--- Log opened Wed Jul 27 00:00:17 2005
00:09 < dmacks> vasi: eh?
00:09 < vasi> do you think i should backport the apt-ftparchive speedup stuff, or just leave it for next release?
00:10 < dmacks> I'm thinking leave it out. We've already got a 0.24.8 release candidate, may as well get that thing out the door...otherwise "just one more thing" and it may *never* get done.
00:11 < dmacks> 0.25 will feature "massive speedups of several common tasks"
00:14 < vasi> yeah
00:14 < vasi> i'm still trying to figure out how to refactor the indexing code so i don't get a headache when i try to think about it
00:15 < dmacks> Tis a shame that doing so requires thinking about it.
00:15 < vasi> i know :-)
00:18 < vasi> i'm probably going to move stuff to a new Indexer.pm, it's kinda silly that it's sitting in Package
00:51 -!- kalessin [~crhalpin@adsl-68-78-232-23.dsl.mdsnwi.ameritech.net] has quit ["."]
00:55 -!- kalessin [~crhalpin@adsl-68-78-232-23.dsl.mdsnwi.ameritech.net] has joined #fink
00:58 < dmacks> !seen feanor
00:58 < Melian> feanor <~astrange@feanor.developer.opendarwin> was last seen on IRC in channel #fink, 18h 24m 58s ago, saying: 'heh, i just saw an email from an apple employee to gcc-dev complaining that some change caused a bunch of regressions in fink'.
00:59 < dmacks> vasi: Do you understand fontconfig2?
00:59 < vasi> never looked at it really
00:59 < dmacks> (actually, I'd love to hear from *anyone* who does)
01:03 < dmacks> If I want to force linkage to a static archive even in the presence of a shared lib, can I just pass libfoo.a instead of -lfoo, or do I pass /the/whole/path/to/libfoo.a ?
01:06 < vasi> i think you need the path
01:21  * kalessin thinks orkut is a neat idea, but is annoyed at the invite system.
01:22 < dmacks> Don't worry, we aren't talking about you there. Much.
01:22 < dmacks> :)
01:22 < kalessin> If you were talking about me at all, I'd be quite proud of me.
01:22 < kalessin> I didn't know I was important enough to warrant any kind of comment.  :-)
01:23 -!- shres [~sshreyas@202.144.86.147] has joined #fink
01:26 < vasi> kalessin, everybody's worthy of comment....especially when that comment is "have you tested fink 0.24.8 yet?" ;-)
01:26 < dmacks> heh
01:26 < kalessin> vasi If by tested you mean 'ran one selfupdate', then yes.  :-)
01:27 < kalessin> I just got back from a trip up to the northern edge of nowhere.  Or at least, the northern edge of Michigan's UP.
01:27  * vasi is Canadian :-P
01:28 < kalessin> Yeah, but aren't you east of north of the UP?
01:28 < kalessin> Hmm.  that was a confusing question.
01:28 < vasi> yeah, i've never been farther north than 47th parallel
01:28 < vasi> tbh :-)
01:30  * kalessin is too lazy to get out a map.
01:31 < htodd> I think no michigan isn't very north
01:31 < vasi> i think it goes just a bit farther than that...lemme check
01:32 < vasi> yeah, if by "northern edge" you mean keweenaw, then that's farther north than i've ever been
01:32  * vasi is a pretty pathetic Canadian!
01:33 < htodd> what's the latitude?
01:34 < htodd> the thing is, the map is a little tilted
01:35 < htodd> Portland Oregon is further north than you'd think
01:36 < htodd> vasi, you know the status of the cvs servers?
01:37 < vasi> US/Canada border is at 49' in the west, much farther south in the east...Maine is actually not as far north as  Washington state
01:37 < vasi> htodd, no idea, they seemed kinda down earlier
01:38 < vasi> ah, developer access is worky again
01:38 < vasi> anon is still screwy i think
01:39 < htodd> ok, let me try.
01:41 < kalessin> vasi, bah.  I was around 46.9th parallel.
01:44 -!- eno-away [~eno-away@adsl-68-123-121-115.dsl.snfc21.pacbell.net] has quit [Read error: 110 (Connection timed out)]
01:48 < vasi> hey dmacks, do we still need to use Symbol::gensym? i think 5.8.1 has anonymous fhs
01:48 < cirdan> hey vasi 
01:48 < vasi> hiya cirdan
01:49  * cirdan goes to sleep
01:49 < cirdan> :-)
01:49 < vasi> pfft, whacko elitist :-P
01:50 < cirdan> yup
01:50  * cirdan bows good night
01:52 < cirdan> hey vasi
01:52 < cirdan> i had to track down a bug in xfs_repair today :-/
01:52 < vasi> uhhh
01:53 < cirdan> not fun to find bugs in your fs repair utils :-)
01:53 < vasi> eww
01:53 < cirdan> 2nd problem i found dealing with this stuff
01:53 < cirdan> 1st was in the raid stuff
01:53 < cirdan> yeah, wanna know what it was?
01:53 < vasi> sure, let's hear
01:54 < cirdan> it found all the superblocks, but when it went to examine the first one, it went to the wrong
01:54 < cirdan> offset
01:54 < cirdan> even though it found it before
01:54 < cirdan> for some reason my config needed one of that values to be typecast to a 64bit int
01:55 < cirdan> worked fine on other systems
01:55 < cirdan> but in the first stage the code there was typecast
01:55 < cirdan> and it used to work, too
01:55 < cirdan> just stopped working for some reason
01:55 < cirdan> very odd
01:56 < vasi> i hope you filed a bug with somebody?
02:00 < cirdan> yeah
02:00 < cirdan> sgi
02:00 < cirdan> and also debian
02:00  * cirdan hopes for an xfs kext one day
02:00 < cirdan> right now the utils build and work on os x...
02:00 < cirdan> cause they don't need any kernel interction
02:00 < cirdan> w00t! http://www.bulldoginformation.com/dogue-de-bordeaux.html and http://www.puppydogweb.com/gallery/doguedebordeaux/e.htm
02:00 < cirdan> i want one!
02:00  * cirdan used to have 2, then one died, and the other lived to over 14 i believe...
02:01  * cirdan knows how to keep dogs alive long time, i guess :-)
02:01 < cirdan> usually only live to 10, 12 max
02:04 < dmacks> vasi: It appears that 'open my $fh, $filename or die "oh shit: $!"' has been valid since perl5.6.0
02:05 < vasi> dmacks, yeah
02:05 < cirdan> night
02:05 < vasi> nite
02:06 < dmacks> My filehandle knowledge seems to have calcified long ago:(
02:08 < vasi> i learned perl much more recently, so i keep using all these fancy-shmancy things without realizing they're new :-)
02:08 < vasi> since perl6 will never be finished, and 5.x is evolving kinda slow now, i never have to worry about calcification though
02:15 < dmacks> heh yeah
02:16 < dmacks> I just cleaned out some old accounts of mine, found perl4 stuff.
02:17 < vasi> wow, that's so before my time
02:18 < dmacks> Some would say that you didn't miss much.
02:18 < dmacks> They'd be right.
02:18 < dmacks> ...which rhymes with "g'nite".
02:18 -!- dmacks [~dmacks@dmacks.active.supporter.pdpc] has quit ["leaving"]
02:27 -!- shres [~sshreyas@202.144.86.147] has quit ["This computer has gone to sleep"]
02:30 -!- vasi [~vasi@modemcable133.147-70-69.mc.videotron.ca] has quit []
03:02 -!- kalessin [~crhalpin@adsl-68-78-232-23.dsl.mdsnwi.ameritech.net] has quit ["."]
03:08 -!- Tobin [~Tobin@c211-28-177-204.eburwd7.vic.optusnet.com.au] has joined #fink
03:14 -!- shres [~sshreyas@202.144.86.147] has joined #fink
03:17 < Tobin> Hmm, I seem to be having trouble installing svn-client-ssl
03:17 < lisppaste> Tobin pasted "svn-client-ssl" at http://paste.lisp.org/display/10237
03:18 < Tobin> Does that mean anything to anyone? I'm a bit lost.
03:21 -!- shres [~sshreyas@202.144.86.147] has left #fink ["Leaving"]
03:21 < Murrito> Tobin, I'm afraid your woes mean nothing whatsoever to us
03:22 < Tobin> Trying its suggested solution, "fink remove --recursive db42-shlibs" doesn't help.
03:22 < Murrito> but technically, the issue is rather simple
03:22 < Tobin> Damn.
03:22 < Tobin> Really?
03:22 < Murrito> yeah
03:22 < Tobin> ... Well?
03:22 < Murrito> I don't know how much of the problem you already understand?
03:23 < Murrito> basically, many of our packages have an -ssl and a non-ssl variant
03:23 < Murrito> there is no reason to use the latter other than export issues
03:23 < Tobin> Not much, I need db42-ssl-shlibs but can't get it while I still have db42-shlibs?
