slangasekwhich is a poorly-named package; if it were being done from scratch, it should be libgtkglext-1.0-0 or libgtkglext-x11-1.0-0, according to taste00:00
DaveMorrisso surely the c++ bindings should be similar00:00
sistpotyDaveMorris: now you've completely confused me, as I can't find a gutsy package (only an edgy one)#00:00
slangaseksistpoty: he's referring to gtkglext, not gtkglextmm00:00
HighNothanks to DaveMorris and the other reviewer there is a new version at revu: http://revu.tauware.de/details.py?package=blueproximity - Reviewers wanted :-)00:00
sistpotyah, ok00:00
DaveMorrisslangasek: yeah, the C package should also be sorted out at the same time IMO (and poss the other bindings available )00:01
slangasekDaveMorris: it shouldn't; once a library runtime package name has been chosen, even if it's poorly chosen, changing it is worse than leaving it alone00:02
DaveMorristhen shouldn't we stick with the name that was in dapper/edgy ?00:02
slangasekDaveMorris: but if the soname changes (are you packaging a new version of gtkglext with a new soname, too?), then the package name can be fixed in the process since the name has to change anyway00:02
slangasekDaveMorris: no, because it's *not* the same version of the library. sistpoty just pointed out the difference in soname00:02
DaveMorristhe version just changes from -1.0 to -1.200:03
slangasekso the package name *must* be changed, and therefore it can be fixed however is most appropriate as long as it doesn't collide with another package name00:03
slangasektherefore, libgtkglextmm-x11-1.2-0 or libgtkglextmm-1.2-0, whichever you find more acceptable00:04
slangasekbut gtkglext's package name should change *only* if a new version is being packaged that changes the ABI00:04
geserHighNo: you need a verbatim copy of the license in the orig.tar.gz (just listing it in debian/copyright is not enough). without a license file it will be rejected by the archive admins.00:04
DaveMorrisslangasek: but the others have jumped to 1.2 in the gutsy->hardy change00:05
DaveMorrisfrom 1.000:05
DaveMorrisor is it to late to fix that now00:05
HighNogeser: so upstream has to include it? If so what would be the quickest and easiest way?00:06
slangasekDaveMorris: in that case, the gtkglext binary package names are currently wrong and must be changed00:06
slangasek(and, who uploaded a new upstream version of this library without understanding that?)00:07
DaveMorriswhich will cause alot of dependices to break00:07
DaveMorrissounds fun00:07
sistpotyDaveMorris: apart from the soname stuff, your debian/copyright doesn't name all licenses used (e.g. ./examples/trackball.c)00:07
DaveMorristhanks sistpoty00:08
slangasekDaveMorris: uh, no - gtkglext has an upstream version bump to 1.2, but the soname did not change00:08
DaveMorrismy mistake,00:09
geserHighNo: yes, as somebody who just downloads the .orig.tar.gz has to know the license.00:09
DaveMorrisso you want 2 binaries00:10
geserHighNo: can you easily redo a new upstream tarball with the added license file?00:10
HighNogeser: yepp00:10
HighNogeser: Can I do that without a version number change in upstream?00:11
geserHighNo: I guess so, but I'm not sure if it might confuse others.00:12
geserand I also don't know if sourceforge let you00:12
slangasekDaveMorris: what do you mean, two binaries?00:12
DaveMorrisone for each so object00:13
slangasekthat's not absolutely necessary00:13
* DaveMorris makes my job easier :)00:13
slangasekit's the "safer" way to package it, but there are plenty of other library packages that include multiple libraries from the same upstream source00:14
slangasekgtk+2.0 is one of those, gtkglext is another...00:14
sistpotywhich means that one ABI version certainly won't change with the ABI version of the other shared object, right?00:15
geserHighNo: about the "/usr/bin/env python/" vs. "/usr/bin/python" issue: I've seen several upstream use env but it has the disadvantage that you can't add addtional options to the shebang. So it is usually changed in Debian/Ubuntu packages.00:17
HighNogeser: ok, thanks for the info00:17
slangaseksistpoty: that's what it implies if you package them together that way, yes00:17
geserHighNo: what problems did you have with blueproxmity being a python script and the desktop file?00:19
HighNogeser: the proximity.py mentions the GPL - but that's not enough I guess?00:19
geserHighNo: unless you put there the whole GPL, I guess it's not00:20
HighNogeser: don't know anymore, I think that was an old problem where I had to find out the directory with the data files that was formerly solved via the startup script. Now the path is included in the path, so that won't count against it anymore.00:21
HighNogeser: so all I should do is include the complete GPL source? wouldn't I have to remove that file for packaging again since there should be a reference to the license files as in control(?)00:22
sistpotyHighNo: the GPL (at least 2, haven't checked 3 yet) has the restriction, that the complete license text must accompany the source00:23
sistpotyHighNo: hence a source package (which s.o. can download individually) must contain the GPL00:24
geserHighNo: the .orig.tar.gz must have a verbatim copy of the GPL (in the right version) but you shouldn't install it in the binary package (there is the reference to /usr/share/common-licenses enough)00:24
HighNosistpoty: ouch, I did not know that. but again - wouldn't I have to delete it when packaged?00:24
HighNogeser: ok, can I do the file deletion with a patch?00:25
sistpotyHighNo: for a *binary* package: yes. there you can refer simply to the already installed version of the GPL under /usr/share/common-licenses00:25
geserHighNo: a file deletion isn't needed, simply don't list it in debian/blueproximity.install (you don't have to put all files from a source package into binary package)00:26
HighNogeser: ah, I forgot that - has been to long ago I created it I guess00:27
HighNowould that license file need a starter like "all package contents is licensed by this license: ...GPL2 original text..."00:28
geserHighNo: just look how the others package on revu do it, e.g. http://revu.ubuntuwire.com/revu1-incoming/termlauncher-applet-0802021100/termlauncher-applet-0.0.3/00:30
geserCOPYING is the GPL license00:30
geserHighNo: or http://revu.ubuntuwire.com/revu1-incoming/libini4j-java-0802011440/libini4j-java-0.2.6/ where License.txt is the Apache license00:31
geserHighNo: there is no requirement on how the file is named00:32
HighNogeser: OK, I'll stick to the examples00:35
DaveMorrissistpoty: if you had 5 mins spare, I've uploaded a new version00:52
DaveMorrisbut I'm off to bed now.00:52
sistpotyDaveMorris: give me a few minutes, and I'll take a look00:52
sistpotygood night DaveMorris00:52
HighNogn and thx DaveMorris00:53
HighNogeser: a new version is on the way, just the usual ten minute delay...00:59
HighNogeser: I would like to hold the new upstream version until we hav all upstream problems eliminated01:00
HighNogeser: so once upstream seems allright I'll upload it to sourceforge01:01
HighNodoes anybody know how big the userbase for feisty and gutsy still is?01:02
HighNoor should I let backports handle the old versions?01:02
HighNobackporting should be a breeze...01:03
RAOFBackports is the only way that new packages get into older releases.01:04
jdongwhy was I just pinged 3 times within the past minute?01:05
* jdong looks up01:05
sistpotyjdong: you're getting pinged if s.o. mentions backports?01:06
jdongsistpoty: I must.01:06
jdongsistpoty: I had a lot of fun with regex one day01:06
* sistpoty imagines01:07
HighNoso you  are the backports man? :-)01:07
jdongPici: not that kind of fun.01:07
jdongPici: that kind of fun encourages me to fix flashplugin-nonfree01:07
HighNoanybody up for a hopefully final review: http://revu.tauware.de/details.py?package=blueproximity01:08
pochustgraber: libfprint is now in debian experimental. You may want to request a sync for it ;) http://packages.debian.org/source/libfprint01:13
* sistpoty goes to bed now01:18
sistpotygn8 everoyne01:18
=== RAOF_ is now known as RAOF
blueyedIs it ok to push changes to http://bazaar.launchpad.net/%7Eubuntu-dev/ubuntu-dev-tools/trunk/ directly - or should I offer the branch? These are changes/improvements to requestsync.. (bug 190351)01:58
ubotuLaunchpad bug 190351 in ubuntu-dev-tools "requestsync crashed with EOFError, when ending input with ctrl-d" [Undecided,New] https://launchpad.net/bugs/19035101:58
blueyedThe branch is locked currently anyway (since > 200 hours).. :/01:59
bddebianHeya gang02:40
RAOFHowdie bddebian.02:40
persiahey bddebian02:41
emgentheya people :)02:41
bddebianHi RAOF, persia, emgent02:42
* ScottK cheers.02:55
ScottKmok0 now TIL courier.02:56
bddebianHeya ScottK02:57
ScottKheya bddebian02:57
ScottKSo I decided to go ahead and sync your testresources nmu over my fix.  You fixed it upstream, so you win.02:58
bddebianmwuhahaha ;-P02:58
RAOFWhat do people think of the Debian machine-readable copyright format proposed here?  http://wiki.debian.org/Proposals/CopyrightFormat03:10
blueyedRAOF: I like it and have used it for the two packages I've created (jedit and tvbrowser on revu)03:11
RAOFIn particular; I think it looks like a reasonable idea, but I'd like to know how people would react to packages hitting the archive with deban/copyright in that format.03:12
ScottKRAOF: No one I know of has objected to it.03:12
* ScottK would get grumpy if people started insisting on it though. It's not policy.03:12
RAOFdebian/copyright should be partially automateable.  Does anyone have scripts in that direction (scraping headers for copyright holders, etc)?03:16
* jdong kneels in front of the confessions window.03:17
jdongDear MOTU deities, I have sinned.03:17
jdongSomeone was trying to add a launcher for gnome-terminal that starts irssi with a custom profile03:17
jdong--window-with-profile looks promising but starts an empty window along with the desired one03:18
jdongI suggested the hack: gnome-terminal -e /bin/true --window-with-profile=irssi -x irssi03:18
* jdong gets up and waits for enlightenment03:18
jdong(i.e. a valid solution)03:18
persiaRAOF: You might start looking at suspicious-source to see if that might help.  On the other hand, licensing is hard, especially because of whitespace issues.03:25
tuxmaniacheya. morning all04:08
=== asac_ is now known as asac
slytherinHi, I want to make a sync request for libxstream-java, but it's dependency is still stuck in 'new' queue so I can't test if it builds. And libxstream-java is needed by groovy which is in depwait.06:49
=== allan__ is now known as Hit3k_
=== Hit3k_ is now known as Hit3k
persiaslytherin: You might try manually building in a chroot to test the build.  Alternately, wait a bit (although only a few more archive-admin days before feature freeze).06:52
HighNogood morning!07:23
pochujdong: will you forward the irssi manpage patch upstream? :)09:05
=== LifeHacker is now known as tuxmaniac
=== allan__ is now known as Hit3k
