--- Log opened Thu Jun 30 00:00:24 2005
--- Day changed Thu Jun 30 2005
00:00 < lisppaste> pogma pasted "bootstrapsucking" at http://paste.lisp.org/display/9526
00:01 < vasi> pogma, looks symptomatic of something earlier?
00:02 < pogma> vasi: dude those files are in /sw4/bootstrap
00:02 < vasi> oh yeah...ugh
00:02 -!- josh_mf [~josh_mf@covad-mozilla.meer.net] has quit []
00:02 < vasi> er, i dunno why it does that
00:03 < pogma> dunno either
00:04 < pogma> it should be going into /sw4/lib, odd
00:04 < vasi> i'm kinda in the middle of something
00:05 < pogma> okay, I'll look for 20 minutes before I have to go to work
00:06 < vasi> i'll take a look after if you can't find it
00:07 < pogma> totally doing ./configure --prefix=/sw4/bootstrap
00:07 < pogma> should just be /sw4
00:08 < vasi> is this phase 2 or 3?
00:08 < vasi> i think phase 2 does go into bootstrap, no?
00:09 < pogma> it is building debs
00:09 < vasi> phase 3
00:09 < pogma> I don't have the whole log, but the previous 3 packages go into %p
00:09 < vasi> pretty strange then
00:16 < dmacks> I wonder if set_percent_c should be moved into _set_destdirs?
00:17 < dmacks> (not sure it's relevant, but it seems like an intelligent change)
00:19 -!- msachs [~msachs@c-67-170-224-126.hsd1.ca.comcast.net] has joined #fink
00:20 < vasi> up to you dmacks
00:20 < dmacks> (OTOH, this does sound like a failure to reset %c after switching bootstrap mode?
00:20 < vasi> that's possible!
00:20 < vasi> in which case, DO move it
00:21  * dmacks wonders what the beginning of pogma's CompileScript showed...
00:21 < pogma> ./configure --prefix=/sw4/bootstrap
00:22 < dmacks> So that's wrong, right?
00:22 < pogma> right, wrong
00:22 < pogma> in installscript %d is correct, and %i is %d/sw4
00:23 < pogma> which is correct, just totally wrong in compilescript
00:23 < dmacks> Yup, %c gets uber-special treatment (b/c it has to handle conditionals)
00:25 < pogma> okay, fine, can you fix it, please :)
00:33 < pHatidic> is there a JDK for OS X?
00:36 < vasi> pHatidic, it comes with XCode
00:37 < vasi> which comes with OS X
00:37 < vasi> just install the Java stuff from your XCode disk
00:37 < pogma> dmacks: Fixed?
00:37 < pogma> :-p
00:37  * pogma goes to work
00:38 < dmacks> It's fairly intertwined...am working on it.
00:39 < dmacks> Acually, screw it...I was working on solving some corner cases, but we completely control what is a bootstrap so "if any -core does something wacky in a bootstrap pkg and breaks it, I'll kill you". There.
00:39  * msachs starts writing horsehead.info
00:40 < dmacks> hehe:)
00:40 -!- joshmoz [~joshmoz@covad-mozilla.meer.net] has joined #fink
00:42 < pHatidic> vasi: thanks
00:42 < vasi> np
00:43 < dmacks> pogma: Fixed.
00:43 < vasi> tnx dmacks
00:43 < dmacks> ...er...as soon as network lag lets the commit go through...
00:43 < vasi> fixed += tested?
00:44 < dmacks> Dude, I'm on a linux box.
00:54 < vasi> oh well
00:54 < vasi> so when are we getting LinFink? :-)
00:56 -!- Murr [~neeri@A17-202-20-71.apple.com] has joined #fink
01:00 < cirdan> hey Murr 
01:00 < dmacks> Friend of mine dated a woman Lynn Fink a while back.
01:00 < cirdan> haha
01:01 < dmacks> At some point should check through to find places where we hardcode platform strings.
01:07 -!- dmacks [~dmacks@dmacks.active.supporter.pdpc] has quit ["leaving"]
01:08 < cirdan> heh
01:08 < cirdan> my post got a +4 insitful on /. :-)
01:08 -!- xhrl [~ThomasW@24.80.39.250] has joined #fink
01:09 < cirdan> i should reply to a reply...but maybe i wont
01:09 < vasi> cirdan, link?
01:12 < htodd> woo
01:13 < cirdan> http://slashdot.org/comments.pl?sid=154333&threshold=1&commentsort=0&tid=126&mode=nested&cid=12945532
01:13 < cirdan> codemonkey has excellent points
01:13 < cirdan> other replies aren't so bad either
01:14 < htodd> oly 3 now
01:14 < cirdan> ?
01:18  * vasi goes to mod cirdan down (j/k)
01:20 < htodd> dunno. I just checked it and it was 3
01:21 < htodd> I don't know how the moderation system works
01:23 < htodd> (not my fault)
01:23 < cirdan> oh
01:23 < cirdan> weird
01:23 < cirdan> old cache maybe
01:23 < cirdan> sleep time
01:23 < cirdan> night
01:24 < htodd> see ya
02:26 -!- till [till@tilladam.kde] has joined #fink
02:31 < till> Morning guys, fink 0.24.7 on Tiger has trouble finding qt includes in /sw/include/qt/, configure fails when compiling arts.
02:32 < msachs> till: What's it say?
02:32 < till> Well, config.log shows it failing to find qtglobal.h and therefor the version check fails.
02:32 < till> Seems like the qt dir is in not in the include path.
02:33 < msachs> Which version of arts are you trying to build?
02:34 < till> 1.4.1-21, the one in unstable, apparently.
02:34 < till> Was pulled in from amarok.
02:34 < till> Which is what I'm trying to build, actually, hey muesli. ;)
02:35 < msachs> Do you have qt3 installed or just qt3-shlibs?
02:35 < till> qtglobal.h is there, if that's what you mean.
02:36 < till> qt3, qt3-dev, qt3-shlibs
02:37 < msachs> So is configure just not adding -I/sw/include/qt to its CFLAGS when looking for arts?
02:39 < till> that appears to be the case, yes.
02:40 < till> [08:24] <till> Seems like the qt dir is in not in the include path.
02:40 < till> :)
02:40 < msachs> heh, gotcha.
02:41 < msachs> RangerRick is the maintainer of that package, but he's RangerAway now.  Sounds like a job for him.
02:42 < till> Seems straight forward enough to fix, but I know too litle about how fink works, so I'll just hack my cflags locally.
02:42 < till> Just wanted to report.
02:43 < msachs> Okay, I'll pass that on, thanks.
02:43 -!- mbroeken [~broeken@hst32165.phys.uu.nl] has joined #fink
02:46 < till> Thank you.
02:46 -!- joshmoz [~joshmoz@covad-mozilla.meer.net] has quit []
03:13 -!- msachs [~msachs@c-67-170-224-126.hsd1.ca.comcast.net] has quit []
03:15 -!- Feanor [~astrange@feanor.developer.opendarwin] has quit [Read error: 110 (Connection timed out)]
03:18 -!- dmacks [~dmacks@netspace.org] has joined #fink
03:42 -!- dmacks [~dmacks@dmacks.active.supporter.pdpc] has quit ["leaving"]
03:48 -!- Murr [~neeri@A17-202-20-71.apple.com] has quit [Read error: 145 (Connection timed out)]
04:32 -!- BleedAway [whocares@saus04.usc.es] has quit ["changing servers"]
04:34 -!- joshmoz [~joshmoz@covad-mozilla.meer.net] has joined #fink
04:46 -!- BleedAway [~stickman@saus04.usc.es] has joined #fink
04:50 -!- _mcp [~hightower@wolk-project.de] has joined #fink
04:50 -!- mcp [~hightower@wolk-project.de] has quit [Connection reset by peer]
04:55 -!- dreamind [~dreamind@C2107.campino.wh.tu-darmstadt.de] has joined #fink
05:07 -!- mcp [~hightower@wolk-project.de] has joined #fink
05:07 -!- _mcp [~hightower@wolk-project.de] has quit [Connection reset by peer]
05:18 -!- mcp [~hightower@wolk-project.de] has quit [Read error: 113 (No route to host)]
05:25 -!- vasi [~vasi@modemcable214.145-70-69.mc.videotron.ca] has quit [Read error: 110 (Connection timed out)]
05:31 -!- mcp [~hightower@wolk-project.de] has joined #fink
05:41 -!- linuxmaniac [~maniac@199.Red-83-35-243.pooles.rima-tde.net] has joined #fink
05:41 -!- vasi [~vasi@modemcable214.145-70-69.mc.videotron.ca] has joined #fink
06:44 -!- dalibor [~Topic@p54A55A67.dip.t-dialin.net] has quit ["Leaving"]
06:52 -!- joshmoz [~joshmoz@covad-mozilla.meer.net] has quit []
06:54 -!- linuxmaniac [~maniac@199.Red-83-35-243.pooles.rima-tde.net] has quit [Read error: 113 (No route to host)]
07:16 -!- Feanor [~astrange@feanor.developer.opendarwin] has joined #fink
07:37 -!- linuxmaniac [~maniac@199.Red-83-35-243.pooles.rima-tde.net] has joined #fink
08:11 -!- linuxmaniac [~maniac@199.Red-83-35-243.pooles.rima-tde.net] has quit [Read error: 110 (Connection timed out)]
08:11 -!- linuxmaniac [~maniac@199.Red-83-35-243.pooles.rima-tde.net] has joined #fink
08:19 < mbroeken> hi Clef!
08:40 -!- zizban [~chris@24-52-0-219.sbtnvt.adelphia.net] has joined #fink
08:44  * pogma is sad, bootstrap still fails
08:45 < zizban> what are you trying to bootstrap?
08:56 < cirdan> fink
08:57 < cirdan> pogma: what are u trying, gcc4 only/
08:57 < cirdan> ?
08:58 < zizban> ah
09:01 < pogma> cirdan: someone with a developer transition system and a reason to need fink was here earlier
09:02 < cirdan> yes
09:02 < pogma> cirdan: fink fails to bootstrap everywhere
09:02 < pogma> cirdan: not just gcc4
09:02 < cirdan> ah, it wasn't just darwin x86 then
09:02 < pogma> darwinx86 has gcc3.3
09:02 < cirdan> aww, wtf...someone modded me as flamebait :-/
09:02 < pogma> gcc4 is the only supported compiler on intel
09:03 < cirdan> yeah
09:03 < zizban> that sucks
09:03 < zizban> both gcc ans flamebait
09:03 < vasi> grrr, iTMS never has the songs i want
09:04 < zizban> someone on /. was complaining that something wouldn't work with Linux. I pointed out that it worked it fine with solaris x86 and maybe it was his computer. got modded as flamebait
09:04 < vasi> really annoying when the USA iTMS has a song by a canadian group, that's NOT on the canadian iTMS
09:04 < zizban> idiots
09:04 < cirdan> vasi: what song?
