2022-09-05 00:00:17 <thrig       > deployment I could agree with. but that's not portability
2022-09-05 00:06:15 <selene_     > mmm that's definitely fair
2022-09-05 00:06:25 <selene_     > i had kind of conflated the two in terms of terminology
2022-09-05 00:07:36 <thrig       > https://www.sqlite.org/whyc.html doesn't even mention javascript as a contender (oversight much??)
2022-09-05 00:09:46 <vms14       > ansi C code is very portable
2022-09-05 00:10:17 <vms14       > thrig: do you really think the sqlite author would consider javascript?
2022-09-05 00:10:52 <thrig       > no, because javascript (especially that tied to a bloat browser) is not portable
2022-09-05 00:10:57 <vms14       > node is supposed to be fast, but I don't think someone looking for performance would go for node
2022-09-05 00:11:35 <eris[m]123  > but javascript is portable
2022-09-05 00:11:58 <selene_     > there's also https://bun.sh/ which i only know of through zig
2022-09-05 00:12:09 <eris[m]123  > it runs on linux, windows, macos, freebsd, openbsd, netbsd umodified if attached to a browser
2022-09-05 00:12:23 <thrig       > https://raw.githubusercontent.com/thrig/dotfiles/master/imgs/my-seventh-node-script.png
2022-09-05 00:12:39 <vms14       > and android, iphone, etc
2022-09-05 00:13:16 <vms14       > lol I had the same look for xterm in the past
2022-09-05 00:13:58 <thrig       > Firefox don't compile on 32-bit openbsd. Chrome runs like a turd, if at all, on anything slightly old
2022-09-05 00:14:16 <vms14       > I miss chrome a lot :/
2022-09-05 00:14:48 <eris[m]123  > thrig: but it runs
2022-09-05 00:14:51 <thrig       > 2008 dell with openbsd and an nvidia card? yeah no chrome for you. all my software runs there just fine
2022-09-05 00:15:06 <eris[m]123  > i get the point youre trying to make
2022-09-05 00:15:12 <eris[m]123  > but like
2022-09-05 00:15:23 <eris[m]123  > javascript works for most people
2022-09-05 00:15:42 <thrig       > that's deployment, not portability
2022-09-05 00:17:00 <eris[m]123  > then define your terms
2022-09-05 00:19:46 <thrig       > how many systems the language compiles and runs on. C == lots. Javascript... it's not terrible like some languages, but not really noteworthy compared to lots of other languages
2022-09-05 00:20:28 <thrig       > shackle it with a bloat browser and it edges pretty close to terrible
2022-09-05 00:20:30 <eris[m]123  > brainfuck #1 :)
2022-09-05 00:24:48 <thrig       > node is probably less portable than say pforth, and otherwise might get lost in the luatclperlpythonruby crowd
2022-09-05 00:32:35 <eris[m]123  > is pforth all that portable?
2022-09-05 00:36:53 <selene_     > i mean, it is the name :p
2022-09-05 00:37:49 <eris[m]123  > does it run on plan 9?
2022-09-05 00:41:51 <vms14       > does it run in netbsd
2022-09-05 00:42:25 <vms14       > in netbsd there are only 3 forths
2022-09-05 00:42:43 <vms14       > ficl pforth and retro
2022-09-05 00:43:30 <vms14       > gforth seems to fail to compile lately, but there was a binary, now the 64 bits binary does not exist in the repo, they removed it
2022-09-05 00:43:50 <vms14       > I remember to try to compile gforth with pkgsrc when they removed the binary and failed
2022-09-05 00:44:04 <vms14       > also it was never able to work with the c ffi it has
2022-09-05 02:57:17 <vms14       > I have a square moving on the screen xD
2022-09-05 02:57:18 <vms14       > https://termbin.com/u33p
2022-09-05 02:57:38 <vms14       > I quite dislike the code :/
2022-09-05 18:05:46 <veltas      > Yeah you've got two broad ways of approaching 'macros' in Forth, in the lisp sense.
2022-09-05 18:05:53 <veltas      > Immediate words, and quotations
2022-09-05 18:06:20 <veltas      > And modern forth can't seem to make its mind up which it prefers, well actually it's quite obvious the latest Forth stuff seems to want quotations
2022-09-05 18:06:35 <veltas      > or otherwise the passing of execution tokens
2022-09-05 18:07:08 <veltas      > And old-school forth is all about immediate words and inserting your data, jumps and conditionals inline
2022-09-05 18:07:37 <veltas      > It sucks, it should be consistent
2022-09-05 18:07:45 <veltas      > retro is at least consistent
2022-09-05 18:09:07 <veltas      > Like exceptions for instance, CATCH takes an execution token to execute, this is highly unnatural for forth
2022-09-05 18:09:19 <veltas      > And the code you get using CATCH just doesn't feel very forthy
2022-09-05 18:09:43 <eris[m]123  > catch is ans right?
2022-09-05 18:09:46 <veltas      > Yes
2022-09-05 18:09:49 <eris[m]123  > its always bloody exceptions
2022-09-05 18:10:00 <eris[m]123  > hate exceptions
2022-09-05 18:10:11 <veltas      > The syntax should have been:   TRY .... CATCH
2022-09-05 18:10:26 <veltas      > i.e. just encapsulate the code that runs with TRY and CATCH
2022-09-05 18:10:30 <veltas      > don't pass an xt
2022-09-05 18:11:45 <veltas      > Which is easier to use and more powerful, i.e. you can implement ANS CATCH with these words:  : CATCH TRY EXECUTE CATCH ;
2022-09-05 18:12:36 <veltas      > eris[m]123: What's wrong with exceptions?
2022-09-05 18:13:20 <thrig       > they can be very trying
2022-09-05 18:13:40 <veltas      > IMO exceptions had a really good opportunity to make RDROP trickery easier and portable, and they are a great way to write i.e. recursive descent parsers
2022-09-05 18:14:00 <veltas      > But the ANS exception library is painful and unforthy
2022-09-05 18:30:07 <veltas      > Such is the fate of anything designed by a committee
2022-09-05 18:30:50 <thrig       > the common lisp committee settled on a pretty neat conditions system
2022-09-05 18:40:53 <KipIngram   > I've formalized some return stack mark and restore ability in my system.
2022-09-05 18:41:12 <KipIngram   > As well as some double-return words.
2022-09-05 18:42:28 <KipIngram   > The double returns obviously can only be run under circumstances where the return stack is "ready" for them, but the mark and restore thing will work regardless of what's been done to the return stack in between.
2022-09-05 18:54:48 <veltas      > thrig: Was it invented by the committee or pulled from the wild?
2022-09-05 18:56:24 <thrig       > the design they specified. it could have gone worse. usually it does
2022-09-05 19:00:43 <veltas      > I think Forth would be in a stronger place if Chuck Moore had remained at Forth Inc and carried on steering the ship
2022-09-05 19:00:55 <veltas      > Of course I don't fault him for leaving, it's his business
2022-09-05 19:01:33 <veltas      > Not that he 'ran' forth inc, but obviously his interaction with the business seemed to play a big role in the evolving design of Forth
2022-09-05 19:02:25 <veltas      > And I don't fault the people who took over from him either, rather I would just say Chuck has quite exceptional taste for these things
2022-09-05 19:02:55 <veltas      > If he had focused on the programming side more generally I think we'd have seen some interesting stuff
2022-09-05 19:48:27 <KipIngram   > I think you're right.  Guess he got bitten by the hardware bug.