00:00:00 --- log: started forth/04.12.17
03:38:54 --- quit: yumehito ()
03:38:55 --- quit: onetom_ (Read error: 104 (Connection reset by peer))
03:47:56 --- join: Raystm2 (~Ray@adsl-69-149-45-208.dsl.rcsntx.swbell.net) joined #forth
04:07:55 --- join: yumehito (yumehito@giguz.turbo.nsk.su) joined #forth
06:04:11 --- quit: Raystm2 ("User pushed the X - because it's Xtra, baby")
07:27:32 --- join: onetom (~tom@cab.bio.u-szeged.hu) joined #forth
08:40:52 --- quit: modesto (Remote closed the connection)
09:11:44 --- join: arke (apache@11.198.216.81.dre.siw.siwnet.net) joined #forth
09:23:43 --- join: saon (Ecoder@c-24-129-95-254.se.client2.attbi.com) joined #forth
09:27:20 <OrngeTide> immediates on ARM seem tricky to do generically. I can only think of:  mov r0, [pc]! ; .word 1234567
09:28:34 <arke> i dont know arm, but if its a RISCy-like, you can also do
09:28:42 <arke> xor r0, r0, r0
09:28:55 <arke> ori r0, 1234
09:28:59 <arke> shl r0, 16
09:29:08 <arke> ori r0, 5678
09:29:23 <arke> ...except that that is hell on the pipeline :)
09:29:31 <OrngeTide> mov r0, #255  is easy though. immediates are kinda wierd. : imm ( op shift -- imm ) 2* shr ; .. or something like that.
09:29:39 --- join: qFox (C00K13S@82-169-140-229-mx.xdsl.tiscali.nl) joined #forth
09:29:45 <OrngeTide> arke, well i can do the shift in the immediate...
09:30:04 <arke> aah.
09:30:09 <arke> qFox: c00k13s!
09:30:32 <qFox> yo
09:30:34 <qFox> wazap
09:30:54 <OrngeTide> arke, i can't actually do:  ori r0, #1234 ... #1234 is too big
09:31:29 <OrngeTide> i have 8 bits for my immediate. and 4 bits for shift/rotate. (0-31, in multiples of 2)
09:31:44 <arke> OrngeTide: is it a 32 bit or 16 bit processor?
09:31:49 <OrngeTide> 32-bit. this is ARM
09:31:56 <arke> hrm.
09:32:08 <OrngeTide> branch instructions have 24-bits for immediates. hehe
09:32:09 <arke> mov r0, 12<<4
09:32:17 <arke> ori r0, 34<<2
09:32:18 <arke> etc.
09:32:19 <arke> :)
09:32:27 <arke> hehe
09:32:35 <OrngeTide> yea. it would take 5 instructions to do that.
09:32:42 <OrngeTide> wait. i could do it in 4. 
09:32:46 <OrngeTide> (or takes 3 args)
09:33:16 <OrngeTide> or r0, r0, 34 ROR 24 
09:33:19 <OrngeTide> or something like that.
09:34:07 <arke> what a pain :)
09:34:35 <OrngeTide> i wonder if the pc trick will work on arms with harvard cache. if it does and doesn't blow on performance i should use that.
09:34:54 <arke> yeah.
09:35:00 <OrngeTide> mov r0, [pc]!  would read from memory at pc and increment pc.
09:35:14 <arke> neat.
09:35:24 <arke> I'm sure some C compilers do it like that :)P
09:35:29 <OrngeTide> i know on intels mixing code and data is a no no. 
09:35:38 <OrngeTide> arke, oh. that's a good idea. i should just see what gcc generates!
09:36:43 <arke> :)
09:37:02 <arke> yeah, on intels it'll flush the cache after every instruction if you do. thats bad. I think you can turn that off though.
09:38:06 <OrngeTide> oops. it was ldr not mov
09:38:12 <OrngeTide> i'm retarded. :P
09:38:46 <OrngeTide> mov can't do what i was just talking about. ehhee.
