00:03:09 -!- syntax_tn has joined.
00:16:12 -!- kipple_ has quit (Read error: 110 (Connection timed out)).
00:47:27 -!- ihope has joined.
00:47:37 <ihope> SimonRC seems to love ##quantum.
00:54:45 * jix compiles wireshark
00:54:51 <jix> it takes too long
00:55:06 <jix> even tho i'm compiling it with make -j 2 on a dual core 2.16ghz machine :(
00:55:19 <jix> why can't people write software that compiles faster....
00:55:54 * ihope looks around to make sure nobody's looking
00:56:09 <ihope> They don't want you to compile it. (Shh, don't tell anyone...)
00:58:15 <jix> hmm warning: pointer targets in assignment differ in signedness seems to be the most common one
00:59:00 <jix> i always try to keep my code warning free... helps spotting warnings that could reduce in fatal runtime errors
00:59:06 <jix> it's done!
00:59:57 <ihope> Pointer assign--whatwhatwhat?
01:00:03 <ihope> See, Haskell actually makes sense :-P
01:00:20 <jix> i'm talking about c....
01:00:52 -!- syntax_tn has left (?).
01:03:13 <Razor-X> I'd have to say that pointers are more intuitive than monads. Not much more intuitive, but they are.
01:04:07 <ihope> Pointers are a monad :-P
01:04:11 <ihope> Actually, they're in a monad.
01:04:26 <Razor-X> Yeah, and in the end it's all using machine code :P
01:04:32 <Razor-X> I'm rating the abstraction, not what it does.
01:04:44 <Asztal> does anyone remember the news about the dynamic weapon pricing system in counter strike? I think they broke it somewhat... a glock 19 costs $-476
01:05:28 <Razor-X> Does Pascal implement pointers the exact same way?
01:22:01 * pikhq is currently down a whole 993 characters from his original effort at basm.b
01:25:37 <pikhq> This is all without having rewritten my strings setting code (fully).
01:27:38 <Razor-X> I have decided to ``cheat'' like BitTorrent.
01:28:05 <pikhq> I think we can call my original attempt "inefficient as fuck".
01:29:08 <Razor-X> Program it in ASM and it'll be ``effecient as f***'', but on the other hand it will be ``f*** annoying'' to code.
01:31:44 <pikhq> http://pikhq.nonlogic.org/basm.opt.b
01:48:49 -!- GregorR-L has quit ("Leaving").
01:55:01 -!- jix has quit ("Bitte waehlen Sie eine Beerdigungnachricht").
02:06:06 -!- Sgeo has joined.
02:25:22 -!- CXI has joined.
02:47:47 <dbc> No, I mean...is it still your plan to set up the strings, complete, in memory and then output them from there, as opposed to using the usual string-printing tactics?
02:48:24 <pikhq> IMO, makes for simpler logic (once you've got the strings set up in memory).
02:48:39 <pikhq> And makes my BFM code a bit easier to understand. ;)
02:50:13 <pikhq> Although it still leaves making the strings assloads of pain.
02:58:08 -!- CXII has joined.
02:59:41 -!- GregorR-L has joined.
03:14:23 -!- CXI has quit (Connection timed out).
03:20:22 -!- calamari has joined.
03:36:07 -!- CXII has changed nick to CXI.
04:13:26 -!- CakeProphet has quit (Read error: 113 (No route to host)).
04:18:07 -!- oerjan has joined.
04:34:22 -!- ivan` has joined.
04:56:02 <dbc> You're buying a certain amount of modularization. If you put the string-printing code inside the loops for each brainfuck command, it would complicate things in a way.
04:56:38 <dbc> On the other hand, the string-printing code would then be only about six or ten times the length of the strings to be printed.
04:57:10 <dbc> It's a tradeoff...I hate those :)
05:02:59 -!- Eidolos has quit (Read error: 104 (Connection reset by peer)).
05:04:44 -!- Eidolos has joined.