09:04 < vasi> that was a week ago, i forget that one...i think it was a Sloan song?
09:05 < cirdan> oh
09:05 < vasi> but this time it's Postal Service - Such Great Heights
09:05 < zizban> ahhh licensing. I (heart) the riaa :/
09:06 < vasi> CRIA up here
09:06 < zizban> ah. probably same modis operandi
09:06 < vasi> but at least file sharing's sorta-mostly-legal here :-)
09:06 < cirdan> Crackhead Rec. Industry?
09:06 < cirdan> :-)
09:06 < vasi> (for now)
09:07 < cirdan> pogma: so...anything I can do to help with bootstrap?
09:07 < vasi> cirdan, turn dpkg.info in stable into dpkg-bootstrap.info, then put both dpkg*.info/patch in fink/10.4-transitional
09:08 < cirdan> 1if it was that simple, why does it make pogma sad?
09:09 < vasi> that's just for normal bootstrap, not g++4
09:09 < cirdan> thats what I was talking about
09:09 < zizban> curse apple intel! :P
09:11 < cirdan> so is it time to finally make a gcc4 tree?
09:12 < vasi> not much harm in *starting* one, but we'd like to get GCC: in all packages first
09:12 < cirdan> right
09:12 < vasi> and also start putting GCC: in debs
09:12 < vasi> er, and maybe decide on an upgrade policy
09:12 < cirdan> sounds almost like a plan :-p
09:13 < vasi> since currently any group of N devels has N+3 policies
09:13 < cirdan> vasi: can't we just grep mschas's buildlogs for [cg]++
09:14 < vasi> that's what drm did, probably mostly worked
09:14 < cirdan> or is that what's going on :-)
09:14 < cirdan> ok
09:14 < vasi> we could also use my gcc-field script on .debs
09:14 < cirdan> ok, now hows to having the gcc version in the deb help us?
09:14 < vasi> BUT i don't think we actually changed GCC: fields yet, drm just notified devels
09:15 < vasi> cirdan, it means that after an upgrade, it's pretty easy to detect packages with an old ABI
09:15 < cirdan> will dpkg/apt know not to consider mis-matched gcc fields when calculating deps?
09:15 < cirdan> (and fink)
09:15 < vasi> (or even NOT after an upgrade...ie: if some packages stay with 3.3 for awhile because they can't get 4.0 working, we can detect when they switch)
09:15 < vasi> cirdan, dmacks wanted to patch into the dep engine...i'm not sure if he still does, and i'm not convinced that's the right place to do it
09:16 < cirdan> well, since we let people use apt/dpkg directly we'd need to patch them also
09:17 < vasi> cirdan, i don't think we'll be using it as a restriction on installing
09:17 < cirdan> oh
09:17 < vasi> more for post-hoc detection (on dist-upgrade?) or for building
09:17 < vasi> if we want to restrict installs, we'd probably be better off using implict Provides: foo-gcc40 and Depends: foo-gcc40 or something
09:18 < cirdan> hrm
09:18 < vasi> you can read the log of dmacks and myself discussing a couple days ago....if you can handle both of us strongly disliking your dummy-packages plan :-)
09:19 < zizban> heh
09:19 < cirdan> heh
09:20 < cirdan> i dont really care too much, it just seemed to be thae best way to make sure no gcc3.3 libs were left behind to cause build failures
09:20 < cirdan> but, with perfect versioned depends, it shouldn't be a problem
09:20 < cirdan> other than users complaining that instead of foo not being found, foo crashes with a bus error ;-)
09:21 < vasi> we gotta arrange to get all of us in #fink at the same time so we can duke it out :-)
09:21 < cirdan> hehehe
09:21 < zizban> let's get ready to ruuuuumble!
09:21 < cirdan> round 1: *ding!*
09:21 < vasi> well i LIKE the idea of having a dist-up of some sort that enforces update-all
09:21 < vasi> or at least asking the user to update-all and telling them if they don't, they should expect trouble
09:22 < vasi> i just really dislike the idea of using fake, empty packages to accomplish that
09:22 < zizban> and not to come on here and whine about it :P
09:22 < cirdan> vasi: i made dist-up as a framework for us to use when we need to make weird changes on system/dist upgrades
09:22 < vasi> tell them to email gnome-core ;-)
09:22 < zizban> heh
09:22 < vasi> cirdan, yes that part is good...well, theoretically, i'll check the code eventually
09:22 < cirdan> changing trees line, removing old dev tools, etc.
09:23 < vasi> yah
09:23 < vasi> but it's the ABI changing part that's totally up in the air atm
09:23 < cirdan> we need to make a map of old system/tree to new system/tree and the subs that need to be called for each upgrade
09:24 < cirdan> going from 10.3/foo might have different steps than going from 10.4/trans -> 10.4
09:24 < cirdan> like dev tools won't need to be purged
09:26 < vasi> hmm, one kinda simple option for the ABI change is to just 1) not do anything super-special, just bump revisions  2) politely tell the user to update-all   3) when fink is started, check if there's any GCC: 4.0 packages that depend on GCC: 3.3 packages or vice versa, and warn the user that they really should get around to update-all already or @packages will break
09:26 < vasi> that way we don't really need to bump the revisions in Depends so much
09:26 < zizban> I like #3
09:26 < vasi> er, those are all part of the same thing zizban :-)
09:26 < cirdan> vasi: won't work
09:26 < zizban> oh
09:26 < cirdan> :-(
09:27 < vasi> cirdan, which way?
09:27 -!- You're now known as RangerRick
09:27 < cirdan> we must force the order of building
09:27 < vasi> apt and dpkg we can patch into if we want
09:27 < cirdan> which we can only do with versioned deps
09:27 < vasi> ah...well we can have update-all force the order based on GCC: instead
09:27 < cirdan> still not good enough
09:28 < cirdan> :-)
09:28 < vasi> why not?
09:28 < cirdan> fink builds a-z normally
09:28 < vasi> just say "if one of our deps has the wrong GCC, then push this build to the back of the list"
09:28 < zizban> cirdsn is a tough taksmaster
09:28 < cirdan> if we push gcc3 packages to the front good
09:28 < vasi> er, that's pretty much the effect
09:28 < cirdan> but abba depends on c++ libs 0ace and zeta
09:29 < RangerRick> abba depends on drum machines
09:29 < vasi> an eurovision
09:29 < zizban> and bell bottoms
09:29 < cirdan> 0ace gets built ok, then abba fails because zeta isn't upgraded yet
09:30 < RangerRick> any time one of abba's deps is not the latest version, it pushes itself to the end of the stack
09:30  * zizban wonders what the next beos clone has to do with all this :)
09:30 < vasi> 0ace gets built, then fink gets to abba and says "hmm, it depends on zeta, but abba and installed zeta don't have matching GCC:, better postpone it"
09:30 < vasi> exactly RR
09:30 < cirdan> and if a user *really* needs abba to work, fink install abba will always fail to buil, unless they stumble on the fact that zeta needs to be updated first
09:30 < RangerRick> not if fink checks zeta's abi
09:30 < cirdan> but then our depends lines are wrong
09:31 < cirdan> we might as well abandon any versioned deps in fink at all if we go that route
09:31 < RangerRick> well, yeah, that's true... means it's broken for apt then
09:31 < cirdan> seriously broken
09:31 < RangerRick> maybe spit out a warning that the maintainer should be e-mailed
09:32 < zizban> oh goody
09:32 < zizban> :)
09:32 < RangerRick> hehe
09:32 < RangerRick> I can't believe I both maintain KDE, and made that suggestion  ;)
09:32 < vasi> RR you agreed with pogma earlier that we simply let ppl upgrade as-is
09:32 < zizban> me either :)
09:33 < vasi> and deal with it when 'fink update some-dependency' breaks 'fink update whatever-depends-on-it'
09:33 < cirdan> or have a validate script that parses the {Build}Depends lines and yells if known g++ packages are depended on and not versioned
09:33 < RangerRick> true, but that doesn't mean we can't warn that ultimately, the deps should be updated to depend on the new version, at least in validate
09:33 < vasi> which is MUCH more broken
09:33 -!- Feanor [~astrange@feanor.developer.opendarwin] has quit [Read error: 110 (Connection timed out)]
09:33 < cirdan> RangerRick: hehe
09:33 < RangerRick> maybe have validate check against the version that's in the bindist?
09:33 < vasi> dammit gnome sucks
09:34 < RangerRick> otherwise validate will mostly pass, 'cause developer tend to be very up-to-date  :)
09:34 < RangerRick> we'll just require all end-users to run validate!
09:34  * RangerRick solves the problem!
09:34 < vasi> the whole re-bumping and forcing versioned depends is ridiculously difficult
09:34 < vasi> rev-bumping
09:34 < cirdan> well, yeah
09:34 < RangerRick> vasi: true, but I'm talking more about a leisurely, as-you-find-it approach, rather than a requirement for release  :)
09:35 < cirdan> but it's the RightWay :-)
09:35 < RangerRick> it will work for people who update-all
09:35 < RangerRick> and it will eventually work more and more for people who don't  ;)
09:35 < vasi> oy
09:35 < cirdan> vasi: i say, anyone who wants to take on using the final 10.4 tree (intel users esp.) need to help out
09:35 < vasi> why not just disallow upgrades?
09:35 < RangerRick> hah!
09:35 < zizban> no upgrade for you!
09:36 < cirdan> no packages are moved unless a user fixes the deps and adds it
09:36 < vasi> have an upgrade script that just 1) moves %p to %p.bak  2) installs a new %p  3) copies over all non-GCC-field debs
09:36 < vasi> 4) does 'fink install everything-that-used-to-be-installed'
09:36 < cirdan> vasi: heh, yeah, well, dist-upgrade could do that too...
09:36 < vasi> tada!
09:36 < cirdan> it'd juck real hard though
09:37 < cirdan> vasi: and the users w/ 6G drives and 500M free space? :-)
09:37 < RangerRick> I think I read that as "fucktard"
09:37 < cirdan> RangerRick: sure, same thing ;)
09:37 < vasi> heh
09:37 < vasi> i guess let the user either keep or delete the old %p
09:37 < cirdan> btw, juck = just + suck
09:37 < zizban> cool
09:38 < cirdan> naa, we gotta hardlink 100k files
09:38 < vasi> ftp://ftp.gnome.org/pub/GNOME/sources/atk/1.9/atk-1.10.1.tar.bz2   something looks wrong with that
09:39 < RangerRick> heh
09:39 < cirdan> heh
09:40 < vasi> tsk tsk dmacks
09:40 < pogma> vasi: bootstrap is not breaking on dpkg, I have dpkg working
09:41 < RangerRick> he already had the tarball on his drive, how was he to know?  :)
09:41 < vasi> where's it borked now?
09:41 < pogma> vasi: still gettext, dmacks' fix didn't
09:41 < vasi> pogma, argh really?
