2024-05-29 09:53:39 <GeDaMo      > If anyone is interested, this appears to be the source code for the Canon Cat https://archive.org/download/CAT240SC
2024-05-29 10:01:36 <veltas      > Noice
2024-05-29 10:05:36 <GeDaMo      > There seems to be a fair amount of M68K asm
2024-05-29 10:06:29 <veltas      > The oldest code at my company is all M68K stuff
2024-05-29 10:07:06 <veltas      > In K&R C
2024-05-29 10:08:07 <dave0       > i can code in m68k assembly :-)
2024-05-29 10:08:27 <dave0       > also i found the sas/c compiler for amiga on archive.org
2024-05-29 10:08:42 <dave0       > it comes with an assembler
2024-05-29 10:09:12 <GeDaMo      > There's an M68K assembler in the Cat source code :P
2024-05-29 10:09:21 <dave0       > nice :-)
2024-05-29 10:09:38 <dave0       > i had a quick look ... it's a reverse polish assembler!
2024-05-29 10:10:37 <GeDaMo      > I thought most Forth assemblers were RPN
2024-05-29 10:10:51 <dave0       > i've only used the sas/c c compiler on an amiga emulator, but it's possible that it works in 1/2 meg of memory off a floppy disk
2024-05-29 10:12:42 <veltas      > I actually can't code in M68K but I doubt it would take me long to learn
2024-05-29 10:13:16 <dave0       > veltas: it's comfortable
2024-05-29 10:13:59 <GeDaMo      > https://ia801308.us.archive.org/23/items/CatForthInside/Cat%20Forth%20Inside.txt
2024-05-29 10:14:59 <dave0       > veltas: there are 8 data registers that you can do arithmetic on, and 8 address registers that you can use as pointers, and a generous amount of addressing modes
2024-05-29 10:15:59 <veltas      > Yeah I just looked at the wikipedia page
2024-05-29 10:16:13 <veltas      > I can guess what 90% of the mnemonics mean already from other archs
2024-05-29 10:16:15 <dave0       > veltas: also it's big endian
2024-05-29 10:16:17 <veltas      > Very conventional
2024-05-29 10:16:28 <veltas      > Yeah reminds me of PPC
2024-05-29 10:16:39 <veltas      > Big endian sucks but I can live with it :)
2024-05-29 10:16:48 <veltas      > Controversial moment lol
2024-05-29 10:17:21 <dave0       > veltas: one more thing is the assembly language is   <src>,<destination>   which is backwards from x86 and arm
2024-05-29 10:17:51 <dave0       > big endian and src,dest order is probably the hardest to un-learn when switching from x86
2024-05-29 10:18:11 <dave0       > just beause it's probably in your muscle memory
2024-05-29 10:18:27 <veltas      > I've done plenty of PPC so I can handle other endianness
2024-05-29 10:18:32 <veltas      > And it usually doesn't matter anyway
2024-05-29 10:18:42 <veltas      > Does it have IBM bit order?
2024-05-29 10:18:55 <veltas      > I.e. 0 is most significant bit
2024-05-29 10:20:41 <veltas      > Looks like it uses normal bit order
2024-05-29 10:20:43 <veltas      > This is good
2024-05-29 10:20:48 <veltas      > IBM bit order sucks
2024-05-29 10:20:57 <dave0       > veltas: for bit instructions 0 is least significant, but for bit field 0 is most significant (bit is 68000 BSET BCLR BCHG BTST bit set/clear/change/test) and field is 68020
2024-05-29 10:21:32 <dave0       > hmm
2024-05-29 10:21:39 <veltas      > Oh dear
2024-05-29 10:21:43 <dave0       > so it's inconsisent between 68000 and 68020
2024-05-29 10:21:48 <veltas      > Many such cases :P
2024-05-29 10:21:54 <dave0       > :-/
2024-05-29 10:22:12 <dave0       > oh and it enforces alignment
2024-05-29 10:22:19 <veltas      > But you get what you're given, it was one of the early big 32-bit archs
2024-05-29 10:22:28 <veltas      > I would have been quite happy with it back in the day
2024-05-29 10:22:32 <dave0       > 16bit and 32bit fetches must be on even addresses and code has to be even too
2024-05-29 10:22:35 <olle        > Ah poor 68k
2024-05-29 10:22:43 <dave0       > veltas: it was real nice to program
2024-05-29 10:23:02 <dave0       > olle: i have fond memories :-)
2024-05-29 10:23:20 <dave0       > flat address space
2024-05-29 10:23:37 <dave0       > completely steam rolled by intel
2024-05-29 10:23:43 <olle        > Yep
2024-05-29 10:23:47 <veltas      > Aligned access is faster on all modern archs anyway
2024-05-29 10:23:51 <veltas      > So we're all doing aligned now anyway
2024-05-29 10:28:11 <olle        > Was it smart to abandon 68k for PPC tho?
2024-05-29 10:28:20 <olle        > Even Apple had to abandon PPC in the end anyway.
2024-05-29 10:28:38 <olle        > And now they might abandon x86 for ARM, which is kewl.
2024-05-29 10:30:42 <dave0       > i think apple spent a lot of time and money going through each different cpu
2024-05-29 10:30:51 <dave0       > maybe linux has an advantage of being portable to everything
2024-05-29 10:31:17 <dave0       > i keep reading that windows NT is portable .. in particular alpha cpu's .. but is it really true?
