| imbrandon | LaserJock: its per user in /var/spool/mail/user iirc | 00:10 |
|---|---|---|
| LaserJock | even if /usr/sbin/sendmail exists? | 00:11 |
| imbrandon | no, you ask where it was sent too | 00:12 |
| imbrandon | s/too/to | 00:12 |
| imbrandon | e.g. how does it know, it knows by user@system and delivers to /var/spool/mail | 00:12 |
| StevenK | Sendmail (or variant) is what writes to /var/mail/<user> ... | 00:12 |
| LaserJock | right | 00:12 |
| imbrandon | err yea | 00:12 |
| LaserJock | so I think I need to set up msmtp to do that :/ | 00:13 |
| imbrandon | LaserJock: why not just install postfix? | 00:13 |
| LaserJock | bah, maybe I should just install exim4/postfix | 00:13 |
| imbrandon | if you want local only, there is very little if any config for postfix | 00:13 |
| imbrandon | on install | 00:14 |
| LaserJock | well, I want system mail local, but I want to be able to send email out via SMTP | 00:14 |
| LaserJock | I was using msmtp for sending out email to the smarthost | 00:14 |
| imbrandon | smtp from your local box? most people will make that as spam imediately | 00:14 |
| imbrandon | ahh ok | 00:14 |
| LaserJock | but I think that meant that system mail got dropped | 00:15 |
| imbrandon | smarthost, yea you can config postfix for a smarthost on install via debconf questions | 00:15 |
| imbrandon | fairly simply | 00:15 |
| LaserJock | yeah, I was just trying to go "lightweight" ;-) | 00:15 |
| LaserJock | although I was just messing around with exim4 and I couldn't get it to work with my smarthost | 00:16 |
| LaserJock | I'll see if postfix does any better | 00:17 |
| imbrandon | :) | 00:17 |
| StevenK | LaserJock: ssmtpd | 00:20 |
| LaserJock | oh for goodness sakes, postfix is messy | 00:28 |
| Fujitsu | LaserJock: Is it? | 00:31 |
| StevenK | I don't think it is. | 00:31 |
| Fujitsu | For setting up local with smarthost, debconf should do it all. | 00:31 |
| Fujitsu | But even so, I find postfix' configuration to be not at all bad. | 00:31 |
| LaserJock | my smarthost uses SSL authentication | 00:31 |
| Fujitsu | Ahahaha. | 00:31 |
| LaserJock | both exim4 and postfix seem to choke when I get to SSl | 00:32 |
| LaserJock | I'm just gonna set up a system-wide msmtp config and see if that works | 00:33 |
| === bperry is now known as brandonperry | ||
| === brandonperry is now known as bperry | ||
| phomes | is there a reason for gnome-games being stuck at 2.20.1 for gutsy? (and 2.20.0.1 for hardy btw) We are getting a ton of crasher reports from bugs in these old versions | 01:07 |
| Burgundavia | phomes: we usually don't update, but talk with seb128 | 01:09 |
| slangasek | well, practically speaking for hardy things are stuck at 2.20.0.1 because 2.21.4 pulls in a new build-dependency which hasn't gone through the main inclusion process yet | 01:13 |
| phomes | Burgundavia: okay, thanks | 01:14 |
| slangasek | oh, correction; the problem for hardy now seems to be that the last build attempt happened right while LP was being ornery, and installing all build-deps failed | 01:16 |
| phomes | slangasek: ok. No chance to get in alpha 3 then right? Too bad. I get the impression most testing of unstable versions come from people trying the alpha/betas of ubuntu | 01:18 |
| slangasek | phomes: not much chance of getting it in alpha 3; depends on me snagging someone who can requeue it | 01:19 |
| slangasek | Hobbsee: ping :) | 01:20 |
| Burgundavia | slangasek: what needs to be done with the hardy alpha 3 page? | 01:24 |
| Burgundavia | what new features need to be added? | 01:24 |
| ScottK | For Alpha release notes is it appropriate to mention new stuff that needs testing. | 01:26 |
| ScottK | ? | 01:26 |
| slangasek | Burgundavia: good question. the new OOo might be worth a mention, if that sneaks under the wire | 01:27 |
| slangasek | Burgundavia: otherwise, it's anything that looks good from -changes :) | 01:28 |
| slangasek | ScottK: sure | 01:28 |
| ScottK | I ask because we recently did the libclamav2 -> libclamav3 transition and went from 0.91 to 0.92, so anything that uses clamav could use a good workout. | 01:28 |
| ScottK | slangasek: That effort made me appreciate the Debian BinNMU process a lot more. | 01:29 |
| ScottK | Not that it was a lot of packages, but it was my first library transition. | 01:29 |
| Hobbsee | slangasek: pong | 01:32 |
| Burgundavia | slangasek: I will poke this evening | 01:33 |
| slangasek | Hobbsee: hi; do you want to try to send gnome-games back through the ringer? It looks to me like these build failures were tied to the mirror issue | 01:34 |
| Hobbsee | slangasek: done, & reprio'd. | 01:34 |
| slangasek | Hobbsee: cheers | 01:34 |
| slangasek | ScottK: yes, it's been interesting for me to readjust to not having binNMUs... :) | 01:34 |
| phomes | slangasek, Hobbsee: nice, thanks :) | 01:35 |
| ScottK | Burgundavia: It'd be nice to get a mention the we've switched to libclamav3 and packages that use it really need some testing to make sure they still work as expected. | 01:36 |
| bmm | Hi everybody. I've written a debian file indexing page, and I'm looking for people who would like to comment on it and people who would like to help adding debian archive urls. Any takers? | 01:38 |
| ScottK | bmm: What problem are you trying to solve? | 01:39 |
| Burgundavia | ScottK: some of that is release notes stuff, but maybe we should merge the release notes and this page, I don't know | 01:39 |
| Burgundavia | slangasek: ^ | 01:39 |
| slangasek | Burgundavia: surely this is the release notes we're talking about? | 01:40 |
| bmm | ScottK: My brother asked me if there was a lgenerals-data package somewhere, and how he should get that data on his computer. I couldn't find a way of finding out, so I decided to create an index of debian file urls. | 01:40 |
| ScottK | bmm: Asking Google isn't enough? | 01:40 |
| Burgundavia | slangasek: are we still doing seperate release notes detailing issues with the release? | 01:41 |
| slangasek | Burgundavia: by "the release" do you mean hardy, or hardy alpha 3? | 01:41 |
| bmm | ScottK: nope, didn't solve that for me.. but of course asking google is the next best thing. | 01:41 |
| slangasek | Burgundavia: https://wiki.ubuntu.com/HardyHeron/Alpha3 is the draft for the alpha 3 release notes. For final release notes, I'm assuming there's not much point in having anything separate yet since issues will come and go | 01:42 |
| bmm | ScottK: I'm not sure google is that good at indexing debian files. Matching on filenames is not easy and knowing that it's a debian file is not really possible. I tried filetype:.deb but it didn't really result in anything ;-) | 01:42 |
| ScottK | bmm: While for a data package it probably doesn't matter so much, mixing and matching Debian packages built for different releases can be problematic. | 01:44 |
| bmm | ScottK: yes, that's true. It may be a good idea to post that on my page :-) | 01:44 |
| bmm | Ok, I've added a notice to the page :-) | 01:47 |
| bmm | ScottK: Thanks for you comment! | 01:47 |
| ScottK | bmm: You're welcome. | 01:48 |
| === _nuu is now known as nuu | ||
| slangasek | phomes: right, so, on the next round the buildd itself works fine and shows again that one of the build-dependencies is not available in main, so. | 01:54 |
| slangasek | nor do I see that anyone's started the process of trying to get it accepted in main | 01:55 |
| bmm | Am I reading this correctly: "#!/usr/bin/python or #!/usr/bin/env python (the former is preferred)" means that I should uphold #!/usr/bin/python in favor of the env version right? | 01:55 |
| bmm | (I would suspect the env python method being more flexible) | 01:56 |
| slangasek | you should use #!/usr/bin/python, yes | 01:56 |
| ScottK | bmm: #!/usr/bin/python is preferred | 01:56 |
| ScottK | bmm: #!/usr/bin/python is more reliable. | 01:56 |
| bddebian | Heya gang | 01:56 |
| StevenK | '#!/usr/bin/env python' will search PATH, so could use /usr/local/bin/python leading to unexpected behavior | 01:56 |
| slangasek | /usr/bin/env is only more flexible for supporting systems that don't support .deb packages | 01:57 |
| bmm | Hihi, every time I start typing a comment, somebody answers it :-D | 01:58 |
| bddebian | Ah, I see no response as usual from the games team on asc.. | 01:58 |
| slangasek | bddebian: er, I saw one response earlier when I looked at the list archive | 01:59 |
| ScottK | Heya bddebian. | 01:59 |
| bddebian | Hi ScottK | 01:59 |
| slangasek | bddebian: anyway, it's come to my attention that the only uploading the current "Uploaders" ever did was to adopt the package into the games team | 01:59 |
| bddebian | slangasek: Yeah, just Miriam | 01:59 |
| phomes | slangasek: so that person starting that process should be me I guess (co-maintainer of gnome-games) any hints on where I should go to do this? | 02:03 |
| phomes | bddebian: asc? | 02:04 |
| slangasek | phomes: https://wiki.ubuntu.com/MainInclusionProcess needs to be started for ggz-gtk-client, if gnome-games is going to be using ggz in Ubuntu | 02:05 |
| slangasek | bddebian: btw, does asc end up with a binary dep on libboost-regexp with this new upstream version? | 02:06 |
| bddebian | phomes: Advanced Strategic Command | 02:06 |
| slangasek | or do I dare hope that maybe it just needs a few templates? | 02:06 |
| somerville32 | I love asc! :D | 02:06 |
| bddebian | slangasek: Hmm, good question | 02:07 |
| phomes | is the games team a ubuntu team or would that be gnome-games team (me...)? Cause I should read up on asc then :) | 02:07 |
| bddebian | Debian Games Team | 02:07 |
| bddebian | We have several Ubuntu members :-) | 02:08 |
| phomes | good - as long as it's not me not doing what I should :) | 02:08 |
| bddebian | slangasek: Depends: asc-data (>= 2.0.1.0), libboost-regex1.34.1 (>= 1.34.1-2.1) :-( | 02:10 |
| slangasek | bddebian: what a horrible thing | 02:11 |
| bddebian | I'm sure that's pulled in by shlib deps though, no? | 02:13 |
| slangasek | yes | 02:13 |
| bddebian | Might not really be necessary? | 02:13 |
| slangasek | it's /using/ boost that's the horrible thing | 02:13 |
| bddebian | Well... | 02:13 |
| slangasek | especially since boost_regexp adds no significant functionality over regcomp() regexec() regfree() | 02:16 |
| slangasek | well, ok - it gets you C++ memory management. I don't think that's a very compelling reason to get one's application tangled up in boost. :) | 02:16 |
| bddebian | Heh, aye | 02:17 |
| StevenK | slangasek: "C offers you enough rope to hang yourself. C++ offers a fully equipped firing squad, a last cigarette and a blindfold." | 02:18 |
| jscinoz | hey again everyone | 02:21 |
| slangasek | StevenK: oh, but it gets better | 02:21 |
| StevenK | slangasek: Oh? | 02:21 |
| jscinoz | My preinst automatically downloads a 1.3mb file, its all working well but obviously it only shows the file download progress in the terminal window (its downloading using wget) is there a way to have it also output it graphically for users installing via synaptic or gdebi-gtk? | 02:21 |
| slangasek | what's the one regexp in the whole program that asc uses boost for? | 02:22 |
| slangasek | strrchr('.') | 02:22 |
| * StevenK sobs | 02:22 | |
| slangasek | behold, the C++ Way | 02:22 |
| slangasek | oh, no, my bad; there are others here that I overlooked | 02:24 |
| slangasek | yes, yes - there's also one here that does dirname() | 02:25 |
| bddebian | heh | 02:25 |
| slangasek | bddebian: well if I were you I would hit your upstream with a shovel for doing this; but it's largely going to be your problem, not mine, that the package depends on boost and will be doomed to eternal pain where Debian testing is concerned, so - uploading | 02:30 |
| bddebian | slangasek: Why is boost getting removed? | 02:30 |