09:38:53 <arke> muahahahaaaaa
09:38:53 <OrngeTide> (i'm still learning)
09:38:54 <arke> :)
09:39:03 * arke tihnks
09:39:23 <arke> you know, if you don't mind the complexity, you should add a compiler-time instruction queue
09:39:36 <OrngeTide> ldr has 12-bits for it's immediates.
09:39:54 <arke> oh neat. that means you can do it in 3 steps
09:39:58 <OrngeTide> mov can load 8-bit immediates and shift them like i said. but to read from pc i use ldr r0, [pc]!
09:40:36 --- join: ASau (~user@83.102.133.66) joined #forth
09:40:39 <OrngeTide> well it still takes 4 steps with mov. and 1 step with ldr
09:40:44 <arke> hehe.
09:40:47 <arke> hi arke 
09:40:49 <arke> err
09:40:50 <arke> hi ASau 
09:40:55 <OrngeTide> by 12-bit immediate i mean a 12-bit offset from pc.
09:41:01 <ASau> Dobry vecer, ASau!
09:41:20 <ASau> I mean
09:41:26 <arke> OrngeTide: neat
09:41:27 <OrngeTide> but i'd be using the register w/ update mode of ldr not the immediate address mode.
09:41:32 <ASau> dobry vecer, arke!
09:41:39 <arke> ASau: terve :)
09:42:25 <OrngeTide> arke, what's really neat is ldm .. it lets me load 1 to 15 registers (using a bitmask). handy if i want to load double words or quad words.
09:43:12 <OrngeTide> and it has the same semantics, i thin. so ldm pc!, {r0-r7}  .. would load up 64 bytes at pc and update pc.
09:44:10 <arke> OrngeTide: you can have numbers representing how long ago a certain register was locked. Then, after each word interpreted or compiled, you go insert one instruction, which would be the instruction whos register has the longest lock time
09:44:14 <arke> OrngeTide: thats neat :)
09:48:26 --- quit: crc2 (Read error: 104 (Connection reset by peer))
09:50:24 <OrngeTide> locked?
09:50:57 <arke> OrngeTide: yeah.
09:51:39 <arke> OrngeTide: basically, you keep a count for each register. when it is used as a writing target, you set its lock count to 1
09:51:52 <arke> OrngeTide: after each iknstruction compiled, you increment the lock count for every register unless its 0
09:54:13 <arke> OrngeTide: when you encounter something like dup, you look for instructions that have the highest lock count for its registers
09:54:29 <arke> OrngeTide: then, you set the lock counts to 0 and compile the instruciton
09:54:59 <arke> OrngeTide: if you jump, you need to flush, and for each read, you need to make sure that there are no previous writes in the queue.
09:55:19 <arke> OrngeTide: it can probably be done in 2-3 screens of forth code
10:07:57 --- quit: arke ("CGI:IRC (EOF)")
10:08:07 --- join: ows (~ows@gtDEI-NATgw.dei.uc.pt) joined #forth
10:19:45 --- quit: saon ("Leaving")
10:22:37 --- join: saon (1000@c-24-129-90-197.se.client2.attbi.com) joined #forth
10:44:53 --- join: fridge (~fridge@dsl-220-253-77-81.NSW.netspace.net.au) joined #forth
10:54:43 --- quit: ows ("Client Exiting")
11:00:26 --- join: Herkamire (~jason@h000094d30ba2.ne.client2.attbi.com) joined #forth
11:37:22 --- join: arke (apache@11.198.216.81.dre.siw.siwnet.net) joined #forth
11:40:38 --- quit: arke (Client Quit)
11:46:29 --- join: arke (apache@11.198.216.81.dre.siw.siwnet.net) joined #forth
12:38:04 --- join: Novocaine (~RalphL@213-48-226-28.hay.cvx.blueyonder.co.uk) joined #forth
12:40:44 --- part: Novocaine left #forth
12:52:29 --- quit: arke ("CGI:IRC (EOF)")
13:08:38 --- join: arke (apache@11.198.216.81.dre.siw.siwnet.net) joined #forth
13:18:05 --- quit: arke ("CGI:IRC (EOF)")
13:37:01 --- quit: qFox ("this quit is sponsored by somebody!")
