--- Log opened Thu Oct 06 00:00:20 2005
00:12:54-!- cmeme [n=cmeme@boa.b9.com] has quit [Read error: 110 (Connection timed out)]
00:15:18-!- beniamino [n=ben@adsl-68-123-122-243.dsl.snfc21.pacbell.net] has quit ["Leaving"]
00:16:15-!- RLD_osx [n=rldempse@66-190-76-181.dhcp.dntn.tx.charter.com] has joined #fink
00:17:13< asari> shres, building modules is not very large, i think install set is the problem
00:17:52< asari> see %b/instsetoo_native/unxmacxp.pro/OpenOffice/install/ after build with-lang=ALL
00:20:40< asari> there are installation set of each language, like en-US, en-GB,... up to 30.
00:20:42< asari> and each one consumes 450MB.
00:21:45-!- cmeme [n=cmeme@boa.b9.com] has joined #fink
00:21:50-!- RLD_osx [n=rldempse@66-190-76-181.dhcp.dntn.tx.charter.com] has quit [Read error: 104 (Connection reset by peer)]
00:24:51< asari> additionally, on Mac OS X, disk images (*.dmg) were also created on building, I thought.
00:28:02-!- shreyas [n=sshreyas@59.92.154.170] has joined #fink
00:30:04< asari> Did I said something non-understandable?
00:30:36< asari> ah.. shres, welcome back :)
00:31:29< shreyas> asari: sorry, got disconnected. Accidently pulled the wire
00:31:39-!- RLD_osx [n=rldempse@66-190-76-181.dhcp.dntn.tx.charter.com] has joined #fink
00:31:47< asari> !
00:36:42-!- shres [n=sshreyas@59.92.151.222] has quit [Read error: 110 (Connection timed out)]
00:38:58-!- RLD_osx_ [n=rldempse@66-190-76-181.dhcp.dntn.tx.charter.com] has joined #fink
00:39:13-!- RLD_osx [n=rldempse@66-190-76-181.dhcp.dntn.tx.charter.com] has quit [Nick collision from services.]
00:39:33-!- RLD_osx_ is now known as RLD_osx
01:01:49-!- chris01 [n=chris01@212.126.165.246] has joined #fink
01:12:20-!- mjayant [n=mj@202.144.95.244] has joined #fink
01:17:55< shreyas> which java does 10.4.2 have ? version ?
01:21:07< dmacks> Probably 1.5 and 1.4.2
01:22:06< shreyas> thats by default, without s/w upgrades ?
01:22:26< dmacks> No idea.
01:22:48-!- dmacks [n=dmacks@pdpc/supporter/active/dmacks] has quit ["leaving"]
01:29:11< chris01> shreyas: http://docs.info.apple.com/article.html?artnum=301073
01:29:23< asari> shreyas:  /System/Library/Frameworks/JavaVM.framework/Versions/
01:30:22< shreyas> asari: i have 3 versions there :-)
01:30:25< shreyas> chris01: thanks
01:30:28-!- lisppaste [n=lisppast@common-lisp.net] has quit [Connection reset by peer]
01:30:33-!- lisppaste [n=lisppast@common-lisp.net] has joined #fink
01:32:13< asari> shreyas: also you can check it with `fink list system-java`
01:32:35-!- cmeme [n=cmeme@boa.b9.com] has quit [Connection timed out]
01:32:55-!- cmeme [n=cmeme@boa.b9.com] has joined #fink
01:33:07< shreyas> asari: cool, never knew that
01:34:13< asari> you must have Depends: system-java15 written in .info, if you want to use Java 5.0 for your packages.
01:42:05 * asari -> swimming
02:19:21-!- sid77 [n=sid77@host-84-222-60-75.cust-adsl.tiscali.it] has joined #fink
02:19:33< dkg> hi
02:19:51< dkg> i installed mysql.
02:20:28< dkg> but its just the client - not the server.
02:20:33< dkg> where to get the server?
02:21:11< mcp> dkg: fink apropos mysql
02:21:16< dkg> ic, i have to run mysqld.
02:21:18< mcp> and see the package name yourself ;)
02:21:23< dkg> but is there no start/stop script?
02:21:56< dkg> mysqld_safe?
02:24:00< mcp> dkg: fink describe daemonic
02:24:13< chris01> dkg: I think you can do: 'sudo daemonic enable mysql'
02:25:22< dkg> chris01: this looks nice.
02:25:25< dkg> chris01: i did it.
02:25:28< dkg> chris01: whats next?
02:25:36< dkg> chris01: how to start/stop?
02:27:33< dkg> there is a xml for mysql now.
02:27:40< chris01> dkg: daemonic will make sure that mysql ist startet next time you start you mac
02:27:57< chris01> so, you can either restart and check, or start mysql manually this time
02:28:43< dkg> chris01: how to start/stop?
02:28:48< chris01> dkg: read the info in 'fink info mysql'
02:28:55< dkg> thanks!
02:29:47< mcp> dkg: you might run into trouble if you want to connect with phpmyadmin to the mysql in fink. last time I tried (~1 month) ago it wasn't able because of too old libraries
02:30:26< dkg> mcp: thanks for this hint.
02:32:21< mcp> dkg: kein thema :)
02:34:27-!- kane_ [n=kane@perl.xs4all.nl] has joined #fink
02:40:28-!- shreyas is now known as shres
02:49:21-!- shres [n=sshreyas@59.92.154.170] has quit ["What can i do? I am just being me"]
03:05:42-!- lisppaste [n=lisppast@common-lisp.net] has quit [Read error: 104 (Connection reset by peer)]
03:19:55-!- lisppaste [n=lisppast@common-lisp.net] has joined #fink
03:38:13-!- spiral [n=spiral@laf31-2-82-224-249-43.fbx.proxad.net] has joined #fink
03:44:58-!- geewz [n=gregreed@ppp122-142.static.internode.on.net] has joined #fink
03:56:17-!- geewz [n=gregreed@ppp122-142.static.internode.on.net] has quit ["Leaving"]
04:21:12-!- Fang [n=Fang@2002:56c2:e14c:1:0:0:0:1] has joined #fink
04:30:39-!- gecko2 [i=gecko@unaffiliated/gecko2] has quit [Read error: 104 (Connection reset by peer)]
04:38:49-!- gecko2 [i=gecko@unaffiliated/gecko2] has joined #fink
05:11:19-!- spiral [n=spiral@laf31-2-82-224-249-43.fbx.proxad.net] has quit [Remote closed the connection]
05:30:16-!- lisppaste [n=lisppast@common-lisp.net] has quit ["Common Lisp IRC library - http://common-lisp.net/project/cl-irc"]
05:30:22-!- lisppaste [n=lisppast@common-lisp.net] has joined #fink
05:52:29< muesli> hey guus
05:52:34< muesli> guys even
05:53:20< muesli> i have a little mac osx problem... maybe someone knows a trick / help
05:53:21< muesli> i'm compiling and linking a program to the CoreAudio libs on 10.3.4
05:53:21< muesli> passing this binary to a 10.2.8 user makes the app crash on audio access
05:53:27< muesli> is there some way to enable a "backwards compatibility" mode?
05:57:42-!- geewz [n=gregreed@ppp122-142.static.internode.on.net] has joined #fink
06:11:25-!- geewz [n=gregreed@ppp122-142.static.internode.on.net] has quit ["Leaving"]
06:21:49-!- michal_ [n=michal@dsl-58-6-123-9.qld.westnet.com.au] has left #fink ["Leaving"]
07:36:59-!- sid77 [n=sid77@host-84-222-60-75.cust-adsl.tiscali.it] has quit ["http://slackintosh.workaround.ch/"]
07:41:39-!- shres [n=sshreyas@59.92.151.121] has joined #fink
07:43:32-!- sid77 [n=sid77@host-84-222-60-75.cust-adsl.tiscali.it] has joined #fink
07:43:33-!- sid77 [n=sid77@host-84-222-60-75.cust-adsl.tiscali.it] has quit [Remote closed the connection]
08:08:47-!- asari [n=chatzill@p2204-ipbf509marunouchi.tokyo.ocn.ne.jp] has quit ["Chatzilla 0.9.68.5 [Firefox 1.0.7/20050919]"]
08:09:10-!- asari [n=asari@p2204-ipbf509marunouchi.tokyo.ocn.ne.jp] has joined #Fink
08:25:21-!- dmacks [n=dmacks@pool-70-22-58-73.balt.east.verizon.net] has joined #fink
08:26:46-!- sid77 [n=sid77@host-84-222-60-75.cust-adsl.tiscali.it] has joined #fink
08:28:14-!- RangerAway is now known as RangerRick
08:28:39< dmacks> muesli: Apple provides SDKs of old OS X versions.
08:28:45< dmacks> Link to the libs there.
08:29:00< RangerRick> muesli: you need to link the libs in /Developer/SDKs/MacOSX10.2.8.sdk and such
08:29:10< dmacks> RangerRick: libftw0 fixed for tiger, unless Apple screwed up its headers more than I thought:)
08:29:15< RangerRick> although I don't know the specifics of doing so
08:29:26< RangerRick> dmacks: alright, I'll give it a shot
08:29:56< muesli> RangerRick, dmacks: you guys rock, once again
08:30:06< dmacks> :)
08:30:20< RangerRick> once again?  that implies that we *stopped* rocking at some point
08:30:21 * RangerRick grins
08:30:54< dmacks> You took a vacation few wks ago. Net rockitude in #fink was reduced.
