--- Log opened Tue Aug 23 20:03:37 2005
20:03 < newmanbe> Yet another test.
20:03 < newmanbe> I need to rm the folders that don't get served...
20:07 -!- msachs [n=msachs@c-24-34-72-223.hsd1.ma.comcast.net] has quit [Read error: 110 (Connection timed out)]
20:08 -!- citizen_0 [n=user@va-sterling-u1-c5c-a-146.chvlva.adelphia.net] has quit ["This computer has gone to sleep"]
20:08 < gopherd> Hi, I'm gopherd the logging bot! You can view todays logs at gopher://newmanbe.homeunix.net/0/irclogs/freenode/fink/2005/08/fink-20050823 Enjoy!
20:08 < gopherd> 8)
20:08 < newmanbe> It seems to be working.
20:09 < newmanbe> It's not really automated though.
20:09 -!- zizban [n=chris@24-52-0-219.sbtnvt.adelphia.net] has joined #fink
20:10 < newmanbe> Comments?
20:11 < akh> zizban's the one for comment.
20:12 < akh> So search is via the standard text search?
20:12 < zizban> ummm what was the question?
20:13 < akh> None--I just made reference to that package which you maintain.
20:13 < akh> (wordplay)
20:13 < akh> gopherd: logs
20:13 < zizban> ah :)
20:13 < akh> gopherd:  logs?
20:15 -!- vasi [n=vasi@modemcable133.147-70-69.mc.videotron.ca] has joined #fink
20:15 < msachs_> "Note that the training sets for [data sets foo, bar, and baz] are obviously noisy since [our heuristic] outperforms [the control heuristic.]"   Riiight...
20:16 < msachs_> er, "[the control heuristic] outperforms [our heuristic]." rather.
20:16 < akh> Damn, damn, damn.  I knew I shouldn't have installed the legacy tools.  Now Fink thinks it needs to install gcc3.1
20:16 < akh> And why do I have gcc2 but not gcc2.95 ?
20:17 < msachs_> akh: Why'd you decide to install them, just out of curiousity?
20:17 < akh> msachs_: curiosity ;-)
20:17 < msachs_> Mmm.  We'd love to get rid of them, but we think that maybe someone needs them for some reason.
20:18 < akh> dmacks and I wanted to investigate if they were the same files from XCode 1.5:  I had a Tiger box, and he was on Panther.
20:18 < akh> Just wild craziness.  :D
20:18 < msachs_> That's probably the major download driver...
20:19 < akh> Well, we do have a few packages that need 3.1...
20:19 < msachs_> Just because nobody's bothered to make them work with 3.3 or 4.0 because those compilers are stricter about some things, or what?
20:21 < akh> I know of one that's ancient and would require recoding for any later gcc.
20:21 < akh> (so the former)
20:21 < newmanbe> akh: Yeah, that's it for search.
20:21 < msachs_> k
20:21 -!- msachs_ is now known as msachs
20:21 < newmanbe> UWM's gopherd has built in search.
20:22 < newmanbe> But you need a case-senstitive file system to build it.
20:23 < newmanbe> Which Apple apparently doesn't like.
20:24 -!- TheSin [n=TheSin@iphost-64-56-130-194.edm.wiband.net] has joined #fink
20:24 < akh> msachs: I'm not sure how many packages that need gcc3.1 that Fink still supports, to be honest.
20:25 < msachs> newmanbe: I think you could make a UFS dmg or something if you don't have a spare partition and you really want to build it.
20:25 < newmanbe> I've tried the disk image.
20:25 < vasi> newmanbe, you can make a case-sensitive HFS+ partition
20:25 < newmanbe> And Disk Utility won't let me have a UFS partition with HFS+ on another partition.
20:25 -!- TheSin [n=TheSin@iphost-64-56-130-194.edm.wiband.net] has quit [Client Quit]
20:26 < newmanbe> Which I suppose I could do using CLI...
20:26 < newmanbe> It could have been a different issue that kept it from building. Dunn'o.
20:26 -!- akh [n=akhansen@68-118-244-23.dhcp.oxfr.ma.charter.com] has quit []
20:29 < newmanbe> Oh yeah, that wasn't the problem.
20:29 < newmanbe> Well, it was, but I got past it.