13:40:19 --- join: qFox (C00K13S@82-169-140-229-mx.xdsl.tiscali.nl) joined #forth
13:47:49 --- join: ows (~ows@83.132.99.46) joined #forth
13:48:16 --- join: MacCall (~chatzilla@nas-cbv-5-213-228-48-117.dial.proxad.net) joined #forth
13:48:23 <MacCall> hello
14:00:53 <MacCall> i'm working on a bytecode translator from mips to x86
14:01:43 <MacCall> i translate the mips asm to another language before compiling it to the x86
14:02:21 <MacCall> i'm just want to know you advice about using forth for the translation
14:03:03 <MacCall> i mean : asm code use a lot the stack pointer an the asm translation is a little difficult
14:03:28 <MacCall> but forth base seems to be the stack utilisation
14:04:01 <MacCall> so have you ever heard about a project like mine using forth ?
14:10:10 <Herkamire> sounds like something forth would be good at
14:12:41 <Herkamire> why are you translating twice? do you plan on expanding it later to translate to/from other architectures?
14:12:54 --- join: arke (apache@11.198.216.81.dre.siw.siwnet.net) joined #forth
14:15:10 <MacCall> the 1st pass is a translation to the asm code to an xml presentation because i'm not already fixed to the output language after "bytecode" analyze (to C ? to forth ? other ?)
14:15:28 --- join: SDO (~SDO@chcgil2-ar7-4-34-143-217.chcgil2.dsl-verizon.net) joined #forth
14:15:40 <MacCall> the 2nd pass is to "compile" the output langguage to the x86 host
14:16:01 <Herkamire> xml? are you mad?
14:16:39 <MacCall> xml is just a "first" tools to experiment language choice
14:16:56 <MacCall> don't worry ;)
14:17:40 <Herkamire> sorry, xml is a little pet peeve
14:18:37 <MacCall> i use xml with apache digester in java because the apache digester acting as a "stack runtime" when analysing xml
14:18:38 <Herkamire> xml is difficult to read and write by both humans and computers
14:19:03 <Herkamire> dear god
14:19:16 <MacCall> lol
14:19:29 <Herkamire> what are you writing?
14:19:48 <Herkamire> or are you just tinkering with things to make the most complex system possible?
14:20:26 <Herkamire> (I have this (sometimes annoying) tendancy to assume that people are trying to write something useful)
14:20:38 <Herkamire> when in fact often people are just tinkering/exploring trying to learn something
14:21:25 <MacCall> my xml solution i just a temporary solution to translate mips code >> xml >> C language >> gcc >> x86 asm
14:21:41 <Herkamire> hehe
14:21:59 <MacCall> it's works for now but not for all the operation code implemented in mips32
14:22:36 <Herkamire> heh, that's just so very different from what I imagined
14:22:51 <MacCall> ^^
14:23:10 <Herkamire> I was figuring on translating directly from mips instructions to x86 instructions
14:23:33 <MacCall> my xml choice is because i'm planning to use forth code in C language place
14:23:55 <Herkamire> how does xml help with forth?
14:24:00 <MacCall> ok, i understand why you were scared ^^
14:24:05 <Herkamire> do you have forth code to deal with XML?
14:25:09 <Herkamire> what's your goal? you writing something that will translate a mips binary into an x86 binary?
14:25:10 <MacCall> from xml i can write a translator to C Language or to the forth language
14:25:31 <Herkamire> oh, sorry
14:25:31 <MacCall> yes
14:25:41 <Herkamire> I thought you meant a forth program would interpret the XML
14:26:24 <MacCall> my english is a little ugly :)
14:26:36 <Herkamire> not that, I just assumed or something
14:27:13 <Herkamire> you occationally leave out little useless words, but your meaning is always clear I think.