08:30:54< kane-xs> you may have been unrocking between the hours of 2am and 7am, at which point you were more 'dozing'
08:30:59< RangerRick> hahaha
08:31:10< kane-xs> ... unless in a rocking chair of course, at which points the observation obviously fails :)
08:31:29< dmacks> Or unless....okay, let's not get into hat discussion, shall we?
08:31:32< dmacks> *that
08:31:39< muesli> okok
08:31:47< muesli> "you guys still keep on rocking" ;)
08:32:36 * kane-xs eyes dmacks
08:32:36< kane-xs> dirty dirty boy!
08:32:55< dmacks> Looks like debian is having some sort of ABI upgrading crisis. "please don't file 380 bug reports yet" someone wrote.
08:33:09< kane-xs> ABI?
08:33:45 * kane-xs waits for a bot with a clever reply
08:33:47< RangerRick> we don't have clever bots
08:33:47< RangerRick> clever bots don't forget everything every 3 days
08:33:47< dmacks> heh
08:34:04< kane-xs> hmm. there used to be some finkbot here iirc
08:34:10< kane-xs> it forgot to join the channel i take it?
08:34:16< dmacks> (also, melian stopped listening to foo?, now only hears !foo)
08:34:24< RangerRick> all our bots have various mental or endurance problems
08:34:25< RangerRick> :)
08:34:39< kane-xs> !ABI
08:34:55< kane-xs> hmmm
08:34:59< dmacks> Melian sorta rocks for a few hours or days at most before falling over completely.
08:35:12 * RangerRick just needs to set up his own damn infobot
08:35:26< kane-xs> google suggets it's the "Assocatiation of British Insurers" -- although I can see how this is a problem for debian, presumably it's not what was intended
08:35:46< RangerRick> hah
08:35:55< dmacks> heh
08:36:07< muesli> hmm
08:36:08< muesli> can't find this SDK dir
08:36:14< muesli> downloaded  the old CoreAudio SDK, though
08:36:29< RangerRick> muesli: if you don't have it, you need to install the optional stuff from Xcode
08:36:30< dmacks> muesli: It's probably an optional install in the XCode cd.
08:36:50< muesli> ah
08:36:51< muesli> hm
08:36:51< muesli> downloading xcode again
08:36:57< muesli> arg
08:37:04< muesli> :(
08:37:04< RangerRick> don't delete it this time  ;)
08:39:25< RangerRick> uh, wait a minute
08:39:27< RangerRick> SF is changing databases without migrating old data?!?
08:39:32< dmacks> !!
08:39:34< RangerRick> what the bloody hell
08:40:33< chris01> hi dmacks, RangerRick: are there any news about moving libgettext3 to stable?
08:40:33< dmacks> After you set up an infobot, could you set up a developers' site that doesn't continually screw the developers?
08:40:43< kane-xs> you actually /use/ sf.net for fink?
08:40:45< RangerRick> https://sourceforge.net/docs/A03/
08:40:47< kane-xs> i've always found it less than hopeless... only thing running theere still is our mailinglists..
08:41:11< RangerRick> and even then, they use an old mailman, so it doesn't have the nicer list management tools  :P
08:41:32< kane-xs> yeah, they suck... they removed the 'explicit white list' option too
08:41:44< kane-xs> but for histarical raisins we're still there
08:41:52< kane-xs> also a PITA to move the lists to another address etc
08:41:54< dmacks> Damn, I had to go to the second hit for google-images for "abi" before finding something pr0nish.
08:42:13< kane-xs> but they don't suck *as much* as their CVS setup, or shell machines
08:43:22< dmacks> Every time they screw us, we start looking for alternatives, but the effort peters out before we actually *do* anything.
08:44:15< kane-xs> dmacks: hypothetically speaking, if magically a machine dropped out of nowhere with  nice bandwith, raid config, backup storage, a p4 open source license and someone maintaining the hardware for you, would that Fix It?
08:45:16< RangerRick> kane-xs: that'd be pretty sweet
08:45:21< kane-xs> you need some mailinglist software to go with that?
08:45:25< RangerRick> :)
08:45:27< dmacks> I think drm recently tried to figure out "what resources we use".
08:45:31< RangerRick> we could always leave the lists on SF for that matter
08:45:45< RangerRick> kane-xs: if you're willing to donate, please send a message to fink-core
08:45:51< RangerRick> if nothing else it will get the discussion going again
08:47:02< kane-xs> rangerrick: i have the privilige of working for a company that's very pro open source and the like.. we sponsor among others debian.org, pythong.org and a host of perl things.. i can always inquire if there's room for more
08:48:22< kane-xs> hmm... it's actually pythoN.org... not sure how that g crept in there...
08:51:10< sid77> kane-xs, rofl
08:51:13-!- shreyas [n=sshreyas@59.92.139.218] has joined #fink
08:53:29< RangerRick> I don't want to know what that python is doing in your thong, thankyouverymuch
08:54:59< dmacks> Hopefully it's not this: http://www.miami.com/mld/miamiherald/news/weather/environment/12820947.htm
08:55:35 * dmacks helpfully redirects #fink's mind ever-so-slightly out of the gutter.
08:55:40< RangerRick> heh
08:56:51-!- akh [n=akhansen@ldx3.psfc.mit.edu] has joined #fink
08:57:37< akh> Hi, all!
08:58:33< asari> hi, akh
09:00:47-!- msachs [n=msachs@c-24-34-72-223.hsd1.ma.comcast.net] has joined #fink
09:01:02< dmacks> Well *that* was a bloody slow summoning
09:01:12< msachs> Hey dmacks
09:01:15< msachs> Looking for me?
09:01:27-!- shres [n=sshreyas@59.92.151.121] has quit [Read error: 110 (Connection timed out)]
09:01:35-!- sid77 [n=sid77@host-84-222-60-75.cust-adsl.tiscali.it] has quit [Remote closed the connection]
09:01:45 * dmacks noticed the logs from your 'make fink world' runs were gone:(
09:01:52< dmacks> drm was lookin' too, dunno why.
09:02:25< msachs> Ah, I think they moved the website over to another box and didn't migrate the data.
09:02:39< msachs> Let me put my last one up there...
09:02:42< dmacks> Hm, that'd do it, yup:)
09:03:22-!- drm [n=drm@ip68-108-245-119.sb.sd.cox.net] has joined #fink
09:03:41< drm> hey, bbraun
09:03:44< dmacks> cooOOool!
09:03:51< pogma> on opendarwin?
09:04:16< pogma> should still be available at sancho.opendarwin.org if they moved www to sampson
09:04:29< pogma> hey drm!
09:04:33< drm> hi pogma
09:04:42 * drm wonders whats cool
09:05:18-!- spiral [n=spiral@laf31-2-82-224-249-43.fbx.proxad.net] has joined #fink
09:05:38< dmacks> msachs appeared, I told him "drm was looking you you", then you appeared.
09:05:40< pogma> drm: fink, of course :)
09:05:47< msachs> Crazy.
09:05:53< drm> hi msachs
09:06:08< spiral> hi
09:06:25< msachs> Okay, I don't appear to have access to sampson, I'll mail bbraun.
09:06:29< msachs> heya drm
09:06:31< msachs> How's it going?
09:06:45< dmacks> pogma: the URL I was using was fink.od, not www.od; I don't know how virtual hosts are mapped:(
09:07:00< msachs> Yeah, http://sancho/fink/build doesn't work.
09:07:05< drm> msachs: are you looking for your logs?
09:07:24< msachs> drm: Well *I* know where they are, they just happen to be somewhere where other people can't get to them...
09:07:31< drm> msachs: i complained to bbraun about them being missing, like, two weeks ago
09:07:42< msachs> I'll poke him, then.
09:07:48< drm> and apparently there is some issue to do with getting ldap enabled
09:07:53< drm> which is kvv's problem
09:07:57 * dmacks can't even remember my op passwords...hope I had ssh keys set...
09:07:58< drm> that he's working on :)
09:10:51< msachs> Consider him poked.
09:11:36< dmacks> msachs: also related to logs, fink will soon automatically save a copy of STDOUT/STDERR if you want to capture it.
09:12:04< msachs> Ah, useful feature.
09:13:36 * dmacks ponders committing it with a horrible internal API vs spending time learning how to make it not suck first.
09:14:10< pogma> dmacks: you can't make it not suck, can only reduce the suckiness :)
09:14:11< drm> msachs: on another matter, is your team familiar with the comments at www.swox.com/gmp , particularly the badmouthing of apple's gcc releases?
09:16:04< pogma> drm: The gmp folks assume that their code is perfect and seem to blame everyone else when it breaks, not an uncommon trait :)
09:16:45< msachs> Not that I'm aware, I'll pass that along, thanks.
09:17:16< drm> pogma: even worse, the gmp folks tell you that stuff may be miscompiled because it fails "make check" on their non-public CVS , not because it failed "make check" in their released code
09:17:27< drm> so there is no way to even tell... it is very frustrating
09:17:33< dmacks> yuck
09:18:21< drm> i'm going to follow their advice and force compiling with gcc 3.0 even though make check works with gcc 4.0.0, and reevaluate after gcc 4.0.1 has been released by apple
09:18:55< pogma> drm: There is a bad hack in the old info file to allow intel compilation
09:19:05< pogma> drm: presubamly still needed
09:19:10-!- vasi [n=vasi@modemcable133.147-70-69.mc.videotron.ca] has joined #fink
09:19:29< drm> yes, i left it in place
09:19:34< drm> having no way to test it of course :)
09:20:48< drm> unfortunately, to handle a situation with the "mpfr" library, i have to revise the gmp package in a non-backwards-compatible way, which means investigating all of the packages that depend on gmp with an eye to revising them all at once...big pain
09:21:28< dmacks> Give the new gmp a new name, then others can get upgraded as time permits?
