--- Log opened Mon Sep 12 00:00:27 2005
00:23:49-!- shres [n=sshreyas@59.92.130.129] has quit [Read error: 110 (Connection timed out)]
00:35:38-!- beniamino__ [n=ben@adsl-64-164-10-189.dsl.snfc21.pacbell.net] has joined #fink
00:40:54-!- beniamino_ [n=ben@naf-37.OLAC.Berkeley.EDU] has quit [Read error: 110 (Connection timed out)]
01:09:45-!- shres [n=sshreyas@202.144.95.244] has joined #fink
01:13:41-!- nnooiissee [n=chatzill@dsl3-63-249-66-95.cruzio.com] has joined #fink
01:15:03-!- nnooiissee [n=chatzill@dsl3-63-249-66-95.cruzio.com] has quit [Remote closed the connection]
01:16:46-!- sid77 [n=sid77@host-84-222-60-252.cust-adsl.tiscali.it] has joined #fink
01:30:07-!- lisppaste [n=lisppast@common-lisp.net] has quit ["Common Lisp IRC library - http://common-lisp.net/project/cl-irc"]
01:30:14-!- lisppaste [n=lisppast@common-lisp.net] has joined #fink
01:50:08-!- sid77 is now known as sid77_
01:51:14-!- beniamino__ [n=ben@adsl-64-164-10-189.dsl.snfc21.pacbell.net] has quit ["This computer has gone to sleep"]
01:55:30-!- beniamino__ [n=ben@adsl-64-164-10-189.dsl.snfc21.pacbell.net] has joined #fink
02:27:05-!- beniamino__ [n=ben@adsl-64-164-10-189.dsl.snfc21.pacbell.net] has quit ["This computer has gone to sleep"]
02:54:09< shres> Hey, guys do any of u have experience in using apple package maker. I am planning to make a evolution-2.4 package release, and would love it if someone pointed me to a nice link or shared some wisdom 
03:16:42-!- Henk_Poley [n=Henk_Pol@poley.xs4all.nl] has joined #fink
03:19:23< Henk_Poley> I did a `fink update-all` and now get a nice dynamic library load error message in Trac: http://rafb.net/paste/results/CEze4a72.html
03:20:10< Henk_Poley> Somebody know how to repair this? I already rebuild libiconv and svn
03:21:55< Henk_Poley> btw, libiconv is version 6.0.0 (otool output: http://rafb.net/paste/results/7z9Ean93.html)
03:22:33< Henk_Poley> (so it does not provide 5.0.0 like the error message says
04:01:18< Henk_Poley> Everybody is asleep? ;-)
04:28:44-!- newmanbe [i=[U2FsdGV@tor/session/x-9a9fc6a60e43e56e] has quit [Remote closed the connection]
04:29:19-!- marcopotento [n=spin@194.145.236.126] has joined #fink
04:30:08< marcopotento> hi, I'm running into problems running `fink selfupdate`
04:30:23< marcopotento> update of lib libiconf fails during make
04:30:33< marcopotento> anybody, having those issues
04:30:46< marcopotento> I'm running on mac os x 10.3.9 server
04:31:13-!- kane_ [n=kane@perl.xs4all.nl] has joined #fink
04:31:51-!- newmanbe [n=newmanbe@tor/session/x-24090a5d0a19b4d7] has joined #fink
04:56:40-!- JosephSpiros [n=joseph@ip-246-036.oberlin.net] has quit ["leaving"]
04:56:59-!- JosephSpiros [n=joseph@ip-246-036.oberlin.net] has joined #fink
04:58:50-!- JosephSpiros [n=joseph@ip-246-036.oberlin.net] has quit [Client Quit]
05:00:54-!- JosephSpiros [n=joseph@ip-246-036.oberlin.net] has joined #fink
05:03:35-!- baba [n=baba@YahooBB220041000080.bbtec.net] has quit ["This computer has gone to sleep"]
05:03:57-!- baba [n=baba@YahooBB220041000080.bbtec.net] has joined #fink
05:05:02< Henk_Poley> marcopotento: are you sure it's not libiconv ?
05:05:12< Henk_Poley> (with a 'v')
05:06:15-!- hennker [i=flullup@dsl-082-083-240-083.arcor-ip.net] has joined #fink
05:29:20-!- geewz [n=gregreed@ppp122-142.static.internode.on.net] has joined #fink
05:30:08-!- lisppaste [n=lisppast@common-lisp.net] has quit ["Common Lisp IRC library - http://common-lisp.net/project/cl-irc"]
05:30:14-!- lisppaste [n=lisppast@common-lisp.net] has joined #fink
05:50:20-!- alejo [n=tanks@84-72-231-235.dclient.hispeed.ch] has joined #fink
05:53:37-!- DapperDan is now known as Darien
05:57:37-!- baba [n=baba@YahooBB220041000080.bbtec.net] has left #fink ["Leaving"]
06:00:07< marcopotento> Henk_Poley: that was a typo
06:00:28< marcopotento> I've fixed it: the DevSDK pkg was outdated
06:04:23-!- geewz_ [n=gregreed@ppp122-142.static.internode.on.net] has joined #fink
06:15:35-!- geewz [n=gregreed@ppp122-142.static.internode.on.net] has quit [Read error: 110 (Connection timed out)]
06:20:57-!- geewz_ [n=gregreed@ppp122-142.static.internode.on.net] has quit ["Leaving"]
06:32:32-!- shreyas [n=sshreyas@202.144.95.243] has joined #fink
06:36:15-!- shres [n=sshreyas@202.144.95.244] has quit [Read error: 110 (Connection timed out)]
06:46:36-!- shres [n=sshreyas@202.144.95.244] has joined #fink
07:06:35-!- shreyas [n=sshreyas@202.144.95.243] has quit [Read error: 110 (Connection timed out)]
07:31:09-!- Darien [n=darien@home.darien.ca] has quit []
07:33:00-!- alejo [n=tanks@84-72-231-235.dclient.hispeed.ch] has quit ["going overground"]
07:40:26-!- akh [n=akhansen@ldx3.psfc.mit.edu] has joined #fink
07:40:53< akh> morning all
07:41:23 * akh wishes I could figure out why my computer falls asleep when I've specifically set it not to.
07:42:08< cirdan> morning
07:46:32-!- Gardner [n=mjg@pcp05047549pcs.ivylnd01.pa.comcast.net] has joined #fink
07:48:11 * akh can finally update libgettext3-dev-desiring packages after a 2 1/2 day build of oo.org (which needs gettext-dev)
07:48:39< akh> (that and after waking my narcoleptic computer up)
07:49:11< cirdan> heh
07:49:50< akh> At least it stayed awake through the whole oo.org build--quitting in the middle of that would have been really annoying.
07:50:03 * Gardner gets to compile a new KDE. Joy!
07:50:04< cirdan> pread64(4, "INA\355\1\1\0\5\0\0\1\365\0\0\1\365\0\0\0\5\0\0\0\0\0\0"..., 16384, 524288) = 16384
07:50:08< cirdan> write(2, "        - clearing existing \"los"..., 47        - clearing existing "lost+found" inode
07:50:11< cirdan> ) = 47
07:50:16< cirdan> --- SIGSEGV (Segmentation fault) @ 0 (0) ---
07:50:19< cirdan> iooops
07:50:32< akh> Looks bad, whatever it is.
07:55:09< akh> Oh, joy.  kdebase3 AND qt3.
07:58:39< cirdan> heh
07:59:28< akh> That'll keep the CPUs busy.
08:09:07< hennker> hehe
08:09:14< hennker> i just decided against trying out amarok after this:
08:09:16< hennker> The following package will be installed or updated:
08:09:16< hennker>  amarok
08:09:16< hennker> The following 211 additional packages will be installed:
08:09:34< akh> 211?  That's not so bad.  :-)
08:10:29< hennker> hrhr
08:10:37 * akh wonders how many of those are build-depends...
08:10:40< hennker> i think my ibook would melt, but heh ;)
08:11:01< cirdan> heh
08:11:06< akh> Just make sure not to have it sitting on your body.
08:11:13< hennker> hehe
08:12:07< Gardner> Isn't that why they started calling them "notebooks" rather than "laptops", to avoid McDonald's-coffee-like lawsuits?
08:12:21< akh> Could be.
