IRC Logs for #crux-devel Monday, 2006-10-09

*** jaeger has quit IRC00:05
*** cptn has joined #crux-devel01:03
cptnhey there01:03
aonhi cptn01:08
*** jaeger has joined #crux-devel08:29
*** j^2 has joined #crux-devel08:47
j^2morning all08:48
*** garbeam has joined #crux-devel08:50
aonhi j^208:50
j^2how was everyones weekend?08:50
aonhousewarming party fri 19:00-> sat 13:00, pretty normal otherwise :)08:51
garbeamhi there, anyone knows if there have been IRC logs of the last friday meeting? I was unable to attend08:51
jaegerNot bad, just didn't get as much done as I wanted to08:51
aongarbeam: it wasn't held at all since some of us weren't able to make it08:51
jaegergarbeam: the meeting didn't happen, we had too many time conflicts08:51
j^2jaeger: that's my weekend too...08:51
garbeamah good then ;)08:51
j^2aon: was it a new place for you? ;)08:52
aoni didn't move :)08:53
aonbut also hadn't visited the place before :)08:53
j^2where did you move to? how's to working for you now?08:54
aoni didn't move at all08:55
aoni sort of misunderstood the question08:55
j^2ah no worries08:55
aona friend of mine moved to a smaller and more expensive apartment :)08:56
j^2so with crux-devel growing, is it true that crux is in a schism?08:56
garbeamok I will have an open eye for the next attempt! cu guys!08:56
jaegertake care08:56
aongarbeam: today 20:00 CEST iirc08:56
j^2aon: is it closer to school/work?08:56
*** garbeam has quit IRC08:56
aonj^2: nah08:56
j^2lol :p08:56
aonit's in a better condition, though08:56
j^2ah nice08:58
Romster20:00 CEST is 4am my time +10GMT iirc, UK time?10:30
aon2000CEST = UTC+210:32
aon17:32 now10:32
aon(dunno why i put the 2000 there)10:32
Romsterlol heh10:33
Romster1:33am here.10:33
Romstertuesday though when your still in monday.10:34
Romsterwell soon to be in monday10:34
Romsterif not already10:34
Romsterwhat happend to using swatch time :/10:34
aonthink pretty much everyone is in monday now10:34
Romsterwell some might be still on sunday still10:35
deus_exI wish.10:35
aonif gmt = 13:32, and timezones go as far as -1210:35
cptnthat would mean there's more that 24h on the globe?10:35
aonheh :)10:36
deus_exOr Wednesday, when BSG is on ;)10:36
j^2deus_ex: first of Season 3 for you?10:38
Romsteryeah -12 is the max amount afaik10:40
deus_exj^2: Season 110:40
deus_exI'm hooked :)10:40
j^2nice!, careful it gets better10:41
deus_exI remember original series.This is better.10:41
j^2season 3 started friday here in the US10:41
aonwhat's bsg?10:41
deus_exaon: BattleStar Galactica10:42
j^2aon: it's crack for nerds ;)10:42
Romstersort of simaler to farscape, and strar trek mixed into one?10:45
deus_exRomster: Farscape is in the league of its own, imho10:46
j^2kinda, it's...a soap but a good soap10:46
Romsteri'll have to watch some and see if i like it, i've heared of batlestar... but not seen any of it yet.10:48
*** prologic has joined #crux-devel11:27
*** treach has joined #crux-devel11:56
*** deus_ex has quit IRC12:12
*** jue has joined #crux-devel12:33
*** deus_ex has joined #crux-devel12:33
Viper_cptn: in which channel does this meeting take place?12:51
tilmansip didn't chicken, did he?12:51
tilmanViper_: this one ;)12:51
Viper_ok good :)12:51
Viper_i hope it doesn't take a long time.. i still have to do some boring stuff for school12:57
tilmandidn't your mom teach you to do that stuff right when you get home from school?12:57
*** sip has joined #crux-devel12:57
Romsterhi sip12:59
Romsterok before the meeting starts, sorry I have not attended any previously, but I'll try to attend from now on. (didn't know of this channel till recently from the ML)13:00
tilmanevening sip13:03
jaegerbad timing, I've got to go. will be back in about an hour. sorry :(13:03
siplater jaeger13:04
Romsterhmmz 4:04am, hy does 404 curse me...13:04
Romsterlater jaeger13:04
Viper_tilman: haha yes she often tells me that :)13:09
sipOk, let's start.13:15
sipso, I think we can start with unmaintained ports, since we have to get rid of them this week13:17
cptnor rather want to :-)13:18
tilmanwhy is that?13:18
Viper_it's a good idea; i proposed yesterday as the date to remove them13:18
Viper_are there any important orphaned ports left?13:18
sipI think Oct, 15 was the semi-official deadline13:18
cptnI'd like to orphan docboox-*13:18
cptnerr, docbook13:19
Viper_sip: ah ok, didn't know that13:19
cptnI'm just not using it often enough13:19
cptnanyone interested?13:19
sipcptn, but, it's in your CV!13:19
tilmanno thanks13:19
tilmani don't think we have any important ports that depend on docbook either13:20
sipI suspect gnome has something to to with it13:20
tilmannot according to prt-get dependent13:20
cptnscrollkeeper depends at least on the dtd13:21
cptnI guess I can keep them for the time being13:21
cptnbut I don't follow upstream anymore13:21
Viper_i will create an up-to-date list of orphaned ports now.. so we can look through it13:21
cptnViper_: good13:22
sipMy take: we should avoid half-maintaining stuff from the time being13:22
cptnsip: I agree in principle13:22
cptnwe have a couple of those ports in opt too13:22
cptnsome even from Per13:22
sipof course there will be some stuff we cannot avoid13:23
Viper_ pine wmclock13:23
Viper_all ports which are still maintained from steeves, heino or walpole13:24
cptnI can pick up mercurial13:24
tilmani vote for removing the wmaker ports cptn just mentioned13:24
sipthanks Viper_13:24
tilmanunless someone will pick them up13:24
cptnalthough I don't use it that often13:24
cptnbut I follow the ML13:24
sipbut I use wmclock13:24
Viper_i suggest that we move all those video related stuff to attic (like ffmpeg)13:25
Viper_nobody seems to need it13:25
cptnViper_: maybe we can submit the list to crux-contrib before moving them13:25
cptnso contrib maintainers can pick up what they want13:25
Viper_cptn: yes of course13:25
sipor after13:25
juewe should find someone for: cdrtools, cvs, dialog, libungif13:25
