[00:50] daja77 (n=daja77@dslb-088-072-033-224.pools.arcor-ip.net) joined #rocklinux.
[00:50] _BoS_ (n=BoS@dslb-088-072-037-239.pools.arcor-ip.net) left irc: Remote closed the connection
[00:50] BoS (n=BoS@dslb-088-072-037-037.pools.arcor-ip.net) joined #rocklinux.
[01:04] daja77_ (n=daja77@dslb-088-072-034-102.pools.arcor-ip.net) left irc: Read error: 110 (Connection timed out)
[01:13] kasc_ (n=kasc@dslb-084-060-098-193.pools.arcor-ip.net) joined #rocklinux.
[01:20] kasc (n=kasc@dslb-084-060-106-063.pools.arcor-ip.net) left irc: Read error: 110 (Connection timed out)
[01:20] Nick change: kasc_ -> kasc
[02:25] blindcod1r (n=blindcod@tor/session/x-1c31b14c462112fb) joined #rocklinux.
[02:25] blindcoder (i=id@tor/session/x-781a46fb456194a3) left irc: Nick collision from services.
[02:26] Nick change: blindcod1r -> blindcoder
[03:55] daja77_ (n=daja77@dslb-088-072-033-224.pools.arcor-ip.net) joined #rocklinux.
[04:02] daja77 (n=daja77@dslb-088-072-033-224.pools.arcor-ip.net) left irc: Read error: 110 (Connection timed out)
[04:33] netrunner (n=andreas@anvame.net) left irc: Remote closed the connection
[04:33] netrunner (n=andreas@anvame.net) joined #rocklinux.
[05:20] madtux (i=miguel@pf0.hostarica.com) left irc: "Leaving"
[07:04] <blindcoder> moin
[07:22] <annesh> moin blindcoder 
[07:22] <blindcoder> moin annesh 
[07:22] <blindcoder> everything fine for todaY?
[07:23] <annesh> i hope so...
[07:23] <annesh> i'll see at 10 o'clock
[07:23] <annesh> hmpf
[07:25] <blindcoder> good luck!
[07:25] <annesh> thanks :)
[07:29] <blindcoder> anyway, I'm off to work now
[07:29] <blindcoder> brb
[08:12] <blindcoder> he
[08:12] <blindcoder> re
[09:05] andyrtr__ (n=andyrtr@p54B6B48A.dip0.t-ipconnect.de) joined #rocklinux.
[09:21] andyrtr_ (n=andyrtr@p54B6B012.dip0.t-ipconnect.de) left irc: Read error: 110 (Connection timed out)
[09:26] ija_ (n=ija@ joined #rocklinux.
[09:39] ija (n=ija@ left irc: Read error: 110 (Connection timed out)
[09:46] Nick change: ija_ -> ija
[10:42] <annesh> re
[10:53] owl (n=owl@ joined #rocklinux.
[10:56] <owl> moin
[11:40] [raphael] (n=raphael@raphael.netpark.at) left irc: "using sirc version 2.211+KSIRC/1.3.12"
[12:10] <blindcoder> clifford: smng_server?
[12:11] <clifford> as discussed at CLT.
[12:12] <blindcoder> must have missed it mentally
[12:12] <th> oh clifford!
[12:12] <th> you live!
[12:13] <clifford> th: no. not really.
[12:13] <clifford> I just checked in the stuff I had on my laptop..
[12:13] <th> clifford: you got mail
[12:13] <th> clifford: at least 3 from my side
[12:13] <clifford> yep. the sm/ssl stuff.
[12:13] <th> yea
[12:13] <clifford> I have no idea..  
[12:14] <th> clifford: i bet that's something with our subversion build
[12:14] <blindcoder> well, when I should pay attention, please tell me :)
[12:14] <th> blindcoder: i tried to get all the funky sm stuff workin
[12:15] <blindcoder> leather, lace and whips?
[12:15] <blindcoder> note: I never used the sm command line tool
[12:17] <th> blindcoder: me neither.
