2021-11-30 02:34:44 <remexre     > has anyone thought about a "superscalar forth"?
2021-11-30 02:35:34 <remexre     > by which I mean, one that makes the superscalar nature of modern CPUs explicit, so it's ergonomic to write code that keeps all the ports (idr the non-Intel term) busy?
2021-11-30 02:44:25 <imode       > you'd be thinking of something like a transport triggered architecture lang.
2021-11-30 02:45:33 <remexre     > how do those handle superscalarness? just by making the ports explicit?
2021-11-30 02:57:43 <imode       > yeah, and the timing.
2021-11-30 04:31:41 <veltas      > logand``: The standard draft doesn't have anything about string literals
2021-11-30 04:32:22 <veltas      > But string literals make some sense, I'm not sure exactly how they work. They don't really need 'recognizers' to build it into a FORTH itself, but to add it to a FORTH something like that would help
2021-11-30 04:33:24 <veltas      > I would tend to advise against using stuff like that when the standard syntax is so simple/easy
2021-11-30 04:33:59 <veltas      > parse-name and parse are more portable for getting strings at interpreter, and s" is portable for compiling
2021-11-30 05:32:12 <veltas      > remexre: I might be totally misintepreting but polyFORTH and similar multi-task FORTHs are certainly able to keep 'ports' busy on one thread
2021-11-30 06:31:39 <ecraven     > if I wanted to try this year's advent-of-code in forth, what's a good implementation on linux (ideally with some sort of Emacs REPL) to do that? gforth?
2021-11-30 06:31:58 <f-a         > people here kind of dislike gforth
2021-11-30 06:32:03 <f-a         > but it offers a lot of convenience
2021-11-30 06:32:08 <veltas      > ecraven: Yes, gforth is probably what you want
2021-11-30 06:32:24 <f-a         > you won’t program for your embedded thingos, but gforth is solid. Don’t know about emacs integration
2021-11-30 06:32:26 <ecraven     > I'd love to use colorforth, but I'm not sure that's a good "introductory" system ;)
2021-11-30 06:32:29 <f-a         > just live in forth repl :P
2021-11-30 06:32:50 <veltas      > I dislike gforth but I think it's a good FORTH to start with, and a good FORTH for many things
2021-11-30 06:33:16 <veltas      > You can't dislike something you don't use, well I don't tend to anyway
2021-11-30 06:33:19 <ecraven     > yea, this is mainly to get *started* and to get better at solving actual problems ;)
2021-11-30 06:33:25 <f-a         > It is well documented, shows some practical tricks (OO in 12 lines of forth) etc.
2021-11-30 06:33:36 <f-a         > but yeah fails on the «ethos» checkbox
2021-11-30 06:33:40 <veltas      > ecraven: gforth is a good starting choice, they have an emacs forth-mode too although I don't know how good it is
2021-11-30 06:34:04 <ecraven     > nice, so I'll just play around with that. is it ok to ask questions about advent-of-code in here a day later than the problem appears?
2021-11-30 06:34:23 <veltas      > It's okay to ask any time
2021-11-30 06:34:38 <veltas      > We don't really care about AoC that much, or spoilers. For us the hard part is FORTH, not the problem itself
2021-11-30 06:35:01 <ecraven     > great, thanks!