09:41 < vasi> i'll look into it, but later
09:41 < vasi> sorry :-(
09:41 < pogma> no problem
09:42 < pogma> I'll go back to beer and dvd now :-)
09:42 < cirdan> hehe
09:44 < vasi> wait a sec, i think i've got it
09:45 < vasi> hmm, or not
09:46 < zizban> heh
09:46 < zizban> bbl
09:46 -!- zizban [~chris@24-52-0-219.sbtnvt.adelphia.net] has quit ["Leaving"]
09:47 -!- dreamind [~dreamind@C2107.campino.wh.tu-darmstadt.de] has quit []
10:02 -!- mbroeken [~broeken@hst32165.phys.uu.nl] has quit []
10:04 -!- baba [~baba@YahooBB220041001026.bbtec.net] has joined #fink
10:16 -!- akh_laptop [~akhansen@jove.psfc.mit.edu] has joined #fink
10:21 -!- akh_laptop is now known as akh
10:21 < akh> !lart gettext-dev for thinking it's essential.
10:21  * Melian strangles gettext-dev with a 9-pole serial cable for thinking it's essential.
10:22 < akh> mmmm....grammar parser.  ;-)
10:23 < pogma> that's a grammatically correct sentence
10:25 < akh> Yeah, I know.  ;-)
10:25 < akh> Just remembering the old days.  :-)
10:26 < akh> Actually, I guess this one would have been OK even when the lart text was left unparsed.
10:26  * RangerRick made a country on nationstates.net http://www.nationstates.net/cgi-bin/index.cgi/target=display_nation/nation=finkatar
10:26 < till> RangerRick: got a minute?
10:27 < akh> (Melian: strangles gettext-dev for thinking it's essentia with a 9-pole serial cable) 
10:27 < RangerRick> till: sure
10:27 < till> RangerRick: I'm in the process of getting fink and kde up on my mac mini, as part of the "we need a Kolab client for Mac OS X" whining of my boss and the "I want amarok back" whining of my wife management. :)
10:28 < till> RangerRick: so I hooked myself up with unstable, no problems, building arts fails, though, due to it no picking up the qtglobal.h in/sw/include/qt, since that dir is no in the includes, it seems.
10:29 < till> RangerRick: My bad, or is there something missing in the build environment?
10:29 < RangerRick> till: I think it's a missing builddep on one of the autotools, thanks for the reminder  :)
10:29 < RangerRick> just a sec
10:30 < RangerRick> hrm, no, it's there now
10:30 < RangerRick> could you lisppaste the whole output?
10:30 < RangerRick> lisppaste: url
10:30 < lisppaste> To use the lisppaste bot, visit http://paste.lisp.org/new/fink and enter your paste.
10:30 < till> RangerRick: in a minute, my wife insist on using the mac for her petty work concerns, atm.
10:30 < RangerRick> =)
10:31 < till> RangerRick: do you want config.log, or the fink build arts output?
10:32 < RangerRick> might as well do both, but the build output first  :)
10:32 < till> Ok. Will be a few minutes.
10:35 -!- ringerc [~craig@dsl-202-72-144-62.wa.westnet.com.au] has joined #fink
10:41 < akh> Thriving economy?
10:42 < akh> I guess we're not running a deficit...
10:44 < cirdan> yes we are...
10:44 < akh> Ah--but we're not running a multibillion dollar deficit.
10:45 < cirdan> yes we are..
10:45 -!- jtyler [~jtyler@iphost-64-56-130-194.edm.wiband.net] has joined #fink
10:45 < cirdan> ;-)
10:46 < akh> Where'd the money go?  Clearly not to bindist servers.  :-P
10:46 < cirdan> clearly...
10:51 < akh> That'd be a heck of a lot of "entertainment", too.  ;-)
10:52 -!- linuxmaniac [~maniac@199.Red-83-35-243.pooles.rima-tde.net] has left #fink ["Leaving"]
10:53 < lisppaste> till pasted "arts build failure" at http://paste.lisp.org/display/9533
10:53 < till> RangerRick: there you go
10:54 -!- Feanor [~astrange@feanor.developer.opendarwin] has joined #fink
10:54 -!- zizban [~chris@24-52-0-219.sbtnvt.adelphia.net] has joined #fink
10:54 -!- akh [~akhansen@jove.psfc.mit.edu] has quit [Remote closed the connection]
10:54 < RangerRick> weird, yeah, what's in config.log?
10:54 < RangerRick> also, what does "fink list qt" print?
10:56 < till> Coming up
10:56 -!- akh [~akhansen@jove.psfc.mit.edu] has joined #fink
10:56 < till> RangerRick: config log says that QT_VERSION is too low, since it can't find qlobal.h, since -I/sw/include/qt is nowhere in sight. :)
10:57 < till> RangerRick: qt3, qt3-dev and friends are installed, of course.
10:57 < pogma> till: I think he wants you to lisppaste the config.log
10:57 < RangerRick> that's weird, since the configure should be passing --with-qt-includes and such
10:58 < lisppaste> till annotated #9533 with "list output" at http://paste.lisp.org/display/9533#1
10:58 < till> I'll lisppaste he whole log, hold on.
11:00 -!- baba [~baba@YahooBB220041001026.bbtec.net] has quit ["Leaving"]
11:00 < till> arf, cut buffer doesn't cope with that.
11:00 < RangerRick> hehe
11:00 < till> RangerRick: want me to mail it?
11:00 < RangerRick> just the relevant bits is fine
11:00 < till> Ok, sec.
11:01 < lisppaste> till annotated #9533 with "config.log snippet" at http://paste.lisp.org/display/9533#2
11:02 < RangerRick> I can augment the info file to do it, I just wonder why it's not working
11:02 < till> Me too.
11:03 < RangerRick> oh well, doesn't hurt to add it to CPPFLAGS  :)
11:05 < till> Thank you.
11:09 < pogma> checkout from May15th bootstraps
11:09 < till> RangerRick: Righto, so what do I need to do to get the fix, simply try building again?
11:19 < runelind> gah.....gaim....loading....soooo...slow!
11:20 < pogma> love the dlsym()
11:20 < zizban> heh
11:21 < zizban> try loading mozilla on soalris x86...woah its sloooooooooooooooow
11:21 < zizban> soalris even
11:21 < zizban> solaris
11:21 < zizban> damn it
11:22 < RangerRick> till: selfupdate in the next hour or so and you should get the fix
11:23 < runelind> RangerRick: did you say that the slow gaim loading was due to glib?
11:23 < till> RangerRick: Thanks.
11:24 -!- pixelfairy [~pixel@dsl081-036-140.lax1.dsl.speakeasy.net] has joined #fink
11:25 < pixelfairy> got minicom working
11:25 < pixelfairy> but it takes a lil work
11:26 < zizban> wow, cool
11:28 < pogma> what's with this crap in base-files? "if [ `uname -r | cut -d. -f1` -gt 7 ]; then"... for manpath stuff. I thought manpath stuff changed in darwin7, not darwin 8??
11:29 < pogma> should that have been -gt 6 ?
11:39 -!- Fang [~Fang@2002:56c1:6039:1:0:0:0:1] has joined #fink
12:00 < cirdan> pixelfairy: what did u have to do?
12:00 < pixelfairy> im working on a how to
12:00 < pixelfairy> http://pixelfairy.org/quickies/minicom-osx.html
12:00 < cirdan> cause i didnt have the problem you did :-p
12:00 < pixelfairy> then im going to mail the maintainer
12:00 < cirdan> i'll be waiting :-)
12:00 < pixelfairy> oh ok
12:00  * akh thinks we need a more appropriate flag for Finkatar
12:00 < pixelfairy> well you already know the other details
12:01 < pixelfairy> really, such a howto belongs on your site
12:01  * zizban is so glad he gets zero emails about his package
12:01 < cirdan> pixelfairy: i have no troubles runnin sudo minicom -s
12:01 < pixelfairy> cirdan: did you set up minicom with minicom -s?
12:01 < cirdan> yes
12:01 < pixelfairy> oh
12:01 < akh> zizban:  nobody wants to comment on your package.
12:01 < pixelfairy> him...
12:02 < zizban> akh: ha! :)
12:02 < pixelfairy> i am going to uninstall and reinstall a couple times so ill try that
12:02 < cirdan> which is why i couldn't figure out how yours want working :-)
12:02 < pixelfairy> do you run it as root?
12:02 < cirdan> yes, or as a user
12:02 < pixelfairy> and it doesnt matter?
12:02 < cirdan> but lockdir needs to be changed as a user
12:02 < pixelfairy> as a user it tells me it cant make the lock fire
12:02 < pixelfairy> oh
12:02 < pixelfairy> ok
12:02 < pixelfairy> so thats another way
12:02 < cirdan> no, i get no such device :-)
12:03 < cirdan> it's very off
12:03 < cirdan> odd
12:03 < pixelfairy> yes, thats why im going to try a couple more times
12:03 < pixelfairy> and see what happens
12:04 < pixelfairy> the web is full of requests of people for a serial console on os x
12:04 < pixelfairy> so its not just us two
12:06 < vasi> pogma, i think i have the bootstrap issue figured out
12:06 < zizban> mac addict had a how to about a serial console years ago
12:07 < pixelfairy> im glad dillo is in fink, i like running the x server in full screen
12:07 < zizban> mozilla is there too
12:07 < zizban> I just installed it the other day for kicks
12:07 < vasi> firefox also
12:07 < vasi> and galeon
12:07 < pixelfairy> oh cool
12:07 < vasi> epiphany
12:07 < zizban> oh ya firefox
12:07 < vasi> konqueror
12:07 < pixelfairy> i like epiphany too
12:07 < zizban> I love Konqueror
12:08 < vasi> lynx :-)
12:08 < pixelfairy> already have that
12:08 < zizban> I'd love RangerRick to update his KDE/<ac port
12:08 < zizban> we all have dreams :)
12:08 < vasi> pssst, test galeon and then send me an "it works!" report :-)
12:08 < vasi> zizban, do it yerself :-P
12:09 < zizban> vasi: sure apt-get install rangerricksbrain
12:09 < zizban> oooo I'd love to break Galeon, I mean, test for you :)
12:09 < pixelfairy> as long as were dreaming i want the serial port to work in darwin in qemu
12:10 < zizban> hmmm lemme guess, galeon is in unstable?
12:10 < akh> vasi:  just a sec...
12:10 < vasi> yep zizban
12:10  * akh has it on my apt sever
12:11 < zizban> d'oh! okay I'll switch to unstable
12:11 < akh> I'll install it here.
12:11 < pixelfairy> is all that in unstable?
