00:03:14 -!- timotiis has quit (Read error: 110 (Connection timed out)). 00:24:25 -!- RedDak has quit (Remote closed the connection). 00:29:58 happy 1 july 00:33:47 -!- tusho has left (?). 00:33:50 -!- tusho has joined. 00:33:50 -!- tusho has left (?). 00:34:01 -!- tusho has joined. 00:34:12 -!- tusho has left (?). 00:34:16 -!- tusho has joined. 01:13:16 its not july first yet here! 01:13:58 yes it is 01:14:02 now goodbye :) 01:14:07 -!- tusho has quit. 01:14:08 im in florida 01:59:49 -!- Judofyr has quit. 03:48:10 http://www.amazon.com/Million-Random-Digits-Normal-Deviates/dp/0833030477/ 03:48:16 A bargain! 05:21:20 -!- Corun has quit ("This computer has gone to sleep"). 06:27:08 lol 06:59:20 -!- cc_toide has joined. 07:17:11 -!- cctoide has quit (Read error: 110 (Connection timed out)). 07:22:08 -!- Slereah has quit (Read error: 110 (Connection timed out)). 07:35:15 augur, happy 1 July too 07:35:17 no idea why 07:35:25 lolk 07:35:35 but it is 1 July in Sweden 07:35:36 i leave or europe at 11 07:35:37 in oh 07:35:38 8 hours 07:35:46 hm why? 07:37:23 fun and profit 07:43:46 AnMaster: see, good thing that I don't update the Mycology comparison, what with all these bugs still in your interpreter ;-) 07:44:04 Deewiant, none of them affected current mycology 07:44:30 Deewiant, anyway last was a pre-release so heh 07:44:33 ah, but I'd do some extra bug hunting just to spite you, and then add anything I find to Mycology ;-) 07:44:58 before next release I plan to create a test program for TURT on my own 07:45:02 will do that later today 07:45:07 going swimming shortly 07:47:37 Deewiant, well I will be happy to accept any bug reports 07:59:59 -!- clog has quit (ended). 08:00:00 -!- clog has joined. 08:00:40 -!- augur has quit (Read error: 110 (Connection timed out)). 08:10:11 -!- oklofok has joined. 08:14:31 -!- oklopol has quit (Read error: 110 (Connection timed out)). 08:38:01 -!- lament has quit (clarke.freenode.net irc.freenode.net). 08:38:01 -!- AAA_AAA has quit (clarke.freenode.net irc.freenode.net). 08:38:02 -!- bsmntbombdood has quit (clarke.freenode.net irc.freenode.net). 08:38:03 -!- atsampson has quit (clarke.freenode.net irc.freenode.net). 08:38:04 -!- mtve has quit (clarke.freenode.net irc.freenode.net). 08:38:05 -!- cherez has quit (clarke.freenode.net irc.freenode.net). 08:38:05 -!- GregorR has quit (clarke.freenode.net irc.freenode.net). 08:38:05 -!- AnMaster has quit (clarke.freenode.net irc.freenode.net). 08:38:05 -!- Ilari has quit (clarke.freenode.net irc.freenode.net). 08:38:05 -!- cc_toide has quit (clarke.freenode.net irc.freenode.net). 08:38:06 -!- sebbu has quit (clarke.freenode.net irc.freenode.net). 08:38:07 -!- cmeme has quit (clarke.freenode.net irc.freenode.net). 08:38:08 -!- Polar has quit (clarke.freenode.net irc.freenode.net). 08:38:08 -!- dbc has quit (clarke.freenode.net irc.freenode.net). 08:38:08 -!- SimonRC has quit (clarke.freenode.net irc.freenode.net). 08:38:09 -!- fizzie has quit (clarke.freenode.net irc.freenode.net). 08:38:09 -!- shachaf has quit (clarke.freenode.net irc.freenode.net). 08:38:09 -!- lifthras1ir has quit (clarke.freenode.net irc.freenode.net). 08:38:10 -!- Deewiant has quit (clarke.freenode.net irc.freenode.net). 08:38:11 -!- jamesstanley has quit (clarke.freenode.net irc.freenode.net). 08:38:11 -!- Dewi has quit (clarke.freenode.net irc.freenode.net). 08:38:12 -!- puzzlet has quit (clarke.freenode.net irc.freenode.net). 08:39:08 -!- cc_toide has joined. 08:39:08 -!- AAA_AAA has joined. 08:39:08 -!- lament has joined. 08:39:08 -!- cmeme has joined. 08:39:08 -!- bsmntbombdood has joined. 08:39:08 -!- cherez has joined. 08:39:08 -!- dbc has joined. 08:39:08 -!- GregorR has joined. 08:39:08 -!- sebbu has joined. 08:39:08 -!- Dewi has joined. 08:39:08 -!- Polar has joined. 08:39:08 -!- AnMaster has joined. 08:39:08 -!- mtve has joined. 08:39:08 -!- shachaf has joined. 08:39:08 -!- jamesstanley has joined. 08:39:08 -!- atsampson has joined. 08:39:08 -!- Deewiant has joined. 08:39:08 -!- puzzlet has joined. 08:39:08 -!- fizzie has joined. 08:39:08 -!- lifthras1ir has joined. 08:39:08 -!- Ilari has joined. 08:39:08 -!- SimonRC has joined. 08:57:14 -!- Slereah_ has joined. 09:01:57 ESOTERIA 09:52:07 -!- Slereah_ has quit (Read error: 110 (Connection timed out)). 10:22:52 -!- Judofyr has joined. 10:50:03 Deewiant, there is some kind of error in ccbi TURT as well as cfunge TURT 10:50:05 different errors 10:50:24 but putting down pen and going forward by 10 pixels shouldn't result in a viewbox like this: 10:50:31 viewBox="-163839.0999 -0.0010 327679.0998 0.0020" 10:50:34 which it does for ccbi 10:50:39 "TRUT"4( 1P 5F D 0P I @ 10:50:41 was the program 11:00:35 Deewiant, also I get a closed path not an open one it seems 11:00:58 closed and filled 11:32:01 Deewiant, further: the instruction to clear the paper with some color doesn't work correctly 11:39:57 -!- Slereah_ has joined. 11:40:35 GLORIA ESOTERIA 12:02:42 Slereah_, ? 12:07:09 -!- cc_toide has changed nick to cctoide. 12:13:29 Hello people 12:27:06 -!- ais523 has joined. 12:27:10 hrrm 12:27:21 well I found quite a few bugs in TURT, both CCBI's and mine 12:27:25 mine is more or less fixed now 12:27:35 still the margins in the generated file are quite weird 12:27:39 but working on that now 12:28:01 what to learn from this: fixed point sucks 12:28:17 and hi ais523 12:28:22 hi AnMaster 12:30:13 Deewiant, why do you use fixed point in TURT btw? 12:30:42 (p.d.p.x < 0) ? "-" : "", getInt(p.d.p.x), getDec(p.d.p.x) <-- seems quite messy to me 12:30:50 just to print a fixed point number 12:31:18 * AnMaster added code to convert to a double 12:44:27 floating-point suffers from rounding errors when large and small numbers are combined 12:44:59 I've had that problem before; I was trying to cause something to change every second, but the time was in epoch-seconds and stored in a floating-point number 12:45:07 it worked fine when I changed it to fixed-point 12:54:36 wtf 12:54:44 I think Deewiant confused turn right and turn left 12:54:51 I can't explain it in any other way 12:54:54 in TURT? 12:54:58 yes 12:55:04 I'm pretty sure Deewiant wouldn't confuse [ and ] 12:55:10 well it is in TURT 12:55:14 not normal funge 12:55:21 void turnLeft() { turt.heading += toRad(ip.stack.pop); turt.normalize(); } 12:55:21 void turnRight() { turt.heading -= toRad(ip.stack.pop); turt.normalize(); } 12:55:33 is heading clockwise or anticlockwise? 12:55:38 ais523, trying to figure out that 12:56:03 ais523, problem it does it totally wrong anyway 12:57:06 0 is east 12:57:12 is all the specs say 12:57:21 when people set 0=east, then they tend to use anticlockwise angles 12:57:30 that's mathematician angle measurement 12:57:41 0=east, pi/2=north, pi=west, 3*pi/2=south 12:57:41 ais523, so 90 should be straight up? 12:57:47 AnMaster: radians 12:57:48 ais523, it is in degrees 12:57:51 according to specs 12:57:54 well, there's a toRad in that code 12:58:01 so Deewiant's storing it in radians internally 12:58:06 but 90 would be straight up in degrees, yes 12:58:10 wtf 12:58:16 ais523, the code treats that as downwards 12:58:37 I suppose you have to look at the original code for TURT 12:58:41 anyway 0 in his code is equal to going diagonally down 12:58:47 that's wrong 12:58:48 ais523, there is nothing but the specs 12:58:56 ais523, yes 12:58:58 http://catseye.tc/projects/funge98/library/TURT.html 12:59:04 in that case, how did J^4 do the turt-quine? 12:59:11 "H 'Set Heading' (angle in degrees, relative to 0deg, east)" 12:59:28 ais523, well he implemented TURT in *his* way in *his own* interpreter 12:59:39 ah 12:59:46 problem is that no implementations agree on this simple test: 12:59:54 "TRUT"4( 29*N 0H 1P 55*F 9a*H 5F 9a*L 5F 0P 5B 0a*R aB 1P I @ 13:00:10 9a*L should turn 90 degrees to the left 13:00:14 * ais523 wonders what a non-esoprogrammer would think of your definition of a "simple test" 13:00:45 ais523, if 90 is straight down then it should draw as the ascii art: 13:00:46 ---. 13:00:48 | 13:00:53 ._ 13:01:06 and a dot a bit above the vertical line 13:01:29 except no interpreter agrees about this 13:02:21 ais523, the last direction change there, is 90 degrees to the left right? 13:02:41 I don't know TURT 13:02:47 what does H do? 13:02:49 and B? 13:02:54 H 'Set Heading' (angle in degrees, relative to 0deg, east) 13:02:57 B 'Back' (distance in pixels) 13:03:01 F 'Forward' (distance in pixels) 13:03:08 I guessed F 13:03:09 L 'Turn Left' (angle in degrees) 13:03:09 R 'Turn Right' (angle in degrees) 13:03:22 P 'Pen Position' (0 = up, 1 = down) 13:03:36 N? 13:03:40 N 'Clear Paper with Colour' (24-bit RGB) 13:03:43 * I 'Print current Drawing' (if possible) 13:04:03 anyway N is currently broken in both cfunge and ccbi 13:04:15 it clears but doesn't set bg color at all 13:04:54 0a*R is a NOP, surely? 