00:01 gcr`: For the racket FFI, what type should I use for a size_t?
00:10 (quit) kofno: Remote host closed the connection
00:16 (join) sw2wolf
00:16 jonrafkind: good question. i dont see anything obvious in the source, you should ask on the list
00:19 (join) RacketCommitBot
00:19 RacketCommitBot: [racket] plt pushed 1 new commit to master: http://git.io/i666sQ
00:19 RacketCommitBot: racket/master 5a2aa22 Neil Toronto: Initial commit for `typed/plot' module...
00:19 (part) RacketCommitBot
00:56 (join) mye
01:04 (quit) zyoung: Remote host closed the connection
01:11 (quit) jackhammer2022: Quit: Textual IRC Client: http://www.textualapp.com/
01:12 (join) Nisstyre
01:19 (quit) jao: Ping timeout: 245 seconds
01:23 (join) jeapostrophe
01:23 (quit) jeapostrophe: Changing host
01:23 (join) jeapostrophe
01:33 (join) Rajah
01:42 (quit) dnolen: Remote host closed the connection
01:46 (join) eli
01:46 (quit) eli: Changing host
01:46 (join) eli
01:55 (quit) jaimef: Ping timeout: 276 seconds
02:22 (quit) ambrosebs: Ping timeout: 246 seconds
02:47 gcr`: So I'm working on some FFmpeg bindings. So far I have it to the point where it can (hackily!) decode video. I have the ARGB video data here as a bytestring; this can do about 66 frames per second on my computer.
02:48 gcr`: Now the tricky part is sending those bytes to a bitmap%. On my system, a call to (send bitmap-dc set-argb-pixels) takes about 80ms for a 400x300 image, which limits me to ~10fps
02:49 gcr`: I know that without hardware assistance, there's going to be limits, but is there any way to make set-argb-pixels faster?
02:59 (quit) jeapostrophe: Ping timeout: 250 seconds
03:01 sw2wolf: After installing gerser and `M-x run-geiser` can open a racket REPL. However, in a racket source file it seems any gerser-eval* function doesnot work ?
03:04 (quit) Rajah: Remote host closed the connection
03:05 (join) Rajah
03:07 asvil: sw2wolf: M-x geiser-mode in buffer?
03:08 sw2wolf: i have ("\\.rkt$" . geiser-mode) in auto-mode-alist
03:10 sw2wolf: asvil: strange! after running `M-x geiser-mode` manually, it works now! thx
03:18 (join) ambrosebs
03:21 (quit) sw2wolf: Remote host closed the connection
03:22 asvil: sw2wolf: probably file was opened before auto-mode-alist, or file name is not matched
03:23 gcr`: hm. After tweaking argb-bytes to remove some silly bottlenecks, I've got my set-argb-bytes call down to ~20msec, which is approx. a 4x speedup.
03:23 gcr`: i'll submit a benchmark + patch tomorrow when i wake up
03:24 gcr`: Ain't realtime yet but it's darn close. <3 FFmpeg+racket
03:58 (join) PCChris
04:16 (join) Pylipala
04:20 (part) Pylipala
04:24 (quit) Nisstyre: Read error: Connection reset by peer
04:27 (quit) Rajah: Remote host closed the connection
04:28 (join) Rajah
04:28 (quit) PCChris: Quit: Leaving
04:30 (quit) jonrafkind: Ping timeout: 264 seconds
04:49 (quit) Rajah: Remote host closed the connection
04:57 (join) cdidd
05:01 (join) mithos28
05:01 (quit) ambrosebs: Ping timeout: 264 seconds
05:05 (quit) mithos28: Client Quit
05:21 (join) ambrosebs
05:26 (join) noctuorare
05:40 noctuorare: Excuse me, I tried #!r6rs (import (rnrs base(6))(rnrs io simple (6)))(display 1) (display (read)) in racket 5.2.1 and found it won't display 1 before waiting me to input something. Is this intended to be so? Thank you.
05:40 (quit) ambrosebs: Ping timeout: 255 seconds
05:45 (join) bitonic
05:51 (join) masm
06:03 (join) sw2wolf
06:05 (quit) mye: Ping timeout: 245 seconds
06:22 (quit) zuurr: Quit: computer sleeping
06:25 (join) ambrosebs
06:26 (quit) noctuorare: Quit: KVIrc 4.1.3 Equilibrium http://www.kvirc.net/
06:47 (join) mye
06:58 (quit) jrslepak: Quit: What happened to Systems A through E?
