[14:59] <dholbach> hello everybody
[14:59] <dholbach> welcome to the MOTU Q&A Session!
[14:59] <dholbach> who do we have here - can we go through a round of introductions?
[14:59] <Amaranth> Oh, this isn't Open Week yet
[15:00] <dholbach> I'm Daniel Holbach, MOTU for quite a while now and trying to make becoming a MOTU as easy and enjoyable as possible.
[15:01] <dholbach> who else do we have here?
[15:01] <persia> I'm Emmet Hikory, MOTU, and active with QA and library porting.
[15:01] <fernando> I'm Fernando Ribeiro, trying to be a MOTU.
[15:02] <dholbach> hey rulus, ntaylor0909, nalioth_, welcome to the session - we're just going through a round of introductions
[15:02] <sommer> me... Adam Sommer I'm relatively new to the community and looking for ways to contribute. Was at another QA meeting, but was called away before the end.
[15:02] <zul_> im chuck short, member of the motu team amongst other things
=== dholbach high-fives sommer
[15:02] <geser> I'm Michael Bienia, MOTU
[15:03] <dholbach> who of you is interested in becoming a MOTU and has prepared some questions for us?
[15:03] <rulus> dholbach: I'm just listening
[15:03] <gaspa> andrea gasparini, not motu.
[15:03] <gaspa> :-P
[15:03] <sommer> I'm interested, but didn't come prepared with anything specific.
[15:03] <dholbach> sommer, rulus, gaspa: great to have you around
[15:04] <dholbach> anything you'd like to know about MOTU? contributing or ubuntu development?
[15:04] <gaspa> there' s already a roadmap, or a list of things that i can do to help ubuntu development?
[15:05] <dholbach> gaspa: sure
[15:05] <dholbach> on http://wiki.ubuntu.com/MOTU/TODO we list quite a lot of different things you can start looking into
[15:05] <dholbach> the 'bitesize' bugs for example are suited for new contributors for example ===might* be easy to fix
[15:05] <gaspa> dholbach: ok, i saw only the hardy roadmap, but it's quite vague.
[15:06] <dholbach> gaspa: we'll be fixing a lot of bugs, so I think that's a good way to get involved in the MOTU team
[15:06] <gaspa> yep, just starting in it
[15:07] <dholbach> http://wiki.ubuntu.com/MOTU/Recipes will help you to play with the tools we use most
[15:07] <norsetto> perhaps gaspa is referring to this: https://wiki.ubuntu.com/UbuntuDevelopment ?
[15:07] <huats> gaspa: and sometimes if you are doing stuffs on a bug with a MOTU called norsetto as a mentor on this bug, be prepared to do a lot of things, even for a bitesize bug... But you'll learn a lot...
[15:08] <norsetto> huats: I never asked you to do a triple somersault .....
[15:08] <dholbach> hey nealmcb
[15:08] <dholbach> norsetto: not yet
[15:08] <persia> gaspa: You might also take a look at https://wiki.ubuntu.com/MOTU/Contributing, which has some suggestions on using the tools, and some of the basic processes for getting things uploaded.
[15:08] <huats> dholbach: :-D
[15:08] <gaspa> norsetto: I was referring to https://wiki.ubuntu.com/HardyReleaseSchedule?highlight=%28hardy%29
[15:09] <dholbach> gaspa: right, that page only explains the various phases of the release cycle we'll go through
[15:09] <nealmcb> dholbach: howdy. I was late to the other discussion, just caught up on the log, and figured I'd lurk some more here
[15:09] <dholbach> and it's not even a final document yet
[15:09] <norsetto> gaspa: yes, but I think if you want to know more about ubuntu development, you really should look at the wiki link I posted
[15:09] <gaspa> dholbach: in fact i wasn't satisfied.
[15:09] <dholbach> nealmcb: excellent
[15:09] <persia> gaspa: That's more a schedule to keep track of how release management affects our work. The goals are more to implement as many features as we want prior to Feature Freeze, and fix as many bugs as possible.
[15:10] <dholbach> any other questions around packaging, MOTU, Ubuntu Development, processes?
