [12:47] note that for libhibernate3-java I'll fix it the other way around. Doesn't build with openJDK so it requires gcj ;) [12:47] slytherin: bug 328391 if you want to follow that story [12:47] Launchpad bug 328391 in libhibernate3-java "Please merge libhibernate3-java 3.3.1.GA+dak1-3 (universe) from Debian unstable" [Wishlist,In progress] https://launchpad.net/bugs/328391 [12:47] Koon: right. [12:48] i'll push that one as soon as libjboss-cache2-java clears binary NEW [12:49] Koon: meanwhile if you are not too busy, can you please file movetouniverse bugs for remaining jboss/hibernate related packages? [12:50] slytherin: will do if I can enjoy some free time. Unfortunately it's been 10 days that I rush, not sure I can do that before FF [12:51] Koon: Ok. If you don't do it by weekend I will take care of it. [12:56] slytherin, Koon: for moving libhibernate3-java to universe see also my last comment in bug 252318 [12:56] Launchpad bug 252318 in ehcache "Please move to multiverse" [Medium,New] https://launchpad.net/bugs/252318 [12:56] (yes, the bug should be retitled) [12:56] geser: and also we should add 'also affects' for oscache, libhibernate3-java [12:57] geser: interesting. I missed that. Thx ! [13:06] I've reuploaded new version of hexdiff, a tool to visualize the hexadecimal differences between 2 files (http://revu.ubuntuwire.com/details.py?package=hexdiff). If it could get some REVU before the 19th :p [13:18] quadrispro: lazy huh? ;-) [13:22] quadrispro: I don't like the way you are fixing the .desktop entry [13:27] mok0: I've uploaded new version (using cdbs) [13:28] quadrispro: you need to put some dependencies in control, did you do that? [13:28] http://home.alessiotreglia.com/jaunty/result/nfoview_1.2.1-0ubuntu1/ [13:28] mmm [13:29] it needs python-gtk2 and python-glade2 [13:29] binary-install/nfoview:: [13:29] dh_desktop -pnfoview [13:29] quadrispro: huh? [13:29] you need to depend on cdbs and python [13:30] mok0: eh, the package build-depends on cdbs and python... [13:30] mok0: dh_desktop is necessary, otherwise it doesn't update the MIME cache [13:31] quadrispro: right, so you can add a rule binary-install/nfoview:: that calls it [13:32] quadrispro: http://pastebin.com/f7e14efad [13:32] done [13:32] dputting to revu [13:32] quadrispro: and in copyright, after the GPL license text, add the following: [13:33] X-Comment: On Debian GNU/Linux systems, the complete text of the GNU [13:33] General Public License can be found in /usr/share/common-licenses/GPL-3 [13:33] oops [13:33] 2 lines [13:33] there's "On Debian systems, the complete text of the GNU General\nPublic License can be found in `/usr/share/common-licenses/GPL-3'. [13:33] isn't it right? [13:33] yes. there is? [13:34] yes [13:34] mok0: http://revu.ubuntuwire.com/revu1-incoming/nfoview-0902121433/nfoview-1.2.1/debian/rules [13:34] and http://revu.ubuntuwire.com/revu1-incoming/nfoview-0902121433/nfoview-1.2.1/debian/copyright [13:34] quadrispro: I was talking about the other copyright file you made [13:35] ah ok ok [13:35] http://pastebin.com/f3b4f627 [13:35] mok0: but I've tried to make the copyright file according with that new template, but lintian gaves me a lot of warnings :( [13:35] quadrispro: ^ [13:35] gave * [13:36] quadrispro: I just tried building with the above file, it gives no warnings [13:36] oh ok :) thanks [13:36] quadrispro: I was wondering about your message before lunch [13:36] quadrispro: so I tried it [13:37] quadrispro: The call to dh_desktop is there automatically if you include the gnome cdbs class. [13:37] ah good [13:37] soren: yeah but we're not building a gnome app, so why add the builddepens? [13:38] Ah, sorry, I thought it was a gnome-ish thing. What's it for, then? [13:38] s/ns/nds [13:38] soren: just a python program [13:39] soren: with a python-gtk2 gui [13:39] mok0: the gnome class is in cdbs itself. [13:39] soren: ah, ok [13:39] :) [13:39] soren: I stand corrected and humble [13:40] :) [13:42] I've started to build with calling gnome class [13:42] mok0: uploading to REVU again [13:43] mok0: done [13:43] soren: thanks! [13:43] sure [13:43] quadrispro: is it there already? [13:44] eh, no [13:45] mok0: REVU's showing