persiaSo glancing at REVU, I'm seeing an explosion in package updates.  Packagers packaging updates should submit the diff.gz in a bug to the sponsors teams.10:02
LucidFoxSo once again, my attempt to migrate from KMail to Thunderbird got botched...10:17
persiaLucidFox: Which were the packages on REVU you wanted to advocate again?10:19
HobbseeLucidFox: why?10:20
HighNooh, you wanted to advocate blueproximity, right?10:20
HighNo(sorry, just a try :-) )10:20
LucidFoxpersia> I didn't review them, so I wouldn't advocate them on the spot - but those are qdevelop and gtkglextmm10:21
Hobbseeblueproximity looked reasonably good to me, although the lack of python-{central/support} stuff looked suspicious10:22
persiaLucidFox: OK.  If you happen to review them, complain here looking for an advocate.10:22
persiaHobbsee: ?  It lists python-support.10:23
gesergood morning10:23
Hobbseepersia: it didnt when i looked, i didn't think10:24
Hobbseeor i missed it10:24
persiamorning geser10:24
persiaHobbsee: Ah.  That might be it.10:24
HighNoHobbsee: right, python-support - but there's not too much stuff in there as it does not create seperate modules for other python stuff...10:24
HighNoHobbsee: might be, I uploaded it several times as one can see...10:24
Hobbseeah right10:24
* Hobbsee only saw the first, iirc10:24
persiaNightrose: Do you happen to use 2ch, or were you just providing advice for kita2?10:27
Nightrosepersia: ? are you sure you mean me?10:28
HighNoJust another question regarding updated translations. If my package is being approved, is there a quick way to update the translations only? I mean I could implement them as a patch and note then when puting the package on revu. Would that be the correct way? Feature Freeze means there will be no more packages accepted right?10:28
persiaNightrose: No.  The pointer is second hand (from comments in http://revu.ubuntuwire.com/details.py?package=kita2).  Based on your response, I suspect you don't use 2ch :)10:29
geserHighNo: FF means that no new packages and no new upstream version will be accepted without an exception10:30
persiaHighNo: Once a package is accepted, updates are submitted as patches in bugs.  Translation updates are welcome until quite late in the cycle.10:30
Nightrosepersia: hehe ok ;-) he just asked me to have a look at the description and check for problems10:30
persiaNightrose: OK.  I just was hoping you were both a KDE user and 2ch user, as while I think the package should be in the archives, this is from altruism, and I haven't actually tested in the ideal target environment :)10:31
Nightrosewell I am a kde user but hinestly have no idea what ch2 is ;-)10:32
persiahttp://2ch.net/ but it's optimised for use on phones, so the interface leaves something to be desired.  No worries :)10:35
HighNopersia: that's good news as I started a translation day on many channels and got some new interesting languages.10:37
persiaHighNo: Great!  More languages is better :)10:38
LucidFoxpersia> commented on qdevelop10:39
persiaLucidFox: Added to my queue.  I'll likely ACK then.  Thanks.10:40
LucidFoxpersia> and speaking of DEB_DH_INSTALL_SOURCEDIR, I now understand the point - it has the same effect as the --sourcedir parameter of dh_install. Basically, it allows to drop debian/tmp/ prefixes from debian/*.install.10:57
persiaLucidFox: Yes, precisely.  Personally, I don't like it because it's not an easy trick to learn, and only saves a few keystrokes, without gaining the advantages of cross-package utility that the use of something like CDBS or debhelper provides.10:59
=== Seveaz is now known as Seveas
* DaveMorris requests are review of http://revu.tauware.de/details.py?package=gtkglextmm (should be there now)11:12
crevettehello there11:14
crevetteI need some help with a package I'm building11:14
crevettelintian returns "W: obex-data-server: package-contains-empty-directory usr/sbin/" and a dpkg -c my.deb shows me it creates an empty /usr/sbin11:15
crevettewhy ?11:15
persiacrevette: Because you used dh-make, and didn't clean up all the template files.  Your issue is with debian/dirs11:16
crevetteah thanks11:16
persiacrevette: Getting that error typically means that there are other issues with the templates as well.  Please check carefully to make sure you've edited all the files, removed any you don't need, and are sure you want to keep the parts remaining.11:19
crevettepersia: okay thanks, I'm running over the other files11:20
pochuAnyone from the MOTU Science team? You might be interested in bug 10547311:39
ubotuLaunchpad bug 105473 in ubuntu "[needs-packaging] Extrema" [Wishlist,Confirmed] https://launchpad.net/bugs/10547311:39
affluxlibotr (main) has a hidden dependency on libgcrypt11-dev, which was fixed in debian. Is a missing dependency sufficient for a sync request?11:52
Georgex96j15qi4gavailable they it's come is part VR of and be completely the could its ultimate in concept likely11:53
Georgex96j15qi4gnever experience part time. they that term fancy helmet of used than why be you're to and could11:53
Georgex96j15qi4ghas helmet after now. doesn't surge VR in available, popular they is user the can there", of films,11:53
Georgex96j15qi4gunable made more from more in the than its had still reality but virtual that the actually actually11:53
Georgex96j15qi4gand the basic to all vision hard way Demise they saw that budget? the of stereoscopic industry on11:53
jpatrick!ops | Georgex96j15qi4g11:53
Georgex96j15qi4ghelmet. imagine, association tell virtual or with pop virtual a be this culture without virtual handshake get watching11:53
ubotuGeorgex96j15qi4g: Help! Hobbsee, Riddell, sladen, fbond, mneptok, gnomefreak, Seveas, dholbach, elkbuntu or PriceChild!11:53
Georgex96j15qi4gyou of may virtual their idea term so have technology even it, a that As include 1990s, watching11:53
Georgex96j15qi4ga commercially up helmets It It's senses term's reason so the of can close they the So course,11:53
Georgex96j15qi4ggoal without senses helmets a the systems didn't available much let's take even reason embracing Of in us11:53
persiaafflux: Depends.  If there are other useful fixes as well, surely.  If there are other things that you don't think should sync, then just a patch would work.  Generally syncs are preferred if they don't cause any issues.11:54
affluxpersia: there were no other changes11:54
HobbseeSeveas: them again?11:54
persiaafflux: Sounds like a sync then.11:54
HobbseeSeveas: they need a kline11:54
HobbseeSeveas: thta nick looks similar to #kubuntu11:55
affluxpersia: we'll, yes, but I wonder if it's enough for a sync after debian-import-freeze11:55
SeveasHobbsee, tor should be shot. Like any improperly implemented privacy thing it's abuse too much11:55
HobbseeSeveas: this is true11:55
persiaafflux: Unless I'm mistaken, there's a bug in Ubuntu libotr that needs fixing.  As it's already fixed in Debian and there were no other changes in Debian, a sync is the easiest way to fix it.  DIF only means "Don't sync stuff just because it's updated in Debian: think about it first".11:56
affluxokay, good11:57
jpatrickpersia: ah, it does work11:58
Hobbseejpatrick: they're being killed12:06
jpatrickHobbsee: they better12:06
persiaLucidFox: qdevelop rejection ACK'd.12:07
LucidFoxpersia> just commented on gtkglextmm as well12:09
persiaDaveMorris: Do you need explicit rejection, or are you on that?12:09
DaveMorrisjust gonna read it12:10
sbodenIs there a way to check what happened when you upload something to revu, but you don't see it appearing on the website?12:10
persiasboden: Ask here.  Which package?12:11
sbodenIt's a first try, maybe I miss permissions somewhere12:11
persiasboden: What is your LP id?12:11
sbodensame ... sboden12:12
sbodenI can see what dput puts in the /incoming directory but after that it gets "erased"12:12
persiasboden: https://launchpad.net/~sboden doesn't show much.  Are you sure?12:13
sbodenmy bad : https://launchpad.net/~svenboden12:13
sbodentoo many userid's12:13
persiasboden: You just joined the team, and the keyring sync hasn't happened yet.  Please try again in about half an hour.12:14
sbodenk thx12:14
DaveMorrisLucidFox: it is currently arch independent, I guess nearly all dev packages could be arch all as well then, I've never realised that12:16
LucidFoxDaveMorris> Well, some -dev packages include .a files12:17
LucidFoxor utilities, like libqt4-dev12:17
DaveMorrisyeah, I've been told that new packages shouldn't include static libs12:18
DaveMorrisso basically my last 3 packages (all libs) could of done this as well12:18
* HighNo wonders how fast you guys are. You look/create tens to hundreds of packages per day. kudos!12:21
HighNojust getting my first package done seem like a full time job to me...12:23
persiaHighNo: teamwork.  Lots of people spending lots of time, each doing one thing.12:23
Hobbseeor whinging, as the case may be12:24
* persia suspects whinging doesn't lead to package updates or new packages, but might be considered part of "teamwork" in some special cases.12:26
persiasboden: Should be good now.  Try again.  Please note that the last official REVU day for hardy has passed, so your package may not get reviewed until May.12:31
DaveMorrisLucidFox: linda gives me one warning I don't understand when using arch all on the dev package - W: libgtkglextmm-x11-dev; File /usr/lib/pkgconfig/gdkglextmm-x11-1.2.pc contained in /usr/lib of Architecture: all package.12:33
webwolf_27does anybody have any idea where I can find an example debian/rules file for a bin-only package12:34
LucidFoxDaveMorris> I think this isn't a blocker.12:35
DaveMorrisI've checked the file created and it doesn't contain any arch specfic stuff, I'll upload it12:36
pefwebwolf_27: *firmware packages ?12:37
webwolf_27pef, no binary-only drivers12:38
pefwebwolf_27: they behave the same way, don't they ?12:39
webwolf_27pef, also a point12:39
pefwebwolf_27: check opera package12:40
pefor unrar non free12:40
LucidFoxI don't frankly understand the point of the Opera package12:41
LucidFoxI can understand proprietary drivers with no free replacement, but a web browser?12:41
webwolf_27LucidFox, I installed opera for the sole purpose of testing webpages with opera12:42
HighNopersia: when as the "last official REVU day" been?12:44
webwolf_27pef, thanks for the tip. The unrar-nonfree looks promising12:44
persiaHighNo: For hardy, REVU Days were every Monday (all timezones) from 5th November to 4th February.  The schedule for the next release has yet to be determined.12:45
persiaAs a very rough guess, I suspect the first REVU Day to happen sometime during the week containing May 29th, but there are many factors that could adjust that significantly.12:46
HighNogosh  - so there's not even a slightest chance my package will make it into hardy?12:47
DaveMorrisLucidFox: it's updated, thanks for looking at the package12:48
persiaHighNo: Basically, no.  Maybe if some MOTU happens not to be busy with last minute things for Feature Freeze and needs your package for something, they might grab it, but it's a slim chance indeed.12:48