Romsterboost I use that...13:25
Romsterand ffmpeg but that dosn't compile anymore now..13:26
* tilman took flac13:27
Viper_Romster: there are also many orphaned dependencies of ffmpeg13:27
Romsteri notice alot there i use..13:27
Romsteri can take up what i use, out of that list.13:27
sipI'll take zsnes (hey, everybody needs some fun)13:27
tilmanjue: could you take mesa3d until x11r7 is replacing opt/x11?13:28
Romsterbut i'm not in contrib atm since i've moved my repo, see ML13:28
cptnsip: znes _and_ wmclock?13:28
Viper_i take cdrtools.. use it a lot13:28
cptnaon: what about cdrkit13:28
sipI think wmclock is already ok13:28
cptncould it replace cdrtools already?13:28
juetilman: ok13:28
aonwell, it works fine for me13:28
aonmaybe i should sync it into contrib13:28
aonso it might get a bit more testing13:29
Viper_cdrtools is also a dependency of dvd+rw-tools13:29
cptnyeah :-)13:29
Viper_which is quite nice13:29
cptnViper_: cdrkit is a fork of cdrtools13:29
aonand they've renamed some binaries13:29
Viper_ah ok13:29
cptnso I guess it would work with dvd+tw-tools too13:29
Viper_then it's ok13:29
aonlike cdrecord -> wodim13:29
cptnbut we'd need more testing13:29
aonso it probably will break something13:29
Viper_until we replace cdrtools i will keep it up-to-date :)13:30
sipI'd live a it for CRUX 2.313:30
cptnViper_: good :-)13:30
tilmanjaeger: do you mind adopting libart_lgpl? it's pretty gtk/gnome'ish ;)13:30
Romsterfaac, faad, ffmpeg, flac, boost, dosbox, fuseiso, id3lib, ketchup, kqemu, sdl*, xine, zsnes (into emulation)13:30
Romsterif no one else wants them i'll take them up and most of them libs13:31
tilmani just said i'd taken flac...13:31
Romsterok no probs. i was look though the list, i'll mark that one off13:31
tilmanRomster: maybe you should wait until the dust settles ;)13:31
Romsteryeah ok, good idea lol13:32
tilmanjue: mmh. i can probably take libungif (or giflib, considering the patent ran out...)13:32
tilmanthough i don't think gif support is that important to have in opt at all cost(?)13:32
aoni'll take netcat if nobody wants it13:32
sipremaining stuff goes to jaeger :)13:33
cptnmaybe we should also check reverse dependencies before moving anything to attic?13:33
cptnsip: heh13:33
Viper_I've also added Per's ports in opt13:35
tilmanre. opt/cvs13:35
tilmanshould we roll a dice?13:35
aoni've been running it with disable-server for some time now13:35
aonand it doesn't seem to break it13:35
tilmanjust wanted to mention that13:35
tilmanwe could make it client-only13:36
tilmanto make the port easier to maintain13:36
tilman(no need to consider server security holes etc)13:36
juegood idea13:36
Viper_would be a good idea.. for servers there is still svn13:36
Romsterimo svn beats cvs.13:37
aonbut that's not the point13:38
tilmanthat's not the point13:38
aonthere's still a lot of stuff in cvs repositories13:38
aonlike mostly everything on sf/nongnu13:38
tilmanwhat about kicking out pine?13:38
sipok for me13:38
tilmanit was suggested to replace it with mutt on the iso, i'm not sure whether there have been objections to his the last time we discussed it13:39
Romsterok cvs client only for the fact of alot of cvs repos are out there.13:39
tilmanbut i think that's a good idea13:39
aonyes, pine is icky13:39
j^2i'm inclined to agree with aon13:39
tilmani hear torvalds uses pine ;)13:40
cptnpine brings along pico though13:40
aontilman: but he doesn't use crux13:40
cptnso we might bring mutt+nano13:40
sipI think pine was useful as a backup editor (pico) for vi-impaired people13:40
aoncptn: good idea13:40
tilmancptn: ah, right (nano)13:40
Viper_cptn: why not mutt+vim?13:40
cptnViper_: :-)13:40
cptnViper_: pico was the editor for the vim haters13:40
aonor why not emacs + emacs!113:40
Viper_hmm.. :)13:41
tilmani _might_ adopt freeglut then13:41
cptn <-- dependent counts13:41
sipI'll take sdl stuff13:41
j^2sip: i agree about the nano bit, i know one of my buddies loved crux, and hated vim so he remastered it with nano13:41
j^2or something along those lines13:42
tilmani'll move pine to attic then unless jaeger objects?13:43
cptntilman: sounds good13:43
cptnmaybe we can create a summary page?13:43
cptnwhich those not attending can then review13:43
tilmanirc log summary?13:43
sipheh. btw, if you're adopting some port, please chenge the maintainer field now so we can have a real-time list13:43
tilmancptn: i'll do that13:43
siphmm, what about jdolan ports?13:44
cptnnot sure13:45
tilmanwe should ask him when he's gonna have more time on his hands to spend on crux, and then decide whether it's better to temporarily remove them?13:45
sipSounds good. I had no response even on the java ports though (with attached patches) :/13:46
*** bd2 has joined #crux-devel13:47
Viper_yeah a real-time list of would be a really nice thing13:47
RomsterI'll hold off and take up what i use if its left over/out, after the dust has setteled from this.13:48
Viper_good idea13:48
sipRomster, yes, I think it's better this way.13:48
sipIt'll be easier to simply have a look at the attic changelog13:49
*** pitillo has joined #crux-devel13:50
Romsteryeah and then fetch out of the attic.13:50
treachwhat about the kde stuff? It's apparently still in opt and it's getting old. 3.5.5 should be out any day.13:50
sipI think there was a volunteer for a kde-specific repo13:50
cptnwell I fear we have no one to take over KDE in core, opt or contrib13:50
tilmanthat alan guy ;)13:50
treachyeah, heard 0 about that since then.13:51
juetreach seems to be a good candidate ;-)13:51
Romstermay i rase a question about private repos, there is some that i've seen not been maintained for a long time or there badly made and missing files, (md5sum, footprint, and meta fields), would some of the very old ones be removed from the ports list?13:51
sipI think it'll be one of the next topics: dedicated repos13:51
Romsteri'm using KDE 3.4.4 currently.13:52
Viper_is it really so bad if we don't have any kde?13:52
