[05:30] <dholbach> hi everyone
[05:30] <ogra> hi everybody, lets start the meeting then
[05:30] <zul> hey
[05:31] <ogra> i would first like to do some administrativa before we come to the agenda
[05:31] <dholbach> which you can read here http://www.ubuntulinux.org/wiki/MOTUMeeting
[05:31] <ogra> how often do we want to do this meeting
[05:31] <ajmitch> at this hour? not too often 
[05:32] <ogra> we can have a rotating timeschedule, but i meant how frequently
[05:32] <pitti> just see how it goes, how much of the agenda will still be left, and decide that afterwards
[05:32] <ogra> weekly, 2 weekly, monthly
[05:32] <ogra> ok
[05:33] <ogra> so lets go to the agenda...
[05:33] <ogra> how do we handle packages in universe that are newer than in Sid or will never enter debian but ubuntu etc
[05:33] <herve> we can still append ubuntu1 to the revision 
[05:33] <ogra> there was something on the wiki a while back, cant look t up right now..
[05:33] <herve> in case thy enter Debian some day
[05:33] <pitti> you should start with ubuntu0
[05:34] <ogra> ...it said something like upstream must be actively maintained
[05:34] <pitti> so that the first Debian version will be never
[05:34] <ajmitch> pitti: we've been doing -0ubuntu1
[05:34] <pitti> -> then you can sync
[05:34] <dholbach> not -0ubuntu1?
[05:34] <pitti> ajmitch: argh, right
[05:34] <pitti> I meant this
[05:34] <pitti> sorry
[05:34] <dholbach> ok
[05:34] <dholbach> but the problem is the interaction with debian people
[05:35] <pitti> ^ a technical or a social one?
[05:35] <ogra> i think its up to the DD to choose that for his pkg
[05:35] <dholbach> atm i can't say anything positive about it
[05:35] <dholbach> yeah... i had no reaction until this point
[05:35] <herve> why Debian wouldn't do the opposite of what we do?
[05:36] <herve> sync Ubuntu packages into Debian
[05:36] <herve> our definition of free software is close enough
[05:36] <dholbach> herve: because they don't have the syncing mechanism
[05:36] <ajmitch> herve: they might, if people upload into debian by their procedures
[05:36] <Kamion> it's a matter for individual maintainers
[05:36] <herve> I meant manually :)
[05:36] <ogra> herve, that up to them
[05:36] <dholbach> it's each DD's decision
[05:36] <ogra> tahts even
[05:36] <ogra> argl
[05:37] <herve> so we would need some mentor like new packagers :-)
[05:37] <pitti> the problem is, if you have a NEW package, then there is no default DD for it
[05:37] <dholbach> if people used cdbs-tarball-dpatch/simple-patchsys (ok, only in the cases where it's possible), giving back would be a lot easier
[05:37] <pitti> if you want it in Debian, you have to file an RFP
[05:38] <pitti> dholbach: ++
[05:38] <ajmitch> dholbach: not everything fits well with cdbs :)
[05:38] <ajmitch> or an ITP if you intend to maintain it
[05:38] <dholbach> ajmitch: that's what i just said :-)
[05:38] <pitti> ajmitch: no, but most packages do
[05:38] <herve> dholbach, got the message ;)
[05:38] <dholbach> but that doesnt apply to NEW packages
[05:38] <pitti> ajmitch: right, if you want to maintain it in Debian yourself, there is no problem, right?
[05:38] <ajmitch> pitti: yep
[05:39] <pitti> I thought the problem were if you don't want/can't maintain it in Debian
[05:39] <ogra> but we will have a lot MOTUs that are no DD in the future i guess
[05:39] <dholbach> can't any of you ubuntu-DDs be the sync-NEW-from-ubuntu--guy? :-)
[05:39] <pitti> btw, Debian can't sync pacakges because they don't want ubuntu version numbers
[05:39] <ogra> they would have to take maintainership
[05:39] <pitti> dholbach: no, that would be a PITA
[05:39] <ajmitch> ogra: sure, you don't have to be a DD to maintain debian packages, just to upload
[05:39] <dholbach> pitti: yes... i can see that
[05:40] <pitti> dholbach: we can certainly sponsor uploads, though
[05:40] <pitti> but not maintain
[05:40] <ogra> but lets not discuss the debian part here, its up to them
[05:40] <pitti> for the record, I'm fine with sponsoring Debian package uploads
[05:40] <haggai> as long as it is easy for debian to use them
[05:40] <pitti> as long as I don't have to review them every time
[05:40] <dholbach> pitti: i'll get back to you :-)
[05:40] <ogra> our part shold be maintainig a list of such packages
[05:41] <dholbach> ogra: which ones?
[05:41] <herve> pitti, you could trust a package that was trusted to enter Ubuntu?
[05:41] <haggai> good idea.  Time for a new wiki page listing new packages not in Debian?
[05:41] <ogra> so if debian is interested they can request a sync
[05:41] <pitti> herve: "trust" in which sense?
[05:41] <ogra> dholbach: NEW
[05:41] <dholbach> ok
[05:41] <dholbach> ubuntu-NEW :-)
[05:41] <herve> pitti, uploading without reviewing
[05:41] <pitti> herve: I don't speak about NEW packages
[05:41] <pitti> herve: just updates
[05:41] <ogra> dholbach: yeah
[05:41] <ajmitch> herve: I wouldn't expect ubuntu packaging to always be perfect
[05:41] <pitti> herve: I can't process NEW packages any way
[05:41] <haggai> having such a wiki page would also be a reference for people wondering what additional stuff ubuntu has
[05:42] <ogra> ajmitch: i would...(after some time)
[05:42] <herve> pitti, ok you would prefer to review each update, ideally
[05:42] <pitti> herve: I mean, the actual maintainer has to care for his bugs, I only would sponsor the upload
[05:42] <ajmitch> ogra: only when someone has time to review the packages :)
[05:42] <Mithrandir> sorry, was off for food.
[05:42] <pitti> however, I don't think that we should automatically upload all new Ubuntu packages to Debian
[05:42] <ogra> ajmitch: i hope we dont drop the crossreviewing
[05:43] <ogra> pitti: never ever
[05:43] <herve> could some ubuntu platform or mailing list be set as the maintainer email of such packages?
[05:43] <pitti> filing RFPs with a link to the Ubuntu package is more than enough IMHO
[05:43] <ajmitch> ogra: yes, but when people are MOTUs they can upload without getting reviews
[05:43] <pitti> then the Debian community can decide 
[05:43] <ogra> ajmitch: we will have a poicy for NEW packages....
[05:43] <pitti> ++ for the wiki page with new stuff
[05:44] <ogra> ok..
[05:44] <ogra> noted
[05:44] <ogra> ok, next point ? 
[05:44] <dholbach> ubuntu-NEW packages
[05:44] <dholbach> and their process
[05:44] <ogra> yup
[05:45] <ogra> i find the suggestion of dholbach very tempting
[05:45] <dholbach> the idea is quite clear: we need to have a thorough review of those packages
[05:45] <dholbach> 3 MOTUs is ok in my eyes
[05:45] <pitti> what about mandating, or at least urging a sane pacakging format?
[05:45] <Mithrandir> ogra: to have a couple of MOTUs reviewing packages before they are uploaded?
[05:45] <ajmitch> pitti: later in the agenda :)
[05:45] <pitti> I mean for new ubuntu packages
[05:45] <pitti> ok
[05:46] <ogra> Mithrandir: yup
[05:46] <pitti> ajmitch: "urge Debian people..." -> that's not my point, but nevermind
[05:46] <ajmitch> pitti: ah, sorry, misread it
[05:46] <ajmitch> Mithrandir: yes, any of the MOTU team, not just the leads I think
[05:47] <dholbach> any other ideas, proposals?
[05:47] <pitti> I'd mandate cdbs+tarball and a proper patch system (simple-patchsys, quilt, whatever); this comes close to the future dpkg format and is really good
[05:47] <ogra> just to make sure 6 or more eyes have seen it
[05:47] <pitti> ogra: we should collect a list of more folks who are willing to do review
[05:47] <Mithrandir> pitti: I find making cdbs packages really icky when you have a bit of complexity.  It has no docs.
[05:47] <pitti> i. e. I'd be willing to review packaging
[05:48] <jani> pitti, so cdbs preferred over plain debhelper?
[05:48] <pitti> for _some_ pacakges
[05:48] <ogra> pitti: we do a lot of cross reviewing in #ubuntu-motu
[05:48] <pitti> Mithrandir: right, the docs should be improved, but it rocks; however, dbs is fine, too
[05:48] <pitti> jani: it's only a proposal, but cdbs really rocks and avoids many errors (and makes packaging easy, btw)
[05:48] <ajmitch> perhaps jbailey or someone could actually document cdbs one day :)
[05:49] <pitti> my point is, if we have one common packaging format, review is easy
[05:49] <Mithrandir> pitti: we don't want to get people to use dbs.