Hobbseepersia: another option, then, is presumably debian-mentors, then sync after our feature freeze12:50
persiaHobbsee: If the package is important enough to allow a FeatureFreeze exception, I'm not sure how pushing through mentors is different from pushing through REVU, unless someone is planning to maintain the package for Debian.12:51
Hobbseethis is true - but new packages from debian carry little risk, and little time taken12:52
* HighNo thinks of a hopelessly overworked MOTU often leaving his/her Notebook or computer unlocked in a possibly insecure environment so that he/she needs my package to lock the pc whenever he leaves it. (Or as DaveMorris put it: You could use it to see your boss coming before he's in your room so you can quickly move from the 'must finish MOTU work desktop' over to the 'work desktop')12:52
LucidFoxHighNo> lol12:52
LucidFoxDaveMorris> Thanks. I'm sorry for not looking at the debian dir first, or I would have brought more things to your attention at once. I'll post some more objections, but these are trivial to fix.12:53
persiaHobbsee: Maybe, but I'd think two ACKs from REVU to be similar to a pass into Debian.  Also, adjusting the build and test environment completely is often more work for the packager.12:53
paaspersia: thanks for reviewing my package. I've fixed your comments and uploaded a new version http://revu.tauware.de/details.py?package=libtuxcap cheers12:53
Hobbseepersia: mmm, true.12:54
persiapaas: You've made it impossible for me to review it, but thanks for the quick response.  Good luck!12:54
LucidFoxDaveMorris> commented12:54
HobbseeHighNo: that's when you train to autolock your machine whenever you walk away12:54
* Hobbsee can't seem to do that under gnome12:55
paaspersia: why's that?12:55
persiapaas: architecture limitations12:55
paaspersia: I see12:55
HighNoHobbsee: see - you need it :-)12:56
* DaveMorris likes commented out deb calls, as it's easier to re-implement them, and for new people to learn from IMO12:56
HobbseeHighNo: no, now i just make sure i flip the cube, then lcok the screen via the mouse, whenever i go away :)12:56
HighNoHobbsee: you could save that precious time :-)12:56
StevenKHobbsee: Or use a screensaver that doesn't show the contents of the desktop12:57
* persia thinks it's not a cube, and that is still a manual action, and admires HighNo's imaginative advertisement12:57
HobbseeHighNo: :)12:57
HobbseeStevenK: i don't.  it's blank12:57
* StevenK uses electric sheep12:57
LucidFoxpersia> please explain to DaveMorris why commented out dh_ calls are bad :)12:58
StevenKBecause it bloats the rules file and makes it hard to read12:59
persiaDaveMorris: It makes debian/rules longer, and therefore is more troublesome to read.  It also looks ugly, and can be confusing to people working on their first patch when they fix a bug.12:59
LucidFoxStevenK, are you the upstream developer of linda or just its Debian maintainer?13:00
LucidFoxAwesome. When can we expect a version that doesn't complain about Standards-Version 3.7.3? :)13:00
HighNoright - that's the last warning on my list too :-)13:00
StevenKA week after the last person asked13:01
LucidFoxI get the hint, thanks13:01
Hobbseemaybe someone should just nmu13:01
StevenKYou know, I'm very close to asking for it's removal.13:02
Hobbseesee if StevenK gets angry13:02
Hobbseeoh?  why?13:02
LucidFoxwell, linda is ubuntuX anyway, so someone could patch it for Ubuntu13:02
StevenKBecause it didn't serve it's purpose, and everybody has pulled my ideas from it and put it into lintian13:02
* StevenK sighs13:03
Hobbseethe fade to black screensaver, when you have stuff inverted, is white!13:03
StevenKErr, yeah. That's what invert means :-P13:03
Hobbseeyeah, but i didn't think it'd go for screensavers too13:04
Hobbseeit doesn't go for desktop backgrounds, for eg13:04
StevenKBecause compiz doesn't draw it13:04
StevenKNautilus does13:04
Hobbseethis reminds me of why i should go back to kde13:05
StevenKIt takes every RGB value it draws, and minuses it from (255, 255, 255)13:05
StevenKSince it has all of the GL textures it draws in memory, inverting colour maps is trivial13:06
StevenK(Geez compiz doesn't seem like magic when you can figure out how it does its crack)13:06
LucidFoxHobbsee> Go back to KDE? Why?13:07
Hobbseemissng bits in gnome13:07
StevenKSuch as?13:08
Hobbseea world clock that works13:08
HighNoI'm at GNOME because I think it's missing the missing bits :-)13:08
Hobbseerunning kde/qt apps natively13:09
LucidFoxthe one in Hardy doesn't?13:09
Hobbseenot really13:09
Hobbseeit doesn't actually show the times atm13:09
Hobbseeyay for bugs13:09
StevenKSo patch it? :-P13:09
HobbseeStevenK: that would requrie being active13:09
Hobbseebut this is true13:09
Hobbseeit's a known bug13:10
StevenKI'm fairly certain that bug is because it always check the local time, not the time for that row13:10
* Hobbsee nods13:10
Hobbseevarious bits of config missing in gnome13:10
persiaHobbsee: Fixing things that annoy you is good.  Fixing things that annoy other people is nice, but not fixing things that annoy you just compounds it.13:10
Hobbseepersia: true13:10
DaveMorrisLucidFox: updated13:11
geserHi bddebian!13:11
HighNopersia: If I would have done that all my life I'd have no windows experience at all...13:11
bddebianHeya gang13:11
bddebianHi geser13:11
persiaHighNo: No reason not to start now :)13:12
persiabddebian: Hey13:12
bddebianHi persia13:12
LucidFoxDaveMorris> Thanks. I have no more objections. (But IANAMOTU.)13:12
HighNopersia: true. I should even not be here at the moment. I got another exam on monday and I've learned about - ehm - like 2 hours...13:13
DaveMorrisyour still waiting then13:13
persiawhich package?13:13
LucidFoxpersia> https://launchpad.net/ubuntu/+source/gtkglextmm13:13
smarterIs it normal that CDBS symlinks things in /usr/share/doc/<pkg>?13:18
LucidFoxsmarter> I've commented on qdevelop13:18
LucidFoxsmarter> yes, it's normal13:18
smarterLucidFox: thanks13:19
LucidFoxyou can ignore lintian warnings about it13:19
smarterbut I've three binary packages(extremetuxracer extremetuxracer-data and extremetuxracer-gimp-dev), extremetuxracer doc files link to extremetuxracer-data but extremetuxracer-gimp-dev has its own docs files13:20
LucidFoxbut CDBS only symlinks identical files present in hard dependencies, ne?13:21
persiasmarter: Wasn't it just -data that got hit by the CDBS symlink attack?  You can make the warning go away by manually linking the directory as advised by lintian, but it's not typically important.13:21
smarterpersia: how can I do this?13:23
persiasmarter: delete the contents of /usr/share/doc/extremetuxracer-data/ and symlink it to /usr/share/doc/extremetuxracer in debian/rules.  I'm not telling you that you should do this, only that it makes lintian quiet.13:25
smarterthat seems a bit overkill for a lintian warning ;)13:26
smarterI'll let cdbs do his stuff13:26
smarterQDevelop fixed and uploaded ;)13:32
=== rzr is now known as rZr
smarterIf I use a .menu file, should I build-dep on menu?13:33
persiaDaveMorris: s/binary:Version/source:Version/, also, might it be good to but the library names in a shlibs file for packages that later build-depend on libgtkglextmm-x11-dev?13:37
persiasmarter: You don't need to build-dep, and dh_installmenu maintainer script bits disable silently if the menu package is not installed.13:37
DaveMorrispersia: ok13:38
aquowhat is the difference between dput and dupload? i want to upload something to ppa13:43
mok0aquo: same thing13:46
mok0dput is written in Python, dupload in Perl...13:46
aquoare there any reasons to prefer one over another?13:48
=== LifeHacker is now known as tuxmaniac
mok0aquo: no13:57
=== jekil2 is now known as jekil
=== highvoltage is now known as highvoltage-bot
=== highvoltage-bot is now known as highvoltage
frafuHello, I am reading the wiki documentation about syncing packages and I wonder what UpstreamVersionFreeze means. Maybe the upstream Feature Freeze?14:52
tonyyarussofrafu: Upstream Version Freeze means that is the deadline for incorporating any upstream versions, ie OpenOffice.org version
persiafrafu: Before hardy, there were separate freezes for "FeatureFreeze", "UpstreamVersionFreeze", and "UniverseNewPackagesFreeze".  Now, they are all "FeatureFreeze".14:55
persiafrafu: Please update the outdated page :)14:55
frafupersia: so I have to replace UpstreamVersionFreeze with FeatureFreeze on that page!? By the way, thanks to both of you for the explanation.14:58
persiafrafu: Exactly.  Thanks for helping to keep the documentation up to date.14:59
frafuok; I will do it14:59
LucidFoxpersia> qdevelop looks good to me now15:04
persiaLucidFox: Best get someone else to ACK for the next bit: it's getting late here.15:04
LifeHackerhi. there has been a package sync from debian unstable but it FTBFS15:11
=== _czessi is now known as Czessi
LifeHackerI have changed a few missing build-deps. How do I go about submitting the patch?15:12
=== LifeHacker is now known as tuxmaniac
gesertuxmaniac: create a debdiff (see https://wiki.ubuntu.com/PackagingGuide/Recipes/Debdiff) and attach it to a bug15:14
tuxmaniacgeser, ok. the bug is already created in Debian BTS. I recreate it in LP also?15:15
gesertuxmaniac: yes15:15
geserbut you can link the LP bug to the BTS one15:15
=== lando_ is now known as lando
DaveMorrishmm, with an shlibs file.  Lintian complains if it's done one way, I change to fix that way and then Linda complains.15:44
DaveMorrisI can't make both of them happy, so which one is wrong?15:44
webwolf_27an somebody please tell me what "dpkg-genchanges: failure: cannot read files list file: No such file or directory" means and evtl. how to fix it15:50
geserwebwolf_27: my guess is that you try to build a (binary) package on an arch where it shouldn't be build15:53
webwolf_27geser, yes it's a binary package but the arch is set correctly15:55
crevettewhere should I propose new soft to package in ubuntu ?15:55
rZrhippu: hi thx for taking care of tuxguitar15:55
rZrhippu: I'll try to port it to icedtea soon15:55
geserwebwolf_27: can you make the package somewhere available to look at?15:56
webwolf_27geser, I can paste the rules file in pastebin15:56
webwolf_27geser, or I can upload a tared package15:57
vemondoes simple-patchsys need anything else for patching than the patch file in debian/patches directory?15:58
geservemon: afaik no15:59
vemoni've seen a "series" file used in some packages. is that required for sps?15:59
geserseries is used by quilt15:59
vemonseems like the package i'm creating at the moment doesn't patch up even though i've included the simple-patchsys.mk in rules and added the patch to debian/patches15:59
kdubwhen i run debuild on a package i'm trying to create, it fails, saying i do not have a secret pgp key. i do indeed have a secret pgp key, so why isnt debuild finding it?16:00
webwolf_27geser, which would you prefer16:01
geserwebwolf_27: let's try first with rules16:01