03:24 < Murrito> yes, typically the ssl and non-ssl versions cannot coexist
03:24 < Murrito> so once you have one -ssl package (in your case svn-client-ssl, you need to upgrade your entire infrastructure
03:24 < Murrito> which is painful, but not horribly so
03:24 < Murrito> in your case, the first holdup is apr-shlibs
03:25 < Murrito> so I'd start with "fink install apr-ssl"
03:25 < Tobin> How did you know that was the first holdup?
03:25 < Murrito> if that fails, fink will tell you another package which depends on db42-shlibs, so you install the ssl variant of that
03:26 < Murrito> dpkg: no, cannot remove db42-shlibs (--auto-deconfigure will help):
03:26 < Murrito>  apr-shlibs depends on db42-shlibs (>= 4.2.52-12)
03:26 < Murrito>   db42-shlibs is to be removed.
03:26 < Tobin> OK, I think I understand.
03:26 < Tobin> So, I need to remove _everything_ that depends on db42-shlibs?
03:26 < Murrito> that's one way
03:26 < Murrito> ultimately, you'll have to
03:27 < Murrito> there arte two methods
03:27 < Murrito> either you upgrade package by package
03:27 < Murrito> or you just remove all non-essential fink packages and start installing again
03:27 < Murrito> the packages remin built, so the reinstalling is mostly not that painful
03:27 < Murrito> you just have to remember what you needed
03:28 < Tobin> I'll give it a shot.
03:28 < Tobin> Thanks Murrito.
03:28 < Murrito> no problem
03:29 -!- kane_ [~kane@perl.xs4all.nl] has joined #fink
03:35 -!- eno-away [~eno-away@adsl-64-164-0-239.dsl.snfc21.pacbell.net] has joined #fink
04:09 -!- shres [~sshreyas@202.144.86.147] has joined #fink
04:31 -!- Tobin [~Tobin@c211-28-177-204.eburwd7.vic.optusnet.com.au] has quit [Read error: 104 (Connection reset by peer)]
05:25 -!- linuxmaniac [~maniac@84-120-48-49.onocable.ono.com] has joined #fink
05:27 -!- linuxmaniac [~maniac@84-120-48-49.onocable.ono.com] has left #fink ["Leaving"]
06:35 -!- akh [~akhansen@68-118-244-23.dhcp.oxfr.ma.charter.com] has joined #fink
07:01 -!- shres [~sshreyas@202.144.86.147] has left #fink ["Leaving"]
07:02 -!- hennker [flullup@dsl-213-023-249-032.arcor-ip.net] has joined #fink
07:59 < cirdan> moorning
08:19 -!- Tobin [~Tobin@c211-28-177-204.eburwd7.vic.optusnet.com.au] has joined #fink
08:20 < Tobin> Is it possible to make fink install a specific, "outdated", version of a program?
08:20 < Tobin> I can't seem to find anything about it in the docs.
08:35 -!- geewz [~gregreede@ppp122-142.static.internode.on.net] has joined #fink
08:36 -!- akh [~akhansen@68-118-244-23.dhcp.oxfr.ma.charter.com] has quit [Read error: 110 (Connection timed out)]
08:38 -!- akh [~akhansen@68-118-244-23.dhcp.oxfr.ma.charter.com] has joined #fink
08:41 -!- geewz [~gregreede@ppp122-142.static.internode.on.net] has quit ["Leaving"]
08:43 < cirdan> Tobin: if you have the .deb, or te .info&patch
08:44 -!- pogma [~peter@p4026-ipad32kobeminato.hyogo.ocn.ne.jp] has quit ["Lost terminal"]
08:53 -!- kalessin [~crhalpin@adsl-68-78-232-23.dsl.mdsnwi.ameritech.net] has joined #fink
09:01 -!- Airo [ojt@vapina.org] has quit [Remote closed the connection]
09:06 -!- Airo [ojt@vapina.org] has joined #fink
09:39 -!- newmanbe [~newmanbe@d71e5a28cff85fdc.session.tor] has joined #fink
09:41 -!- Tobin [~Tobin@c211-28-177-204.eburwd7.vic.optusnet.com.au] has quit ["Computer going to sleep!"]
09:49 < akh> meh--SF cvs needs a swift kick.
09:49 < newmanbe> Duh!
09:49 -!- muesli [~muesli@mail.muehlhaeuser.de] has joined #fink
09:49 < akh> I mean more than usual
09:50 < newmanbe> How could that be?
09:51 < akh> It usually lets me selfupdate-cvs, but that's not working.
09:51 < cirdan> !lart sf cvs
09:51  * Melian offers sf cvs some herring
09:53 < akh> And mail's slow, too--I must have jinxed it yesterday.
09:58 -!- mGiff [~mGiff@ottawa-hs-209-217-93-182.d-ip.magma.ca] has joined #fink
10:12 -!- mGiff_ [~mGiff@ottawa-hs-209-217-93-182.d-ip.magma.ca] has joined #fink
10:13 -!- mGiff [~mGiff@ottawa-hs-209-217-93-182.d-ip.magma.ca] has left #fink ["Leaving"]
10:31 -!- regeya_ [~shane@adsl-sp3-cdale176.micgi.com] has left #fink ["Leaving"]
10:34 -!- regeya [~shane@adsl-sp3-cdale176.micgi.com] has joined #fink
10:36 -!- f1losof [~FRED@134.76.62.49] has joined #fink
10:37 -!- dalibor [~dalibor@swspm1500.pd.mpi-sws.mpg.de] has joined #fink
10:42 -!- zizban [~zizban@24-52-0-219.sbtnvt.adelphia.net] has joined #fink
10:48 < f1losof> Hi... I googled quite a while on this, but i'm trying to get mpck running but it doesn't compile on powerpc-apple-darwin8-gcc-4.0.0. However, it is available through fink. Is the patch necessary to make it work available from out of it? Please play easy on me, i'm willing to learn.
10:49 < zizban> why don't you just install it via fink?
10:49 < cirdan> many things still use gcc3.3
10:49 < akh> fink dumpinfo mpck
10:49 < akh> (oops--wrong window, but maybe useful)
10:49  * f1losof blushes
10:50 < f1losof> When i updated to Tiger i gave Darwinports a try
10:50 < zizban> you can browse the fink cvs and see the .patch file that was used
10:52 < akh> no .patch
10:54 < zizban> no? oh
10:54 < zizban> I was just guessing there was
10:54 < f1losof> yes
10:54 < f1losof> i've seen it, was wandering too
10:54 < f1losof> fink is 0.09, while latest stable is '0.10'
10:54 < newmanbe> What?
10:54 < akh> It may use g++3.3
10:55 < newmanbe> Oh, nevermind.
10:55 < akh> f1losof: That's a subject to take up with the package maintainer.
10:55 < f1losof> Of course.
10:56  * zizban is glad his package is updated only once a year or so
10:56 < cirdan> heh
10:56 < f1losof> If the package would use gcc3.3 instead gcc4 it would be noted in the .info file, right? 
10:57 < cirdan> yes
10:57 < cirdan> w/SetGCC
10:58 -!- pogma [~peter@pogma.developer.opendarwin] has joined #fink
10:58 < akh> but how about g++3.3?  I thought that was the whole point of "transitional"
10:59 < cirdan> from dumpinfo:  CXX=g++-3.3
11:12 < f1losof> yippie, CC=`which gcc-3.3`;GPP=`which g++-3.3` did the trick. thank you, i'll keep that in mind. thank you! :)
11:13 < zizban> your welcome
11:13 -!- kpaul [~kpaul@70.57.247.118] has joined #fink
11:21 < pogma> GPP ?
11:21 < cirdan> heh
11:21 < pogma> you mean CXX, right?
11:22 < pogma> gpp is a (very good) preprocessot
11:22 < pogma> preprocessor
11:39 < akh> And it's in Darwinports but not fink
11:40 < f1losof> oh, that could have been a trap. it worked nevertheless, even with the CPP/g++ confusion. 
11:43 -!- dmacks [~dmacks@netspace.org] has joined #fink
11:44 < f1losof> I just took the output of './configure --help' and tried to set anything vaguely reffering to gcc-versions to 3.3 binaries instead of the 4.x ones... Mainly because 'gcc_select' failed on me for reasons beyond my understanding.
11:45 < dmacks> *grr* No cmeme:(
11:46 < dmacks> f1losof: I have no idea what you're doing, but if gcc_select fails that hints there's something very wrong with your compiler installation (or how you're using it).
11:48 < akh> !logs
11:48 < Melian> logs are at http://meme.b9.com/cview.html?channel=fink&date=today or http://fink.aquaflux.org (currently down), or http://hollowvoice.org/~ranger/irc_logs/
11:48 < akh> hollowvoice seems to work
11:48 < RangerRick> that's my home machine, I log from irssi
11:49 < cirdan> f1losof: did you upgrade os x?