14:27:41 <MacCall> thank you ^^
14:27:49 <Herkamire> your english is quite good
14:28:15 <Herkamire> do you have a forth system that compiles to an x86 binary?
14:28:29 <MacCall> no
14:28:41 <Herkamire> what OS are you working on?
14:29:04 <MacCall> windows/cygwin or linux
14:29:10 <Herkamire> how much do you care about the how fast the resulting binary runs?
14:29:12 <MacCall> i've planed to use gforth
14:29:40 <MacCall> a first step is make the project run
14:30:32 <Herkamire> of course, but if speed is a high priority, it can effect these decisions
14:30:39 <MacCall> i'm gonna looking for optimisation ^^
14:30:55 <Herkamire> afaik there is no forth compiler that can compare with gcc as far as optomization.
14:30:56 <MacCall> speed is high priority yes
14:31:28 <MacCall> perhaps i'm gonna use forth only on some parts that use stack pointer a lot
14:32:19 <Herkamire> what's the connection between forth and stack pointer manipulation?
14:33:51 <MacCall> some time i can't "calculate" how many args are on the stack 
14:34:18 <MacCall> so i can't extract the informations i need to construct a C function
14:35:43 <Herkamire> uhuh
14:36:29 <Herkamire> I'm trying to think if it would even be possible to translate asm into C
14:37:07 <Herkamire> suppose you could
14:37:23 <Herkamire> but it seems like you'd just be creating all sorts of problems for yourself
14:37:43 <MacCall> what kind of problems ?
14:39:09 <Herkamire> like trying to rework branches and calls and such into controll constructs such as if/else for while goto switch etc
14:39:24 <Herkamire> even just figuring out where the end of functions are
14:40:51 <MacCall> for extracting function, it's 99% of time a "jal ra" opcode and for branch case/if are similare and for/while too
14:41:06 --- quit: saon ("Leaving")
14:43:57 <Herkamire> are these binaries generated from C code?
14:44:30 <Herkamire> (the mips ones)
14:44:30 --- quit: arke ("CGI:IRC (EOF)")
14:45:35 <MacCall> yes (but not all)
14:46:13 <MacCall> some routine seems to be written directly in mips asm but they easily identifiable
14:46:14 <Herkamire> mmm... I think you could do the last phase in C and get gcc to do major optomising for you
14:46:55 <Herkamire> you could make a global for each register
14:47:04 <Herkamire> and have gcc figure out what it'll keep in registers
14:47:16 <Herkamire> hmmm... x86 doesn't really have many registers though
14:48:42 <MacCall> it's why i think to but i went in this chan to know if the forth language can offer me some help... what do you think ?
14:50:03 <Herkamire> afaik there is very little in the way of serious optomizing compilers for forth
14:50:29 <Herkamire> I don't think forth is well suited to replace C in your diagram
14:50:49 <Herkamire> but forth would probably be quite good at translating mips instructions directly to C code
14:51:29 <Herkamire> or directly to x86 instructions if you aren't so concered about efficiency
14:51:49 <MacCall> it's what i think to
14:52:28 <MacCall> perhaps i'm gonna replace my "xml >> java digester" toolchain by forth in the futur ^^ 
14:52:44 <Herkamire> what's a digester?
14:54:36 <MacCall> it's a sort of xml parser that can instantiate/push/pop java object when reading a tag. To make this work, you just have to give the xml file and the tagRules
14:55:14 <MacCall> i think that principe i very close to forth programmation
14:55:19 <MacCall> but i can be wrong
14:55:32 <MacCall> because i'm a little newbee in forth ^^
14:55:45 <Herkamire> not sure I follow
14:56:04 <Herkamire> xml parser that turns the xml document into a bunch of nested java objects?