13:05:03 should it say 9a*R? 13:05:07 AnMaster: I think SVG is the reason for fixed point, I might misremember though 13:05:08 yes it should 13:05:11 in any case, accuracy 13:05:19 but it breaks before that 13:05:39 well I think ccbi is quite broken on "TRUT"4( 29*N 0H 1P 55*F 9a*H 5F 9a*L 5F 0P 5B 9a*R aB 1P I @ 13:05:53 I think 90's more likely to be straight up 13:06:01 cfunge is slightly broken but not as broken 13:06:01 but then, Befunge uses mathematical notation with up and down exchanges 13:06:07 s/s$/d$/ 13:06:11 so maybe TURT does the same? 13:06:20 Deewiant, should 90 degrees in TURT be up or down? 13:06:31 I'm happy to use either but I need to know 13:08:27 well now cfunge does the right if 90 is down 13:08:35 still my margins are all messed up 13:10:35 Deewiant, it is clear you haven't tested your TURT ;P 13:10:57 as the generated paths are filled and closed and so on 13:11:01 we need a Turt version of Acid2 13:11:20 ais523, haha well mine test some stuff 13:11:36 as for your statement about it working differently on every interpreter 13:11:48 there's a bit of INTERCAL-72 that the original spec didn't define 13:12:09 and it turned out that J-INTERCAL, C-INTERCAL, and CLC-INTERCAL (the three INTERCAL interps I could get hold of) all implement it differently 13:15:38 hm 13:19:58 so it's sort of a handprint for INTERCAL interps 13:20:16 I've publically stated that I'd prefer it if all future INTERCAL interps do something different yet again upon encountering it 13:20:39 heh? 13:20:41 what is this thing 13:20:56 basically when you STASH a variable, IGNORE it, then RETRIEVE that variable 13:21:02 it's the interaction of read-only-ness and scoping 13:24:05 -!- oklofok has changed nick to oklopol. 13:25:30 how can you figure out how to get a large number in befunge? 13:25:35 I need 0xFF0000 13:25:44 how do you write that out in Befunge... 13:25:53 generally speaking, I factorise and multiply 13:25:58 but there may be better ways 13:26:12 $ factor $((0xFF0000 )) 13:26:12 16711680: 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 3 5 17 13:26:12 aha 13:26:26 well 17 is too large so... 13:26:33 add that up 13:26:34 also 2s can be bunched into 8s 13:26:37 yes 13:27:00 2f+8*8*8*8*8*6*5* 13:29:04 well well 13:29:24 Deewiant, your colors are broken 13:29:37 0xff0000 gets output as "#0000ff" 13:31:51 blergh 13:31:56 this is even more broken 13:32:18 I support that blergh 13:32:19 sorry 13:41:57 ok that is partly fixed now 13:42:19 need to close a path correctly 13:51:57 I found an old marble of mine :D 13:53:20 you lost them? 13:54:34 I didn't even know I still had one. 13:55:38 that's because you don't keep them in your head, remember 13:57:07 OR DO I 13:57:51 "Remember Adolf Hitler, the most famous Black Magick wizard in modern history?" 13:57:56 I'm not sure I do D: 14:00:20 -!- Corun has joined. 14:00:44 Deewiant, ais523: well cfunge is more conforming to TURT specs than ccbi now :P 14:00:51 still not perfect 14:00:54 heh 14:01:59 1) margins totally messed up, 2) it sometimes misses to add path segments when outputting (but not as much as ccbi does) 3) adding dots after lines are even more random operation it seems 14:02:05 is an even* 14:07:25 well my test program works in cfunge now 14:07:45 not saying that other programs will work 14:08:44 ais523, you may want to pull 14:08:56 did you update only TURT? 14:09:03 yes 14:09:18 well TODO and CMakeList.txt too 14:09:22 as I added a man page 14:16:43 -!- Corun has quit ("This computer has gone to sleep"). 14:18:00 -!- pikhq has joined. 14:32:35 -!- atsampson has quit ("back in a minute"). 14:36:03 -!- atsampson has joined. 14:51:10 ais523, there? 14:51:15 yes 14:51:19 I got ccbi to generate an invalid svg file 14:51:23 at least according to inkscape 14:51:29 and konqueror 14:51:39 L0.0020,0.0020 14:51:39 "/> 14:51:40 oh yes 14:51:44 that isn't valid xml 14:51:52 no, the quotes aren't matched properly 14:51:53 cfunge does this right 14:51:56 ais523, indeed 14:52:22 Deewiant, http://rafb.net/p/tZQ61687.html <-- ccbi generates invalid code on that 14:52:31 cfunge can handle it 14:53:24 actually take http://rafb.net/p/o2X0l033.html, the ASCII art had an error 14:53:38 it would be nice to get the v> trick working in a vertical column rather than drifting to the left 14:53:43 I think it might be possible with flying IPs 14:54:01 instead of v, set the IP going diagonally down and to the right 14:54:02 ais523, well yes it would work in a jump table 14:54:09 but that needs more code 14:54:22 something like 11x instead of x 14:54:28 err "instead of v" 14:54:47 yep, 11x> at the end of each line should work 14:54:57 to make Befunge work more-or-less like 1D programming languages 14:55:03 haha 14:55:59 well flying ip is slower I think because it needs complex checks for wrapping 14:56:00 :P 14:56:17 AnMaster: most people writing Befunge programs don't optimise for speed 14:56:19 actually shouldn't matter as long as it doesn't actually wrap while flying 14:56:20 :P 14:56:31 but hehe 14:56:43 ais523, well removing white spaces would help with that 14:59:23 All in favor of me writing a Brainfuck interpreter that runs on raw hardware? 14:59:37 pikhq: you mean, creating hardware that runs BF natively? 14:59:41 I think that's been done before 14:59:43 but in favour anyway 14:59:49 ais523: No, just a kernel that runs Brainfuck. 15:00:06 oh, in that case you'd just write a BF interp that was also an OS 15:00:09 that isn't all that difficult 15:00:13 pikhq, make it portable! 15:00:13 ;P 15:00:18 I/O would probably be the hardest part 15:00:19 ais523: You're right. 15:00:29 Especially since I already have half of a kernel written. ;) 15:04:11 Deewiant, ais523: http://rafb.net/p/wPQrY673.html 15:04:14 new version 15:04:22 ais523, you will love that one ;P 15:04:41 heh 15:04:50 comments at the end too 15:05:18 the 11x> looks so much neater than the v> IMO 15:05:43 well since it stays in the same column it is useful indeed 15:05:56 and for a test suite I don't have time write compact code 15:05:57 I can do that too 15:05:59 a sec 15:06:17 ais523, http://rafb.net/p/t5usdB97.html 15:06:53 $ ./cfunge examples/count.b98 15:06:53 Enter a number: 34 15:06:53 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 15:09:18 http://rafb.net/p/rCRPk325.html 15:09:20 ais523, Deewiant ^ 15:10:18 stroke-width:0.00005px 15:10:24 that seems a bit small 15:10:34 I would have expected 1px 15:11:39 does TURT have a fill, by the way? 15:11:44 and does the TURT quine work in cfunge yet? 15:17:01 http://rafb.net/p/Haj5AB54.html <-- more extended 15:17:17 ais523, as for stroke, it is correct because everything is too small scale 15:17:19 it is a bug 15:17:25 I plan to rescale everything btw 15:17:38 ais523, as for turt quine I think it maybe be !Befunge specific 15:17:54 iirc !Befunge's TURT isn't completely correct 15:17:59 I don't think it uses any !Befunge-specific features except TURT 15:18:14 -!- Corun has joined. 15:18:15 ais523, it depends on differing implementation of TURT 15:18:19 yes 15:18:31 also debugging that quine is not something I plan to do 15:18:47 I make a test program, reason about it, then run it check if result is same 15:18:50 if it isn't it is a bug 15:19:00 or I reasoned wrongly 15:19:29 (that happened just now, what if you change pen color when it is down and then use T to teleport to another location, logically you got to lift the pen 15:20:03 anyway when Deewiant gets back I hope he can fix ccbi :) 15:20:12 * AnMaster ducks 15:20:55 ais523, also my turt is likely to still have bugs apart from very bad margins and stupid scale 15:21:18 well, most esocode has bugs 15:21:29 even something as simple as the original Malbolge interp had lots of vunerabilities 15:24:14 at least my margins doesn't cut off the image like ccbi does :P 15:28:40 ais523, btw you asked for function to execute on a fingerprint being unloaded? 15:28:46 that won't work 15:28:47 I don't plan to use it 15:28:50 I was just wondering 15:28:55 you can unload a fingerprint that isn't loaded 15:29:01 it is perfectly valid 15:29:24 consider A-Z as a set of stacks of function pointers 15:29:27 got that? 15:29:30 oh, does that rollback all the fingerprint commands it would define if it were loaded? 