07:21 (quit) sw2wolf: Quit: ERC Version 5.3 (IRC client for Emacs)
07:25 (join) nathanpc
07:33 (quit) ambrosebs: Ping timeout: 248 seconds
08:02 (join) kofno
08:05 (join) mye_
08:07 (quit) mye: Ping timeout: 245 seconds
08:07 (nick) mye_ -> mye
08:14 (join) mye
08:14 (join) MayDaniel
08:15 (quit) Nisstyre-laptop: Ping timeout: 250 seconds
08:18 (join) soegaard
08:24 (join) jao
08:24 (quit) jao: Changing host
08:24 (join) jao
08:26 (join) jrslepak
09:07 mye: meh, I don't understand how a lexer-src-pos from parser-tools/lex puts position information in tokens it produces.
09:08 (quit) jrslepak: Read error: No route to host
09:08 mye: It seems to just use the info from the matched lex-sre regexp, but that breaks when a sub-lexer is used to lex e.g. string literals
09:08 (join) jrslepak
09:09 mye: I don't see a way to influence what it puts in the token for the position info
09:25 soegaard: mye: Here is an example of how to keep positions using a lexer-src-pos.
09:25 soegaard: https://github.com/soegaard/bracket/blob/master/bracket/lang/parser.rkt
09:26 soegaard: The macro on line 204 is used to transfer the positions.
09:26 soegaard: It might not be the best example though.
09:29 (join) jeapostrophe
09:29 (quit) jeapostrophe: Changing host
09:29 (join) jeapostrophe
09:35 (quit) jrslepak: Quit: What happened to Systems A through E?
09:37 mye: soegaard: I have adapted a similar macro from the example for a scheme parser. But my problem is that the $1-start-pos and $1-end-pos contain the wrong information for the tokens where the value comes from a lexer called from the main lexer (I have sub-lexers that return a list of chars that can recognize unicode escapes like \u00ee)
09:39 mye: I tried to make them into src-pos lexers too but then they return a list with nested position-token instances
09:39 soegaard: Is the sub-lexer a lexer-src-pos too?
09:39 mye: soegaard: not yet
09:40 mye: this is what I get for a value with a lexer-src-pos:
09:40 mye: (position-token (cons #\f (position-token (cons #\o (position-token (cons #\o (position-token '() (position 10 2 4) ...
09:40 mye: I can imagine a function that extract the corrent positions from that
09:40 mye: but then how can I put them in the token?
09:41 mye: I must use e.g. token-LITERAL and that accepts only one argument
09:42 soegaard: hmm
09:43 mye: soegaard: It's still opaque to me how the lexer macro actually puts the position? struct into the position-tokens it generates.
09:44 soegaard: I am reading https://github.com/plt/racket/blob/master/collects/parser-tools/private-lex/token.rkt to see how.
09:44 rudybot: http://tinyurl.com/dynglru
09:45 mye: I had a look at the (make-define-tokens macro, but it didn't enlighten me :-)
09:46 soegaard: Perhaps (make-position-token is too be used?
09:50 mye: All I could figure out it just uses make-token, but this is not exported by lex.rkt, so to make a position-token I would have to create the (token 'LITERAL "foo") manually, but the constructor seems to be missing
09:51 soegaard: Shouldn't (token-LITERAL "foo") work?
09:54 soegaard: Wait a minute. Did you intended for the sub-lexer to return a single token LITERAL with the value "foo"?
09:54 mye: soegaard: that's what I use now: match #\" on the left side then on the right: (token-LITERAL (list->string (string-lexer input-port)). The problem is this makes a position-token with the position info from _only_ the #\"
09:54 soegaard: If so I think it is the sublexer, that needs a change.
09:57 mye: I'll try to rewrite the sub lexers to produce a token with token-LITERAL and then wrap this in a position-token with the corrent positions then return this directly in the main lexer
09:57 (quit) bjz: Quit: Leaving...
09:57 mye: soegaard: thanks, really grateful for your assistance so far
09:58 soegaard: If it doesn't work, then make a minimal example and post it to the mailing list.