20:31 -!- akh [n=akhansen@68-118-244-23.dhcp.oxfr.ma.charter.com] has joined #fink
--- Log closed Tue Aug 23 20:35:09 2005
--- Log opened Tue Aug 23 21:05:00 2005
21:05 < gopherd> Yay! I'm logging again! Silly permissions issues.
21:08 < baba_> oh dmacks, i was writing an email to you
21:09 < baba_> (also to -devel)
21:09 < dmacks> Okay.
21:10 < dmacks> vasi: RangerAway posted an interesting engine regression on the 0.25-blockers wiki page.
21:10 < vasi> the one about build and apt?
21:11 < dmacks> Naw, about reinstalling splitoffs during a rebuild.
21:12 < vasi> the darn RSS feed never seems to work for me
21:12 < vasi> is it verified?
21:13  * dmacks goes to try it.
21:14  * vasi too
21:14 < akh> vasi:  dunno--works for me.
21:15 < dmacks> Seems to work for me also.
21:15 -!- msachs [n=msachs@pool-141-154-68-131.bos.east.verizon.net] has quit []
21:15 < vasi> work the way it's supposed to?
21:16 -!- baba_ [n=baba@YahooBB220041000122.bbtec.net] has quit ["Leaving"]
21:18 < akh> vasi:  yeah--on Safari at least.
21:18 < vasi> oh woops, you mean the RSS feed works
21:19 < vasi> i can read it fine, that's not the problem
21:19 < vasi> the thing is that bloglines never seems to notice when it's updated
21:19 < vasi> i dunno why
21:19 < akh> hmm...haven't looked into it.
21:20  * dmacks was talking about RangerAway's regression; I don't play rss games.
21:20 < vasi> dmacks, can you look at http://sourceforge.net/tracker/index.php?func=detail&aid=766799&group_id=17203&atid=117203
21:20 < dmacks> (I just created a package+splitoff; installed both; rebuilding either one triggers reinstallatioin of both)
21:22 < dmacks> vasi: I get "Failed: no package found for specification '%s'" in all cases it appears.
21:22 < vasi> i'll remove it from the blockers page, RangerAway should use the tracker
21:23 < dmacks> I'll close that tracker item.
21:23 < vasi> woo, another one bites the dust!
21:23 < akh> yay
21:24 < vasi> hmm, here's a stumper: let's say there's foo 1.0-1, and also bar 1.0-1 which provides foo
21:24 < vasi> if the user does 'fink remove foo', should it remove A) just foo, B) just bar, C) both?
21:25 < newmanbe> Whatever would surprise the user the least.
21:25 < newmanbe> Which depends a lot on the user...
21:26 < dmacks> There is a hair's-width of wisdom in debian-devel's position that one should not Provides things that are also real packages.
21:26 -!- zizban [n=chris@24-52-0-219.sbtnvt.adelphia.net] has quit ["Leaving"]
21:26 < akh> yah
21:27 < vasi> well given that in the real world, we have packages that do that
21:27 < vasi> what should happen?
21:28 < dmacks> Remove the real one only. That's the way it's always been.
21:29 < dmacks> User has ways to find out that there's also a provided pkg and what provides it, and can go from there.
21:30 < dmacks> While we're considering crazy package-naming overlaps, what happens when user has installed a real package that is also an existing VirtPackage?
21:30 < akh> like gcc3.1
21:30 < dmacks> Roight.
21:30 < vasi> gcc* is a special case, i hacked it so the virtpackages don't overwrite the real ones
21:30 < vasi> because apt has trouble with that
21:31 < vasi> are there any other cases where this is an issue?
21:32 -!- newmanbe [n=newmanbe@tor/session/x-fac145742787e70e] has quit ["Returning shortly from screen"]
21:33 -!- newmanbe [i=[U2FsdGV@tor/session/x-613c29f86243c0d7] has joined #fink
21:33 < dmacks> I think that's the major one. *maybe* cups-dev?
21:35 < dmacks> Is there any reason to leave this one open? http://sourceforge.net/tracker/index.php?func=detail&aid=1208178&group_id=17203&atid=117203
21:36 < vasi> well it's still a bug, eh?
21:37  * gopherd wonders if he can have irssi scream at people if if someone says something with http:// in it...