15:29:41 I think I get how fingerprints work 15:29:44 now load: push on the relevant stacks 15:29:47 the commands work the same way as variables in INTERCAL 15:29:58 a different stack for each command 15:30:01 but unload: pop *top item* if possible from the relevant stacks 15:30:18 so there is no way to know if it is the same fingerprint in fact that is unloaded 15:30:26 consider the NULL fingerprint for example 15:30:40 you can unload that a few time to clear anything loaded 15:31:01 so in effect "function on unload" is pointless 15:31:40 yes, I suppose so 15:31:50 unless you're using it as a 27th command rather than an unload hook 15:31:58 eh? 15:32:06 oh I see 15:32:13 how would that work? 15:32:17 Å? 15:32:31 well, the unload would do something, as would all the commands in the fingerprint 15:32:51 a fingerprint trying to unload itself in cfunge would cause havoc I bet 15:32:58 or rather, that depends 15:33:09 it couldn't even be sure it unloaded itself 15:33:52 the opcode stacks are quite simple: struct with size, top used item, pointer to memory block 15:34:56 -!- tusho has joined. 15:35:02 Dude fellows. 15:35:06 I wonder 15:35:07 hi ais523 15:35:08 hi tusho 15:35:09 ha 15:35:11 you won 15:35:11 I won that one any way you look at it 15:35:17 hi Slereah_ 15:35:22 ais523: i noticed you were in #canada when connecting 15:35:23 very amusing 15:35:24 Was Brainfuck developed independently of P''? 15:35:25 I was actually about to type hi to Slereah_ when you joined 15:35:30 Because if it is. 15:35:33 but I couldn't greet you until #esoteric was here, of course. 15:35:36 it was a bit offputting 15:35:41 Slereah_: i doubt it 15:35:48 he probably knew of it 15:35:49 With P'', BF and Spoon, it would have been invented three times D: 15:35:51 and thought it could be a could basis 15:35:58 So simple is the language. 15:36:06 and uh Slereah_ 15:36:08 spoon was based on BF 15:36:12 Was it? 15:36:16 yes 15:36:19 Slereah_, iirc BF was based on P'' 15:36:22 I seem to recall the guy saying that he did it independantly 15:36:36 Lemme check 15:36:53 "To be correct, I *re*discovered BrainF*** (known henceforth as simply 'Brain'). I thought wow! I thought oh! I thought damn - somebody's beaten me to it." 15:39:09 ais523, btw I wrote a few other tests for things that ccbi doesn't test. they are in tests/ in cfunge repo 15:39:33 some are cfunge specifc, like concurrent-issues.b98 iirc 15:39:57 (specifc as no other interpreter is likely to ever have the same issue) 15:55:35 -!- Sgeo has joined. 15:55:49 wowee 15:55:53 zzo invented another language 15:55:59 he's a machine 15:56:59 ais523: lawl 15:57:07 a guy on the esolang wiki talked about a language on the Inflection page 15:57:07 -!- ais523 has quit (Remote closed the connection). 15:57:13 -!- Corun has quit ("This computer has gone to sleep"). 15:57:15 without disclosing that it was his 15:57:17 and he linked to a wikipedia page 15:57:20 to add credibility I guess 15:57:22 it's been deleted 15:57:22 :p 15:57:24 -!- ais523 has joined. 15:57:30 well, you missed all of that 15:57:52 paste again? 15:57:54 oh, and hi tusho 15:58:03 lawl 15:58:04 tusho: a guy on the esolang wiki talked about a language on the Inflection page 15:58:08 tusho: without disclosing that it was his 15:58:08 [15:58] tusho: and he linked to a wikipedia page 15:58:08 [15:58] tusho: to add credibility I guess 15:58:08 [15:58] tusho: it's been deleted 15:58:08 [15:58] tusho: :p 15:58:26 zzo invented another language 15:58:29 what language? 15:58:36 AnMaster: see recent changes 15:58:36 varsig 15:58:37 i think 15:58:57 the deletion debate is gold 15:58:59 "The fact it exists makes it notable." 15:59:04 false 15:59:08 my right nostril's hairs are notable 15:59:08 :D 15:59:14 * tusho creates seventy articles about them posthaste 15:59:25 tusho: don't, that's probably speediable 15:59:33 ais523: duh, really?? i would never have guessed 15:59:34 under the "stop vandalising" criterion 15:59:36 which actually exists 15:59:37 I wish we were paid per pound of language. 15:59:44 Imagine what we could earn! 15:59:50 tusho, where on the esolang wiki? 15:59:59 http://esolangs.org/wiki/varsig 16:00:03 although I'm not tusho 16:00:11 yeah ignore him 16:00:14 he's an imposter 16:00:16 because he's not me 16:00:20 AnMaster: http://esolangs.org/wiki/varsig 16:00:22 ah 16:00:23 I am you. 16:00:24 there, I properly answered your question 16:00:27 Listen to me. 16:00:29 hahah 16:00:35 Slereah_: yes but you're the only one who's me 16:00:38 ais523 isn't me 16:00:46 if he was then we wouldn't argue so much 16:00:55 unless i have multiple personality disorder 16:01:01 I like the way varsig defines a convoluted way to do variables, but none of the examples use them, probably because they're not very easy to use 16:01:03 but that would be multi-threaded multiple personality disorder 16:01:10 which I don't think exists 16:01:17 seems like a rather crazy language 16:01:28 however what has this got to do with wikipedia I don't get 16:01:39 oh, and is it just me and tusho, or is ESME just random ramblings of nonsense? 16:01:47 ais523: yes 16:01:51 that's why I put it in the shame category 16:01:58 it's shameful 16:02:05 shame category? 16:02:05 Slereah_ suggested putting it in that category 16:02:06 so I did it 16:02:11 http://esolangs.org/wiki/ESME 16:02:12 ais523: oh, and also 16:02:18 he made a link like 16:02:20 [[Wikipedia:foo]] 16:02:21 or whatever 16:02:22 which is valid 16:02:23 then 16:02:24 "There is currently no text in this page, you can search for this page title in other pages or edit this page." 16:02:24 he changed it to 16:02:26 -!- Corun has joined. 16:02:29 [[WikiPedia:foo]] 16:02:33 i reverted, saying he was right the first time 16:02:38 he re-reverted, saying "I prefer it this way" 16:02:43 maybe it's http://esolangs.org/wiki/Esme 16:02:46 That's Like The People Who Type Like This 16:02:49 And If You Complain 16:02:50 I'm not sure, HTTP just stopped working for me 16:02:54 They Tell You To Stop Insulting Their Style 16:02:57 for no apparent reason 16:03:21 * ais523 resets eir Internet connection 16:03:24 ah yes that is it 16:03:26 -!- ais523 has quit ("resetting my Internet connection"). 16:03:43 http://esolangs.org/wiki/Esme 16:03:45 i vandalised it a bit 16:03:48 since it'll probably be deleted soon 16:04:30 it lacks specs... 16:04:39 wow really 16:04:41 i didn't notice. 16:04:45 your eye is keen, AnMaster. 16:04:55 well wtf is it doing on the wiki without specs or link to specs? 16:05:07 AnMaster: Someone added it. 16:05:08 (WOW) 16:05:20 well seems to be way below quality standard 16:05:22 http://esolangs.org/wiki/Category:Shameful :D 16:05:24 I think that he figured that, because it was a "joke language", it didn't need anything. 16:05:35 AnMaster: So is FURscript, but we keep it because it's funny. 16:05:36 "Boy I will be so random, and it will be amusing!" 16:05:37 Slereah_, even they need specs 16:05:52 "The structure is based off a mix of html, turbo pascal, and BASIC. " 16:05:54 heheh!? 16:05:55 -!- ais523 has joined. 16:06:00 AnMaster: furscript is totally serious 16:06:01 :| 16:06:04 read the talk page 16:06:15 a guy transferred it to esolang because someone put it on their wiki 16:06:22 but it was too bad to stay there. 16:06:27 oh and none of the lesser known programming languages have specs 16:06:29 they still have articles 16:06:33 ais523: http://esolangs.org/wiki/Category:Shameful 16:06:46 Well, the lesser known programming languages are somehow fun. 16:06:52 tusho: VALGOL and SARTRE both have specs 16:06:54 And they got sources, I think. 16:07:02 ais523: after the fact 16:07:07 tusho: yes 16:07:23 btw last time someone tried to create categories on Esolang they got blocked for it, by Graue 16:07:27 oh and none of the lesser known programming languages have specs 16:07:27 they still have articles 16:07:29 which ones? 16:07:33 I love the edit summary. 16:07:34 "(shame)" 16:07:39 so we codified it into policy "don't create categories without discussion" 16:07:42 either implementation, specs or link(s) to spec 16:07:46 are needed 16:07:48 ais523: i didn't create a category 16:07:51 just a warning, I won't block you for it 16:07:51 i never touched a category page 16:07:54 I just added category links 16:07:55 ais523 : Well, we discussed it here! 16:07:57 tusho: well, adding a redlinked category 16:07:59 so ha 16:08:03 And we all agree and all. 16:08:16 i'll revert "Esme is a shameful esoteric programming language created by User:Dagoth Ur, Mad God because he has no language creation talent." though 16:08:20 because shame should be untarnished 16:08:22 and authentic 16:08:35 -!- Slereah_ has set topic: Esme is a shameful esoteric programming language created by User:Dagoth Ur, Mad God because he has no language creation talent. | #esoteric - the international hub for esoteric language design, development and deployment | logs: http://tunes.org/~nef/logs/esoteric/. 