| slangasek | but the jwz quote is very apt here | 02:30 |
| slangasek | bddebian: it's not, it's just constantly undergoing stupid ABI changes | 02:30 |
| bddebian | Oh, aye, OK | 02:31 |
| slangasek | and it has lots of other large reverse-dependencies which are never ready to go at the same time | 02:33 |
| Hobbsee | hey, would the person who keeps feeling the need to smash me over the head with a large hammer every time i get up please stop? | 02:34 |
| * bddebian hides the hammer | 02:34 | |
| Hobbsee | it's YOU! | 02:35 |
| * ScottK relaxes | 02:36 | |
| * Fujitsu sneaks up behind ScottK. | 02:38 | |
| * ScottK looks in the rear view mirror | 02:41 | |
| lifeless | boost is not managed for ABI | 02:50 |
| lifeless | AFAICT, by upstream | 02:50 |
| bddebian | Rock on, thanks slangasek! | 02:54 |
| bddebian | slangasek: Wanna join the games team, we NEED DDs? ;-P | 02:55 |
| slangasek | bddebian: thanks, I'm on enough teams | 03:09 |
| slangasek | lifeless: yes, which means that every time there's a release there's an soname change, which is an ABI change :) | 03:09 |
| === bigon is now known as bigon` | ||
| === Am|NickTaken is now known as Amaranth | ||
| TheMuso | c | 03:34 |
| TheMuso | ugh | 03:34 |
| TheMuso | c | 03:36 |
| TheMuso | argh orca let go of my keys | 03:36 |
| StevenK | I suspect it did, because we saw that. | 03:37 |
| rexbron | hmm, this is still troubling me | 03:50 |
| rexbron | The configure script changes the lib directory to lib64 on 64bit archs | 03:50 |
| rexbron | and makes the install fail | 03:51 |
| jscinoz | hey guys, in my package there is a shell script for launching the application "urbanterror" the shellscript is to be placed in /usr/games/urbanterror. However since one of the two binary packages created by this source package is also called urbanterror, dh_install tries to copy the temporary build dir debian/urbanterror rather than the script which is also debian/urbanterror (but a file rather than a directory). any ideas how i can | 04:57 |
| jscinoz | get around this, short of renaming the shellscript or package | 04:57 |
| slangasek | Burgundavia: OOo is almost certainly not going to make the cut for alpha3 at this point, since it just ftbfs on amd64. | 05:06 |
| Hobbsee | oh goody | 05:07 |
| Hobbsee | you can mention kde4, if it actually works though | 05:07 |
| * Hobbsee heads to work | 05:07 | |
| Burgundavia | slangasek: lovely | 05:08 |
| jscinoz | anyone know how i can deal with my dh_install problem? | 05:11 |
| somerville32 | jscinoz, use the -p option | 05:12 |
| StevenK | jscinoz: Don't call the script debian/urbanterror | 05:12 |
| StevenK | jscinoz: Call it debian/launcher or something, and then install it as debian/urbanterror/usr/games/urbanterror | 05:12 |
| jscinoz | i am using the -p option somerville, | 05:13 |
| jscinoz | stevenk, i was unware that dh_install would rename files | 05:13 |
| StevenK | jscinoz: dh_install can't rename files. | 05:15 |
| jscinoz | then how do i get it from srcdir/debian/launcher to usr/games/urbanterror | 05:16 |
| slangasek | dh_install copies files, which means it can copy them to arbitrary locations | 05:18 |
| === bigon` is now known as bigon | ||
| jscinoz | i know, but i think what stevenk was suggesting is i have it named one thing in debian/ and another thing when its actually installed, and not sure how that would work | 05:26 |
| jscinoz | >_< | 05:36 |
| tonyyarusso | Hi, I was wondering why the fix for flash is taking so long still. Anyone know? | 06:02 |
| LaserJock | because nobody has figured out what to do with it? | 06:31 |
| LaserJock | tonyyarusso: you know something I don't? :-) | 06:37 |
| dholbach | good morning | 06:39 |
| boomer | good morning | 06:41 |
| TheMuso | Hey dholbach. | 06:41 |
| dholbach | hey TheMuso, heya boomer | 06:42 |
| LaserJock | dholbach! | 06:44 |
| dholbach | hey LaserJock! | 06:45 |
| jsgotangco | hello | 06:45 |
| LaserJock | dholbach: I sent you an email | 06:46 |
| dholbach | LaserJock: I noticed but didn't get to it yet - I'll take a look at it in a bit | 06:47 |
| tonyyarusso | LaserJock: Really? I thought imbrandon said it would only be another day or two - like two weeks ago. Thought maybe something strange had come up that was unforseen, b/c he sounded like it was under control (and fixed in Hardy already, just not released vers) | 06:50 |
| LaserJock | tonyyarusso: yes, I'd say something came up: https://lists.ubuntu.com/archives/ubuntu-devel/2007-December/024877.html | 06:50 |
| Fujitsu | Yes, Adobe is silly. | 06:51 |
| Fujitsu | And Konqueror doesn't like XEmbed. | 06:51 |
| tonyyarusso | LaserJock: ty | 06:51 |
| nixternal | BAH! | 07:01 |
| nixternal | I wanted to check on getting another domain name, something a bit more professional sounding than 'nixternal.com', so I said hell, let me check johnson.com, and I could do richard@johnson.com, especially since my middle initial is A. | 07:02 |
| nixternal | so I go to godaddy.com and enter Johnson.com and it tells me it is already taken, but here are some other options | 07:02 |
| nixternal | http://www.nixternal.com/~rj/johnson.png <-- these are the other options | 07:02 |
| nixternal | wtf has this world come to? | 07:02 |
| Fujitsu | Hahahahaah. | 07:02 |
| LaserJock | lol | 07:03 |
| Fujitsu | Nice, nice. | 07:03 |
| nixternal | I am just going to change my name | 07:03 |
| nixternal | my first name is Richard and my last name is Johnson, so I hear it all of the time, but a damn domain registrant telling me the same | 07:03 |
| nixternal | this has gone to far | 07:03 |
| LaserJock | I was thinking donjohnson.com but that might date me | 07:04 |
| nixternal | you dated don johnson? | 07:04 |
| * Fujitsu once knew a Richard Head. | 07:04 | |
| nixternal | sick | 07:04 |
| nixternal | I know a Richard Head | 07:04 |
| nixternal | went to high school with one | 07:04 |
| Fujitsu | Same. | 07:04 |
| nixternal | our football coach even called him dickhead | 07:04 |
| Fujitsu | Haha. | 07:04 |
| * Fujitsu notes that Firefox 3 is being even safer than IE7 wrt. untrusted certificates. | 07:05 | |
| Fujitsu | It doesn't even say where to go to add an exception. | 07:05 |
| Fujitsu | Well, it does, but it's not right. | 07:05 |
| nixternal | and if everyone has seen the old Little Ceasers pizza commercials, they will remember 'Pizza Pizza!' | 07:05 |
| nixternal | well my Peurto Rican buddy started 'Pinga Pinga!' | 07:05 |
| tonyyarusso | Fujitsu: when is FF3 final btw? | 07:06 |
| Fujitsu | tonyyarusso: No idea whatsoever. | 07:06 |
| * tonyyarusso remembers Pizza Pizza! | 07:06 | |
| nixternal | ok, back to bed for me...g'nite all | 07:06 |
| Fujitsu | Night nixternal. | 07:06 |
| Fujitsu | All I know about FF at the moment is that FF3b3 is actually worth using, unlike earlier FF releases. | 07:06 |
| tonyyarusso | "Mozilla is on track to release the final version of Firefox 3 during the first quarter of 2008...we can expect the Beta 3 release sometime in February" | 07:08 |
| tonyyarusso | So, quite likely before Hardy is released, but not necessarily by much? | 07:08 |
| RAOF | bigon: Ping! Re: gnome-keyring-sharp. | 07:12 |
| bigon | RAOF: mmm I will recontact upstream | 07:13 |
| tonyyarusso | what is gnome-keyring-sharp? | 07:14 |
| RAOF | bigon: Oh, why? | 07:14 |
| RAOF | bigon: I'm just hacking on your packages as found on LP bzr. | 07:14 |
| RAOF | Adding a get-orig-source target, updating the versioning, etc. | 07:14 |
| tonyyarusso | seems to not be in gutsy, all right | 07:15 |
| bigon | svn://svn.debian.org/svn/pkg-cli-libs/packages/gnome-keyring-sharp/trunk | 07:15 |
| bigon | actually | 07:15 |
| bigon | RAOF: I don't think the bzr branch is uptodate | 07:15 |
| RAOF | I thought I went there and found nothing. Maybe the websvn was lying. | 07:15 |
| RAOF | Ok, let's see what's changed... | 07:16 |
| RAOF | Right. Not that much. | 07:18 |
| bigon | RAOF: upstream is waiting some feedback before making an official release | 07:19 |
| RAOF | bigon: Right. From whom? | 07:20 |
| bigon | somebody who test his software :o | 07:20 |
| RAOF | Such as may be garnered by putting a package into Debian Experimental, say? :) | 07:20 |
| RAOF | bigon: I don't have write access to pkg-cli-libs svn. If I supply you a diff would you kindly apply it? | 07:23 |
| bigon | RAOF: ok if you could send it by mail | 07:23 |
| RAOF | Yup, certainly. | 07:24 |
| RAOF | I'll just check that the package as it is currently fails to build current SVN, then fix it :) | 07:24 |
| RAOF | Yup. Good. | 07:27 |
| === asac_ is now known as asac | ||
| StevenK | RAOF: Finally reached level 60 :-D | 07:35 |
| RAOF | StevenK: In not so impressive news, I've hit 30 :P | 07:37 |
| StevenK | Heh | 07:37 |
| === cassidy_ is now known as cassidy | ||
| RAOF | bigon: Patch sent. | 07:50 |
| bigon | RAOF: thx | 07:51 |
| RAOF | bigon: Do you remember offhand what else was required for banter (which is what I was originally interested in g-k-s)? | 07:53 |
| bigon | telepathy-sharp(already pkg) gnome-keyring# | 07:54 |
| dholbach | apachelogger_: thanks a lot for your mail - you should be on the net in a bit :) | 07:54 |
| bigon | and notify@ | 07:54 |
| bigon | # | 07:54 |
| RAOF | Thanks. | 07:55 |
| bigon | RAOF: if I remember slomo was working on notify# | 07:55 |
| RAOF | As in, libnotify-sharp? | 07:55 |
| RAOF | Ah, inotify. | 07:56 |
| bigon | no libnotify | 07:58 |
| bigon | RAOF: I'm not sure if the .dll must go in /usr/lib/cli or /usr/lib/mono | 08:00 |
| bigon | gtk is in /usr/lib/mono | 08:01 |
| soren | pochu: http://git.fedorahosted.org/git/cobbler.git/?p=cobbler;a=blob;f=setup.py;h=c8a6b8136e2e6b7ba97290e8321c69defabedd13;hb=HEAD | 08:02 |
| RAOF | bigon: Fair enough. | 08:06 |
| dholbach | LaserJock: ROCK! BehindMOTU works now! | 08:07 |
| LaserJock | ;-) | 08:11 |
| RAOF | bigon: Ahhh... that'd be because gtk-sharp has dllmaps to native code, which is mono specific. gnome-keyring-sharp doesn't, so I think it's meant to go in /cli. | 08:14 |
| bigon | asking upstream | 08:15 |
| RAOF | bigon: That'd be a #debian-mono question, surely? | 08:15 |
| bigon | right :o | 08:16 |
| bigon | http://pkg-mono.alioth.debian.org/cli-policy/ch-packaging.html | 08:16 |
| bigon | "The commonly seen /usr/lib/mono/packagename path should only be used for Mono project packages." | 08:17 |
| bigon | RAOF: about notify-sharp https://bugs.edge.launchpad.net/ubuntu/+bug/139356 | 08:20 |
| lucas | hi, is Lukas Fittl around? | 08:48 |
| dholbach | lucas: his nick is 'lfittl' | 08:50 |
| huats | good morning | 08:57 |
| slytherin | man-di: ping | 09:13 |
| man-di | slytherin: pong | 09:14 |
| man-di | slytherin: please write direct content when pinging me | 09:14 |
| man-di | slytherin: I will answer when online or read backlog | 09:15 |
| slytherin | man-di: Ok. Did you find any time to build lucene2 on debian? | 09:15 |
| man-di | nope, sorry | 09:15 |
| === ember_ is now known as ember | ||
| === cprov-out is now known as cprov | ||
| === rzr is now known as rZr | ||
| DaveMorris | Hi, I'm still looking for a revu of my package - http://revu.tauware.de/details.py?package=libserial - all previous comments have been addressed | 11:03 |
| LucidFox | When creating an .orig.tar.gz from SVN, should I remove .svn directories? | 11:12 |
| Kmos | LucidFox: yes | 11:21 |
| Kmos | lintian gives you an error about that | 11:22 |
| TheMuso | Greetings norsetto. | 11:26 |
| norsetto | heya TheMuso | 11:27 |
| norsetto | themuso: hope to see you soon in the core team ;-) | 11:27 |
| TheMuso | norsetto: I hope so too. | 11:27 |
| TheMuso | Its a Hobbsee! | 11:39 |
| Hobbsee | heya! | 11:40 |
| Hobbsee | it is! | 11:40 |
| Hobbsee | anything interesting happen? | 11:40 |
| TheMuso | No, been rather quiet the last few hours. | 11:40 |