09:21:46< msachs> drm: I can check it if you'd like.
09:22:29< drm> msachs: there is no way to check it, since the gmp's "make check" doesn't have the tests that fail
09:22:45< msachs> Oh, *that's* why you have no way to test it.  Lovely.
09:23:45< drm> somebody asked on the gmp list in july what they were supposed to do to test, and the response was "maybe we'll backport some more tests to the release branch", but they have yet to do so
09:24:13< pogma> drm: But msachs can still test intel build for you...
09:24:13< drm> oh, sorry...yes, that would be good
09:24:40< drm> too much multitasking too early in the morning!
09:24:45< drm> bbraun?
09:25:52< drm> guess its kinda early for him, too
09:25:55< pogma> he's been celebrating his new job a bit recently
09:25:55< msachs> What one's that?
09:26:06-!- shreyas_ [n=sshreyas@59.92.155.204] has joined #fink
09:26:07< drm> i was gonna ask him if sampson is running mysql
09:26:56< pogma> he got something working for a company that does top-secret govt contracts is all he's said about it, don't think he's started yet though
09:27:06< msachs> drm, cvs.sf.net doesn't love me, is gmp 4.1.3-11 new enough for you?
09:27:10< msachs> Fun.
09:27:55< drm> msachs: i am packaging gmp 4.1.4, but it won't be on fink for a while yet... i'll send you info/patch files if you'd like to try it on intel for me
09:28:07< msachs> drm: Sounds good.
09:30:12-!- shreyas_ is now known as shres
09:30:12-!- lisppaste [n=lisppast@common-lisp.net] has quit ["Common Lisp IRC library - http://common-lisp.net/project/cl-irc"]
09:30:15-!- lisppaste [n=lisppast@common-lisp.net] has joined #fink
09:32:45< vasi> mornin' all
09:32:49< dmacks> h'lo vasi
09:32:59< msachs> heya vasi
09:33:58< vasi> drm, you want 'make' to become base & essential?
09:34:13-!- baba [n=baba@YahooBB220041000212.bbtec.net] has joined #fink
09:34:14< drm> vasi: i don't see why
09:34:50< drm> vasi: can't somebody be a binary-only user without it?
09:35:18< vasi> drm, yes i think so
09:35:39-!- shreyas [n=sshreyas@59.92.139.218] has quit [Read error: 110 (Connection timed out)]
09:36:09 * vasi must be confoozled, lemme re-read some of that thread
09:36:18< drm> the real problem, as i was discussing with dmacks last night, is with letting somebody try to use the rsync or cvs update methods if they don't have the dev tools installed
09:36:39< drm> any such person should be forced to stick to "point" updates
09:37:04< dmacks> (...which is now prohibited in HEAD)
09:37:19< drm> dmacks: good...haven't been following the commit logs
09:38:38< vasi> dmacks, just read your patch
09:39:19< vasi> i think it's not so cool to let the user choose 'cvs', then die and say 'loser, your choice is wrong!' :-)
09:39:39< vasi> maybe instead we should just remove the wrong choices?
09:39:41< drm> "fink roulette"
09:40:00< dmacks> Yeah probably good plan.
09:40:34< vasi> oh i was reading up on how debian and debian-derivatives are handling the G++ transition
09:40:44< drm> do tell
09:41:01< vasi> apparently they're actually renaming all the packages from foo to foo-c2!
09:41:06< vasi> well, all the ones that we would consider 'GCC:' packages
09:41:46< dmacks> (/me was trying to put off firing up the pdb until as late as possible)
09:42:01< vasi> i suppose that's easier for them because very few debian users build things from source...so nobody but the build-team has to do the remove-everything-and-rebuild ugliness
09:42:14< drm> yeah
09:42:37< vasi> dmacks, does the mysql switch affect us at all?
09:43:07< drm> vasi: our online package database uses sourceforge's mysql
09:43:19< vasi> drm, yes that's why i'm asking
09:43:32< vasi> i don't know if we use any features that have been deprecated or anything
09:43:43< drm> but we have to do the data conversion
09:44:13< vasi> oh my
09:44:21< drm> my feeling is, that sf's mysql has been flaky sufficiently often, that this should motivate us to move mysql operations to opendarwin if that's possible
09:44:33< vasi> sounds like a plan :-)
09:45:08< pogma> 'cause opendarwin is never flaky :-p
09:45:16< drm> pogma: hehe
09:45:23< vasi> shhhh pogma, everything's gonna be perfect
09:45:54 * vasi puts his hands over his ears and sings as loud as he can
09:46:50< pogma> really need just three -core members, one constantly with hands over ears, one with hands over eyes and one covering mouth...
09:47:02< muesli> dmacks, RangerRick: ok, i got the sdk
09:47:14< muesli> how can i force the gcc to use this coreaudio framework from 10.2.8?
09:47:18< dmacks> pogma: Need a fourth, bent over a chair, as a liason to SF
09:48:06< dmacks> muesli: -F
09:48:08< vasi> lol
09:48:22< pogma> dmacks: would his hands be covering the orifice? Or enlarging it?
09:49:43< dmacks> Heh, no need for goatse here.
09:49:56 * drm barfs
09:50:14< msachs> drm: gmp builds.
09:50:20< drm> msachs: thanks
09:50:28< drm> see y'all later
09:50:35< msachs> *waves*
09:50:36-!- drm [n=drm@ip68-108-245-119.sb.sd.cox.net] has quit ["Leaving"]
09:50:43< dmacks> ANYway...given that we have to migrate our DB's "somewhere" may as well try whatever options we have.
09:50:58< muesli> dmacks: -F?
09:51:17< dmacks> (RTFgccM:)
09:51:20< muesli> heh
09:51:24< muesli> yeah, i do atm
09:51:32< muesli> but the only thing it says about -F is "apple only"
09:52:02< RangerRick> -F is the equivalent of -L for frameworks
09:52:06< muesli> ah
09:52:09< RangerRick> -F /path/to/Library/Frameworks directory
09:52:11< muesli> and it overrides the default location?
09:52:27< msachs> muesli: And also -framework CoreAudio or what-have-you.
09:52:32< dmacks> Mine says "In Apple's version of GCC only, add the directory dir to the head of the list of directories to be searched for frameworks."
09:52:34< RangerRick> it becomes part of the framework search path, yes, before any system ones
09:52:52< muesli> cool
09:52:54< muesli> thanks
09:53:46< muesli> hmm
09:53:51< muesli> still the same binary size
09:54:02< muesli> anyway to check i it really used this lib?
09:54:13< RangerRick> not sure
09:54:17< RangerRick> otool -L
09:54:19< RangerRick> look at the versions
09:54:35< RangerRick> not sure if that will help or not though :)
09:55:33< dmacks> There's got to be *some* way to know if you used the lib. Otherwise it wouldn't matter if you did or not.
09:55:40< muesli> heh, always has been 1.0.0 for coreaudio
09:56:00< muesli> dmacks: well, i would need a 10.2.x system to test it ;)
09:56:56< dmacks> If your attempt to use it vs your known not using it attempts are indistinguishable, then either you didn't use it or it doesn't matter.
09:58:35< asari> !paste
09:58:35< Melian> lisppaste is a bot that lets you post large chunks of text without flooding the channel. Use it at http://paste.lisp.org/new/fink
10:03:43< lisppaste> asari pasted "Data conversion?" at http://paste.lisp.org/display/12299
10:04:11< asari> I tried with my mysql, and it seems that no data conversion needed.
10:04:59< asari> *.sql are taken from shell.sourceforge.net:/home/groups/f/fi/fink/pdb
10:05:42< dmacks> asari: Great!
10:07:05< vasi> woohoo
10:07:31< dmacks> Wow vasi, get_options is complicated.
10:07:32< asari> only 10 minutes to migrate ;)
10:07:47< vasi> dmacks, i prefer to call it 'powerful' :-)
10:08:19< RangerRick> dmacks: the new libftw0 worked fine, btw
10:08:23< RangerRick> anyone else up for doing a smoke test of the apt and dpkg in my exp?
10:08:24< RangerRick> I'd really like input on upgrades
10:08:43< vasi> i'll try 'em out sometime soon
10:08:56< vasi> i'd especially like to find out how synaptic works with the new dpkg :-)
10:09:07< dmacks> Should I put libftw0 in base/? Essential:true?
10:09:49< RangerRick> is apt essential?
10:09:57 * asari already in bed
10:10:07< vasi> g'night asari :-)
10:10:12< asari> see you all
10:10:29< asari> night!
10:10:32-!- asari [n=asari@p2204-ipbf509marunouchi.tokyo.ocn.ne.jp] has quit ["Quitting!"]
10:10:34< dmacks> Or would that be screwing ourselves, since (drm would say) we need to hardcode the BDO:true/Essential:no headers pkg into core?