16:08:44 yes, it can stay there 16:10:45 tusho, IMO every language on the esoteric wiki should have specs, links reference implementation(s) or links to specs to be useful 16:10:58 AnMaster: uh what about unimplemented languages 16:10:59 even HQ8++ or whatever they are called got that 16:11:05 tusho, well then there are specs 16:11:08 under-construction ones, underspecified ones that are still interesting,... 16:11:10 or != and 16:11:15 hm true 16:11:19 but there is a limit for it 16:11:20 it's pointless to specify such rigorous standards for bloomin' esoteric languages 16:11:27 esme is way below that limit 16:11:34 AnMaster: yes, but it's rather amusing 16:11:41 it _should_ be deleted, but it's like a work of modern art 16:11:41 not really 16:11:45 hah 16:11:46 you can sit there and admire it 16:11:55 and try to take in the mental damage used to create it 16:11:56 well how does the example work? 16:11:58 but you can never envelop it all 16:12:03 AnMaster: that's the zen part of it 16:12:11 or what does the example do? 16:12:13 you don't know until you forget 16:12:19 Esme should stay there. 16:12:21 I think it's a language with a concept but no spec 16:12:27 And it should have the tag "NEVER FORGET" 16:12:33 basically the author things that the example should be what an Esme program should look like 16:12:38 ais523: not much of a concept 16:12:47 i think we should protect it so he can't flesh it out 16:12:51 Like a reminder for future generations. 16:12:51 well, it's an idea for an art-language 16:12:52 tusho, why? 16:12:54 :( 16:12:56 AnMaster: it would lose its appeal 16:12:58 and tusho, that's against the idea of a wiki 16:13:03 he should flesh it out IMO 16:13:21 AnMaster: well, maybe he'd make it less shameful 16:13:24 which would be a shame [ha] 16:13:31 good idea if he did 16:13:51 yeah but then Category:Shameful would only include furscript 16:13:56 :p 16:14:05 well furscript should be deleted 16:14:17 no it shouldn't 16:14:17 it isn't esoteric at all, it is just a bad failure 16:14:18 it actually has a spec 16:14:24 and it certainly is esoteric 16:14:28 just not in a good way 16:14:35 tusho, well less esoteric than the Perl entry 16:14:36 ais523: http://esolangs.org/wiki/User:Dagoth_Ur%2C_Mad_God/monobook.js wtf? what does that do? 16:14:40 change 'special page' to 'special'? 16:14:43 AnMaster: hell no 16:14:45 perl is just concise 16:14:52 furscript is esoteric because it can't do anything useful 16:14:56 but it can do really weird things 16:15:01 tusho, http://esolangs.org/wiki/Perl 16:15:01 and it does them in a strange, awkward, and crazy way 16:15:05 i know 16:15:06 i've seen 16:15:13 it is very accurate 16:15:15 and funny 16:15:21 AnMaster: no it's not 16:15:28 the Interpretation is funny 16:15:34 but the program isn't funny 16:15:38 and the implication isn't true 16:15:42 tusho, well it is obscure 16:15:52 so is furscript 16:16:03 tusho: the interpretation is correct, you just don't understand the language it's written in 16:16:05 and obfuscated 16:16:14 it's a SARTRE-like language 16:16:15 ais523: *g* 16:16:20 and yes I know it counts lines in files and subtract files or something like that 16:16:23 AnMaster: not intentionally obfuscated 16:16:26 it looks pretty basic 16:16:30 tusho, that perl? 16:16:32 ais523: http://esolangs.org/wiki/Talk:Esme look, he actually tried to explain it 16:16:34 zzo38 asked 16:16:58 hah 16:17:10 tusho: that explanation strikes me as being a joke, more or less, like the language 16:17:14 oh I think I see 16:17:16 it's clear to me he's describing a paradigm 16:17:22 rather than an actual specced-out language 16:17:23 it is *output* of the program 16:17:28 Is he? 16:17:35 so you can only type "esme" + some ! 16:17:36 ais523: the paradigm of 'lol, the name "esme" is funny and ESME!!esmeMEMEESMESMSME is funnier' 16:17:44 ahh I see 16:17:46 'esme esme' is the program 16:17:50 no wait 16:17:52 nope 16:17:59 it's on one line in the source 16:18:03 tusho, I *think* it may be like this: 16:18:05 emse 2 16:18:11 ESME!! 16:18:12 err 16:18:12 tusho: I didn't say it was a good paradigm 16:18:14 AnMaster: what does 'HE output' mean 16:18:15 esme* 16:18:19 Stop pondering, and let it remains in its category of SHAME 16:18:19 tusho, no clue 16:18:26 maybe THE output? 16:18:29 could be a typo 16:18:45 It's also the same "paradigm" as ook, cow or AAAAAAH 16:18:51 it's a SARTRE-like language <-- wtf is that? 16:19:08 AnMaster: SARTRE was one of the lesser-known langs, but Chris Pressey specced it 16:19:13 it's on catseye somewhere 16:19:17 ais523, hm 16:19:40 found it 16:19:47 http://catseye.tc/projects/sartre/doc/sartre.html 16:20:55 that's the first lang I've seen that mandates that comments must not be misspelled and allows compilers to spellcheck them 16:20:56 ais523: you know wikipedia, can you tell me why tony sidaway never stops changing his name 16:21:04 tusho: no, I can't 16:21:04 and why is it always something strange 16:21:11 I don't know 16:21:22 it's very confusing 16:21:27 ais523, haha 16:22:37 "A special command which, due to the resignation of the programmer, is permitted to perform a wide variety of tasks, among them, alter the direction of program flow, execute a random function, terminate the program, or positionally invert the bits in the data region." 16:23:00 that reminds me so much of the low-quality esolangs that some people turn out 16:23:06 have a command that can do more or less anything at random 16:23:52 has sartre ever been implemented? 16:24:27 I doubt it 16:24:41 normally there's an impl on catseye if it's been implemented and there's a spec 16:24:54 however, looking at that lang, it looks like it might potentially be TC 16:24:57 which disappoints me 16:25:08 I wanted a lang where every possible program was a NOP 16:25:19 that would be much more interesting IMO 16:25:33 there is some such program 16:25:56 oh, it wasn't Chris Pressey, apparently, even though it's on their website, it's John Colagioia 16:26:34 their? 16:26:42 singular they 16:26:48 I'm not sure of Chris Pressey's gender 16:26:57 ais523: i'm pretty sure he's a he. 16:26:57 the name doesn't give a clue eitehr 16:27:01 Chris is he isn't it? 16:27:03 tusho: well, it would seem likely 16:27:05 AnMaster: it's both 16:27:40 ais523: i mean, it's an unfortunate but true fact that the number of females doing esolangs is quite a bit less than males 16:27:45 and I'm sure he might refer to himself as male somewhere on his site 16:27:49 either way, it seems very likely 16:27:54 s/esolangs/programming/, probably, I suspect that's the reason 16:28:04 ais523: but esolangs even more, I'd say 16:28:13 but esolangs are an art form 16:28:15 or at least can be 16:28:24 most of the female programmers i've heard of generally program to get things done 16:28:29 instead of messing around with esolangs and similar 16:28:32 shrug 16:28:39 there's always nerds :p 16:28:42 ais523: Programming is an art form, IMO. 16:29:03 Of course, it's an art form filled with people who don't know a damned thing about artistry. 16:29:16 have you never seen a Perl koan? 16:29:23 s/koan/haiku/ 16:29:26 sorry, wrong art form 16:29:30 perl koan 16:29:31 wow 16:29:36 ais523: heh, pressey dislikes wolfram 16:29:36 '# pedlars of profundity (Penrose, Wolfram, Hofstadter...)' 16:29:44 (under Things I Could Do WIthout on his personal page) 16:29:53 I've seen quite a few gorgeous Perl hacks. . . 16:30:01 But not any haiku that I can think of. 16:30:25 http://aspn.activestate.com/ASPN/Perl/Haiku/InPerl 16:30:26 Well, Penrose is okay. 16:30:33 His tensorial notation makes me smile 16:30:37 that seems to have been a challenge to write haiku that was also legal Perl 16:31:07 :) 16:31:26 ' no less can I say; 16:31:26 require strict, close attention 16:31:26 while you ... write haiku' 16:31:30 can you get that to run? 16:31:43 I suspect so 16:31:53 the first line unloads a library called "less", with three string arguments 16:31:59 yes 16:32:02 then require strict verifies that use strict is possible 16:32:04 which it is 16:32:05 require strict loads strict.pm? 16:32:07 ah, okay 16:32:07 so no problems there 16:32:10 close attention... 16:32:12 we need an attention file 16:32:15 but that should work 16:32:21 and it requires strict and closes attention 16:32:25 while you ... write haiku 16:32:30 My favorite one? 16:32:31 you and write haiku must evaluate to something rangable 16:32:32 wait, that's all one command 16:32:35 ais523: no 16:32:36 it repeatedly closes attention 16:32:38 'no less I can say;' 16:32:40 two commands 16:32:41 $my_args = shift;system("gcc $my_args");print "I prefer C\n"; 16:32:43 that's what the while is doing 16:33:02 "The Sartre scoping rules are somewhat complex in that it may only utilize data which has been accessed previously or any data which it makes up itself. Data which has not yet been accessed is unknown to the Sartre nihilist, however." 