12:11 < pixelfairy> dillo and lynx were the only web browsers i saw
12:11 < vasi> oooh yummy akh, i wasn't aware anybody had actually used galeon 
12:11 < pixelfairy> vasi: i often use it in linux
12:11 < zizban> Mozill and Konqueror are in stable, that I know
12:11 < vasi> more like a masochistic exercise for me, maintaining a package nobody uses
12:11 < vasi> i mean in fink, pixelfairy
12:12 < vasi> and yeah, most stuff is in unstable
12:12 < vasi> but unstable isn't really that unstable
12:13 < pixelfairy> found mozilla
12:13 < pixelfairy> and crossfire while looking for firefox
12:13 < pixelfairy> that might be fun...
12:13 < zizban> whats crossfire?
12:13 < pixelfairy> apperantly a game
12:13 < pixelfairy> i think its a mud
12:13 < zizban> ahhhh
12:13 < pixelfairy> yes, its a mud
12:13 < pixelfairy> now i remember playing it
12:14 < vasi> pogma, gettext bootstrapping bug fixed!
12:14 < zizban> unstable enabled, waiting for update
12:15 < vasi> try it before I break things again :-)
12:15 < zizban> I can't code so this is how you can help fink, btw, by testing stuff for developers/maintainers
12:16 < cirdan> vasi: i say, while we're at it, get rid of old gettext fron bootstrap
12:16 < cirdan> put it in libs
12:16 < cirdan> require stuff to use new gettext in the new tree
12:16 < vasi> cirdan, bring it up with devel
12:17 < cirdan> but at least rm it from bootstrap
12:17 < vasi> as they say in french, c'est pas ma job :-)
12:17 < cirdan> heh
12:17 < zizban> vasi: this will be a while, I'll send a report
12:18 < akh> vasi:  works, other than that stupid crypto linking crap.
12:18 < vasi> zizban, cool thanks
12:18 < vasi> akh, heh
12:18 < zizban> np
12:18 < vasi> wonderful that
12:18 < vasi> k, i'm off now...take care
12:18 -!- vasi [~vasi@modemcable214.145-70-69.mc.videotron.ca] has quit []
12:19 < pixelfairy> how hard was porting firefox?
12:20 < akh> Probably no worse than mozilla
12:20 -!- emp__ [~emp@70.57.239.38] has quit [Read error: 104 (Connection reset by peer)]
12:21 -!- emp__ [~emp@70.57.239.38] has joined #fink
12:21 -!- Feanor [~astrange@feanor.developer.opendarwin] has quit []
12:30 < RangerRick> runelind: well, that's the most likely culprit anyways
12:31  * RangerRick despairs that he has to make koffice -j1  :P
12:31 < RangerRick> slooooooow
12:31 < cirdan> lol
12:31 < cirdan> sucks
12:32 < cirdan> why?
12:32 < cirdan> make 'em fix it
12:32 < cirdan> !
12:32 -!- msachs [~msachs@17.255.96.116] has joined #fink
12:33 < akh> RangerRick: ugh
12:35 < RangerRick> I think it's more tiger's linker
12:35 < RangerRick> or gcc
12:35 < RangerRick> which somehow just kind of.... freezes up
12:35 < RangerRick> or a bug in distcc/ccache
12:35 < RangerRick> which I suppose is more likely
12:35 < RangerRick> it just gets deadlocked
12:35 < RangerRick> if I stop a compile and then start it again, the same 3 files that were building and stopped before continue just fine
12:38 < akh> Maybe you just need to BuildConflict ccache then.
12:38 < akh> ;-)
12:39 < RangerRick> hah
12:39 < RangerRick> maybe you just need to shut up!
12:39  * RangerRick looks at akh menacingly
12:40  * akh whistles innocently.
12:41 < msachs> RangerRick: Did you see the thing from last night about KDE not finding arts?
12:42 < pixelfairy> crossfire seems to just be broken
12:42 < RangerRick> msachs: nope, no one must have RangerAway:'d
12:43 < msachs> I did, sans the colon.
12:43 < RangerRick> at least, my login highlight didn't catch it
12:43 < RangerRick> that or I got disconnected, which happens sometimes
12:45 < msachs> "fink 0.24.7 on Tiger has trouble finding qt includes in /sw/include/qt/, configure fails when compiling arts (1.4.1-21)."
12:46 < akh> There's a mailing list message in a similar vein.
12:46 < akh> wait-nope.  /me confused myself.
12:47  * akh was conflating with till's lisppaste.
12:54 < cirdan> hey msachs 
12:54 < msachs> hey cirdan
12:54 < msachs> How's it going?
12:55 < cirdan> allright
12:55 < cirdan> we've been trying to get an upgrade plan going to 10.4 tree
12:55 < msachs> Yeah?
12:55 < cirdan> vasi had some good ideas, but i shot holes in most of 'em
12:56 < cirdan> our policies are not too bad, but in many cases we dont enforce them
12:56 < msachs> What are the issues with the transition?
12:57 < cirdan> my suggestions are either dummy packages (maybe) or just explicit versioned deps
12:57 < cirdan> msachs: c++ abi code
12:57 < cirdan> also our depends policies are not enforced
12:57 < msachs> Right, what about it?  I mean is it just figuring out how to make sure we never try to link a C++ 10.4-trans package with a C++ 10.4 package?
12:58 < cirdan> well, one suggestion is to just move everyhting to 10.4 and tell ppl to update-all
12:58 < cirdan> and maybe push the g++ marked packages to the end of the update all
12:58 < cirdan> will kinda work, but depends will be all wrong
12:58 < cirdan> for shlibs anyway
12:59 < cirdan> leat amount of work for the maintainer though
12:59 < msachs> Mmhm.
12:59 < cirdan> my suggestion is to put somehting in the start of the rev to mark either the os or the tree, then the rev
12:59 < cirdan> so it's easy to tell where a .info belongs
12:59 < cirdan> *but* no more exact copying info files between trees
13:00 < msachs> That'd be annoying.
13:00 < cirdan> which is a good thing, if we wanna have strict complience with versions
13:00 < msachs> Going forward, we could put a field in build packages saying what dist was used to build them.
13:00 < cirdan> it's possible to have a deb built on 10.3 and 10.4 to have the same version and revision, and the same deps, but much different actual deps
13:01 < msachs> True.
13:01 < cirdan> so, in order to comply with our own policies, we are not allowed to do what we are doing anyway
13:01 < msachs> heh
13:01 < cirdan> some people want it both ways, and it's a pain :-)
13:01 < msachs> What else is new?
13:02 < cirdan> "can't do this, cause it'll change the deb" "can't force this, it's too much work, even though the deb is different" ;-)
13:02 < msachs> Which policies are people running into here?
13:03 < cirdan> about having a deb be the same no matter where it's built
13:03 < cirdan> if there are -rev overlaps, then the same looking deb is different
13:03 < cirdan> i can't really see how hard it is to force ppl to change the rev when they cp a .info from another tree
13:04 < cirdan> really, it forces ppl to look at what they are doing (dmack's recent oops as an example) :-)
13:04 < cirdan> cause fink can also then search for any rev that doesn't match the current os/tree
13:05 < cirdan> i'd like to start a 10.4 tree, with only packages in it that ppl need/want
13:05 < cirdan> you want it in 10.4? you put it there and make sure it works
13:05 < cirdan> instead of mass copying and such
13:05 < cirdan> this way packages can be slowly added, and checked
13:06 < cirdan> but we need good guidelines as to what needs to be changed in the info files
13:06 < cirdan> and upgrade plan can be worked on while this is going on, but for ppl who wanna develop on intel, they can't wait
13:06 < cirdan> for now this could be for fresh installs only.
13:07 < msachs> Sounds like a similar situation to the current unstable/stable divide.  How well does that work?  Do packages get pulled into stable appropriately?
13:07 < cirdan> no
13:07 -!- zizban [~chris@24-52-0-219.sbtnvt.adelphia.net] has quit ["Leaving"]
13:07 < cirdan> usually we don't get feedback unless it's an important package or the maintainer just decides to move it anyway
13:08 < cirdan> we used to have feedback drives, where we would post a list of our packages and ask -users if it works ok
13:08 < cirdan> the only time we get feedback is if there is a problem somewhere
13:08 < msachs> Yeah.
13:09 < msachs> So for 10.4 non-transitional, fresh installs work fine.
13:09 < cirdan> they would, yes
13:09 < cirdan> mainly only a few ppl would be using them for now
13:10 < cirdan> me, intel ppl, and a few others
13:10 < msachs> The only issues are with upgrading.
13:10 < msachs> And me :)
13:10 < cirdan> since you'd need cvs to put files in it
13:10 < cirdan> :-)
13:10 < cirdan> mmm....http://www.blackholeinc.com
13:10 < cirdan> i want one!
13:11 < cirdan> can't spend $3k on a cube though :-(
13:11 < msachs> ha
13:11 < cirdan> fink-NeXT!
13:11 < cirdan> w00t!
13:12 < msachs> So for upgrades, problems arise when something is going to be newly built or installed and it depends or builddepends on something which was built or installed with trans.
13:12 < cirdan> yes
13:12 < pixelfairy> i thought oSx was the continuation of NeXTsTeP
13:12 < cirdan> also there are probs with mixing old debs and new
13:12 < msachs> Do we have a way of determining whether a package was built with trans or not?
13:12 < cirdan> pixelfairy: yeah, kinda
13:12 < cirdan> msachs: no
13:12 < cirdan> versioned deps are our best friend
13:13 < cirdan> but lots of work
13:13 < msachs> What if we started adding a field to the debs of packages now?
13:13 -!- pixelfairy [~pixel@dsl081-036-140.lax1.dsl.speakeasy.net] has quit ["leaving"]
13:13 < msachs> GCC: 3.3 or GCC: 4.0, or ABI: -1 or ABI: 2
13:13 < cirdan> and easiest to do if we have the tree in the -rev
13:13 < msachs> If that field is missing, assume it was built trans.
13:13 < cirdan> msachs: i believe it was just added to head
13:14 < cirdan> msachs: i think it uses gcc vers, not actual abi
13:14 < cirdan> i'd rather have the abi
13:14 < msachs> Okay, so then we /do/ have a way of identifying problematic packages, right?
13:14 < cirdan> but we either need versioned deps or need major hacking on apt/dpkg/aptitude
13:14 < msachs> How about this.
13:15 < msachs> When building or installing a package which has the GCC: field, any of its (build)deps which have a GCC field in the .info must have GCC: 4.0 in the .deb.
13:16 < msachs> If not, the dep gets rebuilt.
13:16 < cirdan> right
13:16 < cirdan> but that'd harder than versioned deps/tree-based %r
13:16 < cirdan> :-)
13:16 < msachs> What's hard about it?
13:16 < cirdan> major hacking on apt/dpkg/aptitude
13:17 < cirdan> can't forget about the bindist and installing debs correctly
13:17 < cirdan> err, directly
13:17 < msachs> Mm, that's true...
13:18  * cirdan really can't see the problem everyone has with modifing each .info... :-)
13:18 < cirdan> i must be really blind
13:18 < msachs> What if we said "doing that is problematic if you've upgraded from trans to non-trans without doing update-all, so we recommend doing an update-all or sticking with 'fink'.  And here's the error message you'll see if you get it wrong, and here's how to fix it."
13:19 < msachs> The more things people have to remember when packaging, the more chances they have to get it wrong.
13:19 < cirdan> msachs: the mailinglists will still be jammed
13:19 < akh> They want everything to be completely "under the hood"
13:19 < cirdan> msachs: the scricter the polict, the easier it is to catch mistakes
13:19 < cirdan> policy
13:20 < msachs> We have tools to help them remember, "fink validate" for instance, but they don't always get used.
13:20 < cirdan> msachs: plus, we need to be away that update-all fails all the time, because of muilddeps
13:20 < cirdan> builddeps
13:20 < msachs> Not being able to cp info files between trees is also a significant workflow change, so people will either not know about it or forget.
13:20 < cirdan> msachs: right, but in unstable, we can have mass validate run on unstable, and stable, and warn ppl if somethig is amiss
13:21 < msachs> It takes time for all the maintainers to fix everything, so we'd wind up having to do a lot of the work ourselves.
13:21 < cirdan> like if foo-1.0.1-10.4t.1 is in the 10.4 tree, easy to catch
13:21 < cirdan> msachs: we do the work ourselves anyway
13:21 < msachs> heh, true
13:23 < cirdan> or if foo-1.0.1-10.4t-1 has a dep of >= bar (2.3.1-10.3.1), that can also easily be caught
13:23 < cirdan> by a script, a human isn't even needed
13:24 < msachs> If it's something we can script and just do, that's a lot easier.
13:24 < msachs> Maybe we can even put a commithook into CVS to automatically do it if someone copies an info file over from trans.
13:24 < cirdan> right
13:24 < cirdan> eventually we can move to something like debian has, a ftp queue where a script checks, builds and adds a package
13:25 < msachs> Yeah.
13:25 < cirdan> then we'd need some way of tagging the intended tree anyway
13:25 < msachs> I've been meaning to write that, actually, but I'm going to be really busy with non-Fink stuff at least through late July.
13:25  * cirdan usually thinks 2-3 years ahead for fink :-)
13:25 -!- Murr [~neeri@A17-202-20-71.apple.com] has joined #fink
13:25 < cirdan> in the 10.2 upgrade i was arguing the same things as now :-)
13:27 < akh> cirdan: So are you ready for x86?  ;-)
13:28  * cirdan will be crying like a baby when he has to move
13:28 < msachs> Well I think people are starting to see the light on QA, both automated and non, people have been generally pleased with the emails I've sent out.
13:28 < cirdan> msachs: hmm?
13:29 < cirdan> internal apple?
13:29 < msachs> You know, the "your package is broken in this way" emails.
13:29 < msachs> Especially the ones with patches attached :)
13:29 < cirdan> ah
13:29 < cirdan> heh
13:29 < cirdan> i'm amazed with 100 packages i think i only got 1 mail from u
13:29 < cirdan> :-)
13:29 < msachs> Well that's 'cuz you're awesome.
13:30 < msachs> Man, that damn fink-devel, what a horrid maintainer, though.  Never gets anything right...
13:30 < akh> heh
13:31 < akh> Better than fink-gnome-core
13:31 < cirdan> hehe
13:31 < cirdan> grep fink-gnome-core /etc/aliases
13:31 < cirdan>  | /dev/null
13:31 < msachs> Hopefully I'll have time to write the build queue sometime in 2005H2.  Or someone else could do it of course, I wouldn't object :)
13:32 < cirdan> :-)
13:32 < cirdan> we need a few more apple devs on board ;-)
13:32  * cirdan can see the job app now...
13:33 -!- yaver [~yaver@pD9F689A5.dip0.t-ipconnect.de] has joined #fink
13:33 < yaver> hi @ all
13:33 < cirdan> "Programmer experienced in all computer languages to work closely with seceretly supported 3rd party package management system. Haskell and INTERCAL required"
13:33 < msachs> Package lots of software which exposes compiler bugs.
13:34 < akh> Pay: $0
13:34 < akh> And all the abuse #fink can dish out.
13:35 < cirdan> akh: no, we have $3 in our bank account...
13:35 < msachs> Damn, I'm not qualified, need to brush up on my INTERCAL.
13:35 < cirdan> we could afford to give them *something* more than abuse :-)
13:35 < cirdan> msachs: feanor can help ya with it ;-)
13:35 < akh> cirdan: That's Infinitely better.
13:35 < msachs> Oh good.
13:35 < akh> biab
13:35 -!- akh [~akhansen@jove.psfc.mit.edu] has quit []
13:36 -!- cmeme [~cmeme@216.184.11.2] has quit [Remote closed the connection]
13:37 -!- cmeme [~cmeme@216.184.11.2] has joined #fink
13:37 < yaver> i have a problem with the fink installen on OS X 10.4.0, on setup i can't choose my Volume (it means: volume does not support symlinks) but it's an hfs+
13:37 < yaver> installer 
13:37 -!- cmeme [~cmeme@216.184.11.2] has quit [Remote closed the connection]
13:37 < msachs> yaver: You need to be an admin users.
13:37 < msachs> *user
13:37 < yaver> oh
13:38 -!- cmeme [~cmeme@216.184.11.2] has joined #fink
13:38 < yaver> root or normal admin user
13:38 < cirdan> normal admin
13:38 < msachs> !symlinks
13:38 < Melian> symlinks is, like, http://www.finkproject.org/faq/usage-fink.php?phpLang=en#non-admin-installer
13:38 < yaver> i've start it as normal admin
13:40 -!- akh [~akhansen@jove.psfc.mit.edu] has joined #fink
13:40 < yaver> ok thx i try it
13:43 -!- yaver [~yaver@pD9F689A5.dip0.t-ipconnect.de] has quit [Read error: 113 (No route to host)]
13:51 < cirdan> msachs: you build glib with new xcode?
13:52 < msachs> cirdan: Yep, successfully
13:52 < cirdan> hrm
13:53 < akh> I did too.
13:53 < cirdan> odd
13:53 < cirdan> akh: i thought yours was xcode 2.0
13:53 < akh> I've updated.
13:53 < cirdan> oh
13:53 < akh> I needed c++filt and figured that was the way to go.
13:53 < cirdan> heh
13:55 < akh> And I like it on the edge.  ;-)
13:59 -!- dreamind [~dreamind@C2107.campino.wh.tu-darmstadt.de] has joined #fink
13:59 < cirdan> msachs: 
14:00 < cirdan>  gcc -DHAVE_CONFIG_H -I. -I. -I.. -I.. -DG_LOG_DOMAIN=\"GLib\" -DG_DISABLE_CAST_CHECKS -DG_DISABLE_DEPRECATED -DGLIB_COMPILATION -no-cpp-precomp -isystem /sw/include -D_POLL_EMUL_H_ -O3 -funroll-loops -fstrict-aliasing -Wall -c gasyncqueue.c  -fno-common -DPIC -o .libs/gasyncqueue.o
14:00 < cirdan> In file included from garray.c:762:
14:00 < cirdan> galiasdef.c:18: error: only weak aliases are supported in this configuration
14:00 < cirdan> like 30-40 lines of that last one
14:00 < cirdan> glib2
14:02 < cirdan> i do see this in configure...checking for GNUC visibility attribute... conftest.c: In function 'f_internal':
14:02 < cirdan> conftest.c:11: warning: internal and protected visibility attributes not supported in this configuration; ignored
14:02 -!- Fang [~Fang@2002:56c1:6039:1:0:0:0:1] has quit ["Famous last word #1: "Oh yeah?! Watch me!""]
14:03 < msachs> -fstrict-aliasing
14:03 < msachs> Weak aliases are not strict.
14:04 < cirdan> so...why would it be failing on my machine/
14:04 < msachs> Is this glib or glib2?
14:04 < cirdan> glib2
14:04 < msachs> ah
14:05 < msachs> hm
14:06 < msachs> Maybe it's -fno-common.  I don't have that in my build.
14:06 < msachs> Oh,k yes I do.
14:07 < msachs> weird...
14:07 < cirdan> yeah
14:07 < msachs> send me your config.log and i'll diff it with mine
14:07 < cirdan> ok
14:09 < msachs> reproducing mine, my builds don't keep build roots around for packages which succeed...
14:09 < cirdan> heh
14:11 < msachs> -configure:2556: found /usr/bin/gcc
14:11 < msachs> +configure:2556: found /sw/bin/gcc
14:12 < msachs>  /usr/bin/ld: Undefined symbols:
14:12 < msachs> -_ngettext
14:12 < msachs> +_libintl_ngettext
14:13 < akh> mmmm...ccache-default
14:13 < cirdan> heh
14:13 < cirdan> http://websrvr40nj.audiovideoweb.com/avwebdsnjwebsrvr4501/portal/media/media-050516-pregnancy.html
14:13 < msachs> In configure, your check for GNU visibility attribute gets yes, mine gets no.
14:13 < cirdan> msachs: i have colorgcc installed :-)
14:13 < cirdan> odd
14:14 < cirdan> wonder why
14:15 < msachs> Your system environment is slightly different than mine.
14:15 < msachs> Don't see anything suspicious, but there it is.
14:15  * cirdan removed colorgcc
14:15 < msachs> It's not 8.2.0 vs. 8.1.0, since the machine I do my build runs on has 8.1.0.
14:15 < cirdan> akh: that vid is dreat
14:15 < msachs> You also have a bunch of other /sw/bin stuff that I don't have, like sed and install.
14:16 < msachs> And fortran.
14:16 < msachs> I'm DCCing you the diff.
14:16 < cirdan> yeah, coreutils
14:16 < cirdan> k
14:19 < cirdan> my abttery need a recalib :-/
14:19 -!- joshmoz [~joshmoz@covad-mozilla.meer.net] has joined #fink
14:19 < cirdan> it should have died like 10-15 mins ago :-)
14:19 < cirdan> msachs: odd, it was colorgcc that confisued it
14:20 < msachs> That's bad.
14:20 < cirdan> yeah
14:20 < cirdan> cause i used colorgcc for a few months now
14:20 < msachs> Mm.