14:56:36 <MacCall> some thing like that
14:57:22 <MacCall> for info : http://jakarta.apache.org/commons/digester/
14:57:29 <Herkamire> I don't see any connection between that and forth programming
14:58:04 <MacCall> perhaps i'm wrong
14:58:56 <MacCall> sorry but i've to go
14:59:08 <MacCall> thank you for your precious advices
14:59:27 <Herkamire> np
15:00:49 --- quit: MacCall ("no reason")
15:19:33 <Herkamire> I don't know what java people see in forth
15:19:43 <Herkamire> forth and java seem like total oposites to me
15:29:45 <crc> hi all
15:35:57 --- quit: ows (Excess Flood)
15:36:20 --- join: ows (~ows@83.132.99.46) joined #forth
15:49:52 <OrngeTide> hi herk
15:50:59 <OrngeTide> herk, http://www.forth.org/svfig/next.html .. mentions HerkForth
16:20:13 --- join: Sonarman (~snofs@adsl-64-169-94-230.dsl.snfc21.pacbell.net) joined #forth
16:20:55 <crc> OrngeTide, it's interesting that they say that retroforth is a descendent of herkforth on that page
16:21:25 * crc wishes there was contact information for the person doing that topic; I'd let him know
16:22:16 <OrngeTide> crc, yea. i thought that was kind of strange.
16:22:37 --- quit: Sonarman (Client Quit)
16:23:56 <OrngeTide> crc, jaffe at roses dot stanford dot edu
16:25:36 <crc> thanks OrngeTide
16:31:52 <crc> Herkamire, how long have you been working on herkforth?
16:32:04 <crc> (Just out of curiosity)
16:45:05 <OrngeTide> i'm finally breaking down and writing a forth. actually i'm writing a forth cross assembler. then i'll write a metacompiler that uses the cross assembler. i think
16:45:53 <OrngeTide> i found out eforth and others don't have DOES> .. my friend said there is some "controversy" about DOES> .. i assumed every proper forth had DOES> .. are people actually against using it?
16:46:28 <crc> I plan to eliminate does> in RetroForth 8.0
16:47:00 <OrngeTide> really? why?
16:47:10 <crc> it makes the code more confusing
16:47:19 <OrngeTide> crc, is there any relationship between retroforth and herkforth at all?
16:47:50 <OrngeTide> that whole paragraph on svfig confuses me. cuz herkforth is a colorforth, i didn't think retroforth was a colorforth
16:47:53 <crc> I like some of the words in herkforth, and I've run it quite a bit (on a shell account and with qemu)
16:48:04 <crc> retroforth is like a punctuated colorforth in design
16:48:09 <OrngeTide> oh i see
16:48:15 <OrngeTide> i should look into it sometime.
16:49:09 <crc> :)
16:49:27 --- join: Novocaine (~RalphL@213-48-225-219.hay.cvx.blueyonder.co.uk) joined #forth
16:50:11 --- quit: Novocaine (Client Quit)
16:50:38 <OrngeTide> hrm. arke isn't around. i was going to tell him i figured out how to do big literals on arm asm.
16:53:52 --- quit: qFox ("this quit is sponsored by somebody!")
16:56:24 <crc> do you know any decent arm assemblers?
16:59:18 * crc dislikes GAS
17:04:48 <OrngeTide> crc, there is that goldroad one. and the one that comes with that BASIC for risc os is nice too.
17:05:02 <OrngeTide> but otherwise there are no good arm assemblers. which is why i'm writing one.
17:05:21 <OrngeTide> also i don't believe in using a c based assembler to write forth when writing an assembler in forth is easy.
17:07:29 <OrngeTide> well i got some work stuff to take care of. ttyl.
17:19:09 * crc wonders how to handle constants without does> 
17:22:56 --- join: crc2 (~crc2@pool-70-20-138-220.phil.east.verizon.net) joined #forth
17:34:27 --- join: Sonarman (~snofs@adsl-64-160-166-19.dsl.snfc21.pacbell.net) joined #forth
17:35:34 <crc> got it :)
17:35:38 <crc> : constant create -5 allot m: literal m: ;; ;
17:36:07 <crc> instead of:
17:36:17 <crc> : constant create , does> @ ;
17:36:40 <OrngeTide> i'd rather have does>
17:36:46 <crc> why?