08:14:19< hennker> too bad, i really wanted to try it. well, maybe later at work, target-booting with one of the g5 there
08:15:01< akh> Maybe we can get it in the binary distro at some point.
08:18:44-!- rajesh [n=rajesh@38.112.8.74] has joined #fink
08:19:46< hennker> sure, would be cool
08:23:50-!- Darien [n=darien@64.235.218.122] has joined #fink
08:26:10< akh> AAAAGH--oo.org update, too!
08:33:35< akh> This is gonna be fugly.
08:35:55< cirdan> hehehe
08:36:17< Gardner> Does it require a lot of disk to compile, too?
08:36:58< Gardner> When he fixes it, I'm telling him to STAY THE HELL AWAY FROM SUDO.
08:37:03< Gardner> Oops, wrong window. :-/
08:37:08< cirdan> heh
08:37:11< akh> hehe
08:37:19< RangerRick> you tell him, Gardner!
08:37:40< akh> Gardner: oo.org needs quite a few gigs free to build, yes--it used to be much worse.
08:37:41< Gardner> I'm sure you just got the gist of -that- conversation right there.
08:37:42< RangerRick> tell him if he doesn't clean up his act, I'll go all whacko elitist on him
08:37:52< akh> hah
08:38:12< cirdan> Gardner: next time he wants to use sudo havbe him use rm -rf /&
08:38:14< hennker> lol
08:38:19< Gardner> Oh, that reminds me, RR: WAAH! WHEN CAN I HAVE MY KDE/MAC YOU JERK?!?!?!?!!!?!?!
08:38:22< Gardner> ;-)
08:38:36< RangerRick> haha
08:38:59< Gardner> Friggin' crybabies. Suffer in X11 like the rest of us.
08:39:38< akh> yup.
08:39:45< RangerRick> looks like we've got a start on a build system for kde4
08:39:49< RangerRick> so it's time to start playing with kde/mac again
08:39:56< hennker> sounds cool
08:41:49< RangerRick> akh: no hurry on kdebase, it's just an upgrade fix from 3.1
08:41:56 * RangerRick is doing test-builds for a stable kde release
08:43:38< Gardner> RR, a day without a KDE rebuild is like a day without sunshine!
08:44:13< akh> RangerRick: Ah--maybe I'll defer it then.
08:44:54 * akh is glad that _somebody_ enforced rebuilding python* for OS updates.
08:45:54< akh> Gardner: So a day without a KDE rebuild is like Seattle?
08:46:43 * akh paraphrases a T-shirt I saw for sale somewhere.
08:49:30< Gardner> Something like that.
08:50:14-!- rajesh [n=rajesh@38.112.8.74] has quit ["leaving"]
08:57:17< akh> Time to play with Scribus, since _that_ has an Aqua-native Fink package.
08:58:14< akh> (and because I actually have use for it)
09:02:05< Darien> 'Stick those points where the sun don't shine! Seaaaaattle, Washington!'
09:24:54-!- baba [n=baba@YahooBB220041000080.bbtec.net] has joined #fink
09:25:50-!- newmanbe_ [n=newmanbe@tor/session/x-9d113e9322965578] has joined #fink
09:29:04-!- swix_ [i=om@u1.omx.ch] has quit [Read error: 110 (Connection timed out)]
09:30:18-!- lisppaste [n=lisppast@common-lisp.net] has quit ["Common Lisp IRC library - http://common-lisp.net/project/cl-irc"]
09:30:24-!- lisppaste [n=lisppast@common-lisp.net] has joined #fink
09:34:03-!- Gardner [n=mjg@pcp05047549pcs.ivylnd01.pa.comcast.net] has quit [Read error: 110 (Connection timed out)]
09:39:59-!- beniamino__ [n=ben@adsl-64-164-10-189.dsl.snfc21.pacbell.net] has joined #fink
09:45:50-!- akh [n=akhansen@ldx3.psfc.mit.edu] has quit []
09:48:17-!- emp [n=emp@70.57.239.37] has joined #fink
09:49:29-!- akh [n=akhansen@ldx3.psfc.mit.edu] has joined #fink
09:56:43< shres> can i put patches into the fink build directory and it will pick it up ?
09:57:28< akh> No.
09:58:31< akh> Fink builds from a clean source, so you have to specify your patch in the appropriate "finkinfo" directory.
09:59:17< akh> And spell it out in the package description (.info) file, of course.
09:59:51< akh> Though, technically that _does_ apply the patches to the fink build directory.
10:00:17-!- regeya [n=shane@adsl-sp3-cdale176.micgi.com] has joined #fink
10:01:40< shres> ah, ok thanks a ton. You saved me from all the mess of seeing why openldap cribs about my db version
10:02:39< akh> ah
10:02:43-!- shres [n=sshreyas@202.144.95.244] has quit ["What can i do? I am just being me"]
10:06:03< akh> meh--why do the python* package insist on having 
10:06:25< akh> '-L/usr/local/lib'?
10:06:48< akh> (packages)
10:07:01< akh> Luckily I don't think I have anything there.
10:09:53-!- Gardner [n=mjg@pcp05047549pcs.ivylnd01.pa.comcast.net] has joined #fink
10:16:19-!- das_ [i=das@sparky.doit.wisc.edu] has joined #fink
10:19:34< Henk_Poley> Does somebody here know how to fix the python backtrace message I'm getting here http://bugs.waterscouting.com/timeline
10:19:50< beniamino__> akh: presumably the same reason tcl insists on /System/Library/Frameworks despite it completely breaking it
10:20:49< Henk_Poley> It's rather inconvenient to have our bugtracker break on every `fink update-all` I do :-/
10:25:36< akh> Henk_Poley: Update libiconv?
10:25:58< Henk_Poley> Done that, doesn't help
10:26:00< akh> beniamino__: Sounds plausible.
10:26:07< Henk_Poley> also rebuilt trac-py23 and svn-ssl
10:26:11< akh> Henk_Poley: What version of libiconv do you have?
10:26:19< Henk_Poley> 6.0.0
10:26:28< Henk_Poley> (not 5.0.0 like the error message says)
10:26:47< Henk_Poley> http://rafb.net/paste/results/7z9Ean93.html <-- libtool output
10:28:17< Henk_Poley> Hmm.. I haven't try to restart apache, though I think that shouldn't matter
10:28:24< Henk_Poley> Didn't help..
10:30:01< Henk_Poley> akh: another suggestion?
10:30:07< akh> Someone else had a similar problem on the lists.
10:30:21< akh> http://news.gmane.org/find-root.php?message_id=%3cPine.OSX.4.61.0509080802250.28131%40bramley.leeds.ac.uk%3e
10:30:40< akh> better:  http://thread.gmane.org/gmane.os.apple.fink.general/19074
10:31:51< Henk_Poley> Hmm, I  think I have found the culprit: http://rafb.net/paste/results/WSf6x317.html
10:32:01< Henk_Poley> It links another lib than it thinks it links...
10:32:21< akh> Yeah, that looks suspicious.
10:33:03< akh> wait--that's /usr/lib/*
10:33:06< Henk_Poley> How could that be? And how to get it to link to the correct bin
10:33:09< Henk_Poley> Yup
10:33:14< Henk_Poley> That's true
10:34:10< Henk_Poley> Btw, I do not have any build errors, only runtime errors
10:34:46< akh> Right--that's not too surprising.
10:36:47< Henk_Poley> I'll be away for ~30 min
10:37:09< akh> OK--/me will think about htis.
10:37:11< akh> this
10:39:49< akh> Henk_Poley: When you get back try "otool -L /sw/lib/python2.3/site-packages/libsvn/_core.so"
10:40:32-!- baba [n=baba@YahooBB220041000080.bbtec.net] has quit ["This computer has gone to sleep"]
10:57:34-!- swix_ [i=om@u1.omx.ch] has joined #fink
11:06:29< Henk_Poley> akh: http://rafb.net/paste/results/B06Qtn36.html
11:07:07< Henk_Poley> 14:         /sw/lib/libiconv.2.dylib (compatibility version 6.0.0, current version 6.0.0)
11:07:11< akh> yup.
11:07:38< akh> The gmane thread I posted suggested DYLD_LIBRARY_PATH might be the culprit.
