[00:01] awe: yeah, i think so [00:01] ok [00:02] see you in dublin! === asac__ is now known as asac === nhandler_ is now known as nhandler [00:57] hmm [00:57] wonder what happened to that eclipse guy === tuantub_ is now known as tuantub === nhandler_ is now known as nhandler === heHATEme is now known as they [00:59] Packaging Perl Module session will be starting soon in #ubuntu-classroom [01:00] nhandler: who would want to waste time on that? [01:00] :P [01:02] someone who's not enlightened as to the wonders of python packaging & the associated fun? [01:03] <3 python packages [01:04] python is nice, but the mess surrounding packaging isn't so hot [01:04] partly due to how many different things have come through from upstream [01:16] * directhex runs dh_clilibs [01:17] there was an eclipse guy? === they is now known as vorian [01:23] vorian, someone who expressed an interest in helping eclipse packaging [01:24] i see [01:24] last log ftw [01:24] indeed === Pici` is now known as Pici [01:39] turning the eclipse package into a script that just downloads it from eclipse would be nice :D [01:39] s/nice/kludgy/ [02:03] This package REVU only needs one more advocation http://revu.ubuntuwire.com/p/xwax Anyone have a second? [02:08] stochastic, you seem to be specifying debhelper 7, but at a glance, i don't see anything in your file which isn't from dh5... am i failing to see something obvious? [02:58] directhex, I think I got some lintian warning about dh_prep that prompted the debhelper 7 req (IIRC), but I could try 5 and see... [03:02] directhex, yes, here's the Lintian warning I get with debhelper 5: xwax source: debhelper-script-needs-versioned-build-depends dh_prep (>= 7) === MobileMyles6o7 is now known as TwoToneSpirit === bdefreese2 is now known as bddebian [04:23] nhandler, would you please have a look at bug #389654 and give me some hints on how to correct the mentioned lintian error :) [04:23] Launchpad bug 389654 in lyricue "Please upgrade lyricue" [Wishlist,Confirmed] https://launchpad.net/bugs/389654 [04:26] porthose: Sure, let me take a look === WelshDragon is now known as Fluffles [04:38] porthose: I really haven't done much with translations/debconf translations, so in order to avoid guiding you incorrectly, I would suggest asking someone else. I am sorry [04:39] nhandler, np I'll post a message the the ml :) [04:40] s/the/to [04:40] porthose: I'll be watching to see the answer :) [05:30] I just noticed this wiki page: https://help.ubuntu.com/community/UnsignedGpgKey If someone gets a chance, it should probably be updated [05:36] anyone would like share the dput.cf for upload to ubuntu? [05:36] or is there any changes? [05:36] thanks [05:37] freeflying: It should be the standard one that came with dput. [05:38] ScottK: its strange, I uploaded twice, nothing happened, no email tell accept or reject [05:39] freeflying: Launchpad was down for maintenance earlier today. Maybe stuff is backed up. [05:40] ScottK: first try was done about 20 hrs ago [05:40] Oh. [05:40] I don't know then. [05:40] freeflying: I'd ask in #launchpad then. If anyone is around they can look it up. === cprov is now known as cprov-zzz [05:45] ScottK: thanks === santiago-pgsql is now known as santiago-ve === rgreening_ is now known as rgreening === apachelogger is now known as birthdaylogger === ejat is now known as e-jat === Quintasan_ is now known as Quintasan [09:49] slytherin: ping [10:11] ttx: pong [10:12] slytherin: about the jetty6 thing... It's not really a package rename. It's rather the idea of uploading the new major release under a separate source package [10:12] slytherin: since we expect to do a separate jetty7 in the future anyway [10:14] from an Ubuntu perspective, we need the jetty 6 libraries in karmic but the jetty 6 work in progress in debian isn't mature enough to be used... Having a jetty6 package solves this, we can upload our own and merge with debian when ready [10:14] ttx: What do you mean by not mature enough? [10:16] slytherin: it hasn't hit experimental yet. I identified a few upgrade issues, and I expect other issues when someone will start using it. And we need the jetty6 libraries in Ubuntu main. [10:18] and you think it will not be in good shape before FF? [10:18] slytherin: I doubt it. Not to main QA standards [10:19] so I'd upload our own simplified jetty6 thing, and ideally we would just merge the debian jetty6 in karmic+1 [10:19] but if it's called "jetty" the whole process is a lot more complicated. [10:21] ttx: hmm. How about let Debian devs upload it with name jetty (it's already in new), sync