10:10:57< dmacks> Whoops, no apt isn't. So ne'ermind:)
10:11:07< vasi> i'm pretty sure that 'base' is supposed to be self-contained though
10:11:23< dmacks> Right. /me was concerned about Essential flag
10:12:07< vasi> also, apt is built as part of bootstrap
10:12:23< RangerRick> apt's not essential
10:12:28< RangerRick> just base
10:12:40< RangerRick> at least, it's not in the info file  :)
10:12:44< vasi> it's not essential (though i think it is 'important' in the debian sense)
10:12:49< RangerRick> yeah
10:12:52< vasi> but it's in Bootstrap.pm as a standard package
10:13:12< vasi> so ftw.info would need to go in base, and fink/$dist/
10:13:12< akh> We don' need no stinkin' apt!
10:13:27< akh> (other than to UseBinaryDist, of course)
10:13:38< dmacks> So when this new apt goes into fink/$dist/, libftw0 will have to as well (obviously), and also get added to Bootstrap.pm
10:13:47< RangerRick> yeah, it doesn't stop fink from functioning
10:14:00< vasi> do any bad things happen if you run fink without apt? i haven't tried that lately
10:14:13< dmacks> (And the whole point of "bootstrapping" is to build from scratch, not download precompiled binaries:)
10:17:20< vasi> dmacks, so is there anything you'd like changed about get_options? i could re-arrange stuff if needed
10:18:42< dmacks> vasi: What I was trying to do was to avoid needing an explicit entry for --no-foo if --foo is specified as 'foo!'
10:18:51< dmacks> RangerRick: Okay, it's in base now.
10:19:35< vasi> dmacks, er yeah unfortunately i'm not about to re-implement Getopt :-)
10:19:38< dmacks> ??
10:19:38< dmacks> It's a long-standing feature of Getopt.
10:19:52< vasi> yeah...you can still pass any Getopt option to get_options
10:20:09< dmacks> My concern is with the --help output.
10:20:38< vasi> right, it doesn't understand how to do help output for anything but the more basic cases
10:20:52< dmacks> If 'foo!', then it should automatically document that --no-foo is the opposite of --foo
10:21:22< vasi> that can happen
10:22:00< dmacks> That way we don't need to have an explicit --no-foo option declared.
10:22:02< vasi> (i just meant before that i'm not about to make get_options understand EVERY Getopt syntax, cuz that's overkill)
10:22:12< dmacks> Oh. Right:)
10:23:06< dmacks> The whole reason I was so insistent about using --no-foo as the opposite of --foo was exactly because it was automatically handled and we wouldn't need to write our own code for the negation flag.
10:24:26< vasi> we only have a single use of --no-something i think...maybe two?
10:24:36 * dmacks was about to add a new one:)
10:25:01< vasi> can you verify that --no-pedantic is the only one that we currently have?
10:25:12< dmacks> --no-use-bindist
10:25:52< dmacks> (The new one would be --no-log-output)
10:26:30< vasi> ok, so i'll hack on get_options to add support for that
10:26:37< dmacks> danke.
10:26:56 * RangerRick wonders if we could just use Getopt and figure out a way to get it to use something other than @ARGV  :)
10:27:59< RangerRick> or is that what you're doing?  it sounds like you keep talking about htings getopt does that get_options does not
10:29:06< dmacks> get_options takes both the Getopt parameters and also the help strings, and then handles --help automatically so we don't need to have a dedicated handler for --help in every Getopt call.
10:29:53< dmacks> (...from the caller's perspective). And also auto-formats the help screen so we don't need to juggle it for screen size and column widths.
10:30:27-!- shreyas [n=sshreyas@59.92.138.24] has joined #fink
10:31:06< RangerRick> ah, I see
10:34:57-!- chris01 [n=chris01@212.126.165.246] has quit ["bye"]
10:39:55-!- baba [n=baba@YahooBB220041000212.bbtec.net] has quit ["Leaving"]
10:41:23-!- shres [n=sshreyas@59.92.155.204] has quit [Connection timed out]
10:59:27< vasi> dmacks, do you know of any way to invoke the debugger from within a perl module?
11:01:03< kane_> vasi: you want the interactive debugger, or access to the debugger information?
11:01:51< vasi> the interactive debugger, like 'perl -d'....it would be nice to 'set a breakpoint' by just calling some function or something
11:02:04< vasi> cuz 'perl -d' doesn't really like modules
11:02:13< kane_> vasi: if you want a nice interface to that use Devel::ebug (on the CPAN)
11:02:57< kane_> and i happen to be working a module that does just that what you ask, but it's not done :) it's sadly not that easy, as the perl5db doesn't have a clean api to talk to
11:03:25< kane_> if you have +w to the file in question, dropping a $DB::single = 1 on that line will make the debugger break there
11:04:51< vasi> oh cool, that's exactly what i need
11:04:56< kane_> since it's a perl statement you can even say:
11:05:05< vasi> * if $something;
11:05:07< kane_> $DB::single = 1 if $var eq 'some content';
11:05:08< vasi> yeah :-)
11:05:11< kane_> :)
11:05:25< vasi> i'll take a look at ebug later
11:06:00< vasi> hopefully it's already in fink?
11:06:00< vasi> oh well, guess not
11:06:12< kane_> not sure about that.. some *other* code of mine is working on creating third party packages from CPAN packages
11:06:25< kane_> got debian packages working.. which means adapting it to fink sholdn't be rocket science (any more :)
11:06:30< vasi> woot
11:06:56< kane_> i poked dmacks about this a while ago, but according to him, not too many people (probably) used the fink perl ports
11:07:38< RangerRick> we really need to get the other 5% of getting popcon working
11:07:54< kane_> what's popcon?
11:08:16< RangerRick> popularity contest; people can submit what packages they have installed
11:08:24< kane_> ah.. as debian does :)
11:08:30< kane_> yeah, helps give focus on the projects at hand
11:08:47< kane_> perhaps obtaining download stats from fink mirrors would help
11:10:39< vasi> kane_, we share our mirrors with dports
11:11:08< vasi> dmacks, something's wrong with the chown'ing and build-as-nobody again
11:11:12< kane_> vasi: i'm sure a clever grep can do wonders there
11:11:15< vasi> cuz it's got apt-ftparchive whining
11:11:49< vasi> kane_, uh it's just 'curl $url' in both cases....no way to really distinguish between distros i don't think?
11:12:46< kane_> vasi: i'm not sure i understand.. does dports use the same (.deb) files as you? and even if they do, doesn't that prove even more these dists are popular? :)
11:13:11< vasi> kane_, well fink has both .debs and build-from-source
11:13:30< vasi> so for the .debs we could probably get statistics...but the source downloads are the same for us and dports
11:15:48< vasi> uh yeah, dmacks...we're still leaving symlinks owned by nobody!
11:19:08< vasi> dagnabbit
11:19:30< cirdan> morning
11:21:10< vasi> hi cirdan
11:21:40< akh> cirdan: your time and further west, that is.  ;-)
11:25:09< dmacks> symlinks have no meaningful ownership value.
11:25:31< vasi> dmacks, well it's meaningful to apt-ftparchive at least
11:25:36< dmacks> ?
11:25:53< vasi> and OS X certainly considers them to have SOME ownership value, whether or not it's meaningless
11:26:48< vasi> if you do 'chown -h root some-symlink', you get a different ownership according to 'ls -l', try it!
11:27:06< dmacks> Fals.
11:27:18< vasi> so if we don't chown symlinks, they remain owned by nobody...which make apt-ftparchive get confoozled
11:27:31< vasi> dmacks, you're on panther then, just do 'chown root some-symlink;
11:28:06< vasi> s/;/'/
11:28:08< vasi> that was the whole reason i was asking for 'chowname_hr', because Tiger and Panther have different behavior in their chown's
11:28:56< dmacks> Both with and without -h, /usr/sbin/chown alters the actual file not the symlink.
11:30:48< vasi> yuk, really?
11:30:53< vasi> (on panther?)
11:32:26< dmacks> Ayup. 10.3.9
11:32:27< dmacks> Disk is whatever was standard when I formatted it under 10.2
11:32:46< vasi> hmm, maybe i'll have to hack apt to deal with the issue
11:32:59 * vasi goes to look through apt's code again
11:44:28-!- cmeme [n=cmeme@boa.b9.com] has quit [Connection timed out]
11:48:29-!- cmeme [n=cmeme@boa.b9.com] has joined #fink
11:51:25 * cirdan is bored
11:51:40< RangerRick> vasi: look at the new one then  :)
11:51:58< vasi> RR, patience, i'm already doing three things at once
11:52:02< RangerRick> woot, got aptitude happy with the new apt  :)
11:52:15< RangerRick> vasi: just saying, patching an old apt just means coming up with the changes twice  :P
11:52:27< vasi> i know
11:53:03< cirdan> new dpkg!
11:53:03< cirdan> :-p
11:53:24< RangerRick> cirdan: got that too
11:53:24< RangerRick> test!  test!
11:54:35< cirdan> hehe
12:05:22-!- shreyas_ [n=sshreyas@59.92.145.90] has joined #fink
12:06:23-!- hennker [i=flullup@dslb-084-063-006-110.pools.arcor-ip.net] has joined #fink
12:08:26-!- sid77 [n=sid77@host-84-222-60-75.cust-adsl.tiscali.it] has joined #fink
12:15:26-!- shreyas [n=sshreyas@59.92.138.24] has quit [Read error: 110 (Connection timed out)]
12:20:25-!- JosephSpiros [i=foobar@ip-246-036.oberlin.net] has quit [Remote closed the connection]
12:20:40-!- JosephSpiros [i=foobar@ip-246-036.oberlin.net] has joined #fink
12:20:43-!- shreyas_ is now known as shres
12:25:31< dmacks> RangerRick: apt0.6.40-dev needs BuildDependsOnly flag. apt0.6.40-shlibs Shlibs field second entry has bogus compatibility_version
12:25:32-!- Fang [n=Fang@2002:56c2:e14c:1:0:0:0:1] has quit ["L'humanité est un cafard, la jeunesse est son ver blanc."]