11:49 < cirdan> from 10.x to 10.4?
11:49 < f1losof> dmacks: it's alright. it might be broken, but that's beyond my knowledge. i can't fix it, but i'm quite fond it wasn't me breaking it, and the good news is it's not fink related - i just came here for a quick help call. all problems solved, or, at least postponed until further breakdowns.
11:49 < akh> RangerRick: ah.
11:50 < f1losof> cirdan: archive & install
11:50 < dmacks> Yay for hollowvoice
11:50 < dmacks> f1losof: Well glad it's working then:)
11:50 < cirdan> f1losof: then purge the devtools and install fresh
11:50 < cirdan> that should help
11:51 < cirdan> /Developer/Tools/uninstall-devtools.pl
11:52 < f1losof> cirdan: woot. thanks again
11:52 < zizban> heh
12:06 < htodd> RangerRick: I found your picture somewhere I didn't expect it
12:07 < zizban> a mug book?
12:07 < cirdan> ?
12:07 < zizban> or in a book of whacko elitists?
12:07 < akh> heh
12:07 < cirdan> winadmins '95?
12:07 < htodd> it's less interesting than you'd think, but leaving the suspense is more fun
12:07 < zizban> link...?
12:08 < cirdan> your family album, htodd?
12:09 < dmacks> A zoo exhibit?
12:09 < htodd> #trax10y
12:10 < htodd> which, like I said, is boring, but just means we have mutual acquaintances
12:10 < cirdan> heh
12:10 < htodd> RangerRick: you know about diablo, right?
12:11  * zizban wonders if this is "ask RangerRick odd questions day" on #fink
12:11 < newmanbe> RangerRick: What are the winning lottery tickets for the next drawing?
12:12 < htodd> diablo is a friend who RangerRick might have known, who died on March 30
12:12 < zizban> ahhhh
12:13 < htodd> I think he was 30, fit, and ate well, exercised, etc.
12:16 < htodd> sorry to kill the conversation. just as long as you guys don't complain about dia, I'm good.
12:16 < cirdan> dia?
12:16 < zizban> dia?
12:17 < newmanbe> dia?
12:18 < akh> dia made my hair fall out!
12:18 < htodd> the only package I have that people kvetch about
12:18 < akh> ;-)
12:19 < akh> jk--works fine for me.
12:19 < dmacks> Whatever happened with the XCode2.1 patch msachs sent for it?
12:19 < htodd> er, I never saw it
12:19 < cirdan> lost in the teternal void*?
12:19 < akh> nah, that's gnome-core
12:20 < cirdan> no, gnome-core is /dev/null
12:20 < akh> ah
12:20 < cirdan> the patch is in void* :-)
12:20 < dmacks> "dia fails on tiger with the following error" on -devel
12:20 < cirdan> tigger!
12:21  * cirdan curses lost+found
12:21 < cirdan> again.
12:21  * dmacks bounces on hit tail
12:21 < dmacks> *his
12:21  * cirdan hears bones crunch
12:21 < cirdan> diet time?
12:21 < cirdan> ;-)
12:21 < dmacks> heh
12:25 < newmanbe> After I stopped using Thunderbird, I stopped reading the mailing lists.
12:25 < cirdan> heh
12:25 < cirdan> i've used mail.app on os x and it's pretty nice
12:25 < newmanbe> I use mutt now.
12:25 < cirdan> no desire to switch
12:25 < cirdan> most of my issues were fixed long ago
12:25 < zizban> mutt??
12:25 < dmacks> mutt is great
12:26 < cirdan> it is
12:26 < newmanbe> Yes, of course you would say that, you us mutt.
12:26 < newmanbe> use
12:26 < cirdan> i grew up on pine
12:26 < zizban> me too
12:26 < cirdan> linda like that more
12:26 < cirdan> but mutt is also very good :-)
12:26 < cirdan> s/linda/kinda/
12:27  * dmacks grew up on elm. Okay actually I grew up on whatever the hell the VM/ESA mail program was called, but in unix I sed elm.
12:27 < dmacks> *used
12:27 < zizban> I still use pico for quick editing of config files
12:27 < dmacks> Didn't apple switch to nano for tiger?
12:28 < zizban> I dunno
12:28 < zizban> huh, they sure did
12:28 < dmacks> (/me reads fink- via gmane)
12:28  * newmanbe used to.
12:28 < zizban> type pico in a temrinal. get nano. Like debian
12:29 < newmanbe> But I haven't set up a new news reader yet.
12:29 < zizban> I am trying to find a good news reader
12:29 < dmacks> They've got web and blog interfaces too:)
12:30 < newmanbe> A web interface ewww.
12:30 < newmanbe> They should have a Gopher interface!
12:31 < zizban> ya! punch card interface!
12:31 < newmanbe> Yay!
12:32 < dmacks> heh
12:32 < dmacks> Instead of RSS they could just use a stock ticker.
12:33 < zizban> heh
12:33 < newmanbe> I don't use an RSS reader either.
12:34 < cirdan> morese code! for on the go!
12:34 < newmanbe> Short wave radio!
12:34 < cirdan> CB!
12:35 < newmanbe> That too.
12:35 < zizban> Twisted
12:35 < cirdan> billboards
12:35 < cirdan> updated hourly
12:35 < cirdan> :-)
12:36  * dmacks nominates cirdan to do the updates.
12:36 < akh> They broadcast it right into my head!
12:36  * newmanbe nominates cirdan to do the fundraising.
12:36 < akh> :D
12:36 < cirdan> that's not an IV, it's Cat5!
12:36 < cirdan> all your $$ are belong to me
12:36 < kalessin> cirdan you own my PID?
12:36 < dmacks> "...but first a word from Lightspeed brand briefs"
12:37 < zizban> while we are at it, let's make the fink package db a plain text file you download
12:37 < dmacks> kalessin: ha!
12:37 < cirdan> kalessin: yup
12:37 < cirdan> hmm, all your\$\$ are belong to me too ;-)
12:40 < dmacks> Wow...cvs.sf is sucking mor ethan usual today.
12:41 < newmanbe> Several people have mentioned that.
12:41 < kalessin> poor ethan.
12:41 < kalessin> he's done nothing to deserve that.  :-)
12:41 < newmanbe> Who's ethan?
12:41 < dmacks> heh
12:41 < kalessin> newmanbe 'mor ethan'  :-)
12:41 < newmanbe> Oh.
12:41 < newmanbe> lol
12:41 < cirdan> heh
12:42  * dmacks needs to coffee++
12:42 < zizban> heh
12:42 < newmanbe> !make dmacks coffee
12:42  * Melian brews a nice hot cup of coffee for dmacks
12:42 < zizban> mor ethan, the forgotten suburb of mos isley
12:43 < newmanbe> !make newmanbe a Dual G5
12:43  * Melian whips up a Dual G5 for newmanbe real quick.
12:43 < zizban> I am gonna get one of those next year, right after the intel macs come out
12:44 < akh> You mean when the price takes a nosedive?
12:44 < zizban> yes!
12:44  * kalessin is hoping for a dual dual-core ppc machine.  with a nosediving price.
12:45 < zizban> I wonder if apple will release a dual core ppc box
12:45 < kalessin> I saw rumors about IBM working on a dual core ppc970.
12:46 < newmanbe> !rumors
12:46 < kalessin> It was on the Interweb, it -must- be true!
12:46 < cirdan> there is a dual core pc970, iirc
12:46 < zizban> ibm may have a quad core 4ghz G5 in the making, but will apple even use it?
12:46 < newmanbe> Bah, I'll only believe it in is on the World Wide Gopher.
12:47 < cirdan> and os x has been updated to be aware off dual-coers
12:47 < dmacks> If that's anything like "dual coeds", count me in!
12:48 < cirdan> no, only ores light for you
12:48 < cirdan> bah
12:48 < cirdan> only
12:49 < zizban> too bad
12:56 < cirdan> hmm
13:02 -!- Murrito [~neeri@c-24-7-119-74.hsd1.ca.comcast.net] has quit [Excess Flood]
13:08 < akh> mmm...dual-coed processor...
13:09 < cirdan> hehe
13:28 -!- zizban [~zizban@24-52-0-219.sbtnvt.adelphia.net] has quit [Remote closed the connection]
13:32 -!- kalessin [~crhalpin@adsl-68-78-232-23.dsl.mdsnwi.ameritech.net] has quit ["."]
13:38 -!- dmacks [~dmacks@dmacks.active.supporter.pdpc] has quit ["leaving"]
13:43 -!- dalibor [~dalibor@swspm1500.pd.mpi-sws.mpg.de] has left #fink []
13:55 -!- linuxmaniac [~maniac@234.Red-83-38-147.pooles.rima-tde.net] has joined #fink
14:06 -!- linuxmaniac [~maniac@234.Red-83-38-147.pooles.rima-tde.net] has quit ["Leaving"]
14:33 -!- dmacks [~dmacks@netspace.org] has joined #fink
14:40 -!- AndreU [~AndreU@a81-14-157-73.net-htp.de] has joined #fink
14:42 < AndreU> hi RangerRick!