16:33:04 well 16:33:14 that means it is a NOP I think? 16:33:18 and you ... write is rangable, presumably, because you can range two strings 16:33:30 AnMaster: no, it can create variables and then access them 16:33:36 hrrm 16:33:45 ah, presumably write haiku has a return value? 16:33:49 that would be rangeable 16:35:14 presumably, programming language haiku only works properly in langs which allow lots of barewords 16:35:58 ais523: ruby poems are nice 16:36:07 does ruby have barewords too? 16:37:01 -!- timotiis has joined. 16:38:19 -!- augur has joined. 16:38:49 You know a package's build system is bad when you have to write a patch just to make the build system cross-compile. . . 16:38:56 Or use a different C compiler, for that matter. 16:39:07 ... Or use different *arguments* for said C compiler. . . 16:39:20 pikhq: I wonder if C-INTERCAL cross-compiles, I've never tried 16:39:27 ais523: Do you use autotools? 16:39:29 haha 16:39:35 it would probably need different arguments for config.sh 16:39:37 pikhq, autoconf but not automake iir 16:39:40 irc* 16:39:41 pikhq: autoconf but not the others 16:39:44 Ah. 16:39:49 That ought to suffice. 16:39:55 * tusho needs to get on his autotools-replacement thing sometime 16:40:10 Though using the rest of autotools would make it much easier. 16:40:10 pikhq, well you need some other stuff in config.sh too 16:40:11 tusho: what would you do differently from autotools? 16:40:19 ais523: Not be insane? :p 16:40:27 GET_CANNONICAL_TARGET or whatever it was 16:40:33 Just ./configure --target=some-other-target would work perfectly. 16:40:40 I don't find autotools that insane, it appears insane because it's trying to do something insane 16:40:50 pikhq, well you need that macro in configure.ac then 16:40:54 whatever it was 16:40:57 ais523: what it's doing isn't as insane as how insane it is, though 16:41:13 hmm... I have a cross-compiler to ARM here, maybe I can try seeing if C-INTERCAL works with that 16:41:15 AnMaster: Fine, so I assume that you have used Autotools *right*. 16:41:24 pikhq, a lot yes 16:41:30 Like, say, up to GNU's packaging standards. ;) 16:41:34 one problem is that it isn't in my path 16:41:35 pikhq, no 16:41:48 ... No? 16:41:52 pikhq, I used it but not "up to gnu's packaging standards" whatever they are 16:41:53 pikhq: well, C-INTERCAL had a configure script when I came to it but mostly ignored its output 16:42:06 I've redone the build system at least twice since then 16:43:51 ah wasn't it AC_CANONICAL_TARGET that was needed? 16:43:53 pikhq, no? 16:43:59 Dunno. 16:44:04 I'm not an Autotools expert. 16:44:09 (I should learn it this summer) 16:44:21 cmake is better 16:44:46 Probably. 16:44:47 I don't have AC_CANONICAL_TARGET in my config.ac for C-INTERCAL 16:44:52 But Autotools is fairly ubiquitous. 16:45:50 Though Cmake is probably going to become much more so, now that KDE uses it. 16:46:12 ais523, I may be wrong 16:46:40 AnMaster: well, I haven't tried at all 16:46:45 maybe I should persuade pikhq to try 16:46:56 e'd get a top-tier modern INTERCAL compiler too 16:46:59 hmm... I have a cross-compiler to ARM here, maybe I can try seeing if C-INTERCAL works with that <-- good idea 16:47:09 AnMaster: yes, but it was set up weirdly 16:47:15 oh? 16:47:27 Lemme learn Autotools, and then I'll go ahead and try Autotoolising C-INTERCAL. :p 16:47:32 I got it by building gcc from source in a subdir deep in my home dir 16:47:46 pikhq: that could be fun, given the way the build system currently works 16:47:50 it already uses autotools 16:47:55 AnMaster: no, only autoconf 16:48:02 I could probably change a sane build system to use automake 16:48:08 I have a lot of experience with it 16:48:28 ais523, however I'm not sure if c-intercal's build system qualify as sane 16:48:33 AnMaster: well, does it deal with having to compile your own compilers to compile the source into C, then compile the C into the finished version? 16:48:51 ais523, it can be done but not fun 16:49:07 you can add custom targets easily enough 16:49:20 also there's a point where one .oil file splits into lots of .c files 16:49:28 all of which have to be compiled and linked back into one executable 16:49:29 ais523, what sort of names? 16:49:44 AnMaster: they follow a pattern, oilout00.c, oilout01.c, oilout02.c and so on 16:49:45 in hex 16:49:51 ok 16:50:09 should be doable 16:50:39 oh, I imagine it's all doable 16:50:43 ais523, however I would begin with converting other directories 16:50:56 what do you mean by that? 16:51:17 ais523, you can use automake in one dir and just autoconf in another 16:51:31 so you don't need to convert all at once 16:51:43 well, all the source is in the same directory, /src 16:51:51 it compiles into things in /tmp 16:51:59 out of tree builds? 16:52:02 and the output goes in /bin and /lib 16:52:03 is that supported? 16:52:08 AnMaster: it doesn't do out of tree at present 16:52:13 ais523, ah 16:52:23 ais523, why the compile into tmp? 16:52:24 for much the same reason cfunge doesn't build if you lose the tree structure in its sources 16:52:26 is there any reason 16:52:33 AnMaster: because there are lots of temporary files that need to be created 16:52:39 all the .o files are kept out of src 16:52:43 ais523, cfunge builds out of tree however 16:52:45 also there are .c files to be generated, and .h files 16:52:48 or at least did recently 16:53:24 well, arguably C-INTERCAL always builds out of tree because the src directory is untouched 16:53:39 but it's always in the same out-of-tree place 16:53:44 well I mean: mkdir build; cd build; cmake ..; make 16:53:47 should work for cfunge 16:53:49 mixing the results of compilation up with the sources is ugly IMO 16:54:15 phone brb 16:55:21 -!- sebbu2 has joined. 16:57:25 -!- augur has quit (Read error: 113 (No route to host)). 17:02:56 back 17:03:15 ais523, well just do a true out of tree build IMO 17:03:47 well, you can simply duplicate the tree, that's easy enough, right? 17:04:01 hm? 17:04:05 cp -r 17:04:18 um I said out of tree build 17:04:23 * ais523 wonders if there's a way to do a recursive ln 17:04:29 AnMaster: I'm talking about how to do the same thing 17:04:33 you don't need to ln 17:04:45 make will overlay the build dir and the real dir 17:04:48 internally 17:04:51 well, yes 17:04:57 and you don't need to create any subdirs 17:05:04 just an empty dir and run like: 17:05:09 I'm trying to think of a simple way to do out-of-tree builds when the source wasn't set up for them 17:05:12 ../ick/configure 17:05:24 and I know how out-of-tree builds work normally 17:05:26 ais523, well imo it should support it :) 17:05:51 -!- augur has joined. 17:05:56 ais523, where is the makefile.in in subdirs!? 17:06:04 probably in sr 17:06:05 oh you only use one single top makefile? 17:06:06 s/sr/src/ 17:06:12 if it isn't in the top 17:06:16 yes, there's only one single makefile 17:06:20 what subdirs are you thinking of? 17:06:21 ah 17:06:25 well src 17:06:30 oh, there's a makefile in doc, but it's independent 17:06:33 * AnMaster is used to autotools + recursive make 17:06:44 not autotools + one top makefile 17:07:02 -!- sebbu has quit (No route to host). 17:07:02 -!- sebbu2 has changed nick to sebbu. 17:08:25 ais523, oh btw I can convert ick to use automake yes, however no idea about your @OBJEXT@ mess 17:08:36 -!- Corun has quit ("This computer has gone to sleep"). 