12:27:58< dmacks> -shlibs needs a DocFiles
12:28:23< RangerRick> I thought splitoffs inherited docfiles if you didn't specify...
12:28:24< vasi> dmacks, it's ok to assume that Storable exists now, right?
12:28:35< vasi> RR, for some reason docfiles isn't inherited
12:28:37-!- cmeme [n=cmeme@boa.b9.com] has quit [Connection timed out]
12:28:40< dmacks> Not yet it isn't.
12:28:48< dmacks> vasi: Yes.
12:29:08< dmacks> None of the new apt* needs Depends:libiconv
12:29:31< dmacks> (though still need BDep:libiconv-dev)
12:29:46< vasi> k, the new magical get_options is done
12:29:50< vasi> go add your option
12:29:59-!- cmeme [n=cmeme@boa.b9.com] has joined #fink
12:30:03< vasi> (my initial design for get_options sucked :-( )
12:30:28< dmacks> new bin/apt-ftparchive is still linking db42 from /sw/lib/libdb-4.2.dylib
12:30:43< dmacks> Thanks vasi, will check it in few hrs.
12:31:17< RangerRick> dang
12:32:20< vasi> grrr, why do we build apt without -g ?
12:33:03< vasi> RR, if the new apt doesn't build with debug symbols, please add them :-)
12:33:13< RangerRick> vasi: ok
12:33:59< vasi> we really ought to have a policy on whether packages should have debug symbols or not
12:34:17-!- JosephSpiros [i=foobar@ip-246-036.oberlin.net] has quit ["leaving"]
12:38:04-!- JosephSpiros [i=foobar@ip-246-036.oberlin.net] has joined #fink
12:48:55-!- KraMer [n=mark@adsl-70-240-229-107.dsl.hstntx.swbell.net] has quit [Read error: 110 (Connection timed out)]
12:49:18-!- cmeme [n=cmeme@boa.b9.com] has quit [Connection timed out]
12:53:22-!- cmeme [n=cmeme@boa.b9.com] has joined #fink
12:54:14< akh> vasi:  make a separate distro for them.
12:55:10< vasi> apparently debian has 'foo-dbg' for many packages
12:55:18< vasi> no idea how they deal with the dependency issues
12:57:53< akh> Very carefully.  ;-)
13:00:14-!- KraMer [n=mark@adsl-70-240-229-107.dsl.hstntx.swbell.net] has joined #fink
13:10:26-!- shreyas [n=sshreyas@59.92.132.29] has joined #fink
13:12:35< dmacks> vasi: ! magic works great!
13:12:41< vasi> yay
13:13:11< dmacks> Guess I oughta actually implement the back end that the flag controls now, huh?
13:13:33< cirdan> heh
13:17:03-!- cmeme [n=cmeme@boa.b9.com] has quit [Read error: 110 (Connection timed out)]
13:18:07-!- cmeme [n=cmeme@boa.b9.com] has joined #fink
13:18:17-!- Albie [n=ambs@bl6-37-12.dsl.telepac.pt] has joined #fink
13:20:53-!- shres [n=sshreyas@59.92.145.90] has quit [Read error: 110 (Connection timed out)]
13:25:03< vasi> that would be an idea, dmacks :-)
13:26:16< cirdan> anyone know a windows util to turn a wireless card into an access point?
13:26:27< dmacks> vasi: What should I call the .pm that implements it? It's a tied handle that tees output. FilterHandle? HandleTee? 
13:26:56< dmacks> OutputTee?
13:27:05< vasi> er...what does it do?
13:27:09< cirdan> TeeAndCake!
13:27:10< dmacks> DansFuglyHackNumberBIGINT?
13:27:35< vasi> you could actually put an apostrophe in there...
13:27:44< vasi> since it's considered just like ::   :-)
13:28:04< dmacks> Yeah! In fact, why don't we backport *all* of fink core to perl4?
13:28:28< vasi> nah, port it to perl6....i'm sure it'll be useful in 2030
13:28:49< dmacks> (the apostrophe you mentioned is a perl4ism)
13:28:54< vasi> yeah i know :-)
13:29:08< vasi> so, wtf does your new module do?
13:29:32< vasi> you haven't told me anything except it has something to do with logging
13:29:40< dmacks> 'tie *STDOUT, "MyThing", "somefile", STDOUT'; and now STDOUT is teed into somefile.
13:30:05-!- lisppaste [n=lisppast@common-lisp.net] has quit ["Common Lisp IRC library - http://common-lisp.net/project/cl-irc"]
13:30:14-!- lisppaste [n=lisppast@common-lisp.net] has joined #fink
13:30:58< vasi> ah ok...
13:31:04< dmacks> (it's got a cleaner interface than that, but that's the internal so you can see that it's really a Tie::Handle at heart)
13:31:39< vasi> while you're working on I/O, think you could implement --really-quiet
13:31:46< vasi> (--im-hunting-wabbits ?)
13:32:34< akh> ...uhuhuhuhuh!
13:32:54< akh> (/me isn't sure how to Elmer Fudd-laugh phonetically)
13:33:24-!- lullabud [n=lullabud@64.193.227.60] has joined #fink
13:33:42< vasi> big philosphical problem, there
13:34:15< dmacks> heh
13:35:33-!- chris01 [n=chris01@84-73-54-203.dclient.hispeed.ch] has joined #fink
13:36:21< dmacks> OutputTee it is.
13:36:34< vasi> sounds good :-)
13:36:53< lullabud> I'm looking for a developer in the SF bay area who is familiar with socket layer programming for some ongoing contractual work.  Anybody interested?
13:37:05< cirdan> heh
13:37:20< cirdan> guess not
13:37:26< lullabud> *Mac developer, that is...
13:39:19-!- lullabud [n=lullabud@64.193.227.60] has left #fink ["Leaving"]
13:43:09 * akh wonders if people really think I can see their screens or read their log files.  ;-)
13:43:55< dmacks> Dunno...
13:44:28 * dmacks quickly iconifies quantum-physics.milfhunter.com
13:50:45-!- kahdgarxi [n=jdgray@216-160-216-64.phnx.qwest.net] has joined #fink
13:51:14< kahdgarxi> Hi. :)
13:52:03-!- kahdgarxi [n=jdgray@216-160-216-64.phnx.qwest.net] has quit [Client Quit]
13:52:29-!- kahdgarxi [n=jdgray@216-160-216-64.phnx.qwest.net] has joined #fink
13:53:12< kahdgarxi> I'm getting an error when compiling libgnomeui2-2.6.1.1-10; it can't fine the Xinerama library
13:53:17< kahdgarxi> I'm using X11.app
13:53:25< kahdgarxi> would that be the cause of my problem?
13:55:04< akh> Could be.
13:56:09< akh> Actually, technically, no.
13:56:35< kahdgarxi> So, how can i make libgnomeui compile?
13:56:37< kahdgarxi> force it to disable Xinerama support somehow?
13:56:37< akh> The problem is most likely that your install of Apple's X11 distribution (not X11.app, which is just the display server) is incomplete.
13:57:04< kahdgarxi> I see.
13:57:06-!- shreyas [n=sshreyas@59.92.132.29] has quit ["I am sleepy"]
13:57:06< akh> My X11.app package has Xinerama libs in it.
13:57:14< akh> (sorry X11USer)
13:57:22< akh> Try reinstalling X11User.pkg.
13:57:40< kahdgarxi> is that on the OSX DVD? (I'm on Tiger btw)
13:57:47< dmacks> (also X11SDK)
13:58:02< kahdgarxi> ok
13:58:06< akh> Yes.
13:59:51< kahdgarxi> Out of curiosity, which is generally 'recommended' by the fink devs, X11.app, or Xfree?
14:00:08< akh> X.org :-)
14:00:22< kahdgarxi> X.org is in fink?
14:00:25< kahdgarxi> I couldn't fine it x.x
14:00:52< dmacks> Yeah, Apple's or xorg's are the way to go.
14:00:53< kahdgarxi> alrihty
14:00:53< kahdgarxi> alrighty*
14:00:55< akh> Neither it, nor XFree86 are in stable on Tiger yet.  Hint, hint, to the maintainer...
14:01:10< kahdgarxi> hehehe :)
14:01:22< dmacks> Woohoo! Not me!
14:01:47< RangerRick> akh: I've been told of some weirdness with ABI stuff in the GLU bits I haven't been able to confirm nor deny, so I've been afraid of omving them to stable
14:02:05< akh> Ah.
14:02:06< RangerRick> but I'm not sure if it's worth waiting  :)
14:03:15< akh> The only really strange thing I heard of was that something in tetex* required a rebuild on Xorg--the bindist version didn't work.
14:07:16< kahdgarxi> So, how 'bout them dodgers?
14:07:25< dmacks> Dodgy.
14:07:38< kahdgarxi> hehe
14:07:39< akh> Wrong coast.
14:08:15< dmacks> Subway Series is out of the picture...maybe we'll have an Acela Series?
14:09:18< akh> So it'll have to stop for several months?
14:09:26< dmacks> ...again...
14:09:49< dmacks> Hey, it got back to work faster than the NHL.