17:36:51 <OrngeTide> it's readable
17:37:19 <OrngeTide> it's the simple solution to a lot of the core words.
17:37:27 <crc> such as?
17:37:34 <OrngeTide> such as constant and variable
17:37:57 <crc> does> is not a simple word though
17:38:20 <OrngeTide> the problem is you can either shove the complexity in does> or smear the complexity throughout your entire program.
17:38:30 <crc> constant was the only word in RetroForth that used does>
17:38:50 <OrngeTide> some of my object oriented junk i was experimenting with use does>
17:39:00 <crc> I don't do OO :)
17:39:10 <OrngeTide> forth itself is OO to begin with. 
17:39:23 <crc> hmm
17:39:25 <OrngeTide> at least the forth dictionary is a very restricted sort of object
17:39:37 <crc> fair enough
17:39:56 <crc> but forth as OO isn't what most people would consider OO these days
17:40:20 <OrngeTide> well c++ people argue with me that objc isn't OO. which is just wrong.
17:40:22 <fridge> namespaces and really_long_function_calls
17:40:45 <OrngeTide> yea. that has nothing to do with actual OO. it's just a feature in a couple of languages marketed as OO
17:41:12 <OrngeTide> fridge, well the purpose of namespaces is you don't ened really_long_function_calls. it's kind of sad how C++ does both.
17:42:01 <OrngeTide> well i'm going to try and sneak out of work so i can go home and play with my free tivo
17:42:20 <crc> :)
18:09:51 --- part: SDO left #forth
18:14:54 --- join: SDO (~SDO@chcgil2-ar7-4-34-143-217.chcgil2.dsl-verizon.net) joined #forth
18:33:40 --- join: saon (1000@c-24-129-90-197.se.client2.attbi.com) joined #forth
18:55:52 --- quit: SDO (Read error: 110 (Connection timed out))
19:07:53 --- join: arke (apache@11.198.216.81.dre.siw.siwnet.net) joined #forth
19:24:12 --- quit: saon (Remote closed the connection)
19:25:29 <madgarden> OrngeTide, WTF::IsWrongWithReally.LongCPlusPlus.Names::Huh()?
19:26:06 <arke> madgarden: hey. :)
19:26:13 <arke> working on it, dude :)
19:31:39 <madgarden> Hey.
19:31:44 <madgarden> Wrong window, dude. ;)
19:32:52 <arke> :)
19:33:08 <madgarden> :D
19:35:22 --- quit: arke ("CGI:IRC")
19:39:34 --- join: saon (1000@c-24-129-90-197.se.client2.attbi.com) joined #forth
20:31:45 <Herkamire> oh good :) you're still on
20:31:46 <Herkamire> I'm back
20:31:57 <Herkamire> very wierd they list retroforth as a herkforth descendant
20:32:14 <Herkamire> I started herkforth about a year and a half ago I think
20:32:27 <Herkamire> I asume retroforth is older
20:35:59 <Herkamire> crc: what's the status on retroforth 7.x running under bochs?
20:38:33 <Herkamire> actually, the linux port might run really well under qemu
20:42:21 <Herkamire> mmm... 1998... yeah, that was before I started herkforth
20:56:16 --- quit: ows ("Client Exiting")
21:07:08 <OrngeTide> hi
21:07:17 <OrngeTide> finally got my stupid tivo to work
21:12:12 <OrngeTide> how come people at svfig don't know anything?
21:13:28 <Herkamire> retroforth was started in 98, herkforth in 03
21:14:30 <Herkamire> yessssss
21:15:51 <Herkamire> I fixed my attractor
21:53:36 <madgarden> attractor?
22:05:52 <Herkamire> coool :)
22:05:58 <Herkamire> check this out: http://jason.herkamire.com:8888/mangledmandelbrot.png
22:06:17 <Herkamire> I've been fiddling with fixed point math (and thoroughly confusing myself :))
22:06:32 <Herkamire> I decided to test the math by running my little mandelbrot generator
22:08:08 <Herkamire> then I thought "how can I know that the mandelbrot will look different if the fixed point math doesn't work quite right" so I tried running the mandelbrot with the fixed point math multiply I was working on set to do 15.17 (while the rest of the mandelbrot code presumes 16.16)
22:08:17 <Herkamire> above is the result. looks cool :)
22:08:45 <madgarden> Heh, funky.