17:08:37 doesn't automake handle that itself? I'm unsure 17:08:37 AnMaster: that's simple enough 17:08:52 object files end .o on Linux and .obj on DOS 17:09:00 likewise for @EXEEXT@ which puts on the .exe extensions if needed 17:09:05 right 17:09:09 it's a feature built into autoconf 17:09:10 what about .com then? 17:09:11 AnMaster: your TURT working yet? 17:09:30 AnMaster: gcc doesn't generate .com files 17:09:35 and C-INTERCAL doesn't use them 17:09:48 Deewiant, yes it does in trunk 17:09:56 Deewiant, better than yours at least :P 17:09:58 good 17:10:00 yep 17:10:09 like said, it was hardly tested at all :-) 17:10:13 Deewiant, mine doesn't handle bg colors yet but nor does your 17:10:29 Deewiant, see the link I gave earlier 17:10:38 I'll see everything later 17:10:44 can't be bothered today 17:11:08 I'll fix it on the weekend the latest 17:11:46 ais523, would this be correct for ick: 17:11:49 bin_PROGRAMS = ick convickt ial 17:11:49 noinst_PROGRAMS = oil 17:11:49 lib_LIBRARIES = libick.a libickmt.a libyuk.a libickec.a 17:11:59 yep, looks right 17:12:09 ais523, what is ial? 17:12:13 oh, no 17:12:17 ial's a dummy target 17:12:21 that handles the includes and libraries 17:12:24 it isn't an executable 17:12:28 apart from that it's right 17:12:29 I see 17:12:44 then apart from oil it should be pretty simple 17:13:16 ais523, does SOURCES contain files for ick or for all? 17:13:29 I need variables with source files for each target basically 17:13:38 AnMaster: for everything 17:13:43 some sources go in multiple targets 17:13:43 hm 17:13:49 ais523, right 17:13:53 look at the link lines for each library and executable 17:13:57 ah 17:13:58 that'll explain what goes where 17:14:08 temp/parser.o temp/lexer.o temp/feh2.o temp/dekludge.o temp/oilout-m.o temp/ick_lose.o temp/fiddle.o temp/perpet.o temp/uncommon.o 17:14:11 that is all for ick? 17:14:18 yes, looks like it 17:14:27 oh wait, there's another noinst_PROGRAM 17:14:31 that generates oilout-m.c 17:14:41 I think I called it bin2c 17:14:53 it just takes a binary file and converts it into a C file defining one variable 17:15:05 ah I see 17:15:29 what exactly is ial for? 17:15:33 I don't get iot 17:15:34 it* 17:15:48 AnMaster: it's a dummy target that causes all the includes and libraries to be copied into appropriate locations in the tree 17:15:55 I see 17:16:04 ick runs straight from the tree, checking ../lib and so on if it can't find things in the PREFIX 17:16:19 ais523, why the need to copy 17:16:31 basically, there are three forms of the distribution 17:16:37 before it's compiled, everything's in src 17:16:56 once it's compiled, it builds in /lib and /include and /bin 17:17:00 including copying things over if needed 17:17:03 and it can run from those 17:17:07 so that's a binary version in-tree 17:17:13 then make install copies the files from there into the PREFIX 17:17:25 basically it can run off a make with no make install 17:17:31 hrrm does the top oil.c include all the other parts or? 17:17:32 and there are cases in the code to check if that's happened 17:17:36 temp/oil.c: src/oil.y temp/config.h 17:17:38 you still got that 17:17:40 AnMaster: oil.c is output from oil.y 17:17:46 oh right 17:17:49 that is simple :) 17:18:31 yep, just a straightforward yacc parser that defines OIL, the translation, etc., is in the same source file 17:19:00 temp/parser.h temp/parser.c 17:19:05 how do you generate the header? 17:19:20 AnMaster: yacc does it automatically when compiling 17:19:24 it outputs both a .h and a .c 17:19:33 with silly filenames IIRC but I mv them into the correct place 17:19:43 aaargh 17:20:01 AnMaster: that isn't even my silliness 17:20:05 that's what yacc/bison do by default 17:20:13 yes but moving it is silly 17:20:15 and autoconf has checks to find out what filename it uses for its output 17:20:27 AnMaster: not if you have two files to build, both using yacc 17:20:34 moving it is not silly because it avoids name collisions 17:21:08 well automake will keep track of how the files depend on each other for you 17:21:12 no need to list headers 17:21:29 even when the headers depend on other headers? 17:21:34 -!- sebbu2 has joined. 17:21:38 and when the headers depend on config.h too? 17:21:45 yep 17:21:51 and which headers are included depends on config.h 17:21:53 it uses the compiler to extract the info 17:22:04 although admittedly that's system headers, not my headers 17:22:07 on first time the file is built 17:22:13 or after a re-configure 17:22:24 also, I think some of the headers include different files based on which file they're included into 17:22:26 so incremental builds work 17:22:55 ais523, this will happen on a C file by C file basis 17:22:59 ok 17:23:18 what can cause trouble is the oil splitting thing 17:23:40 I did that because not doing it was causing Debian trouble 17:23:47 the files were getting to large to reliably compile 17:23:55 s/to/too/ 17:24:02 that's not with /g so it only affects the first to 17:24:52 ais523, I'm not sure how to express that a unknown set of files is generated from one file 17:25:04 wildcards? 17:25:15 besides, Info does that too 17:25:17 not sure if they work at the time it is expanded 17:25:25 ais523, wtf is lextest? 17:25:38 AnMaster: it basically built the lexer with a main(), it's not used nowadays 17:25:43 it was used early on to test the lexer 17:25:48 INTERCAL is not trivial to lex... 17:26:05 temp/oilout-m.o? 17:26:07 what is that 17:26:22 it's the main file 17:26:28 that calls all the other oilout files 17:26:37 each other file contains a function 17:26:44 and oilout-m just calls all the functions in order 17:26:51 to effectively make one big function 17:26:55 hm 17:28:10 lib/syslibc.c:pit/explib/syslibc.c 17:28:11 -cp pit/explib/syslibc.c lib/syslibc.c 17:28:14 what is that good for? 17:28:32 AnMaster: basically, syslibc.c and some other files (like syslib.i) are used by the compiler 17:28:46 but are also example INTERCAL programs, or examples of the syscall thing, or whatever 17:28:51 basically, /src holds files for the compiler 17:29:00 /pit holds example programs 17:29:06 but why not just use it directly, why copy it around? 17:29:08 but some of those programs, like the system library, are needed for compilation 17:29:21 AnMaster: because eventually it has to be installed into /usr/share 17:29:42 and because the /bin /lib /include form a working binary distribution 17:30:04 basically, it's source distribution -make-> binary distribution -make install-> binary distribution in the correct place 17:30:13 the binary distribution runs from /bin /lib and /include 17:30:14 ah, you could just copy it directly, would be easier even with automake 17:30:19 which means all the files have to be there 17:30:33 and not requiring a make install is very useful for people just trying out INTERCAL 17:30:38 saves having to use prefixes and all that 17:30:44 besides, they used not to work 17:40:44 -!- sebbu has quit (No route to host). 17:41:00 -!- sebbu2 has changed nick to sebbu. 17:41:03 -!- tusho_ has joined. 17:41:31 stupid fucking router 17:47:53 -!- augur has quit (Read error: 113 (No route to host)). 17:51:45 ais523, I can figure out how to express a fixed set of files for oil 17:52:00 but not a variable set, wildcard doesn't work before the file is generated 17:52:06 we got a bootstrap issue in fact 17:52:07 AnMaster: well, the number of files used will increase over time as more idioms are added 17:52:19 exactly 17:52:32 ais523, apart from that I almost finished converting it 17:53:56 -!- timotiis has quit (Read error: 110 (Connection timed out)). 17:54:12 ais523, I think this would work apart from install and from the oil thing: http://rafb.net/p/z5P3TG52.html 17:54:19 maybe a few more variables but almost only that 17:55:09 ais523, much shorter as you can see 17:55:26 *.y -> *.c is handled automatically 17:55:28 that doesn't handle the install of things like coopt.sh and syslib.i 17:55:31 same for *.l 17:55:34 ais523, indeed 17:55:40 I said apart from install and oil 17:55:40 does it handle *.y -> *.h too? 17:55:47 ais523, well it should 17:56:09 what about all the command-line args needed 17:56:20 there are quite a few -Ds involved 17:56:23 -!- tusho has quit (Read error: 110 (Connection timed out)). 17:56:42 ais523, aren't they defined CFLAGS? 17:56:44 or such? 17:57:24 AnMaster: I use CFLAGS as well 17:57:24 -!- tusho_ has changed nick to tusho. 17:57:37 ah that would end up as: 17:57:37 AM_CFLAGS = -O2 -W -Wall -DICKINCLUDEDIR=\"$(incdir)\" -DICKDATADIR=\"$(datadir)\" -DICKBINDIR=\"$(bindir)\" -DICKLIBDIR=\"$(libdir)\" -DYYDEBUG -DICK_HAVE_STDINT_H=@HAVE_STDINT_H@ -I./src -I./temp 17:57:46 or CPPFLAGS for defines 17:57:56 as CPPFLAGS = for precompiler and CFLAGS for compiler 17:57:56 AnMaster: but things like $(incdir) rely on things higher up the makefile 17:58:00 which are set by autoconf 17:58:00 they should be split 17:58:04 how does automake handle those 17:58:19 well as autoconf is still used that will be set above 17:58:50 ais523, automake doesn't replace the syntax of autoconf, it simply extends it quite a bit 17:59:06 AnMaster: things like datadir are set in Makefile.in 17:59:09 from autoconf variables 17:59:19 I'm not sure if automake would have variables with the same name 17:59:20 ais523, yes that will still work 17:59:22 trust me 17:59:28 I could have called it pinkfluffyponies and it would still work 17:59:45 well if you had a line like: pinkfluffyponies = @pinkfluffyponies@ 17:59:47 then it would 17:59:50 :P 18:00:00 or whatever 18:00:02 hm, ok 18:00:47 ais523, automake will lump together lots of variables at the start (all those you don't explicitly define) and then put your defines and finally the targets 18:00:51 mostly that is what happen 18:01:23 ais523, anyway installing would take some time to convert 18:01:44 well, it isn't that hard, just making dirs and copying files 18:01:53 although I am disappointed that your way requires an install 18:02:01 ais523, well that could be changed 18:02:14 yes, I suppose so 18:02:19 ais523, however it wouldn't be as trivial 18:02:24 I'm not sure *I* could pull it off 18:02:27 also, where does automake put the obj files? 