14:27 -!- Snaggle [~nieder@128.252.206.171] has joined #fink
14:27 < akh> strange
14:28 < Snaggle> Can someone look at <http://sourceforge.net/tracker/index.php?func=detail&aid=1222737&group_id=17203&atid=414256>?  Been sitting idle for 2 weeks.  Thanks,
14:30 < akh> Only 2 weeks?  ;-)
14:32 < cirdan> it's a new one
14:32 -!- Feanor_ [~astrange@64.207.61.218] has joined #fink
14:38 -!- hennker [flullup@dsl-082-082-239-083.arcor-ip.net] has joined #fink
15:05 -!- Murr [~neeri@A17-202-20-71.apple.com] has quit [Read error: 145 (Connection timed out)]
15:33 -!- msachs [~msachs@17.255.96.116] has quit []
15:34 -!- ken [~ken@ken.codeweavers.com] has joined #fink
15:38 -!- hennker [flullup@dsl-082-082-239-083.arcor-ip.net] has quit [Read error: 145 (Connection timed out)]
15:44 -!- Murr [~neeri@A17-202-20-71.apple.com] has joined #fink
15:50 -!- msachs [~msachs@17.255.96.116] has joined #fink
16:01 -!- msachs [~msachs@17.255.96.116] has quit []
16:01 -!- ringerc [~craig@dsl-202-72-144-62.wa.westnet.com.au] has quit ["avoiding keyboard prints - zzzzz"]
16:04 -!- ken [~ken@ken.codeweavers.com] has left #fink []
16:04 -!- dhamilto [~dhamilto@141.222.38.5] has joined #fink
16:05 -!- cirdan_ [~chris@pcp04356153pcs.glstrt01.nj.comcast.net] has joined #fink
16:05 -!- cirdan_ [~chris@pcp04356153pcs.glstrt01.nj.comcast.net] has quit [Client Quit]
16:06 -!- msachs [~msachs@17.255.96.116] has joined #fink
16:13 -!- msachs [~msachs@17.255.96.116] has quit []
16:16 < akh> lull
16:21 -!- dhamilto [~dhamilto@141.222.38.5] has left #fink []
16:27 -!- msachs [~msachs@17.255.96.116] has joined #fink
16:28 -!- JesseW [~chatzilla@JesseW.student.supporter.pdpc] has joined #fink
16:29 -!- zizban [~chris@24-52-0-219.sbtnvt.adelphia.net] has joined #fink
16:33 < akh> zizban:  comment broke my system.
16:33 < akh> :-P
16:33 < akh> There's your feedback!
16:35 -!- msachs [~msachs@17.255.96.116] has quit []
16:36 < zizban> thanks, man :)
16:42 < zizban> hmmmm everything I am trying to build from unstable is failing with "Buildlock failure"
16:47 < zizban> yet building the package seperately works. weird
16:48 < akh> fink list -i buildlock
16:48 < zizban> hold on, I went on to building scrollkeeper
16:48 -!- msachs [~msachs@17.255.96.116] has joined #fink
16:48 < akh> You should only have one such package at any time.
16:49  * akh thinks "fink cleanup" should remove stale locks.
16:49 < zizban> nothing showing up
16:49 < cirdan> heh
16:49 < cirdan> http://www.omegnet.net/daisuke744/macos.html
16:51 < zizban> he makes it sound so easy
16:53 < akh> zizban: what sort of buildlock failure--failure to set, failure to remove...
16:54 -!- dmacks [~dmacks@netspace.org] has joined #fink
16:56 < zizban> akh: hold on
16:57 -!- msachs [~msachs@17.255.96.116] has quit []
16:57  * dmacks is racing RangerRick to get user's CPU to catch fire with rebuilds of gnome before KDE does it:)
16:58 < akh> fire!  heh, heh,heh
16:59 < dmacks> akh: upstream of pygtk just told me "just don't distribute compiled files and you won't have trouble due to our compiler-script mis-compiling them"
17:00 < zizban> heh
17:01 < akh> Sounds like grounds for a major league ass-kicking.
17:01 -!- linuxmaniac [~maniac@199.Red-83-35-243.pooles.rima-tde.net] has joined #fink
17:02 -!- linuxmaniac [~maniac@199.Red-83-35-243.pooles.rima-tde.net] has left #fink ["Leaving"]
17:03 < dmacks> cluebat pygtk folks
17:03 < dmacks> !cluebat pygtk folks
17:03  * Melian beats pygtk folks with a clue-by-four. *WHAM* get a clue! *WHAM* get a clue! *WHAM*
17:05 < dmacks> Anyone know the blootbot syntax to connect jumper-cables to the target's crotch?
17:07 -!- msachs [~msachs@17.255.96.116] has joined #fink
17:08 < akh> Probably an alicekill would be sufficient
17:09 -!- Snaggle [~nieder@128.252.206.171] has quit ["Yale still sucks"]
17:13 -!- Feanor_ [~astrange@64.207.61.218] has quit []
17:14 -!- newmanbe [~newmanbe@aba7b467afbaf4ca.session.tor] has joined #fink
17:15 < newmanbe> !lart Apple branded mice
17:15  * Melian forces Apple branded mice to use Outlook Express
17:15 < newmanbe> It feels funny...
17:15 < dmacks> Hrm, "Apple brandied rice" doesn't sound so good either:(
17:15 < cirdan> ?
17:16 < zizban> heh
17:16 < cirdan> dmacks: how does debian handle the files?
17:18 < dmacks> cirdan: They distribute the modules as source-only
17:19 < cirdan> oh
17:19 < cirdan> so build in the post-install?
17:19 < dmacks> That means I gotta make at least some of the build dependencies into runtime dependencies.
17:20 < cirdan> oh
17:20 < cirdan> eww
17:20 < cirdan> whats the problem?
17:20 < cirdan> everything hardcoded?
17:21 < dmacks> Their build system considers DESTDIR as part of the ultimate prefix, not as a temp location. There's no way to say "install [here], but it's really going to be used [here instead]", so binaries hardcode the buildpath into themselves.
17:21 < zizban> finally got the buildlocke error again:
17:21 < zizban> fink-buildlock-foo
17:22 < dmacks> (it doesn't help that compiling happens during 'make install' not 'make')
17:22 < zizban> Failed: buildlock failure
17:27 < dmacks> zizban: That's it? Nothing about conflicts with other pkgs? Are you using 'fink' in a pipeline (sending output through 'less' or 'tee' or somesuch)?
17:27 < cirdan> dmacks: oooh fun
17:27 < zizban> hold on lemme go on irc on my mac and show you what I have
17:28 -!- RangerRick [~ranger@cpe-024-168-176-124.nc.res.rr.com] has joined #fink
17:28 -!- Topic for #fink: See http://fink.sf.net/ for information on using Fink with Tiger || Have a question?  First, check the faq: http://fink.sf.net/faq/ || Latest Installer: Fink 0.6.4 (for 10.2) and 0.7.2 (for 10.3) and 0.8.0 (for 10.4) || Latest fink 0.24.7 || Carina || Check New Map in http://fink.sf.net/map
17:28 -!- Topic set by akh [] [Sun Jun  5 21:31:13 2005]
17:28 [Users #fink]
17:28 [ Airo     ] [ dreamind] [ JesseW        ] [ mee_bot ] [ pogma     ] [ uncon     ] 
17:28 [ akh      ] [ dsias_  ] [ JosephSpiros  ] [ Melian  ] [ RangerRick] [ usataway  ] 
17:28 [ bender|  ] [ emp__   ] [ joshmoz       ] [ Miaz    ] [ regeya    ] [ xhrl      ] 
17:28 [ BleedAway] [ eno-away] [ jtyler        ] [ msachs  ] [ RLD_osx   ] [ zizban    ] 
17:28 [ cirdan   ] [ gecko2  ] [ kane-xs       ] [ muesli  ] [ robilad   ] [ zizban_mac] 
17:28 [ Clef     ] [ gzl     ] [ KraMer        ] [ Murr    ] [ runelind  ] 
17:28 [ cmeme    ] [ hramrach] [ lisppaste     ] [ neo     ] [ swix_     ] 
17:28 [ das_     ] [ htodd   ] [ mcp           ] [ pHatidic] [ The_Tick  ] 
17:28 [ dmacks   ] [ jack-   ] [ mdmonk_nothere] [ pnorman_] [ till      ] 
17:28 -!- Irssi: #fink: Total of 50 nicks [0 ops, 0 halfops, 0 voices, 50 normal]
17:28 -!- Channel #fink created Sun Aug  3 18:57:20 2003
17:28 < dmacks> zizban: So if you now try 'fink build libgnomecanvas2' does it still do that?
17:28 < zizban> trying
17:28 -!- Irssi: Join to #fink was synced in 43 secs
17:28 < zizban> well its building, we'll see
17:29  * dmacks tries to distinguish an actual buildlock or package-build ordering problem from a bug in the error message, (should probably mention not-installed pkgs, not just conflicts in paragraph 1 line 1)
17:30 < zizban> heh
17:31 < zizban> nope, works fine
17:31 < dmacks> *phew* /me tries to avoid futzing with the actual build routines at all cost
17:32 -!- dmacks is now known as dmacks_away
17:32 < zizban> mmmmkay, that's it?
17:33  * zizban watches dmacks fling himself on his horse and ride off madly off in all directions
17:33 < akh> ah, the old libgettext3-dev swapping with gettext-dev error!
17:33 < zizban> an old bug, eh?
17:33 < akh> Not that old.
17:33 < akh> The fix is to keep doing updates.
17:34 < zizban> selfupdate?
17:34 < akh> No, just keep updating the packages.
17:34 < zizban> ?
17:34 < akh> Basically just do what you've been doing.
17:34 < zizban> ah
17:34 < zizban> okay
17:35  * akh doesn't seem to have this problem anymore.
17:35 < zizban> lucky you :)
17:35 < akh> fink from HEAD
17:36 < zizban> yummy
17:36 < zizban> I'm doing all this just to test Galeon for vasi
17:36 -!- zizban_mac [~zizban@24-52-0-219.sbtnvt.adelphia.net] has quit [Remote closed the connection]
17:36 < akh> If you're on Tiger you'll get  a runtime error.
17:37 < zizban> oh fun
17:37 < akh> http://fink.sourceforge.net/faq/usage-packages.php?phpLang=en#tiger-gtk
17:37 < akh> They're ubiquitous.
17:37 < zizban> I musy have missed that
17:38 < akh> Or you haven't used a GTK app.
17:38 < zizban> true
17:38 < zizban> I haven't had a lot of time for X11 these days
17:40 < akh> Luckily the workaround isn't too heinous--one could probably set the environment variable at login, but I didn't want to create "unforeseen consequences" by advocating that.
17:41 < zizban> true
17:48 -!- msachs [~msachs@17.255.96.116] has quit []
17:53 -!- You're now known as RangerAway
17:59 < dmacks_away> akh: we're prolly gonna force FALLBACK in everyone's env Soon.
17:59 < akh> w00t!
18:00 < zizban> heh
18:04 < akh> Will that be done in a way such that Apple's X11 is aware of it?
18:04 < akh> (as if I care)
18:04 -!- rudy_ [~rudy@ip68-224-184-250.hr.hr.cox.net] has joined #fink
18:04 < zizban> heh
18:04 < zizban> some people would care
18:04 < akh> Right--it doesn't affect my ability to run stuff.
18:05 -!- dmacks_away is now known as dmacks
18:05 < rudy_> akh: you a physics nut?