[12:17] <th> blindcoder: but i tried to get it working
[12:17] <th> blindcoder: but the perl-svn bindings are broken
[12:17] <th> (i have submaster+swig+perl in my build)
[12:18] <blindcoder> I see
[12:18] <th> Can't load '/usr/lib/perl5/site_perl/5.8.7/i686-unknown-linux-gnu/auto/SVN/_Core/_Core.so' for module SVN::_Core:
[12:18] <th> +/usr/lib/libsvn_ra_dav-1.so.0: undefined symbol: SSL_load_error_strings at
[12:18] <th> that's the error
[12:19] <blindcoder> oh
[12:20] <th> clifford: perhaps you want to add a flag to patches "accepted"
[12:20] <th> clifford: of course that would be the result of the approvals. but approval patterns might change
[12:21] <th> clifford: but acceptance should be persistent
[12:21] <clifford> th: full ack.
[12:21] Action: blindcoder trying a fix for desktop files in forked packages
[12:22] <clifford> accepted == applied?
[12:22] <th> clifford: accepted is the requirement for being applied
[12:22] <th> clifford: so the command-line tool we spoke of (e.g. cron job) would try to apply accepted patches
[12:22] <clifford> thats the 'approved' thing.
[12:23] <th> clifford: no
[12:23] <clifford> but the open/accepted/rejected/discarded thing is missing.
[12:23] <th> clifford: multiple apprivals might be necessary for acceptance
[12:23] <clifford> no that's the votes.
[12:23] <th> oh
[12:23] <th> wait
[12:23] <clifford> there can only be one approved by one person.
[12:24] <clifford> the idea is that one person approves the patch and then it is applied automatically.
[12:24] <th> nono
[12:24] <clifford> nono?
[12:24] <th> there are "votes"
[12:24] <th> anyone can vote
[12:24] <clifford> yes. like now.
[12:24] <th> now patterns define which votes are approvals
[12:24] <th> and how many approvals are necessary
[12:24] <th> and if all approvals are given. it is accepted.
[12:24] <clifford> ah! the patterns table is missing!
[12:25] <th> that was the terminology i states in my mail
[12:25] <clifford> mail?
[12:25] <th> Date: Mon, 6 Mar 2006 15:15:50 +0100
[12:25] <th> From: Tobias Hintze <th@hbs-solutions.de>
[12:25] <th> To: fake@rapidnetworks.de, clifford@clifford.at
[12:25] <th> Subject: new-submaster-interface.txt
[12:25] <clifford> when?
[12:25] <th> 12:27:45 < th> Date: Mon, 6 Mar 2006 15:15:50 +0100
[12:26] <th> clifford: the distinction between "vote" and "approval" seems important to me. anyone can vote. but patterns define whoms vote is an approval
[12:26] <th> clifford: an negative-approval would be a strong veto
[12:26] <th> clifford: a negative vote is just a negative-vote
[12:27] <clifford> ok. that's exactly how it is now and how I planed it.
[12:28] <clifford> just the pattern thable is missing in the schema file right now.
[12:28] <th> and a flag for acceptance.
[12:28] <th> and....
[12:28] <th> .   user_id_approver INTEGER NOT NULL
[12:28] <th> a patch can have multiple approvers?
[12:28] <clifford> yes. the state (open, rejected, applied, discarded).
[12:29] <th> ok
[12:29] <clifford> that's just mixing up terminologies because I didn't know your mail when I wrote that.
[12:29] <th> ok
[12:29] <th> very well
[12:29] <th> just another thing regarding this.
[12:29] <th> will patches be able to do svn meta ops?
[12:29] <clifford> the database just needs to know about votes.
[12:29] <th> svn rm
[12:30] <clifford> that some votes are stronger than others is implied by the yet missing pattern table.
[12:30] <clifford> on a file: yes.
[12:30] <clifford> on a directory: no.
[12:30] <clifford> (as it is now.
[12:30] <th> ok - well, i could not get sm cmd-line util to work yet.
[12:31] <clifford> the idea of the approved_by thing is that at the end one person still needs to push the "do it" button.
[12:31] <th> that would always be the last person who is approving
[12:31] <clifford> .. that can be done automatically for the extra packages, but not for stuff that changes core.
[12:32] <th> ahhh
[12:32] <th> yea
[12:32] <th> i'm mixing up votes again.