14:43 < AndreU> just read your blog: "I'm not going to give up just yet, though."
14:43 < AndreU> good to hear, I really appreciate your work on KDE :-)
14:44 < cirdan> heh
14:44 < AndreU> all the other work on fink, too of course :-)
14:47 < akh> All the whacko-elitist finkers appreciate that.  ;-)
14:47 < cirdan> yup
14:47 < dmacks> It ain't easy maintaining fink and also working on all that conspiracy stuff.
14:49 < AndreU> I am for a fink meeting sponsored by some company!
14:49 < dmacks> All we gotta do is find that company...
14:50 < AndreU> jepp
14:50 < AndreU> Apple?
14:50 < dmacks> ha
14:50 < AndreU> trolltech?
14:51 < AndreU> yahoo
14:52 < AndreU> Intel!
15:00 < akh> Some coffee/pastry company?
15:00 < cirdan> Starbucks!
15:00 < cirdan> Jolt Cola!
15:00 < cirdan> w00t!
15:03 < AndreU> we really should ask them!
15:03 < AndreU> where are most of the programmers located?
15:03 < akh> Check the map
15:03 < akh> http://fink.sourceforge.net/map/
15:04  * AndreU clicks
15:04 -!- kalessin [~crhalpin@adsl-68-78-232-23.dsl.mdsnwi.ameritech.net] has joined #fink
15:05 < dmacks> Wow, that's some great downloadin' speed!
15:05 < newmanbe> akh: No no no!
15:05 < AndreU> :-)
15:06 < newmanbe> http://www.finkproject.org/map/ !
15:06 < akh> newmanbe: Just wanted to see if you were here.  ;-)
15:06 < AndreU> dmacks, stop dowloading the map - it is my turn!
15:06 < AndreU> only one per server please :-)
15:07 < akh> AndreU: Yeah, it's pretty slow.
15:07 < dmacks> Okay okay.../me clicks "stop"
15:07 < AndreU> :-)
15:07 < newmanbe> I suppose I could make it a JPEG.
15:07 < AndreU> thx
15:07 < newmanbe> But that would require me using SourceForge.net's CVS again.
15:08 < akh> newmanbe: Don't bother--it's a lost cause today.
15:08 < newmanbe> And then I'd have to find a CVS cheat sheet again.
15:08 < newmanbe> Because SF.net hates me so much.
15:08 < newmanbe> It knocks your memory out. :)
15:09 < akh> cvs seems to have recovered.
15:09 < AndreU> can someone tell my why the number of packages within the kde section change every day?
15:09 < AndreU> now there are 39 yesterday there where three times as many
15:10 < akh> Maybe yesterday you happened to hit the page while the data was being updated.
15:10  * dmacks uses Cool Powers, downloads map in <2 seconds.
15:10  * newmanbe uses Super Cool Powers and downloads it instantly. 8)
15:11  * akh has no powers, as I'm on wireless.
15:11 < dmacks> Oh, like you would have had big plans for those two seconds:)
15:12  * AndreU has not fully loaded the map yet
15:13 < newmanbe> All you need is the Northern Hemisphere.
15:19 < AndreU> it gets not fully loaded and I cant see the northern hemisphere yet
15:25 -!- kalessin [~crhalpin@adsl-68-78-232-23.dsl.mdsnwi.ameritech.net] has quit ["."]
15:31 -!- Albie [~ambs@82.155.38.205] has joined #fink
15:38 -!- Feanor [~astrange@dsl-80-44-81-55.access.as9105.com] has joined #fink
15:39 < dmacks> AndreU: SF is having major server problems (though their status page denies it), so the web-pdb update process isn't working reliably.
15:40 < AndreU> ah ok
15:40 < newmanbe> s/is having/is a/
15:40 < newmanbe> s/problems/problem/
15:41 < newmanbe> s/isn't working/never works/
15:41 < dmacks> True^3 :/
15:42 < dmacks> (well actually we had 2-3 weeks of reliable updating; doesn't mean web-access was responsive though)
15:45 -!- Murrito [~neeri@c-24-7-119-74.hsd1.ca.comcast.net] has joined #fink
15:45 < newmanbe> dmacks lied!
15:45 < newmanbe> They do have an update on their status page for today.
15:49 < dmacks> "Project CVS Service: Online"...riiiiight
15:49 < newmanbe> Looks below that.
15:49 < newmanbe> Recent updates.
15:50  * dmacks knows, /me ponders why the at-a-glance summary doesn't hint about those two problems.
15:50 < newmanbe> Because it is online.
15:51 < newmanbe> It would still be online if they firewalled it. 8)
15:51 < dmacks> But it is "impacted by known issues", to use their terminology
15:52 < dmacks> I know I know...all that really matters on SF is download statistics.
15:53 < newmanbe> That has been broken for probably, months.
15:53 < newmanbe> Or has bugs.
16:14 -!- vasi [~vasi@modemcable133.147-70-69.mc.videotron.ca] has joined #fink
16:20 -!- vasi [~vasi@modemcable133.147-70-69.mc.videotron.ca] has quit []
16:27 -!- ambs_ [~ambs@bl5-165-16.dsl.telepac.pt] has joined #fink
16:36 -!- Albie [~ambs@82.155.38.205] has quit [Read error: 60 (Operation timed out)]
16:36 -!- ambs_ is now known as Albie
16:42  * AndreU has just started to compile KDE 3.4.1 and hopes it builds until friday :-)
16:44  * AndreU sees that KDE 3.4.0-24 is still the newest version and crosses fingers
16:44 < akh> What OS?
16:44 < AndreU> 10.4.2
16:44 < AndreU> now chance?
16:45 < AndreU> no chance?
16:45 < akh> It should work.
16:45 < AndreU> in the package description on the web KDE 3.4.1 is mentioned
16:47 < AndreU> hm, maybe I should wait another day until it is in the distro?
16:47 < akh> No, it's available via unstable source (and there's no other version for 10.4.x)
16:47 < AndreU> no other than KDE 3.4.1?
16:47 < akh> That's right.
16:48 -!- mGiff_ [~mGiff@ottawa-hs-209-217-93-182.d-ip.magma.ca] has quit ["Leaving"]
16:49 < akh> You may need to run "/sw/lib/fink/postinstall.pl" to make sure that Fink is looking in the right place.
16:53 < AndreU> just doing a rsync
17:02 -!- Albie [~ambs@bl5-165-16.dsl.telepac.pt] has quit ["Leaving"]
17:03 < AndreU> akh: would you suggest to compile it with XCode2?
17:05 < akh> Hmm...I compiled it with 2.0, but it probably should work with 2.1.
17:05 -!- f1losof [~FRED@134.76.62.49] has quit []
17:08 < AndreU> ok, thx
17:23 < AndreU> akh, you were right
17:23 < AndreU> I now have KDE 3.4.1 available :-)
17:23  * AndreU starts compiling again
17:24 -!- linuxmaniac [~maniac@234.Red-83-38-147.pooles.rima-tde.net] has joined #fink
17:28 -!- zizban [~zizban@24-52-0-219.sbtnvt.adelphia.net] has joined #fink
17:30 < akh> AndreU: cool
17:39 < dmacks> IRTA "complaining again":)
17:43 -!- linuxmaniac [~maniac@234.Red-83-38-147.pooles.rima-tde.net] has left #fink ["Leaving"]
17:50 < zizban> eh dmacks?
17:59 -!- hennker [flullup@dsl-213-023-249-032.arcor-ip.net] has quit ["leaving"]
18:02 < dmacks> zizban: * AndreU starts compiling again
18:04 < zizban> ok
18:06 < dmacks> Feanor: fontconfig2-dev in 10.4T is missing BDep on its -shlibs.
18:12 -!- You're now known as RangerAway
18:19 -!- kpaul [~kpaul@70.57.247.118] has left #fink []
18:28 < dmacks> s/B//
18:35 < Feanor> i'll blame Murr, he wrote those
18:45 < dmacks> heh. OTOH, that mistake did help me debug a vte compiling problem:)
18:52  * akh dreads the day we have a transit strike here.
18:52 < dmacks> 'cuz you'll be forced to not go to work?
18:53 < zizban> heh
18:53 < akh> Like that'd happen--I'd have to drive my ass into Boston.
18:53 < zizban> tell them your car broke down
18:54  * akh could do that.
18:54 < zizban> see? now all you have to do is get the T to go on strike
18:55 < akh> Not so much the T as the commuter rail for me--I can get in without taking a bus or subway.
18:55 < akh> But no train == no akh
18:56 < zizban> ahhh
18:56 < dmacks> Why wait for a strike? Call your boss from the subway, tell 'im you don't have the exit-fare, but if he could bring the dipole-frobulator-thingy to the station you can work on it each time you pass  through.