geservemon: perhaps the naming of the patches (file extension)16:02
webwolf_27geser, ok just a sec16:02
vemongeser, you're probably right. i don't have any extension there atm16:03
vemonmaybe i'll try to use .patch16:03
webwolf_27geser, http://pastebin.org/18822 not much too it though16:04
geserkdub: does the changed-by field from the .changes file match your key uid (including any comments)?16:04
geserwebwolf_27: that's a very short rules file. what are you trying to do?16:05
webwolf_27geser, I'm package the closed-source brother-lpr printer drivers16:06
geserwebwolf_27: even then you will need some dh_* calls in binary-arch16:07
webwolf_27geser, ok16:07
kdubgeser: it detected my name as "Kevin" when the name on my key id is my full name, and when i change it in the .changes file, it gets overwritten to the wrong value16:08
vemonshould the get-orig-source target download the same upstream version as the package version implies or the newest (use watch file & uscan)16:08
geserkdub: the value is taken from your changelog entry (in debian/changelog)16:10
kdubgeser: there it goes, thanks!16:11
ScottKDaveMorris: If you have to pick, keep lintian happy.  It's more current.16:11
protonchrisScottK: thanks for answering my question the other day.  You said that I should use postgresql-8.3 | postgresql-8.2 .  However, the package requires a configure option to point to the postgresql bin directory.  What is the best way to handle this?16:14
DaveMorrisScottK: thanks, did you want me to file a bug on it?  I think it's because the soname is libgtkglextmm-x11-1.2 Linda seems to want libgtkglextmm-x1116:14
webwolf_27geser, thanks it got built16:15
kdubshould pbuilder generate a .deb file?16:17
ScottKDaveMorris: Linda is currently unmaintained. I wouldn't worry about it.  If it gets picked up again I think the author knows what he needs to do.16:17
webwolf_27geser, now it's only missing the included files list16:17
ScottKprotonchris: I'm not a postgresql expert, but there is a postgresql-common package that deals with multiple versions installed side by side.  I'd look into that.16:18
protonchrisScottK: thanks.16:18
geserwebwolf_27: you should copy the files to the correct location below debian/<pkgname>/ so they got included in the binary deb16:18
webwolf_27geser, thanks. you probably noticed that this is my first deb ( I used to build RPMs but this is a little different16:19
ScottKmok0: Thanks for the courier merge.  Was that fun (I uploaded it last night)?16:20
=== r00f is now known as f00t3r
crevettehow can I create an ITP on debian from a ubuntu system as reportbug was patched ?16:27
spectiecrevette, email submit@bugs.debian.org16:28
crevetteokay I found16:29
jdongpochu: my original reporter was planning on doing that... I don't know if he's still doing so. If not, I'll do it :)16:29
pochujdong: cool :)16:30
gesercrevette: tell reportbug to use the Debian BTS and use reportbug then16:31
=== LifeHacker is now known as tuxmaniac
asantonihi all16:34
asantoniCan I prod someone to review my updated Mixxx package sometime? http://revu.tauware.de/details.py?package=mixxx16:35
asantoni(please) :)16:36
jdongHow can we get people to stop using Launchpad as a discussion board?16:36
Seveasjdong, with a hammer16:36
jdongI'm increasingly seeing bugs where it turns into a forum thread of people telling each other to grab stuff from git, compile vanilla kernels....16:37
asantonijdong: in bug report comments?16:37
jdongasantoni: right.16:37
jdongor confirm/say exactly what the person above said16:37
asantoniit's a consequence of having an open bug tracker16:37
ubotuLaunchpad bug 177570 in hal "[hardy] two batteries display when left clicking on g-p-m" [Medium,Confirmed]16:37
jdongfor example, on that bug, only about 25% of those comments were actually necessary16:38
asantoni(not that that sheds any light on a solution)16:38
jdongcomment #23 was what ultimately set me off.16:38
asantoniyeah, I've seen it happen tons of times too.... the nautilus/file-roller drag-and-drop bug was a circus16:38
jdongI don't know if providing a way for LP to link to a discussion medium would help16:39
jdonglike linking a bug to a forum thread and being able to move posts back and forth?16:39
asantoniYeah, perhaps16:39
jdongthere was a backports bug too where by the time I saw it there were 50+ replies and I was really tempted to just say "file it again"16:39
asantoniPeople (including myself at times, I admit) like looking at bugreports for "the quick fix"16:40
asantoniIt's kind of a bastardized use case16:40
jdongasantoni: I don't blame you. As a user, that's all you really care about when you see a bug.16:40
jdongasantoni: I'm not here to discourage people from doing that, I just want to find a better medium for it16:40
asantoniright, so the question is, what's the best way to address that use case16:40
jdongbecause the way it's being done now *gets in the way* of developers trying to fix the bug16:40
asantonioh yeah, I understand completely16:41
asantoniI don't know if adding something like a "temporary workaround" link would help at all16:42
asantonibecause while that would make it easy for people to find the quick fix, it probably wouldn't stop LP from becoming a forum16:42
jdongprobably something like "Link this to a {discussion}"16:42
jdongwhere Discussion can be a forum thread, mailing list, launchad answer ticket....16:42
asantoniYeah, that might work16:43
jdongand also, for posts to be migrate-able by the QA team across that link16:43
asantoniPerhaps there could be some streamlined way to add a "temporary workaround" Launchpad answer16:43
geserjdong: in your comment on that bug you forgot to tell that those who try this workaround will have to fix their system themselves if the workaround breaks something now or in the future16:45
jdonggeser: I dodn't want to open that can of worms, but yeah, that workaround can cause real hell16:45
asantoniit's true16:46
vemonany motus could advocate & upload this: http://revu.tauware.de/details.py?package=lashwrap16:46
jdongrandomly grabbing the moment's hal from git and overwriting the local installation with it16:46
jdongis it kosher to depend on libc6-dev (>= foo) to force a rebuild against a newer libc6?16:46
geserjdong: afaik yes16:48
jdongmmmkay :)16:49
jdongand I'd like to vent some anger at mldonkey16:49
jdongwhich seems to need the entire ocaml stack to just debian/rules clean.16:49
geserjdong: in such cases I use dpkg-source -b to get a new source package (and dpkg-genchanges if I need also the .changes file)16:51
jdonggeser: good idea16:54
tuxmaniacdebdiff -v old dsc_old dsc_new should be sufficent for submitting patches?16:54
tuxmaniac* debdiff -v dsc_old dsc_new16:55
ScottKtuxmaniac: Yes16:56
ScottKHowever look at it to make sure it's sane.16:56
stanipochu & scottk: I'll continue here as it is ubuntu specific... I see that ubuntu hardy still ships winpdb 1.3.2, while debian ships 1.3.4. Any chance of updating that as well? Probably only a sync.16:56
ScottKShouldn't be a problem.16:58
crevettewhere should I submit ITP for ubuntu ?16:58
tuxmaniacScottK, no diff is generated17:01
tuxmaniacScottK, only if the control file is changed?17:01
=== bmk789 is now known as bmk789_shower
kdubcan anyone help me with my pbuilder problem? http://pastebin.ca/89750317:03
ScottKstani: winpdb sync requested.17:04
ScottKtuxmaniac: You did build the new source package, right?17:04
asantonikdub: The "make install" is failing17:05
ScottKdpkg-buildpackage/debuild -S17:05
tuxmaniacScottK, aah wdiff is not available it seems17:05
kdubi did do debuild -S, but i can try again17:05
asantonikdub: Does running "make install" in the entropy source directory work normally?17:05
ScottKkdub: I was talking to tuxmaniac.  Sorry for the confusion.17:05
tuxmaniacScottK, I did this. pbuilder --debbuildopts -sa17:06
kdubasantoni: make install does not work17:06
tuxmaniacScottK, I meant s/pbuilder/pdebuild17:07
ScottKstani and pochu: pychecker still FTBFS in Ubuntu.17:07
asantonikdub: ok, so... is that the correct directory to be executing "make install" in?17:07
ScottK(with python 2.5)17:07
gesercrevette: https://wiki.ubuntu.com/UbuntuDevelopment/NewPackages and in case of an ITP assign it to you17:09
crevetteokay thanks17:09
kdubit is the correct directory, however, this particular configure script is doing funny things (i.e. not working right) so i'll look into that now17:11
crevetteif someone is interested to review this http://revu.tauware.de/details.py?package=obex-data-server17:15
=== LifeHacker is now known as tuxmaniac
tuxmaniaccan somone check bug 190473 and ack the diff? Its the first one that is much better17:27
ubotuLaunchpad bug 190473 in geda-xgsch2pcb "[FTBFS] geda-xgsch2pcb 0.1.2-1 build fails at checking for XML:Parser" [Undecided,Confirmed] https://launchpad.net/bugs/19047317:27
gesertuxmaniac: your last debdiff misses a changelog entry (0.1.2-1ubuntu1) with your changes17:31
gesertuxmaniac: you need to change also the maintainer field in debian/control17:32
tuxmaniacgeser, I know. But the previous one which peter has uploaded contains them17:32
gesertuxmaniac: but not a quite correct one: wrong version, wrong distribution17:35
gesertuxmaniac: and the maintainer change is also missing there17:35
mok0When writing the lintian overrides file, do you need to include the entire lintian message, or just a regexp?17:35
tuxmaniacgeser, aah yes. if we are taking this for ubuntu then we need to update all those17:35
tuxmaniacgeser, I guess he uploaded the same one as for debian bts bug17:36
tuxmaniacgeser, do we wait for debian to upload the changes or make local changes?17:36
gesertuxmaniac: we can fix it now for Ubuntu17:36
tuxmaniacgeser, ok17:36
gesertuxmaniac: if you want to fix this bug, could you prepare a debdiff with all needed changes and subscribe then the ubuntu-universe-sponsors team for sponsoring?17:37
tuxmaniacgeser, am doing that already :)17:37
webwolf_27geser, Sorry to bother you again. I created the dirs and copied the files into debian/packname but the deb still doesn't include the files17:48
tuxmaniacgeser, mainatiner field will be MOTU right?17:52
DaveMorristuxmaniac: yeah17:53
gesertuxmaniac: yes, you can use the updatemaintainer script from ubuntu-dev-tools to do it17:53
kdubthe package i'm trying to build turns out to use an "install.sh" script, not a make install. can i tell pbuilder about this somehow?17:56
tuxmaniacgeser, oh cool. thanks17:56
geserkdub: sure, replace the make install call in your debian/rules with the correct call of that install.sh script17:56
kdubgeser: thanks, for some reason i didnt think of that...17:57
linux__alienhi people18:00
linux__alienhey mok018:00
linux__alienhey LucidFox18:00
mok0Hey linux__alien18:00
linux__alienthanks for your help yesterday finally got the package :)18:00
linux__alieni will now try doing an other package . those given in the recipe18:01
linux__alieni am interested in fixing bugs and packaging too so how do i go about it18:01