[05:49] <ogra> pitti, but we always have packages nobody feels smart enough to review, like hula
[05:49] <dholbach> no other opinions on the ubuntu-NEW thing?
[05:49] <pitti> :-)
[05:49] <haggai> dholbach: might be worth noting it only applies to MOTUs, not maintainers?
[05:49] <pitti> ogra: during my security work, I've seen so many braindead packaging formats, it really hurts
[05:49] <pitti> ogra: we must not make this error again
[05:49] <ogra> heh
[05:49] <ogra> ok
[05:49] <dholbach> haggai: this is MOTUMeeting :-)
[05:49] <pitti> so enforcing a strong packaging policy would really rock
[05:50] <dholbach> pitti: ++
[05:50] <haggai> dholbach: I know but you're talking about documenting a process
[05:50] <Mithrandir> I think that having patches in arch can also be fine, if the arch repo is mirrorer publically
[05:50] <pitti> Mithrandir: you mean patches which are directly in the debian diff.gz?
[05:50] <dholbach> can we please get to the patch-thing later?
[05:51] <pitti> ok
[05:51] <Mithrandir> yes, and {arch} directories in there too
[05:51] <pitti> argh
[05:51] <dholbach> haggai: you're right... there should be exceptions to it
[05:51] <pitti> Mithrandir: arch dirs in diff.gz? that sucks
[05:51] <dholbach> haggai: i wouldn't tell seb128 to let me review his packages :-)
[05:52] <Mithrandir> t would be interesting to have a MOTU review some of my packages. :)
[05:52] <ajmitch> Mithrandir: if we're up to the task :)
[05:52] <ogra> haggai, i think forcing a policy that NEW packages must get reviewed is a good first set in a process
[05:52] <pitti> ++
[05:52] <ogra> s/reviewed/crossreviewed
[05:53] <herve> by at least 3 persons?
[05:53] <pitti> ogra: the lintian output should be published somewhere
[05:53] <ogra> additionally i would like to have the "must actively be maintained upstream or you take it yourself" in
[05:53] <mvo> yes, I think this review is good because we have much less "ownership" on packages than e.g. debian
[05:53] <ajmitch> herve: yes, we've often had reviews from 2-3 people already
[05:53] <ogra> s/take it/take it over
[05:54] <dholbach> ogra: ++
[05:54] <herve> ok so you can conclude 3 is enough?
[05:54] <Mithrandir> yes, more seems a bit excessive
[05:54] <ajmitch> yes, we don't want too many
[05:54] <crimsun> 3's sufficient to me
[05:54] <ogra> for a initila review, its ok
[05:54] <herve> approved!
[05:54] <ajmitch> it'd get too slow to do 4 or more reviews for each new package
[05:54] <dholbach> and keeping a list of those who "signed off" is good, those 4 guys have to fix the bugs then :-)
[05:54] <pitti> ogra: special exception for suid/sgid binaries
[05:54] <pitti> ogra: I'd like to be notified about them
[05:55] <ajmitch> pitti: we can pass those on to you
[05:55] <pitti> so that I can do a more insight review
[05:55] <ogra> hehe, piti, i assume one of the three will raise a hand then
[05:55] <dholbach> pitti: yeah
[05:55] <pitti> ogra: can we please formalize this?
[05:55] <ajmitch> ogra: well it only took a couple of minutes to spot the suid applet :)
[05:55] <ogra> pitti: you will get them all ;)
[05:55] <pitti> ogra: no suid/sgid without approval from security@ubuntu.com
[05:55] <pitti> okay, thanks
[05:55] <ogra> ok, all suid binary packages go to the security team first
[05:55] <dholbach> Mithrandir: about the public mirror
[05:56] <dholbach> pitti: about the lintian warnings
[05:56] <dholbach> we have no mechanisms at all for doing that
[05:56] <pitti> dholbach: I'm more concerned about certain errors
[05:56] <crimsun> ajmitch: (it's also logged concurrently by fabbione)
[05:56] <pitti> dholbach: okay, forget the publishing
[05:56] <ajmitch> crimsun: great
[05:56] <dholbach> and doing it manually is a pain in the ass
[05:56] <pitti> dholbach: but lintian finds a lot of errors, at least the packager/the reviewers should pay attention to it
[05:56] <pitti> dholbach: oh, not manually
[05:57] <ogra> pitti: i normally dont approove packages with linian errors/warnings....
[05:57] <pitti> dholbach: I thought about integrating it into the buildd process
[05:57] <Mithrandir> dholbach: public mirror and ask jblack to mirror it on the supermirror (sourcecontrol.net)
[05:57] <dholbach> pitti: we will set up docs of common "pitfalls"
[05:57] <pitti> okay
[05:57] <pitti> dholbach: debian automatically publishes the lintian errors, we can certainly set this up, too
[05:57] <dholbach> yeah... we need some architectural changes
[05:57] <dholbach> because the motu crowd will grow
[05:57] <jani> what about lintian warnings when sweeping across universe (i.e. python transition)
[05:57] <dholbach> were what? 10 now?
[05:58] <pitti> jani: in particular?
[05:58] <jani> should we stop for them or postpone for separate task
[05:58] <dholbach> but we'll 50-100 soon, i guess
[05:58] <jani> and do just the focused changes
[05:58] <ajmitch> dholbach: trying to coordinate that many will be fun
[05:58] <ogra> so are we done with NEW Packages ?
[05:58] <dholbach> ajmitch: that's why we have to make proper decisions now
[05:58] <jani> python transition or the new .desktop file missing
[05:58] <ajmitch> ogra: I think so
[05:58] <dholbach> ogra: yeah
[05:58] <crimsun> pitti: something equivalent to Debian's packages.qa would be wonderful, but that seems to be quite some infrastructure.
[05:59] <ogra> ok, next is Teams
[05:59] <pitti> jani: oh, lintian complains about old pythong versions? neat
[05:59] <ajmitch> currently there are 2 teams listed
[05:59] <Mithrandir> crimsun: we'll have that, it's called launchpad
[05:59] <ajmitch> pitti: I don't think it does
[05:59] <pitti> crimsun: no worries, that's already being worked on
[05:59] <jani> pitti, no for instance unrelated warings
[05:59] <crimsun> Mithrandir: / pitti: ah, excellent.
[05:59] <jani> pitti, for each transitioned package say 5 warnings
[05:59] <dholbach> ogra: write in CAPITAL letters
[05:59] <ogra> TEAMS NOW
[05:59] <pitti> jani: well, common sense applies :-)
[05:59] <jani> pitti, stop and solve them or go on with transitioning packages (the current goal)
[06:00] <dholbach> jani, pitti: that's something we'll discuss in "patches"
[06:00] <pitti> jani: I don't care about speling errrors, I care about broken libraries and suid files and such
[06:00] <jani> dholbach,ok :)
[06:00] <jani> pitti, ok then
[06:00] <ogra> we currently have two active teams and a bunch of proposals
[06:01] <dholbach> namely java and mono?
[06:01] <ajmitch> ogra: well the zope/plone page was just put up yesterday, we've got 3 people willing to do zope stuff for now 
[06:01] <dholbach> oh cool
[06:01] <ogra> i personally would like to see the packaging efforts shared in a team then having individual maintainers
[06:01] <dholbach> yeah
[06:01] <jani> crimsun, xfce4 team?
[06:01] <herve> no Python team or merged with Zope?
[06:01] <ogra> but we need teamleaders.... to care for them
[06:02] <crimsun> jani: sure, something of that ilk
[06:02] <ajmitch> herve: separate
[06:02] <ajmitch> see MOTUTeams
[06:02] <ogra> herve: propse one :)
[06:02] <ajmitch> noone has created the wiki page, so there's no team :)
[06:02] <dholbach> we'll have more overview over the packages (with a chance of weeding out)
[06:02] <ogra> crimsun: could you make a page below desktop teams...
[06:02] <ogra> for xfce
[06:02] <crimsun> ogra: absolutely
[06:02] <ogra> great :)
[06:03] <dholbach> but how do we get people in teams? you have any ideas?
[06:03] <ogra> as i wrote on the agenda, i think its up to the teamleaders.....
[06:03] <ajmitch> pitti: you want people in the MOTUSecurity team?
[06:03] <ogra> but we need these first
[06:03] <pitti> ajmitch: in any case, right now we only have one person caring for universe security updates
[06:03] <pitti> (Gerardo di Giacomo, Astarot)
[06:03] <ajmitch> ogra: how will team leaders be elected/appointed/volunteered?
[06:03] <ogra> btw: http://www.ubuntulinux.org/wiki/MOTUTeams
[06:04] <ajmitch> pitti: alright
[06:04] <pitti> but he's no MOTU
[06:04] <ajmitch> a few more would be good then
[06:04] <pitti> in any case
[06:04] <dholbach> ajmitch: activity, that's what i'd say
[06:04] <ogra> ajmitch: who comes first....
[06:04] <ogra> ajmitch: later the team should sort it themselves
[06:04] <pitti> I will coordinate vulnerabilities and I have tools to support me
[06:04] <pitti> but I would appreciate a MOTU to do the uploads
[06:05] <ogra> pitti: the plan was that the team is lead by a MOTU, but the members dont need to be...
