2021-12-18 00:05:38 <dave0       > forth project on hacker news: https://news.ycombinator.com/item?id=29593032     https://github.com/RickCarlino/fbbs2
2021-12-18 03:51:56 <veltas      > joe9: Because if the strings contain same content up to min(n1,n2), then on next access you will check next character which is out-of-bounds
2021-12-18 03:52:14 <veltas      > Offset min(n1,n2)-1 I should say
2021-12-18 03:53:02 <veltas      > And the point of compare is to decide which string comes first, not just whether they are the same string
2021-12-18 05:30:23 <joe9        > http://okturing.com/src/12824/body I am trying to figure out how restore-input here works?
2021-12-18 05:30:31 <joe9        > It seems to expect an n in the stack.
2021-12-18 05:30:43 <joe9        > to do the 5 <>
2021-12-18 05:30:58 <joe9        > but, I cannot figure out who is providing the n
2021-12-18 05:31:06 <joe9        > s/th/that/
2021-12-18 05:31:32 <joe9        > the query when it calls ?restore-input, does not have anything on the stack.
2021-12-18 05:39:04 <neuro_sys_  > What is this from? Looks neat.
2021-12-18 05:56:37 <veltas      > joe9: restore-input is a standard forth word
2021-12-18 05:57:09 <veltas      > Its signature is ( x_n ... x_1 n - flag)
2021-12-18 05:58:48 <veltas      > There is also save-input which has sig ( - x_n ... x_1 n)
2021-12-18 06:16:27 <veltas      > Took me a while to figure out that's a paste website
2021-12-18 06:23:35 <neuro_sys_  > Oh, part of CORE EXT in ANS Forth
2021-12-18 06:23:57 <neuro_sys_  > I almost didn't pay much attention to them, but apparently they're useful in building the core word set.
2021-12-18 06:24:02 <neuro_sys_  > Good to know.
2021-12-18 06:25:02 <veltas      > They're something I don't know if I fully understand how to use, or what their implications are
2021-12-18 06:25:39 <veltas      > Like for example, if I'm parsing inside `include` am I going to be able to restore the line I'm on with it?
2021-12-18 06:26:01 <veltas      > I don't think I could restore the line I'm on at the REPL
2021-12-18 06:26:58 <veltas      > Maybe it's up to the implementation, and the point is just to let you nest contexts of parsing for i.e. `execute` so you can share stuff like `>in`
2021-12-18 06:29:36 <veltas      > neuro_sys_, joe9: They're paired with n>r and nr> by the way, which can send/retrieve a context laid out in that format to/from the return stack
2021-12-18 06:33:43 <veltas      > Sorry I mean `evaluate`, not `execute`
2021-12-18 07:57:20 <neuro_sys_  > Not too relevant, but I do remember reading INCLUDE should allow nesting up to 8 levels.
2021-12-18 07:58:01 <neuro_sys_  > I haven't implemented it yet, but it sounds like you need a stack to store the previous line buffer and file handle stuff.
2021-12-18 07:59:42 <neuro_sys_  > SAVE-INPUT description seem to imply the stack/recursive nature of it.
2021-12-18 08:00:20 <veltas      > Not sure why you prefixed with "Not too relevant", because what you said seems highly relevant
2021-12-18 08:00:31 <neuro_sys_  > Indeed :D
2021-12-18 08:00:50 <veltas      > Not too relevant, but: yes.
2021-12-18 08:00:56 <neuro_sys_  > Haha
2021-12-18 14:26:57 <potatoalienof13> are there any other general-architecture compiled forths besides movforth?
2021-12-18 17:22:40 <MrMobius    > potatoalienof13, how do you mean?
2021-12-18 17:38:35 <potatoalienof13> MrMobious: Forth implementations that compile to machine code, and are not restricted to only one architecture.
2021-12-18 18:44:43 <neuro_sys_  > I suppose there's nothing in the standard to print integers right justified with N digits like %2d in printf?
2021-12-18 18:50:31 <neuro_sys_  > Ah it's doable with <# # #> alright.
2021-12-18 19:26:47 <neuro_sys_  > https://github.com/neuro-sys/advent-of-code-2021/blob/main/4-1.fs.md