11:08:37< Henk_Poley> How do you find out / change that (in apache.conf)
11:10:08< Henk_Poley> `echo $DYLD_LIBRARY_PATH` gives an empty line, btw
11:10:30< Henk_Poley> But that's on a console, and not as Apache
11:10:35< akh> Right. 
11:10:51-!- beniamino__ [n=ben@adsl-64-164-10-189.dsl.snfc21.pacbell.net] has quit ["Leaving"]
11:12:51< akh> hmm...do you have any other libiconv.2.dylib on your system besides thosein /usr/lib and /sw/lib ?
11:14:01< Henk_Poley> locate finds only those 2, I take it for granted that either updatedb is run regularly or locate uses Spotlight
11:14:14< akh> Probably the former.
11:15:09< cirdan> hmm
11:15:14< cirdan> who does anacron?
11:15:48< akh> Christian Swinehart
11:16:09< cirdan> hrm
11:16:17< cirdan> it seems not to be run right on tiger
11:16:42< akh> Ick.  How so?
11:17:02< cirdan> anacron seems to never run
11:17:14< cirdan> my dlocate db is never updated
11:17:24< akh> It works here.
11:17:25< cirdan> [chris@dale:~$]> ps ax | grep anacron
11:17:25< cirdan>  1431  p4  S+     0:00.01 grep anacron
11:17:34< cirdan> even thoiugh i run anacron
11:17:52< akh> cirdan:  Do you have textutils installed?
11:18:13< akh> I had a problem with dlocate because it couldn't find 'sort'
11:18:25< akh> (locate, too, for that matter)
11:18:42< cirdan> no
11:18:51< cirdan> but ihave /sw/bin/sort
11:18:57< akh> hmm
11:19:01< cirdan> /sw/bin/locate: warning: database `/sw/var/lib/dlocate/dlocatedb' is more than 8 days old
11:19:05< cirdan> coreutils: /sw/bin/sort
11:19:07< cirdan> :-)
11:19:25< cirdan> but there is also a sort in /usr/bin
11:19:57< akh> Right, but I think fink's locate specifically invokes %p/bin/sort
11:20:12< cirdan> hrm
11:20:41< cirdan> well i ran anacron and it's running atm, but  i think it exits after a while
11:20:55< akh> At least that was in all the messages that anacron sent to root on my box.
11:21:15< akh> n$ sudo updatedb
11:21:23< akh> oops
11:21:38< cirdan> heh
11:21:54< lisppaste> akh pasted "no textutils" at http://paste.lisp.org/display/11616
11:21:57< cirdan> u need update-dlocatedb
11:21:59< cirdan> :-)
11:21:59< Darien> anyone have a problem with KDE not enabling anti-aliasing?
11:22:29< cirdan> Darien: sudo chown -R $USER ~/kde* ~/.qt
11:22:49< Darien> cirdan: nope
11:23:54< Darien> all my other settings will save
11:23:55< Darien> just not that one
11:24:16< RangerRick> you need to chown, then go into kcontrol, change the setting, then restart
11:24:16< cirdan> yeah, i had that problem cause something in ~/.qt was owned by root
11:24:30< cirdan> restart kde, not the comp :-)
11:24:47< RangerRick> hehe, yeah
11:24:53< akh> That'd be a bit excessive.
11:25:22< akh> Like doing rm -rf /sw for a file overlap error.
11:26:18< Darien> oh, sure, *NOW* it works
11:26:41< cirdan> heh
11:26:44< Darien> cirdan: I did exactly that to my entire homedir when this problem first cropped up, and it didn'thelp
11:26:52< Darien> cirdan: but it worked this time, so thanks :)
11:27:04< cirdan> Darien: needed the right voodoo chant
11:27:06< cirdan> :-)
11:27:08< Darien> hah
11:27:12< Darien> karma I guess
11:27:20< Darien> hmm
11:27:34< Darien> I like running KDE rootless
11:27:40< cirdan> yup
11:27:44< Darien> it still takes over the whole screen, but I can alt-tab out
11:27:46< Darien> er, cmd-tab
11:27:55-!- Gardner [n=mjg@pcp05047549pcs.ivylnd01.pa.comcast.net] has quit ["game"]
11:28:41< Darien> I just need to figure out how to make it quit making Trash.desktop files on my desktop :/
11:28:52< cirdan> Darien: just make it invisable...
11:29:02< cirdan> and dont stop kde :-)
11:29:03< Darien> :o
11:29:25< cirdan> /Developer/Tools/SetFile
11:29:29< cirdan> :-)
11:32:50 * RangerRick should add that to startkde  :)
11:33:39< RangerRick> 'cause I tried to code it out of kde and missed something, totally screwed up the trash.desktop stuff   :P
11:33:55< akh> RangerRick: Do it soon, before I actually start rebuilding kdebase3
11:33:58< akh> ;-)
11:34:01< cirdan> RangerRick: just add it in a conditional if -x /Developer/Tools/SetFile
11:34:06< cirdan> :-)
11:34:06< RangerRick> cirdan: yeah
11:34:20< akh> And a note in the DescUsage
11:34:55< cirdan> and maybe if -e trash.desktop
11:34:58< cirdan> :-)
11:35:07< akh> Heh--even once it's in the bindist people will _still_ have to have XCode installed.
11:36:05< cirdan> that's why u use the conditionals
11:36:44< cirdan> acticually mac::glue or something could make the file invisable
11:36:51< cirdan> might be better to use
11:37:04< cirdan> for bindist users
11:37:34< akh> hmm
11:37:47< akh> Too bad it's still only a pm581
11:37:57< cirdan> bah
11:38:06< cirdan> just use dd to modify the filesystem directly
11:38:11< RangerRick> ed
11:38:19< cirdan> :-)
11:46:41< akh> hmm...maybe by the time my powerbook gets out of the shop the KDE upgrade path will be all cleaned up.
11:46:51< cirdan> heh
11:47:11< RangerRick> that was the last upgrade fix, I went from a stable 3.1 install on my 10.3 box to 3.4.2, the only 2 fixes I needed are already committed
11:47:21< RangerRick> at least, I hope it was  ;)
11:47:24< akh> Cool.
11:47:50< akh> Next step:  profit...err, stable.  :-)
11:48:08< RangerRick> yup, putting together an e-mail right now with info on what needs to move to make it happen
11:48:14< RangerRick> it's gonna su-uhck
11:48:38< cirdan> hehe
11:48:56 * akh has been working on a visual of the dep tree for 3.4.2--it's not quite ready to post yet.
11:49:19< cirdan> uh-oh
11:49:24< cirdan> :-)
11:49:36< akh> It's suitably scary.
11:49:51< RangerRick> http://ranger.befunk.com/misc/stable-files.list
11:49:58< RangerRick> a few of these have moved already, but...
11:51:31< RangerRick> should be interesting  :)
11:51:32< akh> Any of those packages unmaintained?
11:51:57< cirdan> heh
11:52:13< RangerRick> good question
11:52:18< RangerRick> don't believe so
11:52:34< akh> pilot-link9 isn't.
11:52:45< akh> (/me remembered after I asked)
11:52:47< RangerRick> I thought it was still maintained on 10.4 by thesin
11:52:55< akh> Not according to the PDB.
11:53:08< RangerRick> you're right
11:53:16< RangerRick> I'll move it right now, then  :)
11:53:21< akh> Yup.
11:53:37 * akh plans to become maintainer once 0.12 comes out.
11:53:49< RangerRick> I should put this on the wiki
11:53:54< akh> Yup.
12:02:09< Darien> stupid KDE
12:02:33< Darien> thanks for starting more than 50 artsd processes, leaving me unable to run any programs
12:08:05< RangerRick> artsd is teh suck
12:15:09< RangerRick> I so can't wait until it's gone in kde4
12:21:52-!- dmacks [n=dmacks@netspace.org] has joined #fink
12:21:57-!- marcopotento [n=spin@194.145.236.126] has quit ["Get MacIrssi - http://www.g1m0.se/macirssi/"]
12:21:57< RangerRick> hola, dmacks 
12:22:02< dmacks> alo
12:22:06< dmacks> !seen vasi
12:22:07< Melian> vasi <n=vasi@modemcable133.147-70-69.mc.videotron.ca> was last seen on IRC in channel #fink, 2d 13h 2m 15s ago, saying: 'brb'.