[06:05] <pitti> fine for me
[06:05] <ogra> pitti: as a team will need at least one uploader
[06:05] <ajmitch> ogra: ok, so the zope team lead would currently be doko (although he's a full maintainer, not a MOTU)
[06:05] <herve> ok so I can be member of a Python team
[06:05] <ajmitch> herve: you can 
[06:05] <crimsun> herve: indeed
[06:05] <ajmitch> I'll also be in a python team, I think
[06:06] <herve> so I'll create the MotuPyton page and I myself to the members
[06:06] <ogra> i'll join pygtk :)
[06:06] <ogra> herve: go on
[06:06] <dholbach> we need ideas what those teams have on their goal list
[06:06] <ajmitch> herve: just copy the text from the other team page :)
[06:06] <dholbach> so people have a clearer meaning of why joining the team
[06:06] <ajmitch> dholbach: agreed
[06:06] <herve> ajmitch, you read my mind :-)
[06:06] <ogra> dholbach: that schould be written on every team page
[06:07] <dholbach> yeah... but we have to move :-)
[06:07] <ajmitch> herve: well I just copied the appropriate text from the mono team page :)
[06:07] <crimsun> dholbach: true. I've forwarded jani a brief (by no means conclusive) list of goals for xfce; that will be on the wiki page for the xfce team.
[06:07] <dholbach> cool
[06:08] <dholbach> what should the group separation be based on? skill? "package section"? ...?
[06:08] <dholbach> or whatever 2-3 people agree on? ;-)
[06:08] <ogra> separation ?
[06:08] <ogra> let the groups sort it themselves....
[06:08] <dholbach> ok
[06:08] <jani> focus on a specific software?
[06:09] <ajmitch> there are going to be people in more than one group anyway
[06:09] <dholbach> was just an idea to help new guys to get into it
[06:09] <ogra> lets just see that we get one MTU as lead for every group
[06:09] <crimsun> there's an implicit separation based on packages if you're referring to overarching themes - or do you mean within a team?
[06:09] <crimsun> s/themes/teams/
[06:09] <dholbach> no... not within a team
[06:09] <herve> what about a team able to help and solve on porting problems
[06:09] <herve> ?
[06:09] <ogra> the teams should have a package list ....
[06:10] <dholbach> just as proposal... i could imagine there's someone who wants to get involved and doesnt know where to start
[06:10] <ogra> so every othe team can see what they are working on
[06:10] <herve> Python team officialy created!
[06:10] <jani> well teams should not be exhaustive
[06:10] <ajmitch> herve: renaming your page, sorry :)
[06:10] <jani> just specialized groups
[06:10] <ogra> dholbach: he asks and gets pointed to the teamleader
[06:10] <dholbach> if (s)he says: "i know a bit of perl", (s)he could get into a PerlGroup *shrug*
[06:10] <herve> ajmitch, ?
[06:10] <jani> other MOTU activities might go on outside of specific teams
[06:10] <dholbach> i was just wondering if this was any good
[06:11] <ogra> dholbach, yup, so you poing him/her to the perl team
[06:11] <ajmitch> herve: hmm, I think we should probably keep the MOTU pages with the MOTU prefix consistent?
[06:11] <dholbach> yeah
[06:11] <crimsun> dholbach: yes, that sounds fine
[06:11] <crimsun> ajmitch: agreed
[06:11] <ajmitch> although the zope & mono team pages aren't..
[06:11] <jani> a METAMOTUTeam
[06:11] <jani> to coordinate where new people go :)
[06:11] <dholbach> so a division of groups based on skill and "interest"/"package section"?
[06:11] <herve> ajmitch, ok I retitle the Zope page too then
[06:12] <ogra> interest and package selection
[06:12] <dholbach> jani: MOTUTeams :-)
[06:12] <crimsun> dholbach: moreso the latter
[06:12] <ajmitch> herve: ok
[06:12] <ogra> anybdy else who wants to lead/join a tema ?
[06:12] <dholbach> i'd say perl/mono/java is a "skill" :-)
[06:12] <ogra> man...my typing is crap
[06:12] <ajmitch> ogra: I just joined python team
[06:12] <jani> ogra, we got used to it, no problem :)
[06:13] <herve> as a beginner, I prefer not giving myself too much burden
[06:13] <ajmitch> herve: we'll soon thrash you into shape ;)
[06:13] <ogra> hehe
[06:13] <dholbach> ok... what did we agree on?
[06:13] <ogra> we have teams
[06:13] <crimsun> herve: the mentoring process is a vital part; the more experienced among us are helping each other and the less experienced; the less experienced will then become more experience and are expected to assist newer, etc.
[06:13] <jani> I'd join xfce
[06:13] <ogra> led by a motu
[06:14] <herve> crimsun, sounds like a mailing-list and a plain wiki section for MOTUs...
[06:14] <ajmitch> herve: and irc, of course :)
[06:15] <herve> I can't but agree on the sharing of knowledge
[06:15] <dholbach> yeah
[06:15] <herve> ajmitch, yes but it's getting harder with more and more people
[06:15] <dholbach> and we should have sexy RSS feeds
[06:16] <dredg> damn work (specifically tomcat being on crack) has wiped out any chance of me paying attention to what's going on. where can i get logs of this for later reading?
[06:16] <crimsun> ogra: teams led by a MOTU(s) with possibly major contributions from members and non-members, correct?
[06:16] <dholbach> dredg: people.u.c/~fabbione/irclogs
[06:16] <ajmitch> crimsun: yep
[06:16] <crimsun> ajmitch: ok
[06:17] <ajmitch> herve: are you a MOTU yet, or going through the process?
[06:17] <dholbach> ok, is the MOTUTeam thing settled?
[06:17] <Mithrandir> dredg: http://err.no/tmp/foo.txt is just today's log, very up-to-date.
[06:17] <dholbach> i dont think we have to form the teams right now :-)
[06:18] <dholbach> ogra: alright?
[06:18] <crimsun> I'm clear wrt MOTUTeam
[06:18] <dredg> cheers
[06:18] <dholbach> ok
[06:19] <dholbach> the next point will be the most-eagerly discussed one: patches going back to debian
[06:19] <ajmitch> ah yes
[06:19] <ogra> sorry, got disconnected
[06:19] <dholbach> as i proposed on MOTUMeeting: we need some form of easy-to-apply patches (publically, maybe with mail/rss notification), where changes are absolutely obvious
[06:20] <ogra> my summary of the last topic:
[06:20] <pitti> ... or documented in the changelog
 teams are led by a motu
 the teams are separated by interests/package selection
 eery team needs a short description on the teams page
 and a packagelist
 team members dont need to be MOTU
[06:20] <ajmitch> dholbach: we can't expect debian packagers to change their packaging ways, either :)
[06:21] <dholbach> no we can't
[06:21] <herve> ajmitch, maintaing is enough for now but being trusted a MOTU some day would be an honor
[06:21] <dholbach> but we can urge them to, whenever we touch packages and submit out patches to    *.u.c/patches
[06:21] <crimsun> ogra: great, agreed.
[06:21] <ogra> k
[06:21] <ajmitch> they're a notoriously stubborn lot, the debian maintainers ;)
[06:21] <dholbach> but it's the best we can do... as i see it
[06:21] <ajmitch> yes
[06:22] <dholbach> any other plans? ideas?
[06:22] <ogra> you wont get them doing that i suspect
[06:22] <ajmitch> and sometimes they'll reject patches because they don't fit well 
[06:22] <Mithrandir> we really want to get our patches back into debian, since we then don't have to maintain them ourselves.
[06:22] <herve> hu... sorry I'm catching up...
[06:22] <ogra> i would like (as i wrote) just make a list of changes we made
[06:22] <dholbach> Mithrandir: what about the syncing aspect?
[06:22] <ajmitch> Mithrandir: agreed, the less work we have to do to keep packages in sync, the better
[06:22] <herve> what's the difference with patches for Ubuntu main packages from Debian?
[06:22] <ogra> its the up to the DD to pick is particular ones
[06:22] <pitti> as a general rule of thumb, debdiffs are the most universal form of patch submission
[06:23] <pitti> they will fit everytime
[06:23] <pitti> and include changelogs and such
[06:23] <Mithrandir> dholbach: MOM helps a lot, but it's still a bit of manual work.
[06:23] <dholbach> Mithrandir: ok... will have to learn about that at some stage
[06:23] <crimsun> ogra: that should be listed in the changelog - for instance, I enumerate in debian/changelog what each thing I've added to debian/patches/ does, so one can search the hoary-changes archive
[06:23] <ogra> malone should help there as well
[06:23] <dholbach> pitti: yeah... but easy to "select" patches would be rocking even more
[06:24] <pitti> dholbach: one debdiff per change
[06:24] <pitti> dholbach: unless you changed several things in one upload
[06:24] <crimsun> pitti: ah, excellent idea
[06:24] <pitti> dholbach: then of course this needs to be split
[06:24] <ogra> crimsun: i thougth about a more viewable thing...like a list/websearch etc
[06:24] <crimsun> ogra: oh ok
[06:24] <herve> crimsun, is there a policy or procedure on modifying Debian packages yet?