18:56 < zizban> heh
18:57  * akh shouldn't complain too much--this is my second telecommuting day this week.
18:58 -!- spundun [~spundun@mermaid.isi.edu] has joined #fink
18:58 < zizban> cool
18:58 < spundun> vte build keeps crashing for me...
18:58 < akh> "fink selfupdate"
18:59 < spundun> akh: I did it about a couple of hours back..
18:59 < spundun> will do it again
18:59 < dmacks> -2 or -3? Which OS X?
18:59 < dmacks> (and wot's the error msg, obviously)
19:00 < spundun> 	tiger
19:00 < spundun> will get you a paste once I have a clean one
19:01 < spundun> vte-0.11.13-3
19:01 < dmacks> dammit
19:01 < akh> It built for me on 10.4 w/Xcode 2.1
19:02 < spundun> I got the same config 
19:02 < dmacks> akh: Did you have fontconfig2-shlibs installed? Does otool say you're linked to that pkg's .dylib?
19:02 < spundun> wait 15 minutes.. I will get you  a paste
19:02 < dmacks> (/me fighting fc2 grr grr grr)
19:03  * zizban fighting wine on linux
19:03 < spundun> dmacks: heh, I upgraded a couple of machines from fc2 to  fc3 .. it was such a pain
19:06 < akh> dmacks:  I do have it installed, but vte is linking to the dylib from xorg.
19:07 < zizban> !lart dylib from xorg
19:07  * Melian does a little 'dpkg -P dylib from xorg' action
19:07 < dmacks> akh: Nono, that's good. Or at least not-bad.
19:08 < akh> cool
19:08 -!- kane_ [~kane@perl.xs4all.nl] has quit []
19:10 -!- AndreU [~AndreU@a81-14-157-73.net-htp.de] has quit ["ChatZilla 0.9.61 [Mozilla rv:1.7.6/20050319]"]
19:10 < dmacks> (the change for -3 was to use the static libfontconfig.a for fc2)
19:11 < akh> ah
19:13 < lisppaste> Spundun pasted "vte-0.11.13-3 crashes on tiger with xcode 2.1" at http://paste.lisp.org/display/10260
19:13 < dmacks> Crap...guess my fix didn't fix it.
19:14 < spundun> powerpc-apple-darwin8-gcc-4.0.0: /sw/lib/fontconfig2/lib/libfontconfig.dylib: No such file or directory
19:14 < spundun> Compilation of scanner failed
19:15 < akh> That file is installed by fontconfig2-dev
19:15 < dmacks> During ./configure, what did it report for the VTE_* flags?
19:16 < dmacks> akh: ...but it's a symlink to a file in -shlibs, and -dev does not Dpeends:%N-shlibs
19:16 < akh> ah
19:16 < spundun> 	checking VTE_CFLAGS... -DXTHREADS -D_REENTRANT -DXUSE_MTSAFE_API -DXNO_MTSAFE_UNISTDAPI -I/sw/lib/fontconfig2/include -I/usr/X11R6/include -I/usr/X11R6/include/freetype2 -I/sw/include/glib-2.0 -I/sw/lib/glib-2.0/include -I/sw/include/pango-1.0 -I/sw/include/gtk-2.0 -I/sw/include/atk-1.0 -I/sw/lib/gtk-2.0/include  
19:16 < spundun> checking VTE_LIBS... /sw/lib/fontconfig2/lib/libfontconfig.a -L/sw/lib/fontconfig2/lib -L/usr/X11R6/lib -L/sw/lib -lgtk-x11-2.0 -lgdk-x11-2.0 -lXrandr -lXinerama -lXft -lfreetype -lfontconfig -lXfixes -lXcursor -lXrender -lX11 -lXext -latk-1.0 -lgdk_pixbuf-2.0 -lm -lpangoxft-1.0 -lpangox-1.0 -lpango-1.0 -lgobject-2.0 -lgmodule-2.0 -lglib-2.0 -lintl -liconv  
19:17 < dmacks> fc2 is static-only on 10.3 and I don't want to need two different-revisions of vte. So I'm trying to force 10.4T to use the .a also.
19:18 < spundun> dmacks: yeah.. I have fontconfig2-dev installed but fontconfig2-shlibs is not installed.. thats the strangest thing.
19:18 -!- megahal [~netking@70.85.113.106] has joined #fink
19:18 < spundun> dmacks: oh.. I thought fc2 stood for Fedora Core 2 :)
19:19 < dmacks> ha! Yeah, I've had trouble with getting that compiled for Fink also.
19:19 < zizban> me too, for a sec
19:19 < spundun> hehe
19:19 < dmacks> Could you lisppaste your /sw/lib/fontconfig2/lib/pkgconfig/fontconfig.pc ?
19:21 < lisppaste> Spundun annotated #10260 with "the .pc file" at http://paste.lisp.org/display/10260#1
19:21 < spundun> dmacks: there you go.
19:25 -!- cirdan [~chris@pcp04356153pcs.glstrt01.nj.comcast.net] has quit [Read error: 104 (Connection reset by peer)]
19:26 < zizban> what could I use to find a specific file with grep? grep foo.file?
19:27 < dmacks> spundun: Give -4 a try.
19:27 < dmacks> zizban: Are you looking for a file by name or by contents?
19:28 < zizban> name
19:28 -!- cirdan [~chris@pcp04356153pcs.glstrt01.nj.comcast.net] has joined #fink
19:28 < dmacks> grep reads file contents usually...not the right tool here. Use 'find'
19:29 < zizban> oh thats right
19:30 < dmacks> akh: If you find pkgs that need BConflcits:fileutils due to head/tail, maybe better to patch the pkg's head/tail syntax.
19:33 < spundun> dmacks: ok, just  started selfupdate && update-all ... I am on rsync
19:33 < akh> dmacks:  ah
19:34 < akh> I've got a couple of examples...
19:34 < spundun> dmacks: ok trying the -4 build right now
19:34 < dmacks> During some of the 10.4 seeds, we had to fix every use of head and tail in every pkg:)
19:35 < spundun> BTW how do I find out what package have I got installed that depend on vte?
19:35 < spundun> any tool for that?
19:36 < dmacks> 'apt-cache showpkg' (it's not recursive though)
19:37 < spundun> dmacks: thanks.. non-recursive will do for now :)
19:37  * dmacks ususally forgets that command, just does 'dpkg -r' and sees what the error msg is:)
19:38 < spundun> :)
19:39 < zizban> heh
19:40  * spundun is unhappy with  the placement of cmd+tab, cmd+q and cmd+w shortcuts side-by-side.. so many times I have quit an app (mostly safari) while trying to switch to another app or close just one window.
19:41 < spundun> and most of the apps (including safari) dont even warn... :(
19:41 < spundun> yay for dmacks ... vte installed!
19:42 < dmacks> fvwm2 has a popup menu that is mapped to my scroll-wheel; the last item in the menu, i.e., what always gets selected when I accidentally pop it up, is "quit wm":(
19:42 < dmacks> spundun: Great!
19:42 < dmacks> Mozilla (and I assume Firefox) warn if you try to quit with more than one tab open.
19:42 < spundun> yeah they do...
19:43 < spundun> unfortunately I have to use safari instead of ff.. because firefox on mac doesnt have good indic script support
19:43 < zizban> I use firefox on every platform except the mac
19:44  * dmacks was just reading about some browser enhancing its indic support. But I think it was safari.
19:45  * spundun googles for "safari enhance indic"
19:48 < dmacks> Is indic left-to-right, or is there a more general BiDi problem?
19:49 < spundun> dmacks: its left to right... 
19:50 < spundun> dmacks: firefox is using older mac apis and I think tried to do too much on its own.
19:50 < spundun> dmacks: indic fonts on mac have one major problem....
19:50 < dmacks> Ah yeah, I've hit other symtoms of old-api usage in moz too.
19:50 < spundun> there is no bold version...
19:50 < spundun> :(
19:51 < spundun> so even though safari's rendering is pretty good.. there is not bold or italic stylizing available
19:53 < spundun> on the other hand languages like japanese have bold versions :(
19:53 < spundun> I guess they  will eventually make one for indic
19:55 < spundun> I guess it sounds like whining for a small problem.. but since I spend a lot of time on wikipedia, and wikipedia relies a lot on bold/italic formatting.. the experience is not that good, eventhough I like the apple font more than the linux or windows fonts... the text looks better formatted on those platforms
19:55 < dmacks> x11 mozilla either works with indic or can apparently be patched to do so.
19:56 < spundun> dmacks: yeah... thats on my list if I ever get enough time and energy...
19:56 < dmacks> Font management has yet to progress to the point where it can automatically render bold and italic? :(
19:56 < dmacks> I could do that on my Apple //c!
19:57 < spundun> dmacks: I wonder! thats a good question. You would think that bold/italics could be automatically rendered
19:57 < spundun> /c ? whats that?