05:10:52 -!- calamari has quit ("Leaving").
05:22:45 -!- Sgeo has quit (Remote closed the connection).
05:46:24 -!- puzzlet_ has joined.
05:46:28 -!- puzzlet has quit (Read error: 110 (Connection timed out)).
05:48:21 -!- CXI has quit (Connection timed out).
07:36:59 -!- Asztal has quit (Read error: 54 (Connection reset by peer)).
07:37:15 -!- Asztal has joined.
07:56:48 -!- GregorR-L has quit ("Leaving").
07:59:59 -!- clog has quit (ended).
08:00:00 -!- clog has joined.
09:31:00 -!- Arrogant has joined.
09:57:05 -!- oerjan has quit ("Later").
10:39:42 -!- Arrogant has quit ("Leaving").
12:53:31 -!- jix has joined.
13:25:36 <pikhq> dbc: IMO, the code simplification is worth it. . .
13:25:52 <pikhq> Especially when one considers that the string output code is all in BFM macros, anyways.
14:47:33 -!- ivan` has quit (Read error: 104 (Connection reset by peer)).
14:47:41 -!- ivan`_ has joined.
14:47:49 -!- ivan`_ has changed nick to ivan`.
14:49:26 -!- ivan` has quit (Connection reset by peer).
14:49:30 -!- ivan`_ has joined.
14:49:44 -!- ivan`_ has changed nick to ivan`.
15:17:57 <SimonRC> pikhq: I feel that "+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++" might be possible to optimise futher.
15:18:41 <SimonRC> pikhq: wouldn't the text-to-bf program help too?
15:29:12 -!- tgwizard has joined.
15:39:38 <EgoBot> 34 ++++++++++++[>++++++++>>><<<<-]>+. [26]
15:39:53 <GregorR> And that, my friends, is why the text-to-bf program doesn't help :P
16:21:07 -!- oerjan has joined.
16:34:00 -!- kipple_ has joined.
16:53:29 -!- GregorR-W has joined.
17:12:46 -!- ihope_ has joined.
17:24:25 <pikhq> SimonRC: Um. . . That's what I'm currently optimizing.
17:25:03 <pikhq> oerjan: Your earlier suggestion came after I had implemented exactly that (re: [.[-]]).
17:27:02 -!- ihope has quit (Read error: 110 (Connection timed out)).
17:27:04 <oerjan> fine. as for those 105 plusses, take a look at Brainfuck constants on the wiki.
17:27:21 <pikhq> oerjan: I've been rewriting all of that.
17:27:51 <pikhq> Assloads of pain, becuase that's 158 cells to set. . .
17:27:59 <oerjan> (essentially 105 = 2/5 + 3)
17:28:34 * pikhq is making the loops write to multiple points in the array whenever it adds to efficiency
17:32:01 <pikhq> This would be infinitely easier if stdcons.bfm existed.
17:32:19 <pikhq> Especially if stdcons.bfm had wrapping cells.
17:34:58 <oerjan> i have the list in brainfuck so the question is how to convert it to BFM
17:37:14 <pikhq> Change your script to generate BFM instead of Brainfuck. :p
17:37:35 <pikhq> (if you hand it to me, I'd be willing to make the necessary changes)
17:37:38 <oerjan> i am not generating it, i am extracting it from the wiki list
17:38:02 <pikhq> I could probably change brainfucktobfm.tcl for the job. . .
17:39:08 <oerjan> or i'll whip up something in haskell which works for this special case
17:39:55 <pikhq> Does it pick out the shortest wrapping variants, or just the shortest nonwrapping?
17:40:59 <pikhq> Ideally, it'd pick out the shortest 2-cell wrapping variants. . .
17:41:13 <pikhq> I *think* those are usually the shortest. . . Right?
17:41:31 <oerjan> except when the 1-cell are shortest
17:42:09 <GregorR-W> If 1-cell is the shortest, then 2-cell is just 1-cell except the second cell isn't used :P
17:42:17 <pikhq> GregorR-W: Indeed.