[06:24] <dholbach> herve: it's happening now :-)
[06:25] <ogra> herve: dont break it to much
[06:25] <herve> yes but some draft somewhere on the wiki or the wild internet
[06:25] <herve> I don't ask for the approved one :)
[06:25] <ajmitch> herve: it mainly depends on the changes needed & the package, I guess
[06:26] <ajmitch> there are policies for versioning & the like
[06:26] <pitti> yes, you should follow the versioning policy in all cases
[06:26] <Mithrandir> we also want to decide whether we should fix minor bugs or not.
[06:26] <herve> ok nothing for more advanced cases
[06:26] <pitti> otherwise this will become a mess
[06:26] <Mithrandir> like spelling errors and such
[06:26] <pitti> why not?
[06:26] <pitti> if somebody wants to?
[06:26] <ogra> sure
[06:26] <Mithrandir> pitti: more work to maintain.
[06:26] <Mithrandir> basically
[06:26] <dholbach> i guess this is something which would be better to decide we had a cooler architecture
[06:27] <pitti> Mithrandir: MOM?
[06:27] <Mithrandir> pitti: mergeomatic
[06:27] <herve> I think those can wait another revision with more important patches?
[06:27] <pitti> Mithrandir: if it's _only_ a speling mstake, then it might not be worth the trouble, agreed
[06:27] <ogra> if i pick a package to maintain. i want it to be as clean as i can get it
[06:27] <pitti> Mithrandir: but fixing it in addition is fine imho
[06:28] <Mithrandir> it's just a matter of policy -- I can live with either.
[06:28] <dholbach> ogra: there's nothing wrong with fixing stuff and with being nitpicking ,but how do we have to change the architecture to make it as easy as possible for debian and us?
[06:28] <crimsun> Mithrandir: I think minor errors are game; in that case one dpatch in debian/patches/ fixing the like would suffice, no? (easily sent back to Debian)
[06:28] <pitti> it is easiest if you don't change the architecture :-)
[06:28] <Mithrandir> (but forcing them to be _clean_ would make my work easier, since I can just continue being really stict when reviewing packages)
[06:29] <dholbach> pitti: <b>I</b> wouldnt dare it
[06:29] <ogra> dholbach: as i said before, a list/webinterface where a DD can look for changes on his/her pkg
[06:29] <ogra> and decide which to adopt
[06:29] <pitti> dholbach: in some cases I successfully bitched the Debian maintainer to do it :-)
[06:29] <ogra> so they dont feel we steal their work
[06:30] <ogra> but rather feel we helped....
[06:30] <dholbach> i think before hoary is released we should focus on getting things done, after that we need to *magically* get that architecture
[06:30] <dholbach> so we can do everything as nicely as we can
[06:31] <ogra> as i said before, i think malone is much of that tool
[06:31] <dholbach> i heard much about keeping patches in baz
[06:31] <dholbach> what do you think?
[06:31] <ajmitch> dholbach: we haven't got much time to get stuff done for hoary :)
[06:32] <Kamion> dholbach: won't really work until we have imports of all the packages
[06:32] <ogra> that would require additional baz knowledge for new MOTUs
[06:32] <Kamion> dholbach: when we do, that'll be great, but it's a dead loss until then
[06:32] <dholbach> Kamion: ok
[06:32] <ogra> which is an additional hurdle
[06:32] <Kamion> Canonical has an employee working on a next-generation baz, which should be much less painful to use
[06:32] <dholbach> ogra: right... but if it makes things <ever> most easy, that's the way to go
[06:33] <ogra> dholbach: but that would lower the rate of new MOTUs a lot i guess....
[06:33] <dholbach> ogra: depends on the tool :-)
[06:33] <jani> ogra, if done right using it will be easier than learning debian packaging 
[06:34] <dholbach> ogra: maybe there'll be some automatism
[06:34] <ogra> guys, baz isnt even fully documented 
[06:34] <jani> ogra,  true
[06:34] <ogra> how should one that has never used a versioning tool step in there easily
[06:35] <pitti> hct
[06:35] <pitti> :-)
[06:35] <ogra> i think the learning curve is already quite high if you never did packaging....
[06:35] <jani> what I'm saying is that if the tool has a nice interface it should be less of a hurdle than dpkg :)
[06:35] <jani> tla/baz is far from that :(
[06:36] <pitti> ogra: cdbs is great to make the learnign curve less steep :-)
[06:36] <herve> ogra, when I started, I inspected other packages and copied some stuff
[06:36] <ajmitch> pitti: as long as it's documented
[06:36] <herve> is cdbs documented enough? 
[06:36] <ajmitch> pitti: I've had to read through the cdbs source a number of times :)
[06:36] <pitti> well, enough to understand it
[06:36] <pitti> but not very detailled
[06:36] <pitti> ajmitch: me too :-)
[06:36] <ajmitch> and I've had great success harassing jbailey on irc ;)
[06:37] <ogra> pitti: i have cdbs as suggestion for NEW packages in my notes already ;)
[06:37] <Mithrandir> pitti: I disagree, I think cdbs is terrible and dh_make-generated stuff is a lot easier to read and understand,.
[06:37] <ajmitch> but he was also my sponsor for some debian packages
[06:37] <tritium> Perhaps canonical could set up a bounty for cdbs documentation.
[06:37] <Mithrandir> that is, it's terrible because you have to read through the source to figure out how it works.
[06:37] <pitti> Mithrandir: I've seen too many poorly written debian/rules files which weren't more than the dh_make output :-(
[06:38] <pitti> okay, but then this is a documentation problem
[06:38] <tseng> dh_make makes a pretty nasty package
[06:38] <herve> I lost the topic we're talking about :o
[06:38] <pitti> Mithrandir: and cdbs has tarball.mk :-)
[06:38] <dholbach> herve: ++
[06:38] <pitti> which rocks
[06:38] <ogra> herve,  The right way to handle upstream bugfixes
[06:38] <ajmitch> pitti: agreed, I like using that
[06:38] <ogra> but its a bit offtopic indeed
[06:38] <herve> ogra, rather sounds like "packaging made easy"
[06:38] <ogra> heh
[06:38] <jani> pitti , any link to dpkg-ng talks?
[06:38] <ajmitch> herve: it is :)
[06:39] <ogra> ok, did we have any consensus on the topic ?
[06:39] <Mithrandir> I don't think a rules file with:
[06:39] <Mithrandir>  #! /usr/bin/make -f
[06:39] <Mithrandir> include /usr/share/cdbs/1/class/makefile.mk
[06:39] <Mithrandir> include /usr/share/cdbs/1/rules/debhelper.mk
[06:39] <Mithrandir> is very understandable either.  Unless you happen to understand cdbs.
[06:39] <pitti> jani: hmm, ask Keybuk, he has a nice page about it
[06:39] <dholbach> ogra: not yet
[06:39] <jani> pitti, ok thanks
[06:40] <Mithrandir> jani: possibly on dpkg.org
[06:40] <herve> can we sum up the decisions or proposals about handling upstream bugfixes?
[06:40] <tritium> But for a new packager, it's less intimidating.
[06:40] <pitti> Mithrandir: but you have to understand debhelper, too
[06:40] <Mithrandir> pitti: it has man pages and sensible names. :)
[06:40] <ogra> i think its easier to understand the basics of packaging without cdbs....
[06:40] <pitti> (okay, off-topic, agreed)
[06:40] <dholbach> ogra: we won't sort it out today
[06:40] <Mithrandir> yeah, off-topic. :)
[06:40] <ogra> even it may be easier to maintain
[06:40] <ogra> OK
[06:41] <ajmitch> ogra: yes, but currently we don't have the equivalent of NM
[06:41] <tseng> herve: in bugzilla, or packages?
[06:41] <ogra> NEXT POINT !
[06:41] <herve> tseng, I lost it long ago
[06:41] <Mithrandir> can we get a summary first?
[06:41] <ogra> hmm, did we have any consensus?
[06:41] <ogra> i didnt see one
[06:41] <ajmitch> on bugfixes/patches?
[06:42] <ogra> yup
[06:42] <dholbach> we want to have cool web-based/mail/rss/baz-architecture
[06:42] <dholbach> i think that was all
[06:42] <Mithrandir> dholbach: at some point in the future.
