2021-12-14 04:13:27 <remexre     > is it just me, or does it seem kinda really annoying to implement CATCH with a top-of-return-stack register?
2021-12-14 05:26:56 <neuro_sys_  > Where can I find the definition of T{ and }T also used in forth-standard.org pages?
2021-12-14 05:27:19 <neuro_sys_  > Nevermind, I found it: https://forth-standard.org/standard/testsuite
2021-12-14 05:30:17 <dave0       > i found an interesting bug in my forth... when evaluating a string with line endings in it, if you put the comment word \ at the end of the line, it skips the newline as whitespace, and then drops the characters to the end of the next line... so \ at the end of the previous line actually comments out the *next* line aswell!
2021-12-14 05:31:50 <dave0       > doesn't happen interactively because there's only 1 line, and it strips the return when inputting
2021-12-14 05:32:05 <neuro_sys_  > I had a similar bug too. Not sure if I fixed it. But it was an off by one issue with PARSE I suppose.
2021-12-14 05:32:24 <neuro_sys_  > I had defined \ as : \ refill drop ; though
2021-12-14 05:33:26 <dave0       > i did   : \ 13 parse 2drop ;
2021-12-14 05:34:08 <neuro_sys_  > Was it so that on linux, line endings are only 10, and on windows it's 10 and 13?
2021-12-14 05:34:35 <neuro_sys_  > I'm developing on linux at the moment, and only check for 10, 13 is pretty much ignored.
2021-12-14 05:34:40 <dave0       > ah no, my parse stops at 10 or 13 so both line endings from linux & windows work
2021-12-14 05:35:16 <dave0       > words just considers 10 and 13 as whitespace
2021-12-14 05:35:20 <dave0       > word
2021-12-14 05:35:49 <dave0       > parse-name to use the standard name
2021-12-14 05:36:15 <dave0       > there was a long discussion on the forth standard page about refill
2021-12-14 05:36:30 <dave0       > someone wanted to slurp a whole file into memory and just evaluate it
2021-12-14 05:38:05 <dave0       > neuro_sys_: i think your method is the correct way
2021-12-14 05:38:30 <neuro_sys_  > I haven't given it much thought yet, I'll have to come back to "white space" problem at some point.
2021-12-14 05:39:37 <dave0       > i spent some time to make it robust against reading outside of the terminal buffer
2021-12-14 05:40:30 <dave0       > i come from c where buffer overruns are endemic
2021-12-14 05:41:37 <neuro_sys_  > I have no memory protection whatsoever, but yeah buffer overruns and off-by-ones are always tricky.
2021-12-14 05:42:07 <dave0       > oy off-by-one .. don't get me started lol
2021-12-14 05:42:23 <neuro_sys_  > When working on the algorithm, drawing on paper with little boxes designating the memory helps a lot.
2021-12-14 05:43:06 <neuro_sys_  > With this Forth, I somehow ended up writing code mostly on paper as it turned out to be more efficient
2021-12-14 05:43:18 <dave0       > cool
2021-12-14 05:43:32 <dave0       > when i get stuck i like to draw a picture
2021-12-14 05:43:56 <neuro_sys_  > Yesterday I wrote do loop on my little notebook on the tram.
2021-12-14 05:44:06 <neuro_sys_  > When I got home and typed it in, it just worked
2021-12-14 05:44:40 <neuro_sys_  > I'm hesitating if I should add support for I, J and K though, as the standard requires it.
2021-12-14 05:45:18 <dave0       > i saw some code where it used I as an index in an array
2021-12-14 05:45:25 <dave0       > in a do loop
2021-12-14 05:47:13 <neuro_sys_  > I really dislike DO LOOP now, it's unnecessarily complex. I'll possibly add FOR NEXT and just keep using it. Actually most of the time I just use BEGIN/WHILE/RPEAT/UNTIL.
2021-12-14 05:48:00 <dave0       > this is why forth is good :-)
2021-12-14 05:49:59 <neuro_sys_  > It is so much fun
2021-12-14 06:11:16 <neuro_sys_  > The output of WORDS: https://pastebin.com/raw/tuBQmmEJ
2021-12-14 06:12:08 <neuro_sys_  > Anything after close-file was defined in a separate ascii source file, but before that is in assembler mixed with hand coded Forth.