it in karmic and do the modifications as per our need. [10:22] slytherin: I'm slightly concerned in updating the "jetty" package with something that doesn't upgrade properly [10:23] also I see value in keeping jetty5 libraries available [10:23] slytherin: but yes, we could follow that route [10:24] it's just that I don't have so much time to dedicate to jetty package QA [10:24] slytherin: do you think you would ? [10:25] slytherin: i don't agree with the "it's not used at all anyway" approach from Marcus [10:26] slytherin: i don't know of the daemon, but the libraries are used by some Ubuntu-powered Java developers. [10:27] for them having jetty(5) and jetty6 available makes sense. [10:27] tough decision. [10:28] slytherin: and it's theirs rather than ours. Though you sit on both tables :) [10:28] Just wanted to make you understand why I was asking. [10:29] Thing is Debian Java has revived recently and there is lot of activity. So I want to play nice with them as well. It's people like Marcus, Matthew, Ludovic, Vincent that are driving force these days in D-J. [10:29] Note that the developers working on jetty wanted to call it jetty6 originally. They changed their mind on Marcus request. [10:31] In the end they (Ludovic and Torsten) are the ones who will do the work. [10:35] That is probably because Marcus mentioned that current jetty package is buggy and there is no phasing out. [10:42] will anything break if I just rename a git repository? [10:46] mv foo.git bar.git [10:55] slytherin, Hi! I've changed source package name to "cobertura" and corrected it as you suggested yesterday. Now the package is here: http://revu.ubuntuwire.com/p/cobertura. Could you, please, confirm that you are ok with the changes. Thanks in advance. === yofel_ is now known as yofel [12:44] gaspa: hello Andrea. No news regarding ocaml upload? [12:45] it's been done hasn't it? [12:45] david_mentre: ^ [12:51] Laney: Hello Laney. As far as I know, ocaml packages are still in sync-blacklist.txt and ocaml source package in Karmic is still 3.11.0-5 [12:52] the sync has been approved; all you can do is wait for it to be processed [12:53] I suggest that you reopen bug 387943 and ask for those changes to be undone too to avoid confusion [12:53] Launchpad bug 387943 in ocaml "Karmic: please do NOT synchronize following packages" [Undecided,Fix released] https://launchpad.net/bugs/387943 === cprov-zzz is now known as cprov [13:21] Laney: ok, I'm waiting for the sync to be processed. I reopened bug 387943. [13:21] Launchpad bug 387943 in ocaml "Karmic: please do NOT synchronize following packages" [Undecided,Fix released] https://launchpad.net/bugs/387943 [13:22] david_mentre: cool, nie one [13:22] nice === noodles775 is now known as noodles775-afk === vorian is now known as rofl === noodles775-afk is now known as noodles775 === rofl is now known as vorian === tuantub is now known as spam01 [15:06] juli_: I added a comment. If you fixed that within 2 hours. I will try building it. [15:06] ttx: still there? [15:06] slytherin: yes [15:07] ttx: whenever you get time, can you check swtcalendar and monajat on revu? I have already added advocations on those. [15:08] slytherin: I add it to my TODO list, eta unknown :) [15:08] thanks [15:16] Hi dyfet: I saw you updated the ubuntu-mobile-voice-user-interface regarding the packaging... Have you had a chance to try the updated packages? I could only try the executables in a chroot env. (so no audio etc.). [15:20] That is one thing I was going to do some testing. I simply noted that the package has been updated :) [15:20] slytherin, Thanks! Could you tell ms which license I should use for debian/*? GPL-2+? [15:21] I was also reformatting the blueprint's wb for progress tracking... [15:22] juli_: your choice. I usually keep it in sync with the license for application/library source. [15:23] slytherin, ok, thanks. [15:23] slytherin, will upload corrected version soon [15:24] http://loldebian.wordpress.com/2009/07/23/mono-rejected/ [15:25] highvoltage: what is that? [15:26] slytherin: a loldebian post. [15:32] dyfet: ah ok, yes, please let me know the results (or add them to the revu if you get a chance) [15:35] noodles775: I will add my results to revu [15:36] Thanks! [15:40] slytherin, done: http://revu.ubuntuwire.com/details.py?upid=6444. Thanks for helping me to make the package better! [15:43] juli_: I am in office currently. Expect reply within 3 hours. [15:44] slytherin, ok. I'm waiting for. [15:59] StevenK: ping [16:02] awe: Yes? [16:02] i finished a couple of merge that you previously uploaded, could