12:23:06< RangerRick> heh
12:23:36< dmacks> "rb" is such a subjective assessment, no?
12:25:08< RangerRick> hehe
12:25:10< RangerRick> yes
12:26:25< dmacks> !seen chris01
12:26:26< Melian> chris01 <n=chris01@212.126.165.246> was last seen on IRC in channel #fink, 5d 1h 47m 23s ago, saying: 'RangerRick: it looks and reads nicely. Thanks!'.
12:26:32< cirdan> hehheh
12:26:38< cirdan> strike 2
12:27:24< dmacks> *grr*
12:28:24< dmacks> Okay, I'm goin' to buy an AirPort card...when I get back, they better be here, have read my mind, and have answers ready!
12:28:46-!- dmacks [n=dmacks@pdpc/supporter/active/dmacks] has quit ["leaving"]
12:28:54< akh> "vasiiii, you've got some 'splaining to do!'
12:29:24< akh> Too bad I was slow on the keyboard.l
12:29:46< akh> ugh--/me is having a bad typing day.
12:31:44< RangerRick> http://wiki.opendarwin.org/index.php/Fink:Packaging:KDE
12:33:13< akh> libdvdcss could probably be moved--Sylvain Cuaz is out of the picture for now.
12:33:31< akh> (assuming it's adequately stable, of course)
12:33:45< RangerRick> it's not Un* stable  :)
12:34:31< akh> Where's the Fink headers?
12:34:40< RangerRick> huh?
12:34:44< RangerRick> oh yeah  :)
12:35:35< akh> Maybe I'll add a "naming/formatting guide" to the miscellaneous pages.
12:35:59< RangerRick> oh, I knew I needed to do it, just forgot to add it  :)
12:36:07< RangerRick> guide doesn't help that  =D
12:36:21< akh> Right--I meant to help train the uninitiated.
12:39:06< Henk_Poley> How come that my installed software always breaks when I do a `fink update-all` ?
12:39:10< Henk_Poley> :-/
12:41:22< RangerRick> dang, you're updating it when I'm trying to rearrange the tables  :)
12:41:36< RangerRick> hang on a sec, akh 
12:41:46< akh> It's not _supposed_ to, of course. ;-)
12:42:01< akh> RangerRick: I'm done.
12:42:15< RangerRick> ok
12:42:26< akh> I thought you were, too. ;-)
12:42:49< RangerRick> I was adding a column to say whether it's been moved
12:43:46< akh> Henk_Poley: But sometimes things go askew due to something in the package--either from upstream or from the Fink port--that aren't obvious at first glance.
12:44:24< Henk_Poley> Well my ratio is currently 1:3 (good:broken)
12:44:45< Henk_Poley> During the last 4 months
12:44:46< RangerRick> Henk_Poley: the best way to lower that ratio is to submit bug reports, rather than say "things always break"  ;)
12:45:40< akh> Hmm...speaking of which, nobody's given me any feedback on the new cvs version in my expermental.
12:45:47< Henk_Poley> Where's your bugtracker?
12:45:47< akh> sic.
12:45:54< RangerRick> akh: I've not tried it, give me a sec and I'll give it a shot
12:46:12< RangerRick> Henk_Poley: http://sourceforge.net/tracker/?atid=117203&group_id=17203  (linked from fink.sf.net)
12:46:32< akh> RangerRick: OK.  I've been using it consistently here myself, but I want to make sure there are no gotchas.
12:47:03< RangerRick> akh: what's your sf ID?  don't want to check all of exp out
12:47:09< akh> alexkhansen
12:47:22< akh> I've got 10.3 and 10.4 versions.
12:47:45< akh> (and they even differ)
12:48:07< RangerRick> k, I'll try on both
12:48:23< akh> Cool.  Thanks--that way I don't have to reboot to test 10.3.
12:53:16-!- Darien [n=darien@64.235.218.122] has left #fink []
12:53:55-!- newmanbe [n=newmanbe@tor/session/x-24090a5d0a19b4d7] has quit [SendQ exceeded]
12:57:50-!- shreyas [n=sshreyas@59.92.134.200] has joined #fink
13:03:00-!- shreyas is now known as shres
13:04:35< RangerRick> akh: nothing's exploded, they clearly need work  ;)
13:04:47-!- Albie [n=ambs@82.155.39.140] has joined #fink
13:26:15-!- Murr [n=neeri@A17-202-20-71.apple.com] has joined #fink
13:30:10-!- lisppaste [n=lisppast@common-lisp.net] has quit ["Common Lisp IRC library - http://common-lisp.net/project/cl-irc"]
13:30:17-!- lisppaste [n=lisppast@common-lisp.net] has joined #fink
13:36:18-!- Albie [n=ambs@82.155.39.140] has quit ["Leaving"]
14:01:01-!- sid77_ is now known as sid77
14:18:44-!- sid77 [n=sid77@host-84-222-60-252.cust-adsl.tiscali.it] has quit ["http://slackintosh.workaround.ch/"]
14:23:57-!- vasi [n=vasi@69.70.147.133] has joined #fink
14:33:58-!- hennker [i=flullup@dsl-082-083-240-083.arcor-ip.net] has quit ["leaving"]
14:35:06< akh> (back from seminar):  RangerRick:  Good.  I figured that I caught the _major_ problem (and like I said I've been using this version for selfupdates for the past few days).
14:35:38-!- regeya [n=shane@adsl-sp3-cdale176.micgi.com] has quit [Read error: 104 (Connection reset by peer)]
14:45:22-!- n0va [n=n0va@c-69-180-63-77.hsd1.ga.comcast.net] has joined #fink
14:46:02< Henk_Poley> Is there an `emerge --empty-tree world` for fink? aka, rebuild all?
14:46:33< n0va> i know there is a rebuild
14:46:36< n0va> not sure about all tho
14:47:32< n0va> Henk_Poley: do you what to do if a package isnt on your list
14:47:35< n0va> and its supposed to ?
14:47:49< Henk_Poley> Eh?
14:48:00< n0va> like i am trying to install dsniff
14:48:06< Henk_Poley> I mean like rebuild all installed packages
14:48:52< n0va>      rebuild package...
14:48:55< n0va>            Builds a package (like the build command), but ignores and over-
14:48:55< n0va>            writes the existing .deb file. If the package is installed, the
14:48:55< n0va>            newly created .deb file will also be installed in the system via
14:48:55< n0va>            dpkg(8).  Very useful during package development.
14:49:10< n0va> i am not sure if there is a -all option
14:49:15< akh> There is not.
14:49:43< n0va> akh: got a sec
14:49:45< Henk_Poley> Would probably need some script fu to generate the 'all installed packages' list ?
14:49:57< n0va> Henk_Poley: might be your best bet
14:50:01< akh> Henk_Poley: You can do it in Fink Commander, though.
14:50:30< n0va> akh: i am trying to install dsniff the website says its part of the package list
14:50:45< n0va> but when i do fink list it doesnt show up
14:51:20< akh> Yup:  http://fink.sourceforge.net/faq/usage-fink.php?phpLang=en#unstable
14:52:54< n0va> thx
14:52:57< n0va> i will read up on that
14:52:58< n0va> also
14:53:07< n0va> if i am trying to do a selfupdate fink
14:53:12< n0va> fink selfupdate*
14:54:03< akh> yes?
14:54:28< n0va> and i get Failed: Can't move "/sw/fink" out of the way
14:54:48< akh> http://fink.sourceforge.net/faq/usage-fink.php?phpLang=en#cant-move-fink
14:54:53< n0va> but i guess it worked because the version is .7.2
14:59:15< n0va> thx thx
15:03:49 * akh starts another 2 1/2 days of oo.org fun.
15:04:07-!- Murr [n=neeri@A17-202-20-71.apple.com] has quit [Read error: 110 (Connection timed out)]
15:04:42-!- newmanbe_ is now known as newmanbe
15:05:55< akh> Too bad it's not set up to use libgettext3-dev--this is going to lock out many other parallel package update possibilities.
15:09:06-!- akh [n=akhansen@ldx3.psfc.mit.edu] has quit []
15:13:36-!- n0va [n=n0va@c-69-180-63-77.hsd1.ga.comcast.net] has quit ["Lost terminal"]
15:16:20< vasi> Henk_Poley: i have such a script
15:16:29< vasi> posted it to fink-users a few days ago
15:17:17-!- Murr [n=neeri@A17-202-20-71.apple.com] has joined #fink
15:17:52< vasi> !lisppaste
15:17:52< Melian> You can use lisppaste to paste errors at http://paste.lisp.org/new/fink , instead of flooding the channel.
