00:34 HavokMalo: Hooray!
00:34 HavokMalo: You folks might want to update the website - worked just fine with Ubuntu 10.04 x64
00:34 jonrafkind: eli, maybe you could amke that clearer?
00:40 HavokMalo: eli has access to the site directly? Woah, cool
00:43 eli: jonrafkind: Make what clearer?  (Didn't follow everything.)
00:43 eli: HavokMalo: Which site?
00:43 jonrafkind: download.racket-lang.org
00:44 jonrafkind: make the entry for jaunty clear that it works on 10.04 as well, or add another entry for 10.04 that points to the same file or something
00:44 eli: I have no idea what's the difference between jaunty or whatever.
00:44 jonrafkind: me neither, but apparently racket from jaunty works fine on it
00:46 eli: I have no idea how to even tell which ubuntu version the machine is using, I just write what I'm told.
00:46 HavokMalo: lol
00:46 eli: And of course the fact that it compiled on one version doesn't mean that it will run exactly on that version.
00:47 jonrafkind: are you saying having a jaunty release is pointless?
00:47 HavokMalo: Yeah
00:47 HavokMalo: Or it won't run without complications. But still - it is disconcerting to newer people if the distribution versions don't match.
00:50 eli: jonrafkind: The label is only saying which configurayion it was compiled on, not which configuration it's expected to work with.
00:50 jonrafkind: ok, but thats compeltely worthless information
00:50 jonrafkind: unless you assume that compiled on X will work on X
00:51 eli: I have no idea what you mean.
00:51 eli: If it's completely worthless then what's the point of writing any platform names at all?
00:51 jonrafkind: why do I care if you compiled on a windows machine or a freebsd machine, as long as it works on my linux machine? maybe you used a cross compiler
00:51 HavokMalo: I thought it was simply what platforms had been tested
00:51 jonrafkind: because your suppsoed to write names that mean "it will work on X"
00:51 HavokMalo: I just wanted to add that it appears to work on 10.04
00:52 eli: jonrafkind: If it says X, then it was built and tested on X.  That's all.
00:52 eli: There's nothing in there that says that it will not work on any other Y.
00:52 jonrafkind: irght.. but making users figure that out is not nice for them
00:52 HavokMalo: I was confused.
00:52 eli: Figure what out?
00:52 jonrafkind: that X will work on Y
00:53 jonrafkind: especially when you have direct evidence that it is so
00:53 jonrafkind: "will jaunty work on 10.04? i have no idea! lets find out. oh it does"
00:53 eli: Are *you* going to maintain this table?
00:53 jonrafkind: well if it does then the website could say that
00:53 jonrafkind: no
00:54 jonrafkind: it could say "usually works on 10.04" maybe
00:54 eli: No.
00:54 eli: Ubuntu and Fedora versions change every ~6 months.
00:54 HavokMalo: 1 line, every 6 months.
00:54 jonrafkind: i dont think 10.04 is going to change
00:54 jonrafkind: they will come out with 10.10, thats a different version
00:54 eli: I'm not going to maintain that table.
00:55 HavokMalo: lol
00:55 eli: I'm not going to even try now to run 5^2 installation to figure out the current table.
00:55 jonrafkind: it seems like a one-time edit to me, when someone says "yes, the jaunty build works on $foo" then you update the table
00:55 eli: Actually, 2*5^2, since the textual installers are likely to be different.
00:55 eli: No, it's not a one-time edit.
00:56 eli: What happens when some dependencies change?
00:56 eli: For example, when it switches to use gtk?
00:56 eli: (it = gracket)
00:56 jonrafkind: then remove the table at that point
00:57 eli: What you're suggestion requires running 2*5^2 full tests, on every build.
00:57 eli: If you're going to do that, then I'll include that.
00:57 eli: I'm not going to do that even now.
00:57 eli: s/now/once/.
00:58 jonrafkind: HavokMalo, sorry, you'll just have to guess at stuff
00:58 HavokMalo: Couldn't we put a package in the repository to ensure dependencies are handled properly?
00:58 eli: HavokMalo: That depends on how fast the ubuntu repository gets updated.
00:58 eli: jonrafkind: What's so special about that?
00:58 eli: jonrafkind: Can you find a single place that provides that kind of information?
00:59 jonrafkind: yea, HavokMalo just said it worked
00:59 eli: jonrafkind: Can you find a single place that provides that kind of information?
00:59 jonrafkind: no
00:59 jonrafkind: look theres a 99.9% chance the jaunty build is going to work on 10.04
00:59 jonrafkind: in the exceptionally rare case that it doesnt, then update the website
00:59 jonrafkind: otherwise you're going to have confused people coming around
01:00 eli: OK, I'm dropping this conversation.
01:00 HavokMalo: I do like the idea of a repository entry.
01:00 HavokMalo: k
01:01 eli: HavokMalo: Sorry for the confusion, but what jonrafkind is suggesting (and is not in any way volunteering to do or even help with) is simply insane.
01:01 jonrafkind: second best suggestion is to put that information on the web somewehre, but we dont have a wiki and i dont know a place people would look
01:02 jonrafkind: unless our racket logs get indexed by google
01:02 HavokMalo: A wiki would be awesome.
01:02 jonrafkind: a wiki is sort of being worked on (by me, but not recently), no promises though
01:39 (join) superjudge
01:45 (quit) HavokMalo: Ping timeout: 252 seconds
02:15 (quit) jonrafkind: Ping timeout: 276 seconds
02:45 (quit) superjudge: Quit: superjudge
03:56 (join) superjudge
05:06 (join) masm
05:09 (quit) superjudge: Quit: superjudge
06:23 (quit) masm: Quit: Leaving.