14:10:37< kahdgarxi> I love watching make messages whiz by. :)
14:15:41< RangerRick> dmacks: I think you just put Copyright (c) 2005 in OutputTee.pm, chris didn't work on it in 2001  ;)
14:16:39< kahdgarxi> Thanks for the help ahk, libgnomeui compiled successfully. :)
14:16:39< dmacks> Ooh, good point:) I guess it's dmacks 2005 and Fink 2005-present?
14:16:43< RangerRick> dmacks: also, with the tee thing, how are you going to handle stderr?
14:17:02< dmacks> (was based on some stuff I wrote before getting involved in fink)
14:17:16< RangerRick> yeah, seems reasonable
14:17:24< akh> kahdgarxi: Glad it works.
14:17:27< dmacks> Both STDERR and STDOUT go to the file.
14:17:30< kahdgarxi> :)
14:18:11< RangerRick> dmacks: ok, the example showed just STDOUT, and I wasn't sure if you could tie multiple handles to a single output
14:18:48< dmacks> Yeah. The only clue I think is that the tee "appends" to a file. Look at the last few lines of PkgVersion.
14:19:06< RangerRick> was just looking at the tee thing itself
14:19:33< RangerRick> so that will even grab stdout/stderr of child processes?  I never understood how all that worked
14:20:52< dmacks> Yeah. When you system(), the spawned stdfoo are connected to the caller process's stdfoo.
14:22:11< kahdgarxi> what's the recommended method for setting/changing your gtk2 theme with fink?
14:22:22< dmacks> So we (PkgVersion) redirect stdfoo to a file, which means child processes' stdfoo goes to parent's, which goes to file.
14:22:40< RangerRick> kahdgarxi: same as any other OS, I assume, I don't know the mechanics of changing gtk themes
14:23:22< kahdgarxi> hmm...
14:24:18< RangerRick> but there wouldn't be anything osx-specific about it
14:25:16< kahdgarxi> the program i use in linux isn't in fink 
14:25:18< kahdgarxi> :(
14:26:13< dmacks> Wot program's that?
14:26:17< mcp> good evening
14:26:32< kahdgarxi> i don't remember the exact name of the app
14:26:33< mcp> anyone some knowledge on 10.4.2 and ipfw?
14:26:39< kahdgarxi> the executable is switch2
14:26:59< kahdgarxi> it's like gtk-theme-switch or something to that effect
14:27:51< kahdgarxi> other than that, the only way i know of to switch is with gnome-settings-manager
14:27:51< kahdgarxi> which i can't seem to find.. :(
14:31:58< dmacks> There's a theme-selector-capplet in control-center
14:32:29 * dmacks doesn't have a full gnome install either linux or fink to examine.
14:34:45< dmacks> We don't have a gtk-theme-switch package at this time.
14:34:45< vasi> dmacks, generally we just credit chrisp individually
14:34:48< vasi> (in the copyright in *.pm)
14:34:58< vasi> because we can't assume that he wished to sign his (C) over to Fink-the-team
14:35:47< dmacks> Okay...
14:36:24< vasi> (apparently it's considered "safer" to have the (c) for something under the team's name (or the FSF)...since that way if the GPL is ever invalidated, we can say "well we're changing the license to <whatever>")
14:37:04< vasi> (but if it was individual, we'd have to track down each individual)
14:37:16< vasi> kahdgarxi: i think you set something in your ~/.gtkrc-2.0
14:38:10< dmacks> Makes sense. I'm happy to leave its current form up to GPL and Fink, but it's based on some stuff that was not written for fink (random crap I played with a while ago on my own).
14:38:23< vasi> up to you who you want to assign the (c) to, since it's your code
14:38:23< dmacks> If Fink Team wants to cover that with GPL that's fine too:)
14:38:51< vasi> no reason you can't license the version in fink as GPL, and your own hacky stuff as something else
14:39:14< vasi> (the part in Fink *better* be GPL, otherwise we'll have problems!)
14:39:31< dmacks> Hmm...I guess since its current form is GPL and I'm placing my old-crap parts there, it doesn't matter where it came from so much.
14:40:08< vasi> btw, did we ever figure out the license-situation of .info/.patch files?
14:40:20< dmacks> RangerRick: So Notify* doesn't need chrisp's copyright either?
14:40:24< dmacks> vasi: Nope.
14:40:40-!- regeya [n=shane@adsl-sp3-cdale176.micgi.com] has quit ["Leaving"]
14:40:45< RangerRick> hm, it shouldn't, I wrote it
14:40:50< RangerRick> I must have copied too
14:41:18< vasi> we really should establish a policy on that asap...right now it's quite possible that a dev could say 'fink has no license to distribute my .info' and sue us
14:41:42< dmacks> <druggie teenager>I learned it by watching you!<holds up cigar-box of wacky perl>
14:41:52< vasi> heh
14:44:44< dmacks> My position has been "by contributing it to us to distribute, you're giving us permission to distribute it, why else would you be contributing it?" But that didn't stand up to -devel scrutiny so well.
14:45:15< vasi> yeah
14:45:50< vasi> cuz maybe the user meant we could distribute, but not modify?
14:46:39< dmacks> Or distribute for fink but not let others access to hand-build.
14:46:39< vasi> we have to make it explicit that either 1) they're assigning (c) to fink  or 2) they're licensing it under license <BLAH>
14:46:57< vasi> where <BLAH> could either be by dictat of -core, or one of a list of GPL-compatible licenses
14:49:12< vasi> probably the best way to do things would be to enforce that all .infos submitted to fink have a line 'InfoLicense: This .info file and any associated .patch file are licensed under <BLAH>'
14:49:23< vasi> s/submitted/commited to CVS/
14:49:36< dmacks> Also have to make sure people state that the thing is their own work or is covered by a license at least as permissive as [whatever we decide]
14:50:17< vasi> yeah, have them state the copyright holder too
14:51:17< vasi> anybody involved with fink who's very familiar with licensing, so we can vet this?
14:51:43< dmacks> It hella easier if they assign to us...otherwise would we have to track who submitted each change to Maintainer (or to some other -devel who committed it)?
14:52:37< dmacks> drm sounded like he knew what he was talking about:)
14:53:54< dmacks> WWDD?
14:54:04< akh> OTOH, maybe this would give us a weasel way out of dealing with people who try to build packages made by "some guy".
14:54:15< akh> (meaning not in Fink, but using our methods)
14:56:28< dmacks> Adding the new field would make it easier to track legacy pkgs.
14:57:19< vasi> yeah exactly
14:57:34< dmacks> Slap something on -devel?
14:57:52< dmacks> post this bit of #fink log on Wiki?
14:58:08< akh> Sure (to both)
14:59:25< dmacks> [If we email current maintainers to get permission, I bet we'll shake loose some broken email addies as well]
14:59:46 * dmacks still finds jswhit's old one sometimes.
15:00:10< dmacks> Okay, coffee'n'kinko's...bbl
15:00:11-!- dmacks [n=dmacks@pdpc/supporter/active/dmacks] has quit ["leaving"]
15:00:45< akh> coffees and copies :-)
15:15:02-!- akh [n=akhansen@ldx3.psfc.mit.edu] has quit []
15:21:36< vasi> RangerRick: i believe i have a patch for you
15:22:26< RangerRick> vas	ok
15:23:03< RangerRick> erp
15:23:27< RangerRick> stupid tab bug  :)
15:24:07< RangerRick> vasi: ok
15:25:24< vasi>  http://vasi.dyndns.org/files/fink/extracttar.cc.patch
15:25:30< vasi> do you intend the new apt to be finished and committed soonish, or will it still be awhile?
15:31:19< RangerRick> would like it to be soonish, just to get it out of my hair
15:32:49< RangerRick> hm, not getting a response from vasi.dyndns.org
15:32:52< vasi> oh woops, vasi.webhop.org
15:33:01< RangerRick> there we go
15:33:07< vasi> my ISP doesn't like port 80, gotta do auto-redirection :-/
15:35:58 * newmanbe would say something about that, but thinks that he probably doesn't need to. :)
15:37:57 * newmanbe rears e-mail from SF.net
15:38:08< newmanbe> s/ar/ad/
15:38:31< newmanbe> They "fixed" the MySQL performance problems.
15:44:18< RangerRick> vasi: didn't apply to 0.6.41 for some reason, but could be done by hand easily enough
15:46:48< vasi> yeah, it looks like UID is now Itm.UID
15:47:08< vasi> sorry 'bout that
15:47:31-!- Albie [n=ambs@bl6-37-12.dsl.telepac.pt] has quit ["Leaving"]
15:47:48-!- akh [n=akhansen@ldx3.psfc.mit.edu] has joined #fink
15:51:03< RangerRick> no biggie
15:51:19< RangerRick> just glad you caught it, I didn't change it in my patch  :)
15:51:22< RangerRick> gonna try building now
15:57:39-!- Snaggle [n=nieder@128.252.206.171] has joined #fink
15:58:21< Snaggle> has anyone used the latest xchat update (2.4.5)?  I can't get it to display channels in tabs (was working fine up to 2.4.4)
15:58:41-!- sid77 [n=sid77@host-84-222-60-75.cust-adsl.tiscali.it] has quit ["http://slackintosh.workaround.ch/"]
15:59:23-!- akh_x11 [n=akhansen@ldx3.psfc.mit.edu] has joined #fink
15:59:59-!- akh_x11 [n=akhansen@ldx3.psfc.mit.edu] has quit [Client Quit]
16:00:54< akh> Snaggle: I see the same thing.