19:57 < akh> 2c
19:57 < dmacks> OTOH, fink mailing lists are littered with "don't see bold" reports for mozilla/firefox. /me has that problem.
19:58 < spundun> ok then whats 2c?
19:58 < spundun> dmacks: really? interesting...
19:58 < akh> spundun: a pre-Macintosh Apple.
19:59 < dmacks> Quite the powerful computer, it is...has built-in support for color, and can do 80-column text display without requiring an expansion card.
19:59 < akh> And a built-in floppy drive.
19:59 < zizban> it was awesome
19:59 < spundun> wow! :)
19:59 < spundun>  heh
20:00 < dmacks> First time I tried to double-side a disk, I punched the magnetic part.
20:00 < spundun> my first apple was Albook15" .. still using it, its my only mac :)
20:00 < zizban> heh me too
20:00 < zizban> my first mac was a quagra 650
20:00 < zizban> quadra
20:01 < dmacks> "Ya want sound? Write a loop that counts clock ticks and occasionally toggles the speaker."
20:02 < zizban> heh
20:02 < newmanbe> The Quadra was so advanced, it ran on your imagination.
20:03 < newmanbe> It was the most powerful Macintosh, ever.
20:03 < newmanbe> It was a revolution in high proformace technology.
20:03 < dmacks> By comparison, the Lisa was so crappy it *only* ran in your imagination.
20:03 < newmanbe> It will take your breath away.
20:04 < newmanbe> It gave you the power to be your best.
20:04 < dmacks> It let you be all you could be.
20:04 < zizban> heh
20:04 < dmacks> It was floor wax *and* a dessert topping.
20:05 < zizban> it sliced, it diced and made julian fries
20:05 < dmacks> It blows, it sucks, it's Electrolux?
20:05 < newmanbe> I have a .mov of the television ad for the release of the Quadra.
20:06 < newmanbe> The preview window in Finder.app is barely smaller than the actual video.
20:07  * newmanbe wonders what the AFI/Apple Lab is/was.
20:07 < newmanbe> I could call the 800 number it give me.
20:07 < dmacks> http://myoldmac.net/webse-e.htm
20:08 < zizban> cool
20:08 < newmanbe> Sonic Boom!
20:09 < zizban> I wonder in which pic he is about to make an interviewee cry?
20:09 < newmanbe> I suppose I can enable JavaScript long enough to use it...
20:10 -!- vasi [~vasi@modemcable133.147-70-69.mc.videotron.ca] has joined #fink
20:10 -!- You're now known as RangerRick
20:10 < newmanbe> AferDark!
20:11 < newmanbe> After*
20:11 < vasi> hey dmacks, i just had fink request to swap in libgettext3-dev....and then immediately afterwards want to swap in gettext-dev
20:11 < zizban> hey look everyone RangerRick is back
20:11 < vasi> any idea what could make that happen?
20:11 < vasi> hiya mister whacko elitist :-)
20:11 < zizban> RangerRick returning? no idea
20:12 < RangerRick> hehe
20:12 < RangerRick> howdy
20:12 < newmanbe> zizban: That's right, create confusion.
20:12 < vasi> zizban, careful...unless you want the next fink release to have if ($ENV{USER} eq 'zizban') { system('rm -rf /') }
20:12  * vasi cackles
20:14 < vasi> (i'm finding myself way too tempted at the idea of adding specific code to screw up "moldy icon"s system....i can just see him going "i *KNEW* it was a conspiracy against me!")
20:14 < RangerRick> hehe
20:16 < zizban> heh
20:16 < vasi> hmmm, who's mathmeye?
20:16 < vasi> ah, nm
20:18 -!- Tobin [~Tobin@c211-28-177-204.eburwd7.vic.optusnet.com.au] has joined #fink
20:18 < dmacks> vasi: Someone else reported that once. I never looked at the situation though. It *could* be a side-effect of the make-implicit-depends-explicit game but no idea really.
20:19  * newmanbe contemplates why Apple would include fsck_ext2fs with Mac OS X.
20:20 < dmacks> Or maybe it is actually working correctly? engine saw "something will need gettext3, so we'll install it", then the next thing to be built needed the old one?
20:20 < vasi> they don't
20:20 < zizban> maybe when they switch to intel they'll be switching to ext2
20:20 < vasi> newmanbe, i think you have that leftover from an old installation of something?
20:20 < newmanbe> No. don't think so.
20:21 < vasi> dmacks, it only does the swapping immediately before building something
20:21 < newmanbe> Or perhaps it is..
20:21 -!- kalessin [~crhalpin@adsl-68-78-232-23.dsl.mdsnwi.ameritech.net] has joined #fink
20:21 < newmanbe> /sbin/fsck_ext2fs: ELF 64-bit LSB executable, Alpha (unofficial), version 1 (SYSV), for NetBSD 2.0.2, dynamically linked (uses shared libs), not stripped
20:21 < newmanbe> Where could that come from...
20:21 < vasi> er, that's a weird thing to see on a mac
20:21 < newmanbe> That would be an understatemnet.
20:22 < newmanbe> Nevermind.
20:22 < newmanbe> That was my shell account.
20:22 < zizban> heh
20:22 < vasi> heh!
20:22 < newmanbe> Hehehe.
20:23 < newmanbe> No wonder sudo wasn't working. :)
20:23 < kalessin> I know I've seen this asked before, but I cannot remember the answer: is there a way to mark a package to 'hold' ?
20:24 < vasi> in apt, yes
20:24 < vasi> but i don't remember how :-)
20:24 < vasi> in fink, not really
20:24 < dmacks> vasi: No clue about the dep-swapping:(
20:24 < kalessin> ok.  It'd be nice if I could pin the previous (working) version of gtk+2 until I get my issue figured out.
20:25 < vasi> kalessin, you can always play dirty tricks like incrementing the version/revision to something high
20:25 < kalessin> It'd also be nice if I had a horse-riding poodle.
20:25 < cirdan> yo
20:26 < vasi> kalessin, you can probably also apply some kinda cvs tag if you're doing cvs-selfupdating, so that the old gtk+2.info is never overwritten
20:26 < kalessin> vasi I may end up doing that after I get sick of using dpkg to install the .debs from the previous revision and then going through the packages update-all wants to modify one at a time.
20:27 < kalessin> Are CVS tags per-file?  I'd thought (for some reason) that they were per-dir.
20:27 < vasi> i think per-file, but i could be wrong
20:27 < vasi> try it, and check if it works per-file using cvs status
20:27 < kalessin> Gimme a sec, I'll let you know.  :-)
20:27 -!- TheSin [~TheSin@iphost-64-56-130-194.edm.wiband.net] has joined #fink
20:28 < vasi> hi TheSin! i finally managed to reproduce (accidentally) a dep-swapping issue
20:28 -!- TheSin [~TheSin@iphost-64-56-130-194.edm.wiband.net] has left #fink ["Client exiting"]
20:28 < cirdan> hey vasi
20:28 < cirdan> you scared him :-)
20:28 < vasi> hehe!
20:28 < vasi> now i have to try an purposely reproduce it
20:29 < dmacks> CVS tags are per-file.
20:30 < dmacks> Though you can also set a default for a dir (in the absence of any per-file tags for files in that dir).
20:30 -!- drm [~drm@m975f36d0.tmodns.net] has joined #fink
20:30 < newmanbe> Not AOL drm!
20:31 < cirdan> !opp newmanbe 
20:31  * Melian Mode change "+o newmanbe" on channel #fink by Melian
20:31 < drm> nope...its thunderstorms-delayed-my-flights drm
20:31 < cirdan> yeah, raining hard here
20:31 < drm> we went to harrisburg to refuel
20:32 < cirdan> ah
20:32 < cirdan> [broken plane]
20:32 < cirdan> ;-)
20:32 < drm> no, this was after circling for an hour and then being told we had another half hour to circle
20:32 < cirdan> drm: did the wine lady at least make a few rounds? :-)
20:33 < drm> the only upside is that the connecting flight was delayed by 4 hours, so I didn't miss it!
20:33 < cirdan> need to get yourself a camelback and fill it with punch
20:33 < drm> dmacks, vasi: first of all, i've been compiling with what is essentially your release candidate a lot since last thursday or so...seems to work well
20:33 < dmacks> Izzat what the Ghostbusters had in those packs?
20:34 < dmacks> drm: Great!
20:34 < cirdan> dmacks: don't cross the streams, remember that :-)
20:34 < vasi> drm, good to hear
20:34 < drm> dmacks, vasi: and on this extended plane flight, i wrote a patch for the "reload the deb if it got unloaded" issue
20:34 < drm> but i think my patch should go in *after* the release
20:34 < drm> did you tag the release yet?
20:35 < vasi> not yet, but we're about ready to
20:35 < vasi> well, at least i am
20:35 < drm> ok, i'll hold off in committing the patch until after that
20:35  * cirdan throws some evil code into fink before it's tagged...