juenot bad, but would be nice to have :-)13:52
treachjue: I'm not unwilling, just unworthy and unfit. :-)13:52
Viper_that's true :)13:52
bd2I'm using it at least (only qt+kdelibs+kdebase, for the konqueror)13:52
tilmanshould we give alan a dedicated kde repo on _now_ or first check whether he's doing an okay job?13:52
Romsterat a pinch i'll take it over but i planed to diverge to other window managers.13:53
tilmanbefore promoting the kde as being official of some sort13:53
aonthere seem to be three kde's in the private repos13:53
aonalthough lithium has just base and libs13:53
sipwell, if it's on a separate repo, even on, it's not official either (as contrib)13:53
tilmanwhat does that mean for x11?13:54
tilmanmmh well13:54
sipx11 is official13:54
Romstersip, if its in a svn for kde then there could be a group that can maintain kde ports?13:54
aonbut what i mean is that even if we don't hear from alan that doesn't make us KDEless13:54
tilmansip: i mean x11 as in "" ;)13:54
Viper_concerning the repo db: i always remove repos which aren't existing anymore13:54
cptnaon: yeah13:55
sipRomster, yes, that's the plan: one or more people maintaining kde-related ports13:55
Romsterwill crux 2.3 be using xorg 7.1?13:55
Romstergood :)13:56
siptilman, I'm afaraid I don't understand13:56
cptnsip: well, x11r7 is a separate repo currently13:56
tilmanis the plan to use my x11r7 repo for an x11 implementation in crux in the future?13:56
tilmanto replace opt/x11?13:56
cptnif we host it in a dedicated repo, how can the user see which one is official?13:56
tilmanthat's why i was asking13:57
sip1) It 'll be on the cd13:57
cptnfrom a tech POV, it would probably make sense since it may be nice to branch for new major releases13:57
sip2) Docs on the website13:57
cptn(make sense to have its own repo)13:57
tilmani agree13:58
juebut should be hosted on IMHO13:58
cptnyes, definitely13:58
Romstergnome, kde, xorg all hosted on crux.nu13:59
cptnthere's one issue though:13:59
cptnwill "opt" depend on the "x11" repo?13:59
cptnI mean, currently we have a clear hierarchy13:59
cptncore > opt > contrib13:59
tilmanx11 would be between opt and contrib i think(?)14:00
tilmansince it depends on a couple of opt ports14:00
cptnbut opt apps use x1114:00
sipcore > opt = xorg > contrib14:00
cptnI wonder if we could merge it, so the users don't realize x11 is a separate git repo14:00
Romsterlet x11r6 goto contrib and add xorg as default.14:01
tilmancptn: is the repo-dependency really a problem?14:01
Romsterall of x11r7 could go into opt?14:01
Romsterand x11r6 be labeled x11-legecy?14:01
tilmanmerging them would be trivial, but is it worth the (minimal) effort? ;)14:01
cptntilman: it's just slightly annoying on prt-get level14:02
cptni.e. also if we'll have more repos in the future14:02
tilmani see14:02
cptni.e. an official gnome one14:02
tilman+1 on merging then14:02
cptnand maybe an official kde one at some point in time14:02
tilmanin the rsync-phase14:02
cptnwe do git -> rsync+httpup already14:03
sipI beg to differ :)14:03
tilmanopt will be huuuge :)14:03
cptnsip: yeah, we wanted to simplify originally :-/14:03
Romsternar lol14:03
*** ChanServ sets mode: +o cptn14:03
Romsterwe don't want a server/desktop install cd repo setup?14:04
tilmancptn: it's annoying in prt-get because there's no clear hierarchy?14:05
cptntilman: because you'll have to add it to prt-get.conf14:05
j^2i dont think we should have multiple cds, or repos, if you want a server you should build it, just like a desktop14:05
cptntilman: if it's just three, I'd say it's okay14:05
sipCould we please go back one step to the unmaintained ports question?14:05
cptnbut the interdependency is a bit unclean14:06
cptnsip: sure14:06
sipJust to be sure we can finish at least one thing :)14:06
cptnyou wish :-)14:06
* sip prepares some coffee14:06
juelast point was to replace pine with mutt/nano -> +114:07
Romstercore, opt, (xorg, kde, gnome), contrib14:07
Romstersince some setups won't need xorg and won't want to keep syncing it all the time.14:07
sipok for removing pine14:07
Romsterheh sip can finish preparing coffee :)14:08
Viper_removing pine is ok14:08
tilmanyeah, that's awesome romster14:08
aonhow about {,k}qemu?14:09
Romsterno objections, i'm a vim user. *looks around*14:10
cptnsip wanted that14:10
siponly in binary form though14:10
sipqemu is broken with gcc414:10
Romsternever thought i'd ever get used to vim.14:10
tilmanRomster: sorry, but this is a crux developer's meeting mainly, so don't comment on each and every point that's brought up14:10
tilmanit's kinda distracting ;)14:10
sip(see the README)14:10
aonsip: so are you planning on going back to source (after|if) it starts working properly with gcc4?14:11
sipaon, right14:11
cptnsip: I think I saw a gcc4 / qemu post on a distro ML somewhere14:12
cptnI'll try to find the URL14:12
cptnhaven't tried it myself though14:13
sipI think that are the same patches we use in qemu now14:13
cptnoh, okay14:14
sipthey're useful for going past compilation, but the resulting binary is somehow b0rked14:14
treachfwiw, qemu works to some extent here with gcc 4.1, but it's a bit random.14:14
treachsome guests seem to work, others doesn't.14:15
jaegereven the guests that work have stability issues14:15
jaegerI used to build the updated ISOs in qemu, can't anymore14:15
treachIC. /me goes back to lurking in his corner.14:15
jaegerpine can go, fine with me14:18
cptnshould we warn crux@ first?14:18
tilmanok. i'll do that when i remove the port14:19
cptnnot sure if there are any pine users remaining, but if yes, they might be happy to switch14:19
jaegerinformation and disclosure are good :)14:19
sipWe may send a more comprehensive email when we move stuff to the attic14:19
tilmanmaybe that should be even automated?14:20
cptnmaybe not to crux@, but it would be nice14:20
tilmanwhen a port is moved from core/opt/any other official repo to attic, queue a notification14:21