you take a peek and ack 'em for me? [16:02] s/merge/merges/ [16:02] please? [16:03] awe: Not now? It's 1 am :-) [16:03] doh [16:04] guess i need to start using that world time widget mgz mentioned the other day! [16:04] i'll catch you later ( much ) [16:04] Heya gang [16:06] awe: have you already filed bugs and added debdiffs there? [16:07] slytherin: yes...i also meant syncs, not merges... [16:08] awe: make sure you have subscribed appropriate sponsors teams. So if StevenK does not have time, someone else will take a look. [16:09] i used requestsync, so they should be all set. i was pinging steven cause he was the last uploader, and i figured i could buy him a guinness or three next time i see him for expedited service. ;) [16:10] slytherin: no rush. thanks! [16:10] awe: Send me a mail, and I'm happy to look when I start work [16:10] StevenK: cool [16:33] fta, gwibber appears to be nice from the PPA. [16:34] good [16:34] is it approximately what's in karmic? [16:35] yes [16:40] will will do a few contribution merges and then cut off 1.2 branch; then trunk will get a big bomb ;) [16:42] fta, hey, --enable-plugins almost sorts works ish! [16:43] (on chrome) [16:43] directhex, unless you depend on gmail [16:43] and s/chrome/chromium/g [16:43] fta, i don't. i do, however, have the silverlight 1.0 version of bubblemark displaying. albeit very jerkily === noodles775 is now known as noodles775-afk === ApOgEE__ is now known as ApOgEE [18:54] Hi. Any MOTU willing to sponsor the merge of gmerlin-avdecoder (Bug #283213)? It's a mandatory step to update openmovieeditor to latest version. [18:54] Launchpad bug 283213 in openmovieeditor "Please merge gmerlin-avdecoder 1.0.0-0.2 from Debian Multimedia (Stable)" [Wishlist,Confirmed] https://launchpad.net/bugs/283213 [18:55] siretart, I saw you made some comments on Debian's mailing list on gmerlin-avdecoder. Do you have more info about that? === micahg1 is now known as micahg [19:05] <_andre> hyperair: hi :) did you have some time to review those new packages? [20:00] _andre: eh what? [20:02] <_andre> hyperair: http://revu.ubuntuwire.com/p/watchcatd === cprov is now known as cprov-afk [20:15] i'll go look at it now [20:15] _andre: you should change the Maintainer to Ubuntu Developers , if i'm not mistaken. [20:17] _andre: priority should be optional, rather than extra. [20:18] don't add COPYRIGHT to debian/docs [20:22] _andre: get rid of debian/dirs [20:23] _andre: and use a .install file to specify which files go into which package [20:23] man dh_install for more info [20:24] for installing the upstream ChangeLog, use override_dh_installchangelogs, and in that rule call dh_installchangelogs ChangeLog. man dh_installchangelogs for more info [20:24] also drop the dh_install call, it'll be done automatically. [20:29] <_andre> hyperair: thanks, i'll do that [20:30] <_andre> i think i saw "Maintainer: Ubuntu MOTU Developers " in an example at the wiki, though [20:30] _andre: also, rather than duplicating all the commands in the Makefile's install rule in debian/rules, why don't you add DESTDIR support to your Makefile? you're the upstream, aren't you? [20:33] <_andre> i wanted to avoid making a new release just for that... [20:34] ah [20:34] then how about adding a patch system? [20:34] and patching the Makefile? [20:34] <_andre> because it's more complicated than just installing the files :) [20:35] hmm tru [20:35] e [20:35] then use dh_install. =) [20:35] it's much simpler [20:37] <_andre> is it just "dh_install file destdir"? [20:37] no [20:37] just list your files in debian/watchcatd.install [20:37] see the manpage [20:38] <_andre> hmm [20:38] what's dh_install -s? [20:38] i haven't seen the -s option before [20:39] <_andre> it was there from the dh_make template [20:39] <_andre> i just left it [20:40] ah. [20:40] toss it [20:40] dh_make is obsolete! =p [20:41] <_andre> i noticed :P [20:41] dh_make is obsolete? [20:41] well not really [20:42] it's just that the rules file produced is old [20:42] not that it can't be used, just that it's longer and more cluttered [20:42] oh, that you mean [20:43] <_andre> hyperair: so if i use a .install file i can remove the override_dh_auto_install target? [20:44] yes. [20:45] also remove your debian/dirs file [20:45] <_andre> ok [20:46] another thing i noticed... your preinst and postrm scripts appear to create a user and a group [20:46] what's the group used for? [20:47] <_andre> well, it's just an unprivileged user, i think it makes sense for it to have its own user and group like most daemons that drop privileges have [20:49] <_andre> so