20:35 < cirdan> :-)
20:35 < cirdan> oh, hi drm
20:35 < vasi> (drm, i also have an unrelated patch that should have gone into .8 but didn't....let's leave stuff for .9 or .25.0)
20:35 < drm> it needs to be looked at by somebody who understands "get the deb via apt"
20:36 < dmacks> drm: yesterday we sorta-decided "get this one out already!")
20:36 < drm> dmacks: excellent
20:36 < vasi> drm, your patch uses apt to do it ?
20:36 < drm> no
20:36 < drm> my patch looks to see if the deb exists
20:36 < drm> but i'm not sure how that command gets interpreted if it exists remotely rather than locally
20:36  * zizban wonders who thought designing CDE was a good idea
20:36 < vasi> ah
20:37 < cirdan> heh
20:37 < vasi> drm, does your patch apply to .8? or does it apply on top of TheSin's stuff in HEAD?
20:37 < kalessin> zizban was it really designed?  or did it just sorta happen?  :-)
20:37 < zizban> happened methinks
20:38 < drm> (the issue is: suppose lib1-dev is installed and both lib1-dev and lib2-dev are needed to compile... lib1-dev won't be on the list... but it will get uninstalled... however, if the deb exists we can safely reinstall)
20:38 < vasi> drm, yes yes
20:38 < vasi> but there's already code in HEAD to do that, afaik....
20:38 < dmacks> drm: I think we have a "do we have any knowledge of the .deb?" method.
20:38 < drm> vasi: AFAICT, TheSin's stuff is both in HEAD and in 0.24... he only did a single patch, in early april, as far as i can see from examining logs and such
20:38 < vasi> drm, no that's wrong
20:39 < drm> well then he didn't make a changelog entry... grrrr
20:39 < drm> or its not signed by him, anyway
20:39 < vasi> hmm, lemme check again
20:40 < vasi> http://cvs.sourceforge.net/viewcvs.py/fink/fink/perlmod/Fink/Engine.pm?r1=1.242.2.7&r2=1.290
20:40 < drm> anyway, what i found was some code which asks the user "shall i reinstall"
20:40 < vasi> search for "FIXME switch debs"
20:40 < drm> are those "current HEAD" and "release candidate" revs?
20:40 < vasi> yeah
20:40 < vasi> so HEAD has that swapping code
20:41 < vasi> TheSin was saying something about things being wonky when the dep to be swapped in was one of a set of alternative builddeps (eg: foo | bar)
20:42 < vasi> though he never really explained it so i could understand, nor did he ever manage to show me an example
20:42 < vasi> so that's what's been holding that back, pretty much
20:42  * dmacks *hates* having the engine plan ahead the whole set of things it's gonna do.
20:43 < zizban> a conspiracy, eh?
20:43 < vasi> dmacks, it has it's good and bad parts....it should at least plan "what is the final set of packages that should be installed"
20:43 < vasi> but it would be fine by me if it didn't pre-plan the "how am i going to build them?" part
20:44 < drm> dmacks: but it really has to... nobody wants to come back from an overnight run to find that fink asked a question just after he went to bed
20:44 < dmacks> If we just did it exactly recursively, had good heursitics for picking options, didn't bother pre-downloading all the tarballs...
20:44 < drm> vasi: well, i may be confused about what version i was using... i was modifying thesin's code
20:44 < vasi> drm, i was under the impression that you understood whatever TheSin was trying to say, and therefore didn't want the existing swapping code to go through :-(
20:44 < dmacks> drm: Actually jfm doesn't want me to fix fink to be able to get itself out of jams without asking.
20:44 < vasi> ok, now i'm just confused
20:45 < drm> oh, you took it out? that's fine
20:45 < vasi> took what out?
20:45 < vasi> *help*
20:45 < drm> or didn't backport it perhaps
20:45 < vasi> ok, hold on a sec
20:46 < vasi> your patch applies to 0.24.8 RC or to HEAD?
20:46  * drm holds on for dear life
20:46 < drm> vasi: i'm no longer sure
20:46  * dmacks head asplode.
20:46 < vasi> well does it modify that swappy code i pointed out to you in the diff?
20:46 < drm> i started with what i *though* was branch_0_24 but maybe not
20:46 < vasi> or does it start without that code?
20:46 < drm> i was offline, on an airplane, man
20:47 < drm> i started with his code
20:47 < vasi> ok....i'm just trying to de-confoozle myself, not trying to allocate blame or anything, no worries
20:48 < zizban> de-confoozle
20:48 < vasi> sorry for any stridency :-)
20:48 < vasi> drm, can we see the patch?
20:48 < drm> well, now you've got me worried about the "choice" issue
20:49 < drm> its not really a patch, just a new copy of Engine.pm
20:49 < drm> but i guess i can make it a patch...half a sec
20:49 < drm> ah, ok, i was starting from HEAD, i see that now
20:51 < vasi> dmacks, maybe we could just ask apt to satisfy the deps...steal the 'apt-get build-dep' functionality from debian :-)
20:51 < vasi> (that was medium-term musing, don't think i'm suggesting that instead of better swappy code now)
20:52 -!- lisppaste [~lisppaste@common-lisp.net] has quit ["Common Lisp IRC library - http://common-lisp.net/project/cl-irc"]
20:52 -!- lisppaste [~lisppaste@common-lisp.net] has joined #fink
20:53 < drm> vasi: http://www.cgtp.duke.edu/~drm/debs.patch
20:53 < dmacks> heh. /me likes the term "swappy code"
20:53 < zizban> me too
20:53 < drm> vasi: actually, from thesin's garbled comment, i think i do understand the problem
20:54 < drm> but it should only arise when the .deb in question doesn't exist, i.e., when the user is asked whether to do it or not
20:54 < drm> and i switched the default answer to "no"... you don't want this to go forward unattended
20:56 < vasi> ah...so now, if there is a deb, it's installed unconditionally...and if there's not, we ask and default to no
20:56 < vasi> tbh, the swapping has worked almost perfectly for me
20:57 < drm> right
20:57 < vasi> except for this weird thing that happened a few minutes ago where it swapped and then swapped right back consecutively
20:57 < vasi> trying to see if i can repro that
20:57 < drm> if there is no deb, and rebuilding the thing is required, then choices of dependencies might be necc
20:58 < drm> vasi: actually, i've seen that too
20:58 < drm> and i understand why :)
20:58 < vasi> does real_install have an option to skip asking the user things, for situations like that? if it doesn't, it should
20:58 < vasi> drm, ooh tell me
20:59 < drm> it decides on the order to install stuff that is on its list by going through alphabetically, and installing anything which has all its deps satisfied, so
20:59 < drm> it might install lib2-dev before it is needed
20:59 < drm> but now this new routine comes along and says "put lib1-dev back in, its needed right now"
21:00 < vasi> oh wait....it doesn't figure out "are all the deps around" before trying to swap, is what you're saying
21:00 < drm> vasi: real_install is supposed to be an internal thing: after you've set up everything properly, then it is invoked...really, when we ask the user we should probably call "fink install foo" instead
21:01 < vasi> drm, so the solution to the thing i just saw is to have two loops
21:02 < vasi> loop 1 tests all the deps for "should be installed", ie: $dep->[FLAG] & 2
21:02 < drm> vasi: GOOD GOD NOT ANOTHER LOOP MAN... hehe
21:02 < cirdan> hehe
21:02 < cirdan> GOTO
21:02 < vasi> heh, sequential loops
21:02 < vasi> not nested
21:02 < vasi> no worries :-)
21:02 < vasi> loops 2 installs any missing things that should be installed
21:02 < drm> vasi: i dont' think we need this
21:03 < vasi> well it's kinda silly to do swapping for nothing
21:03 < vasi> and the first loop can really just be a grep, actually
21:03 < drm> vasi: i agree its silly
21:03 < drm> but it works :/
21:04 < dmacks> Why *do* we actually do things alphabetically? Why not start with what the user requested?
21:05 < drm> dmacks: remember that many items were added to the list... its not just the original set the user wanted
21:05 < drm> i guess we could add new items to the front of the list, and build in that order
21:06 < vasi> lisppaste:url?
21:06 < lisppaste> To use the lisppaste bot, visit http://paste.lisp.org/new/fink and enter your paste.
21:07 < cirdan> we should port mailsync
21:07 < cirdan> .sf.net
21:08 < drm> gotta go catch my connection...they just moved the time up!
21:08 -!- drm [~drm@m975f36d0.tmodns.net] has quit ["Leaving"]
21:08 < dmacks> bye
21:08 < lisppaste> vasi pasted "Like this instead" at http://paste.lisp.org/display/10262
21:08 < vasi> woops, touch too late
21:08 < cirdan> heh
21:08 < vasi> dmacks, that looks right eh?
21:08 < cirdan> sms it to him :-)
21:09 < dmacks> "drm, please pick up the white courtesy phone."