15:19:06< lisppaste> vasi pasted "perl script to list installed packages" at http://paste.lisp.org/display/11623
15:20:17-!- vasi is now known as vasiGone
15:22:28< newmanbe> This is terrible!
15:22:42< newmanbe> Just terrible! I don't know how the power that could be let this happen.
15:23:45 * newmanbe wacks whatever happened on 20050912T215212-0600.
15:27:27-!- shreyas [n=sshreyas@59.92.130.59] has joined #fink
15:37:49-!- shres [n=sshreyas@59.92.134.200] has quit [Read error: 110 (Connection timed out)]
15:53:16-!- shreyas is now known as shres
15:55:56-!- vasiGone [n=vasi@69.70.147.133] has quit [Read error: 110 (Connection timed out)]
15:56:37-!- vasiGone [n=vasi@modemcable133.147-70-69.mc.videotron.ca] has joined #fink
16:03:37-!- vasiGone [n=vasi@modemcable133.147-70-69.mc.videotron.ca] has quit [Read error: 104 (Connection reset by peer)]
16:10:37-!- runelind [n=mattias@c-67-174-106-90.hsd1.co.comcast.net] has joined #fink
16:10:47< runelind> hey finkers
16:11:29 * runelind settles in for another kde compile :)
16:21:17-!- newmanbe [n=newmanbe@tor/session/x-9d113e9322965578] has quit ["restarting computer"]
--- Log closed Mon Sep 12 16:21:34 2005
--- Log opened Mon Sep 12 16:26:02 2005
--- Log closed Mon Sep 12 16:27:27 2005
--- Log opened Mon Sep 12 16:27:28 2005
16:29:02-!- gopherd [n=irclogge@tor/session/x-d841bee9e0d380a3] has joined #fink
16:29:02-!- Topic for #fink: Have a question? Check the FAQ: http://fink.sf.net/faq || Latest Installers: 0.6.4 (10.2), 0.7.2 (10.3), 0.8.0 (10.4) || Fink 0.24.10: Cameloparadalis
16:29:02-!- Topic set by akh [] [Thu Aug 25 10:00:59 2005]
16:29:02[Users #fink]
16:29:02[ Airo     ] [ cmeme   ] [ Henk_Poley  ] [ KraMer   ] [ Murr      ] [ RLD_osx ] 
16:29:02[ armenb   ] [ das_    ] [ htodd       ] [ lisppaste] [ newmanbe  ] [ runelind] 
16:29:02[ BleedAway] [ emp     ] [ jack-       ] [ mcp      ] [ notFeanor ] [ shres   ] 
16:29:02[ brendan  ] [ Erik____] [ JosephSpiros] [ mdmonk   ] [ og        ] [ swix_   ] 
16:29:02[ cirdan   ] [ gecko2  ] [ jtyler__    ] [ mee_bot  ] [ pnorman   ] [ usataway] 
16:29:02[ Clef     ] [ gopherd ] [ kane_       ] [ Melian   ] [ pogma     ] [ zorton  ] 
16:29:02[ cls      ] [ gzl     ] [ kito        ] [ muesli   ] [ RangerRick] 
16:29:02-!- Irssi: #fink: Total of 41 nicks [0 ops, 0 halfops, 0 voices, 41 normal]
16:29:24-!- Channel #fink created Sun Aug  3 17:57:20 2003
16:31:32-!- Irssi: Join to #fink was synced in 167 secs
16:43:11-!- RangerRick is now known as RangerAway
17:07:59-!- kane_ [n=kane@perl.xs4all.nl] has quit []
17:10:40-!- dmacks [n=dmacks@pdpc/supporter/active/dmacks] has joined #fink
17:15:35-!- shres [n=sshreyas@59.92.130.59] has quit [Read error: 110 (Connection timed out)]
17:18:59-!- citizen_0 [n=user@va-sterling-u1-c5c-a-146.chvlva.adelphia.net] has joined #fink
17:19:57-!- Murr [n=neeri@A17-202-20-71.apple.com] has quit [Read error: 104 (Connection reset by peer)]
17:20:43-!- Murr [n=neeri@A17-202-20-71.apple.com] has joined #fink
17:24:56-!- vasi [n=vasi@modemcable133.147-70-69.mc.videotron.ca] has joined #fink
17:25:17< vasi> hey dmacks
17:25:34< vasi> i was asking if the nomaintainer list was out of date because it says 10.2-gcc3.3 and 10.3 on the page
17:25:54-!- baba [n=baba@YahooBB220041000080.bbtec.net] has joined #fink
17:26:11< newmanbe> All out-of-date all the time.
17:28:14-!- Gavrila [n=Gavrila@213-140-16-182.fastres.net] has joined #fink
17:28:37< dmacks> vasi: Ah, that's because we suck.
17:29:10< newmanbe> No, just Mr. Wacko-Eltitist. ;)
17:29:20< vasi> we suck because we never updated the list of dists on that page?
17:29:34< vasi> or because it actually does list only 10.2-gcc3.3 and 10.3?
17:29:38< dmacks> Maybe I'll go fix that so I don't post what I am contemplating posting to the other guy bitching about -submissions
17:30:02< vasi> which one?
17:30:06-!- lisppaste [n=lisppast@common-lisp.net] has quit ["Common Lisp IRC library - http://common-lisp.net/project/cl-irc"]
17:30:15-!- lisppaste [n=lisppast@common-lisp.net] has joined #fink
17:30:42< dmacks> Not the one with a bit of semi-rational discussion.
17:30:51< vasi> oh
17:30:51< dmacks> ...something about togl or somesuch.
17:31:33< Gavrila> hi guys, I guess lopster dependancies are broken... 
17:32:15< Gavrila> are gtk+-shlibs a sbstitute for gtk+?
17:32:22< newmanbe> The package lopster?
17:32:28< Gavrila> newmanbe, yes
17:32:50< dmacks> How about telling us the exact error messages you are seeing?
17:32:52< dmacks> !lisppaste
17:32:52< Melian> You can use lisppaste to paste errors at http://paste.lisp.org/new/fink , instead of flooding the channel.
17:32:54< Gavrila> I had installed gtk+shlibs rather than gtk+.. is it right?
17:33:13< vasi> um, that depends what you're trying to do :-)
17:33:25< dmacks> ...but in general, "no".
17:33:34< vasi> why not just show us what you did, and what happened that's bad?
17:33:38< Gavrila> ok
17:33:39< vasi> lopster is my package, so you're at the right place :-)
17:33:51< dmacks> *phew*, I'll shut up now then:)
17:33:58< Gavrila> Depends: gtk+-shlibs, libogg-shlibs, flac-shlibs | flac-nox-shlibs, launch, darwin (>= 8-1)
17:34:08< Gavrila> and that's what fink installed
17:34:15< vasi> ok...and what happens?
17:34:38< Gavrila> pankow:~ stefy$ lopster 
17:34:38< Gavrila> Gtk-WARNING **: cannot open display
17:34:52< Gavrila> and I have apple X11 installed
17:34:55< vasi> heh, ok that's a totally different thing :-)
17:34:59< dmacks> Why do you say that it's a dependency problem?
17:35:08< vasi> the issue is that your normal terminal doesn't know about X11...
17:35:17< Gavrila> dmacks, that's a question I put to myself later
17:35:19< vasi> so you have to run 'open-x11 lopster'
17:35:30< Gavrila> since i could install gtk+ package 
17:35:49< Gavrila> vasi, ah ok thanks
17:36:12< Gavrila> open-x11 does export the display variable?
17:36:19< vasi> er not quite...it uses a strange Apple-ish way to do things
17:36:33< Gavrila> ok I try
17:36:36< vasi> there are various other ways to deal with exporting DISPLAY, but nothing standard
17:36:52< vasi> (on my box, i use a couple of scripts involving my .xinitrc)
17:37:21< vasi> !seen drm
17:37:21< Melian> drm <n=drm@ip68-108-245-119.sb.sd.cox.net> was last seen on IRC in channel #fink, 2d 17h 24m 7s ago, saying: 'vasi: i'm going to bed...we can discuss this another time'.