17:45:41 <oerjan> question: for the 1-cell case, do you still want the macro to take a second argument for uniformity?
17:49:05 -!- GregorR-L has joined.
17:49:52 -!- GregorR-W has quit ("Chatzilla 0.9.75 [Firefox 1.5.0.6/0000000000]").
17:52:29 <oerjan> i am blithely assuming set works with negative numbers
18:24:58 -!- bsmntbom1dood has joined.
18:36:33 -!- bsmntbombdood has quit (Connection timed out).
18:44:58 <pikhq> Um. . . Not yet. Will soon.
18:45:22 <oerjan> i will assume it anyhow.
18:48:48 <oerjan> aha. is there a new tarball?
18:50:22 <pikhq> New tarball is now up.
18:50:36 <pikhq> With somewhat large amounts of bugs fixed. :)
18:52:33 <oerjan> except you need to fix the negative numbers: subtract -num, not num
18:53:41 <pikhq> I fixed it in set.
18:54:09 <pikhq> (unless you argue that calling "set foo -1" should execute "subtract foo -1", which should do "+".)
18:54:43 <oerjan> um, no, you definitely got it the wrong way around.
18:55:08 <pikhq> This outputs [-], then executes "subtract foo 1".
18:55:16 <pikhq> subtract foo 1 outputs "-".
18:56:08 <oerjan> are you sure? because the code i downloaded looks like it would do subtract foo -1
19:10:54 <oerjan> very well: http://home.nvg.org/~oerjan/stdcons.bfm
19:26:09 <pikhq> Grr; you're right.
19:27:55 <pikhq> Fixed in current tarball
19:29:08 <pikhq> And, BTW, thank you *very* much.
19:29:34 <oerjan> just one more improvement coming up
19:32:15 <oerjan> i changed those macros that just add a constant to the result of another macro so that they call it
19:33:14 -!- wooby has joined.
19:33:17 -!- wooby has quit (Remote closed the connection).
19:33:47 * pikhq doesn't see what you mean. . .
19:35:21 <oerjan> for example, 26 is calculated by first calculating 25 and then adding 1. the original version converted the whole code, the new one calls cons25 as a subroutine.
19:36:54 <pikhq> I assume you have no issues with it being under the GPL and part of BFM?
19:36:57 <oerjan> all this with some particularly dirty haskell.
19:37:05 <pikhq> Part of the current tarball now. . .
19:39:14 <oerjan> perhaps add a note that it was converted from bf code on the esolang wiki
20:53:41 -!- CakeProphet has joined.
21:43:24 -!- Sgeo has joined.
22:03:05 -!- oerjan has quit ("Leaving").
22:24:28 -!- bsmntbom1dood has changed nick to bsmntbombdood.
22:32:32 -!- bsmntbombdood has left (?).
22:32:34 -!- bsmntbombdood has joined.
22:34:08 -!- bsmntbombdood has quit ("leaving").
23:39:34 <pikhq> 41 cells to go. . .
23:46:09 <pikhq> Finished, and with only one (minor) bug.
23:48:29 <pikhq> A difference of 10476 characters from my initial attempt (7288 characters now).
23:49:02 <pikhq> http://pikhq.nonlogic.org/basm.opt.b
23:51:35 <pikhq> Not quite as small was wib.b (6217), but nearing it. . .
23:51:45 -!- Arrogant has joined.
23:52:25 <pikhq> And basm.opt.b *still* has a lot of obvious things to optimize away.
23:52:55 -!- Arrogant_ has joined.
23:52:59 -!- Arrogant_ has quit (Read error: 104 (Connection reset by peer)).
23:57:15 <pikhq> 7256 characters now. . .
23:57:53 * pikhq gently removes obvious things to cut
23:58:37 -!- tgwizard has quit (Remote closed the connection).