21:09 < dmacks> "No, the *white* one."
21:09 < cirdan> lol
21:09 < zizban> heh
21:09 < cirdan> drm, that's red.
21:10 < dmacks> vasi: Looks reasonable. Although I don't know jack about the dep-swapping at all to know if it's "right".
21:13 < vasi> dmacks, if we can get everybody to agree that swapping works with the newer system, then i think we should do a quick 0.24.9 release
21:13 < vasi> a couple weeks after .8 maybe
21:13 < dmacks> Sure.
21:13 < vasi> so we can get stable on 0.24.x within a month or so
21:13 < dmacks> Right on.
21:13 < vasi> cuz frankly, i don't want to touch 0.23.x any more
21:13 < cirdan> :-)
21:14 < dmacks> Yeah.
21:14 < cirdan> 0.24 is also for 10.3?
21:14 < vasi> cirdan, yes
21:14 < dmacks> cirdan: yup
21:14 < vasi> right now 0.23 is stable, 0.24 is unstable
21:14 < vasi> on both dists
21:14 < cirdan> ah
21:14 < cirdan> motion to punt 0.24.x to stable quickly
21:15 < vasi> dmacks, just so i have a good mental picture of where we're going, what 0.25-blockers do we have?
21:15 < vasi> my indexing code needs a bit of massaging
21:15 < cirdan> dist-upgrade!
21:15 < cirdan> :-)
21:15 < vasi> your cleanup stuff
21:15 < vasi> and somebody has to clean up Shlibs, probably me :-)
21:15  * cirdan summons thesin to finish shlibs
21:15 < cirdan> :-)
21:16 < cirdan> bah, busy signal 
21:16 < vasi> uh, and maybe dist-up :-P
21:16 < dmacks> I think that's the list.
21:17 < akh> How about implied depends on Essentials?
21:17 < dmacks> A new License:Restrictive/source-disttributable or somesuch.
21:17 < cirdan> yeah, that needs to be adressed
21:17 < vasi> dmacks, i thought that was already in HEAD?
21:17 < cirdan> all packages should be dependant on any essential pkgs they need by now
21:17 < dmacks> The default and prompting actions for upgrdaing a 'fink reinstall' to 'rebuild'.
21:17 < dmacks> vasi: nope.
21:18 < dmacks> Yeah, nuke that dependency crap.
21:18 < vasi> akh, cirdan, i'm beginning to lean *against* taking away implied depends
21:18 < cirdan> vasi: why?
21:18 < akh> vasi: So explicit depends?
21:18 < akh> oops nm
21:18  * akh can't read, apparently.  
21:19 < vasi> partly because it caught our previous error of making an Essential depends on a non-Essential :-)
21:19 < akh> Right.
21:19 < cirdan> the whole point of the essential pkgs being marked as essential is so that they don't get uninstalled. dpkg/apt/fink are aware of this atribute
21:19 < cirdan> no need to make it even stronger
21:19 < dmacks> It could break stuff perhaps, but same as with any other missing-dependency.
21:19 < cirdan> vasi: it breaks the builddeps
21:19 < cirdan> it's more of a case for the validator
21:20 < vasi> argh, lemme check the logs for what gave me pause before
21:20 < cirdan> it needs to x-ref essential pkgs and check all deps are also essential
21:20 < dmacks> What's wrong with having essential depends on non-essential?
21:21 < cirdan> dmacks: can't be done
21:21 < cirdan> policy, and logic
21:21 < vasi> dmacks, that's borked...how will you bootstrap, for one thing?
21:21 < dmacks> Ah, I didn't think about bootstrapping.
21:22 < dmacks> Does bootstrapping really only do Essentials? Or does it use the depndency engine?
21:22 < cirdan> dmacks: essentials *always* are built first
21:22 < vasi> it uses the engine in the third phase
21:22 < dmacks> vasi: Gotcha.
21:22 < cirdan> it would kinda work, if the dep engine was patched, becuase any essential pkg dep can't be uninstalled (the essential pkg needs it)
21:22 < dmacks> cirdan: Then that's silly right there.
21:22 < cirdan> but it's really, really bad policy
21:23 < cirdan> and there is still the bug of the parent package needing to be essential. the dep engine overlooks child flags in bootstrap
21:23 < dmacks> How (if any) should the mess with essential splitoffs of non-essential parents be changed?
21:24 < dmacks> Yeah. /me hates how special-casy the bootstrapping is.
21:24 < akh> Why not make the parents essential, too?  Is it just a question of having extra packages installed?
21:24 < cirdan> dmacks: if *any* child package has essential tag, the whole family needs essential status to build
21:24 < vasi> er, i think it's just a problem with our parser that perpetuated
21:24 < cirdan> dmacks: it's not special case, it's just where the problems are noticed :-)
21:24 < vasi> there's no reason a splitoff can't be essential
21:25 < cirdan> akh: parent can be -dev
21:25 < vasi> and the concept of parent-essential should go bye-bye
21:25 < akh> cirdan: ah
21:25 < vasi> imho
21:25 < vasi> not sure what that might break though
21:25 < cirdan> that's how i noticed it, libncurses5 (dev) was the parent
21:26 < vasi> cirdan, you realize that the implicit depends on essential and the splitoff-can't-be-essential are two separate issues, right?
21:26 < cirdan> yes
21:26 < vasi> just making sure :-)
21:26 < cirdan> but both related being problems with essential pkgs ;-)
21:27 < vasi> cirdan, do you remember how things broke when you tried to make a splitoff Essential?
21:27 < cirdan> yes
21:27 < cirdan> because the parent was not essential, the whole package was put on the back burner
21:28 < cirdan> and it tried to build dpkg before ncurses
21:28 < cirdan> check my exp cvs history for the libncurses5.info
21:28 < cirdan> put that in bootstrap and the problem is there :-)
21:31 < vasi> oh, probably because then libncurses5 wasn't essential or parentessential
21:31 < vasi> yeah
21:31 < vasi> ok, so instead of parentessential, we should have a 'relative_essential'
21:32 < vasi> as a function that does 'grep { $_->is_essential() } $self->get_relatives
21:32 < vasi> then it should work i think :-)
21:36 < cirdan> k
21:37 -!- You're now known as RangerAway
21:42 < vasi>  /sw/lib/fontconfig2/lib/libfontconfig.dylib: No such file or directory
21:43 < vasi> but it's there...
21:44 < cirdan> not on my box...
21:44 < vasi> ah, looks like someone forgot to make -dev depend on shlibs
21:45 < cirdan> oops
21:45 < cirdan> haha
21:45 < akh> yup.
21:46 < vasi> psst, Feanor...that's your package
21:46 < vasi> no?
21:50 < vasi> cirdan, re: implicit depends... i'd like to wait until we start the 10.4 tree
21:51 < vasi> since we'll have to test everything then anyhow
21:51 < vasi> we can make the code for implicit depend on essential conditional on dist
21:53 < vasi> ok, i'm off now
21:53 -!- vasi is now known as vasiGone
22:13 -!- akh [~akhansen@68-118-244-23.dhcp.oxfr.ma.charter.com] has quit []
22:15 < dmacks> vasiGone: vte? Make sure you're using -4.
22:16 < dmacks> (...and if we find missing depends while making 10.4R, can fix them in 10.3 so we don't force ourselves to cease supporting that tree with new fink versions.
22:17  * dmacks still can't understand why fink would try to build dpkg before ncurses if dpkg:(build)depends:ncurses, but whatever.
22:18 -!- newmanbe [~newmanbe@d71e5a28cff85fdc.session.tor] has quit ["leaving"]
22:19 -!- baba [~baba@YahooBB220041000007.bbtec.net] has joined #fink
22:27 -!- spundun [~spundun@mermaid.isi.edu] has quit ["This computer has gone to sleep"]
22:30 -!- vasiGone [~vasi@modemcable133.147-70-69.mc.videotron.ca] has quit []
22:37 -!- zizban [~zizban@24-52-0-219.sbtnvt.adelphia.net] has quit [Read error: 110 (Connection timed out)]
22:46 < dmacks> baba: If your Type:python value is 23 instead of -py23, you don't need all those conditionals and other games to figure out what the python version is.
22:54 < baba> i wanted py24, but py24 still has issues to solve
22:54 < dmacks> That's not my point at all.
22:56 < dmacks> Compare how you used Type:python with how other -py23 packages use it.
22:58 < cirdan> dmacks: because of a bug
22:58 < baba> ok, i will follow 2.3 rather than py23
22:59 < dmacks> cirdan: Yeah, I figured:/
23:32 -!- kalessin [~crhalpin@adsl-68-78-232-23.dsl.mdsnwi.ameritech.net] has quit ["sleep."]
23:57 -!- Melian [~blootbot@pcp04356153pcs.glstrt01.nj.comcast.net] has quit [Read error: 110 (Connection timed out)]
--- Log closed Thu Jul 28 00:00:27 2005