linux__alienso much interested in contributing to Ubuntu in either of these 2 forms but the former would be the better :)18:02
LucidFoxlinux__alien> Fixing bugs is great and always welcome. New packages may be too late to submit at this point, we're very near feature freeze18:02
=== bmk789_shower is now known as bmk789
linux__alienyes i agree very true . Fixing bugs the smallest bug would be a real good experience for me dont you think so /18:03
linux__alienhence it would be really good if someone could take me as part of their team and mentor me and i could do all and everything to help them too18:04
linux__alienthats my humble request18:04
linux__alienso that if someone to help me comfortable can start assigning me or rather telling me hey you could do this first and then fix this bug #some number and let me know something like this18:05
linux__alieni would try my level best to fix it and if any issues i would ask my doubts .18:05
linux__aliencan someone here help me in this regard18:05
rZrhi linux__alien , havent you an account on livejournal ?18:06
vemonif i have one advocation for my package in revu then will it "go away" if i upload a new version?18:06
linux__alienrZr, i dont have an account in livejournal18:06
linux__alienrZr, why18:07
rZrsome guy use the same nick i think18:07
linux__alienLucidFox, any help please?18:08
tuxmaniacplease can someone check bug 190473 's debdiff whether they are proper? subscribing universe sponsors18:09
ubotuLaunchpad bug 190473 in geda-xgsch2pcb "[FTBFS] geda-xgsch2pcb 0.1.2-1 build fails at checking for XML:Parser" [Undecided,Confirmed] https://launchpad.net/bugs/19047318:09
LucidFoxvemon> yes18:09
LucidFoxvemon> you can ask that MOTU to readvocate it18:10
LucidFoxlinux__alien> what kind of help?18:10
linux__alienLucidFox, a person who could help me in going forward in terms of bug fixing. i would parallely look into packaging i am already doing it so i would like to start some fixing of small issues18:11
linux__aliento start with18:11
linux__alienbut i wouldnt know which is small and which wouldnt so if someone could tell me where to look at or something like that18:12
linux__alienLucidFox, to be frank i would need a mentor for this :)18:12
linux__alienLucidFox, i was happy that i had joined one team as i had told you couple of days back but not able to contact the mentor in any manner at all18:12
linux__alienits all become silene18:12
rZrhippu: if it's not too late before uploading tuxguitar, i'd like to make some tests18:12
LucidFoxWhat team was it?18:13
linux__alienLucidFox, https://launchpad.net/~gnome-uis18:13
hippuwhat do you mean?18:14
linux__alienLucidFox, i am in need of a real help and i ve subscribed to the MOTU mailing list too18:15
geserlinux__alien: are you still looking for an easy bug to work on?18:15
linux__aliengeser, yes very much to start with yes18:15
linux__aliengeser, i am completely new to this so something to start with so that if i get stuck i could contact the person so i was thrilled when i saw the word Mentor in launchpad so joined immediately18:16
linux__alienbut didnt happen unfortunately :(18:16
geserlinux__alien: unmet dependencies (unmetdeps) are often quite easy to fix (often only a rebuild is needed for a library transition)18:16
geserlinux__alien: see the output of "apt-cache unmet -i" (preferable from a hardy environment)18:17
dcorderoi have a warning from lintian, that say me that an image name license.png of my package is and extra-license-file, but it's only an image that the application use. Must be the file renamed or can i ignore this warning message18:17
geserPackage siproxd version 1:0.5.13-1ubuntu2 has an unmet dep:18:17
linux__aliengeser, i dont have that currently installed18:17
geser Depends: libosip2-3"18:17
tuxmaniacgeser, did you check the latest debdiff that I uploaded. Hope its all right.18:17
geserlinux__alien: it would be good if you did have one hardy environment available as all development happens in hardy. A hardy pbuilder or hardy chroot will be enough (no need to update your system to hardy if you don't want to).18:19
* LucidFox is angry at LP being slow18:20
gesertuxmaniac: almost there: XSBC-Original-Maintainer should be the old value from Maintainer (the Debian maintainer)18:21
linux__aliengeser, i see lot of unmet depends in gutsy itself18:21
ScottKlinux__alien: It's best to concentrate on Hardy.18:22
geserlinux__alien: yes, we never happen to fix all, but it's to late to fix them for gutsy18:22
linux__aliengeser, so i ve to update my system to gutsy ?18:22
gesertuxmaniac: what about the comment on python-gtk2-dev from Peter?18:23
linux__alieni dont have that amount of bandwidth :(18:23
linux__aliento download it so if there are some standalone packages i can download them and try fixing bugs in that and upload it again18:23
LucidFoxlinux__alien> I think the easiest bugs are cosmetic stuff like errors in package descriptions and missing desktop files18:24
linux__alienLucidFox, ok how do i do that18:24
linux__alieni dont know where to look for and what to look for too ?18:24
LucidFoxhere's an easy one, which can also give you some experience in cooperating with upstream:18:24
LucidFoxbug #19005418:24
geserlinux__alien: no, but access to a hardy environment is really helpful (a hardy pbuilder should be sufficient for most cases).18:24
ubotuLaunchpad bug 190054 in tellico "Please remove "Encoding" from tellico.desktop" [Undecided,New] https://launchpad.net/bugs/19005418:24
LucidFoxyou will need to modify the desktop file and put the modified one into the debian/ directory, and add it to the installation rules18:25
linux__alienLucidFox, went through the bug and it says this is deprecated18:25
geserlinux__alien: but for hardy some bandwitdh helps as hardy updates very often (and you need to download the source package and often also do a test-build in a pbuilder and download also the build-depends).18:25
linux__aliengeser, so i ve to update my system to hardy is it?18:26
LucidFoxyou don't have to have your primary system as hardy18:26
ScottKNo.  You can do your testing in a chroot or vm.18:26
LucidFoxalthough, as ScottK says, having an isolated environment helps - like a chroot environment for building packages in pbuilder18:27
=== LifeHacker is now known as tuxmaniac
linux__alienLucidFox, i ve only one system . A laptop which is very new and installed 7.1018:27
tuxmaniacgeser, apart from the XSBC-Original maintainer is there any other issue?18:27
gesertuxmaniac: the only other issue I've is, who is right about the addition python depends?18:28
LucidFoxAs for the bug: you will need to make it so that it installs a desktop file that does not contain an encoding field18:29
geserlinux__alien: with a chroot you can have a hardy environment inside your gutsy installation18:29
linux__alienso i can use both ?18:29
geserlinux__alien: yes18:29
linux__alienso nothing would happen to the 7.10 setup which i ve currently ?18:29
linux__alienif you say nothing would happen i am ready for it . how do i do it and what should i install18:30
linux__alienso if i install that i can contribute to hardy is it?18:30
linux__alienand fix bugs in that environment ?18:30
geserlinux__alien: if you have the option you could install hardy into a separate partition and select at boot time if you want hardy or gutsy18:30
* LucidFox pauses speaking about the bug for now18:31
ScottKWhat's the python depends question?18:31
linux__alieni dont have that too18:31
linux__alienif i install the chroot what all could i do18:31
linux__alienfor hardy18:31
geserScottK: bug #19047318:32
ubotuLaunchpad bug 190473 in geda-xgsch2pcb "[FTBFS] geda-xgsch2pcb 0.1.2-1 build fails at checking for XML:Parser" [Undecided,Fix committed] https://launchpad.net/bugs/19047318:32
LucidFoxlinux__alien> see https://wiki.ubuntu.com/PbuilderHowto18:33
geserScottK: there are two different debdiffs18:33
* ScottK looks18:33
LucidFoxthat page contains information about setting up a chroot environment to build packages in Hardy18:33
LucidFoxwhile you can continue to run Gutsy as your main OS18:34
tuxmaniacgeser, peter is right18:34
* tuxmaniac apologises for so much confusion18:34
linux__alienLucidFox, oh so this would be used only for packaging or could also be used for fixing bugs and compiling them ?18:34
geserlinux__alien: for both, you can login into a pbuilder (pbuilder login) and work then inside the hardy pbuilder18:35
ScottKtuxmaniac: Then would you please update your debdiff with his version (credit him in debian/changelog).  Your debdiff is more comprehensive (maintainer change and all that).18:36
tuxmaniacScottK, ok18:36
geserlinux__alien: there is also https://wiki.ubuntu.com/DebootstrapChroot but this is a little bit harder so setup (for a newcomer) and also not updated for hardy18:36
linux__alienso if i install that i could start fixing issues for hardy is it?18:37
linux__alienthat would be only in packaging right ?18:37
LucidFoxyes, you can then test-build your modified packages for hardy in that environment18:38
geserlinux__alien: pbuilder is always used to test that your modified package still builds, so it's not only used for packaging but also after bug fixes18:39
LucidFoxlinux__alien> When you're preparing a fix for a bug, you prepare a new version of the package. For example, if the previous version was X.Y-0ubuntu1, your modified version will be X.Y-0ubuntu2.18:39
linux__alienbut for testing i cannot use that i would need the complete Hardy environment18:39
linux__alienyes true LucidFox18:39
LucidFoxYou then create a debdiff between the original and modified versions, and attach it to the bug report.18:40
linux__alienfine but to test whether that bug has got fixed or not ?18:40
linux__alieni would need Hardy right?18:40
LucidFoxAs for testing, in simple cases, you can also build the packages for Gutsy and test them there.18:41
tuxmaniacScottK, updated18:41
geserlinux__alien: it depends on the bug you want to fix: if you want to fix a typo in the package description then you don't need a complete hardy environment, but e.g. for fixing a crash that can be different18:41
linux__aliengeser, hmmm ok18:41
linux__alienget it18:41
tuxmaniacScottK, geser let me know whether things are Ok now.18:42
tuxmaniacthanks a lot for the help ScottK and geser18:42
LucidFoxlinux__alien> so, would you like to try and prepare a fix for bug #190054?18:42
ubotuLaunchpad bug 190054 in tellico "Please remove "Encoding" from tellico.desktop" [Low,Triaged] https://launchpad.net/bugs/19005418:42
linux__alienLucidFox, yes sure18:42
linux__alienLucidFox, how do i do that18:43
linux__alienLucidFox, it says deprecated18:43
LucidFoxAllow me to explain.18:43
linux__alienLucidFox, sure :) sorry18:43
LucidFoxThe problem with this package is that the desktop file that upstream ships contains a deprecated field, which should be removed.18:43
LucidFoxDesktop files can be provided in two different ways.18:43
* tuxmaniac can go to sleep a happy man18:44