| Hobbsee | aww | 11:45 |
| * Hobbsee points at http://www.linuxno.de/_data/gallery/nwl7/_medium_DSCN7823.JPG for a bit of amusement | 11:46 | |
| norsetto | heya hobbsee, long time no see | 11:46 |
| * persia questions the safety of such an arrangement, and wonders if a longer lead just wasn't available | 11:46 | |
| Hobbsee | hiya norsetto, how were your holidays? | 11:46 |
| Fujitsu | llery/nwl7/_medium_DSCN7823.JPG for a bit of amusement | 11:46 |
| Fujitsu | 22:46:31 < norsetto> heya hobbsee, long time no see | 11:46 |
| Fujitsu | llery/nwl7/_medium_DSCN7823.JPG for a bit of amusement | 11:46 |
| Fujitsu | 22:46:31 < norsetto> heya hobbsee, long time no see | 11:46 |
| Fujitsu | Oops. | 11:47 |
| Hobbsee | oookay? | 11:47 |
| Fujitsu | Stupid Synaptics. | 11:47 |
| Fujitsu | I tried to three-finger click on the link, but missed. | 11:47 |
| Hobbsee | heh | 11:47 |
| norsetto | I feel pasted :-) | 11:47 |
| * Hobbsee reads kmos report | 11:47 | |
| Hobbsee | LOL! right then. | 11:48 |
| Hobbsee | sync requested but not really checked. Could not provide a justification beside "...it's the only way to get things updated..." | 11:48 |
| * Hobbsee notes that dholbach perhaps needs to check his sponsorships more carefully, toto | 11:50 | |
| norsetto | anyone know of a mouse cursor for kde? | 11:52 |
| * Hobbsee updates the wiki page | 11:52 | |
| norsetto | I mean, a mouse cursor in the shape of a mouse ..... | 11:52 |
| === asac_ is now known as asac | ||
| * Hobbsee hasn't | 11:56 | |
| pochu | soren: I don't know whether you can change variables from debian/rules, sorry. | 12:11 |
| artm | say i want to maintain a package for 2 flavours of ubuntu and for debian unstable, what are common practices for maintainging this? at the moment i just change the release in the top changelog record before making source package for each flavour, but it feels wrong | 12:11 |
| persia | If that's your only change, that's likely the best answer for something external to the repositories. You might define the special release "artm", and just compile it several different ways, but this may make it more difficult for your users to clearly identify against which environment the package they are using was compiled. | 12:13 |
| persia | pochu: You can export make variables to shell variables. If you can get the shell variables into python, you should be all set (unless I have the wrong context). | 12:14 |
| pochu | persia: he wants to change some variables without patching the file | 12:16 |
| artm | persia: well, so far that is the only change. eventually i would like to get the packages into official distributions, but for now I share them with our little community from my PPA so we can test/polish the package before suggesting it | 12:17 |
| persia | pochu: That's harder. Does python autoset internal variables from specially crafted shell variables or accept them on the command line? If neither, you're right. | 12:17 |
| === LucidFox is now known as Sikon_Stargate | ||
| pochu | i.e. changing wwwpath = "/var/www/cobbler/" to "/usr/share/cobbler/www/" | 12:17 |
| persia | artm: Then playing with the changelog is your best bet. | 12:17 |
| soren | pochu: Well.... If I can do it by passing arguments to setup.py, then I can do it from debian/rules. | 12:18 |
| artm | persia: thanks | 12:18 |
| pochu | soren: of course, but I don't know whether you can change just one of those variables with an argument to setup.py :) | 12:18 |
| soren | pochu: Ok, fair enough :) | 12:18 |
| pochu | soren: you can change all of them (or all of data_files), but just one... dunno | 12:19 |
| pochu | Maybe it's possible and I don't know it, who knows :) | 12:19 |
| persia | If you can change all of them, you can change one: just extract them all from setup.py, adjust the one you want, and change them all to the adjusted values. | 12:19 |
| soren | pochu: Ok. Meh.. It's fine to maintain a patch to setup.py. I was just curious is there was a more correct way to do it. | 12:19 |
| soren | persia: That's even more of a maintenance burden, though. | 12:20 |
| persia | (the writing of the limited python parser is left as an excercise for the make programmer) | 12:20 |
| persia | soren: I agree. I'm a fan of patches, personally. | 12:20 |
| persia | The quick & dirty way is to run sed -i against setup.py, but that'd be painful to understand for the next person. | 12:21 |
| soren | :) | 12:21 |
| promag | hi everyone | 12:21 |
| pochu | persia: not if you put a nice comment :) | 12:22 |
| promag | not sure if this is the right place, but I have a problem in makefile.am | 12:23 |
| persia | pochu: Maybe. Depends on the package. I avoid using sed to patch in debian/rules unless there is an active maintainer already doing so. It makes it hard to meet both the build-twice-in-a-row and the debian/rules-clean-is-really-clean tests. | 12:23 |
| promag | what I have: file.xml which will create file.h and file.cpp | 12:23 |
| promag | how could I define this in makefile.am | 12:24 |
| soren | promag: You can put regular make rules in makefile.am. | 12:24 |
| promag | I've tried it but how can I make the dependency work? | 12:25 |
| soren | promag: Er... You haven't given nowhere near enough information for that question to make sense (let alone be answered). | 12:26 |
| persia | promag: file.h file.cpp: file.xml | 12:26 |
| promag | what I put in _SOURCES? | 12:26 |
| promag | file.xml? | 12:27 |
| promag | or file.xml and file.cpp ? | 12:27 |
| soren | Have you tried? | 12:27 |
| promag | yes.. the custom target isn't called... | 12:28 |
| promag | I have tried this: http://unix.derkeiler.com/Newsgroups/comp.unix.questions/2003-04/0157.html | 12:30 |
| soren | I belive #devtools on OFTC is the autotools support channel. | 12:33 |
| promag | thanks | 12:36 |
| mok0 | promag: I've done something similar, hang on | 12:36 |
| promag | mok0: great! | 12:37 |
| mok0 | promag: put the file.cpp in _SOURCES | 12:37 |
| promag | it is already | 12:38 |
| mok0 | promag: and make a rule to generate file.cpp | 12:38 |
| promag | .cpp: %.xml | 12:38 |
| promag | echo $@ && touch $@.cpp | 12:38 |
| mok0 | file.cpp: file.xml | 12:38 |
| promag | I have this | 12:38 |
| mok0 | promag: that rules doesn't do anything? | 12:38 |
| promag | no | 12:39 |
| mok0 | file.cpp: file.xml /n/t xml2cpp < file.xml > file.cpp | 12:39 |
| promag | if I specify file.cpp: file.xml then its ok... but if I use %.cpp: %.xml then it fails | 12:40 |
| promag | I need a generic rule | 12:40 |
| mok0 | promag: hm | 12:40 |
| promag | otherwise I'll have lots of those | 12:40 |
| mok0 | promag: but you are not generating any .cpp files from the .xml (?) | 12:41 |
| promag | yes I am... touch file.cpp :) | 12:41 |
| mok0 | just empty files? | 12:42 |
| promag | well it works :P | 12:42 |
| promag | later I'll switch touch to the generator | 12:42 |
| mok0 | promag: what does this program do? :-P | 12:42 |
| promag | what? | 12:42 |
| promag | what program? | 12:42 |
| promag | the generator? | 12:43 |
| mok0 | promag: it was a joke... a program with null files... | 12:43 |
| promag | :P | 12:43 |
| promag | well it doens't matter what the .cpp has | 12:43 |
| promag | (for now) | 12:43 |
| mok0 | promag: there may be some restriction on using generic rules inside Makefile.am, I'm not sure | 12:43 |
| promag | mok0: yes I think generic makefile rules doens't work in Makefile.am | 12:44 |
| promag | if you see /usr/share/automake-1.10/am/yacc.am | 12:44 |
| mok0 | promag: why don't you make a small script that generates the mass of individual rules | 12:44 |
| promag | because automake should be the solution no? | 12:45 |
| mok0 | promag: just to generate the individual rules to put in Makefile.am | 12:45 |
| promag | I'm not happy with that solution | 12:45 |
| mok0 | promag: I understand. But it would work | 12:46 |
| mok0 | promag: you may zoom in on a better solution later | 12:46 |
| promag | mok0: soren: persia: any idea of a package that could have a similar process? | 12:55 |
| soren | promag: What happens to generic rules you put in Makefile.am? | 12:55 |
| soren | I wouldn't expect automake to just remove them. | 12:56 |
| promag | nothing | 12:56 |
| promag | Ill check makefile | 12:56 |
| promag | its right after .PHONY | 12:57 |
| soren | So it's there? | 12:57 |
| soren | Ok, what was the problem again? | 12:57 |
| mok0 | promag: then it should work | 12:57 |
| promag | hoooooooo | 12:58 |
| promag | %.cpp fails | 12:58 |
| promag | but if I do %Generated.cpp: %.xml then it works | 12:58 |
| persia | promag: Don't put anything after .PHONY. Also, try adjusting %.cpp: %.xml to be $(CPP_FILES_TO_BE_CONSTRUCTED): %.cpp: %.xml | 12:58 |
| promag | I'm not putting anything .PHONY... automake is! | 12:59 |
| * persia disagrees with automake on this stylistic point, but the important part is $(expression): (variable extraction): (variable replacement) | 13:00 | |
| persia | If you don't have both colons, make assumes you are talking about a file called %.cpp | 13:00 |
| === azeem_ is now known as azeem | ||
| mok0 | promag: http://libre.adacore.com/viewvc/trunk/polyorb/docs/Makefile.am?view=markup | 13:01 |
| promag | you are most helpful! | 13:02 |
| Toma- | Hey, im trying to track down a change to the Makefile in linux-ubuntu-modules. Specifically, the reason why rtl8180 is disabled | 13:06 |
| DaveMorris | Toma-: have you looked in the changelog file? | 13:10 |
| Toma- | yep | 13:11 |
| Toma- | https://edge.launchpad.net/ubuntu/+source/linux-ubuntu-modules-2.6.24 | 13:11 |
| Toma- | reading this | 13:11 |
| Toma- | which is really no help at all | 13:12 |
| DaveMorris | I'd suggest looking up the guys irc name who made the change and asking them, or firing off an email to them | 13:13 |
| Toma- | BenC :D | 13:13 |
| jpatrick | Toma-: best poke him in #ubuntu-devel | 13:13 |
| Toma- | ill hassle him in -kernel | 13:14 |
| === Sikon_Stargate is now known as LucidFox | ||
| geser | dmb: Hi, didn't you want to test pdfedit with lesstif2-dev? any results? | 13:29 |
| dholbach | Hobbsee: which one are you talking about? | 13:39 |
| dholbach | hey norsetto :) | 13:39 |
| dholbach | hiya geser | 13:39 |
| Hobbsee | dholbach: the sync request which slangasek threw down? | 13:39 |
| dholbach | Hobbsee: which one was that? | 13:39 |
| Hobbsee | it's listed on kmos report | 13:39 |
| dholbach | Hobbsee: i built it myself and the build-depends doesn't seem to be necessary any more | 13:40 |
| dholbach | Hobbsee: why were you referring to it? | 13:40 |
| Hobbsee | was just surprised - and was looking thru the list | 13:41 |
| dholbach | to the bug I mean? | 13:41 |
| norsetto | hey dholbach! Still all in one piece or missed some bits after your snowboard adventures? :-) | 13:41 |
| dholbach | norsetto: it was absolutely awesome | 13:41 |
| dholbach | unfortunately my camera didn't work so no pictures | 13:41 |
| norsetto | dholbach: glad to hear that, as all good things it ended though, and now you are back ;-) | 13:41 |
| * dholbach hugs norsetto | 13:41 | |
| dholbach | yeah, it's great to be back again | 13:42 |
| dholbach | norsetto: how were your holidays? | 13:42 |
| * norsetto hugs dholbach too (but find him a bit cold and its surprised to find some snow flakes on him ....) | 13:42 | |
| dholbach | hehe, not from berlin | 13:42 |
| * Hobbsee sets dholbach and norsetto on fire | 13:42 | |
| geser | Hi dholbach, Hobbsee | 13:43 |
| norsetto | dholbach: pretty busy with a peesonal project which I had been postponing for too long | 13:43 |
| * norsetto flame hugs Hobbsee | 13:43 | |
| dholbach | norsetto: I hope you're doing alright | 13:44 |
| norsetto | heya geser | 13:44 |
| geser | Hi norsetto | 13:44 |
| BugMaN | Ave norsetto! :) | 13:44 |