2021-12-14 06:21:55 <dave0       > which cpu?
2021-12-14 06:22:51 <neuro_sys_  > x86 for now
2021-12-14 06:23:07 <neuro_sys_  > Ultimately I'd like to port it to z80
2021-12-14 06:23:39 <dave0       > cool
2021-12-14 06:30:18 <dave0       > javascript forth: https://github.com/eatonphil/jsforth
2021-12-14 06:31:25 <dave0       > i don't know javascript but i fumbled my way through a simple program
2021-12-14 06:32:00 <dave0       > i was gonna try and get something that resembles a terminal in the browser
2021-12-14 06:32:24 <dave0       > which that javascript forth has
2021-12-14 06:57:24 <neuro_sys_  > Just in case, there may be already one that's written by someone else as a library, considering it's JavaScript.
2021-12-14 06:57:40 <neuro_sys_  > There's a library for everything :)
2021-12-14 07:38:36 <neuro_sys_  > It seems core.fr and tester.fr contains the ANS Forth test suite: https://github.com/gerryjackson/forth2012-test-suite/tree/master/src
2021-12-14 14:34:19 <neuro_sys_  > This video is fun: https://www.youtube.com/watch?v=1XdgUK1NbpI
2021-12-14 14:34:41 <neuro_sys_  > The guy seems to have coded his recent commercial game in his own Forth
2021-12-14 14:36:09 <neuro_sys_  > This is the game written in Forth: https://www.youtube.com/watch?v=IgDfaBmV3TY
2021-12-14 14:46:43 <MrMobius    > neuro_sys_, looks like that's written in C with scripting in Forth
2021-12-14 14:50:57 <imode       > I thought it was all Forth.
2021-12-14 14:59:31 <MrMobius    > https://webcache.googleusercontent.com/search?q=cache:B1Er_N6-ky4J:https://twitter.com/rocketcatgames/status/800093625547767808+&cd=21&hl=en&ct=clnk&gl=us
2021-12-14 15:06:24 <neuro_sys_  > MrMobius: Ah good find
2021-12-14 15:08:12 <neuro_sys_  > I think Forth is good for game scripting subsystems. I heard Lua is very popular due to ease with interfacing with C/C++ I guess. I suppose Forth would outrun Lua in terms of execution speed.
2021-12-14 15:10:53 <MrMobius    > hmm, maybe. what makes you say that?
2021-12-14 15:11:23 <MrMobius    > forth for a scripting language has the advantage of being very easy to implement. i dont think its about speed
2021-12-14 15:13:09 <neuro_sys_  > It's dynamically typed, but maybe that's also important for ease of development.
2021-12-14 15:15:00 <neuro_sys_  > As for speed, considering many real time games update the world state very often, they'd benefit from speed
2021-12-14 15:18:58 <MrMobius    > games are one of those things where extra speed doesnt really helpas long as your making the timings
2021-12-14 15:19:10 <MrMobius    > at least for game state updates
2021-12-14 17:31:36 <veltas      > Serverless FORTH when?
2021-12-14 17:33:57 <veltas      > Almost certainly LuaJIT would be faster than any FORTH you can embed, unless maybe gforth can be embedded? But then I don't think you could sell the game right?
2021-12-14 17:34:18 <veltas      > Well I mean keep source code proprietary, no rule against 'selling' in GPL per se
2021-12-14 17:34:39 <veltas      > Guys when are we going to get serverless forth?
2021-12-14 18:59:05 <MrMobius    > what does that mean?
2021-12-14 19:10:58 <remexre     > CGI
2021-12-14 19:11:12 <remexre     > CGI in the cloud*
2021-12-14 19:16:53 <crc         > I write web apps using a Forth backend; does that count?
2021-12-14 19:25:58 <imode       > veltas: you can do it on lambda.
2021-12-14 19:27:09 <imode       > AWS lambda, rather.