[15:11] <huats> dholbach: I have one question for you
[15:11] <dholbach> huats: fire away
[15:11] <norsetto> dholbach: how can a contributor submit a patch?
[15:11] <dholbach> norsetto: good question
[15:12] <dholbach> it's all explained at http://wiki.ubuntu.com/SponsorshipProcess
=== norsetto writes that down
[15:12] <huats> right now, do what distribution we put in the changelog ? gutsy or gutsy-proposed ?
[15:12] <dholbach> the gist of it is: if you're fixing a bug, attach the patch that you've produced and subscribe the reviewer team to it
[15:12] <dholbach> for main/restricted that's ubuntu-main-sponsors
[15:12] <norsetto> huats: hardy
[15:12] <dholbach> for universe/multiverse that's ubuntu-universe-sponsors
[15:13] <Amaranth> Whoa even I didn't know about requestsponsor
[15:13] <dholbach> If you've never produced a patch before: https://wiki.ubuntu.com/MOTU/Recipes/Debdiff might help you with that
[15:14] <dholbach> hiya blueyed
[15:15] <blueyed> Hi dholbach
[15:16] <dholbach> huats: if you want to fix a bug in gutsy (it's already release, so we can't upload to 'gutsy' itself any more), you'll need to follow http://wiki.ubuntu.com/StableReleaseUpdates
[15:16] <dholbach> (which has both instructions for main and universe)
[15:16] <gaspa> ok, if i decide to write a patch for a bug in https://wiki.ubuntu.com/MOTU/TODO, what should i do to prevent someone else from making the same work?
[15:17] <norsetto> gaspa: you should assign the bug to yourself
[15:17] <gaspa> norsetto: i can't
[15:17] <gaspa> i launchpad i can't modify nothing of the bug.
[15:17] <norsetto> gaspa: you can't!? Ask for membership in the bugsquad then
[15:17] <dholbach> gaspa: you need to login to launchpad and click for example on the status of the bug
[15:17] <dholbach> norsetto: that's not required
[15:18] <norsetto> dholbach: ah, yes, perhaps he is not logged
[15:18] <gaspa> ... i wasn't able to do it... i'll retry...
[15:18] <gaspa> no ,i'm logged...
[15:18] <dholbach> gaspa: click for example on the status of the line in the middle of page, which has the status/sourcepackage etc information
[15:19] <gaspa> i'll retry later, maybe i lost something.
[15:19] <dholbach> gaspa: and follow up with information on the bug as you have new findings to report, that's also important to not do the same work twice
[15:19] <dholbach> gaspa: alright, let me know if things still go wrong, or ask on #launchpad
[15:19] <gaspa> yep
[15:19] <gaspa> ah, but...
[15:20] <dholbach> welcome nxvl
[15:20] <gaspa> if I assign the bug to me, when does it becomes fixed? when i attach a debdiff?
[15:20] <persia> gaspa: When it gets uploaded. The changelog in your debdiff will close the bug.
[15:21] <persia> gaspa: You'll unassign yourself when you add the debdiff, and use http://wiki.ubuntu.com/SponsorshipProcess to get a sponsored upload.
[15:21] <seb128> dholbach: thanks to the rocking contributors we had in the desktop team this cycle
[15:21] <norsetto> gaspa: when you subscribe u-u-s, if your patch is accepted, the sponsor will mark the bug as fix-committed, and then, as persia said
[15:22] <dholbach> seb128: care to advertise the DesktopTeam and what contributors can do in it?
[15:22] <seb128> the desktop team is a wonderful place to contribute to the Ubuntu desktop
[15:23] <coNP[uni]> Desktop Team definitely rocks
[15:23] <seb128> we have a nice IRC channel (#ubuntu-desktop), a mailing list, a wiki (https://wiki.ubuntu.com/DesktopTeam)
[15:23] <seb128> the wiki has a summary of the team and pointers, we also have a TODO on https://wiki.ubuntu.com/DesktopTeam/TODO
[15:24] <seb128> we welcome questions, etc on #ubuntu-desktop and usually contributors find sponsor easily for their patches or updates so feel free to join
[15:24] <seb128> dholbach: good enough ?