| norsetto | dholbach: as well as a I can be :-) | 13:45 |
| norsetto | BugMaN: morituri me salutant ? | 13:46 |
| BugMaN | norsetto: yes :) | 13:46 |
| norsetto | BugMaN: oh, you told her the truth? Never do that! | 13:46 |
| apachelogger__ | dholbach: hooray, thanks :) | 13:46 |
| === apachelogger__ is now known as apachelogger | ||
| * norsetto goes to eat something | 13:48 | |
| Chipzz | I hope this is not offtopic here, but does anyone around have experience with dotdeb? | 13:49 |
| dholbach | Hobbsee: did you check anything regarding bug 180915? | 13:54 |
| ubotu | Launchpad bug 180915 in big-cursor "Please sync big-cursor 3.7 (universe) from Debian unstable (main)" [Wishlist,Incomplete] https://launchpad.net/bugs/180915 | 13:54 |
| bobbo | I need a bit of help updating a package | 13:56 |
| bobbo | I fixed a man page in a package a few weeks ago (showfsck-1.3ubuntu1) not knowing a new version (showfsck-1.4) had been released upstream in 2004. I want to merge with the upstream version but not sure how i should set up the changelog? | 13:58 |
| dholbach | Hobbsee: I'll remove the "-" (regarding but 180915) in the report because I agree that Kmos should have replied to the objection, but to me it's still a valid sync request - that's why I put "please only add a review (+,0,-) for bug reports that are closed." | 14:01 |
| dholbach | Hobbsee: if you have any objection about the bug being OK, please add your comment to the bug report | 14:01 |
| Hobbsee | ...right | 14:02 |
| slytherin | can anyone please tell me what is meaning of 'tentatively lool'? | 14:11 |
| Hobbsee | lool is a person | 14:11 |
| Hobbsee | so, presumably "possibly lool" | 14:12 |
| Hobbsee | or "i think it's lool that you want" | 14:12 |
| geser | slytherin: /whois lool | 14:12 |
| slytherin | geser: I have no idea. I was just checking DesktopTeam TODO page on wiki. And it has this text against nautilus 2.21.x | 14:13 |
| LucidFox | I've just uploaded package smplayer-themes to REVU, looking for reviews: https://bugs.launchpad.net/ubuntu/+bug/181319 | 14:33 |
| ubotu | Launchpad bug 181319 in ubuntu "[needs-packaging] smplayer-themes" [Wishlist,In progress] | 14:33 |
| LucidFox | oops | 14:33 |
| LucidFox | http://revu.ubuntuwire.com/details.py?package=smplayer-themes | 14:33 |
| bobbo | Could a MOTU check out my fix for Bug #181775 | 14:50 |
| ubotu | Launchpad bug 181775 in showfsck "Showfsck package out of date" [Undecided,In progress] https://launchpad.net/bugs/181775 | 14:50 |
| somerville32 | bobbo, Hi. The Debian Import Freeze is currently in-effect meaning that there needs to be a provided rational for the merge. | 14:53 |
| somerville32 | *rationale | 14:54 |
| ScottK | bobbo: You need an interdiff, not a debdiff for a new upstream version. See https://wiki.ubuntu.com/UbuntuDevelopment/Interdiff | 14:54 |
| ScottK | somerville32: The bug includes a good rationale. | 14:55 |
| somerville32 | ScottK, The fact that it was released in 2006 is good rationale? | 14:55 |
| ScottK | The fact that it's two years old, yes. | 14:56 |
| ScottK | DIF isn't really a freeze as it is turning off the auto-sync with Debian. | 14:56 |
| ScottK | The threshold for 'good' is very low at this point. | 14:56 |
| ScottK | Personally, I'd settle for "I thought it was important enough to invest my time in preparing the update", but that's just me. There is no actual process requirement for justification. | 14:57 |
| somerville32 | ScottK, okay, thanks for clearing that up. | 14:58 |
| somerville32 | bobbo, https://wiki.ubuntu.com/UbuntuDevelopment/Merging <-- for detailed merging instructions | 15:00 |
| bobbo | somerville32: im not merging from Debian, but from upstream | 15:01 |
| bobbo | is the diff.gz the interdiff page is talking about the same as a source .tar.gz? | 15:10 |
| ScottK | bobbo: diff.gz is the one you get after you make the source package. | 15:11 |
| ScottK | Source package being usually made up of orig.tar.gz, diff.gz, and .dsc. | 15:12 |
| bobbo | running debuild -S doesnt create a diff.gz though | 15:12 |
| norsetto | morning scottk | 15:12 |
| ScottK | Morning norsetto. | 15:13 |
| emgent | hello there | 15:13 |
| ScottK | bobbo: It should. | 15:15 |
| ScottK | hello emgent | 15:15 |
| bobbo | ScottK: It jsut makes a showfsck_1.4.tar.gz containing the source of the package | 15:18 |
| ScottK | bobbo: Then you're making a native package and you shouldn't be doing that. | 15:19 |
| ScottK | bobbo: I take that back. | 15:20 |
| ScottK | bobbo: I didn't notice that it actually is a native package. | 15:21 |
| bobbo | ScottK: ah so how should i merge it then? | 15:22 |
| ScottK | bobbo: Where is the source for the upstream release? | 15:22 |
| ScottK | bobbo: Not sure. | 15:22 |
| ScottK | bobbo: Point me at the upstream site and let me see what they have there. | 15:23 |
| bobbo | http://www-id.imag.fr/Laboratoire/Membres/Danjean_Vincent/deb.html#showfsck but i included the 1.3ubuntu1 changes in my debdiff in the bug | 15:23 |
| bobbo | ScottK: http://www-id.imag.fr/Laboratoire/Membres/Danjean_Vincent/debian/pool/main/s/showfsck/showfsck_1.4.tar.gz <-- Direct link to the tar.gz i used to merge | 15:24 |
| ScottK | bobbo: For a native package, interdiff makes no sense. What I'd suggest you do is take his 1.4 package, add your changes and call it 1.4~ubuntu1, and then make a debdiff between the two. | 15:28 |
| bobbo | ok i'll do that and upload it | 15:28 |
| ScottK | bobbo: Think about it based on the idea that whoever sponsors it will retrieve the upstream source themselves. | 15:28 |
| tjaalton | Mez: are you still working on the iFolder ITP's you've filed? | 15:38 |
| Mez | tjaalton, no. Not after novell messed me about - I thought they'd been automatically closed thoug | 15:38 |
| tjaalton | Mez: ok.. I've been planning to do that for quite some time now, but so far haven't had the time | 15:39 |
| === mgallagh__ is now known as mattva01 | ||
| mattva01 | I'm pretty new to packaging so I'm wondering if someone could help me out | 15:44 |
| ScottK | mattva01: Probably. Asking specific questions will get you more help. | 15:44 |
| mattva01 | I have a package that does not create any binaries , it just configures eclipse in a certain way | 15:46 |
| mattva01 | but I get errors similar to this :dh_builddeb: I have no package to build | 15:46 |
| mattva01 | dh_testdir | 15:46 |
| mattva01 | and dpkg-genchanges: failure: cannot read files list file: No such file or directory | 15:47 |
| mattva01 | is there something obvious i'm missing? | 15:47 |
| man-di | mattva01: you always need to build on binary deb | 15:47 |
| man-di | mattva01: you do this by adding an entry to debian/control | 15:48 |
| mattva01 | the entry is correct I believe | 15:48 |
| man-di | mattva01: then you would not get the above error message | 15:48 |
| mattva01 | well let me check then :) | 15:49 |
| siretart | mattva01: not every binary package needs to actually contain binaries. if the package is architecture neutral, make it 'arch: all' | 15:49 |
| man-di | mattva01: paste your debian/control to some paste.bin like http://rafb.net/paste | 15:49 |
| man-di | mattva01: binary package just means, its some deb you can install | 15:50 |
| mattva01 | ah | 15:50 |
| man-di | it can contain everything or nothing | 15:50 |
| man-di | or something in between | 15:50 |
| mattva01 | http://rafb.net/p/lHaGwj36.html | 15:53 |
| bobbo | ScottK: My new debdiff is up at Bug #181775 if you want it | 15:53 |
| ubotu | Launchpad bug 181775 in showfsck "Showfsck package out of date" [Wishlist,In progress] https://launchpad.net/bugs/181775 | 15:53 |
| mattva01 | so is there anything wrong with it? | 15:57 |
| man-di | mattva01: Version: 1.0 should not be in there | 15:58 |
| man-di | but thats most probably not your error | 15:59 |
| mattva01 | hmm | 16:00 |
| man-di | version number comes always from debian/changelog | 16:00 |
| mattva01 | oh yeah | 16:00 |
| Chipzz | I think a space is required after : | 16:01 |
| Chipzz | and the cdbs build dependency is most likely bogus | 16:01 |
| mattva01 | yeah it is, sorry bout that | 16:01 |
| Chipzz | also | 16:01 |
| Chipzz | the dependecies on sun-java6-jre, sun-java6-jdk are most likely bogus | 16:02 |
| mattva01 | no I actually need those | 16:02 |
| Rospo_Zoppo | Riddell: i'm working on a fix for a pk2moto ftbfs; the problem is that the debdiff is very large, can you have a look at it (bug 181581) ? thanks | 16:02 |
| ubotu | Launchpad bug 181581 in p2kmoto "FTBFS on all architectures" [Undecided,Confirmed] https://launchpad.net/bugs/181581 | 16:02 |
| Chipzz | mattva01: is there a very good reason it won't work with for example gcj or sth? | 16:02 |
| mattva01 | the AP board seems to require sun java | 16:03 |
| Chipzz | still | 16:03 |
| Chipzz | you're not *packaging* the AP board | 16:03 |
| Chipzz | you're only *configuring* it | 16:03 |
| rexbron | Is there ap olicy of having .py files in /usr/bin/ ? | 16:03 |
| mattva01 | true | 16:03 |
| Chipzz | if the AP board package will only work with a specific version of java, then IT should have that dependency | 16:04 |
| mattva01 | ah yeah | 16:04 |
| Chipzz | if in the future the AP board software is changed to also work with alternative runtimes, your package will block that | 16:04 |
| Chipzz | >> not good | 16:04 |
| === \sh_away is now known as \sh | ||
| mattva01 | of course | 16:05 |
| Chipzz | hrrrm well actually it won't, it will just force unneeded packages to stay installed | 16:06 |
| Chipzz | not good nevertheless | 16:06 |
| mattva01 | well I see I have much to learn :) | 16:06 |
| mattva01 | still don't understand these errors I'm getting though | 16:07 |
| \sh | moins | 16:07 |
| geser | Hi \sh | 16:08 |
| Chipzz | mattva01: btw, I don't know ap board nor do I particulary care; but, is it a seperate ubuntu package or just some tarball? | 16:08 |
| Riddell | Rospo_Zoppo: hmm, you're modifying autofoo generated files, isn't there a way to fix it without doing that? | 16:09 |
| Chipzz | since you have Source: apjava; assuming apjava is the AP board source package, that would also be incorrect ;) | 16:09 |
| Riddell | Rospo_Zoppo: or how did you create that patch? | 16:09 |
| mattva01 | its a tarball | 16:09 |
| \sh | hmm...is it my ISP or is archive.ubuntu.com slow today? | 16:10 |
| mattva01 | AP board= Advanced Placement (for high school) , the case study for AP Computer Science is something called gridworld | 16:10 |
| Chipzz | mattva01: then another question is, why are you just packaging some postinstall scripts instead of that tarball? ;) | 16:10 |
| Rospo_Zoppo | Riddell: I deleted the old one, then I created a new one with the same name and I did aclocal, autoconf, automake | 16:10 |
| Rospo_Zoppo | Riddell: this makes the package build but the debdiff is very large | 16:12 |
| mattva01 | well I am packaging the tarball for gridworld but it never actually gets untarred | 16:12 |
| Riddell | Rospo_Zoppo: right, fair enough, if that's what it needs than that's what needs to be done | 16:12 |
| Riddell | Rospo_Zoppo: has there been any upstream activity on this do you know? | 16:12 |
| Rospo_Zoppo | Riddell: I don't know this honestly | 16:13 |
| Riddell | Rospo_Zoppo: no, doesn't look like there has been | 16:13 |
| Rospo_Zoppo | ok | 16:14 |
| Riddell | Rospo_Zoppo: well fine with me, I'll upload | 16:14 |
| Rospo_Zoppo | Riddell: thanks :) | 16:14 |
| Riddell | Rospo_Zoppo: builds for me, uploaded, thanks for the fix | 16:16 |
| Rospo_Zoppo | good | 16:16 |
| \sh | wow...guys...thx for all the cheering :) | 16:28 |
| mattva01 | :( failed again | 16:28 |
| bddebian | Heya gang | 16:31 |
| ScottK | Heya bddebian | 16:32 |
| bddebian | Hi ScottK | 16:32 |