16:00:58-!- akh_x11 [n=akhansen@ldx3.psfc.mit.edu] has joined #fink
16:01:30< akh> With a bunch of GARBAGE: messages
16:01:54-!- akh_x11 [n=akhansen@ldx3.psfc.mit.edu] has quit [Client Quit]
16:04:00< akh> I tried moving .xchat2 out of the way to see what happens--no such luck.
16:05:00-!- akh_2 [n=akhansen@ldx3.psfc.mit.edu] has joined #fink
16:05:00< Snaggle> ok. i've gone back to 2.4.4 for now.  can't find anything about this in the xchat forums so far
16:05:02< akh> I think I've got 2.4.5 on my Debian box and it works.
16:05:04-!- akh_2 [n=akhansen@ldx3.psfc.mit.edu] has quit [Client Quit]
16:05:42< akh> strange
16:08:17< cirdan> yo
16:11:13-!- Snaggle [n=nieder@128.252.206.171] has left #fink ["Leaving"]
16:15:03 * akh concentrates on my copy job so that I can close my laptop up immediately when it's done.
16:15:21< akh> but, yo cirdan
16:15:30-!- zizban [n=chris@24-52-0-219.sbtnvt.adelphia.net] has joined #fink
16:16:00 * akh hates catching the last rush-hour train.
16:16:08< akh> (or rather, missing)
16:16:23 * zizban tells akh to move
16:16:43< akh> Will zizban give me enough money for a house closer to Boston? ;-)
16:16:54< zizban> sure, anything for you bud :)
16:17:34< akh> heh.  I couldn't move too close, anyway, because then my wife would have a commute.
16:17:49< zizban> yup
16:18:42< akh> Ah--time to spare!  Laters.
16:18:45-!- akh [n=akhansen@ldx3.psfc.mit.edu] has quit []
16:32:57-!- mlopes [n=mlopes@195-23-159-245.net.novis.pt] has joined #fink
16:33:15< mlopes> hi. dia is available on the package list through the website, but apt-get install dia or fink install dia don't work
16:33:38< mlopes> the faq talks about this problem, but suggests using fink install as a solution.. which still doesn't find the package in my case
16:33:45< cirdan> !tell mlopes about unstable
16:33:53< newmanbe> Is it in unstable?
16:33:54< newmanbe> Do you have unstable enabled.
16:34:07< mlopes> i think it's in the unstable tree
16:35:53< mlopes> now, something not related to fink.. but does someone know how to change the keyboard behaviour to a default pc, like help key working as insert? :-)
16:36:28< cirdan> no
16:37:14< mlopes> or activating home key instead of command + left arrow :P
16:42:10< vasi> mlopes, i don't think all that is possible
16:42:18< vasi> there are just different conventions :-)
16:42:57< vasi> things would get confusing very quickly if you tried to translate between mac <-> windows conventions....eg: what does ctrl-c do in the terminal? :-)
16:45:44< mlopes> well, I meant mac os x/linux
16:45:44< mlopes> I don't use Windows.. for more than 5 years now
16:45:48< mlopes> well, I've never used ctrl + c either. select text and middle button works fine ;-)
16:46:07-!- chris01 [n=chris01@84-73-54-203.dclient.hispeed.ch] has quit ["Quitting!"]
16:48:24< mlopes> back ontopic.. well, fink install dia now works
16:48:34< mlopes> but it has lots of virtual dependecies that have nothing to do with dia!
16:48:44< mlopes> for example lynx..
16:49:16-!- dmacks [n=dmacks@pool-70-22-58-73.balt.east.verizon.net] has joined #fink
16:49:44< dmacks> *grr*...filehandles in system() suck even more than I imagined.
16:54:45< zizban> is that possible?
16:55:26< dmacks> I imagined I had created enough positive pressure to counteract the suck, but then I saw there was still some suck left.
16:55:36< zizban> heh
16:55:40-!- JesseW [n=chatzill@pdpc/supporter/student/JesseW] has joined #fink
17:01:06-!- spiral [n=spiral@laf31-2-82-224-249-43.fbx.proxad.net] has quit [Read error: 104 (Connection reset by peer)]
17:08:51-!- RangerRick is now known as RangerAway
17:10:17< mlopes> for some reason http://distfiles.dub.ie.eu.finkmirrors.net is down, so each time curl tries to run I have to wait ages.. how can I immediatly force curl to chose another mirror without killing the whole process?
17:10:54< dmacks> 'fink configure' will let you change your preferred-mirror selections.
17:13:49< mlopes> well, but if I kill the current process (135 dependencies), will I lose my current work through or will fink resume?
17:14:30-!- kane_ [n=kane@perl.xs4all.nl] has quit []
17:14:59< dmacks> Each package's compile is "atomic": you would lose a partially-compiled package, but a completely compiled package would be preserved.
17:15:34< mlopes> but I'm still retrieving packages from servers.. will fink download them again or just resume?
17:16:29< zizban> resume
17:16:32< zizban> usually
17:16:34< mlopes> great :-)
17:16:35< dmacks> Any completely-download file will be saved; a partially-downloaded file *might* be restartable where it'll canceled
17:16:54< dmacks> So at worst you lose one tarball's download time
17:16:58< mlopes> well, the packages are small so that doesn't matter
17:17:59< mlopes> sending an INT signal seems to also do fine, but I have to chose another mirror each time
17:18:56< dmacks> Yeah. Fink treats INT to curl is the same as if the curl process completed and failed naturally.
17:19:55< mlopes> :-)
17:24:39< vasi> dmacks, any objection to 'fink configure --mirrors', to only change the mirrors?
17:26:21< dmacks> vasi: Go for it.
17:26:21< zizban> good idea
17:26:22< vasi> wunderbar
17:31:06-!- lisppaste [n=lisppast@common-lisp.net] has quit ["Common Lisp IRC library - http://common-lisp.net/project/cl-irc"]
17:31:06-!- lisppaste [n=lisppast@common-lisp.net] has joined #fink
17:33:14< Murr>     what's our policy on ccache-default?
17:33:15< Murr> I have a package that's not compatible with it
17:33:22< Murr> is it the job of the package to list a conflict?
17:33:54< Murr> my understanding is that ccache breaks quite a few package
17:34:51< dmacks> Yeah, if you know it breaks it, add a BuildConflicts
17:37:42< dmacks> So a tie()ed STDOUT is not inherited into a system() but a reopened STDOUT is?
17:38:40< vasi> Murr, i use ccache-default and it's never to my knowledge broken anything
17:38:52< vasi> no that it's impossible or anything, just that i've never seen it happen
17:39:16< vasi> dmacks, system does a fork & exec
17:39:52< vasi> when the exec happens, there's no longer a perl runtime in the process
17:39:55< vasi> so what's going to make the tie-magic work?
17:39:55< vasi> (what's gonna execute the perl code that implements the tied handle?)
17:40:50< vasi> dmacks, is log_output ever actually used yet? or is it just sitting there waiting to be connected?
17:41:05< dmacks> Well, fileno(STDOUT) is still the same and I set cleared close-on-exec on it...
17:41:13-!- mlopes_ [n=mlopes@193.136.33.132] has joined #fink
17:41:59< vasi> dmacks, a tied filehandle is a feature of the perl runtime, which a reopened stdout is a feature of a unix process
17:42:17< vasi> so one of them continues when the perl runtime disappears, and the other doesn't
17:42:45< dmacks> Well crap then, I'm *not* gonna write a whole PerlIO: layer for this
17:43:00< vasi> what you probably would have to do is filter stdout, by turning system($blah) into open(BLAH, "$blah |")
17:43:13< vasi> and then send the results of that wherever
17:44:26-!- vasi is now known as vasiUnfasting
17:44:35< dmacks> Yeah, I've been desperately avoiding merging system()'s STDERR into its STDOUT in case someone uses those differently on the fink command-line.
17:44:36< dmacks> enjoy:)
17:48:04< cirdan> heh
17:51:27-!- mlopes [n=mlopes@195-23-159-245.net.novis.pt] has quit [Read error: 110 (Connection timed out)]
17:54:34< dmacks> ...and binding a filehandle to itself is also a bad idea. /me learns again.
18:05:37-!- hennker [i=flullup@dslb-084-063-006-110.pools.arcor-ip.net] has quit [Read error: 110 (Connection timed out)]
18:25:17-!- dmacks [n=dmacks@pdpc/supporter/active/dmacks] has quit ["leaving"]
18:46:40-!- mlopes_ [n=mlopes@193.136.33.132] has quit ["Leaving"]
19:03:14-!- kahdgarxi [n=jdgray@216-160-216-64.phnx.qwest.net] has quit []
19:12:54-!- JesseW is now known as JesseW_away
19:12:54-!- Feanor [n=astrange@mp1-248-166.dialup.emory.edu] has joined #fink
19:25:37-!- cmeme [n=cmeme@boa.b9.com] has quit [Connection timed out]
19:26:56-!- cmeme [n=cmeme@boa.b9.com] has joined #fink
19:31:32-!- zizban [n=chris@24-52-0-219.sbtnvt.adelphia.net] has quit ["Leaving"]
19:40:28-!- vasiUnfasting is now known as vasi
19:48:02< cirdan> ummm, why does fink not have lyx?
19:51:53< newmanbe> It does.
19:52:08< newmanbe> !tell cirdan about unstable
19:52:10< newmanbe> ;)
19:53:03< newmanbe> It know it used to anyway. *shrug*
19:53:47< newmanbe> And I found it!