10:18 (join) zyoung
10:34 (quit) kofno: Remote host closed the connection
10:38 (quit) vu3rdd: Remote host closed the connection
10:46 (join) anRch
10:55 (join) ijp
10:58 (join) h11r
11:21 (join) zuurr
11:25 (quit) anRch: Quit: anRch
11:35 (join) kofno
11:37 (join) spiderweb
11:50 (quit) jao: Ping timeout: 245 seconds
12:10 (quit) jeapostrophe: Ping timeout: 265 seconds
12:24 (quit) h11r: Ping timeout: 264 seconds
12:24 (join) josdeha_rt
12:24 (quit) josdeha_rt: Remote host closed the connection
12:38 (join) dnolen
12:40 gcr`: Aw, man
12:40 gcr`: So I'm speeding up set-argb-pixels
12:40 gcr`: One of my ideas was to allow the premultiplied? argument to be 'assume-premultiplied, in which case it could simply copy each row with bytes-copy!
12:41 gcr`: However, on little endian systems, the destination byte vector is really in BGRA format, so I have to do some byte swizzling or else it looks wrong
12:41 gcr`: Which is a REAL shame since this would have been a 10x speedup for set-argb-pixels on my workload. :/
12:42 gcr`: Is there a quick way to switch the endianness of byte strings assumed to contain ints?
12:44 (quit) kofno: Read error: Connection reset by peer
12:44 (join) bitonic`
12:47 (quit) bitonic: Ping timeout: 248 seconds
12:56 (join) mithos28
13:08 (nick) bitonic` -> bitonic
13:14 (join) netrino
13:23 (quit) zyoung: Remote host closed the connection
13:23 (join) mizu_no_oto
13:25 mye: why do I keep thinking that (list (values 1 2 3)) should evaluate to '(1 2 3) !?! (Why doesn't it ¿ )
13:26 (quit) mizu_no_oto: Client Quit
13:27 mithos28: Because (list <>) expects the expression <> to return 1 value not three
13:27 (join) h11r
13:29 mye: mithos28: yes, but when I look at (list 1 2 3) it looks like more than one value :-P What is the reader (?) doing with the 1 2 3 to turn it to "one" value?
13:30 mithos28: its not, that is (list <> <> <>) and each <> expects the expression to have 1 value
13:30 offby1 distrusts that "list" expression ... shifty eyes
13:30 offby1: <--<
13:30 (join) mizu_no_oto
13:31 offby1: >-->
13:31 (join) jonrafkind
13:33 (quit) ijp: Ping timeout: 264 seconds
13:36 mye: oh, right, I had too much coffee today (not used to it). I still am sometimes in the situation where I would like to give the multiple return values as separate function arguments.
13:37 mye: (hey, I read on the ML they exists for symmetry with n-ary functions in the first place)
13:37 mithos28: call-with-values
13:37 mithos28: rudybot: init racket
13:37 rudybot: mithos28: your sandbox is ready
13:38 mithos28: rudybot: (call-with-values (lambda () (values 1 2 3)) list)
13:38 rudybot: mithos28: ; Value: (1 2 3)
13:39 mye: mithos28: I really need to intern this symbol, thx
13:50 (quit) mizu_no_oto: Quit: Computer has gone to sleep.
13:56 (join) Nisstyre
14:14 (part) h11r
14:31 (join) mizu_no_oto
14:47 (join) gcr``
14:47 (nick) gcr`` -> gcr
14:49 (quit) gcr`: Ping timeout: 264 seconds
14:50 (join) ijp
15:00 (join) PCChris
15:15 (join) bitonic`
15:19 (join) Blkt
15:19 (join) jao
15:20 (quit) jao: Changing host
15:20 (join) jao
15:21 (join) RacketCommitBot
15:21 RacketCommitBot: [racket] plt pushed 5 new commits to master: http://git.io/ci2Rdw
15:21 RacketCommitBot: racket/master d3e4fbe Matthew Flatt: make scheme_check_threads() run at least on thread quantum...
15:21 RacketCommitBot: racket/master fb7cbcf Matthew Flatt: date struct: improve docs...
15:21 RacketCommitBot: racket/master 59c693d Matthew Flatt: fix docs on scheme_malloc_fail_ok...
15:21 (part) RacketCommitBot
15:24 (join) zyoung
15:37 (quit) mizu_no_oto: Quit: Computer has gone to sleep.
15:43 (quit) mye: Quit: mye
15:46 (nick) bitonic` -> bitonic
15:53 (quit) nathanpc: Quit: Computer has gone to sleep.