if there's a bug it won't have any access it shouldn't have because it shared some other group [20:51] i don't think it requires its own group, does it? =\ [20:52] i mean, do users need to be in a group? [20:52] no right? [20:52] _andre: https://wiki.ubuntu.com/DebianMaintainerField this will resolve your doubts on the Maintainer field in debian/control which now is Ubuntu Developers [20:53] <_andre> hyperair: a user has to be in at least one group... [20:53] it can be in nogroup [20:54] actually nevermind. don't mind me [20:54] i'm not sure about this [20:54] <_andre> it's the same reason people don't use nobody as the daemon user [20:55] i see [20:55] <_andre> can a .install file handle renames? [20:56] <_andre> the upstream file is called watchcatd.prod.conf but it should be installed as watchcatd.conf [20:58] "dh_install cannot rename files or directories" (from the dh_install manpage) [20:58] <_andre> oh [20:59] <_andre> so i guess i'll still need the override [21:00] <_andre> does that mean i need to keel the dirs file with "etc" in it? [21:01] you can still rename it after you called dh_install [21:04] <_andre> where do i do that? [21:05] just add the necessary mv calls to debian/rules [21:06] <_andre> isn't it better to just install them with the correct name in the first place? [21:07] or that [21:09] <_andre> testing the changes with pbuilder now... [21:09] there is no one "true" way to do it, you've several options how you solve it === ripps_ is now known as ripps [21:24] <_andre> hyperair: new package uploaded [22:06] <_andre> is there a dh command that automatically makes the shared library symlinks? [22:10] _andre: no there isn't. the upstream build system should have handled that. [22:10] _andre: you *really* should use autotools [22:10] and possibly combine the packages into one source package [22:11] anyway, is there a reason /etc/watchcatd.conf is to be 0600 rather than 0644? [22:11] <_andre> just paranoia i guess [22:12] hmm [22:12] does watchcatd run as root? [22:12] i remember you created a user right? [22:12] which means that watchcatd should run as that user? [22:12] <_andre> the master process does, the slave runs as user watchcat [22:13] aah i see. [22:13] slave? O_o [22:13] what does the master do? [22:13] and what does the slave do? [22:13] also, is there a reason you duplicated this line: install -m0600 watchcatd.prod.conf $(BUILD_DIR)/etc/watchcatd.conf ? [22:14] <_andre> the slave talks to the client programs who want to be monitored, and when one of those becomes unresponsive, the master kills it [22:14] oh. i see. [22:14] sounds like a rather roundabout way. [22:14] <_andre> oops [22:14] <_andre> no [22:14] ? [22:14] <_andre> shouldn't be duplicated [22:14] ah [22:15] also, you can combine your install -d line with your install -m600 line [22:15] install -D -m0600 bla /etc/bla [22:15] i mean $(BUILD_DIR)/etc/bla [22:15] <_andre> cool, didn't know that [22:16] <_andre> hyperair: the idea is to have as little code running as root as possible [22:20] i see. [22:20] _andre: what's to stop me from sending fake kill requests to the server? [22:26] <_andre> hyperair: the client talks to the server via a unix socket, and it can only request its own pid to be killed [22:27] <_andre> this can be asserted with getsockopt on linux and with a more complicated procedure on freebsd [22:29] _andre: that's between the server and the slave. what about directly sending the request to the master? [22:32] <_andre> hyperair: the master only talks to the slave, it doesn't accept direct connections from other processes [22:32] ah. i se [22:33] <_andre> this is very useful for example in a shared hosting environment [22:33] mmhmm [22:33] i see [22:33] <_andre> you can monitor the apache processes [22:33] <_andre> and bad code by your customers won't kill the server :) [22:34] sounds good =) [22:34] <_andre> but you need the mod_watchcat package too :P [22:34] mmhmm of course [22:35] <_andre> i'll upload watchcatd again without that duplicate line [22:36] chrisccoulson: were you planning on doing bug 403721 [22:36] Launchpad bug 403721 in funkload "Packaged version too old, upstream even warns against it" [Wishlist,Triaged] https://launchpad.net/bugs/403721 [22:42] nellery - i didn't intend to work on it straight away. did you want to take it? [22:42] chrisccoulson: no that's alright, you can take it [22:50] _andre: W: watchcatd: manpage-has-errors-from-man usr/share/man/man5/watchcatd.conf.5.gz 122: warning: `SE' not defined === ian_brasil is now known as ian_brasil_afk === asac_ is now known as asac