17:39:27< Gavrila> vasi thanks it worked like a charm
17:39:31< vasi> yay
17:39:58< dmacks> Bill Scott has some great DISPLAY shell-dotfile snippets.
17:40:17< vasi> url?
17:40:25< Gavrila> anyway open-x11 is fine for me
17:40:36< dmacks> One moment...still fixing nomaintainer
17:42:37< vasi> k
17:44:25< dmacks> Okay, it defaults to 10.3|10.4T now.
17:44:52< vasi> thanks
17:47:29< dmacks> http://xanana.ucsc.edu/Library/init/zsh/environment*
17:52:06< cirdan> hey all
18:08:22< newmanbe> Hello.
18:09:18< vasi> dmacks, ah....what i do is put 'echo $DISPLAY >> /.xdisplay' in my .xinitrc
18:09:32< vasi> then i just have a simple command that reads that for the value of the current display
18:09:38< vasi> and exports it
18:14:13< cirdan> ok, what the hell is a .turd file?
18:14:56< newmanbe> Hey! Only Mr. Wacko Elitist can use language like that.
18:23:58 * cirdan adds newmanbe to .turd
18:24:08< cirdan> hey! i found a use for it :-)
18:25:09 * newmanbe hadds cirdan to .$ENEMIES
18:25:23< newmanbe> Where $ENEMIES=httplovers
18:26:28< newmanbe> The disgrace...
18:26:59< cirdan> hey, i've been hadd!
18:31:29-!- zizban [n=zizban@24-52-0-219.sbtnvt.adelphia.net] has joined #fink
18:52:40-!- dmacks_ [n=dmacks@pool-70-22-41-14.balt.east.verizon.net] has joined #fink
19:07:01-!- dmacks [n=dmacks@pdpc/supporter/active/dmacks] has quit [Read error: 110 (Connection timed out)]
19:21:13-!- Gavrila [n=Gavrila@213-140-16-182.fastres.net] has quit ["Leaving"]
19:32:29-!- baba [n=baba@YahooBB220041000080.bbtec.net] has quit ["This computer has gone to sleep"]
19:38:39-!- jake1 [n=catrin@pool-68-163-177-215.bos.east.verizon.net] has joined #fink
19:49:18-!- dmacks_ is now known as dmacks
20:09:33-!- Netsplit herbert.freenode.net <-> irc.freenode.net quits: Henk_Poley, JosephSpiros, pnorman
20:10:50-!- Netsplit over, joins: JosephSpiros, Henk_Poley, pnorman
20:20:19-!- RangerAway is now known as RangerRick
20:22:17-!- Gardner [n=mjg@pcp05047549pcs.ivylnd01.pa.comcast.net] has joined #fink
20:25:53< lisppaste> Gardner pasted "libao2 won't compile -- any ideas on a fix or workaround?" at http://paste.lisp.org/display/11633
20:28:54< zizban> that's from fink, right?
20:29:33< Gardner> Yes.
20:30:09< RangerRick> weird
20:30:28< RangerRick> I've got no idea, but I'm very surprised that's happening, unless you've got something weird in /usr/local
20:30:28< zizban> build lock error
20:30:48< RangerRick> that's not a bild lock error
20:31:11< Gardner> I started poking around for memset, memcpy, etc and I think there's something wrong with the include path.
20:31:40< Gardner> But hey, what do I know...?
20:32:20< zizban> ah
20:33:08< Gardner> I think I tried doing it with /usr/local moved out of the way. I'll try it again.
20:34:04< Gardner> Yep, blew up same way.
20:35:00-!- dmacks_ [n=dmacks@pool-70-22-44-137.balt.east.verizon.net] has joined #fink
20:39:44< dmacks_> Gardner: Try removing libquicktime0?
20:40:45< Gardner> OK, trying.
20:41:28< dmacks_> That package seems to have headers that mask the "normal" apple ones and break a lot of builds. Not sure if it's the case here, but it's an easy thing to try.
20:41:42< Gardner> Nope, it blew up same way.
20:41:43< dmacks_> D'oh:(
20:41:51< dmacks_> 10.3 or 10.4?
20:41:52< Gardner> 10.4
20:42:08< Gardner> SHRT_MAX is defined in /usr/include/ppc/limits.h
20:42:36< Gardner> Which should get brought in if you #include <limits.h>
20:42:59< dmacks_> What's it's value? (it's 32767 on 10.3)
20:43:24< Gardner> Same.
20:43:25-!- dmacks [n=dmacks@pool-70-22-41-14.balt.east.verizon.net] has quit [Read error: 110 (Connection timed out)]
20:43:26< RangerRick> newmanbe: so have you put your name on that gopher package and submitted it to the package tracker yet?
20:43:39-!- dmacks_ is now known as dmacks
20:45:28< dmacks> RangerRick: I'll let you handle committing it...I assume you'll find it to your liking?
20:46:11< RangerRick> I packaged it.  I just don't want to maintain it.  :)
20:46:20< RangerRick> it needs a name stuck on it, that's about it
20:46:34< zizban> I'd volunteer but I have no clue about it ;)
20:46:38 * dmacks knows, hence /me made joke that was funny in /my head:)
20:46:55< RangerRick> wasn't sure  :)
20:47:24< zizban> heh
20:47:24< dmacks> (that's why I assigned some pkg-requests tracker about gopher to you.
20:48:22< RangerRick> yeah
20:48:23< RangerRick> saw that
20:49:54< dmacks> Are we declaring dmalloc a lost cause for now? I just got email from an active maintainer who: 1) is pissed that zsh is busted, 2) knows the fix, and 3) offered to apply it and take over as Maintainer.
20:50:22< vasi> hmm...Debian user just got a Mac Mini....tried to tell him about Fink, he says "someone at Apple told him it was almost unusable, and that he should use dports instead"
20:50:28< Clef> hi
20:50:35< vasi> now i know how netscape felt :-(
20:51:17 * Clef says give zsh to the guy.
20:51:19 * dmacks went to Apple store today, had to go to 5 employees before I found 2 that agreed about a very basic hardware compatibilty issue.
20:52:03< Gardner> Ha. Darwinports only *just* got usable.
20:52:10 * Clef got a nice letter from verizon offering him FIOS 15 MB down / 2 MB up for $44.95.
20:53:27< vasi> RR, it was in personal email, you saw something else :-)
20:53:56< Gardner> Regarding libao2... It looks like its configure script only includes limits.h if __STDC__ is defined. Is it usually?
20:54:08< vasi> hey, we like dports here...we just like fink more :-)
20:54:50< Clef> [03:18] 	-->	dmalloc (n=mule@b36F2.static.pacific.net.au) has joined #fink
20:54:50< Clef> few days ago
20:54:52< Clef> au?
20:55:24< Clef> [03:40] 	<--	dmalloc has quit ("Si hoc legere scis nimium eruditionis habes")
20:55:40< Clef> libao2 is not a fun pkg to maintain
20:56:13< RangerRick> vasi: what was in personal email?
20:56:20 * RangerRick is not sure what you're talking about
20:56:29< vasi> oh never mind, too many conversations at once
20:56:55< vasi> Gardner, there's a command to dump all macros
20:57:01< vasi> er, a gcc option
20:57:11< vasi> don't quite remember what it is though, check the manpage
20:57:46< Gardner> I'm sure ilibao2 isn't fun, but unfortunately Gaim depends on it.
20:58:22< Clef> use the binary
20:58:28-!- RangerRick is now known as RangerAway
20:58:42< Gardner> I did, I just don't want it barfing at me whenever I do an update-all.
20:58:43< Clef> hmm
20:59:44< Gardner> vasi: -g3 ?
20:59:54-!- zizban [n=zizban@24-52-0-219.sbtnvt.adelphia.net] has quit []
21:00:20< vasi> i dunno
21:01:39-!- baba [n=baba@YahooBB220041000080.bbtec.net] has joined #fink
21:01:52< Gardner> Ah, cpp -dD
21:02:46< Gardner> No __STDC__ here, just __STDC_HOSTED__. Blah.