[12:32] <th> we need more than votes for approval
[12:32] <th> cause votes happen before in-depth testing
[12:34] <th> so voting is only like commenting and flagging for selection
[12:35] <th> but we additionally need the approval-thing 
[12:35] <th> so my mail is not 100% correct.
[12:38] <th> clifford: hmmm *idea*
[12:38] <th> clifford: perhaps we could change vote_type ENUM('+', '-')
[12:38] <th> clifford: into something with more than two options
[12:39] <th> clifford: ENUM('+', '-', 'A', 'V')
[12:39] <th> clifford: adding approval and veto
[12:41] <clifford> and why is this better?
[12:41] <clifford> afais only one approver is needed and I would prever the approving beeing different from voting to avoid usage errors.
[12:43] <th> oh ok
[12:43] <th> i'm fine with one approver
[12:43] <th> i think i got that idea from you
[12:43] <th> (the number of needed votes for each pattern)
[12:45] <th> clifford: the patterns define who may give the approval, right?
[12:45] <clifford> no. the patterns define which votes are recommended to be there before a patch is approved.
[12:45] <th> ah ok
[12:46] <th> so positive-vote-outcome is a requirement for approval. something like that
[12:46] <clifford> patches to the non-core stuff (and no config, etc) are approved automatically when all recommended votes are there.
[12:46] <th> ahh yes. good.
[12:46] <clifford> its a recommendation, not a requirement.
[12:46] <th> ok
[12:47] <clifford> e.g. we also may approve a patch to a package owned by XY without waiting for XY to vote for it.
[12:47] <th> i see
[12:47] <th> well in that case i thought WE would vote for it
[12:47] <clifford> [schema updated in svn]
[12:48] <th> but having it as a recommandation only is fine with me too
[12:48] <th> O D R A
[12:48] <blindcoder> what if XY voted against the patch?
[12:49] <blindcoder> sorry for interrupting...
[12:49] <daja77_> re
[12:49] <blindcoder> moin daja77_ 
[12:49] <th> blindcoder: than i see no reason for applying it.
[12:50] <clifford> blindcoder: then we still can approve it. however: maybe we wouldn't make friends that way..
[12:50] <th> blindcoder: but i would see reason for arguing on the ML
[12:50] <blindcoder> okay, that's what I wanted to know :)
[12:50] <th> clifford: should Qs have stati?
[12:50] <th> clifford: like "highly experimental", "open/closed"
[12:52] <th> clifford: and what's the votepattern_user?
[12:52] <th> clifford: ahhh that's a n:m
[12:53] <blindcoder> on a totally unrelated topic: I'm uncertain what to do with the ksimus and licq packages
[12:53] <th> clifford: forget the last question!
[12:53] <blindcoder> fork vs. split
[12:53] <blindcoder> atm. I make a compiletime decision to add the plugins to the package licq
[12:54] <blindcoder> now: is it better to have licq licq:console licq:qt licq:kde or licq licq-console licq-qt licq-kde?
[12:55] <th> for the user it's only a matter of ":" or "-".
[12:55] <blindcoder> no
[12:55] <blindcoder> rocket emerge licq vs rocket emerge licq{,-console,-qt,-kde}
[12:56] <th> blindcoder: where would you put common files in the case of forking?
[12:56] <blindcoder> licq
[12:56] <th> ah - i can't read
[12:57] <th> rocket emerge installs (by default) all splits?
[12:57] <blindcoder> yes
[12:57] <th> i see highly redundancy here.
[12:57] <th> the main difference is in package creation only
[12:57] <blindcoder> how so?
[12:58] <blindcoder> the splits/forks would only contain the plugin files
[12:58] <blindcoder> yes
[12:58] <blindcoder> for most/all practical purposes, it doesn't matter
[12:58] <blindcoder> but I think we should have a policy/best practice in this case
[12:58] <th> so we have a forks and splits. but for the users. it's almost the same
[12:58] <daja77_> i'd prefer split packages
[12:58] <blindcoder> so one can expect the same behavior from all packages
[12:59] <th> same here.