[06:42] <ogra> which malone might do already
[06:42] <Mithrandir> and we want to push as many changes back to debian as humanly possible
[06:42] <ogra> (list with ubuntu changes)
[06:42] <ajmitch> but for now, use debdiff to get patches to submit to debian maintainers
[06:43] <Mithrandir> we'll do minor changes as well as larger and infrastructural stuff, but try not to rip apart the debian packages completely
[06:43] <tseng> which I think everyone should strive to do, where possible
[06:43] <ajmitch> tseng: either that or be the debian maintainer :)
[06:43] <Mithrandir> (larger and infrastructural is like the python transition)
[06:44] <tseng> hm that will be a whore to get back into debian
[06:44] <ogra> ok, noted for the summary
[06:44] <Kamion> like the conventions for people doing non-maintainer uploads in Debian
[06:44] <Kamion> it's considered polite not to totally change the packaging in an NMU
[06:44] <pitti> ++
[06:44] <Kamion> in our case, it's just silly to totally change the packaging - you're making a lot of merge work for yourself in the future
[06:45] <Kamion> unless you're sure that change will go back to Debian
[06:45] <Treenaks> How about submitting all (useful) patches from universe to the debian bts?
[06:45] <Mithrandir> Kamion: true, but we want to do stuff like correct spelling and such, which you wouldn't do in an NMU
[06:45] <ogra> ok, are we done with the topic ?
[06:45] <Kamion> Treenaks: it's a matter of filtering
[06:45] <ogra> oh
[06:45] <Kamion> Mithrandir: that's true
[06:45] <Treenaks> Kamion: true, but MOTUs need to do coordinate with the debian maintainers, imho
[06:45] <Kamion> Treenaks++
[06:46] <Kamion> but, in the event that you simply can't, you shouldn't let that stop you
[06:46] <ajmitch> morning mdz 
[06:46] <mdz> morning
[06:46] <ajmitch> ogra: yes, that's all we can do really
[06:46] <mdz> I had a reminder set to attend this meeting, but was distracted by kubuntu pressures
[06:47] <Treenaks> ogra: yes, but talking to the d-d is not a bad thing.. knowing who he/she is, knowing a bit about the packages
[06:47] <mdz> have you talked about Malone yet?
[06:47] <ogra> mdz: next 
[06:47] <ajmitch> no, that's next on the list I think
[06:47] <dholbach> mdz: briefly in hoping it'd make patch-offers-for-debian easier
[06:47] <ogra> Treenaks: sure, but what i propose is just having a list, and let them pick themselves...
[06:48] <ogra> ...and notify if iwe made changes
[06:48] <ogra> which i would expect to be possible with malone....so lets move on now :)
[06:48] <ogra> ok for everyone ?
[06:49] <ajmitch> ok
[06:49] <dholbach> ok for me
[06:49] <ogra> Bug tracking using Malone
[06:49] <bradb> yeah
[06:49] <ogra> bradb, tell us about it :)
[06:49] <bradb> so:
[06:49] <bradb> i need to get a feel for what you guys need to get rolling with malone
[06:49] <bradb> here's the goal:
[06:49] <ogra> btw, thanks for coming
[06:49] <mdz> you are now bradb's most important customer :-)
[06:49] <bradb> get universe using Malone ASAP. it's good for you; it's good for me.
[06:50] <ajmitch> it'll be great for us..
[06:50] <dholbach> hi sabdfl 
[06:50] <ajmitch> hi sabdfl 
[06:50] <sabdfl> hi all
[06:50] <ogra> yeah
[06:50] <bradb> so, we have to figure out what needs to happen to get universe on malone
[06:50] <ogra> hi sabdfl 
[06:50] <bradb> hi sabdfl 
[06:50] <bradb> so, first: how many malone users are there for universe?
[06:50] <Mithrandir> as submitters or packagers?
[06:50] <pitti> users in the sense of packagers?
[06:50] <bradb> and from there, what kind of activity could we expect on the system?
[06:50] <ogra> currently about ten, but we expect to grow a lot soon
[06:51] <dholbach> bradb: 10 MOTUs + 10 not-yet-MOTUs + developers
[06:51] <ogra> (MOTU)
[06:51] <bradb> ok
[06:51] <bradb> how many bug reports do you think we'd see per day in the system?
[06:51] <bradb> ballpark
[06:51] <bradb> like, way ballpark
[06:51] <Riddell> is malone intended to replace bugzilla.ubuntu.com?
[06:52] <ogra> yup
[06:52] <dholbach> there should be some kind of a changelog-extraction to know who worked on the package
[06:52] <ogra> but universe is the test iirc
[06:52] <Treenaks> bradb: universe only? 20-ish?
[06:52] <bradb> Riddell: for universe first, yes. it's intended to be the BTS for the MOTU.
[06:52] <dholbach> that'd make things most easy
[06:52] <bradb> Treenaks: ok
[06:52] <bradb> next: when do you really need to get started?
[06:52] <bradb> how urgent is this?
[06:53] <Mithrandir> bradb: bugs.d.o seems to have about 5k bugs/month.
[06:53] <ogra> as soon as we can
[06:53] <dholbach> bradb: i wouldnt want to be a pain in the neck, but the current system is a bit chaotic
[06:53] <ogra> bradb: its a PITA to use ubuntu-users for that
[06:53] <dholbach> :-)
[06:53] <Mithrandir> dholbach: what system? ;)
[06:53] <ajmitch> ogra: or the wiki
[06:53] <dholbach> Mithrandir: the *pointing wildly around* system
[06:53] <ogra> heh
[06:53] <Mithrandir> basically, we don't have any BTS atm.  It sucks, wildly.
[06:53] <ogra> yup
[06:53] <bradb> i ask because one of our guys told me it would take about another week to get universe packages imported into Malone.
[06:53] <Mithrandir> bradb: so three months ago would be nice.
[06:54] <tseng> a few bugs go through bugzilla anyway
[06:54] <dholbach> bradb: what will the bug assignment be based on?
[06:54] <tseng> ive done what I could with them in-place but resolved them UNIVERSE
[06:54] <bradb> Mithrandir: cool, that makes it easier for you to drink the koolaid
[06:54] <bradb> :P
[06:54] <ogra> bradb: if next week is as soon as possible, thats ok ;)
[06:54] <bradb> dholbach: distro + sourcepackagename
[06:55] <dholbach> bradb: but we dont own packages in ubuntu
[06:55] <Mithrandir> bradb: you know this, but running something as big as universe (heck, even just main) without a BTS is madness.  If next week is what we get, then next week is fine.
[06:55] <bradb> next: how do we verify that malone is suitable for your purposes, for a first go? remember, we're not trying to knock you out in round 1.
[06:55] <dholbach> bradb: so everyone who worked on a package should be possible bug fixer
[06:55] <ogra> cros bug fixer
[06:55] <bradb> dholbach: that's how it is right now. wiki-mode task editing.
[06:55] <ogra> cross even
[06:56] <bradb> (for distro side, anyway)
[06:56] <tseng> re that, do we have a policy for bugs which are "upstream"
[06:56] <ajmitch> bradb: universe maintainership is fairly loose, thankfully
[06:56] <tseng> and there isnt an obvious fix/patch
[06:56] <Mithrandir> tseng: somebody adopts the bug and talks to upstream?
[06:56] <tseng> the one im refereing to isnt even reproducable afaict
[06:56] <bradb> ogra: would it be you that would look at malone to see if it's sane for basic universe usage? mdz? sabdfl? all of the above?
[06:56] <Treenaks> tseng: fix bug, send patch to upstream. or tell upstream there's a bug, hope they fix it
[06:57] <tseng> i asked the user to work on an already open bug upstream
[06:57] <tseng> in the case im refering to
[06:57] <ogra> bradb: i don like to be responsible alone as long as i can have more eyes helping out....laets make it a small group ....
[06:57] <dholbach> bradb: does the MOTU crew get a preview? :-)
[06:58] <bradb> i just need a single point of contact between me and universe.
[06:58] <ogra> ok, that would be me or dholbach, since we do the administrativa
[06:58] <Mithrandir> I would be happy with ogra
[06:58] <ajmitch> ogra: lucky vic^Wvolunteers :)
[06:58] <bradb> dholbach: can you all access dogfood? we might need to sort out. i'd be more than happy to schedule something like that in the next day or two.
[06:58] <Mithrandir> I guess we'll whine if stuff passes them but needs fixing still
[06:59] <bradb> ogra: then it's you! :P
[06:59] <ogra> ok
[06:59] <ajmitch> great
[06:59] <dredg> sorry i've missed the past er.. while.. day job got in the way ;)
[06:59] <dredg> where are we?
[06:59] <ajmitch> dredg: malone
[06:59] <bradb> ok, so, to recap, here's what we've figured out so far:
[06:59] <dredg> (i'll catch up on the older stuff in a bit)
[07:00] <bradb> 1. the universe traffic should be easily manageble by malone
[07:00] <bradb> 2. we have stuff to do, but we can hopefully get universe imported within a week (depending on what the gina guys can do for me)
[07:00] <dredg> ajmitch: cheers
[07:00] <bradb> 3. we need a session to demo malone, and ensure it's usable for a first go, for your guys' purposes.
[07:01] <dholbach> cool
[07:01] <bradb> 4. ogra is my point of contact for universe; dholbach is my backup contact for universe.
[07:01] <tseng> can I ask a stupid question? probably already been addressed even
[07:01] <ogra> tseng: no
[07:01] <ogra> tseng: ask a wise one ;-P
[07:01] <tseng> well, im wondering what malone does over bugzilla.