aonin my mailbox there are at least 3 different pine clients14:21
aonper and two others14:21
cptnadded/updated/removed ports14:21
tilmanand flush the queue to some ml once a day. maybe.14:21
tilmancptn: well, if we have all 3 of those, then it's almost crux-commits :)14:21
sipwith an attached toilet-flush.wav14:21
cptntilman: well, sure we'd need some queuing14:22
cptnwe could just make the ML default "digest" for that list ;-)14:22
tilmanshould we put this on the todo list?14:24
juethe new iso installer is still dialog-based ?14:24
sipOh, almost forgot. Please put ideas in DevVision and move approved ideas in ToDo14:25
sip(talking of the wiki)14:25
jueif so we need a maintainer for dialog14:25
sipjue, I think so, if we only want to add group selection14:25
sipbtw, cptn lxdialog lacks some stuff (ie progress) so I dont't think it'is an option14:26
cptnsip: okay14:26
cptnwell, are we done with unmaintained ports for now? :-)14:26
sipNot unless we move them to the attic14:27
tilmanwe probably should nail the opt/cvs issue _now_14:27
sipor will we do it later/tomorrow?14:27
tilmanaon: would you want to maintain a client-only version of cvs? ;)14:27
Viper_i thought oct 1514:27
aontilman: i could do that, yeah14:27
Viper_would be the deadline :)14:27
tilmansip: sticking to the oct 15 deadline seems sensible14:28
tilmanobjections to reducing opt/cvs to client-only?14:28
sipfine for me14:28
Viper_it's ok14:28
cptnif it's a problem, a cvs-server port will should up14:28
cptnnot in opt :-)14:29
aonand we can move to opencvs if that ever gets done :)14:29
Viper_btw please change the maintainer tags in the Pkgfile of adopt ports now because after this meeting i will create another list and send this new one to the mailing list14:30
cptnFYI I decided to _not_ take over mercurial14:31
cptnso feel free to grab it if you want14:31
Romstercan there be a svn client-only subversion to? I know some users have bitched at having to isntall the full subversion just to download the source out of a svn repo to build in some ports.14:31
sipok, so the plan is to adopt ports in the next few days, then move the remining ones into the attic.14:32
tilmanViper_: i'll first have to do some homework re. libungif and giflib14:33
Romsteri know we had a striped down one.14:33
tilmanViper_: before taking over libungif14:33
tilmansip: yes14:33
jaegerI'd like to see that list, since I came in late. I'm sure I can take some of them14:33
sipwho offers to take care of moving the cruft?14:33
cptncan't we just switch to git now? ;-)14:33
Viper_made during this meeting14:34
sipRomster, we had a svup (binary) port14:34
jaegerViper_: aye, I'm looking at that one now, just not sure who is taking what at the moment14:35
Viper_i am taking cdrtools14:35
tilmanViper_: can you 'svn up' and recreate the list?14:35
tilmanor make that url above a cgi? :P14:36
Viper_tilman: of course14:36
sipwait a sec14:36
Romstersip, yeah _had_ i'll figure something if i keep seeing users bitch about that.14:36
cptnwell, as svn ports aren't recommended, I don't see that as a problem14:37
Viper_ok updated the list14:38
cptnViper_: could you add the dependent count too?14:39
cptnfor p in `cat orphaned |sed -e 's|opt/||g'`; do d=`prt-get dependent --all $p|wc -l`; echo "$p: $d"; done > deplog14:39
cptnthat's what I used14:39
aonok, so14:39
aoncvs is ready to be committed14:39
aonperhaps i'll go ahead? :)14:40
tilmando it!14:40
sipjaeger, any comment on the black list?14:43
*** j^2 has quit IRC14:44
jaegerblack list? missed that14:44
aonlibart_lgpl probably needs to stay?14:44
jaegerI've got libart_lgpl in my list to look at14:44
cptngimp needs that one14:44
cptndialog and esound are former ports maintained by Per14:45
Viper_and libungif.. it's needed by imagemagnick14:45
cptnwe probably need both14:45
tilmanesound? :(14:45
jaegerbittorrent, ffmpeg, gamin, libart_lgpl, libdvd{css,nav,read}, libgphoto2, libxslt, ncftp, ntfsprogs, poppler, rar, scorched3d, scummvm, startup-notification14:45
bd2please, can you move contrib on the git rails before moving blacklisted packages to attic? Or rather just move them to contrib.git directly, and mark Maintainer: unmaintained. and wait for a week14:45
cptnbd2: why can't the contrib maintains just get them via svn?14:46
bd2there are really much of BASICS packages there: at, calc, openvpn...14:46
aonjaeger: transcode depends on liba52...14:46
bd2cptn, they can. but its double job, imho. 1) move to attic 2) move from attic to contrib14:47
jaegeryeah, though I haven't used transcode recently and would be willing to concede it to someone who does regularly14:47
cptnbd2: otherwise we move them there, and then move them away if no one wants them14:47
cptnit's just that if we move them all to attic it's just one step for us14:47
tilmanmoving them to attic on the 15th is good[tm]14:47
Viper_i suggest moving them to attic even if it could means double work14:47
tilmannot sure the git migration can happen that fast(?)14:47
cptnand then the maintainers share the workload14:47
sipand we'll be sure only maintained stuff ends up in contrib14:48
Viper_orphaned ports belongs to attic and not to the contrib repository14:48
bd2cptn, that is my point. move them to contrib.git, post a mail saying: "Hey! Contrib members, run grep "unmaintained" /usr/ports/contrib/*/Pkgfile and choose any you'd like to maintain, otherwise they'll removed in a week14:48
tilmanthat's cra14:48
tilmanso if we move a broken cvs port there14:49
Viper_so we never get rid of these ports :)14:49
tilmanand some unsuspecting user breaks his box with the port in that one week14:49
tilmanwe're the idiots14:49
cptnokay, I think you made your point :-)14:49
cptnonce we move to git, contrib maintainers will be reevaluated anyway I guess14:50
bd2in the lesser counts14:50
cptnmaybe to start with yes14:51
cptnmike_k told me that the 'ci' repo should go14:51
cptnsince it's not maintained anymore14:51