21:03:24-!- drm [n=drm@ip68-108-245-119.sb.sd.cox.net] has joined #fink
21:03:42< drm> howdy
21:03:52< vasi> hi drm
21:03:53< dmacks> helo
21:04:03< drm> so, about dist-upgrade
21:04:07< vasi> ah yes
21:04:47< vasi> what's your plan, so far?
21:05:01< vasi> (i promise not to run off this time)
21:05:02< drm> basically, fink has to have a way of checking "what's the latest version I could upgrade to, given the OS that's installed", then it has to create/update the new tree, then
21:05:40< drm> probably update essential packages, and warn the user that he/she should run "fink update-all" to finish the process
21:06:03< drm> creating/updating the new tree will be done differently, depending on the upgrade method
21:06:18< drm> (i.e. cvs vs point vs rsync)
21:06:19< vasi> differently = rsync or cvs?
21:06:22< vasi> ah ok
21:06:37< vasi> um, so we're punting on what happens once the essentials are upped?
21:06:55< drm> well, i don't like forcing an upgrade-all
21:07:02< drm> but i'm willing to debate the point
21:07:07< vasi> yeah, it's just easy for users to break things
21:07:15< vasi> well, for fink to break things for users
21:07:23< drm> (the new tree is supposed to contain higher rev numbers of stuff that needs upgrading)
21:08:06< vasi> eg: spam is installed, eggs and ham depend on spam (and everything's got GCC: <blah>)
21:08:23< vasi> user does 'fink update eggs', so eggs and spam are updated
21:08:50< vasi> now spam is GCC: X and ham is GCC: X+1....so *boom*
21:09:19< drm> good point
21:09:27< vasi> basically, updating piece-meal causes things to break, in about all non-trivial cases
21:09:59< drm> so how about a message like this: "your fink installation will be updated to the new tree BLAH, and all installed packages will be updated...do you want to continue?"
21:10:07< vasi> the debian folks are still working on this issue, in a future version of dpkg you'll be able to do something like Depends: foo (+blah) to ensure that foo has the blah "flag"
21:10:21< vasi> but that's in the future
21:10:39< vasi> well that will cause other breakage, because it's certain that at the start some packages are missing from the new dist
21:10:42< vasi> or broken
21:10:59< drm> why is it certain? :)
21:11:11< vasi> because it's happened every time in the past :-)
21:12:01< vasi> the only thing the would NOT break is to basically remove everything, then re-install in the new dist...keeping the existing .debs for non-GCC package
21:12:03< vasi> i'm not saying we should do that...
21:12:12< dmacks> That's going to be an issue whether we force "top-down" or "bottom-up" upgrading though: if any package doesn't compile, everything "above" it will be broken 'cuz of GCC mismatch.
21:12:18< vasi> just that we have a choice between multiple imperfect solutions
21:12:41< drm> well, for users of the stable tree, i will have checked very carefully that everything with a GCC tag compiles
21:12:51< drm> can't say the same about unstable though
21:13:33< drm> but the plan there is, to not add pkgs that have a GCC tag to the new tree until they compile...that doesn't really address the upgrade prob thoughk, does it?
21:13:47< dmacks> nope
21:14:06< vasi> they'll just sit broken
21:14:23< drm> it also doesn't addrss the issue of pkgs in the unstable tree which should have a GCC tag but don't
21:14:42< drm> however, if my task is to repair the entire unstable tree before creating 10.4, then 10.4 will never get created
21:15:06< vasi> a further problem relates to the .debs the user already has
21:15:23< drm> how so?
21:15:30< vasi> even if we come up with some system that completely updates the current install...
21:15:37< vasi> the user still has debs around with the old GCC field
21:15:45< vasi> which will break if they're ever installed
21:15:46< drm> well, probably not, actually
21:15:58< drm> they would like in the old tree
21:16:10< vasi> ok, we could do that
21:16:18< drm> s/like/live/
21:16:30< vasi> we should definitely have a way for the non-GCC debs to be pushed forwards to the new tree
21:16:31< drm> oh wait, they have symlinks in /sw/fink/debs though, don't they?
21:16:42< dmacks> Those links are via dists/
21:16:43< vasi> i don't think fink looks in there....but we can clean it up
21:17:08< drm> well, i don't think people should reinstall debs from the old tree
21:17:12< dmacks> ...so they would cease to be usable exist as soon as dists/ changes.
21:17:22< dmacks> s/usable//
21:17:28< drm> yeah
21:17:48< drm> (by the way, does "fink cleanup" look for .debs to remove in non-current trees?)
21:17:49< vasi> drm, why not for .debs that are still current?
21:18:00-!- Henk_Poley [n=Henk_Pol@poley.xs4all.nl] has quit [Read error: 104 (Connection reset by peer)]
21:18:02< vasi> drm, i don't believe it does...how would it know they exist?
21:18:34-!- Henk_Poley [n=Henk_Pol@poley.xs4all.nl] has joined #fink
21:18:34< drm> vasi: because there are other changes when a tree changes...for example, the 10.3->10.4foo upgraders will have MACOSX_DEPLOYMENT_TARGET set to a different value, and so on
21:19:00< drm> in 10.4, we really truly abandon prebinding, for the first time
21:19:01< vasi> drm, ok then....so if packages are different, then why are we keeping the installed ones?
21:19:08< drm> (was left active in 10.4-T, due to oversight)
21:19:21< vasi> either the debs and installed pkgs are both good, or both bad
21:20:02< drm> vasi, dmacks: suppose we give users a choice upon 'fink dist-upgrade', of either updating all installed packages, or re-compiling new versions (with a brief explanation about the risk/benefit analysis)
21:20:55< vasi> we'll get lots of confused mail...
21:21:29< dmacks> (cleanup_debs removes .deb from %p/fink/dists/ that do not match a .info in user's active trees (or dpkg status) and all broken links in %p/fink/debs)
21:21:29< drm> i agree, that there is no point in throwing away non-GCC debs, if we are gonna allow installed pkgs to stay
21:21:48< vasi> this is a situation where the dports idea of having multiple libs 'isntalled' but not 'activated' (via symlinks) would be useful
21:22:01< drm> (well, cleanup_debs should also examine inactive trees IMHO)
21:22:28< dmacks> (that would require indexing them)
21:22:39< vasi> yeah, that's what i was about to say :-)
21:23:10< vasi> i think the 'figure out what to do with the old debs on dist-up' should be a separate operation
21:23:11< vasi> from cleanup_debs
21:23:20 * dmacks agrees
21:23:47< drm> (that's why it was parenthetical :)
21:23:51< dmacks> (okay:)
21:24:28< dmacks> Remind me again why we aren't doing lib/ and lib-gcc4?
21:24:45< vasi> that's not a bad idea, actually
21:24:52< drm> ok, how about this: when upgrading, the old %p/fink/debs is copied into the new one with "s/dists/old-tree-name/" applied
21:25:08-!- emes [n=emes@pdpc/supporter/student/emes] has joined #fink
21:25:30< emes> am I the only person who can't update-all on unstable (10.4)?
21:25:35< drm> dmacks: well, as I recall, it was pogma who was violently opposed to this
21:25:42< vasi> drm, how to clean up debs can wait...it's more important to figure out the actual overall strategy
21:25:49< vasi> !summon pogma
21:25:50 * Melian takes out 20 clean, identical-looking phones, some extra hands, and pretends to be a telemarketer for a large corporation, so she gets delivered a phonelist containing pogma's coordinates.
21:26:02< vasi> emes, wanna paste your error?
21:26:02< emes> it always dies on automake1.9 and automake1.7
21:26:02< drm> vasi: the last remark was not about cleaning up debs, it was about reusing them
21:26:15< dmacks> Then we could hang GCC into the %n so user can have both libfoo's installed, and eventually nuke the old-gcc's one when everything's upgraded.
21:26:30< dmacks> But in the mean time he can do piecemeal and never break in the middle.
21:26:55< emes> vasi: http://pastebin.com/362171
21:27:09< dmacks> (we could use the OBSOLETEs mechanism)
21:27:11< vasi> dmacks, that has the annoying aspect of renaming every last package
21:27:16< vasi> well, GCC package
21:27:32 * drm hates to think what happens when somebody sets DYLD_LIBARY_PATH to /sw/fink/lib and half of the libs are elsewhere
21:27:48< vasi> so users can no longer do 'fink install somepackage', they now have to know about the gccX suffix
21:28:27< vasi> emes, you have a stray buildlock....try removing it with dpkg
21:28:40< vasi> 'dpkg -r fink-buildlock-kdeutils3-3.4.1-21'
21:28:57< drm> dmacks: the other problem is that there may be files not in /sw/lib and so there would be overlaps
21:29:00< drm> between the pkgs
21:29:16< drm> vasi, emes: the last command needs "sudo" before "dpkg ..."