[07:01] <dholbach> i didnt follow the development cycle of malone
[07:01] <Treenaks> gina sounds like one of the ftp-admin tools
[07:01] <Treenaks> dogfood is 
[07:01] <Treenaks> uh
[07:01] <bradb> dholbach: gina is the thing that does package imports. dogfood is app we use for internal usage only of the apps.
[07:02] <tseng> im sure it does something crazy cool, its just not apperant from the web interface that I can see
[07:02] <ogra> tseng: the power is under the hod ;)
[07:02] <ogra> hood even
[07:02] <dholbach> bradb: ok, nice to know you're working on it
[07:03] <dholbach> bradb: with the gina and dogfood guys :-)
[07:03] <bradb> heh
[07:03] <herve> tseng, piloting a BTS-like platform from e-mails cannot be seen from a Web interface :-)
[07:03] <ogra> if i understood correctly, malone will care for the flwback of bugfixes to debian
[07:03] <bradb> how should we schedule the demo session then? when should it be scheduled for? who absolutely must attend?
[07:03] <ajmitch> ogra: that will rock
[07:04] <dholbach> bradb: we could meet on irc or mail the concrete date and time
[07:04] <ogra> bradb: 1. i'm free all the time, just tell me a date 2. aks the others that want to attend then 3. as many MOTUs we can get
[07:05] <bradb> I'm thinking a little lead time would be useful, so maybe we should aim for Mondayish.
[07:06] <ogra> bard: ok... i'll sort it out with the others then
[07:06] <crimsun> Monday, 14 March 2005?
[07:06] <bradb> we need to ensure you guys can access dogfood too
[07:06] <ogra> oops s/bard/bradb
[07:06] <ajmitch> bradb: that would be good
[07:06] <Treenaks> sounds ok
[07:06] <dredg> ok, we know what to do... we can work out the finer points after the meeting
[07:06] <dredg> (imo)
[07:06] <ogra> crimsun: 7th ?
[07:06] <bradb> crimsun: march 7th :)
[07:06] <crimsun> bradb: ah, ok.
[07:06] <sabdfl> just reading scrollback, i didn't see an explanation for you guys about why malone is worth doing over bugzilla
[07:07] <sabdfl> should i run through that now, and bradb can correct me if i take a wrong turn?
[07:07] <tseng> please.
[07:07] <ogra> yay
[07:07] <dredg> sabdfl: bugzilla is heinous ;)
[07:07] <bradb> sabdfl: please do :)
[07:07] <sabdfl> ok
[07:07] <mdz> I think the problems with bugzilla are widely acknowledged, but sure :-)
[07:07] <sabdfl> first off, malone is brand new, minimalist, and likely to have bugs. yay
[07:08] <sabdfl> you guys are the distro guinnea pigs, but i hope you will love the vision
[07:08] <sabdfl> so that's what i'll pitch
[07:08] <pitti> will it ease propagating patches to Debian?
[07:08] <sabdfl> pitti: yes, explicitly
[07:08] <pitti> that was the main concern IIRC
[07:08] <sabdfl> malone is a BTS for the whole open source world, if they want it
[07:08] <sabdfl> you can use it for an upstream project
[07:08] <sabdfl> you can use it for ubuntu, and derivatives
[07:08] <sabdfl> and you can easily keep track of the state of a bug anywhere in that system
[07:08] <sabdfl> so
[07:09] <sabdfl> a user files a bug on a source package in universe
[07:09] <pitti> sure, I mean, does that help for the MOTUs right now? I. e. is all of this already implemented?
[07:09] <herve> I guess Malone is supposed to be user-friendly too?
[07:09] <pitti> if so, then it is certainly worth a try
[07:09] <sabdfl> someone here looks at that and says "hey, that's upstream", so they mark it as upstream
[07:09] <sabdfl> this much works already
[07:09] <herve> which bugzilla hardly is with the default interface
[07:09] <pitti> cool
[07:09] <ogra> herve, https://launchpad.ubuntu.com/malone
[07:09] <sabdfl> upstream can see their bugs, and we can see bugs on source packages
[07:09] <sabdfl> comments are shared between everyone who sees the bug
[07:09] <pitti> sabdfl: don't you think a bunch of "real" users would be good to develop Malone further?
[07:10] <dredg> sabdfl: right. the main thing here is that currently it's recommended to report bugs to ubuntu-users. that ml is already *very* high volume, and bug reporting to a mailing list is imo a bad idea
[07:10] <herve> ogra, thanks
[07:10] <sabdfl> there are some real users already - bazaar, and lp internal
[07:10] <ogra> pitti: thats what we are starting right now
[07:10] <sabdfl> dredg: once the basics are working in malone we will direct users to malone
[07:10] <sabdfl> people can mark a bug fixed in *their* part of the open source world
[07:10] <sabdfl> so at a glance you can see "ah, upstream have fixed this bug"
[07:11] <sabdfl> similarly, if we fix a bug, and it's also been flagged as upstream, then upstream will see that we fixed it
[07:11] <pitti> sabdfl: does it interact with debbugs?
[07:11] <sabdfl> pitti: steady on
[07:11] <dredg> sabdfl: ok, this sounds very promising
[07:11] <sabdfl> the idea is to encourage people to collaborate
[07:11] <sabdfl> we will sync all debbugs bugs into malone
[07:11] <sabdfl> and we will keep track of them
[07:12] <sabdfl> so, when debian fixes a bug, we see that it is fixed, and will show that in malone
[07:12] <Treenaks> cool
[07:12] <sabdfl> and we will also spec a feature to tell debian when we fix it 
[07:12] <sabdfl> automatically
[07:12] <dredg> nice
[07:12] <sabdfl> as long as the maintainer wants that
[07:12] <pitti> ^ that's what I meant
[07:12] <pitti> for pushing back patches
[07:12] <ajmitch> sounds useful
[07:12] <pitti> that would rock
[07:12] <sabdfl> yes
[07:12] <dredg> ok, currently how usable is malone?
[07:12] <sabdfl> further down the line, this will integrate with baz
[07:12] <sabdfl> so when you make a patch, it can propagate and be tracked at a code level
[07:13] <sabdfl> dredg: the basic structure is there, but there is a LOT of work to be done
[07:13] <sabdfl> there's been no heavy usage, and so lots of things will irritate you
[07:13] <dredg> fair enough
[07:13] <sabdfl> i am confident it will be pretty polished by the time we get to hoary+1
[07:13] <ogra> since we are the testers ;)
[07:14] <bradb> dredg: we've just doubled the size of the development team though (2 people! :), and indeed, Real Users (TM) is the best way i've found to do the things most useful for users as early as possible.
[07:14] <ogra> it has to be :)
[07:14] <dholbach> i'm already excited about it
[07:14] <dredg> ok. but it seems to me that having active testers doing real things would be a good thing
[07:14] <pitti> but at some point real guinea^testers are just required to see how it behaves in practice
[07:14] <pitti> dredg: :-)
[07:14] <ogra> im excited about it since mataro....but still wait.....
[07:14] <dredg> provided that it's not so young that it can kill work
[07:15] <dredg> ie there's not much point using it at all even for real-world tests if it eats the bugs/kills my cat/whatever
[07:15] <dholbach> dredg: look at the current situation, how could it kill work? :-)
[07:15] <sabdfl> dredg: it wont eat your bugs
[07:15] <bradb> dredg: it's stable. you won't lose data. there will be some things that will annoy you though, like sabdfl said.
[07:15] <dredg> dholbach: point taken :)
[07:15] <sabdfl> it might just require more clicking than you'd like
[07:15] <dredg> sabdfl: i can live with that
[07:15] <sabdfl> s/might/will/ :-)
[07:15] <ajmitch> dholbach: it leaves us stunned & amazed in awe? :)
[07:16] <ajmitch> sabdfl: any BTS will be better than what we have :)
[07:16] <sabdfl> ajmitch: stunned like a brick between the eyes?
[07:16] <dholbach> :-)
[07:16] <ogra> the current situation can only improve....
[07:16] <ajmitch> there is that..
[07:16] <sabdfl> ok, that's enough smileys to win any battle with
[07:16] <dredg> sabdfl: the idea of tracking bugs via a ml gives me The Fear
[07:16] <Treenaks> more clicking... that's not bad, as long as there's a way of reporing it :)
[07:16] <Treenaks> which is probably there :)
[07:16] <Mithrandir> email interface?
[07:16] <sabdfl> dude, luurrve debbugs - it's a mlmgr!
[07:17] <sabdfl> Mithrandir: not yet, but in the works
[07:17] <bradb> Mithrandir: notifications yes, but no reporting interface yet
[07:17] <herve> was there some discussion already done about not developing a user-friendly interface over BTS?
[07:17] <dredg> sabdfl: the idea of tracking bugs via a ml as high volume as ubuntu-users makes me die inside
[07:17] <Mithrandir> it would be _very_ useful to be able to bug manipulation through email
[07:17] <sabdfl> true
[07:17] <dholbach> ajmitch: we'll be rocking universe even harder... 