ScottKtuxmaniac: What do you think python-gtk2 (>=2.8), python-gtk2 (<<2.10) | python-gobject is going to do?18:44
LucidFoxFirst, as part of upstream distribution - and second, as part of the packaging, in which case they go to the debian/ directory.18:44
LucidFoxIn the second case, the packager generally tries to push the changes to desktop files upstream.18:44
LucidFoxYour goal is to create a modified desktop file with the Encoding field removed, and modify the package so that it installs instead of the original desktop file which does contain this field.18:45
LucidFoxAm I being clear so far? :)18:45
tuxmaniacScottK, the version should be greater than 2.8 but less than 2.10?18:45
linux__alienthe goal is very clear :)18:45
LucidFoxSo, you have two options.18:46
linux__alienhow about option 1?18:46
LucidFox1. Put the modified desktop file in the debian directory, and modify debian/install so that it is installed into /usr/share/applications - thus overwriting the upstream desktop file.18:46
ScottKtuxmaniac: And then python-gobject?  When is that wanted?18:46
linux__alienOk where should i download the package from18:47
LucidFoxor 2. Place a patch in debian/patches that would modify the file in place during build.18:47
linux__aliensince this is the first thing i might ask you the silliest question too so please forgive me for that18:47
linux__alienso where do i download the package from ?18:48
LucidFoxlinux__alien> To download the source package, you can use apt-get source, but that requires that your deb-src in sources.list is set to Hardy. Since your main installation doesn't run Hardy, here's an alternative option.18:48
LucidFoxThe bug report has a link that says "Overview".18:48
LucidFoxClicking it will bring you to the source package overview page, which contains the upload history.18:48
tuxmaniacScottK, without python-gobject also it builds :S18:48
tuxmaniaclet me check once again.18:49
linux__alienyes i see some comments18:49
LucidFoxthose are debian/changelog entries.18:49
linux__alienand i see the published versions too18:49
LucidFoxin the Version history area, click the link to the most recent version.18:49
LucidFoxIn our case, 1.3-1ubuntu1.18:50
LucidFoxyou will see links to 3 files: dsc, orig.tar.gz and diff.gz.18:50
* tuxmaniac will look into it tomorrow.18:50
linux__alienLucidFox, i clicked on that pencil icon18:50
linux__alienoh sorry18:50
LucidFoxno, not the pencil icon - it's for upstream links18:50
linux__aliengot it18:50
linux__alienyes i see18:51
linux__alieni get to see those files as you said18:51
LucidFoxThese files constitute the source package. Save them to the directory in which you'll be working on that package.18:51
linux__alienall 3 of them ?18:51
LucidFoxalternatively, you can use dget, but this is the simpler approach - I'll tell you about dget later18:52
linux__alienbut one question i ve18:53
LucidFoxlinux_alien> tell me when the downloads finish :)18:53
linux__aliencan i update my existing system to Hardy is that possible and advisable?18:53
LucidFoxlinux__alien> If you even raise this question, then no.18:53
LucidFoxThe point of unstable releases is that something may break.18:54
LucidFoxSo they aren't recommended for everyday use.18:54
linux__alienoh ok18:54
warp10Hi all!18:54
LucidFoxhi warp1018:54
linux__alienthen how do people here do it18:54
warp10Hey LucidFox18:54
linux__alienassume a person has only one system and wants to contribute and he does not have another partition too18:55
linux__alienin that case how does he manage to do it18:55
LucidFoxthose who do run a full hardy installation (I don't) run it as a secondary system for testing and development only18:55
LucidFoxthis can be done on a separate partition or in a virtual machine18:55
linux__alienhow do you run it ?18:55
LucidFoxlike qemu or VirtualBox18:55
linux__alienoh ok18:56
linux__alienok got it now what do i do its downloaded18:56
linux__alienve got all downloaded18:56
jdonglinux__alien: for most times, testing stuff on hardy can be done in a chroot environment or a pbuilder login18:56
linux__alienoh ok then let me install that and do that18:56
jdonglinux__alien: the only things you can't do in it are test very hardy specific things, like kernel or X bugs.18:56
jdonglinux__alien: but for that, you can just rely on other hardy testers to provide that feedback18:56
linux__alienjdong, other GTK related bugs, Gnome bugs can be done is it ?18:57
jdongI don't think ANY of us here would complain about someone not running Hardy18:57
linux__alienwith chroot ?18:57
jdonglinux__alien: depends on what kind. You can launch even GNOME apps from inside that chroot with a bind-mounted /tmp18:57
jdonglinux__alien: but if the bug is a specific interaction bewteen the app and Hardy's X or kernel, then you're out of luck18:57
jdongbut that's a minority of bugs18:57
jdongI can only think of a handful of times where I actually needed a hardy environment18:58
jdongand even for that, I could just boot a LiveCD and test from there18:58
linux__alienthen its fantastic i would do that . Thats fine with me let me look into those bugs which wouldnt need a complete hardy environment18:58
LucidFoxTo be completely honest, I don't even run Hardy in chroot (expensive traffic): I run Gutsy with a lot of backported tools and libraries18:58
LucidFoxif there's something _really_ hardy-specific, I test-build it in PPA18:58
jdongLucidFox: yeah, once you have enough experience/judgement to know the limitations of your test environment, that works totally fine :)18:58
LucidFoxlinux__alien> Now your goal is to unpack the source, work on it to prepare the new version, and then create a new source package.18:59
linux__alienLucidFox, you are gonna be my mentor ;-)18:59
LucidFoxWhich will be 1.3-0ubuntu2 because the current version is ubuntu118:59
LucidFoxlinux__alien> do you use mc in the console?18:59
linux__alienyou want me to use it ?19:00
LucidFoxno, it's optional, just would be more convenient than a plain console... it's a console file manager19:00
linux__alienLucidFox, i ve opened the file19:01
linux__alienwhich has the problem19:01
LucidFoxwhat file?19:01
LucidFoxso you've already unpacked the source?19:01
LucidFoxdon't modify the file in place19:01
LucidFoxbecause direct changes to the upstream source are unwelcome19:01
linux__alienso what do i do19:02
LucidFoxinstead, copy the desktop file to the debian directory, and edit it there19:02
linux__alienok since i am gonna create a new package i should create a debian directory and copy this file there and edit there right?19:02
LucidFoxThere's an already existing debian directory19:03
LucidFoxfrom the existing packaging19:03
LucidFoxlinux__alien> did you use dpkg-source to unpack?19:03
linux__alieni unpacked the orig.tar.gz19:03
linux__alieni used tar -zxvf19:03
LucidFoxno, that's not how it should be19:03
LucidFoxremove the resulting source directory19:03
LucidFoxand use dpkg-source -x (dsc file)19:04
linux__alienok got it19:04
LucidFoxthis will unpack the orig.tar.gz _and_ apply the diff.gz, so you end up with a source tree and the debian directory19:04
linux__alienoh ok19:04
linux__aliennow i ve copied the file to debian19:05
linux__alienand i ve opened the file19:05
linux__alientoo now ve a question19:05
LucidFoxnow, edit the desktop file in debian and remove the line that shouldn't be there19:05
linux__alienthe bug report says This item is deprecated in Version=1.0 of freedesktop standards, so please remove it.19:05
linux__alienthere are 3 comments in that file19:06
linux__alienthey are talking about those comments ?19:07
LucidFoxleave the comments alone19:07
LucidFoxjust remove the Encoding line19:07
LucidFoxcomments are harmless :)19:07
jdongLucidFox: except when put there by the TA on my english paper19:07
jdongthen it costs me points :)19:07
linux__alienfine but where does the bug say that the encoding is extra19:08
LucidFoxlinux__alien> what do you mean, encoding is extra?19:08
linux__alieni mean for a person who has been working on this might know but for quite a new comer does the bug show that encoding shouldnt be there ?19:08
LucidFoxlinux__alien> the bug report contains a diff which lists the necessary changes19:09
linux__alienthere is a small hyphen before it19:09
LucidFoxyes; it means that this line should be removed19:09
LucidFoxit's a minus sign :)19:09
linux__alienok .. got it first time i didnt notice that19:10
linux__alienok removed it19:10
linux__alienand now i see that the control, copyright, rules file are already in place19:10
LucidFoxif the line was supposed to be added, it would have been +Encoding=UTF-819:10
linux__alienok .... got it19:10
LucidFoxyes, but there are two more things to do19:10
LucidFoxfirst, you need to make this file install19:10
LucidFoxyou can do that by editing debian/install19:10
linux__alienso how do i package without generating those files agin19:10
LucidFoxyou aren't done editing the package files yet :)19:11
linux__alienthere is no install file in debian19:11
LucidFoxthere is a line there: debian/tmp/usr/share/applications/*; it is responsible for installing the original, upstream desktop file (because this is where it resides)19:12
LucidFoxyou, therefore, should remove it, and instead add a line to install your modified desktop file19:12
LucidFoxwhich would be:19:12
LucidFoxdebian/(desktop file).desktop usr/share/application19:13
linux__aliendebian/tmp/usr/share/apps has to be removed19:13
LucidFoxno, not usr/share/apps19:13
LucidFoxthis is what the line says19:13
linux__alieni dont have that in that file19:14
LucidFoxwait a minute19:14
LucidFoxhow come I did? o_O19:14
LucidFoxah, two slashes19:14
linux__alieni dont have applications at all19:15
linux__alienin that file19:15
LucidFoxit's tellico.install, you're probably looking at tellico-data.install19:15
linux__aliengot it sorry19:15
linux__aliendeleted that19:15
LucidFoxnow, you will have to add a line to install your new desktop file19:16
LucidFoxdebian/(desktopfile).desktop usr/share/applications/kde19:16
LucidFoxthe part before the space tells where the file is, the part after the space tells where to install19:16
linux__aliendebian(x-tellico.desktop).desktop usr/share/applications/kde19:18
linux__alienoh sorry19:19
LucidFoxfirst, without the parentheses19:19
linux__alientwo desktops are there19:19
LucidFoxthe bug is about tellico.desktop19:19
LucidFoxnot x-tellico.desktop19:19
LucidFoxso you worked on the wrong desktop file :)19:19
linux__alienoh :(19:20
linux__alienam sorry19:20
linux__alienreally sorry19:20
LucidFoxso it should be19:20
LucidFoxdebian/tellico.desktop usr/share/applications/kde19:20
linux__alienbut that file does not have Encoding line in that file19:20
LucidFoxHow come...?!19:22
linux__alieni guess the bug report is wrong :)19:22
LucidFoxwait a second19:22
LucidFoxlinux__alien> yes, the bug report is indeed invalid :((((19:24
LucidFoxI'm terribly sorry19:24
linux__alieni am happy that i ve found out something so youo gave me an opportunity to find that so thanks to you :)19:25