15:53 (join) nathanpc
15:59 (join) dyoo
16:02 (join) bitonic
16:04 (join) bjz
16:04 (join) bjz_
16:05 (quit) dyoo: Quit: dyoo
16:08 (quit) bjz: Ping timeout: 246 seconds
16:23 (join) nejucomo
16:26 (quit) Nisstyre: Quit: Leaving
17:12 (quit) dnolen: Ping timeout: 246 seconds
17:29 (quit) zyoung: Remote host closed the connection
17:30 (join) anRch
17:42 (join) offby1`
17:43 (join) kofno
17:44 (quit) offby1`: Client Quit
17:47 (quit) mithos28: Quit: mithos28
18:02 (join) jeapostrophe
18:02 (quit) jeapostrophe: Changing host
18:02 (join) jeapostrophe
18:04 (join) RacketCommitBot
18:04 RacketCommitBot: [racket] plt pushed 1 new commit to master: http://git.io/MYsOPA
18:04 RacketCommitBot: racket/master 22f9c85 Matthew Flatt: fix optimizer bug, again...
18:04 (part) RacketCommitBot
18:12 (quit) bjz_: Quit: Leaving...
18:24 (quit) jeapostrophe: Ping timeout: 264 seconds
18:29 (quit) anRch: Quit: anRch
18:36 (quit) kofno: Remote host closed the connection
18:39 (join) tilde`
18:46 (join) gcr`
18:46 (join) kofno
18:48 (join) RacketCommitBot
18:48 RacketCommitBot: [racket] plt pushed 5 new commits to master: http://git.io/VxXMew
18:48 RacketCommitBot: racket/master b6a4a48 Kevin Tew: [Distributed Places] uncomment tests
18:48 RacketCommitBot: racket/master 0876466 Kevin Tew: Fixed broadcast and reduce for non powers of 2. Added alltoall and alltoallv
18:48 RacketCommitBot: racket/master e30fdf0 Kevin Tew: Fixed broadcast and reduce from nodes other that 0. Added gather and allgather
18:48 (part) RacketCommitBot
18:50 (join) bjz
18:57 (quit) netrino: Quit: Ave!
18:59 (join) neilv
19:03 (quit) tilde`: Quit: kthxbai
19:04 (quit) MayDaniel: Read error: Connection reset by peer
19:07 (quit) kofno: Remote host closed the connection
19:13 (join) kofno
19:22 (quit) soegaard: Quit: soegaard
19:23 (quit) Blkt: Read error: Connection reset by peer
20:01 (join) masm1
20:02 (quit) kofno: Remote host closed the connection
20:02 (join) myx
20:02 (quit) masm: Ping timeout: 260 seconds
20:07 (quit) jao: Ping timeout: 272 seconds
20:08 (join) kofno_
20:14 (quit) bitonic: Ping timeout: 255 seconds
20:29 (quit) ChibaPet: Quit: leaving
20:34 (join) zyoung
20:39 (join) mizu_no_oto
20:48 (join) sw2wolf
20:56 (join) ChibaPet
20:56 (quit) ChibaPet: Changing host
20:56 (join) ChibaPet
20:57 (quit) nejucomo: Ping timeout: 276 seconds
21:07 (join) nejucomo
21:29 (join) francisl
21:35 (quit) nejucomo: Ping timeout: 276 seconds
21:37 (join) jeapostrophe
21:37 (quit) jonrafkind: Ping timeout: 265 seconds
21:38 (part) gcr`: "ERC Version 5.3 (IRC client for Emacs)"
21:46 (join) PCChris_
21:47 (join) nejucomo
21:47 (quit) sizz: Quit: No Ping reply in 180 seconds.
21:48 (join) sizz
21:50 (quit) PCChris: Ping timeout: 255 seconds
21:55 (quit) jeapostrophe: Ping timeout: 260 seconds
21:59 (quit) nejucomo: Ping timeout: 276 seconds
22:00 (quit) zyoung: Remote host closed the connection
22:00 (join) jonrafkind
22:01 (quit) dca: Ping timeout: 248 seconds
22:11 (quit) masm1: Quit: Leaving.
22:14 (quit) PCChris_: Read error: Connection reset by peer
22:17 (quit) nathanpc: Quit: Computer has gone to sleep.
22:39 (quit) mizu_no_oto: Quit: ["Textual IRC Client: www.textualapp.com"]
23:01 (quit) Shviller: Ping timeout: 255 seconds
23:02 (join) Shviller
23:03 (join) francisl_
23:04 (quit) francisl_: Client Quit