[07:17] <sabdfl> Mithrandir: you'll have that by bendy, promise
[07:17] <Mithrandir> sabdfl: yay. :)
[07:18] <ajmitch> and an emacs interface as well.. 
[07:18] <ogra> bendy ? is it official now ?
[07:18] <sabdfl> ogra: oops. no.
[07:18] <ogra> he
[07:18] <ogra> h
[07:18] <sabdfl> working title
[07:18] <pitti> ajmitch: rather a command line tool like "bts"
[07:18] <Mithrandir> pitti: both
[07:18] <pitti> "bts tag 12345 pending" is really cool
[07:18] <ajmitch> pitti: yes, but I like to be able to view the list of bugs in emacs when working on the changelog
[07:18] <herve> pitti, rather a graphical tool for end users?
[07:18] <ajmitch> that is useful, I agree
[07:18] <dredg> ajmitch: bah. emacs. no matter what the users say about it i maintain that, like sendmail, lacks a decent editor and mta
[07:19] <sabdfl> for example, here are the bugs in bazaar (the distributed rcs)
[07:19] <sabdfl> https://launchpad.ubuntu.com/malone/products/bazaar/
[07:19] <Treenaks> ajmitch: bts query 12345
[07:19] <ogra> ok, there are still three topics left 
[07:19] <ogra> are we done with malone ?
[07:19] <ajmitch> Treenaks: I know, but I like emacs still :)
[07:20] <ajmitch> ogra: three topics left & the sun is rising here
[07:20] <dredg> ok, i'm happy enough with malone
[07:20] <bradb> i think so. i'll get the demo sorted with you guys by later today. (i.e. a firm date/time)
[07:20] <Treenaks> ogra: I think we're done with malone
[07:20] <dholbach> bradb: cooool
[07:20] <dredg> yes it may require more clicking but if it benefits us long term (and everyone else) then it can't be a bad thing
[07:20] <ajmitch> bradb: thanks
[07:20] <ogra> ok, next is: questions regarding ChanServ? and the like ??
[07:21] <herve> we don't need to be so many ops
[07:21] <dredg> ogra: simple. add a bunch of trusted users (5-10) to the access list who can op when needed
[07:21] <Treenaks> ogra: nobody ops, MOTUs voice, MOTU Masters can get ops to get rid of abuse?
[07:21] <dredg> that many ops are unneeded
[07:21] <dredg> the current access policies are on crack
[07:21] <ogra> Treenaks, ++
[07:21] <dholbach> dredg: ++
[07:21] <ajmitch> dredg: agreed
[07:21] <herve> Treenaks, agree
[07:21] <Treenaks> (voice is for the "visual distinction" Mithrandir asked for)
[07:22] <dholbach> i dont think we need that
[07:22] <ogra> how is it handled in -devel currently
[07:22] <ogra> ?
[07:22] <dholbach> and i dont like it, to be honest
[07:22] <Mithrandir> not
[07:22] <dredg> Treenaks: yeah. the key here is "can get ops"
[07:22] <dredg> i'm not sure we need voice
[07:22] <ajmitch> ogra: topic isn't locked in devel
[07:22] <Treenaks> ajmitch: I'm for unlocking it in -motu
[07:22] <ogra> is it locked in -motu ?
[07:22] <dredg> motu should be a collaborative channel by MOTUs, contributers and j.random stranger
[07:22] <ajmitch> if only a couple of people can get ops, then unlock topic
[07:22] <Treenaks> ogra: mode -t
[07:22] <ogra> ah
[07:23] <sabdfl> quick question - are the MOTU the full set of people who can upload to universe? or the people who approve people to upload?
[07:23] <sabdfl> are they the MMOTU?
[07:23] <ogra> sabdfl: i confused that in the beginning
[07:23] <ogra> sabdfl: so vurrently MOTU is every uploader to universe
[07:23] <ogra> s/v/c
[07:23] <sabdfl> ok
[07:24] <sabdfl> let's leave it that way
[07:24] <dredg> sabdfl: for the most part everone in -motu is a contributer on some level
[07:24] <ogra> and we have some motu masters
[07:24] <sabdfl> and MOTUM are the people who can approve new MOTU
[07:24] <sabdfl> gotcha
[07:24] <dredg> ah, you mean MOTU as opposed to the channel...
[07:24] <ogra> for approval and administrativa
[07:25] <dholbach> i don't think we need voices, you can always look up or ask who's a MOTU
[07:25] <ogra> sabdfl: it just sounds cooler to your friends to say "i became a MOTU today" 
[07:25] <tseng> there need not be a distiction between members and non members
[07:25] <dredg> ogra: meh
[07:25] <dholbach> it's like dredg said: " everone in -motu is a contributer on some level"
[07:25] <tseng> I felt none when i worked on mono stuff for hoary
[07:25] <ogra> so we agree on unlocking the topic, no ops and a list off administartive ops ppl
[07:25] <ajmitch> ogra: agreed, we've got to get those tshirts printed for conferences  ;)
[07:25] <tseng> I was treated the same as everyone else working on ubuntu
[07:26] <ogra> ajmitch: YAY !!! +++
[07:26] <dredg> i don't see the need for separation or 'classes' of users in the channel
[07:26] <Mithrandir> ajmitch: "I am your MASTER"?
[07:26] <dholbach> yeah, dredg ++
[07:26] <ogra> Mithrandir: yeah
[07:26] <ajmitch> Mithrandir: sounds good
[07:26] <dredg> it's pointless. anyone who contributes is doing their job
[07:26] <herve> tseng, I would even think you were already a MOTU at the beginning! 
[07:26] <ogra> Mithrandir: dont punish to much ;)
[07:27] <dholbach> could we decide on the chanserv thing now? :-)
[07:27] <dholbach> please :-)
[07:27] <ogra> so we agree on unlocking the topic, no ops and a list off administartive ops ppl
[07:27] <ogra> ??
[07:27] <dredg> ogra: yes
[07:27] <herve> agree
[07:27] <ogra> ok, noted for the summary
[07:27] <tseng> same as #-devel
[07:27] <Mithrandir> ogra: ok
[07:27] <ogra> next:  what can we do that our approval process doesnt slow down as much as debians NM process if the masses start rushing in
[07:28] <dholbach> we need more reviewers
[07:28] <ajmitch> reviewers of packages?
[07:28] <ogra> do we want a policy that a new MOTU has to review a certain amount of pkgs ?
[07:28] <dholbach> the system sabdfl appointed in last tb meeting is sufficient in my eyes
[07:28] <ajmitch> dholbach: summary of that?
[07:28] <ogra> 5 packages sponsored
[07:29] <ogra> its not enough in my eyes
[07:29] <ogra> but fine in the fast approval process we're currently in
[07:29] <tseng> I would disagree with any sort of policy forcing work on volunteers
[07:29] <Mithrandir> tseng: ack
[07:29] <dredg> i don't think you can gauge it on number of packages..
[07:30] <Mithrandir> some packages are big and hard and evil, while others are small and trivial
[07:30] <sabdfl> new folks should focus on small, then grow, though
[07:30] <ogra> tseng: if its a premise 
[07:30] <dholbach> we should make sure, people communicate nicely, people minding only their own stuff, it's a matter of trusting
[07:30] <dholbach> s/, people minding only their own stuff/
[07:30] <tseng> ogra: would it make more sense to have people specifically allocated to reviewing packages?
[07:30] <dholbach> tseng: anyone who finds the time
[07:31] <ogra> tseng: sure, but we dont have them yet
[07:31] <dholbach> you won't find anyone who is "just the review guy"
[07:31] <ogra> tseng: so forcing newbies will probably help both sides
[07:31] <ogra> tseng: i learned a lot more about packaging by reviewing others packages
[07:32] <Treenaks> ogra: newbies will learn a lot from that, especially if they review together with some other people
[07:32] <ogra> yup
[07:32] <ogra> thats my point
[07:32] <dholbach> furthermore we should find a way of tracking the things people do, not a documentation overhead, but something to get an impression of what people do/work on
[07:32] <dholbach> then it'll be a lot easier to approve people
[07:32] <ajmitch> dholbach: you've done a good start on that with the wiki
[07:32] <ogra> that should be sorted in the groups stuff
[07:32] <dredg> dholbach: so what you're saying is that there should be a better way of communicating with other contributers?
[07:32] <herve> dholbach, their wiki homepage?
[07:32] <dholbach> take wasabi for example: if you saw his plan on wiki/JavaProgress, you knew he was up for the job
[07:33] <dholbach> that added to some nice packages he did should make the decision easy
[07:33] <ogra> dholbach: i think currently your list is enough...