19:56:09-!- akh [n=akhansen@68-118-244-23.dhcp.oxfr.ma.charter.com] has joined #fink
20:01:36< cirdan> it's not in my unstable
20:02:00< cirdan> /sw/fink/10.3/unstable/main/finkinfo/editors/lyx.infonot in 10.4
20:02:23< cirdan> not in 10.4
20:21:57< akh> The old lyx-qt needed gcc2.95 to build.
20:22:21< akh> (/me is too lazy to check the logs to see if that was mentioned.  ;-) )
20:31:35-!- dmacks [n=dmacks@pool-70-22-58-73.balt.east.verizon.net] has joined #fink
20:31:54< dmacks> vasi: --log-output now works. Correctly. Really.
20:32:10< vasi> yay
20:32:29 * vasi cvs up's
20:33:05< vasi> ah, cute strategy
20:33:08< dmacks> A --logfile (that takes %-exps) might be a useful addition, but that's easy enough to add now.
20:33:40< newmanbe> !lart MSDOS
20:33:46 * Melian dumps 42 tons of dirt, manure, and fish heads on MSDOS
20:33:46< newmanbe> s/$/ File System/
20:33:49< akh> ms dos:  what's that
20:33:51< akh> ?
20:33:51< akh> ;-)
20:33:52< vasi> newmanbe: FAT?
20:33:52< newmanbe> Why does the LCD have to be so terrible!
20:33:59< dmacks> Having to set $^F before open()ing is silly no?
20:34:00< newmanbe> Yes, Fat.
20:34:06< newmanbe> NetBSD calls it msdosfs. :)
20:34:40< newmanbe> s/fs//
20:35:09< akh> !lart NetBSD
20:35:09 * Melian blasts NetBSD with a huge firehose then strangles NetBSD with it
20:35:54< newmanbe> !netbsd
20:35:58< Melian> Of course it runs NetBSD, but does NetBSD run Fink?
20:35:58< vasi> i don't understand that code at all
20:36:02< newmanbe> dpkg runs I know.
20:36:09< newmanbe> Well, it's in the ports. :)
20:36:42< akh> vasi: dmacks may have to charge "consulting fees"
20:37:06< newmanbe> Mac OS X doesn't like it when you have different volumes with the same names.
20:37:40< akh> newmanbe: Yeah, that's probably not a good idea
20:37:56< dmacks> Only file descriptors <= $^F are available for direct access in spawned processes (i.e., 'open FOO, "| something >& _YOUR_NUMBER_HERE_"')
20:37:56< msachs> newmanbe: It's perfectly happy to mount them as /Volumes/Foo and /Volumes/Foo 1.
20:38:37< vasi> dmacks, when does fink access a spawned process?
20:38:52< vasi> (build as nobody, maybe?)
20:38:52< dmacks> "something"=="tee $logfile"
20:39:23< newmanbe> msachs: Tell that to the dead mounts. :)
20:39:30< dmacks> tee always filters STDOUT only, so I open STDERR to a tee, then redirect the output of tee to the original STDERR fileno
20:39:44< msachs> dead mounts: It's perfectly happy to mount them as /Volumes/Foo and /Volumes/Foo 1.
20:39:47< msachs> ;p
20:40:22< dmacks> And what matters isn't $^F when the process is spawned, but $^F when the file descriptor to be used in the spawned process is created.
20:40:23< vasi> dmacks, why are you saving fileno's anyhow, instead of filehandles?
20:40:31< newmanbe> And it gets the icons all screwy in Finder.app and friends.
20:41:14< dmacks> I'm actually saving handles. But both $^F and bash >& deal with fileno's so I'm getting sloppy with terminology:(
20:47:48< vasi> i mean...what's wrong with "open my $oldout, '>&', STDOUT"
20:47:50< vasi> ?
20:48:03< vasi> when do you have to access $oldout in a sub-process?
20:48:53< vasi> (changing $^F can be risky i thought, since you might accidentally close a FH you want to keep open that's below $^F...i could be confused though)
20:49:48< dmacks> Whoops, I should have made sure I wuold only raise $^F, not lower it.
20:50:20< dmacks> But my whole goal was to start and stop filtering the normal STD*. I don't want to have to adjust *every* print and warn and system call to send output to my tee filters.
20:51:25< vasi> right...you just save STDOUT with the 'open' i did earlier
20:51:27< vasi> and then re-open it to "| tee ..."
20:51:30< dmacks> Right...that's what I do.
20:51:41< vasi> and i don't understand what's the playing with $^F needed for
20:52:30< vasi> the current STDOUT is *always* below $^F
20:52:30< dmacks> I want to log STDERR but not merge STDERR into fink's STDOUT.
20:52:42< vasi> if you redirect STDOUT, it should still have #1
20:52:43< vasi> right, so you re-open OUT and ERR
20:52:44< vasi> still, they're both always below $^F
20:53:57< dmacks> Right. But my copies aren't. And I need to access the copies (so the STDERR filter goes to the original STDERR)
20:54:56< vasi> when do you have to access the copies in a sub-process?
20:55:29< dmacks> The whole problem is that the filter only operates on its STDOUT, so I need to push STDERR into STDOUT of the filter then push it back to STDERR of the caller. The "push it back" occurs in a sub-process (piped open)
20:56:23< vasi> eugh, ok
20:56:51< vasi> it still sounds to me like that's just redirection....
20:58:27< dmacks> Believe me, this wasn't the first thing I tried:) It was just the cleanest and simplest that worked...
20:59:30< vasi> can you be certain that raising $^F won't have any adverse effects?
21:00:04< dmacks> $^F only has an effect during open()
21:01:36< dmacks> At worst, I think some open files could be accessible to *Script. But I raise $^F only at the very beginning of the 'fink' command, so unlikely to have anything open (and we're pretty good about not keeping files open anyway).
21:01:36< vasi> or system, no?
21:01:39< vasi> anything that forks
21:01:42< dmacks> Right.
21:02:08< vasi> isn't $^F inherited though?
21:02:19< vasi> (ie, if something in a fork itself forks...)
21:03:41< vasi> hey, what's the entry-point to start logging, so i can test it out?
21:05:03< dmacks> $^F does not appear inherited through system()
21:05:27< dmacks> --log-output. Or you want the direct API?
21:05:47< vasi> alright...i know i'm being anal about this, but i've learned that messing with FH's can be High Voodoo
21:07:08< dmacks> No problem:)
21:07:10< dmacks> (hold on, HEAD's busted)
21:07:21< akh> ???
21:10:51< dmacks> I'm reliably getting $^F==5, meaning original STD* are 0,1,2; my copies are 4 and 5. I'll track down 3...
21:11:08< dmacks> HEAD's fixed.
21:12:08< lisppaste> vasi pasted "Break this!" at http://paste.lisp.org/display/12322
21:13:26< dmacks> If STDERR has gotten opened to something other than fileno 2...
21:14:45< dmacks> That's why I use "my dup of STDERR" not "fileno 2"
21:16:10< vasi> uh, STDERR is *always* 2
21:16:10< vasi> if you reopen it, it's still 2
21:16:27< vasi> try putting a 'print fileno(STDERR), "\n"' while the STDERR is reopened in my example
21:16:42< vasi> STDIN is always 0, and STDOUT is always 1
21:16:53< dmacks> In that case, no need to even stash $old*
21:17:31< vasi> you DO have to keep the old ones, because if you later try to reopen STDERR to 2, you'll be  re-opening it to itself (the NEW itself)
21:17:57< vasi> the way i use >&2 is kinda like saying $var = $var + 2
21:18:08< dmacks> (ah right)
21:18:42< dmacks> I got deep recursion when I tried that approach a few hrs ago...can't remember the exact details.
21:21:45< vasi> hmm really? weird
21:22:38 * dmacks tries to recreate...not having luck at doing so...
21:29:05-!- shres [n=sshreyas@59.92.128.70] has joined #fink
21:30:09-!- lisppaste [n=lisppast@common-lisp.net] has quit ["Common Lisp IRC library - http://common-lisp.net/project/cl-irc"]
21:30:15-!- lisppaste [n=lisppast@common-lisp.net] has joined #fink
21:42:48< dmacks> Okay, I'm only sorta-comfortable why that works correctly now, but it does.
21:42:52-!- JesseW_away [n=chatzill@pdpc/supporter/student/JesseW] has quit [Connection timed out]
21:43:03-!- akh [n=akhansen@68-118-244-23.dhcp.oxfr.ma.charter.com] has quit []
21:46:34-!- shres [n=sshreyas@59.92.128.70] has quit [Read error: 104 (Connection reset by peer)]
21:48:01< dmacks> committed, now with 20% less fileno silliness.
21:59:17-!- dmacks is now known as dmacks_away
22:00:29< vasi> great
22:18:10-!- dmacks_away [n=dmacks@pdpc/supporter/active/dmacks] has quit [Read error: 110 (Connection timed out)]
22:35:21-!- dmacks [n=dmacks@pdpc/supporter/active/dmacks] has joined #fink
22:47:01-!- dmacks [n=dmacks@pdpc/supporter/active/dmacks] has quit ["leaving"]
23:47:48-!- dmacks [n=dmacks@pool-70-22-58-73.balt.east.verizon.net] has joined #fink
23:48:20< dmacks> !seen drm
23:48:21< Melian> drm <n=drm@ip68-108-245-119.sb.sd.cox.net> was last seen on IRC in channel #fink, 9h 45m 52s ago, saying: 'see y'all later'.
23:57:16< dmacks> brendan: ping
--- Log closed Fri Oct 07 00:00:20 2005

.