21:29:39< emes> drm: yeah, i figured that out :)
21:29:42< emes> vasi: thx
21:29:50< dmacks> vasi: could be handled automatically by fink though. 
21:29:54< dmacks> drm: Hmm...that's a problem.
21:29:54< vasi> dmacks, not by apt though
21:30:07-!- lisppaste [n=lisppast@common-lisp.net] has quit ["Common Lisp IRC library - http://common-lisp.net/project/cl-irc"]
21:30:12-!- lisppaste [n=lisppast@common-lisp.net] has joined #fink
21:30:14< dmacks> True.
21:31:01< vasi> ok, how about we let the user try to update-all...but we keep all the old debs around for now
21:31:09< vasi> if the update-all succeeds, we get rid of the old debs
21:31:23< drm> the mechanism we have used in the past, and are in the midst of implementing for this transition, is imperfect but it will leave users with something reasonable at the end
21:31:23-!- Gardner [n=mjg@pcp05047549pcs.ivylnd01.pa.comcast.net] has quit ["Bye!"]
21:31:24< drm> (even if they tear out all their hair in the middle)
21:31:24< vasi> otherwise, we restore the old fink installation, and let them try again later
21:31:57 * dmacks bbl
21:31:58< vasi> so we basically do the 'start from scratch' thing, but have a safety valve if things break
21:32:00< vasi> k dmacks
21:32:13< drm> vasi: restore the old one? how? they just installed a bunch of new pkgs, and then failed on some update or other
21:32:30< drm> so they are in a mixed state
21:32:42< vasi> drm, the process looks like this:
21:33:05< vasi> 1. Notice we have a new dist available, give the user the choice to dist-up or wait for now
21:33:28< vasi> 2. User tries dist-up, so we download the new tree and add it to fink.conf, remove the old one from fink.conf
21:33:44< vasi> 3. We update the essentials unconditionally
21:34:07< vasi> 4. We remove everything non-essential, and try to build them all from scratch (or download via apt)
21:34:20< vasi> 5. On success, get rid of the old tree, we're done
21:35:01< vasi> 6. On failure, remove everything non-essential, and use the old .debs from their old tree to restore their setup exactly as it was
21:35:13< drm> what if some old .debs are missing?
21:35:23< vasi> er, yeah that could be a problem :-)
21:35:47< vasi> there's no good reason for them to go missing, if some are we can offer to build them
21:35:48< drm> and what if one of their packages is openoffice.org? or even kdelibs3?
21:36:14< drm> vasi: fink cleanup might have wiped them out
21:36:18< vasi> drm, there is NO solution that doesn't require the user to spend at least some time with a non-working KDE
21:36:52< vasi> either KDE isn't present/working while the new kde is building, or KDE is broken while kdebase and kdelibs have mismatched GCC
21:37:22< vasi> (er, i guess if we had a radical change like %p/lib-gcc4 we could have KDE working always...but that's a BIG change)
21:37:50< drm> yeah, i'm not in favor... i consider that decision to have been made back in feb. or march.
21:38:32< drm> anyway, its dinner time for me...i'll look in again later to see if you're still here, and we can continue...i gotta think some more, too
21:38:48< vasi> okie doke, seeya later
21:38:51< vasi> (dinner? must still be on the west coast)
21:39:08< drm> yeah, i'll be on the west coast until May
21:39:30< drm> except when traveling, of course :)
21:39:41-!- drm [n=drm@ip68-108-245-119.sb.sd.cox.net] has quit ["Leaving"]
21:45:00< newmanbe> RangerAway: No, I haven't fixed the location of the man pages.
21:51:32-!- Murr [n=neeri@A17-202-20-71.apple.com] has quit [Read error: 110 (Connection timed out)]
21:54:11-!- baba [n=baba@YahooBB220041000080.bbtec.net] has quit ["Leaving"]
22:07:46-!- emes [n=emes@pdpc/supporter/student/emes] has left #fink ["Leaving"]
23:04:17-!- dmacks is now known as dmacks_away
23:06:46-!- dmacks_away [n=dmacks@pdpc/supporter/active/dmacks] has quit ["leaving"]
23:09:07-!- drm [n=drm@ip68-108-245-119.sb.sd.cox.net] has joined #fink
23:24:43< vasi> hey drm
23:24:49< drm> hey vasi
23:25:03< vasi> TheSin seems to be back in commission
23:25:09< drm> cool
23:25:16< vasi> so hopefully he'll fix up Shlibs
23:25:25< drm> yeah
23:26:28< vasi> just curious, what's up on the west coast? sabbatical or something?
23:27:33< drm> got it in one...sabbatical
23:28:02< drm> fall semester at a physics institute at U.C.Santa Barbara, spring semester at a math institute at U.C.Berkeley
23:28:42< vasi> sounds cool...i guess in Cali they can't call it a 'winter semester' <g>
23:29:13< drm> well, actually UCSB is on the quarter system, so they have a winter quarter
23:29:56< vasi> ah...damn, ruined my joke
23:30:17< vasi> any further thoughts about dist-up?
23:31:59< drm> i don't think fink can really handle uninstalling everything non-essential, keeping a list of what was installed, and then reinstalling
23:32:40< vasi> justify that statement :-)
23:32:46< drm> what with troubles with buildconflicts and so on
23:32:59< drm> there is a good chance that user intervention will be required
23:34:20< vasi> how about we combine the idea of reversion with a piecemeal update
23:34:22< vasi> ?
23:34:35< vasi> so we update the essentials as per usual
23:34:50< vasi> and then leave everything installed, and counsel the user to do a full update-all
23:35:32< vasi> then we keep a flag set until everything is updated...and until that flag is cleared (by fink noticing it's all upped), you have the option to revert to your previous install, as best as fink can manage
23:36:02-!- shres [n=sshreyas@59.92.139.185] has joined #fink
23:36:07-!- xhrl [n=ThomasW@S0106000f3d5d5bed.vs.shawcable.net] has joined #fink
23:36:39< vasi> so maybe every time you try to use fink, it'll warn "hey, you still have unupdated stuff, if things break you can try 'fink revert'"
23:36:42< drm> unless the "previous install" was generated by making a copy of /sw to /sw.bak, i don't think i trust fink to be able to revert under all circumstances...the attempt to revert could leave you in a worse state
23:37:41< drm> yeah, actually, why don't we do that?
23:38:07< drm> again, the user could opt out (will be given an opporutnity to opt out)
23:38:14< drm> which is necessary in case there isn't a lot of disk space on the machine
23:38:19< pogma> hi
23:38:22< drm> hi pogma
23:38:39< drm> oh, its sept 15 holiday?
23:39:04 * drm tries to remember... respect for aged, perhaps?
23:39:11< pogma> sept 13th, holidays have moved around abit due to the "Happy Mondays" campaign
23:39:21< pogma> so that holiday is next monday
23:39:21< vasi> "happy mondays"?
23:39:21< drm> ah
23:39:57< pogma> vasi: public holidays in japan all used to fall on specific dates, which sucked if the date was a saturday
23:40:12< vasi> so they moved them all to mondays?
23:40:12< drm> this was also true of other countries
23:40:17< pogma> koizumi & co decided to move a bunch of holidays to always fall on mondays
23:40:34< drm> the usa moved a lot of them to mondays too
23:40:35< vasi> yeah, we do it with jewish fast days too....if one falls on the Sabbath, it gets postponed to Sunday
23:42:14< drm> pogma, is the only advice we can give for the libJPEG mess "avoid setting DYLD_LIBRARY_PATH"? wasn't there some other workaround that i'm forgetting?
23:44:04< pogma> use a case sensitive file system :)
23:44:06< drm> yeah, ok :)
23:44:07< drm> this has bitten openoffice.org recently
23:50:10-!- beniamino [n=ben@naf-37.OLAC.Berkeley.EDU] has joined #fink
--- Log closed Tue Sep 13 00:00:28 2005

.