[15:24] <dholbach> seb128: sounds good - especially the kick-ass DesktopTeam/TODO page
[15:24] <huats> seb128: If I wasn't already convinced I would have been....
[15:25] <seb128> coNP[uni] is one of the rocking contributors which joined the team this cycle
[15:25] <seb128> some things we are looking at for hardy BTW
[15:25] <coNP[uni]> It is not impossible to become a MOTU
[15:25] <coNP[uni]> [ 15:26] <seb128> - packaging GNOME 2.21.n to 2.22.1, listing annoy glitches laying around for too long that we should fix for the coming lts, make the wiki documentation nicer, and maybe switch to packaging in bzr if we can make the workflow easy enough
[15:26] <dholbach> that sounds like a lot of action in the hardy cycle
[15:26] <seb128> yes
[15:27] <seb128> would also be nice to have some webpage summary for things like version of the package in ubuntu, debian, upstream
[15:27] <seb128> patches browsable easily for upstream also
[15:27] <dholbach> also experimental versions
[15:28] <seb128> well, I've abused enough this classroom with desktop advertising, I'm waiting for you on #ubuntu-desktop now
[15:28] <dholbach> but yeah - sounds like a nice small project for somebody
[15:28] <dholbach> ok. any other questions?
[15:29] <dholbach> maybe a problem somebody has, we look at together?
=== norsetto welcomes giftnudel to the MOTU Q&A session
[15:31] <sommer> /me admits to still reading the packaging guide
[15:31] <dholbach> sommer: that's no problem
[15:31] <dholbach> norsetto, coNP[uni], huats: do we have an easy bug we can tackle in the remainder of the session?
[15:31] <persia> sommer: That's great. Reading is probably the best way to get questions
=== norsetto checks
[15:32] <sommer> is it better to be proficient with packaging from scratch section before moving on?
[15:33] <sommer> or jump to the more automated tools?
[15:33] <norsetto> bug 154399 seems a nice bitesized one
[15:34] <norsetto> ubotu seems to be sleeping: https://bugs.edge.launchpad.net/ubuntu/+source/util-linux/+bug/154399
[15:34] <dholbach> sommer: I think it's good to work on easy bugs to get a feeling of what you're dealing with
[15:34] <persia> sommer: I'd recommend reading the entire guide to build familiarity, and then working with a bunch of different packages while chasing bugs, to see different implementations. That will likely give a better understanding than trying to understand completely from the guide.
[15:34] <sommer> persia, dholbach: ah, sounds good
[15:34] <dholbach> norsetto: ok, so how do we attempt to fix it?
[15:35] <persia> sommer: That's the big advantage of open-source it's easy to build on others work.
[15:35] <norsetto> dholbach: well, I would first check if nfs-utils is in the repo
[15:35] <norsetto> dholbach: for instance apt-cache search nfs-utils
[15:36] <norsetto> make sure you have an updated cache first
[15:36] <norsetto> anyone has done that?
[15:36] <dholbach> I don't have it
[15:37] <dholbach> I have a source package called nfs-utils though - I used apt-cache showsrc nfs-utils for that
[15:37] <norsetto> anyone know what that is?
[15:38] <norsetto> do you know what is a source package?
[15:38] <norsetto> dholbach: will you or shall I?
[15:38] <dholbach> norsetto: go ahead
=== coNP[uni] votes for norsetto
[15:39] <sommer> isn't a source package a package that will build multiple binaries?
[15:39] <norsetto> ok, I try .....
[15:39] <coNP[uni]> sommer: you can have one or more binaries built from a source package
[15:39] <norsetto> sommer: yes, a source package its a package that contains, as the name says, the sources, and the instructions for the build system on how to make binaries out of it
[15:40] <norsetto> you download source packages with the apt-get command too; for instance apt-get source nfs-utils
[15:40] <geser> if you have a deb-src line in /etc/apt/sources.list
[15:41] <norsetto> so, you see, in this case there is a source package for nfs-utils, but there is no binary package
[15:41] <norsetto> geser: yes
[15:41] <dholbach> and the binary package is the .deb package that gets actually installed
[15:41] <norsetto> if you do an apt-cache search nfs-utils you will see that nfs-utils is actually provided by another package, its a virtual package
[15:42] <coNP[uni]> Actually you can also download a source package for a binary package, with apt-get source <binarypackagename>
[15:42] <coNP[uni]> So you don't have to know the source package name.