07:05 (join) masm
07:23 (join) superjudge
07:25 (quit) superjudge: Client Quit
08:24 (join) superjudge
08:32 (quit) superjudge: Quit: superjudge
08:42 (join) superjudge
10:12 (quit) superjudge: Quit: superjudge
10:58 (join) stamourv
11:42 (join) jonrafkind
12:21 (quit) sstrickl: Quit: sstrickl
12:23 (join) sstrickl
13:51 (nick) samth_away -> samth
14:16 (join) superjudge
15:12 (quit) superjudge: Quit: superjudge
15:55 (join) spoofy
15:58 spoofy: Hello. Will debian package of PLT-Racket be created sometimes? I mean, is it in plans?
15:59 (join) superjudge
15:59 spoofy: I tried to install from shell script - system-wide installation failed: "Error: installation failed".
16:01 stamourv: unix-style installation?
16:04 spoofy: yep
16:05 stamourv: there's a stray README file that the installer doesn't know what to do with
16:06 stamourv: to fix it, suspend the installation after everything has been extracted, delete the file manually, then resume
16:06 stamourv: this workaround should work for now, and the real problem has been fixed in git
16:07 spoofy: oh
16:07 spoofy: ok, thanks
16:22 (join) carleastlund
16:59 (quit) jonrafkind: Ping timeout: 265 seconds
17:04 (quit) superjudge: Quit: superjudge
17:47 (join) kronoz
17:47 kronoz: hi
17:48 kronoz: anyone alive here?
17:48 mattmight: Howdy.
17:48 kronoz: hi!
17:48 kronoz: I'm having embarrassingly lame problems with DrRacket :'-(
17:49 kronoz: basically, I am opening another scm file via load, e.g. (load "foo.scm"), then I reference a proc from that file, e.g. (foo 1 2 3), but drracket complains unbound identifier in module in: foo
17:49 kronoz: what am I doing wrong?! :'-(
17:50 kronoz: I am probably being dumb about this :-)
17:51 kronoz: mattmight: you there? Any ideas?
17:52 mattmight: hmm
17:52 kronoz: hmm indeed! :-)
17:52 mattmight: I can't imagine why that doesn't work.
17:52 kronoz: lol nor can I... grawrgh.
17:53 mattmight: I'm here mostly to ask questions about compiler internals.  I'm afraid I'm not yet a racket expert. ;)
17:53 kronoz: it cooooould be a misinstaled racket, as it died complaining about not being able to rm /etc/lib/plt even when sudo'd
17:53 kronoz: mattmight: lol fair enough :-)
17:53 kronoz: mattmight: as you can tell, I am faaar from being a racket expert myself!
17:55 samth: kronoz, you probably want to use `require', not `load'
17:55 kronoz: samth: ah thanks, will give it a go
17:55 kronoz: samth: can I require a path?
17:55 samth: that may mean that you have to change what's in "foo.scm" to be a module
17:55 samth: yes, you can require a path
17:56 samth: take a look at http://docs.racket-lang.org/quick/index.html#%28part._.Modules%29
17:57 kronoz: samth: thanks
17:59 kronoz: samth: now receiving errors of the 'compile: unbound identifier (and no #%app syntax transformer is bound)' variety
17:59 samth: from what?
18:00 kronoz: samth: this is someone else's library, he declares
18:00 kronoz: "Precedence Climbing" expression parser (like Shunting Yard) to incorp
18:00 kronoz: oops
18:00 kronoz: sorry
18:00 kronoz: grr @ clipboard
18:00 kronoz: samth: (provide (all-defined-out))
18:01 kronoz: samth: how do you mean? I can link you to the code...
18:01 kronoz: samth: sorry to sound like a 'gimme teh codez' kinda guy!!
18:01 samth: if the library you're requiring is a module (starts with #lang)
18:01 samth: then you should be able to require it just fine
18:02 kronoz: samth: yeah 'tis
18:02 kronoz: samth: (eval source-form ((ometa-namespace-getter))) is generating the error.
18:03 kronoz: samth: http://www.eighty-twenty.org/hgwebdir.cgi/ometa-scheme/file/5deee5dae03d/ometa.scm#l1
18:03 samth: ah, ok
18:03 kronoz: samth: just trying to make sense of this damn code... hence wanting to use drracket for some debugging capabilities!
18:04 samth: and you're just requiring that module?
18:04 kronoz: yeah
18:05 kronoz: well actually turns out no :-)
18:05 kronoz: some other code is causing it, gwargh embarrassing!
18:06 samth: does it work now?
18:07 kronoz: samth: no, but I'm able to get up a debug sesh to see what the problem is :-)
18:07 samth: ok
18:08 kronoz: samth: thx for your help! See how this goes... does debugging usually take a long time?
18:08 samth: what do you mean?
18:08 samth: it can take days
18:08 kronoz: samth: I meant as in cpu time... however I had stupidly not clicked on 'run' lol. That might help
18:08 samth: oh
18:09 kronoz: samth: haha yeah, I have spent days debugging before. the pain.
18:25 kronoz: samth: hm looks like it might be caused by (provide (all-defined-out)) - there's a nested define which is the proc it's trying to call when the error is raised, I suspect it is the first instance of the code attempting to access a non-provided proc hence it failing there...
18:35 kronoz: maybe I'll try converting the nested defines into letrecs.
18:36 (quit) stamourv: Quit: ERC Version 5.2 (IRC client for Emacs)
18:49 (quit) kronoz: Quit: ERC Version 5.3 (IRC client for Emacs)
19:12 (nick) samth -> samth_away
19:25 (quit) masm: Quit: Leaving.
20:05 (quit) carleastlund: Quit: carleastlund
20:32 (quit) jao: Ping timeout: 265 seconds
20:55 (join) jonrafkind