21:37 < gopherd> s/he/it/
21:37 < dmacks> Well technically he solved it:)
21:38 < akh> And it looks like a heisenbug
21:38 < dmacks> Yeah, all we gotta do is run time backwards and find the exact descriptor of his initial state.
21:39 < vasi> damn, mozilla is unmaintained? that sucks
21:39 < dmacks> ya
21:39 < newmanbe> vasi: No one wants to spend the time to build it....
21:40 < vasi> oh hey, there's a new fink developer who wants to include multiple versions of unison in fink (unison2.9, unison2.10, etc) since they use different protocols
21:40 < vasi> i like the idea, any objections?
21:40 < dmacks> Makes sense.
21:40 -!- Ksilebo is now known as Mancat_Fish
21:40 < akh> fair enough--though 2.9 doesn't do resource forks, IIRC
21:41 < vasi> akh, some linux distros only have 2.9 though, so it's kinda important for some people to have it in fink
21:41 < dmacks> Does anything Depends on it?
21:41 < akh> vasi:  Yeah, true.
21:41 < akh> dmacks:  I don't believe so.
21:41 < vasi> dmacks, not as far as i know
21:42 < dmacks> cool
21:42 -!- Mancat_Fish is now known as Ksilebo
21:43 -!- Ksilebo is now known as Manfish_Cat
21:45 -!- Manfish_Cat is now known as Ksilebo
21:47 -!- akh [n=akhansen@68-118-244-23.dhcp.oxfr.ma.charter.com] has quit ["Leaving"]
21:48 < dmacks> vasi: 'fink remove foo' could check if 'foo' matches more than one "installed thing" and ask what to do? Same for 'fink install foo'.
21:49 < vasi> yeah, that makes sense
21:52 < lisppaste> dmacks pasted "Inconsistent behavior" at http://paste.lisp.org/display/11057
21:57 < vasi> uh, shouldn't it complain about the fake package?
21:57 < dmacks> Yup
21:58 < dmacks> (or not do any sanity-checking at all and just pass off to dpkg)
21:58 < vasi> uh, what's the different between get_pkglist and expand_packages?
22:01 < dmacks> Different names and argument prototypes?
22:01 < vasi> whoa, 'fink remove' takes a regex!?
22:02 < vasi> grrr, what's the *semantic* difference? when should i use one over the other?
22:02 < dmacks> I'll pretend I knew that and say "yeah, cool 'fink remove' feature eh? I thought *everyone* knew about it!"
22:03 < dmacks> No clue actually.
22:03 < vasi> *sigh*
22:03 < dmacks> phone...bbl
22:03 < vasi> i knew that would be your answer, i just need more righteous indignation in my life or something :-|
22:22 -!- KraMer [n=mark@adsl-70-240-178-94.dsl.hstntx.swbell.net] has quit [Read error: 110 (Connection timed out)]
22:37 -!- dk0r [n=dk0r@cpe-24-194-171-176.nycap.res.rr.com] has quit []
22:37 -!- KraMer [n=mark@adsl-70-240-199-178.dsl.hstntx.swbell.net] has joined #fink
22:40 -!- shres [n=sshreyas@59.92.138.175] has joined #fink
22:41 -!- pogma_ [n=peter@p4026-ipad32kobeminato.hyogo.ocn.ne.jp] has joined #fink
22:53 -!- kbroderick [n=kbroderi@rich-dsl6-236.greenmountainaccess.net] has quit []
22:55 -!- pogma [n=peter@opendarwin/developer/pogma] has quit [Connection timed out]
23:18 < dmacks> vasi: the val-unpack branch is now tagged in the whole perlmod/Fink directory if you want to play with it.
23:19 < vasi> kewl
23:19 < dmacks> Speedup 50-90% for .deb validation and consolidated error reporting. I guess I could put in some sleep() if you like the old behavior:)
23:20 < dmacks> s/and/, and/
23:29 -!- msachs [n=msachs@c-24-34-72-223.hsd1.ma.comcast.net] has joined #fink
23:36 -!- eno-away is now known as eno
23:59 -!- asari [n=asari@p2209-ipbf910marunouchi.tokyo.ocn.ne.jp] has quit ["Quitting!"]
--- Log closed Wed Aug 24 00:00:43 2005

.