| \sh | damn..my body is hurting now...why do I have to go into a gym and do some training...:( | 16:33 |
| LaserJock | \sh: hi | 16:39 |
| LaserJock | nice to see your MOTU app ;-) | 16:39 |
| \sh | LaserJock: thx :) | 16:40 |
| Nightrose | \sh: wohooo you applied again? /me checks | 16:42 |
| * ScottK kicks dholbach to approve \sh today .... | 16:42 | |
| Nightrose | hehe | 16:42 |
| * Nightrose hugs dholbach instead | 16:42 | |
| Nightrose | might work better :P | 16:42 |
| dholbach | Nightrose: thanks, exactly :) | 16:43 |
| Nightrose | ;-) | 16:43 |
| * dholbach looks at it after the CC meeting | 16:43 | |
| \sh | thx a lot to all of you...but let dholbach do his paid work first :) | 16:43 |
| \sh | it's nothing to hurry about :) | 16:44 |
| * ScottK has been trained by hobbsee | 16:47 | |
| leonel | ScottK: hello : For cve-2007-6337 I've made a diff in clamav 0.92 and 0.91.2 for bzlib_private.h and it's a little change should I do the debdiff ?? | 16:48 |
| ubotu | Unspecified vulnerability in the bzip2 decompression algorithm in nsis/bzlib_private.h in ClamAV before 0.92 has unknown impact and remote attack vectors. (http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2007-6337) | 16:48 |
| \sh | ScottK: lol :) | 16:48 |
| leonel | even is an unknown issue | 16:48 |
| ScottK | leonel: CVE says ClamAV before 0.92. What change do we need in 0.92? | 16:49 |
| leonel | in 92 none for 91.2 | 16:51 |
| ScottK | Ah. | 16:51 |
| ScottK | Yes, please | 16:51 |
| ScottK | (debdiffs) | 16:51 |
| leonel | ScottK: http://paste.ubuntu-nl.org/51455/ | 16:52 |
| * ScottK looks | 16:52 | |
| leonel | searching for more info | 16:52 |
| leonel | I've found the same partch for gentoo : http://bugs.gentoo.org/attachment.cgi?id=138930 | 16:53 |
| ScottK | leonel: Where did you find the patch originally? | 16:53 |
| \sh | leonel: is it on the CVE mitre page linked? | 16:54 |
| leonel | as the cve says the bug was in the bzlib_private.h made the diff then looking more | 16:55 |
| \sh | ScottK: I think it comes from here http://security.gentoo.org/glsa/glsa-200712-20.xml | 16:55 |
| leonel | found the gentoo one | 16:55 |
| ScottK | OK. I'd say go for it then. | 16:55 |
| \sh | http://bugs.gentoo.org/show_bug.cgi?id=202762 | 16:55 |
| leonel | http://www.cve.mitre.org/cgi-bin/cvename.cgi?name=2007-6337 | 16:55 |
| ubotu | bugs.gentoo.org bug 202762 in Vulnerabilities "app-antivirus/clamav < 0.91.2-r1 Multiple vulnerabilities (CVE-2007-{6335,6336,6337})" [Major,Resolved: fixed] | 16:55 |
| leonel | and there links to gentoo | 16:55 |
| \sh | it's always good to have a patch from a different distro ready...I trust redhat very much but sometimes as well gentoo ,-) | 16:56 |
| === bigon726 is now known as bigon | ||
| ScottK | It looks like gentoo is the only one with a patch for this yet. | 16:57 |
| \sh | ScottK: fedora is pushing new upstream versions for bug fixes | 16:57 |
| \sh | ScottK: RHEL doesn't have clamav afaik | 16:57 |
| ScottK | OTOH, they released it two weeks ago and haven't had to fess up to any changes. | 16:57 |
| ScottK | Ah. | 16:57 |
| leonel | \sh: same here about redhat security | 16:57 |
| leonel | so .. go for the debdiff for gutsy and feisty ? | 16:58 |
| \sh | ScottK: afaik we and debian are the only two distros with really stable releases and bugfixes for even non enterprise distros ready... | 16:58 |
| ScottK | Well Fedora is going to have fun with 0.92 then as we had to patch at least two packages in addition to rebuilding for the soname change | 16:58 |
| \sh | ScottK: they will push 0.93 to their archives | 16:59 |
| \sh | ScottK: and deal with the rest | 16:59 |
| ScottK | leonel: Yes. And then once the Gutsy one is out, we'll backport that to feisty-backports. | 16:59 |
| leonel | ScottK: Ok.. going to the kitchen to cook the debdiff ... | 16:59 |
| ScottK | Great. | 16:59 |
| \sh | ScottK: btw...are we still nailed for -backports with the packages from the latest developement version or can we also upload to it like to -motu? | 17:00 |
| ScottK | We can do source backports, but it takes a core-dev to upload. | 17:00 |
| ScottK | We can backport from any release though, so a backport from gutsy-security to feisty-backports is no problem (clamav does require a source backport though) | 17:01 |
| \sh | ScottK: so actually we could do a backport of wine to all other releases with a source change... | 17:02 |
| \sh | well, this will be fun...correct time to celebrate: 11th january, 2008: Release Of KDE4.0.0, birthday of kris koehntopp (aka isotopp) and Stephan Hermann (aka \sh) ... if this isn't perfect I don't what is | 17:09 |
| bluefoxicy | why can I perpetually not record in ubuntu | 17:09 |
| bobbo | Does anyone have a tutorial on packaging a Python script with a Distutils setup.py? | 17:10 |
| geser | ScottK: while speaking of clamav: what do we do with the clamav-data package in hardy? should we try to keep it as up-to-date as possible? i.e. sync it from debian regularly? | 17:11 |
| bluefoxicy | https://bugs.launchpad.net/ubuntu/+source/gnome-media/+bug/174550 | 17:12 |
| nxvl_work | does anyone knows where norsetto is? i haven't see him in a long time | 17:12 |
| ubotu | Launchpad bug 174550 in gnome-media "can't record sound with microphone. listening works. tested several applications among them the simple sound recorder!!" [Undecided,New] | 17:12 |
| mok0 | <a href="https://wiki.ubuntu.com/PackagingGuide/Howtos/CDBS?highlight=%28CDBS%29">bobbo: PackagingGuide/Howtos/CDBS - Ubuntu Wiki</a> | 17:12 |
| norsetto | nxvl_work: hmmm, here? | 17:12 |
| bobbo | nxvl_work: He was on about 3pm GMT | 17:12 |
| nxvl_work | norsetto: hi! | 17:12 |
| mok0 | bobbo: <a href="https://wiki.ubuntu.com/PackagingGuide/Howtos/CDBS?highlight=%28CDBS%29">PackagingGuide/Howtos/CDBS - Ubuntu Wiki</a> | 17:12 |
| nxvl_work | norsetto: it's long time i don't see you | 17:12 |
| norsetto | nxvl_work: Hola! | 17:12 |
| bobbo | thanks mok0 | 17:12 |
| norsetto | nxvl_work: yes, was pretty busy, how is it? | 17:13 |
| nxvl_work | norsetto: little boring, really light days on work | 17:13 |
| mok0 | bobbo: it has the rules file for you right there | 17:13 |
| bobbo | mok0: "Packaging Python modules with CDBS"? | 17:14 |
| mok0 | bobbo: yep | 17:14 |
| ScottK | geser: Yes. Please keep it up to date. | 17:14 |
| ScottK | bobbo: It's easy enough to do. The pypolicyd-spf source package is a decent example. | 17:15 |
| jdong | LucidFox: just a status update, mpeg4ip was rejected by archive admins due to debian/copyright issues. I'm going to work on fixing that up but it's a lot of work and will probably take me the rest of today and tomorrow. | 17:39 |
| jdong | LucidFox: it bundles all kinds of works from 10+ sources and some of them don't have live website links anymore so I'm gonna have to do some detective work | 17:40 |
| \sh | dholbach: thx :) | 17:41 |
| dholbach | \sh: :-) | 17:41 |
| \sh | dholbach: what to do now? | 17:42 |
| dholbach | \sh: wait for the other MC members to send in their vote | 17:42 |
| imbrandon | moins all | 17:42 |
| dholbach | if everybody's happy you will get re-added to the team | 17:42 |
| \sh | dholbach: cool...so no work from my side anymore... | 17:43 |
| dholbach | no :-) | 17:43 |
| \sh | would be a cool birthday present...;) | 17:43 |
| nxvl_work | imbrandon: i have just write you and e-mail some minutes ago | 17:43 |
| nxvl_work | :D | 17:43 |
| imbrandon | nxvl_work: i just replyed | 17:43 |
| imbrandon | nxvl_work: you can count on me | 17:44 |
| nxvl_work | imbrandon: thnx | 17:44 |
| nxvl_work | :D | 17:44 |
| nxvl_work | received | 17:44 |
| * nxvl_work *HUGS* imbrandon | 17:44 | |
| imbrandon | :) | 17:44 |
| * ScottK reads dholbach's mail and looks over at geser and soren. | 17:45 | |
| soren | ScottK: Exactly which e-mail are you talking about? | 17:46 |
| ScottK | soren: The one where dholbach +1'ed \sh's MOTU application. Maybe I lost track, but I didn't see a reply from you yet. | 17:47 |
| * ScottK is hoping we can give \sh a good MOTU birthday present | 17:47 | |
| soren | ScottK: I haven't replied yet. | 17:47 |
| soren | On purpose. | 17:47 |
| soren | His birthday is coming up? | 17:47 |
| \sh | soren: http://www.sourcecode.de/content/day-celebrate | 17:48 |
| soren | Ah, I see. | 17:48 |
| * \sh needs a coffee and tries to fix simage... | 17:51 | |
| leonel | ScottK: should I do a new bug report for this cve ?? I have the gutsy debdiff all builded installed and checked bzip2 fine | 17:57 |
| ScottK | leonel: Yes. | 17:58 |
| ScottK | leonel: Let me know when you have it filed and I'll approve the tasks for gutsy/feisty | 17:59 |
| === dholbach_ is now known as dholbach | ||
| leonel | ScottK: ok | 18:00 |
| leonel | ScottK: done bug 181830 | 18:05 |
| ubotu | Launchpad bug 181830 in clamav "CVE-2007-6337 Unknown impact remote attack" [Undecided,New] https://launchpad.net/bugs/181830 | 18:05 |
| leonel | baking feisty's debdiff | 18:05 |
| * ScottK looks | 18:05 | |
| === x-spec-t is now known as Spec | ||
| ScottK | leonel: Done. | 18:09 |
| ScottK | keescook: Would you please have a look at bug 181830 for Gutsy? | 18:09 |
| ubotu | Launchpad bug 181830 in clamav "CVE-2007-6337 Unknown impact remote attack" [Undecided,In progress] https://launchpad.net/bugs/181830 | 18:09 |
| dholbach | see you guys tomorrow! have a nice rest of your day! :-) | 18:10 |
| keescook | ScottK: sure thing | 18:10 |
| ScottK | keescook: Thanks. Patch is from Gentoo. | 18:10 |
| \sh | grmpf...where is the cdbs documentation...the link from DeveloperResources is not working anymore | 18:11 |
| ScottK | \sh: You'll just end up reading the source anyway. May as well start there. | 18:11 |
| \sh | thx soren | 18:11 |
| \sh | ScottK: well I just need the correct line for regenerating autofoo crap | 18:12 |
| \sh | ah it's in the doc dir | 18:12 |
| \sh | ScottK: to fix simage I needed to add a new macro to the source package | 18:13 |
| * ScottK nods | 18:13 | |
| ScottK | geser: Ping | 18:15 |
| \sh | I'll do it differently .. I#ll recreate the autofoo stuff manually (and put it into a patch file) | 18:22 |
| === davi_ is now known as davi | ||
| leonel | ScottK: feisty's clamav does not have libclamav/nsis/bzlib_private.h ... | 18:35 |
| leonel | searching where is the bzip things | 18:35 |
| ScottK | leonel: OK. I think the nsis library may have been new in 0.91. | 18:36 |
| leonel | ScottK: yes | 18:36 |
| leonel | ScottK: there's no #define BZ_GET_FAST | 18:42 |
| ScottK | leonel: Sounds like non-applicable to Feisty then. Thanks for checking. I'll update the bug. | 18:42 |
| === cassidy_ is now known as cassidy | ||
| ScottK | keescook: Could you point me at (or provide) the final version of your Gutsy package for clamav so I can prepare a source backport for feisty-backports? | 18:53 |
| === bigon is now known as bigon` | ||
| bobbo | I just uploaded a package to REVU without the source tar.gz, how should i fix this? | 19:29 |
| ScottK | bobbo: Upload again with the tar.gz | 19:29 |
| bobbo | ScottK: It says there is nothing to upload whe i try to re-up it | 19:35 |
| ScottK | bobbo: dput -f | 19:36 |
| bobbo | ScottK: brilliant thanks | 19:37 |
| ScottK | bobbo: man dput, not brilliance | 19:38 |
| bobbo | ScottK: ive spent so long at the command line today, trying to figure something out for myself would be just too hard | 19:39 |
| ScottK | Understand. No problem. | 19:39 |
| sladen | bobbo: dput revu foo.changes | 19:40 |
| bobbo | ScottK: Thats it uploaded and registered thanks | 19:41 |
| ScottK | bobbo: You're welcome | 19:42 |
| * man-di is too dumb to use launchpad. how can I found the build logs for lucene2 again? | 19:46 | |
| DktrKranz | man-di, https://edge.launchpad.net/ubuntu/+source/lucene2/2.2.0-2 | 19:47 |
| man-di | DktrKranz: thanks | 19:48 |