18:05 < dmacks> How does it affect Apple's X11?
18:06 < akh> rudy_: Well, I commute 2.5 hours each way to work an academic job, so that sounds pretty nutty.
18:06 < zizban> what do you do, commute from Maine?
18:07 < rudy_> akh: want to do my last physics problem? :)
18:07 < akh> Sure.
18:07 < dmacks> rudy_: The answer is "1". All you need to do is assign the units.
18:07 -!- newmanbe_notaclo [~newmanbe@H80.C211.tor.velocet.net] has joined #fink
18:07 -!- newmanbe_notaclo [~newmanbe@H80.C211.tor.velocet.net] has quit [Killed by Idoru. (Sorry, tor connections to freenode are temporarily blocked due to abuse.)]
18:08 < rudy_> "An air mattress is 2.5 m long, 0.71 m wide, and 13 cm deep. If the air mattress itself has a mass of 0.22 kg, what is the maximum mass it can support in fresh water?"
18:08 < dmacks> "1"
18:08 < dmacks> I already told you that.
18:08 < rudy_> i thought i had the right answer, and yet it still says it's wrong
18:08 < rudy_> dmacks: "42" now go ask the question.
18:09 < dmacks> But this isn't life, the universe, and everything...this is academia, an exam, and nothing.
18:09  * rudy_ puts dmacks on his stab list
18:09 < akh> rudy_: the mass is equal to the mass of an equivalent volume of water.
18:10 < akh> or is it...just a sec
18:10 < rudy_> well, sure i can calculate the equiv volume of water for my 0.22 mass, but this is a maximum problem
18:10 < rudy_> so i'm thinking that the matress should have near 0 volume above the water level
18:11 < akh> Right.
18:11  * rudy_ sets the volume to 0.000000001 and calculates it out
18:12 < akh> mass of water = volume of mattress * density of water
18:13 < rudy_> nod, i have that value :)
18:13 < akh> supported mass = mass of water - mass of mattress
18:15 < rudy_> isn't mass of water missing a gravity component?
18:16 < akh> oops--I should have said "weight"
18:16 < rudy_> nod
18:17 < akh> hate it when that happens.
18:17 < akh> Stop having exams for 10 years and everything goes to pot.  ;-)
18:17 < rudy_> weight and mass get confused, especially in NASA. where they enjoy converting SI to english
18:17 < akh> That too.
18:17 < zizban> mass and density get me
18:18 < rudy_> mass and density?!?
18:18 < zizban> ya
18:18 < rudy_> density = m/v
18:18 < zizban> is there a simple way to remember the difference?
18:18 < rudy_> they're not the same thing? :)
18:18 < akh> density doesn't depend on the size of the object.
18:18 < zizban> like staurn, which is massive but not dense
18:19 < akh> A gallon of water has the same density as a glass of water.
18:19 < rudy_> akh: did you run the calculation? :)
18:19 < rudy_> i'm getting like 0.015219164118 kg
18:19 < zizban> but a gallon of water is more massive, right?
18:19 < akh> No--got distracted by $job
18:19 < akh> yes
18:19 < rudy_> damn jobs :)
18:20 < rudy_> oh well, 0.0152 was wrong :)
18:20 < rudy_> 29/30 isn't bad though
18:21 < dmacks> A typical AOLer has a definite mass, 70-100kg or so, but is nearly infinitely dense:)
18:21 < rudy_> heh
18:21 < akh> 0.01075
18:23 < dmacks> zizban: A good distinction among various measurements is between "intrinsic properties" (things specific to the kind of stuff) from "extrinsic properties" (things specific to a particular sample of the stuff).
18:24 < zizban> ah
18:24 < zizban> damn storm bbl
18:24 -!- jtyler [~jtyler@iphost-64-56-130-194.edm.wiband.net] has quit ["Leaving"]
18:24 -!- zizban [~chris@24-52-0-219.sbtnvt.adelphia.net] has quit ["Leaving"]
18:25 < dmacks> akh: I guess we assume rudy_'s supported object is a point mass? And that "supported" doesn't mean "sunk to the bottom but the mattress didn't pop"?
18:26  * rudy_ attaches dmacks
18:29 < akh> dmacks: Of course I assumed that.  ;-)
18:31 < akh> And that the water was at 4 C.
18:31 < rudy_> and that dmacks shaved his head completely bald
18:32  * dmacks got in trouble numerous times in school for answering a too-strict reading of a word problem, or making other assumptions in order to simplify the work.
18:33 < cirdan> hmm
18:33 -!- citizen_0 [~user@va-sterling-u1-c5c-a-146.chvlva.adelphia.net] has joined #fink
18:39 < cirdan> this sux0rs
18:39 -!- JesseW [~chatzilla@JesseW.student.supporter.pdpc] has quit [Remote closed the connection]
18:47 < dmacks> akh: The open-source seminar at Babson the other day has audio+slides available on groklaw.
18:48 < cirdan> cool
18:51 < akh> nice
18:54 < xhrl> has anyone found the compression of p7zip to be better than bzip2?
18:54 < akh> haven't tried it yet.
18:56 -!- Feanor [~astrange@mp1-248-116.dialup.emory.edu] has joined #fink
18:56 < akh> giving it a whirl now
18:56 < xhrl> i am trying to see whether you can encrypt the files too
19:00 < dmacks> "zip" also supports encryption.
19:00 < akh> ewww--zip
19:00 < dmacks> (p7 claims 2-10% higher compression ratio than zip, if I understand their fairly useless marketroid webpages)
19:01  * akh was just going to try a file
19:01 < dmacks> You and your "real world" and "actual data" can just go to hell, m'kay?
19:01 < dmacks> :)
19:02 < akh> heh--already there
19:02 < dmacks> Ah, still at work, huh?
19:03 < akh> yup--there was an 'incident'
19:03 < cirdan> oo
19:03 < dmacks> d'oh!
19:03 < cirdan> damn, win2k isnt even on torrent sites anymore
19:04 < akh> RIP
19:04 -!- newmanbe [~newmanbe@4b8a0da864b59e8e.session.tor] has joined #fink
19:09 < akh> now once I finish building p7zip and apt-get it over here I can try that.
19:11 -!- rudy_ [~rudy@ip68-224-184-250.hr.hr.cox.net] has quit []
19:12 < dmacks> !seen vasi
19:12 < Melian> vasi <~vasi@modemcable214.145-70-69.mc.videotron.ca> was last seen on IRC in channel #fink, 6h 54m 10s ago, saying: 'wonderful that'.
19:12 < akh> he fixed boostrap
19:13 < dmacks> Cool.
19:13 < akh> At least supposedly.
19:15 < xhrl> akh, are do you not usually work on-site?
19:16 < xhrl> err;)
19:16 < xhrl> -are
19:16 < akh> I'm at work--I'm on my laptop in the lab rather than in my office.
19:16 < cirdan> whee...
19:16 < cirdan> gcc -dynamiclib -flat_namespace -undefined suppress -o .libs/liblinc.1.0.0.dylib  .libs/linc.o .libs/linc-connection.o .libs/linc-protocols.o .libs/linc-server.o .libs/linc-source.o  libglib-2.0.dylib -L/sw/lib /sw/lib/libgobject-2.0.dylib /sw/lib/libgthread-2.0.dylib /sw/lib/libglib-2.0.dylib /sw/lib/libintl.dylib /sw/lib/libiconv.dylib -install_name  /sw/lib/liblinc.1.dylib -compatibility_version 2 -current_version 2.0
19:16 < cirdan> powerpc-apple-darwin8-gcc-4.0.0: libglib-2.0.dylib: No such file or directory
19:16 < cirdan> powerpc-apple-darwin8-gcc-4.0.0: /sw/lib/libgobject-2.0.dylib: No such file or directory
19:17 < dmacks> cirdan: You updated glib2-shlibs but not glib2-dev
19:18 < cirdan> dmacks: no, update-all did
19:19 < dmacks> You sure all three glib2* got updated by the time of this error?
19:19 < dmacks> I.e., are you just deflecting blame or pointing out an actual problem? :)
19:21 < cirdan> i dunno
19:23 < dmacks> Are all pkgs in 'dpkg -l glib"2*"' of the same version?
19:23  * dmacks suspects glib2-dev != glib2-shlibs
19:31 < dmacks> vasi and I discussed having 'fink' do an 'apt-cache unmet' (or our own implementation thereof) at various points during fink install/update.
19:31 < akh> which solves...?
19:36 -!- msachs [~msachs@17.255.96.116] has joined #fink
19:36 < msachs> 'lo
19:36 < akh> hey
19:36 < newmanbe> Hello.
19:37 < akh> hmmm....didn't know that Growl had to be installed for all users for Fink to see it.
19:37 < akh> Guess I should have suspected, though...
19:37 < The_Tick> maybe the info documents should be clarified ;)
19:38 -!- Feanor [~astrange@feanor.developer.opendarwin] has quit [Read error: 110 (Connection timed out)]
19:39 < dmacks> akh: cirdan's problem is probably that %N-shlibs got updated but not %N-dev, even though %N-dev:Depends:%N-shlibs(=%v-%r)
19:39 < dmacks> 'unmet' would flag this inconsistency.
19:39 < akh> Ah--makes sense.
19:39  * akh has experienced the same problem before.
19:48 -!- dreamind [~dreamind@C2107.campino.wh.tu-darmstadt.de] has quit []
19:50 -!- baba [~baba@YahooBB220041001026.bbtec.net] has joined #fink
19:52 -!- Murr [~neeri@A17-202-20-71.apple.com] has quit [Read error: 145 (Connection timed out)]
20:00 -!- Feanor [~astrange@mp1-249-82.dialup.emory.edu] has joined #fink
20:08 < pogma> morning
20:11  * cirdan goes for some guiness
20:12 < msachs> hey pogma
20:12 < msachs> bye cirdan
20:12 < msachs> Fair trade ;)
20:12 < akh> bye cirdan, hi pogma
20:13 < dmacks> heh
20:21 < dmacks> msachs: You have xml-parser-pm58X marked as a failures/unanalyzed on intel, but the .log indicates successful build.
20:23 < msachs> No it doesn't, is my comparison link wrong?
20:23 < msachs> Nope.
20:24 < pogma> okay, bootstrap improving, need to patch apt though
20:24 < dmacks> msachs: /me is looking at http://fink.opendarwin.org/build/2005-06-24/out/pkgindex-intel.html
20:24 < pogma> fink code bad :)
20:25 < msachs> dmacks: That doesn't actually link to the Intel ones, I coped that over from the other host to run cmpruns, use /build/2005-06-24.intel/out/pkgindex.html
20:25 < dmacks> Ahhhh. Yes, that's better (in a linking-to-log-with-error sense anyway)
20:26 < msachs> :)
20:26 < msachs> Biggest blocker is probably glib2, which appears to have some sort of endianness issue or something.