22:08:53 <madgarden> Can't really spot what's wrong though. ;)
22:11:26 <Herkamire> here's what it normally looks like: http://jason.herkamire.com:8888/mandelbrot.png
22:12:25 <Herkamire> notice how the colors cut each other off in the mangled
22:12:56 <Herkamire> and in the normal one each color makes a full continuous ring around the shape (however conveluted)
22:13:27 <madgarden> Ahh.
22:13:28 <Herkamire> and of course the set appears a lot bigger
22:14:07 <Herkamire> an attractor is a function that you can keep shoving a few variables through and plotting them on the screen each time
22:14:11 <Herkamire> and it looks cool
22:16:12 <Herkamire> eg, here's the one I made with herkforth:
22:16:18 <Herkamire> http://jason.herkamire.com:8888/attractor.png
22:17:25 <madgarden> That's neat.
22:27:34 <Herkamire> I just discovered that my sin() code takes input with 17 bits after the decimal point, and gives output with 16
22:27:48 <Herkamire> I'm don't know if I did this on purpose
22:29:27 <Sonarman> does your sin() use a table? (just curious how you chose to implement it)
22:30:03 <Herkamire> mmmm... take that back
22:30:10 <Herkamire> input to sin has nothing to do with radians
22:30:22 <Herkamire> table. with interpolation
22:31:33 <Sonarman> ok
22:31:40 <Herkamire> apparently when I wrote the attractor code I forgot what sin() takes as input
22:33:46 <Sonarman> does your sin() just take an angle such that 0 = 0 and $FFFFFFFF (or whatever the maximum is with your fixed-point system) = 2pi?
22:34:01 <Sonarman> i.e., the range of possible angle measures scaled to the word size
22:34:26 <Herkamire> 0..$c9000
22:34:37 <Herkamire> I haven't the foggiest why I chose 0xc9000
22:34:48 <Sonarman> ok, then i won't bother asking why :)
22:34:57 <Sonarman> how large is your table?
22:35:41 <Sonarman> and did you enter your table by hand (looking at another table), or did you have it computed for you?
22:37:12 <Herkamire> this just get's weirder and weirder... looks like my table has 202 entries
22:37:36 <Herkamire> I wrote a C program to spit out the table
22:39:07 <Sonarman> using C sin()?
22:40:27 <Herkamire> yeah
22:40:36 <Herkamire> with doubles probably
22:40:50 <Sonarman> ok
22:40:53 <Herkamire> this sin code makes no sense
22:41:04 <Sonarman> well, it still results in cool images :D
22:41:13 <Sonarman> and that's all that matters, right? :)
22:42:24 <Herkamire> oh, I know
22:42:34 <Herkamire> ok 202 is correct I guess
22:42:57 <Herkamire> it's just a table of the first pi/2 radians (90 degrees)
22:43:56 <Herkamire> 16 bits unsigned
22:53:51 <Herkamire> ok, what I haven't calculated what value one radian is in this scale
22:56:16 <Sonarman> that would be... $c900/(pi/2), right?
22:56:58 <Herkamire> yeah
22:57:59 <Herkamire> well, $c9000/(pi*2)
22:59:01 <Sonarman> oh, right
22:59:13 <Sonarman> but in your table you only go up to pi/2, right?