it [13:45] quadrispro: k [13:49] quadrispro: FAILED [dpkg-buildpackage died] [13:49] =-O [13:49] http://home.alessiotreglia.com/jaunty/result/nfoview_1.2.1-0ubuntu1/nfoview_1.2.1-0ubuntu1.buildlog [13:50] quadrispro: can't explain that, but the gnome think pulls in some autoconf checks that I think you don't want [13:50] s/think/thing [13:51] quadrispro: let me go back to your penultimate upload [13:52] ok [13:52] sure [13:53] quadrispro: your upload from 12 Feb 2009 14:39 builds perfectly [13:54] mok0: ok, is it necessary anohter upload? [13:54] quadrispro: unfortunately, yes [13:54] quadrispro: you should not have listened to soren ;-) [13:55] preparing another upload.... uff! :) [13:55] mok0: build log? [13:55] soren: http://home.alessiotreglia.com/jaunty/result/nfoview_1.2.1-0ubuntu1/nfoview_1.2.1-0ubuntu1.buildlog [13:55] soren: I can pastebin some of it [13:55] ...from the failed build. [13:56] ah sorry :D [13:57] soren http://pastebin.com/f729544d [13:57] soren: it wants to do some autotools fun [13:57] where's the configure script? [13:57] soren: there isn't any [13:58] Ah. [13:58] soren: It probably assumes there is one because it thinks its a gnome project [13:59] soren: in any case, the extra binary-install:: rule works perfectly :-) [13:59] Right. [14:00] In fact the dh_desktop thing is the subject of a flamewar on a bug in BTS [14:01] 439717 [14:01] ubottu, can you please give me a BTS link? [14:01] Error: I am only a bot, please don't think I'm intelligent :) [14:02] ubottu: oh, I forgot [14:02] Sorry, I don't know anything about oh, I forgot [14:02] mok0: uploaded [14:05] quadrispro: +1 [14:05] thank you mok0 [14:05] I've reuploaded new version of hexdiff, a tool to visualize the hexadecimal differences between 2 files (http://revu.ubuntuwire.com/details.py?package=hexdiff). If it could get some REVU before the 19th :p [14:05] quadrispro: now go fix some bugs [14:06] yes :D [14:06] No more new package from you ;-) [14:07] ok ok. I will work hard only on bugufixes :) [14:07] hehe [14:08] dolanor: I'll give it to you straight: there is no COPYING file giving us the permission to distribute the source tree. Without it, we can't distribute :-( [14:08] dolanor: :'( rather [14:10] dolanor: can you contact upstream? [14:15] mok0: i'm in direct contact with him ^^ [14:15] dolanor: cool [14:15] mok0: I'll do the patch myself and give it to him [14:16] dolanor: errrr, ok, but you can't patch a license file into Ubunut [14:16] Ubuntu [14:16] mok0: He hasn't time to reupload the new version, but when I'll harrass him to upload the new version on his site, I think he can take time for this :) [14:16] dolanor: good [14:17] no no, I take the orig.tar.gz, increment the version, add a COPYING, and base my ubuntu package on it :) [14:17] and mail the new tar.gz to upstream [14:17] I always ask him first [14:17] dolanor: sounds reasonable :-) [14:18] so a COPYING file is missing in the root oif the archive ? [14:18] is there any COPYING sample for a really simple source package ? :) [14:28] dolanor: if you look in the file fileinfo.c, there's a URL to the license text [14:29] http://sam.zoy.org/projects/COPYING.WTFPL [14:29] dolanor: ... so you need to change debian/copyright accordingly [14:30] I guess WTFPL is DFSG compliant [14:50] Hi RainCT, how did your exam go? [14:52] mok0: do I put COPYING into the debian/doc ? [14:52] dolanor: no, it will be installed automatically when dh_installdocs is called [14:53] ok [14:53] dolanor: as long as it's called COPYING [14:57] mok0: and, for the debian/copyright, Do i have to put that the packaging is under GPL ? [14:57] probably yes [14:57] dolanor: it's up to you [14:58] dolanor: either that or the same license as the software [15:00] mok0: So I must precise it in the copyright : upstream is wtfpl and packaging is gpl ? [15:00] with copy of each license in it [15:01] dolanor: yes, if you use the new copyright format, you just need a section for Files: debian/* with your license details === bdefreese is now known as bddebian [15:06] Heya gang [15:10] mok0: didn't find the new copyright format, do you have any links ? [15:10] !copyright [15:10] Sorry, I don't know anything