2024-05-29 10:31:28 <olle        > Apple is based on Darwin OS now, which is a *nix
2024-05-29 10:31:33 <GeDaMo      > It was in the 90s, not sure about now
2024-05-29 10:32:27 <GeDaMo      > Windows NT targeted a lot of the platforms that were about in the 90s like MIPS and Sparc
2024-05-29 10:32:36 <GeDaMo      > Most of them no longer exist :/
2024-05-29 10:33:04 <GeDaMo      > Huh, maybe not Sparc
2024-05-29 10:33:24 <GeDaMo      > "Multi-architecture support — initially instruction set architectures including IA-32, MIPS, and DEC Alpha; support for PowerPC, Itanium, x64, and ARM were added later and later still: x86 (including IA-32 and x64) and ARM" https://en.wikipedia.org/wiki/Windows_NT
2024-05-29 10:33:51 <GeDaMo      > Ah, Sparc was planned but never released
2024-05-29 14:56:21 <KipIngram   > I'm not 100% sure, but I do think MacOS is still BSD based.  I'd be surprised if they've completely replaced all that.
2024-05-29 14:57:40 <KipIngram   > Sparc had a nice little day in the sun.
2024-05-29 14:57:48 <KipIngram   > For a little while in there it looked like they might take over.
2024-05-29 14:58:10 <KipIngram   > All of that history is very interesting.
2024-05-29 14:58:30 <KipIngram   > You can waste a lot of Saturday afternoons delving into it on YouTube.
2024-05-29 15:00:06 <olle        > KipIngram: Yea Darwin OS is a BSD fork IIRC
2024-05-29 15:00:17 <olle        > But might be more Frankenstein that that :)
2024-05-29 16:31:06 <rendar      > there were Forths that implemented particular words to intend events? e.g. if i do `: __on_mouse_click "hello!" . ;`  here i'm defining a word that is called every time i move the mouse
2024-05-29 16:33:14 <crc         > I've done Forths with things like that
2024-05-29 16:47:47 <KipIngram   > Oh, I've no doubt it's Frankensteined.
2024-05-29 16:48:27 <KipIngram   > rendar: That's getting into an area that's not "standardized," so you see various approaches out there, I expect.
2024-05-29 16:48:44 <KipIngram   > I think I'll have an event loop of that sort in my next system.
2024-05-29 16:49:03 <KipIngram   > If you ever want to have anything like a standard graphical interface then you need that kind of thing in some form.
2024-05-29 16:49:59 <KipIngram   > You can "describe" regular Forth in that model - QUERY is sort of like "on Enter, give me a string."  And in the meantime it just "manages the keyboard and display."
2024-05-29 16:50:21 <KipIngram   > Since that's the only interesting thing that can then happen, you don't need to have a list of things to wait for.
2024-05-29 16:50:53 <KipIngram   > To wait for any of a set of events that have unknown sequence you need something more general.
2024-05-29 16:54:11 <crc         > Darwin has a userland originally derived from FreeBSD, but (most of) the kernel is definitely not from BSD.
2024-05-29 16:58:40 <KipIngram   > Because it's been re-written, or because it originated elsewhere?
2024-05-29 16:58:49 <crc         > it originated in Mach
2024-05-29 17:00:29 <thrig       > (they also pulled in random bits from OpenBSD, judging by the CVS ID tags in various files)
2024-05-29 19:36:22 <rendar      > KipIngram, i see, thanks for your wise advice
2024-05-29 20:00:54 <zelgomer    > rendar:  :noname ." hello" cr ; when mouse_click  ( define when accordingly)
2024-05-29 20:01:40 <rendar      > interesting
2024-05-29 20:01:53 <rendar      > but how you'd define `when` there?
2024-05-29 20:02:48 <zelgomer    > : when ( xt ccc)  word ( now look up some memory location by string and store xt to it using !) ;
2024-05-29 20:03:33 <zelgomer    > if you want to take the magic out of it, then just  :noname  ." hello" cr ;  on-mouse-click !
2024-05-29 20:03:55 <zelgomer    > on-mouse-click is a variable
2024-05-29 20:10:52 <crc         > rendar: from one of my test systems, http://forth.works/share/DTYLCV9QCo.txt
2024-05-29 20:14:38 <rendar      > cool, crc
2024-05-29 20:14:40 <rendar      > thanks all
2024-05-29 20:16:11 <crc         > (in mine, the mouse actions cause the vm to trigger interrupts
2024-05-29 20:16:43 <crc         > the code sets a few words to handle these (default handler is just a non-operation)
2024-05-29 20:23:30 <veltas      > Do you any have recordings demoing graphical retro programs out there?
2024-05-29 20:23:55 <veltas      > I mean graphical interactive programs
2024-05-29 20:33:26 <crc         > I haven't recorded anything graphical yet. I can do so the next time I'm running the X11 or SDL stuff.
2024-05-29 20:41:27 <crc         > I don't really have a lot of graphical things to show though. Arland & I have been experimenting with small things on and off (mandelbrot viewer, a small "paint" application, some simple drawings), but we've not yet started anything bigger
2024-05-29 20:47:33 <thrig       > mandelbrot sets are pretty big
2024-05-29 21:20:53 <veltas      > crc: I would be interested