22:59:36 <Sonarman> i mean, the equivalent of it under your scaling
23:01:38 <Herkamire> yeah, but $c9000 is 2 pi
23:01:48 <Sonarman> yeah, i see now
23:03:01 <Herkamire> humph
23:03:06 <Herkamire> one radian is $1ffd8
23:03:26 <Herkamire> which is so close to 2 in my 16.16 fixed notation that it's not funn
23:03:27 <Herkamire> y
23:03:33 <Herkamire> I was probably aiming for 1 and got the math wrong
23:04:34 <Sonarman> hehe
23:14:27 <Herkamire> ok, fixed
23:18:49 --- quit: fridge (sterling.freenode.net irc.freenode.net)
23:18:49 --- quit: yumehito (sterling.freenode.net irc.freenode.net)
23:18:49 --- quit: I440r (sterling.freenode.net irc.freenode.net)
23:18:50 --- quit: saon (sterling.freenode.net irc.freenode.net)
23:18:50 --- quit: Herkamire (sterling.freenode.net irc.freenode.net)
23:18:50 --- quit: OrngeTide (sterling.freenode.net irc.freenode.net)
23:18:50 --- quit: warpzero (sterling.freenode.net irc.freenode.net)
23:18:50 --- quit: ianp (sterling.freenode.net irc.freenode.net)
23:18:50 --- quit: Sonarman (sterling.freenode.net irc.freenode.net)
23:18:50 --- quit: crc2 (sterling.freenode.net irc.freenode.net)
23:18:51 --- quit: crc (sterling.freenode.net irc.freenode.net)
23:18:51 --- quit: madgarden (sterling.freenode.net irc.freenode.net)
23:18:51 --- quit: Fractal (sterling.freenode.net irc.freenode.net)
23:18:51 --- quit: cmeme (sterling.freenode.net irc.freenode.net)
23:18:52 --- quit: onetom (sterling.freenode.net irc.freenode.net)
23:19:24 --- join: crc (~crc@pool-70-20-138-220.phil.east.verizon.net) joined #forth
23:19:24 --- join: saon (1000@c-24-129-90-197.se.client2.attbi.com) joined #forth
23:19:24 --- join: Sonarman (~snofs@adsl-64-160-166-19.dsl.snfc21.pacbell.net) joined #forth
23:19:24 --- join: crc2 (~crc2@pool-70-20-138-220.phil.east.verizon.net) joined #forth
23:19:24 --- join: Herkamire (~jason@h000094d30ba2.ne.client2.attbi.com) joined #forth
23:19:24 --- join: fridge (~fridge@dsl-220-253-77-81.NSW.netspace.net.au) joined #forth
23:19:24 --- join: onetom (~tom@cab.bio.u-szeged.hu) joined #forth
23:19:24 --- join: yumehito (yumehito@giguz.turbo.nsk.su) joined #forth
23:19:24 --- join: OrngeTide (orange@rm-f.net) joined #forth
23:19:24 --- join: warpzero (~warpzero@wza.us) joined #forth
23:19:24 --- join: cmeme (~cmeme@216.184.11.2) joined #forth
23:19:24 --- join: I440r (~mark4@216-110-82-59.gen.twtelecom.net) joined #forth
23:19:24 --- join: madgarden (~madgarden@Ottawa-HSE-ppp4082712.sympatico.ca) joined #forth
23:19:24 --- join: ianp (~ian@inpuj.net) joined #forth
23:19:24 --- join: Fractal (jah@selling.kernels.to.linus.torvalds.at.hcsw.org) joined #forth
23:19:24 --- mode: irc.freenode.net set +o crc
23:20:36 --- quit: I440r (sterling.freenode.net irc.freenode.net)
23:20:36 --- quit: fridge (sterling.freenode.net irc.freenode.net)
23:20:36 --- quit: yumehito (sterling.freenode.net irc.freenode.net)
23:21:27 --- join: fridge (~fridge@dsl-220-253-77-81.NSW.netspace.net.au) joined #forth
23:21:27 --- join: yumehito (yumehito@giguz.turbo.nsk.su) joined #forth
23:21:27 --- join: I440r (~mark4@216-110-82-59.gen.twtelecom.net) joined #forth
23:22:57 --- quit: onetom (sterling.freenode.net irc.freenode.net)
23:23:25 --- join: onetom (~tom@cab.bio.u-szeged.hu) joined #forth
23:59:59 --- log: ended forth/04.12.17