| DktrKranz | :) | 19:48 |
| \sh | ScottK: do you have some time to check on something? I need some advice here | 19:58 |
| ScottK | Sure | 19:59 |
| \sh | ScottK: please dget http://ftp.de.debian.org/debian/pool/main/s/simage/simage_1.6.1-3.dsc (new release from debian) and check out aclocal.m4 | 20:01 |
| \sh | ScottK: you will find a libungif rule inside, but when you do a aclocal you won't see it again in the newly generated..even when you add the cfg/ dir (where some simage macros are)...I wonder if they patched it in manually ...or if it's automatically generated..I don't find anything in the cfg/ dir | 20:02 |
| ScottK | You may be assuming I know more about auto**** stuff than I do, but I'll have a look | 20:02 |
| \sh | ScottK: well, I only want another opinion...:) | 20:04 |
| ScottK | OK. Looking | 20:04 |
| \sh | ScottK: what I did was patching aclocal.m4 manually and recreate configure with a patched configure.ac to recognize the new macro...but this is not the normal way of doing things | 20:04 |
| amarillion | Hey there, if anyone has some spare time please review my package speed-game http://revu.tauware.de/details.py?package=speed-game | 20:06 |
| ScottK | \sh: The aclocal.m4 from the Debian package and the upstream tarball identical with the same time stamp. | 20:14 |
| ScottK | It hasn't changed since 2004, so if it's automatically generated, it's not very often. | 20:14 |
| \sh | ScottK: well, yeah...but it looks like that it was generated manually and not via aclocal | 20:17 |
| ScottK | \sh: I'm of the view that if your approach worked, then that's a good sign you're doing the right thing. | 20:18 |
| C10uD | hai all | 20:18 |
| \sh | ScottK: ok...will provide something and test | 20:19 |
| bobbo | Hey, if any MOTU's have some spare time could they check out http://revu.tauware.de/details.py?package=ckkern for me? | 20:19 |
| C10uD | sorry if i start with a question but i have some problems signing a debian package, this is the script i'm actually using http://pastebin.ca/848388 and seems like is building a package but are signed only the dsc and the changes files, any idea in where the problem is? | 20:20 |
| jdong | oh for crying out loud! every file has a different author/contributor and licensing terms!! | 20:22 |
| jdong | can debian/copyright just say see the headers of the source files or do I have to re-list the first 100 lines from each source file? | 20:22 |
| amarillion | head -N 100 * >> debian/copyright | 20:23 |
| amarillion | :) | 20:23 |
| jdong | haha | 20:23 |
| jdong | echo "Go figure it out yourself" >> debian/copyright? :D | 20:23 |
| Burgundavia | jdong: what package is that? | 20:24 |
| ScottK | jdong: debian/copyright MUST list all licenses. It's OK not to list all contributors if there are a lot of minor ones. | 20:25 |
| amarillion | C10uD, I think only those files need to be changed, right? | 20:25 |
| amarillion | s/changed/signed/ | 20:25 |
| C10uD | i don't really know ? :P | 20:25 |
| jdong | Burgundavia: mpeg4ip contains the ISO MPEG-4 reference software | 20:25 |
| C10uD | if i upload all the files to a ftp server | 20:25 |
| jdong | Burgundavia: each file in the reference software has a different set of contributors | 20:26 |
| C10uD | then the people get it's an unsecure package | 20:26 |
| amarillion | C10uD, the changes file contains md5sums of the rest | 20:27 |
| ScottK | jdong: Have fun. | 20:27 |
| amarillion | at least that is how I understand it | 20:27 |
| mok0 | I challenge any MOTU to find errors in http://revu.tauware.de/details.py?package=gpp4 | 20:27 |
| keescook | ScottK: You bet; it'll be in the archives in the next few hours. | 20:27 |
| ScottK | keescook: THanks. | 20:27 |
| keescook | np, thanks for the ping on it. | 20:28 |
| jdong | ScottK: thanks, will do :D | 20:28 |
| \sh | ScottK: looks like that I'm correct...actually it finds libgif :) | 20:28 |
| C10uD | :s i don't really know.. people add the the repository then they download the package but they get the "unsecure" message.. i'm likely a noob in debian packaging i tried to follow some guides but i really can't find the problem :s | 20:28 |
| amarillion | Well, somebody other than me should be better to answer this | 20:29 |
| * \sh would like to see upstream developers to use the documented, correct way of using autofoo crap | 20:29 | |
| amarillion | but I think your problem lies elsewhere. The fact that only the dsc files and changes files are signed is normal | 20:29 |
| C10uD | ok thanks anyway for your interest | 20:29 |
| C10uD | :) | 20:29 |
| \sh | strike..worked | 20:30 |
| amarillion | mok0, you're MOTU hopeful too right? | 20:30 |
| mok0 | amarillion: yup :) | 20:31 |
| amarillion | mok0, are you doing only new packages or updating existing packages as well? | 20:32 |
| mok0 | amarillion: I have mostly been doing new packages | 20:32 |
| amarillion | I'm wondering which of the two is better | 20:32 |
| amarillion | and easier | 20:32 |
| amarillion | Until now I've only tried new packages | 20:32 |
| C10uD | mok0: could you please take a look at my problem lines above? :P | 20:33 |
| mok0 | amarillion: you are more free to do your own stuff with new packages | 20:33 |
| * mok0 looks | 20:33 | |
| LaserJock | anybody know how to do spell checking in irssi/gnome-terminal/konsole-kde4 ? | 20:34 |
| C10uD | thanks :P | 20:34 |
| amarillion | so it's more fun... yeah, I can see that | 20:34 |
| \sh | ScottK: are fake syncs still acceptable? (new debian revision, not new upstream?) | 20:35 |
| mok0 | C10uD: What's the problem? | 20:35 |
| C10uD | the problem is that if i upload all the files (deb, targz, dsc, changes) to the ftp server then people get the "unsecure" message | 20:36 |
| LaserJock | amarillion: I find updating existing packages to be better for people just learning, but new packages tend to be a bit more rewarding | 20:36 |
| C10uD | even if seems like i've signed the dsc and changes files | 20:36 |
| mok0 | Isn't that because you have not signed the "Release" file? | 20:37 |
| broonie | C10uD: It's the Release file needs to be signed for apt | 20:37 |
| LaserJock | C10uD: the archive needs to be signed, Release file | 20:37 |
| LaserJock | darn, little late to the party ;-) | 20:37 |
| mok0 | :) | 20:37 |
| C10uD | well maybe :D i'm a noob packager | 20:37 |
| C10uD | :P | 20:37 |
| C10uD | how can i do that? :D | 20:37 |
| * C10uD points at the script | 20:37 | |
| LaserJock | C10uD: what are you using to make the archive? | 20:37 |
| C10uD | :P | 20:37 |
| mok0 | LaserJock: http://pastebin.ca/848388 | 20:38 |
| C10uD | >> http://pastebin.ca/848388 | 20:38 |
| C10uD | whoops | 20:38 |
| mok0 | Fastest again | 20:38 |
| mok0 | hehe | 20:38 |
| C10uD | :) | 20:38 |
| LaserJock | oh heavens | 20:38 |
| broonie | Use something like | 20:38 |
| minghua | LaserJock: What does irssi/gnome-terminal/konsole-kde4 mean? It's impossible to spellcheck in gnome-terminal without help from other programs, AFAIK. | 20:39 |
| mok0 | C10uD: I found the reprepro utility really nice to build a repo, it signs & everything | 20:39 |
| vud1 | hi! i need add a bash variable to the system (MOZILLA_FIVE_HOME=/usr/lib/firefox). in what file must my postinst file insert it? | 20:39 |
| LaserJock | minghua: I just need *something* that I can easily check spelling when I'm using irssi | 20:40 |
| LaserJock | mok0: falcon might be better for a new person | 20:40 |
| LaserJock | C10uD: I think you can do all that much easier and in fewer steps | 20:40 |
| C10uD | i think so, but as i said i'm a newb :P | 20:40 |
| mok0 | LaserJock: never heard of falcon, will check it out! | 20:40 |
| amarillion | vud1: what kind of package do you need it for? Can't you set it in a wrapper shell script for the app that needs it? | 20:40 |
| LaserJock | mok0: written by Seveas | 20:41 |
| amarillion | I think that i show the firefox run script itself does it | 20:41 |
| LaserJock | C10uD: for instance, to get rid of all the .svn use svn export | 20:41 |
| ScottK | \sh: They should be avoided. | 20:41 |
| ScottK | \sh: Is there a reason a regular sync can't be done? | 20:41 |
| LaserJock | C10uD: you really don't need to use dh_make either. Just copy the old debian/ files | 20:41 |
| * minghua knows a way to spellcheck in irssi/gnome-terminal, though it likely won't help LaserJock. :-P | 20:42 | |
| * C10uD writes the tips :P | 20:42 | |
| Seveas | dh_make sucks anyway :) | 20:42 |
| mok0 | Seveas: +1 from me | 20:42 |
| vud1 | amarillion: is a desktop package writed in python+gtk. i am not sure if i can use a shell script | 20:42 |
| LaserJock | C10uD: then instead of dpkg-buildpackage try using debuild (install fakefoot as well) then you don't have to sudo | 20:42 |
| LaserJock | Seveas: does falcon do archive signing? | 20:43 |
| Seveas | LaserJock, of course | 20:43 |
| C10uD | mm | 20:43 |
| vud1 | amarillion: my debian/rules uses my setup.py file to write the package.. so.. i dont know how to insert a shell script there (sorry for my english) | 20:43 |
| Seveas | LaserJock, building, archiving, signing, mirroring | 20:43 |
| LaserJock | C10uD: then use falcon to manage the repo | 20:43 |
| Seveas | LaserJock, apt-get data, maintainer mailing, linda/lintian | 20:43 |
| Seveas | it does it all :) | 20:43 |
| Seveas | I meant app-install data btw | 20:43 |
| LaserJock | that's what I thought | 20:44 |
| LaserJock | I just wanted to make sure | 20:44 |
| C10uD | ok i'll try these, thanks :P | 20:44 |
| mok0 | Seveas: sponsor, upload, fix bugs... | 20:44 |
| LaserJock | I need to install it and try it out | 20:44 |
| * mok0 too | 20:44 | |
| LaserJock | Seveas: I'll probably be writing about falcon in an upcoming book. I might be picking your brain here soonish | 20:44 |
| Seveas | LaserJock, I released 2.0.0 recently | 20:45 |
| mok0 | LaserJock: What book is that? Sounds interesting | 20:45 |
| Seveas | at 2.0.3 now, quite a few bugs are being found | 20:45 |
| LaserJock | Seveas: and that'll be in Hardy I presume? | 20:45 |
| Seveas | imbrandon is working on that | 20:45 |
| LaserJock | mok0: called Ubuntu Server in Action | 20:45 |
| Seveas | LaserJock, the 3rd reincarnation of it? | 20:45 |
| LaserJock | I'm writing chapters on packaging and repo management | 20:46 |
| LaserJock | Seveas: heh, something like that | 20:46 |
| Seveas | LaserJock, I've been reviewing parts of it since over 2 years :) | 20:46 |
| LaserJock | Hardy man, we've got to wait for an LTS ;-) | 20:46 |
| * mok0 has used Seveas' debomatic with pleasure | 20:46 | |
| Seveas | debomatic?! | 20:46 |
| amarillion | vud1, well you could put a shell script e.g. yourprogram.sh in your debian directory and add "install -m 755 yourprogram.sh /usr/bin/yourprogram" or something like that | 20:46 |
| Seveas | that's not mine for sure :) | 20:46 |
| mok0 | Ah, its not yours ;-/ | 20:46 |
| amarillion | vud1, and then the shell script should set the environment variable and then run the python script | 20:47 |
| vud1 | amarillion: mmm aha i understand | 20:47 |
| mok0 | Its DktrKranz's | 20:47 |
| LaserJock | minghua: would that involve learning an asian language? | 20:47 |
| LaserJock | :-) | 20:47 |
| DktrKranz | mok0, ah. Nice to know :P | 20:47 |
| vud1 | amarillion: so.. my package should insert 2 files in /bin but.. i think it is problematic. | 20:47 |
| amarillion | yeah maybe. It's just one solution | 20:48 |
| vud1 | amarillion: for example the *.desktop file is in the source and is common for all distributions, (not in the package) | 20:48 |
| amarillion | but the alternative of setting MOZILLA_FIVE_HOME globally is not great either | 20:48 |
| vud1 | amarillion: so the desktop file will launch the "program bin" | 20:48 |
| vud1 | amarillion: i see | 20:49 |
| amarillion | yeah, but you can patch the .desktop file to use the new script as well | 20:49 |
| amarillion | that is just one line change | 20:49 |
| mok0 | DktrKranz: it's great, I'd like it to be able to build both amd64 and i386 in one go | 20:49 |
| vud1 | could i do that? | 20:49 |
| vud1 | how can i pathc the desktop file? | 20:49 |
| mok0 | DktrKranz: but I haven't had time to play with it | 20:50 |
| DktrKranz | mok0, me too, but I need to understand cross-compilers or qemubuilder first | 20:50 |
| mok0 | DktrKranz: I think you can do it in a pbuilder on an amd64 machine | 20:51 |
| amarillion | vud1, yeah, but you need to read up a bit I guess | 20:51 |
| mok0 | DktrKranz: you can build i386 there | 20:51 |
| amarillion | better get a second opinion from someone more knowledgeable than me before you commit a lot of time to this | 20:51 |
| mok0 | DktrKranz: but you need to be able to fork the build | 20:51 |
| vud1 | thanks | 20:51 |
| DktrKranz | mok0, on amd64 I can build i386 as well, but my server is i386 only :( | 20:52 |
| amarillion | vud1, here is the documentation on patch system though: https://help.ubuntu.com/6.10/ubuntu/packagingguide/C/ps-chap.html | 20:52 |
| DktrKranz | (and it's down since days now...) | 20:52 |
| mok0 | DktrKranz: out of luck, huh? | 20:52 |
| vud1 | amarillion: i will read it now. thanks a lot | 20:52 |
| DktrKranz | yeah | 20:52 |
| DktrKranz | mok0, but yes, I wrote it just because I started to hate wanna-build | 20:53 |
| mok0 | DktrKranz: I think you can give up on running qemu if you don't have a cpu that does virtualization | 20:53 |
| amarillion | gotta go now. Good night | 20:53 |
| DktrKranz | mok0, I don't think an old celeron can :) | 20:53 |
| mok0 | DktrKranz: me neither :-) | 20:53 |
| mok0 | I'm getting a dual quadcore shortly :-P | 20:54 |
| DktrKranz | so, let's keep it warm until dancer releases something really cool | 20:54 |
| mok0 | DktrKranz: is he still developing pbuilder? | 20:55 |
| DktrKranz | I guess so | 20:55 |
| mok0 | DktrKranz: you can do it in debomatic | 20:55 |
| mok0 | DktrKranz: Allow the user to start 2 (or more) pbuilders | 20:56 |
| DktrKranz | mok0, it does actually | 20:57 |
| DktrKranz | it can run as many parallel builds as you wish | 20:57 |
| mok0 | DktrKranz: Aha? I have to play with it soon again, I guess | 20:58 |
| mok0 | DktrKranz: With different parameters for the pbuilder? | 20:58 |
| DktrKranz | mok0, yes. even two distributions (e.g. ubuntu and debian) | 20:58 |
| DktrKranz | as far as debootstrap is instructed :) | 20:58 |
| mok0 | DktrKranz: cool | 20:59 |
| DktrKranz | mok0, I'd like to show you it in action, but my server is down... | 20:59 |
| mok0 | DktrKranz: bah | 20:59 |
| mok0 | DktrKranz: my mobo is also dead on my own box | 21:00 |
| DktrKranz | my friend's mother destroyed his router during some housekeeping tasks, and now he's out of the net | 21:00 |
| mok0 | DktrKranz: That should teach him to do his own cleaning :-) | 21:00 |
| DktrKranz | Trust me, it's not advisable, last time I lost my mobile! | 21:01 |
| mok0 | DktrKranz: Can't you just call it? | 21:01 |
| DktrKranz | so I did :) | 21:01 |
| mok0 | DktrKranz: ... but it was out of power... | 21:02 |
| DktrKranz | no, it was out of reach... :P | 21:02 |
| === \sh is now known as \sh_away | ||
| === cprov is now known as cprov-out | ||
| === apachelogger_ is now known as apachelogger | ||
| geser | ScottK: pong | 21:21 |
| ScottK | geser: I was hoping you'd +1 \sh's MOTU application so dholbach can give it to him on his birthday tomorrow (later today for you). | 21:22 |
| geser | ScottK: will do | 21:23 |
| ScottK | geser: Thanks. | 21:23 |
| mok0 | No takers for my challenge? ;-) | 21:29 |
| jdong | ok, whew, that was a long copyright file | 21:31 |
| C10uD | ok thanks to your tips i'm improving dramatically the script :P now i have a question: is there a way i can tell debuild to build the package and sign the files without inserting my passphrase twice? e.g. a conf file or whatever | 21:38 |
| jdong | C10uD: use a gpg agent :) | 21:39 |
| C10uD | im messing with /etc/devscripts.conf and i've found an interesting # Do we cache the gpg passphrase by default? This can be dangerous! # DPKGSIG_CACHE_PASS=no | 21:40 |
| C10uD | :P | 21:40 |
| jdong | the warning might be iimportant :) | 21:41 |
| jdong | I'd rather have my passphrase cached by something designed to do so securely :) | 21:42 |
| C10uD | anyway that seems not working for me (fast test.. :P) | 21:42 |
| C10uD | so what was your advice? :P | 21:42 |
| jdong | hahaha | 21:43 |
| jdong | to have users configure a GPG agent | 21:43 |
| jdong | such as gpg-agent at the command line or the Seahorse GPG agent shipped by default with Ubuntu GNOME | 21:44 |
| C10uD | erm isn't that a waste of resources? :P | 21:44 |
| jdong | howtos on both readily available | 21:44 |
| jdong | well it's better than insecurely transporting a GPG passphrase | 21:44 |
| C10uD | well i don't need that security...yet :P | 21:45 |
| jdong | err... suit yourself? | 21:46 |
| slangasek | C10uD: so you'll be discarding your current gpg key and creating a new one when you decide you do need that security, right? | 21:47 |
| jdong | why even GPG sign something with an insecure key like that? | 21:47 |
| jdong | I wouldn't care if your packages are signed if that's how they're signed. | 21:48 |
| C10uD | mmm well maybe you are right lul | 21:48 |
| jdong | at this rate you might as well hard code it into your script and echo it into debuild! | 21:48 |
| C10uD | ok ok gotcha | 21:48 |
| jdong | :) | 21:48 |
| C10uD | i'll follow the REAL MEN's way! | 21:48 |
| C10uD | :D | 21:48 |
| C10uD | do i have to learn how to make up a gpg agent? :P | 21:50 |
| slangasek | C10uD: ... why are you claiming in your IRC client to be "mark shuttleworth"? | 21:51 |
| C10uD | i like trolling | 21:51 |
| jdong | slangasek: everyone likes to be mark :) | 21:51 |
| slangasek | what an odd thing | 21:51 |
| C10uD | do you really think i'm mr.ubuntu asking for packaging? :S | 21:52 |
| jdong | C10uD: I'd like to talk about my raise.... | 21:52 |
| Fujitsu | slangasek: We get them a bit. | 21:52 |
| C10uD | lul | 21:52 |
| * jdong grins at Christian Marillat's reply to his mail | 21:55 | |
| jdong | slangasek: everyone likes to be mark :)Thanks you very much, I hate to write debian/copyright file. | 21:55 |
| jdong | oops | 21:55 |
| jdong | GRR | 21:55 |
| jdong | Dear Synaptic touchpad gods: | 21:56 |
| jdong | Two-finger scroll and two-finger middle click do not mix well. | 21:56 |
| jdong | CC: Santa, Easter Bunny. | 21:56 |
| C10uD | lol | 21:58 |
| === TreMobyl is now known as Solarion | ||
| nxvl_work | how often are the CC meetings taking place? | 22:01 |
| ScottK | nxvl_work: Nominally every 2 weeks, but it varies. | 22:02 |
| persia | nxvl_work: Generally about once a month, but sometimes they are missed due to various conflicts. Longest span I've seen was about 6 weeks. | 22:03 |
| nxvl_work | thnx | 22:03 |
| nxvl_work | persia: i emailed you, did you received it? | 22:04 |
| persia | nxvl_work: Likely, but I haven't read it yet. Looking now... | 22:04 |
| === azeem_ is now known as azeem | ||
| civija | guys, how can I found out does slrnface package has ubuntu maintainer or is it synced from debian? | 22:13 |
| geser | !info slrnface | 22:14 |
| ubotu | slrnface: shows X-Faces from a newsposting on an X11 terminal emulator. In component universe, is optional. Version 2.1.1-6 (gutsy), package size 25 kB, installed size 104 kB | 22:14 |
| jpatrick | civija: look at it's version if it has -*ubuntu* it has ubuntu changes | 22:14 |
| geser | civija: looks like it was taken directly from Debian | 22:14 |
| jdong | am I correct in thinking that for a SRU debdiff, one should NOT adjust the Maintainer field for the maintainer spec? | 22:14 |
| jdong | or is that trivial enough that it's acceptable to put it in too? | 22:14 |
| civija | geser: aha, tnx | 22:15 |
| civija | so i could take slrnface and package it for universe? | 22:15 |
| ScottK | jdong: For Dapper/Edgy no, for Feisty + yes. | 22:15 |
| geser | jdong: for >= feisty (?) it's ok to change the maintainer and also done | 22:15 |
| jdong | ScottK: geser thanks :) | 22:15 |
| geser | civija: what do you mean? it's already in universe | 22:16 |
| civija | geser: i'm thinking of becoming motu and i would like to maintain slrnface | 22:16 |
| slangasek | ScottK: so following the maintainer spec for dapper in an SRU would be cause for rejection? | 22:17 |
| ScottK | slangasek: Yes because the tool chain for Dapper has never been tested with modified maintainers and the new fields | 22:18 |
| geser | slangasek: the tools are not tested if they work with the maintainer change | 22:18 |
| slangasek | ScottK: ah, I see | 22:18 |
| ScottK | Dapper and Edgy | 22:18 |
| geser | civija: have you tried to work with the Debian maintainer on this package as we try to use the Debian packages where possible? | 22:19 |
| mok0 | Isn't the whole point of using rfc822 files that unknown tags are ignored by the tools? I thought there was a forward compatibility in that | 22:19 |
| emgent | hello people ;) | 22:20 |
| civija | geser: yes but didn't get any response | 22:20 |
| civija | btw slrnface doesn't work in ubuntu without litle patch | 22:20 |
| ScottK | mok0: Thats the theory, but for an SRU we don't want any risks that are avoidable. | 22:20 |
| civija | i can try to contact him again ... | 22:21 |
| mok0 | ScottK: I understand | 22:21 |
| geser | civija: prepare a debdiff then and let is sponsor into Ubuntu | 22:21 |
| civija | geser: ok, tnx | 22:21 |
| sommer | /wc/ | 22:22 |
| sommer | woops | 22:22 |
| emgent | DktrKranz, in hardy there is phpbb2 or phpbb3 ? | 22:23 |
| DktrKranz | emgent, IIRC phpbb2 | 22:24 |
| emgent | uhm ok cool. vulnerable too | 22:24 |
| emgent | i go to write diff | 22:24 |
| slangasek | civija: generally you're more likely to get a response from Debian maintainers if you're sending them concrete patches than if you just say you want to help, particularly on smaller packages; I guess you did the latter, since I don't yet see any bug reports or patches from you in the Debian BTS? | 22:27 |
| civija | slangasek: yes you are right | 22:29 |
| civija | i'll make a debdiff | 22:29 |
| ScottK | geser: Thanks. | 22:31 |
| ScottK | \sh_away: You got +1 from all the MC, so now we just wait for dholbach to make it official on your birthday. | 22:31 |
| StevenK | \sh is coming back? | 22:32 |
| ScottK | Yes. He applied for MOTU again. | 22:32 |
| geser | StevenK: if \sh doesn't pull back his application he'll be a MOTU in a few hours again | 22:36 |
| StevenK | Heh | 22:36 |
| mok0 | Good then he can review my perfect package | 22:36 |
| TheMuso | mok0: What is this package that you so boldly claim is "perfect"? | 22:37 |
| mok0 | gpp4 on revu | 22:37 |
| TheMuso | Ok, I'll take a look. | 22:38 |
| mok0 | (he took the bait ;-) | 22:38 |
| persia | mok0: Are you sure I can't find anything if I looked? | 22:38 |
| mok0 | persia: you'd have to be good | 22:38 |
| ScottK | persia is good, so watch out. | 22:39 |
| mok0 | ScottK: yes, but I am upstream :-) | 22:39 |
| ScottK | That helps, but is no guarantee. | 22:39 |
| mok0 | ScottK: heh | 22:40 |
| TheMuso | mok0: For a start, thers the out of date standards version. | 22:40 |
| persia | mok0: Your debian/copyright doesn't match http://lists.debian.org/debian-devel-announce/2006/03/msg00023.html :P Specifically doesn't mention the debianisation, fails to have licensing for the packaging, and uses globs. | 22:40 |
| persia | Anyway, I don't have time for a full review now. | 22:40 |
| mok0 | http://wiki.debian.org/Proposals/CopyrightFormat | 22:41 |
| mok0 | persia: that's what DD's want now | 22:41 |