20:26 < msachs> Fails a unicode test.
20:26 < dmacks> htodd: your string-approx-pm581.info in both 10.3 and 10.4T need Depends:perl581-core
20:28 < dmacks> msachs: glib2 went to 2.6.5 this morning...I wonder if they fixed it?
20:28 < msachs> I'll do a one-off build.
20:28 < dmacks> Hold on...lemme look at the sources first:)
20:30 < akh> You kids and your crazy schemes.  ;-)
20:31 < dmacks> Well, they didn't change the test source, and nothing about it in the changelogs.
20:31 < msachs> Building.
20:31 < msachs> If it still fails I'll take a slightly closer look at itl.
20:33 < dmacks> http://bugzilla.gnome.org/show_bug.cgi?id=143380
20:35 < dmacks> May as well stop that build.
20:36 < msachs> Mmm.
20:40 < dmacks> htodd: If you need perl, you can just use perl581-core, you don't need to give system-perl581 as an alternative to that.
20:40  * dmacks looking at extutils-pkgconfig-pm581 (which also needs Depends:pkgconfig probably)
20:49 -!- driftkop [~driftkop@user-0c8hrip.cable.mindspring.com] has joined #fink
20:53 -!- zizban [~chris@24-52-0-219.sbtnvt.adelphia.net] has joined #fink
20:54 < dmacks> msachs: system-vhia will never be able to build automatically...you may as well skip it.
20:54 < msachs> Okay.
20:55 < dmacks> (it's a placeholder for a Framework that must be manually installed)
20:55 -!- pogma [~peter@pogma.developer.opendarwin] has quit [Remote closed the connection]
20:56 < msachs> Gotcha.
20:56 -!- pogma [~peter@p1174-ipad212kobeminato.hyogo.ocn.ne.jp] has joined #fink
20:56 < akh> !wb pogma
20:56 < Melian> Welcome back pogma, o lonely traveller amongst the TCP/IP packets of chaos.
20:58 < zizban> wow
20:58 < zizban> whatta welcome
20:59 < akh> Somebody apparently had a lot of free time.
20:59 < zizban> I guess
21:02 < dmacks> It's a blootbot built-in I think.
21:03 < akh> ah
21:03 < akh> Train time.  Laters
21:03 -!- akh [~akhansen@jove.psfc.mit.edu] has quit []
21:04 < newmanbe> Chuga-chuga choo choo!
21:19 -!- driftkop [~driftkop@user-0c8hrip.cable.mindspring.com] has quit ["Client exiting"]
21:25 < pogma> python does not build on intel?
21:26 < newmanbe> Mac OS X Intel or just Intel?
21:27 < zizban> or darwin x86?
21:28 < pogma> does nobody read msachs' messages to -devel
21:28 < newmanbe> Not recently.
21:28 < dmacks> Who's messages to what, no?
21:29 < pogma> Ah, gmp, yeah, bbraun ran into that
21:29 < pogma> I think he just hacked configure so it would not try and build the x86 optimized bits
21:29 < pogma> since the assempler does not understand the assembly :)
21:30 < dmacks> pogma: there are some updates sitting in the -submissions tracker in response to msachs's emails that nobody's had time to deal with:(
21:31 < pogma> doubt that gmp is there :)
21:31 < pogma> but I'll take a look after shower
21:51 -!- zizban [~chris@24-52-0-219.sbtnvt.adelphia.net] has quit ["Leaving"]
21:55 -!- vasi [~vasi@modemcable214.145-70-69.mc.videotron.ca] has joined #fink
21:56 -!- msachs [~msachs@17.255.96.116] has quit []
21:57 < vasi> dammit
21:57 < dmacks> Yeah, I always feel like crap when I first join #fink too:)
21:58 < vasi> heh....i just came to ask msachs something...and he ran away as soon as i joined :-)
21:59 < newmanbe> dmacks: Of course you feel like that!
21:59 < newmanbe> There's a war going on at Freenode as we speak!
21:59 < vasi> ooh fun
21:59 < newmanbe> And at OFTC.
22:02 < vasi> hey, does anybody know if msachs is doing anything very weird with his buildfink? cuz i can't repro some of his errors
22:02 < pogma> some of his errors aren't
22:03 < vasi> it's very strange
22:03 < vasi> oh wait...you said he symlinked gcc3.3 -> gcc4.0, that would explain it
22:04 < pogma> yes, only gcc4
22:10 < vasi> hmph
22:11 < pogma> well, intel only has gcc4
22:12 < pogma> so, maybe it's a good thing that he's got those links
22:19 -!- Murr [~neeri@A17-202-20-71.apple.com] has joined #fink
22:21 < vasi> pogma, but i don't think we CAN change perl581 to gcc4 unless we want all perlmods to become invalid....
22:22 < vasi> so for now it just means that all pm581 packages aren't tested well by buildfink for the current dist
22:23 < pogma> sigh, I hate that fink has perl packages
22:23 < vasi> as opposed to using CPAN?
22:23 < vasi> er, "perl packages" == perlmods, or perl58*-core?
22:25 < pogma> perl58*-core
22:26 < vasi> i think it's good that we keep backwards compat for a little while, so ppl have a chance to update their perlmods...and it's good to provide new updated versions of perl with bugfixes and such
22:26 < pogma> but there is not really a "better" way, anything we did would have caused problems whenever apple updated their perl
22:26 < vasi> but i wouldn't mind phasing out 581 and 584 in tiger
22:28 < vasi> (heh, cuz what we do now never caused problems when apple updated their perl? :-)
22:35 < dmacks> Unless people know to install perl581-core when updating to Tiger, they're scrod anyway.
22:35 < newmanbe> You could tell them.
22:36 < dmacks> Yes, yes we *could*.
22:36 < pogma> scrod?
22:36 < dmacks> Okay, wise-guy, what do *you* think the past tense of "screw" is?
22:37  * pogma checks reference books
22:37 < dmacks> (the fish-slapping connotation is an added bonus:)
22:38 < pogma> dmacks: apparently the past tense of screw is screwed, except for a small portion of the internet where any old made up word will do
22:38 < dmacks> "will screw" "screw" "was scrod" "used gentoo"
22:39 < vasi> heh
22:41 < pogma> heh, getting mad, calling the engineers idiots and shouting at devbugs@apple got my bug reopened
22:41 < dmacks> sweet! Which bug?
22:42 < pogma> dmacks: well, thing is, they closed it twice as correct behavior (an ld bug that fink never sees) and I disagreed
22:42 < dmacks> Ah.
22:42 < pogma> dmacks: but now they have explained why and I feel crap for calling them idiots :/
22:42 < dmacks> Oh:(
22:43 < dmacks> Tell them you were confused, and thought you were talking about the FALLBACK bug?
22:43  * vasi hides from the scary linker
22:46 < dmacks> pogma: How does libtool determine if linker can make shared libs?
22:46 < pogma> dmacks: a very big case statement in libtool.m4
22:48 < dmacks> I've got a pkg that undefs _LT_AC_TAGCONFIG (it knows it doesn't need C++ or other tagged languages), but that makes AC_PROG_LIBTOOL think "no shared libs". Apparently only on darwin.
22:49 < dmacks> Any thoughts, other than "that's a dumb hack to solve a non-problem"?
22:50 < pogma> is joe orton @ redhat.com the author?
22:50 < dmacks> Nope. It's a gnome thing.
22:50 < dmacks> (they seem responsive to fixing it, but don't understand it)
22:51 < pogma> anyway, there was a problem with some version of libtool where I had put the darwin bits in the wrong section and joe's patch to not use c++/f77 broke darwin
22:51 < vasi> dmacks, which pkg?
22:51 < dmacks> libglade2
22:52 < pogma> dmacks: do you have a patch for it?
22:53 < dmacks> Yeah, I un-undef'ed _LT_AC_TAGCONFIG
22:53 < pogma> that should be fine, what version of libtool did it have?
22:54 < dmacks> ltmain.sh says VERSION=1.5.6
22:56 < pogma> sigh, current is 1.5.18
22:56 < pogma> I fixed the configure dies if no c++ compiler in 1.5.10
22:56 < pogma> 1.5.20 is due soon
22:56 < pogma> Ah, author uses debian, I bet
22:57 < dmacks> 'TIMESTAMP=" (1.1220.2.95 2004/04/11 05:50:42) Debian$Rev: 220 $"'
22:58 < dmacks> Ya wanna post a reply to bugzilla, or should I?
22:58 < pogma> yeah, the debian libtool maintainer gave up maintaining the package and it hasn't been updated for ages
22:58 < vasi> oops
22:58 < dmacks> d'oh^2
22:58 < dmacks> Good thing nothing ever changes in libtool.
23:02 < vasi> argh, i keep getting SPOD
23:02 < vasi> !alicekill Safari's SPOD
23:02  * Melian I want to kill Safari's SPOD, I mean, I wanna, I wanna kill Safari's SPOD. Kill. I wanna see Safari's SPOD's blood and gore and guts and have Safari's SPOD's veins in my teeth.  Eat Safari's SPOD's dead burnt body.  I mean kill, Kill, KILL, KILL.  and I jump around with vasi, yelling "KILL, KILL"
23:03 < newmanbe> cirdan: That still shouldn't be using /me.
23:04 < vasi> why not?
23:07 < dmacks> As written, it's something she's saying, not an action.
23:07 < dmacks> "* Melian wants to kill..." or "< Melian> I want to kill..."
23:09 < vasi> oh ok, i didn't even notice it was an action....i thought you meant the nick of the requesting user shouldn't be in it
23:14 < dmacks> pogma: Do you mind if I just cut'n'paste a section of your comments into the bugzilla?
23:15 < pogma> go ahead
23:31 -!- gecko2 [gecko@gecko2.user] has quit [Excess Flood]
23:32 -!- dmacks is now known as dmacks_away
23:32 -!- gecko2 [gecko@freeforge.net] has joined #fink
23:34 -!- citizen_0 [~user@va-sterling-u1-c5c-a-146.chvlva.adelphia.net] has quit [Read error: 145 (Connection timed out)]
23:45 < cirdan> !alicekill newmanbe 
23:45 < Melian> I want to kill newmanbe, I mean, I wanna, I wanna kill newmanbe. Kill. I wanna see newmanbe's blood and gore and guts and have newmanbe's veins in my teeth.  Eat newmanbe's dead burnt body.  I mean kill, Kill, KILL, KILL! and I jump around with cirdan, yelling "KILL, KILL"
23:45 < cirdan> grr
23:52 < pogma> joshmoz: let us know if any packages fail to build along the way
23:52 < pogma> but looking at msachs build report it should be okay
23:53 < joshmoz> everything I need works
23:53 < joshmoz> it all built
23:53 < joshmoz> now I find out if Firefox builds :)
23:53 < joshmoz> I think the port is done, but I haven't tested it until now - flying blind
--- Log closed Fri Jul 01 00:00:04 2005