[15:43] <norsetto> now you just have to check if the man page object of the bug is really provided by the nfs-common package
[15:43] <norsetto> one way to do it (there are many) is to download it, and check its content with a dpkg -c command
[15:44] <giftnudel> isn't it sufficient to look at the file list for that?
[15:45] <norsetto> giftnudel: you mean the one at http://packages.ubuntu.com/? As I said, there are many ways to do it
[15:45] <giftnudel> well, you explained it already
[15:45] <sommer> so to address the bug, would you change the mount man page to say install nfs-common instead of nfs-util?
[15:45] <norsetto> sommer: if the man page is indeed in nfs-common, yes
[15:46] <sommer> norsetto: but it may not be?
[15:46] <sommer> ah you check the nfs-util package, correct?
[15:46] <norsetto> sommer: I would also check the options section, to see that it indeed contains the mount options for nfs and nfs4
[15:47] <giftnudel> ls
[15:47] <giftnudel> ups
[15:47] <geser> sommer: this is also a check if the information from the bug is correct
[15:48] <sommer> yep, mount has options for nfs and nfs4
[15:49] <norsetto> sommer: we should check that in the nfs man page too
[15:50] <sommer> norsetto: okay, I'm not finding anything in the mount man page about nfs-utils though.
[15:51] <norsetto> sommer: what distribution are you using?
[15:51] <sommer> norsetto: Gutsy
[15:51] <giftnudel> well I can see it there
=== geser too
[15:51] <giftnudel> right at the bottom
[15:51] <sommer> I did man mount... is that right
[15:51] <giftnudel> yes
[15:52] <giftnudel> then go to the end and you should see it
[15:52] <sommer> woops I'm blind
[15:52] <sommer> I see it now
[15:53] <norsetto> sommer: I think it is because using space you end up with the header just at the end, so you may miss it
[15:54] <sommer> norsetto: I used /nfs to do a search and when it highlighted the nfs part I skipped over the -utils at the end... heh
[15:55] <norsetto> has any of you tried to check the nfs-common package, and its nfs page?
[15:55] <giftnudel> at least the dpkg seems to contain it and debian/nfs-common.manpages wants to install it
[15:55] <sommer> norsetto: trying to install it now.
[15:56] <norsetto> also checking it with: http://packages.ubuntu.com/cgi-bin/search_contents.pl?word=nfs.5.gz&searchmode=searchfilesanddirs&case=insensitive&version=gutsy&arch=i386
[15:57] <norsetto> ok, so we checked the correctness of the bug report, the solution given was correct, do we all agree?
[15:57] <sommer> I concur
[15:58] <dholbach> what do we do now?
[15:58] <norsetto> how do we actually patch this out, anyone?
[15:58] <giftnudel> well, in the mount source package change the string?
[15:58] <sommer> move the man page back to nfs-util, or change the man page of mount?
[15:59] <giftnudel> the first idea is also nice
[15:59] <norsetto> sommer: well, I think changing the man page of mount is more appropriate, but, is there anything else we should check first?
[15:59] <sommer> why it was changed?
[16:00] <norsetto> sommer: exactly .....
[16:00] <sommer> probably in nfs-util changelog?
[16:00] <norsetto> sommer: could be, we could also ask upstream, perhaps is something specific to ubuntu
[16:01] <giftnudel> maybe there's a patch in debian already ...
[16:01] <norsetto> giftnudel: good point
[16:01] <norsetto> norsetto: anyhow, for the sake of the exercise, lets suppose we need to patch this
[16:02] <norsetto> how would we go about it?
[16:03] <sommer> apt-get source mount
[16:03] <norsetto> in which source package is the mount man page?
[16:03] <sommer> then find the man page file and edit it
[16:04] <sommer> util-linux?
[16:04] <giftnudel> sommer: yes, it's in there
[16:04] <norsetto> sommer: thats what the reported said, seems like he knows what is doing, but lets check it
[16:04] <proppy> hi, Q&A still going on ?