LucidFoxwhile you won't get the honor of closing the bug, you can finish your experience by pretending that the issue did exist :)19:25
LucidFoxjust instead of LP, you'll upload your debdiff to pastebin19:25
LucidFoxfor me to review19:25
linux__alienwhats LP ?19:26
LucidFoxjust copy the desktop file unmodified (pretend that the original one had an encoding field and you removed it :))19:26
LucidFoxand make the change to the install file19:26
LucidFox(as I said, I'm really sorry, and will give you a real bug the next time around)19:27
linux__alienyou want me to copy the tellico.desktop file contents and paste it in launchpad ?19:27
LucidFoxno, not Launchpad19:27
LucidFoxand don't take action just yet19:27
LucidFoxI want you to prepare a debdiff just like you would if you were actually fixing a bug19:27
linux__alienok so how do i prepare that i ve not used that :(19:28
linux__alieni am sorry i guess i am irritating you :(19:28
LucidFoxyou're not :)19:28
LucidFoxdid you modify the install file?19:28
linux__alienits half baked19:29
linux__alienlet me delete this source tree and again untar it ?19:29
LucidFoxOkay, so now the (fictional) bug is squashed and you're almost there. Now you'll need to bump the version from -0ubuntu1 to -0ubuntu2.19:29
LucidFoxto do this, you will need to add a new entry to debian/changelog.19:29
linux__alienso what should i do now19:30
linux__alienthe install file is half  baked19:30
LucidFoxto do this, move to the source directory and type dch -i19:30
LucidFoxthis will open a text editor with a new debian/changelog entry19:30
linux__alienIt opens some file and it shows me lot of entries19:30
LucidFoxyou will edit the top entry19:31
LucidFoxfirst, the distribution is gutsy, you have to change it to hardy19:31
linux__alienyes the cursor is in that area19:31
LucidFoxand second, you'll need to actually describe your changes19:31
LucidFoxso make the list of changes something like:19:31
linux__alienif i want to close it and reopen it again19:31
linux__alienhow do i do it19:32
linux__alienone sec19:32
linux__alienthe cursor blinks near to * and top of it i ve gutsy written with urgency - low19:32
LucidFox* Removed deprecated encoding field from desktop file (LP: #190054)19:32
linux__alienso have to change that ?19:32
LucidFoxfirst, you have to change gutsy to hardy19:33
linux__aliendid that19:33
LucidFoxand second, write some actual meaningful text describing your changes19:33
linux__alientellico (1.3-1ubuntu2) hardy; urgency=low19:33
linux__alien  * Removed deprecated encoding field from desktop file (LP #190054)19:33
ubotuLaunchpad bug 190054 in tellico "Please remove "Encoding" from tellico.desktop" [Low,Invalid] https://launchpad.net/bugs/19005419:33
LucidFoxthe magical LP: # field contains the number of the bug you're fixing, and it will ensure tha the bug is autoclosed19:33
LucidFoxwhen this version is uploaded19:34
LucidFoxnow you're set - save the file19:34
linux__alienbut the email address is wrong in that file19:34
linux__alieni mean my email address19:34
linux__alienchange that too ?19:34
linux__aliencan i ?19:34
LucidFoxis it the one used in your PGP key?19:34
LucidFoxthe signature should be of the form "Name <email>" exactly as in the PGP key19:35
LucidFoxso yes, edit it19:35
linux__alienok changed it19:35
LucidFoxnow save19:35
linux__alienyes done19:35
LucidFoxand finally, you can build the modified source package19:36
LucidFoxdebuild -S19:36
linux__alienbut that install file is half baked19:36
linux__alienis that ok ?19:36
LucidFoxwhat do you mean, "half baked"?19:36
linux__alienit does not contain the original line nor the modified line is complete19:36
LucidFoxthen edit it until it's "fully baked" :)19:36
jdongpreheat oven to 450 degrees...19:37
linux__aliendch -i19:37
linux__alienparsechangelog/debian: error: badly formatted trailer line, at file debian/changelog line 519:37
linux__aliendch: fatal error at line 462:19:37
linux__alienProblem executing dpkg-parsechangelog:19:37
linux__alieni get this19:37
ubotupastebin is a service to post large texts so you don't flood the channel. The Ubuntu pastebin is at http://paste.ubuntu-nl.org (make sure you give us the URL for your paste - see also the #ubuntu channel topic)19:37
linux__alienok will do that sorry19:38
LucidFoxlinux__alien> paste your output there19:38
LucidFoxjdong> it's all my fault, I should have checked the bug for validity first :((19:39
jdongLucidFox: huh, what's your fault?19:39
LucidFoxnow we can only simulate fixing the bug because we already went halfway through19:39
linux__alienwhy should we do that LucidFox19:39
LucidFoxlinux__alien> please paste debian/changelog as well19:40
jdonglinux__alien: you should also pastebin debian/changelog19:40
jdongat least up to a few lines after #519:40
linux__alieni am not able to open it using dch -i19:40
LucidFoxno, don't open it with that19:40
linux__alienvi ?19:40
LucidFoxjust open it in a text editor and paste the text from there19:40
LucidFoxgedit/kate will probably be better since you need to paste into a web browser19:41
frafuHello, should the needs-packaging bug of mousetweaks not have been automatically closed since it is now in universe and debian/changelog contained: (LP: #162874) ? Or is it still open because the hppa architecture has not been built yet because it is waiting for a dependency?19:41
LucidFoxfrafu> check if the changesfile contains a Launchpad-Bugs-Fixed line19:41
LucidFoxlinux__alien> there should be a space before the <19:42
jdongfrafu: mark it Fix Committed when it's in source NEW, Fix Released when it gets approved19:42
LucidFoxBalaji G <balajig81@gmail.com>19:42
linux__alienok LucidFox19:42
LucidFoxas for dch, it's only needed to _create_ the new debian/changelog entry - after it's there, you can edit debian/changelog in any text editor19:43
linux__alienoh ok19:43
frafuLucidFox: where can I find the changesfiles?19:43
linux__aliennow how do i get back the original install file19:43
LucidFoxlinux__alien> what for?19:43
linux__alienthe install file is not complete :(19:44
LucidFoxwhat do you mean?19:44
linux__alieni had modified the install file remember because i had to give the path of the desktop file that i modified19:44
LucidFoxif you modified it correctly, there's nothing to worry about19:45
linux__alieni didnt modify it coz we had a problem coz i had modified x-tellico and then we came to know it should be tellico19:45
linux__alienso left it at that point19:45
LucidFoxjust change x-tellico to tellico19:46
linux__alienoh ok19:46
LucidFoxI'm sorry, I'm sleepy and want to see the result before I go to bed19:46
LucidFoxso, I'm sorry, but could you just build it it already? :)19:47
linux__alienso left it at that point19:47
ScottK\sh_away: https://launchpad.net/ubuntu/+source/octave-epstk/2.2-8/+build/506791 looks like an oops, missed a spot from your octave transition.19:47
linux__alienok now i ve to give debuild -S at the source19:47
LucidFoxit will create a new source package - 1.3-1ubuntu2.dsc19:48
ScottKBTW, that one ^^^ looks like it might have an easy fix for a MOTU hopeful to look into.19:48
linux__alienLucidFox, i get an error because of not signing it19:48
linux__alieni ve to give -us and then some other option19:48
linux__alieni forgot :(19:48
frafujdong:So I set the bug to Fix Released,  as mousetweaks is in the universe repo!? Or should I wait until the hppa is also in the universe repository?19:49
LucidFoxyou mean you didn't configure GPG?19:49
* LucidFox headdesks19:49
Toadstoolgood evening!19:49
LucidFoxdid it at least create the dsc file?19:49
ScottKHeya Toadstool.  Long time no see.  Welcome back.19:49
LucidFoxyou don't need it signed for debdiff anyway, I just really want to look at it ASAP19:49
jdongfrafu: set it to Fix Released19:49
ToadstoolScottK: hi!  how's it going?19:49
linux__alienyes its created the dsc19:50
LucidFoxah, good19:50
ScottKToadstool: Pretty good.  I'm in front of the Tech Board on Tuesday for my core-dev application.19:50
Toadstoolgood luck!19:50
LucidFoxlinux__alien> now you run debdiff between the old and new dsc19:50
LucidFoxdebdiff olddscfile newdscfile > somename.debdiff19:50
LucidFoxafter that, paste the contents of the output file to the pastebin19:51
ToadstoolI am so running behind with what happened in MOTU land lately :)19:51
LucidFox(if you were actually fixing a bug, you would download it to Launchpad, but as  I said, it's my fault)19:51
ScottKToadstool: It's good to have you back.  Stick around and catch up.19:53
LucidFoxlinux__alien> That's it!19:53
linux__alienGreat :)19:53
ScottKIs anyone here set up to do GCC 4.3 (GCC Snapshot) test builds?19:53
LucidFoxNow, if you were fixing the bug, you would attach it to the bug report and then subscribe sponsors19:54
linux__alienOh ok19:54
frafujdong: ok, I will do it right away; thanks. Out of curiosity: does it normaly happen automatically? Or is something the reviewers normaly do and it is not automatic?19:54
LucidFox(ubuntu-universe-sponsors or ubuntu-main-sponsors, depending on where the package is)19:54
linux__alieni guess once i  get an other bug to fix i would get a more better idea19:54
LucidFoxlinux__alien> I'll try to find a valid bug for you the next time around :)19:54
ToadstoolScottK: yup, that's the plan19:54
LucidFoxright now I'm going to sleep19:54
linux__aliensure will you be here tomorrow19:54
jdongfrafu: I believe it only doesn't happen automatically for a brand new package that hits NEW19:55
linux__alienme too :)19:55
jdongfrafu: for ordinary uploads with (LP: #foo), it is automatic19:55
ScottKjdong: It's automatic for New packages too, but after they get out of source New.19:55
jdongScottK: oh, hmm I didn't see it happen when clutch passed trhough last week :(19:55
jdongI had to have it manually done19:55
ScottKHmm.  I thought I'd seen it done, but maybe I'm wrong.19:56
LucidFoxlinux__alien> actually, wait :)19:56
LucidFoxyou screwed up a bit19:56
LucidFoxgah, he left19:56
frafuSkottK: it did not happen for mousetweaks either; but perhaps because one architecture is not built yet19:57
ScottKjdong: You're right.  It's because the bug isn't against that package (and it can't be since it doesn't exist yet).19:57
ScottKfrafu: ^^^19:58
jdongScottK: aha, that's why :)19:58
frafuSkottK: that makes sense20:00
ScottKjdong: It actually sounds like an LP bug to me.  If an upload claims to fix a bug that doesn't have a package assigned, then LP should believe it.20:02
jdongScottK: concurred20:04
ScottKjdong: Would you please file it and I'll confirm it (they've heard enough from me lately)?20:05
frafuThis brings me to another question: When doing a package for a sync to a new release: I thought I had to add do debian/changelog only the items relevant to packaging. But obviously it was wrong because this way, the bugs will not be automatically closed. So I will also have to add the fixed bugs to it.20:06
frafuIn other words: what comes into debian/changelog: Everything from the ChangeLog from the package plus the changes necessary in the packaging?20:06