18:02:31 in the same place as the sources? 18:02:35 that strikes me as a bad idea 18:02:39 ais523, well I assume you will do a true out of tree build 18:02:44 :P 18:02:46 what if I have a .o in the sources with the same name as the .c 18:03:09 ais523, why would you have that? and as I said: I assume true out of tree builds will be done 18:03:15 or even enforced like gcc enforces it 18:03:47 AnMaster: it's not beyond the realms of possibility that I might want a .c and a .i with the same name 18:03:51 currently C-INTERCAL doesn't allow that 18:04:14 well... no idea :P 18:04:42 ais523, I have autotools on massive but *sane* projects, nothing like ick's all build system quirks 18:06:44 *.c: *.bin 18:06:45 ./bin2c blergh < $< > $@ 18:06:50 well blergh make that not work 18:06:53 otherwise it would work 18:07:10 ais523, see ? 18:07:14 AnMaster: I have to specify the names of the character sets somehow 18:07:21 as it happens, they're specified in the makefile 18:07:54 xpm makes the name of the variable in the file dependant on the file name iirc 18:08:03 something like that could make sense I guess 18:08:22 yep, I do that somewhere too 18:08:33 the .bin files are used for two things 18:08:41 they're installed, and used as .bin files, by convickt 18:08:55 but they're also converted into .c then .o and linked to the runtime libraries, for use by the I/O code 18:09:01 LOADLIBS = @LIBS@ @LEXLIB@ 18:09:03 it is never used 18:09:06 so what is it for? 18:09:29 that's been there forever, more or less, probably something to do with the last-but-one build system 18:09:44 ais523, well, it doesn't seem to be used at all? 18:10:11 well, quite possibly it isn't at present 18:10:23 I think it was probably used by an old install system, or something like that 18:10:39 ais523, anyway I fail at expressing the mutli-unknown-file dependency on the generated oil files in a way that can be resolved in advance (which is needed) 18:11:12 apart from that and some painful with with install my conversion probably works 18:12:24 so I give up 18:13:00 well, I'm not sure how much of an advantage automake would have over the current build system anyway 18:13:08 the current system is at least nice and expressive of what it does 18:13:13 showing all the steps explicitly 18:13:41 ais523, it would be considerably shorter and take care of tracking dependency on headers automatically, also it would be easier to maintain 18:14:21 ais523, most of the time you can shrink numbers of lines/chars/whatever considerably by using automake 18:14:34 it needs automake, though. Does it run on DOS? (Last I tried, configure scripts had to be built seperately for DOS.) Does it run on systems which don't have automake? 18:14:52 ais523, the files can be generated in advance 18:14:57 what about systems where the default shell isn't sh-compatible? (I have sh -c at various points for that reason.) 18:14:58 automake will generate autoconf files 18:15:06 which then will be processed by configure 18:15:38 ais523, also it should be as compatible as the generated configure 18:16:01 which iirc is pretty well 18:16:24 ais523, it does work on mingw+msys on windows too 18:16:31 I used projects which had to do that 18:16:52 ais523, as for DOS: no clue and I don't think anyone else knows either :P 18:17:19 AnMaster: well, there's evidence in the autoconf changelog that they tried to fix it to run on DOS 18:17:25 also in comments in the source code 18:17:31 ais523, I assume you will put cfunge in some separate source directory? 18:17:32 so I'm hopeful that it'll work next time I tried 18:17:41 AnMaster: I haven't thought about how to distribute cfunge yet 18:17:53 I'm thinking about distributing fffungi separately from ick 18:17:58 oh? 18:18:03 after all, different licences, different packaging, and so on 18:18:11 if you distribute it separately you should probably use cmake 18:18:13 I can help there 18:18:36 ais523, I don't think cmake works on DOS, but it does for about everything else :P 18:18:49 including MSVC project files 18:19:14 but cfunge won't compile under MSVC as it lacks C99 and you would also need to remove some fingerprints that depend on fork() and such 18:19:26 ais523, I guess ick has no problem with vfork() btw? 18:19:28 or fork() 18:19:31 forgot what one I use 18:19:37 AnMaster: if followed by an exec, no 18:19:45 otherwise yes 18:19:49 it is indeed followed by an exec 18:20:11 ais523, well a few pipe operations + an exec 18:20:17 // Do the FD stuff. 18:20:18 // Close unused end 18:20:18 close(outfds[0]); 18:20:18 // Dup the FD 18:20:18 dup2(outfds[1], 2); 18:20:22 yep, that's fine 18:20:23 things like that in the forked side 18:20:35 ais523, oh a strdup to build arguments array too it seems 18:20:37 as long as the forked side never returns 18:20:41 but that is a heap operation 18:20:49 indeed, it doesn't return 18:20:54 so no problem 18:21:25 it either exec() or _Exit()s 18:21:38 _Exit() in case of errors 18:21:43 _Exit? 18:21:46 I thought it was _exit 18:21:57 The function _exit() terminates the calling process "immediately". Any open file descriptors belonging to the process are closed; any children of the 18:21:57 process are inherited by process 1, init, and the process's parent is sent a SIGCHLD signal. 18:22:06 _Exit is C99 18:22:11 while _exit is POSIX 18:22:15 what's the difference? 18:22:16 anything? 18:22:18 Feature Test Macro Requirements for glibc (see feature_test_macros(7)): 18:22:18 _Exit(): _XOPEN_SOURCE >= 600 || _ISOC99_SOURCE; or cc -std=c99 18:22:20 ais523, only the name 18:23:12 and that one is C standard while the other is POSIX standard 18:23:28 "The function _Exit() is equivalent to _exit()." 18:23:28 well, you need both, right? 18:23:34 both C99 and POSIC 18:23:36 ais523, no why would I? 18:23:38 s/POSIC/POSIX/ 18:23:41 I'm C99 18:23:53 and POSIX.1-2001 defines _Exit() too 18:23:54 you use srandom() IIRC? 18:23:59 I thought that was POSIX 18:24:02 C has srand() 18:24:02 ais523, yes I do. I need both yes 18:24:12 I don't need both _exit() and _Exit() 18:24:17 I thought that was what you said :P 18:24:46 ah 18:25:07 and yes environ is POSIX too 18:25:14 there are a few more things like fork() and such 18:25:31 ais523: you know, integrating ccbi would probably have been less crazy.. 18:25:40 tusho, ccbi is D.... 18:25:42 tusho: ccbi's written in D 18:25:47 that would be painful and hard to compile 18:25:56 also, I think cfunge works really well with this 18:25:59 D isn't painful 18:25:59 :p 18:26:07 it's pretty well-behaved from ick's point of view 18:26:09 tusho, to integrate into a C program? yes 18:26:09 dunno how it interfaces with c, though 18:26:18 ok then, someone needs to write a sane funge interp in c 18:26:19 :P 18:26:37 ais523, and yes cfunge is quite well behaved compared to ick 18:26:42 tusho, look it is saner than ick 18:26:52 I never do longjmp() or such 18:27:12 AnMaster: well, longjmp() in C mirrors INTERCAL's FORGET perfectly 18:27:12 longjmp() is pretty sane 18:27:15 so how could I not use it? 18:27:19 hah 18:27:24 tusho, not really 18:27:25 __posix_tell_fuzzy_logic_cpu_central_board ... not so much 18:27:44 from man page: 18:27:47 "longjmp() and siglongjmp() make programs hard to understand and maintain. If possible an alternative should be used." 18:27:47 :P 18:27:57 * ais523 begins to wonder if C-INTERCAL could do with a few _posix_fadvises just to annoy tusho 18:28:07 AnMaster: yes, I know that line's in the man page 18:28:09 ais523, not _ in front iirc 18:28:18 AnMaster: ok 18:28:31 yep, no _ 18:28:34 otherwise I got the right name 18:28:34 ais523, just check the #ifdef and such to see if it is supported, for example FreeBSD 6.2 doesn't support it 18:28:49 AnMaster: I'd use autoconf to check if it's supported, it's simpler that way 18:28:59 ais523, that isn't what the standard says 18:29:04 you try to link it, and if it fails, it isn't supported 18:29:04 #if defined(_POSIX_ADVISORY_INFO) && (_POSIX_ADVISORY_INFO > 0) 18:29:06 code here 18:29:07 #endif 18:29:22 would be the correct way according to man posixoptions 18:29:23 :P 18:29:23 and the reason autoconf's useful is that it works even on things that don't obey the standards 18:29:25 which is most of them 18:29:43 ais523, well 1) only POSIX systems will ever define these 18:29:49 AnMaster: what will you be optimizing for next? cpu cache? 