cptnthings like this14:51
Viper_to sum it up: grab orphaned ports, on oct 15 they will be moved to attic14:53
tilmanit's on the wiki already ;)14:53
bd2no one can grab them except core devels.14:53
bd2so this purpose is meaningless14:53
tilmanno it's not14:53
cptneveryone can grab them14:54
tilmanafter oct 1514:54
bd2cptn, yeah. in their private repos. firsly, everybody will copy attics to their own repos, and will remove them when they will start arriving at contrib...14:54
Viper_tilman: add per's opt ports to the second line14:55
cptnbd2: what exactly is the point you're trying to make?14:55
tilmanViper_: oh, right :)14:55
bd2cptn, move them to contrib.git after 15 Oct and wait.14:55
cptnwell, okay14:56
tilmani think i already made a very good point why that's not good14:57
cptnI think it's been clearly shown why not14:57
tilmanthe horse is dead14:57
cptn^ my words14:57
sipquick, move to another topic!14:57
tilmansip: do you have a schedule for the git migration?14:57
sipok, that is the new topic: git migration, contrib.git etc14:58
sipI have no real schedule, we can start contrib.git right now if we want to14:58
sipthere are a few issues, though:14:59
sipthe first one is slowness of the ssh link on crux.nu14:59
sipthe second is political: who's allowed to contribute?15:00
Viper_will we also discuss something else than the git migiration in this meeting?15:00
Viper_sorry for interrupting you @sip15:00
sipViper_, not interested? :)15:00
Viper_well.. i don't think i will be a great help to you in this topic15:00
Viper_still have to do some work for school :)15:01
Viper_wasn't at home the whole day15:01
sipIf you're in a hurry, we can briefly discuss the new contrib and the portdb now15:01
Viper_yeah that would be fine15:01
Viper_because i am really interested in these topics15:01
sipok, leaving technical difficulties alone, we need a internal maintainer for the contrib repo15:02
sipand probably some changes to the portdb15:02
Viper_do you mean we need a maintainer for the scripts behind the repo or a maintainer of the repo itself15:02
tilmanViper_: "kummertante" ;)15:03
sipIt's more a person who speaks to contrib members15:03
Viper_well i am already doing this job :)15:03
tilmana person contrib members can contact in case of problems, i think?15:04
*** pitillo has quit IRC15:04
sipyes, and also have the power of giving/revoking access15:04
Viper_yeah you gave me a shell account for this :)15:04
sipI'm not sure if he(=Viper_) should talk to a single contrib representative or what15:05
sipany idea on this? shall we nominate a "special" contrib member?15:06
cptnwhat for?15:06
tilmani'm reminded of the gentoo user representive foo15:06
cptntilman: maybe 'cokehabit' has some spare time?15:06
sipto have a 1:1 link between contrib and crux15:06
tilmansip: not sure we need such a person.15:07
sipok, so Viper_ if you feel to, you could play a coordinator role for contrib15:09
Viper_yes but i thaught i was doing this already :)15:11
sipthat's fine, only with git you'd have to deal with shell accounts instead of configuring prtsync15:11
sipjust wanted to make you aware of that :)15:11
tilmanthat's just git-shell, though, right?15:12
siptilman, yes15:12
Viper_ok good but i don't think that this would be a problem :)15:12
sipfine. now, regarding the portdb15:12
sipsince it needs some rewrite or cleanup, and nobody volunteerd so far, I was thinking of a "light" version15:12
sipin whick we simply list the port names (searchable), but we don't actually download all the repos as we do now15:13
*** Romster has quit IRC15:14
sipalso 'cause in my utopia all the good stuff would end up in contrib anyway15:14
sipcomments/ideas on this?15:14
Viper_well it seems to be a good idea15:14
jaegerno objection15:14
cptnsounds good15:14
Viper_many private repositories aren't up-to-date anymore15:14
jueok for me15:14
Viper_so it doesn't make sense to search files in them15:15
*** Romster has joined #crux-devel15:15
sipbe sure to think of it twice, since the protdb is quite useful ;)15:15
bd2if anybody interested, I think that portdb is useless. Personally I don't trust it (except Han repo), and always copy ports found in portdb to my personal directory -> taking maintainership privately15:15
bd2that is why I'm waiting for the contrib.git :-)15:16
jaegersounds like your issue is with maintainer reliability, not the portdb itself15:16
bd2jaeger, right you are. but I don't know most of maintainers -> no trust15:16
sipyes, contrib.git should be a place that you could _almost_ trust15:16
tilmanhan has got quite a lot of crap in his repo :>15:16
aoni think it's good to maintain the portdb15:17
sipor trust a bit more than $unknown_maintainer15:17
bd2tilman, but he is not removing it too often, and especially not BASICS (at, openvpn)... :-)15:17
aonsince how can you prove you're good enough to go to contrib without?15:17
tilmanbd2: :-))15:17
Viper_sip: personally i only need the search function in the portdb for searching opt or core ports15:17
aonwell, of course the repos could be checked individually15:18
aonbut the portdb could cause some complaints from other users if someone starts to push horrible crap into it :)15:18
cptnwell, I think it's obvious we want something like portdb15:19
cptnbut apparently, many packagers don't use it15:19
cptni.e. don't check whether some port exists already15:19
bd2or checks and making dups, because of no trust15:20
bd2as do I15:20
aonyou could also read the pkgfile15:20
aonand then decide whether to use it or not15:20
Viper_well i think the solution sip proposed is quite good: only providing the names of the ports15:21
aonyeah, it seems to be a sane thing to do15:21
Viper_if somebody is interested in it it's no problem to check the private repo15:21
sipok, if everybody's fine with it I'll add the "light" into the todo15:22
bd2I think contrib.git should be much openess, to get more trust from users. the more people in contrib the more chances that port will not be deleted because of maintainer X leaved crux15:22
aonso the people who's pkgfiles you don't trust should be let to fool around in git? :)15:22
cptnwell, let's focus on the agenda15:23
cptnsip: please do15:23
bd2aon, people who can use git can't even laught. :-)15:23
bd2so, no chances that they'll fool around15:23
tilmansip: if you want to, you can clone my x11r7.git repo now, and hook it into crux.nu15:23
tilmansip: or do it whenever you feel like it15:23
siptilman, ok. before going to bed15:24
* bd2 be default thinks about people at the very good attitude. what is that called? optimism?15:24
Viper_is there anything to add to the topic portdb and contrib?15:24
jueone point to contrib: I'd suggerst to add a line like "check your ports with prtverify" to the contrib rules15:24
sipgood idea15:25
tilmanjue: good point, should go in definitely15:25
cptnViper_: we wanted to discuss the new rules for contrib git15:25
cptnbut you may want to skip on that if you have things to do15:25
tilmani had a few other minor points that i don't remember right now (for rules/guidelines)15:25
tilmanbut they probably don't need to be discussed now15:25
treachtilman: communication? ;-)15:26
treachhan and his ignore list.15:26
Viper_cptn: no i think i will stay.. i hope we are discussing this point right now15:26
sipok, so we can discuss the policy for contrib members15:28
sipI really have few ideas here :/15:29
cptnwell, basically the requirements were very low so far15:29
tilmanatm you need one advocate, right?15:29
cptnsince the initial hope was to win many people for the idea15:29
cptntilman: yeah15:29
Viper_that advocate thing is a really good idea15:30
sipcptn proposed to raise the bar to 2 supporters15:30
cptnhowever, as git will be hosted on, some more trust seems appropriate15:30
tilmanwhat about a trial period, in which the new guy sends patches to an advocate15:30
tilmanwho reviews and commits them15:30
tilmanafter the trial period, the new guy would get +w access to the repo15:31
cptntilman: it'll be hard to find this advocate15:31
sipbtw, new wrules should go here:
cptni.e. it's too much work15:31
bd2tilman, pardon. patches to what?15:31
bd2contib ports?15:31
bd2isn't contrib maintained by volunteer maintainers?15:32
cptncrux as a whole is15:32
bd2I mean, "not core"15:32
sipmaybe we can start by granting access for a bunch of maintainers we trust15:32
bd2how someone can apply patch to the port which not maintained by him?15:33
tilmanit was just an idea15:33
tilmanand cptn already bashed it ;))15:33
tilmanso forget about it15:33
bd2ah.. sorry15:33
Viper_are there any other options to raising the bar to 2 supporters?15:34
sipraising it to 3? :)15:35
cptnI'd also say that it _must_ run through some port validator initially15:35
Viper_cptn: i think that's the work of the advocate15:35
cptnI mean, as a hard condition15:35
juecptn: agreed15:36
Viper_i could create a wiki page on how to advocte in the right way :)15:36
Viper_ah we15:36
Viper_not i15:36
sipcptn, you mean only for new members or globally for every contrib 'export'?15:36
cptnthe situation we want to avoid is that people join contrib for the sake of being in there15:36
cptni.e. I inviting my drinking buddies15:36
cptnso if you submit your repo, it won't even get into the stage where someone can advocate for you15:37
cptnunless you pass the validator15:37
cptnthis can even be done automatically15:37
sipisn't it a bit complicated?15:37
cptnwhat's the alternative?15:38
sipI mean, that could be a task the supporter could do15:38
cptnhave all applications go to viper first?15:38
sipmuch as we do now15:38
cptnnow they go directly to crux-contrib15:38
sipno, I mean a request is evaluated by contrib members15:38
Viper_the problem is you cannot advocate somebody by only looking at his Pkgfiles for some minutes15:39
sipwhich are responsible for checking before approving15:39
cptnsure, but if your collegue votes for you, there's no QA done at all15:39
Viper_you will need some time to do this15:39
cptnof course, if we don't trust the current guys in contrib, we're probably screwed anyway15:39
Viper_i.e. download the ports and test sem15:39
sipsure, supporters should have some guideline, as the sort of implicit rules *we* follow now15:40
cptnbut I don't think prologic was as critical towards Romster as I would have been15:40
tilmanprobably not;)15:40
cptnhow about we _assign_ people to verify a repo?15:40
cptni.e. a submission is sent in, two random contrib maintainer get to look at it?15:40
Viper_sounds fair15:40
Viper_if these two maintainer gets some time for it (say one weak) it's ok15:41
sipwell, if those two are on vacation there could be some problem15:41
cptnso we could have a three step procedure:15:41
cptn1. sent to Viper_15:41
Romsteri kind of got thrown in with not much clue to find my own ground...15:41
cptnsmall initial verification15:41
cptn2. Viper_ forwards to two contrib guys15:42
cptnthey confirm within two days, or he chooses a replacement15:42
cptn(vacation situation)15:42
cptn3. the two guys take a decision, submit it to crux-contrib for vetos15:42
aonseems sane15:42
Viper_i like it :-)15:42
cptnViper_: a bit of work for you though15:43
aonand we need to get the initial two contrib guys somehow :)15:43
sipI'd prefer a solution in which Viper_ is involved only for the final check/accound administration15:43
sipbut if it's ok for tyou, nothing against it15:43
cptnsip: should be doable as well15:43
cptnI mean, the least the two guys should do is run prt-val15:44
cptnOTOH then there needs to be some automatic distribution of incoming submissions15:44
cptnotherwise, it'll be always the same guys doing it15:44
cptnalso, there's a catch:15:44
cptnsome maintainers might just not want to do it15:44
cptnalthough I think it's a good training to look at each others ports15:45
deus_exprtverify.log of repo with initial submission?15:45
cptnand learn some tricks from time to time15:45
Viper_well there can be the problem that the maintainer just don't have the time to do this15:45
Viper_but that can be solved by changing with anybody :)15:45
aonit could cause some grudge if someone never answers requests to validate ports15:46
tilmanthe assignment should be randomized i think15:46
cptnthe new rules will have to make sure that this is a requirement15:46
cptndo charity work :-)15:46
aonso will they be removed from contrib if they don't do it?15:46
cptnthat's the only way to enforce this, unfortunately15:47
Viper_i think it will be a good idea to form teams.. that always the same two people are working together15:47
sipwell, if contrib is meant to get people work together, I say we simply let them in charge of basic administration and regulation15:47
bd2that is whole thing so stupid. pardon me, I don't even want to argue about it, but the whole page is stupid. just to you know15:47
aonyou shouldn't have said that at all, then15:47
bd2maintainer is maintain his ports, and it's fine if he don't want do bureaucracy stuff15:48
cptnmaybe we can have crux-branded blinkers15:49
bd2and I'm sure there will a lot contrib member who will like to look for newbies request without enforcing. but removing from contrib for not doing bureaucracy is so stupid.15:50
bd2I'm finished, sorry again.15:50
cptnwell, feel free to do things better15:50
sipbd2, what we proposed here is a way to get a well-organized 3rd party repository15:50
tilmansomeone's gotta take care the whole doesn't go into mayhem15:51
cptnas far as I can see, you don't even maintain a port repo for others to use15:51
cptnso no one expects you to speak in favour of working together15:51
Viper_come on checking another repo for say one or two hours won't kill you15:51
cptnshould you ever start, you'll see that there's some dirty work to do15:51
bd2sip, no need to. give people place and they'll play there nicely.15:51
cptnand equally distribution it might make things easier for everyone involved15:51
tilmanbd2: that's very optimistic15:51
bd2tilman, I agree. but have you tried?15:52
aon...optimistic especially coming from someone who needs to duplicate ports because of not trusting portdb...15:52
cptnportdb is this15:52
cptneveryone could use each others ports easily15:52
cptnbut did anyone care?15:52
Viper_if we do it in the way you proposed (bd2) the quality of the contrib will be very low soon15:52
tilmancptn: yeah well, but that comparison is a bit unfair i think15:52
bd2cptn, portdb is an anarchy. you're purposes monarchy.15:52
tilmanwith the new repo, they are all working in the same working space15:53
tilmanthat's quite a big difference15:53
tilmanbd2: bullshit15:53
cptnbd2: no I don't15:53
bd2I'm purpose the middle of that. Yes, some approval for the newbies should be done, but not so strict15:53
tilmanbd2: they are proposing a self-regulating ... something15:53
bd2tilman, people will self-regulate themselfs!!15:54
tilmanwhen per was the monarch nobody fn complained15:54
tilmanjesus motherfucker!15:54
bd2thay have a mouth to speak, and fingers to type!15:54
tilmangood for them ;)15:54
aonbd2: perhaps you should propose a different scheme15:54
sipbd2, we aim for a community, people playing well together, and sometimes this means somebody should do the burocracy thigs. we're not offering a myspace account simply fill with stuff15:55
jueyep, and first of all we want high qulity ports15:55
jueeven in contrib, which is part of CRUX at all15:56
bd2and what maintainer will get back? beurocracy requests? personally I'm not motivated to contribute just because you're so nice guys. I'm motivated in pkgutils/crux-arm stuff because I'm *interested*, I'm not interested in fear of being removed from contrib, because I'm not doing silly stuff for crux15:58
cptnyeah, but unlike you other users contribute to crux15:59
cptnhttpup repos etc15:59
bd2and of course, you know - I'm nobody here. neither contrib member, not core developer. you can ignore me as you like.15:59
cptnwell, it's not about ignoring15:59
cptnit's about having goals16:00
cptnwhich you haven't shared in the past16:00
cptnand obviously this is still the same16:00
*** jdolan has joined #crux-devel16:01
siphey jdolan16:01
jdolanhehe, hi fellas :)16:02
aonhi jdolan16:02
jdolanlogs anywhere?16:02
cptnjust pick crux-devel there16:03
sipis there enough material to fill up NewContribRules?16:05
jdolanthanks for picking up my slack, sip.16:05
jdolanwrt jdk/jre16:06
tilmani didn't fill in the part about contrib on
tilmanmaybe we can just link to NewContribRules though16:06
sipjdolan, you didn't even reply me when I sent patches ;)16:06
aondidn't seem to be mentioned but i guess it's otherwise quite clear that the maintainers need to sign up on flyspray before they're accepted?16:06
cptnsip: not sure if there's a real agreement here16:06
tilmanyes PLEASE16:06
cptnaon: oh yeah16:06
sipaon, yes16:06
cptnmore bureaucracy!16:07
sipand we should really separate CRUX/contrib there16:07
bd2no, I'm with you here :-)16:07
jdolansip, i know, sorry.  i've been working 10 hour days, 6 days a week.  it sucks.  that's my only excuse :-/16:07
cptnbd2: look, there's another side to this:16:07
sipjdolan, np.16:07
cptnbd2: we have somewhere around 10 repos in contrib16:07
tilmansip: sounds good16:07
cptnin say a year16:07
cptnbd2: so it'll hardly ever hit one of the contributors to help out16:08
cptnbd2: i.e. look after a ports repository16:08
cptnif someone is not willing to invest an hour per 6 month to look at someone else's ports, he's not in the right place in contrib16:08
sipunfortunately flyspray is a bit unfriendly with separate projects.16:09
cptnsip: let's switch, or rewrite!16:09
bd2they are investing their time in maintaining their ports. they are already doing good thing16:09
siplet's patch to death! :)16:09
sipbd2, sometimes it's not enough16:09
jdolangot out of work with barely enough time to get a bike ride in before dark, wouldn't you know it i have a flat tire.16:09
cptnbd2: well, if they can't invest that hour16:10
cptnthey certainly won't care about user bugs16:10
bd2they care about bugs in their ports!16:10
cptnthey did a good thing already, just for themselves16:10
bd2that is job splitting16:10
cptnthey should also care about the health of the collection16:10
cptnsince they're part of it16:10
tilmanbd2: you're boring ;)16:10
cptnwell, to be fair I started it again16:11
cptnI'll shut up now16:11
tilmancptn: you're boring ;)16:11
Romsteri care about the whole crux distro in general, don't know about bd2's views...16:11
bd2why shold they care about health? if anybody will keep their ports healthy, then all repo will be good!16:11
bd2overall health16:11
tilmando we want to discuss anything further?16:11
Romsterbd2, some bugs may go unnoticed, ever thought about that?16:12
jdolannot to be a dick or anything, but i thought non-maintainers were asked to hold their comments/questions for the meeting.16:12
*** bd2 has left #crux-devel16:12
jdolaniow, this is not the place to argue with cptn unless you're ....16:12
Romstererm sorry..16:13
jdolancptn takes enough shit from :)16:14
sipyes, please guys, let us talk a bit so we can put toghether a proposal16:14
sipok, I hasked before if we had enough material for a NewContribRules page because I have a half-plan16:16
sipIf a couple of voulnteers can fill it with the most successful ideas, we can than review it later16:17
sipI say so because I think the new contrib is quite urgent16:18
sipsince we're moving some stuff to the attic16:18
tilmani liked your proposals16:18
tilmanautomatic prtverify run + randomlny picking peers to review16:19
tilmanbut i'm not sure whether that idea was popular with all of us16:19
tilmanthanks to the mess above16:19
cptnI especially like that one as so far, you could submit a repo, and no one would care then16:19
cptnhappened more than once16:19
cptnin the new situation, you'd at least get a "not yet"16:19
cptnit appeared frustration, although it was that way on purpose16:20
cptnnot to be frustration, but to encourage making a name in the community first16:20
cptnfrustration -> frustrating16:21
aoni think some quick non-automatic check before supplying the repository to peer review would be good16:21
tilmani'll be boring and go to bed now16:21
tilmansip: maybe you can outline the basic stuff in the wiki and i'll comment on it16:22
aonso that they aren't bothered with complete insanity that just happens to pass the tests16:22
tilmanor something16:22
jaegerif it's going to be reviewed anyway, why the non-auto check instead of auto?16:22
cptnif Viper can cope with it, it might be nice if the peers would report back to him16:22
cptnso Viper's the bad guy :-)16:22
cptnnight tilman16:22
aonnight tilman16:22
jaegernight, tilman16:22
siptilman, ok. night!16:22
jdolannn tilman16:24
jdolani think i just caught a spurt of motivation.16:24
jdolani'll update my ports tonight!16:24
aon23:41 <@cptn> 1. sent to Viper_16:25
aon23:41 <@cptn> small initial verification16:25
aon23:42 <@cptn> 2. Viper_ forwards to two contrib guys16:25
aon23:42 <@cptn> they confirm within two days, or he chooses a replacement16:25
aon23:42 <@cptn> (vacation situation)16:25
aon23:42 <@cptn> 3. the two guys take a decision, submit it to crux-contrib for vetos16:25
aon23:42 <@cptn> done16:25
aon23:42 < Viper_> i like it :-)16:25
jdolani can dig that.16:25
aonhow do you think the other maintainers would react if they had to crawl through endless amounts of really bad ports?16:25
jdolanyou don't need to crawl through them if the first few are bad16:25
jdolan"hey this looks like shit, try again."16:26
cptnif they're really bad, prtval should catch it, no?16:26
jdolani think it'd also be nice to add something to prtverify [or have another tool] that checks for aged files.16:27
aoni've no idea what that is, but prtverify can let some stuff through that shouldn't16:27
jdolanand just issues a warning, like "Repository jdolan has not changed in 30 days."16:27
jaegerjdolan: this'll help, it makes me laugh every time I look at it :)
jdolanlol, nice.  arr maytee.16:28
jaegeranyway, I've got to go, but that seems like a decent procedure so far. I'll be back on when I get home16:28
*** jaeger has quit IRC16:28
sipso, shall we continue another day? I planned to go to bed an hour ago ;)16:29
cptnyeah, sounds like a plan16:30
aoni'm planning on going to bed now16:30
aongn ->16:30
cptnnight aon16:30
sipnight aon16:30
sipthere's a long list of topics to be discussed. Anybody has some date in mind?16:31
sipor shall we just discuss things sparsely?16:31
cptnmaybe we could establish some regular day?16:31
cptnlike every tuesday?16:32
cptnif we can do it regularily, it should go down to 30 minutes to 1 hour16:32
cptnand if you miss one, it's not that much16:32
sipevery day is fine for me as long as the time is the same16:32
jdolanSee You Next Tuesday16:33
sipsounds good to me16:33
jdolan(C U N T)16:33
jdolani suck.16:33
jdolani'm going to go try to run 5 miles.  should be painful.16:33
sipjoin channel #crux-cunt for some action16:33
jdolansorry i was late, like i said i'll do my best to prt-get update jdolan soon.16:33
jdolanlol :D16:33
jdolanadios :)16:34
cptnbye jdolan16:34
sipsee you jdolan16:34
cptnand good night everyone16:34
cptntime to get some sleep16:34
sipnight cptn16:34
sipleaving too16:34
juenight all16:35
sipsee you tomorrow16:35
*** cptn has quit IRC16:36
*** sip has quit IRC16:36
*** jue has quit IRC16:36
*** Romster has quit IRC17:26
*** Romster has joined #crux-devel17:27
*** jaeger has joined #crux-devel18:18
*** Romster has quit IRC18:56
*** Romster has joined #crux-devel18:58
treach"What do you think?" "STFU"? :-/19:05
*** treach has left #crux-devel20:17
*** Romster has quit IRC23:18

Generated by 2.11.0 by Marius Gedminas - find it at!