2023-06-12 10:31:31 <KipIngram   > So, after reflecting on things, I think what I like is the idea of a vocabulary being an entirely distinct thing - occupy its own memory region(s), etc.  And I'd like to be able to say <vocabulary name> DISCARD, and have that all freed up.  That way I can load tools and use them and then not keep them around.
2023-06-12 10:32:26 <KipIngram   > These screen-like windows that represent different interactions I'll call "sessions."  A session will get it's on main thread, it's own vocabulary, it's own console.
2023-06-12 10:32:46 <KipIngram   > I'll have some way of switching among them.
2023-06-12 14:31:48 <drakonis    > https://compilercrim.es/bootstrap/miniforth/
2023-06-12 14:33:17 <KipIngram   > Ah, a "sector" Forth.
2023-06-12 14:34:19 <drakonis    > yes
2023-06-12 14:34:27 <drakonis    > its part of a series
2023-06-12 14:34:49 <drakonis    > it builds up towards a larger forth
2023-06-12 14:36:05 <thrig       > voltronforth when?
2023-06-12 14:36:19 <drakonis    > NOW
2023-06-12 14:41:09 <drakonis    > https://board.flatassembler.net/topic.php?t=20690 this listing thing is pretty great, dont need to whip out the disassembler all the time
2023-06-12 14:44:19 <drakonis    > look for output.lst btw
2023-06-12 14:44:31 <KipIngram   > We had a link to one of those a while back that was pretty nifty.  It started out executing one-character word names - you just fed it a bizarre looking byte stream.  Few lines of that and all of a sudden you came to
2023-06-12 14:44:38 <KipIngram   > ( now we can include comments!)
2023-06-12 14:44:52 <KipIngram   > And it was just on from there, toward something more "proper."
2023-06-12 14:44:58 <drakonis    > that's rad.
2023-06-12 14:45:08 <KipIngram   > Yeah, I was pretty tickled by it.
2023-06-12 14:45:16 <KipIngram   > Worked on one of my own for a little while in there.
2023-06-12 14:45:25 <drakonis    > doing assembly is tickling my brain more than my dayjob
2023-06-12 14:45:33 <drakonis    > i'm doing that instead of my dayjob right now lol
2023-06-12 14:45:33 <KipIngram   > I was trying to get one that would use the old BIOS calls to get the start-up bytes.
2023-06-12 14:45:45 <thrig       > get back to work!
2023-06-12 14:45:47 <KipIngram   > I get like that at times.
2023-06-12 14:46:44 <drakonis    > work sux
2023-06-12 14:46:53 <drakonis    > i want to get hired to do something more exciting
2023-06-12 14:48:11 <thrig       > gator wrestling?
2023-06-12 14:49:59 <drakonis    > that would be life threatening
2023-06-12 14:50:36 <thrig       > exciting things usually are
2023-06-12 14:51:22 <unjust      > KipIngram: was that planckforth?
2023-06-12 14:59:27 <KipIngram   > I think that's right, yeah.
2023-06-12 14:59:52 <KipIngram   > Oh, that's right - and we SEPARATELY were looking at sectorForth.
2023-06-12 14:59:56 <KipIngram   > About the same time.
2023-06-12 15:00:01 <KipIngram   > I was trying to mash them up.
2023-06-12 15:00:49 <KipIngram   > Tinkering with a sector Forth that worked like planck Forth.
2023-06-12 15:10:36 <trev        > "haha you can write a Forth in a few hours"
2023-06-12 15:10:45 <trev        > 
ACTION has been reading about Forth for over a week

2023-06-12 15:12:25 <drakonis    > which language though?
2023-06-12 15:12:49 <trev        > x86 asm, i guess ._.
2023-06-12 15:13:00 <drakonis    > welcome to the club
2023-06-12 15:13:35 <trev        > ty, glad to be here
2023-06-12 15:13:55 <trev        > should probably just do arm, but i feel like torturing myself
2023-06-12 15:15:25 <drakonis    > do it in whichever you want
2023-06-12 15:48:29 <KipIngram   > trev: you only wind up using a small subset of x86 monstrosity for a Forth.
2023-06-12 15:48:37 <KipIngram   > And it's largely from the more sensible part.
2023-06-12 15:49:48 <KipIngram   > I'm about to tackle writing one using my most recent Forth, which means writing my own assembly words.  Got a good bite taken out of them in the last couple of weeks.
2023-06-12 15:51:19 <trev        > you will provide asm words in your Forth, then write a new Forth using the first one?
2023-06-12 15:51:29 <KipIngram   > Yes.
2023-06-12 15:51:57 <KipIngram   > The goal is to wind up with a system that can rebuild itself and yield the same md5 or something hash before/after.
2023-06-12 15:52:25 <KipIngram   > Then never have to use an external tool (for that at least) again.
2023-06-12 15:53:44 <trev        > that sounds cool
2023-06-12 15:54:30 <thrig       > eforth does the "can build itself" thing
2023-06-12 15:54:48 <KipIngram   > Here's hoping.  A weekend ago I worked up a code generator that does all the binary ops (add, sub, ...) plus mov for reg, reg, and also for the indirect and indirect with offset source and destination cases.
2023-06-12 15:55:03 <KipIngram   > I can't think of any way to exhaustively check it, but it's gotten right everything I have checked so far.
2023-06-12 15:55:32 <KipIngram   > And I've decided how I want to handle jumps.
2023-06-12 16:18:11 <KipIngram   > drakonis: That miniforth article is good.
2023-06-12 16:18:17 <drakonis    > yes.
2023-06-12 16:18:50 <KipIngram   > Lots of nifty tricks.