about copyright [15:11] http://wiki.debian.org/Proposals/CopyrightFormat === _neversfelde is now known as neversfelde [15:14] thank you, modifying the stuff :) I didn't see your remarks on revu :o [15:14] dolanor: ah :-) [15:15] dolanor: I always put them there even when doing IRC reviews [15:15] slomo: is there any plan to sync/merge libdvdread/libdvdnav from Debian experimental? [15:18] dolanor: you don't have fixes for all of my comments === RoAk is now known as RoAkSoAx [15:22] does anyone know what speeds up the boot time, is it building most of the things as modules or building everything 'build-in'? [15:22] slytherin: I don't think it makes a lot of difference [15:23] slytherin: most of the boot time is starting services [15:23] slytherin: ... and the fact they are started serially [15:24] mok0: I have built a custom kernel. And now I plan to tweak it to remove unnecessary parts. [15:24] slytherin: I've tried that several times. Never could get it to work right :-P [15:25] slytherin: now I just use ubuntus generic kernel [15:27] mok0: what problems did you face? [15:27] slytherin: ah, can't remember, it was several years ago [15:27] slytherin: kernel panic was one of them [15:29] jdong, siretart wgrant_ can one of you merge lp:~andreas-wenning/vlc/ubuntu into ~motumedia/vlc/ubuntu? I was going to upload andreas merge (builds fine and solves the bug), but then i realized i cant merge into ~motumedia (What's that team for, shouldn't ubuntu-dev own it?) [15:34] slytherin: last time I did my own kernels (all needed modules built-in), I had to either still use an initramfs (because of the UUIDs) or use the /dev/sdXY in the grub menu [15:35] geser: I have just started to play with it, so can't really say I understand all the things involved. Anyway, will do the experimenting at home now. [15:38] dtchen, too about vlc^ [15:38] slytherin: I used to build my own kernels in the past. The only real benefit was a smaller kernel deb. And you need keep track of kernel updates yourself. [15:40] mok0: I'm not sure about which keys to press to get a minus instead of hyphen ... [15:40] dolanor: just press - on your keyboard, but prefix it with backslash in the file, like so: \-v [15:41] dolanor: it's because in nroff, - means hyphen and \- means minus [15:42] dolanor: when typeset on the screen, you can't tell the difference, but when rendered for a printing device, you can [15:43] mok0: ok, thanks :) [15:47] does feature-freeze apply to packages in the advocation queue? [15:47] It has to be uploaded to Ubuntu by Feature Freeze. [16:04] Where can I find the script that does the merging on DaD? [16:06] mok0: the Copyright review seems very cool :) I hope it will facilitate the revuing :) [16:07] mok0: http://dad.dunnewind.net/grab-merge.sh? [16:07] iulian: errhm not that, the script that generates what grab-merge fetches [16:07] mok0: launchpad.net/dad [16:07] Ah [16:07] Laney: hah, but of course! Thanks! [16:08] * mok0 kicks self [16:08] looks like a massive bash script [16:08] * Laney dies /o\ === thekorn_ is now known as thekorn [16:20] Would a MOTU mind reviewing http://revu.ubuntuwire.com/p/pyproj? It's a python wrapper for PROJ.4 (a map projections library). I've already corrected some initial problems found by fabrice_sp__. === santiago-pgsql is now known as santiago-ve [17:01] superm1: you're right, that branch should probably be in ~ubuntu-dev. feel free to push the merge there! [17:03] chrismurf: pyproj reviewed... [17:03] mok0, thank you [17:03] chrismurf: you're welcome :-) [17:04] the debian/watch I basically copied [17:04] Is there actual documentation on that someplace? [17:04] a manpage or anything? [17:04] chrismurf: intersting... [17:04] google is frustratingly useless [17:04] man uscan [17:04] ah - thank you [17:04] chrismurf: oh, yes, google for watch :-) [17:04] ah! wonderful [17:05] chrismurf: the man page has bad practice in the examples, though [17:05] lol - fantastic [17:05] chrismurf: it mentions (.*) which can give problems [17:05] ok [17:05] chrismurf: it's better to do what you are doing [17:06] chrismurf: then you make sure it is fetching a revision number like 1.2.3 [17:07] makes sense [17:08]