[07:34] <ogra> if everybody cares to update regulary
[07:34] <dholbach> yeah, but we should have an agreement for every MOTU to subscribe to it :-)
[07:34] <ogra> yup
[07:34] <ogra> but it will get a mess in the future
[07:34] <dholbach> and an agreement to do reviews :-)
[07:34] <ogra> we should keep a better system in mind
[07:34] <dholbach> yeah... maybe team-based
[07:34] <ajmitch> dholbach: I think most of the current MOTUs do reviews anyway
[07:35] <ogra> since you dont want a wiki page with a list of 1000s of packages
[07:35] <tritium> personally, I'm a private person.  I don't blog, and don't keep a homepage or wiki page.
[07:35] <ajmitch> tritium: so am I, but it does make it easier to track work
[07:36] <dholbach> i think we should collect ideas on this issue and the provide-patches-for-debian issue and discuss it in the next meeting
[07:36] <ogra> great idea
[07:36] <dholbach> these 2 points are really a lot of work
[07:36] <ogra> dholbach ++
[07:36] <tritium> ajmitch, yes...
[07:36] <tseng> agreed
[07:36] <dholbach> thanks
[07:36] <dholbach> next item would be " Opportunity of a MOTU dedicated mailing-list. " :-)
[07:36] <ajmitch> ah, that's another topic, next meeting time :)
[07:36] <ogra> argl
[07:37] <dholbach> i like the idea too
[07:37] <tseng> im pondering what the usefulness is at present time
[07:37] <dholbach> since the current mailing lists really are a bit crowded
[07:37] <tseng> we get devel topics on -devel list
[07:37] <dholbach> we already have 3-4 people not showing up in IRC often
[07:37] <tseng> and we really want packages to be reviewed to go to the wiki
[07:37] <ogra> i dont like it
[07:38] <dholbach> well... atm it is only package reviewing
[07:38] <ogra> everything can be done through -devel currently....
[07:38] <herve> the idea is also to lower the traffic on -users
[07:38] <dholbach> but the organisation of teams will be another issue soon
[07:38] <tseng> package review should go to our wiki
[07:38] <dholbach> herve: yeah
[07:38] <ogra> it would fragmet the stuff
[07:38] <tseng> we can make a better document on the proceedure
[07:38] <dholbach> ogra: there's nothing wrong with separating things from each other
[07:39] <dholbach> i think
[07:39] <tseng> if it doesnt already exist in the "how to be an motu" docs
[07:39] <ogra> it is, if they arerelated to the same task
[07:39] <tseng> indeed
[07:39] <dholbach> but we'll grow
[07:39] <ogra> and MOTU is simply -devel
[07:39] <tseng> hm im not sure what "simple" part you work on :P
[07:39] <ogra> administrative things can be done through -users or on irc
[07:40] <ajmitch> I don't think we need to separate universe development from main development
[07:40] <dholbach> i think it's no problem subscribing to 2 lists
[07:40] <ogra> i dont see the need for a ML yet....probably later
[07:40] <dredg> er ok.
[07:40] <dredg> 1. irc is not convenient for everyone
[07:40] <tseng> dholbach: the problem is, everyone subscribes to 2 lists, which isnt always convenient
[07:40] <tritium> dholbach, once I return to my job, IRC will not be an option for me.
[07:40] <dredg> 2. users is very high volume. very very very
[07:40] <tseng> and those who are unaware miss out
[07:40] <herve> as for now, MOTU mails must to -devel?
[07:40] <tseng> herve: thats the question.. what are MOTU mails?
[07:41] <ogra> dredg: since users is dedicated to universe, we should read it anyway (at least i do that)
[07:41] <ogra> herve: devel mails must to devel, yes
[07:41] <dholbach> well.. ok... atm we dont seem to have the need for it... we collect pros and cons on the wiki and discuss it in two months, ok? :-)
[07:41] <dredg> ogra: right. i get 500+ mails/day for mailing lists associated with my job. i killed my sub to users cos i'd end up with thousands of unread mails
[07:41] <jani> dredg, consider reading though gmane
[07:42] <jani> so you don;'t get you inbox flooded
[07:42] <ogra> in my definition motu mails are devel mails or administartive motu stuff...
[07:42] <Mithrandir> jani: the problem is the volume.
[07:42] <Mithrandir> jani: and gmane doesn't solve that
[07:42] <dredg> jani: oh, and i hate web interfaces :)
[07:42] <tseng> I admit that ive also unsubbed from -users
[07:42] <jani> mithrandir: you only pick the subjects you care about ;)
[07:42] <tseng> it wasnt useful for me to have it pile up unread
[07:42] <dredg> jani: especially for something that is not web based
[07:43] <jani> gmane through thunderbird ;)
[07:43] <ajmitch> dredg: use nntp then
[07:43] <jani> of course not the web
[07:43] <jani> nntp that's what I implied but not said actually
[07:43] <herve_> I do read lists over gmane.org
[07:43] <ajmitch> that's how I read it
[07:44] <ogra> for me -users is a big pool of things that are needed....there are te suggestions what is missing, and we are the guys to solve the missing bit, since we are more free then the main crew
[07:44] <jani> since I switched to nntp I can follow a lot more lists
[07:44] <herve_> it's far better than regular mail or Web forums (*sigh*)
[07:44] <dholbach> ok... shall we re-discuss it in 2 months?
[07:44] <Mithrandir> jani: my news reader and mail reader is the same, so it doesn't matter.  -users is too high-volume for a lot of MOTU to follow it.
[07:45] <herve> dholbach, in 2 month, Malone will have changed things, moreover
[07:45] <dholbach> ok
[07:45] <ogra> herve: and currently we are about 10 motus...so i still see no need for a own ML
[07:46] <ogra> since amloe will change a lot
[07:46] <dholbach> ok
[07:46] <ogra> malone even
[07:46] <ogra> lets reschedule the task for times we are 50 MOTUs or more 
[07:46] <herve> ogra, it's also for all packagers, not just the reviewers
[07:47] <dholbach> ok
[07:47] <dholbach> conclusion is: there are an awful LOT of things to decide now, because we will grow... i will set up wikipages soon to collect ideas and to re-discuss things. i'd appreciate it, if you'd add your ideas, too.
[07:47] <herve> what topic are left? (I must go soon)
[07:47] <ogra> none, 
[07:47] <ajmitch> herve: that was the last on the page
[07:47] <ogra> ...
[07:47] <herve> good!
[07:47] <herve> :-)
[07:47] <ajmitch> I think we just need to argue about the next meeting time :)
[07:47] <ogra> so about the next meeting
[07:48] <herve> I can't leave you then
[07:48] <ogra> as i asked in the beginning, how often should we hold this
[07:48] <herve> I'm not very productive with my cough anyway
[07:48] <ogra> monthly ? 
[07:48] <tseng> monthly.
[07:48] <dholbach> yeah
[07:48] <jani> ok
[07:48] <ogra> next time an hour later ?
[07:48] <dholbach> nice :-)
[07:48] <ogra> (for Treenaks )
[07:48] <ajmitch> monthly
[07:48] <dholbach> and poor ajmitch 
[07:48] <Treenaks> ogra: uh that's OK
[07:49] <ajmitch> heh
[07:49] <ogra> again on Thursday ?
[07:49] <ajmitch> ok
[07:49] <ogra> ok, next meeting will be on march 31 2005 17:30 utc
[07:50] <ogra> four weeks from now...
[07:50] <dredg> cool
[07:50] <dholbach> very nice
[07:50] <ajmitch> just before hoary release?
[07:50] <Treenaks> ajmitch: why not :)
[07:50] <ogra> i'll write a summary and send it to -devel and -users
[07:50] <dholbach> thanks ogra, thank you very much
[07:50] <dredg> OT: does anyone know of who is in the list of 'strongly signed' keys?
[07:50] <ogra> expect it in half an hour
[07:50] <ajmitch> Treenaks: as long as we can have a release party as well ;)
[07:51] <ogra> dredg: everyone you find on the keyservers
[07:51] <dredg> i need to arrange to get my key signed, even if i have to fax my passport and drivers licence to another country
[07:51] <ajmitch> dredg: most people with a GPG key who are involved in debian or ubuntu :)
[07:51] <ogra> dredg: (signed that is)
[07:51] <ajmitch> eg any DD
[07:51] <ajmitch> nearly any
[07:51] <ogra> dredg, http://keyserver.mine.nu/
[07:51] <dredg> ajmitch: right. i note the lack of any strongly signed keyholders in ireland :)
[07:51] <dredg> though hmmm
[07:51] <Treenaks> ogra: not subkeys.pgp.net?
[07:52] <ogra> Treenaks: works too :)
[07:52] <ajmitch> dredg: I see 5 DDs in ireland
[07:52] <dredg> ajmitch: eh?
[07:52] <ogra> so thank you for everybody attending here...meeting is closed
[07:52] <ajmitch> dredg: see db.debian.org
[07:52] <ajmitch> ogra: thanks
[07:53] <dholbach> thanks ogra, again :-)
[07:53] <tritium> thanks :)
[07:55] <Kamion> dredg: the strong set's pretty trivial to get into
[07:55] <Treenaks> yeah, ask #ubuntu-nl :P
[07:56] <jani> kamion, not in romania apparently ;(
[07:58] <jani> by all