2021-11-30 06:35:02 <veltas      > ecraven: Important! Get the latest version of gforth, a late snapshot, don't use your distro's version of gforth
2021-11-30 06:35:10 <veltas      > It's extremely out of date
2021-11-30 06:35:17 <ecraven     > veltas: thanks, I'm on arch linux, that might be up to date
2021-11-30 06:35:21 <veltas      > Nope
2021-11-30 06:35:28 <veltas      > They don't have gforth, if it's in AUR then not sure
2021-11-30 06:35:39 <ecraven     > yea, I have 0.7.3 from AUR
2021-11-30 06:35:45 <veltas      > That's the out of date one
2021-11-30 06:35:55 <veltas      > The best thing to do is keep that installed, and build a new snapshot
2021-11-30 06:36:07 <ecraven     > wow, gforth is developed at tuwien? never knew that
2021-11-30 06:36:10 <veltas      > (because you need gforth installed to build new gforth, I think)
2021-11-30 06:36:25 <ecraven     > just build this? http://www.complang.tuwien.ac.at/forth/gforth/Snapshots/current/
2021-11-30 06:36:29 <veltas      > Yes
2021-11-30 06:36:55 <f-a         > to be honest I am not 100% sure forth is suited to AoC problems
2021-11-30 06:37:03 <f-a         > *but* you will learn plenty
2021-11-30 06:37:09 <veltas      > I've been doing AoC problems in FORTH and I think it's a good exercise
2021-11-30 06:37:17 <f-a         > and dream of  DUP SWAP2 ROT  in your dreams
2021-11-30 06:37:17 <veltas      > Good luck keeping up with the daily releases though
2021-11-30 06:37:35 <veltas      > Yeah I did learn a lot doing it
2021-11-30 06:37:44 <veltas      > I've not finished last year's yet lol
2021-11-30 06:37:56 <ecraven     > yea, the point is mostly to get better acquainted with forth, and the first few days are usually simple enough in any language
2021-11-30 06:38:09 <veltas      > Please go ahead and ask anything you want about it
2021-11-30 06:38:11 <ecraven     > I've done APL a few years ago (and forgot almost everything since ;)
2021-11-30 06:38:43 <veltas      > Well they seem to assume easy I/O so you will end up writing some kind of file parsing library for FORTH
2021-11-30 06:39:15 <ecraven     > yea, they usually want you to read space-separated numbers or something like that
2021-11-30 06:39:28 <veltas      > Because FORTH has read-file and you sort of have to build it up from that, unless you shove the file input into the forth program itself which is an option
2021-11-30 06:40:27 <veltas      > You can modify / re-use the words you make for that as you go, so it should hopefully become less of a barrier as time goes on
2021-11-30 06:40:43 <veltas      > Also consider doing the big boy data sets
2021-11-30 06:40:54 <ecraven     > what are those?
2021-11-30 06:41:21 <veltas      > They're larger input data files that require you to actually employ proper algorithms / data structures
2021-11-30 06:41:48 <ecraven     > hehe, where's the fun in doing things properly :P :D
2021-11-30 06:42:01 <veltas      > It's a challenge if you get bored
2021-11-30 06:43:16 <veltas      > I ended up writing merge sort and binary search in FORTH for one of them, and it's definitely a stress-test of FORTH
2021-11-30 07:06:07 <ecraven     > sorry to ask about gforth specifically, but does anyone know of a way to disable the ansi color output? running it inside a REPL in emacs shows all the escapes that I don't want ;)
2021-11-30 07:12:30 <veltas      > No worries asking about gforth or any forth in here
2021-11-30 07:13:04 <veltas      > Does -status help?
2021-11-30 07:15:13 <veltas      > I mean running the word '-status', not passing as a command-line argument or anything
2021-11-30 07:19:38 <veltas      > ecraven: Run '-status' and 'default-mode'
2021-11-30 07:20:27 <veltas      > default-mode I believe disables the colors, -status removes status bar (which uses terminal escape sequences, of course)
2021-11-30 07:32:10 <veltas      > Looks like Rather's now interacting with the Forth2020 Facebook group
2021-11-30 08:05:12 <ecraven     > thanks!
2021-11-30 08:05:30 <ecraven     > now to find out how to pass these two on the command line... ;)
2021-11-30 08:05:49 <ecraven     > -e it seems ;)
2021-11-30 08:19:10 <veltas      > Yeah my gforth is alias to 'gforth -e -status' (colors don't seem to be default for me, so I don't need to run default-theme either)
2021-11-30 08:20:04 <veltas      > Here is the AOC big boy solution I did with sort https://github.com/Veltas/aoc20/blob/master/1-2-big.fs
2021-11-30 08:20:39 <veltas      > I'm sure I could improve it somehow now, been nearly a year since and I'm always learning more about how to refactor FORTH
2021-11-30 15:41:15 <logand``    > veltas: i found the string literals in some gforth code, i'll stick to s"
2021-11-30 17:07:31 <Zarutian_HTC> s" is supported by most forths no?
2021-11-30 17:08:42 <Zarutian_HTC> and compiles the string into the current word being defined with (str) or dostr prepended
2021-11-30 17:09:47 <Zarutian_HTC> whose runtime effect is to push straddr and strlen onto the stack and skip execution over it
2021-11-30 18:01:29 <veltas      > Zarutian_HTC: Yes