[16:05] <dholbach> proppy: right now norsetto helps us to fix https://bugs.launchpad.net/ubuntu/+source/util-linux/+bug/154399
[16:06] <norsetto> ok, so everybody agrees is in there?
[16:06] <proppy> ok
[16:06] <giftnudel> finding the offending file is not that difficult, but then
[16:07] <norsetto> giftnudel: yes?
[16:07] <giftnudel> editing is not appropriate, so we need a patch, right?
[16:07] <proppy> using cdbs-edit-patch ?
[16:08] <norsetto> proppy: is the package using it already?
[16:08] <giftnudel> how do you see that?
[16:08] <sommer> giftnudel: why isn't editing the file appropriate?
[16:08] <proppy> by checking dependencies in control and include in rules
[16:09] <norsetto> there is also a nice util you can use to determine what patch system a package is using, its called what-patch, and its in ubuntu-dev-tools
[16:09] <huats> sommer: since you won't easily remember every modification you've done... so there are tools to help you.... norsetto just mentions...
[16:09] <persia> sommer: giftnudel: It depends on the package. In general, patches are preferable to direct edits, but some packages don't use patches, and we avoid adding patch systems. The tool norsetto advises may be the easiest way to decide.
[16:10] <huats> sommer: you will edit the file, but with the tool
[16:10] <sommer> ah... thanks I'm with ya
[16:10] <nealmcb> pardon a query out of the blue: what is u-u-s?
[16:10] <proppy> sommer: if the file you need to edit is not in debian/ then I've been told you need to patch
[16:10] <norsetto> nealmcb: ubuntu-universe-sponsors
=== proppy apt-get sourcing the package
[16:11] <dholbach> nealmcb: http://wiki.ubuntu.com/SponsorshipProcess
[16:11] <giftnudel> it doesn't seem that util-linux has patches in it already, at least I see none and the control file doesn't have dependencies on a patch system
[16:11] <giftnudel> proppy: that's what I remembered, too
[16:12] <sommer> so it's cool to directly edit this package?
[16:12] <norsetto> giftnudel: I think you are correct, so in this case we have two options, either we do an inline patch or we include ourselves a patch system
[16:13] <norsetto> personally, I will never go for inline, even in a simple case like this, but others may have different opinions
[16:13] <nxvl> what are we working at? i losted lot of the conversations :S
[16:13] <persia> norsetto: Please reconsider. For many packages, we try to sync back to Debian, and some Debian maintainers are strongly opposed to patch systems, as it breaks their workflow.
[16:14] <proppy> the package seems to do some man tweaking already
[16:14] <proppy> perl -pi.bak -e 's/agetty/getty/g' debian/util-linux/usr/share/man/man8/getty.8 \
[16:14] <norsetto> persia: yes, in that case I simply revert it back when we merge
[16:14] <proppy> debian/rules:71
=== zul_ is now known as zul
[16:15] <norsetto> but persia is right, we should try to stick with what the debian maintainer is doing, so, in this case we might have to do, as proppy suggested, a little hacking
[16:15] <sommer> sweet... lets hack!
[16:15] <giftnudel> its only one line, so it's probably not bad
[16:16] <proppy> "as proppy suggested, a little hacking" I like the sound of that
[16:17] <proppy> norsetto: by inline you mean using sed/python/perl/awk/cat ?
[16:17] <norsetto> proppy: or simply manually changing the source
[16:18] <norsetto> the change will be in .diff.gz
[16:18] <proppy> norsetto: how can you change something which is not in debian ?
[16:18] <proppy> so you will end with a .diff.gz with 'outside of debian/ changes ?
[16:18] <proppy> norsetto: I thought it was not accepted
[16:19] <norsetto> proppy: its debatable, personally, I don't like it, it makes it difficult to trace your changes
[16:19] <giftnudel> proppy: probably a case of "Never ever do that, but now, it's just easier"
[16:20] <proppy> giftnudel: norsetto: understood thanks for clearing it up
[16:20] <sommer> norsetto: why is it harder to trace changes? Wouldn't it show up in the debdiff? also documented in changelog?