| TheMuso | mok0: There are 2 * in your changelog. | 22:42 |
| TheMuso | * * Initial release (LP: #176211) | 22:42 |
| * persia notes that monsterz was a poor choice as an example | 22:42 | |
| mok0 | TheMuso: arrrgh | 22:43 |
| TheMuso | mok0: Since your address is not @ubuntu.com, you need an XSBC-Original-Maintainer field. | 22:43 |
| mok0 | TheMuso: I am upstream | 22:43 |
| TheMuso | No, you still need it, for packaging reasons. | 22:43 |
| persia | mok0: the ftp masters reject FAQ still lists the other. Can you pass a pointer showing me that this proposal was accepted by a significant number of DDs? | 22:43 |
| TheMuso | The package won't build oterhwise. | 22:43 |
| TheMuso | otherwise | 22:43 |
| mok0 | persia: no, I don't have such documentation | 22:44 |
| mok0 | persia: only from 1 | 22:44 |
| persia | Further, the w.d.o page says "This is not a proposal to change the policy in the short term.", so I don't think it's started yet. | 22:44 |
| mok0 | persia: "to prepare for future changes, you might as well.." | 22:44 |
| * persia still wants to know who did the initial packaging for a given package. | 22:45 | |
| mok0 | persia: that's in changelog | 22:46 |
| persia | mok0: Not reliably, as long as it remains best practice to have a truncated changelog for first introduction. Take a look at some of the packages first in Ubuntu and later brought into Debian. | 22:46 |
| mok0 | persia: ..and in copyright: Files: configure,ac .. debian/* | 22:47 |
| mok0 | persia: I don't like the idea of truncating the changelog for those packages that were first in ubuntu | 22:49 |
| mok0 | persia: like rewriting history... | 22:49 |
| TheMuso | mok0: In libgpp4-0.install, you know you can use wildcards? SO something like | 22:50 |
| TheMuso | usr/lib/libgpp4.so.* | 22:50 |
| persia | mok0: True, but I don't like preserving 37 upstream changelog revisions and small bumps before the package attracted the attention of a distro. There's surely a happy balance somewhere, but I'd prefer the debian/copyright header until one is found. | 22:50 |
| * nxvl *HUGS* persia | 22:50 | |
| mok0 | TheMuso: yeah, ok | 22:50 |
| mok0 | persia: granted, I don't know why they didn't keep that. | 22:51 |
| mok0 | persia: I'll put it back | 22:52 |
| persia | mok0: If you want to put it in a format that is easily machine-readable, I've no objections :) | 22:52 |
| mok0 | persia: Of course, I can make my own tags legally in rfc822 :-) | 22:52 |
| TheMuso | mok0: You don't need dh_link in debian/rules, as you don't need to create any symbolic links other than the library ones do you? Those are automatically created. | 22:52 |
| mok0 | X-First-Packager: blah blah | 22:53 |
| mok0 | TheMuso: that's right | 22:53 |
| mok0 | TheMuso: 1-0 to you | 22:53 |
| TheMuso | mok0: Do you want me to document my comments on the revu page? | 22:54 |
| mok0 | I think I have them in the scrollback | 22:54 |
| TheMuso | Ok. | 22:54 |
| mok0 | TheMuso: I'll fix them right away, and persia's comments on debian/copyright, too | 22:55 |
| TheMuso | ok | 22:55 |
| * mok0 hugs TheMuso and persia | 22:55 | |
| TheMuso | mok0: Oh yeah, you also might consider making sure your library has a soname, and you ahve a libgpp4-0.shlibs file to specify the minimum library version for more tight linking for other programs | 22:57 |
| TheMuso | s/ahve/have/ | 22:57 |
| mok0 | TheMuso: the major and minor version are both 0 | 22:57 |
| TheMuso | mok0: I know, but when symbols change, new versions of the library need to indicate to other packages somehow that they ned to depend on a newer version when being linked against. | 22:58 |
| mok0 | TheMuso: Isn't .shlibs created by dh_makeshlibs? | 22:58 |
| TheMuso | mok0: I suggest you check out the school session next thursday that slangasek and another MOTU are doing about library packaging. | 22:58 |
| mok0 | TheMuso: Yeah, I have it on my calendar already! | 22:59 |
| TheMuso | mok0: Yeah it does, but I've seen instances where packages have had them in the debian dir. | 22:59 |
| mok0 | TheMuso: isn't that a problem only if I bump the soversion? | 23:00 |
| TheMuso | mok0: I think so. I'll be attending that session as I need a few things cleared up myself. At this point, I'm going only on experience. | 23:00 |
| mok0 | TheMuso: I won't do that unless the API changes | 23:00 |
| slangasek | if your symbols have changed, *not* bumping the soversion is a problem... | 23:00 |
| mok0 | slangasek: that would be an ABI change, right? | 23:01 |
| slangasek | yes | 23:01 |
| mok0 | But this is the first public version of the library, so I can't see how problems could arise until I release another version | 23:02 |
| mok0 | Say I change the API/ABI, I would bump the major sonumber and also change the name of the package to libgpp4-1 | 23:03 |
| C10uD | is there a way i can hardcode the dput ftp password somewhere? :P (now, i can do this uh!) | 23:03 |
| slangasek | mok0: right | 23:03 |
| mok0 | That means applications using version 0 and version 1 could live side-by-side | 23:05 |
| mok0 | So I really don't understand what TheMuso talked about before, with the .shlibs file being necessary to write | 23:06 |
| mok0 | C10uD: ~/.dput.cf ? | 23:06 |
| C10uD | i can't find the appropriate option in the manpage | 23:06 |
| mok0 | C10uD: Is it a local ftp site? | 23:07 |
| slangasek | mok0: something has to generate the .shlibs; in some cases it's appropriate to let dh_makeshlibs do this on its own | 23:07 |
| C10uD | nope | 23:07 |
| mok0 | slangasek: yeah that's what I thought | 23:07 |
| slangasek | mok0: in other cases, you need to provide more information; such as when a new .0 version of the lib has added new symbols | 23:07 |
| mok0 | slangasek: I need to think about that | 23:08 |
| nxvl | persia: i have edited mi wiki page, can you take a look and tell if if thats was you where talking about? | 23:09 |
| mok0 | slangasek: my understanding is that any change in API/ABI requires you to bump the major sonumber | 23:09 |
| mok0 | slangasek: but if you release a new version, for example with a bugfix, you bump the minor sonumber | 23:10 |
| slangasek | mok0: any ABI change that breaks compatibility with existing apps built against a previous version of the lib requires you to bump the major sonumber | 23:10 |
| mok0 | slangasek: I think we're on the same page... | 23:10 |
| slangasek | mok0: well, but you said "any change in API/ABI" - adding a new function is a change to the ABI, but it's a backwards-compatible one | 23:11 |
| slangasek | so the soname doesn't change, but you need to make sure any apps built against that new function have a versioned dependency on the library | 23:11 |
| slangasek | which is where you need to give that version information to dh_makeshlibs | 23:11 |
| mok0 | slangasek: got it | 23:12 |
| mok0 | slangasek: so creating a .shlibs file sort of helps you keep track | 23:13 |
| === Ubulette_ is now known as Ubulette | ||
| mok0 | slangasek: ... but then I have to get rid of dh_makeshlibs, yes? | 23:14 |
| slangasek | mok0: in the normal case, you can still use dh_makeshlibs and just pass a -V option to it | 23:16 |
| mok0 | slangasek: ok, I will look at the man page. | 23:16 |
| mok0 | slangasek: thanks for a useful discussion. I will try to attend the motu school session next Thursday | 23:20 |
| slangasek | mok0: ok, cheers | 23:20 |
| === bigon` is now known as bigon | ||
| TheMuso | I must admit, I think all prospective MOTUs should know this stuff. I'm not affraid to admit I don't entirely, and think this needs addressing, as MOTUs are expected to touch many packages, libraries included. | 23:25 |
| slangasek | I think this definitely falls in the realm of "nice if everyone knew, not practical for that to happen any time soon" | 23:26 |
| TheMuso | True, but I think something could be done for future MOTUs. | 23:26 |
| slangasek | well, I think that would mean setting a much higher bar for MOTU admission | 23:28 |
| bobbo | On the REVU wiki page it says i need an elgamal secondary key to recover my password, how do i go about creating one? | 23:28 |
| TheMuso | slangasek: Well, IMO so be it. | 23:29 |
| slangasek | hmm, you want the technical bar for MOTU to be higher than the technical bar for DDs? | 23:31 |
| * TheMuso wasn't aware that this wasn't covered as something that dds should know. | 23:32 | |
| Fujitsu | slangasek: In some ways I think it should be, but perhaps not. | 23:32 |
| slangasek | TheMuso: most DDs are pretty clueless when it comes to packaging libraries, both long-time DDs and NMs. When I went through the NM queue, I remember there being a warning about not picking a library for my first package because libraries are Advanced. :) | 23:36 |
| TheMuso | slangasek: Thats something that the library packaging guide mentions. | 23:37 |
| TheMuso | http://www.netfort.gr.jp/~dancer/column/libpkg-guide/libpkg-guide.html | 23:37 |
| * slangasek makes warding signs | 23:37 | |
| TheMuso | huh? | 23:37 |
| slangasek | TheMuso: I'm not happy with the library packaging guide; as you can see it's an independent work by one particular DD, and a few of the recommendations are insane | 23:38 |
| TheMuso | Warding signs against the guide? | 23:38 |
| TheMuso | slangasek: Ah. I must say the guide does ramble a bit as well, which makes it hard to read./ | 23:39 |
| TheMuso | SO fair enough, I won't be so quick to recommend it then. | 23:39 |
| slangasek | yes, warding signs against the guide | 23:39 |
| Kmos | http://ftp-master.debian.org/~joerg/removals/removals.rss | 23:39 |
| Kmos | if someone interested | 23:39 |
| slangasek | because I usually hear it mentioned in the context of "of course we changed the -dev package name, the library packaging guide said to" | 23:39 |
| StevenK | Ugh. | 23:40 |
| StevenK | BAD. | 23:40 |
| TheMuso | slangasek: Has th ere been any thought o making an official packaging guide, if there isn't one already? | 23:40 |
| slangasek | TheMuso: sure, I've thought about it a lot ;) | 23:40 |
| slangasek | doing anything about it is another question entirely | 23:40 |
| TheMuso | for libraries | 23:40 |
| TheMuso | yep of course | 23:40 |
| slangasek | unfortunately, I myself am behind the times; buxy has done a great job of implementing dpkg-gensymbols, but while I understand all the concepts I haven't had a chance to use it yet in any of my own libs | 23:44 |
| TheMuso | Nice. | 23:44 |
| minghua | Yeah. I think I just saw a Debian package having a non-back-compatible ABI change without SONAME bump yesterday. | 23:48 |
| === _[PUPPETS]Gonzo is now known as [PUPPETS]Gonzo | ||
| minghua | The problem is that most upstream is clueless about SONAMEs too. | 23:48 |
| rexbron | Question: CDBS does not seem to handle .desktop files, does someone know a way to do this? | 23:49 |
| slangasek | yes, that is a problem; distros bear the burden of ensuring ABI consistency as part of their integration work, because many lib upstreams are ignorant or don't care | 23:49 |
| slangasek | but that's exactly why -gensymbols is so valuable once we've grown a community of experts around its use, because it makes checking for ABI breakage an automatic thing | 23:52 |
| * minghua nods | 23:53 | |
| minghua | I need to learn symbol versioning and the dpkg-gensymbols tool and apply them to my package. | 23:54 |
| mok0 | What is the correct Standards-version for hardy? | 23:54 |
| minghua | Although I'm still not sure how good they will work for C++ libraries. | 23:54 |
| Fujitsu | mok0: 3.7.3 at the moment. | 23:55 |
| mok0 | Fujitsu: thx | 23:55 |
| minghua | Right, so this SONAME breakage is real, I can crash my application with it. | 23:59 |
| * minghua goes to report a bug. | 23:59 | |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!