18:29:59 tusho: I thought you liked J 18:30:00 tusho, ooh cachegrind from valgrind? 18:30:01 :P 18:30:10 ais523: yes I do and? 18:30:17 well it's optimised for CPU cache 18:30:28 that's not why I like it, though 18:30:34 i like it for its paradigm & conciseness 18:30:37 it's refreshing 18:31:03 ais523, 2) then check using the way I suggested, as that is the correct way according to man page and a freebsd developer I asked 18:31:27 IIRC some systems don't define those even though they have them 18:31:39 ais523, well that is crazy 18:31:51 why support broken systems though? 18:31:55 not really, it's because they have to have C99-compatible headers too 18:32:09 I've done that by mistake simply by using -ansi in a file which was actually POSIX 18:32:10 eh? 18:32:16 hm 18:32:27 it went and turned off support for all the POSIX functions whose protos were in, say, string.h 18:32:32 or other headers that exist in non-POSIX C 18:32:42 ais523, if you write out a binary file and know how long it will be try posix_fallocate() 18:32:53 that does actually have a use: helps against fragmentation 18:32:54 AnMaster: I only write out text files of unknown length 18:32:59 again see _POSIX_ADVISORY_INFO 18:32:59 so not particularly helpful 18:33:01 ais523, ah 18:33:02 why support broken systems though? 18:33:07 why support any systems? 18:33:11 good question! 18:33:20 is it? 18:33:22 tusho: well, CLC-INTERCAL and C-INTERCAL both still support EBCDIC 18:33:32 and that's obsolete 18:33:39 ais523: I was telling AnMaster that all systems are broken 18:33:40 although for C-INTERCAL you need to use a conversion program 18:33:51 ais523, how would the code for C-INTERCAL be compiled on that? you would need to convert to tri-graphs right? 18:34:03 AnMaster: the C source code's ASCII 18:34:12 it's INTERCAL source it accepts in EBCDIC 18:34:14 yes but could it be converted? 18:34:17 but yes, converting to trigraphs is trivial 18:34:27 and any EBCDIC system should be capable of it nowadays 18:34:52 int main(void)?? 18:34:57 "and any EBCDIC system should be capable of it nowadays" 18:34:58 *g* 18:35:11 tusho: really, or they wouldn't be able to run most C 18:36:50 EBCDIC systems still exist? 18:36:59 AnMaster: probably not, but just in case... 18:37:15 dd still handles EBCDIC IIRC 18:37:21 hah 18:37:33 dd does binary doesn't it? 18:37:36 [[ 18:37:38 Each CONV symbol may be: 18:37:38 ascii from EBCDIC to ASCII 18:37:38 ebcdic from ASCII to EBCDIC 18:37:39 ]] 18:37:41 from man dd 18:37:43 oh 18:37:44 yes, it does binary 18:37:50 man recode :P 18:37:51 but also supports ascii/ebcdic conversion 18:37:58 which is binary, I suppose 18:38:06 because it isn't textmode from either system's point of view 18:38:09 at least, not at both ends 18:38:39 ais523: dd should not be doing that... 18:38:44 it should be a seperate program.. 18:38:55 the unix way, etc 18:39:14 tusho, yes like recode or such 18:39:22 dd isn't very unixy 18:39:30 dd if=foo of=blah bs=1234 18:39:31 AnMaster: O RLY 18:39:34 unixy would be 18:39:43 dd -i foo -o blah -b 1234 18:39:45 :P 18:39:52 so not very unixy indeed 18:39:56 AnMaster: dd itself isn't unix 18:40:01 'dd - convert and copy a file' 18:40:04 it is older yes 18:40:06 it has 'and' in the description of what it does 18:40:10 instant unix fail 18:40:31 tusho, ooh an idea: 18:40:46 funtoken - execute commands of the funge 18:40:53 funadd - add operand 18:40:54 and so on 18:41:06 so just one program calling one other program for each opcode! 18:41:07 XD 18:41:11 heh 18:41:14 with data sent over stdio 18:41:20 to handle changes to stack and such 18:41:27 -!- atsampson has quit (clarke.freenode.net irc.freenode.net). 18:41:27 -!- cmeme has quit (clarke.freenode.net irc.freenode.net). 18:41:30 because main program would need to keep track of that 18:41:39 oh no wait 18:41:49 a separate daemon for funge space and one for stack 18:41:51 tusho, XD 18:41:56 hey, cmeme came back! 18:41:57 talking over unix sockets 18:42:28 yes 18:42:30 you said that yesterday 18:42:35 just a netsplit 18:42:45 -!- atsampson has joined. 18:42:45 AnMaster: no, the netsplit caused it to leave 18:42:52 which is when I noticed that it was here in the first place 18:42:54 it was missing for weeks 18:43:02 but must have come back or the netsplit wouldn't have made it leave 18:44:36 oh, I have to go 18:44:39 bye, everyone 18:44:40 -!- ais523 has quit ("(1) DO COME FROM ".2~.2"~#1 WHILE :1 <- "'?.1$.2'~'"':1/.1$.2'~#0"$#65535'"$"'"'&.1$.2'~'#0$#65535'"$#0'~#32767$#1""). 18:44:40 -!- cmeme has joined. 18:45:09 -!- Judofyr has left (?). 18:45:19 -!- Polar has quit (Read error: 104 (Connection reset by peer)). 18:45:55 why does he have to go 18:45:57 he never said 18:48:38 -!- Polar has joined. 18:48:54 who wants to hear about my evil project 18:49:03 depends on what it is 18:49:08 AnMaster: dude, the guy is at uni and has to go to cafes and stuff regularly to get to computers and stuff. 18:49:16 why does it matter? 18:49:21 hm ok 18:49:26 what is your evil project then 18:49:32 i think you've heard it 18:49:38 not sure 18:49:44 yeah 18:49:44 you have 18:49:46 what one is it 18:49:57 you have so many evil projects 18:50:26 ah I see 18:50:35 well yes I heard about that one 18:56:13 -!- sebbu2 has joined. 19:11:53 -!- sebbu has quit (Connection timed out). 19:57:23 -!- RedDak has joined. 20:28:44 Bye all 20:32:29 -!- Sgeo has quit (Remote closed the connection). 20:41:28 -!- timotiis has joined. 20:43:52 tusho, where is ais I wonder 20:43:53 :/ 20:44:23 AnMaster: at home 20:44:31 i guess 21:14:06 -!- puzzlet has quit (clarke.freenode.net irc.freenode.net). 21:14:06 -!- Dewi has quit (clarke.freenode.net irc.freenode.net). 21:15:34 -!- Dewi has joined. 21:15:34 -!- puzzlet has joined. 21:19:03 -!- pikhq has left (?). 22:06:13 -!- sebbu2 has changed nick to sebbu. 22:26:43 -!- cctoide has quit (clarke.freenode.net irc.freenode.net). 22:27:03 -!- cctoide has joined. 22:36:45 -!- sebbu2 has joined. 22:39:28 -!- sebbu2 has changed nick to sebbu3. 22:40:46 -!- sebbu has quit (Nick collision from services.). 22:43:27 -!- sebbu3 has changed nick to sebbu. 22:48:29 Is there anything you can do to force a brainfuck program to halt, apart from letting it run to the end of it's code? 22:48:52 jamesstanley: no 22:48:56 OK, thanks 22:48:57 + - < > , . [ ] 22:48:58 that's it 22:49:01 learn to use it :-P 22:49:17 I was wondering if you can do anything with those that causes an undefined state which would cause it to halt. 22:49:26 jamesstanley: Nope. 22:49:32 OK 22:50:05 jamesstanley: Can I perhaps define each one for you? :P 22:50:17 I know what the commands do 22:50:25 I've written several small programs now 22:50:42 jamesstanley: Then you'd know that the commands have nothing about halting.. 22:50:42 I was just wondering if there was any way to make execution halt without letting it run to the end. Seems not. 22:51:02 Well, decrementing the memory pointer past 0 might make it stop 22:51:06 All sorts of undefined things like that 22:51:18 There are ways to terminate a C program without calling exit. ;) 22:51:40 Anyway, I'm going to bed now. 22:51:44 jamesstanley: generally that wraps it to 255 22:51:47 or whatever the max is 22:51:51 but it's exactly that undefined 22:54:26 -!- sebbu2 has joined. 23:18:31 -!- olsner has joined. 23:18:45 anyone know french? "Madame Camille obtient la crampe de chatte si vous ne mangez pas de tout son fromage." 23:19:20 she contracted a cramp because someone didn't eat all the cheese? 23:21:17 -!- sebbu has quit (Connection timed out). 23:21:40 olsner: hahahahahah 23:23:14 or she *will* get a cramp *unless* you eat all the cheese? 23:26:18 jamesstanley: well, you can do beginning_of_code maybe_halt_code end_of_code ==> beginning_of_code {not halting_cond}[ end_of_code ] 23:27:23 and if you make an interp consider the end of the program an infinite supply of ]'s, you can have a "context-free" way to halt 23:27:32 hmm 23:27:44 actually not that simple in case you're inside a loop when you wanna halt. 23:27:56 sorry, i didn't think that through 23:28:43 jamesstanley: oklopol always talks like this 23:30:59 yeah, he always fails 23:32:02 hah 23:50:56 o 23:54:13 -!- timotiis has quit (Read error: 110 (Connection timed out)). 23:58:21 http://www.www.extra-www.org/ 23:58:57 GregorR: is this your response to no-www 23:59:03 Yes :P 23:59:04 (that's my pre-click prediction) 23:59:06 Yes it is :P 23:59:42 GregorR: you need a compliance checker 23:59:53 'struth :P 23:59:57 a four grade system: