00:01 (join) j3parker
00:01 j3parker: hey, what is the easiest way in the racket webserver (doing a simple servlet) to add a sort of cronjob functionality? i need to periodically perform tasks and i don't want to wait until someone connects to trigger this cleanup
00:02 j3parker: timers did not work as i had expected, is what i am looking for accomplished by custodians?
00:11 j3parker: hm actually timers may work for me
00:27 (quit) zx80: Quit: Leaving
00:39 (join) masm
00:42 j3parker: slighly akward because the thread that starts the timer sticks around until it goes away but it works..
01:00 (join) jonrafkind
03:14 (quit) jonrafkind: Ping timeout: 260 seconds
04:22 (join) zx80
04:24 (part) zx80
05:02 (join) mije
05:03 mije: hi
05:04 mije: is it possible to apply a mixin on an object and not to the whole class ?
05:36 (quit) sunnyps: Ping timeout: 260 seconds
05:38 (join) sunnyps
05:45 (join) jeapostrophe
06:08 (quit) jeapostrophe: Quit: jeapostrophe
06:30 (join) jeapostrophe
07:54 (quit) jao: Ping timeout: 265 seconds
08:49 (quit) masm: Quit: Leaving.
09:15 (quit) mije: Ping timeout: 252 seconds
09:15 (quit) sunnyps: *.net *.split
09:20 samth: j3parker, you probably want a separate thread that waits a specific amount of time
09:20 samth: custodians are defintely not what you want
09:27 (join) sunnyps
09:38 (join) dzhus
10:01 (quit) ri4a: Quit: ZNC - http://znc.sourceforge.net
10:23 (join) mije
10:31 Baughn: Is there any way to have Racket use multiple OS threads, so I can actually use my CPU?
10:31 Baughn: I've got an eight-core 1.6GHz one, which means that the current setup is a bit..
10:36 jeapostrophe: Baughn: look at futures
10:36 jeapostrophe: another parallel construct, places is on its way
10:36 jeapostrophe: and "raco setup" uses multiple processes with std ipc to parallelize its work
10:37 (quit) jeapostrophe: Quit: jeapostrophe
11:02 Baughn: How portable are modules created with raco ctool --c-mods?
11:02 Baughn: I'm assuming I need to regenerate the file every compile to be safe, but is that actually the case?
11:10 (join) jeapostrophe
11:11 (quit) jeapostrophe: Client Quit
11:27 samth: Baughn, what do you mean by portable?
11:40 Baughn: Well, could I for example check a linux-generated file into a source repository and expect it to work on a windows build?
11:40 Baughn: Oh yes..
11:40 samth: i don't think so
11:40 Baughn: samth: Also, can I somehow tell raco ctool -x to output a file of a *different name* than the .3m.c thing?
11:41 Baughn: Because putting a mv into the build rule is just ugly
11:42 (join) jonrafkind
11:44 samth: i don't think so, although it ought to be easy to add, if you want to
11:46 Baughn: It's kind of obligatory if it's meant to be a serious tool
11:52 (join) masm
11:57 (quit) sunnyps: Quit: Ex-Chat
11:58 (join) carleastlund
12:01 (join) sstrickl
12:29 (quit) jonrafkind: Read error: Connection reset by peer
12:29 (join) jonrafkind
12:34 (quit) mije: Ping timeout: 252 seconds
12:36 (join) jeapostrophe
12:48 (quit) jeapostrophe: Quit: jeapostrophe
13:55 (join) anRch
13:57 (join) jeapostrophe
14:00 (quit) sstrickl: Quit: sstrickl
14:43 (quit) jeapostrophe: Quit: jeapostrophe
14:49 (quit) anRch: Quit: anRch
15:19 Baughn: Anyone know how to get Racket cooperating with gdb or valgrind?
15:19 Baughn: I get quite a lot of sigsegvs in normal operation, which makes finding the sigsegv I /didn't/ want hard
15:24 (join) hanDerPeder
15:27 blake_johnson: Baughn: I think that racket uses sigsegv in a special way, so after you start gdb, you need to type "handle SIGSEGV noprint nostop" and then run racket to get to the actual error
15:27 Baughn: blake_johnson: The problem is that the 'actual error' is a sigsegv
15:28 Baughn: Well, I'll try it with the CGC first. Could be ctool -x is messing up.
16:00 (quit) hanDerPeder: Quit: hanDerPeder
16:14 (quit) dzhus: Ping timeout: 240 seconds
16:19 tewk: Baughn: your sigsegv will show up as an abort with the sigsegv handler on the top of the stack if you follow blake_johnson's suggestion
19:00 Baughn: Turned out to be my own code that was messing up. Heh.
19:00 Baughn: Annotating C++ code for 3m is a pain, and I doubt I really need to use it. Am I correct in assuming that the CGC does /not/ scan the C heap?
19:01 Baughn: (Unless I allocate with scheme_malloc myself, which I don't)
19:08 jonrafkind: I think CGC does scan the C heap, thats the whole point of it right
19:08 jonrafkind: 3m wont scan it, it only scans stuff allocated with scheme_malloc and friends
19:08 (quit) alexsuraci: Remote host closed the connection
19:26 (join) alexsuraci
19:39 (quit) masm: Quit: Leaving.
20:03 (quit) carleastlund: Quit: carleastlund
20:40 Baughn: jonrafkind: No, the point of the CGC is that it works even if some non-pointers are treated as pointers
20:41 Baughn: And so you don't have to point out alll the pointers
20:41 (quit) jonrafkind: Ping timeout: 240 seconds
23:05 Baughn: Is nil scheme_void or scheme_null?
23:05 Baughn: I'll find out, I guess..
23:07 (join) jeapostrophe
23:20 (quit) jeapostrophe: Quit: jeapostrophe
23:32 (join) jonrafkind
23:43 (join) jeapostrophe
23:53 (quit) jeapostrophe: Quit: jeapostrophe