[12:59] <blindcoder> okay, splits it is
[12:59] <daja77_> forked packages would increase build time
[12:59] <blindcoder> I'll add information for this to the wiki
[12:59] <th> clifford: ok - so we can state the status in {queue,journal}_text
[13:00] <clifford> yes.
[13:00] <th> that's ok
[13:00] <th> but who's able to change the text?
[13:01] <clifford> journal_text will be the entire journal, the journal_patch table is automatically updated whenever journal_text changes.
[13:01] <th> who's able to put a patch into a Q?
[13:01] <clifford> hmm..
[13:01] <th> clifford: and journal_text may contain comments and all?
[13:01] <clifford> journal is easy: the owner (and all admins)
[13:01] <th> yea journal is easy
[13:01] <th> but who put's patches into a q...
[13:02] <th> patch creator perhaps
[13:02] <clifford> I'm not sure.
[13:02] <clifford> patch creator and admins?
[13:02] <th> if that is enough...
[13:02] <th> the more important question...
[13:02] <th> if we only have q_text to state the status...
[13:03] <th> then a q wont get closed
[13:03] <th> so new patches could arrive
[13:03] <th> hmmm
[13:03] <th> would that hurt? perhaps not.
[13:03] <th> blindcoder: btw - thanks a lot for polishing the wiki/doc
[13:04] <blindcoder> th: that was stf, not me
[13:04] <th> oups
[13:04] <blindcoder> :)
[13:04] <th> thanks to stf then.
[13:06] <clifford> th: open www.rocklinux.org in konqeror and have a look at the nav bar.
[13:06] <th> apropos
[13:06] <th> clifford: the logo was NOT ok in konquerer before we did the change
[13:07] <clifford> then I remember it incorrectly.
[13:07] <th> clifford: the nav links miss aline on top?
[13:08] <clifford> the entire nav thing is appended at the end of the page after the content.
[13:08] <th> i can't confirm that
[13:09] <th> http://thzn.de/shot.png
[13:09] <daja77_> navbar is empty in konqueror here
[13:09] <daja77_> ah right it is after the content 
[13:09] <clifford> http://www.clifford.at/priv/snapshot2.png
[13:10] <daja77_> different versions of konqueror i guess
[13:10] <daja77_> <- 3.4.2
[13:11] <clifford> 3.5.0 here
[13:12] <th> 3.2.3 here ;)
[13:12] <daja77_> guess that somehow
[13:12] <th> so i've another reason for NOT updating ;)
[13:16] <th> clifford: playing with the margin values should fix this i guess. i dont know why this happens.
[13:16] <th> sucks
[13:16] <th> clifford: perhaps anyone should be able to create Qs
[13:17] <th> clifford: we could restrict it later.
[13:17] <clifford> sure.
[13:17] <th> clifford: if we get to many Qs
[13:17] <clifford> we can remove Qs which aren't needed anymore.
[13:17] <clifford> (maybe Qs should only be marked as removed?)
[13:17] <th> if all patches in a Q are applied... it could remove itself
[13:18] <th> yea
[13:18] <th> we should delete NOTHING
[13:18] <th> well spam perhaps
[13:18] <clifford> .. and the grouping information is lost.
[13:18] <th> if i say "remove" i mean "set state to removed" or something
[13:18] <clifford> yes. so we need some kind of 'removed' flags in the database.
[13:20] <th> yea
[13:20] <clifford> I've now added queue_discarded and journal_discarded.
[13:20] <clifford> And queue_discarded must only be set if no open patches are in the Q
[13:21] <clifford> s/must only/may only/  ;-)
[13:58] <th> hmmmm /usr/lib/libsvn_ra_dav-1.so.0  is not linked against libssl
[14:03] <blindcoder> hmm
[14:03] <blindcoder> I wonder why I added --disable-gpgme to licq
[14:06] <daja77_> ./splrun: error while loading shared libraries: libpcre.so.0
[14:06] <daja77_> interesting
[15:16] <blindcoder>         splitdesc_rms() { desc_I="plugin for rms"; }
[15:16] <blindcoder>         splitreg 85 rms 'licq_rms.so'
[15:16] <blindcoder> what is missing?
[21:15] <th> this channel is idle for almost 6 hours now.
[21:16] <annesh> re
[00:00] --- Thu Mar  9 2006