jdongScottK: a bit busy at the moment but if I remember lateri n the day, yes I will20:07
Toadstoolfrafu: what do you mean by sync? new upstream release or sync with Debian?20:07
frafunew upstream release20:07
Toadstoolfrafu: you don't have to specifically mention every single change in the upstream application then.  Just put the changes closing known LP bugs plus packaging changes, I'd say20:09
vemonfrafu, i think "New upstream version x.x (LP: #xxxxxxx)" is enough for the debian changelog. unless you've also made some changes to the packaging scripts20:10
vemonthis is for an updated (new upstream) package.20:10
ScottKIt is nice, but not required to mention major new features in debian/changelog as that's the one that gets shown to users on upgrade.20:11
Toadstooler... I'd rather use "New upstream version:" and enumerate relevant changes rather than just close random bug reports with (LP: #xxx)20:11
vemonyes, it's nice to mention the relevant new features20:12
vemonbut blind copying upstream ChangeLog changes to debian/changelog is probably not a good idea20:13
ScottKAgreed.  That's excessive20:13
frafuok; I think that I got the idea: the packaging changes have to be present; the LP# have to be present, and for the rest common sense applies by keeping in mind that it is the packaging changelog)20:17
frafuThanks for your input20:17
ScottKYou're welcome.20:18
ScottKKeep up the good work.20:18
frafuScottK: thanks;I will try  ;)20:19
bddebianHeya gang20:36
ScottKHeya bddebian20:38
bddebianHi ScottK20:40
ScottKslangasek: If you are around, I'd really appreciate bin Newing of gpsd to I can get the library transition done (only affects 3 packages, but I'd like to get it done before feature freeze as our GPS stuff is totally broken right now).21:04
mok0ScottK: I need a bit of help here21:11
* ScottK will try.21:13
warp10MOTUs, my package gbemol (a graphical frontend for MPD) is on REVU and waits for your reviews. http://revu.ubuntuwire.com/details.py?package=gbemol21:13
mok0Look at the earlier pastebin: http://pastebin.ubuntu.com/4378/21:13
mok0I got rid of most of it, but I can't figure out lines 15,17,1921:14
mok0those init scripts are part of other packages that this one depends on21:14
mok0the localhost package must simply configure those other packages and start the daemons21:15
ScottKWhy is it a separate binary package?21:15
mok0So when it doesn't own those init scripts, why does it complain?21:15
ScottKBecause it needs an init it doesn't ship.21:16
mok0ScottK: It's a "meta-package" that sets up a simple batch system just on 1 machine21:16
mok0ScottK: it does some reconfiguration of config files and restarts the daemons21:17
mok0ScottK: I could get rid of the package perhaps21:17
ScottKPackages aren't supposed to touch each other's config files.  I'd recommend consolidation.21:17
ScottKThat's without knowing the details of why you split it out though.21:18
mok0ScottK: consolidation?21:18
mok0= zapping? :-)21:18
ScottKPutting the stuff that's in that package in the same binary pacakage as ships the conf files21:19
ScottKerr inits21:19
mok0ScottK: I understand. Thinking about it, I think it is not a good idea to have this package. Perhaps I can supply a script that does the same instead.21:20
mok0ScottK: There's another question. If you have the source tree present, please take a look at torque-base.postinst and .postrm21:21
mok0prerm, sorry21:22
mok0ScottK: I'm doing something bad to /etc/services21:22
mok0(It works great but I don't think it's allowed)21:22
mok0The postinst script inserts a block in /etc/services, and prerm removes that same blcok21:23
ScottKAnd why are you doing this?21:25
ScottKIn general you are not allowed to touch stuff in /etc that your package doesn't own.21:25
mok0ScottK: Afair those entries need to be there or the software doesn't work21:26
mok0ScottK: there really ought to be a /etc/services.d/ where packages could drop entries21:26
torkelmok0: torque needs entries in /etc/services? Why? We have been running it for a couple of years without patching services21:31
mok0torkel: ok?21:31
mok0Then I get rid of that21:32
torkelmok0: are you packaging it from scratch?21:32
mok0I will ask to get the entries defined in net-base21:32
mok0torkel: yes21:32
torkelmok0: you haven't checked the package the SARA guys did?21:33
mok0torkel: nope21:33
mok0didn't know about it21:33
torkelcheck for torque_2_deb21:35
mok0torkel: thx21:35
mok0torkel: my package is very close to finished though21:35
ScottKmok0: I think you aren't supposed to modify /etc/services because I can't find any clear exception to allow it in policy.  I may be wrong though.21:36
mok0ScottK: I just heard from torkel that you don't need to patch /etc/services, so I'll just get rid of it21:37
mok0torkel: My packages are split off in into several, for clients, head-node, compute-nodes, etc.21:44
mok0torkel: perhaps you'd like to test it sometimes21:44
=== apachelogger_ is now known as OneTwink
OneTwinkhellboy195: ye don't wanna know that ;-)22:15
hellboy195OneTwink: ok ^^22:16
hellboy195OneTwink: lost bet?22:16
OneTwinkhellboy195: nah, insanity going round in #amarok22:17
hellboy195OneTwink: poor Harald :P22:18
OneTwinkhellboy195: well, I am party responsible for it ;-)22:18
hellboy195OneTwink: deleted branch for amarok2 alpha? ^^22:19
OneTwinkhellboy195: nope, I was cuddling with a fellow amarok teamster22:20
hellboy195OneTwink: lol. and how long will you be named OneTwink?#22:21
OneTwinkhellboy195: until we go back to normal again :P22:22
=== doko_ is now known as doko
hellboy195OneTwink: xD freaks. As long as you keep hacking on amarok! ^^22:23
OneTwinkhellboy195: we might stop developing and open up a partnership agency ;-)22:24
hellboy195OneTwink: omg xD22:24
frafuHello, I have a package here that has a debian/control and debian/control.in file. Can anybody please confirm that to add a new dependency to the package, I only have to add the dependency name to the dependency list in both packages?22:31
ScottKfrafu: Add it to control.in and control should get rebuilt.22:35
=== boomer` is now known as boomer
frafuScottK: I am creating a debdiff; will I have to rebuild control or will it be done by the person who applies the debdiff later?22:37
geserfrafu: the clean target will probably recreate debian/control from debian/control.in22:39
ScottKIt should get done automatically when you build the source package to be able to make the debdiff.22:39
* ScottK doesn't necessarily have a deep understanding of this, so I'd listen to geser.22:39
man-digeser: I would file bugs against packages doing this22:40
geserScottK: I'm not sure about the clean target, but as clean is called when the source target and the change propagates to debian/control is should happen in the clean target22:40
frafugeser: In other words debuild -S will create the new debian/control file; correct?22:41
man-difrafu: no22:41
geserman-di: some cdbs packages have control.in and @cdbs@ in Build-Depends. Do you know when debian/control gets recreated?22:41
man-difrafu: I know at least three packages with debian/control.in that just need to run "debian/rules debian/control"22:41
geserI remember also seeing package where I modified debian/control to only see in the debdiff that I've missed control.in and the change was lost22:42
man-digeser: problem is that buildd installs build-depends before starting build and debian/control regeneration could change build-depends22:43
frafuI am talking about gnome-control-center22:43
geserman-di: does this also apply to fields like uploaders? some gnome packages use control.in to fill in the team22:44
frafuunfortunately I don't have any experience with cdbs yet22:45
man-digeser: IMO its generally a bad idea to regenerate debian/control at build-time, not matter what fields change22:45
frafuman-di so, if I put the dependency also in debian/control, it will be available for buildd?22:46
frafuman-di:  I don't understand : just need to run "debian/rules debian/control"22:48
man-difrafu: I dont know about gnome-control-center but its this way in all packages I maintain with debian/control.in22:49
frafuSo what should I do apart taking a crash course in cdbs ? ;-)22:50
ScottKgeser: I know Debian Python Modules Team used to generate uploaders that way for a while and gave it up as a bad idea.22:50
geserI've looked at the gnome-control-center package and it does it for uploaders22:51
* ScottK holds his nose.22:51
man-diScottK: yeah, this stinks ;-)22:51
frafuman-di what is what in all packages you maintain with debian/control.in?22:52
man-difrafu: in my packages we generate Depends and Build-Depends differently for Debian and Ubuntu22:53
man-difrafu: and also some Depends can be configured at build time22:54
frafuSkottK, man-di: and I have to cope with it in order to only add a dependency :-/22:54
frafuis there some tutorial/guide about it?22:55
ScottKfrafu: For your purposes if you make the change in both control and control.in and debdiff the resulting package, it should be safe.22:57
=== OneTwink is now known as apachelogger
frafuScottK: I guess that you are right since geser told above that g-c-c does it for the uploaders (I suppose that he meant that it does it not for the dependencies)23:02
frafugeser,  man-di: do you concur?23:02
geserfrafu: simply compare control and control.in to see what gets replaced/filled in23:03
geserfrafu: changing control.in and checked that both files are changed in the debdiff should be enough but with changing both files you are on the safe side23:04
frafuYes, I changed both files, called debuild -S, created the debdiff and both changes were in the debdiff; to be precise there were exactly and only all the changes that I made in the debfile.23:08
frafugeser, SkottK,  man-di: thanks for all your input;  I have to quit now; bye23:09
=== Kopfgeldjaeger is now known as hrist2
=== hrist2 is now known as Kopfgeldjaeger
=== |Baby| is now known as Baby
mok0Where in the package should I document override files?23:31
Babyprobably changelog, I guess23:32
mok0Baby: yeah23:33
minghuaYeah, and I wouldn't call a changelog entry "documenting"...23:33
minghuaNot that an override needs documenting, of course.23:34
mok0minghua: ScottK told me to23:39
mok0minghua: "I think this should either be corrected or explained and have an over-ride"23:40
minghuamok0: I hope a changelog entry would be enough explanation.23:40
ScottKmok0: I meant explained on REVU why the over-ride was appropriate.23:40
ScottKAnd in debian/changelog.23:41
minghuaHey ScottK.23:41
mok0ScottK: I was commenting on minghua's statement that overrides does not need documenting...23:42
minghuaJust to be clear, I meant a debian/changelog entry with explanations would be good enough for an override.23:44
mok0minghua: ok, sorry for misrepresenting your views23:44
mok0This override file, should it contain the whole lintian warning line, or just a unique part of it?23:46

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!