[16:21] <norsetto> sommer: yes, but it is showing up in the .diff.gz too, and that change will be incremental
[16:21] <dholbach> sommer: imagine you have 10 different patches all applied directly to the source
[16:21] <norsetto> sommer: so, suppose you do 10 changes, how do you associate the changelog to the change?
[16:21] <dholbach> sommer: they affect 30 different files, now you want to remove one patch of them, because you understood that it's obsolete
[16:21] <proppy> norsetto: I guess we need also to patch SEE ALSO ?
[16:21] <dholbach> it's easier, if you can just remove one patch file in debian/patches
[16:22] <sommer> ah... I'm with you
[16:22] <persia> sommer: Also consider that upstream may want your patches, and may only apply a subset. It's a lot easier to deapply them separately when you manage them separately.
[16:22] <norsetto> and you know immediately which one
[16:22] <dholbach> great
[16:22] <sommer> thanks, that makes sense now.
[16:22] <dholbach> excellent
[16:24] <giftnudel> so for just a simple fix like that, there's nothing that speaks against it, but for bigger patches, individual patches are better
[16:24] <norsetto> so, what do we do once you have your patch? In this case corrected the line in the man page?
[16:24] <norsetto> somebody mentioned a changelog?
[16:24] <sommer> dch -i?
[16:24] <norsetto> sommer: for instance
[16:25] <norsetto> what is it going to be the new version number?
[16:25] <persia> giftnudel: rather, it's better to follow the package you're working with. Even a one-line change deserves a dedicated patch if the package has an patch system in use.
[16:25] <giftnudel> right, I forgot that
[16:26] <sommer> norsetto: good question... I'm still fuzzy on version numbering.
[16:26] <proppy> norsetto: perl -pi.bak -e 's/nfs \(5\)/nfs-common \(5\)/g' correct ?
[16:27] <norsetto> sommer: its not difficult, check what is in the changelog already
[16:27] <norsetto> proppy: I like it when you talk to me like that
[16:27] <proppy> norsetto: ubuntu2 ?
[16:28] <norsetto> proppy: yes, in this case the old one was 2.13-8ubuntu1, so, since this is an ubuntu revision, we will use 2.13-8ubuntu2
[16:28] <norsetto> what if the old one was 2.13-8 ?
[16:28] <giftnudel> proppy: well, should nfs-utils not be changed to nfs-common?
[16:29] <proppy> giftnudel: you're right, my patch is bad
[16:29] <proppy> giftnudel: I missread the bug report
[16:30] <sommer> in 2.13-8. -8 is the debian package number correct?
[16:30] <proppy> giftnudel: the package is to be changed, not the manpage name
[16:30] <dholbach> sommer: yes
[16:30] <dholbach> sommer: if we do an ubuntu change, we simply add ubuntu1 and then increment it
[16:30] <sommer> dholbach: gotcha
[16:30] <dholbach> sommer: if we were the first to package 2.14 (which is not in debian yet, we'd have 2.14-0ubuntu1
[16:31] <sommer> okay... so the 8 isn't the 8th ubuntu version?
[16:31] <norsetto> ok, anything else we should do?
[16:31] <norsetto> sommer, no its the debian revision
[16:31] <persia> sommer: That's the Debian revision. For any -XubuntuY, X is the Debian revision, and Y is the Ubuntu revision.
=== dholbach is about to call it a day
[16:32] <norsetto> a little thingie that we should not forget .....
[16:32] <dholbach> thanks everybody for showing up, thanks norsetto for running the biggest part of the session
[16:32] <norsetto> since we made an ubuntu modification, we should also change the maintainer
[16:32] <proppy> giftnudel: perl -pi.bak -e 's/nfs-utils package must be installed/nfs-common package must be installed/g' better ?
[16:32] <sommer> cool... I think I'm good on revision numbers now too, thanks fellas
[16:33] <sommer> dholbach: thank you
[16:33] <dholbach> rock on
[16:33] <giftnudel> I wanted to help so often, but everytime I wanted to fix such an easy thing, there was so much I needed to think of and then I resigned again ...
=== dholbach hugs y'all - see you around, guys
[16:34] <norsetto> giftnudel: please come to #ubuntu-motu and ask for help, we are there also for this
Go back to MOTU/Q&A/Logs.