I forked original project and corrected one obvious mistake (, instead of ').
Than I realised I can't compile it, because digging out TASM and trying to run it in a virtual machine is just too much for me. So I ported this to GNU AS.
Surprisingly, I found two differences that are not obvious.
--- /dev/fd/63 2024-02-18 16:04:46.994421948 +0000
+++ /dev/fd/62 2024-02-18 16:04:46.994421948 +0000
@@ -289,7 +289,7 @@
00001200: 8d0b 6b0a 4f0a f407 3a06 6b0a f407 7106 ..k.O...:.k...q.
00001210: 200d 270a 3708 7106 c204 270a 3a06 3708 .'.7.q...'.:.7.
00001220: 7106 c204 c546 4f52 54c8 e511 2c0d 1a12 q....FORT...,...
-00001230: 8120 34fe 0000 8b44 4546 494e 4954 494f . 4....DEFINITIO
+00001230: 81a0 34fe 0000 8b44 4546 494e 4954 494f ..4....DEFINITIO
00001240: 4ed3 2412 b406 3708 3a06 4508 7106 c204 N.$...7.:.E.q...
00001250: c1a8 3612 b406 c401 2900 3d0f be05 c204 ..6.....).=.....
00001260: 8451 5549 d450 12b4 062b 07fe 0771 0682 .QUI.P...+...q..
@@ -339,7 +339,7 @@
00001520: 8d45 4d50 5459 2d42 5546 4645 52d3 1515 .EMPTY-BUFFER...
00001530: b406 6607 7607 af05 8c0a f20e c204 8642 ..f.v..........B
00001540: 5546 4645 d220 15b4 0655 15c2 0485 424c UFFE. ...U....BL
-00001550: 4f43 cb3e 15b4 06c4 01b1 ff13 1429 083a OC.>.........).:
+00001550: 4f43 cb3e 15b4 06c4 0128 0013 1429 083a OC.>.....(...).:
00001560: 064b 0586 07e4 1366 074b 05c2 0483 522f .K.....f.K....R/
00001570: d74d 15b4 0609 093a 06d9 01c2 04b4 06be .M.....:........
00001580: 05be 05be 05c2 0485 464c 5553 c86d 15b4 ........FLUS.m..
At line 0x1230.
Original listing says:
2769 1230 81 20 .BYTE 81h," "+80h
I guess it's an undefined behaviour of concatenating string with number.
After my changes it's: (offset because GAS is numbering within section, not from 0)
3044 10dd 81 A0 .BYTE 81h,' '+80h
At line 0x1550
Original listing says:
0031 0144 DISK_START: .EQU 0A000h ;Pseudo disk buffer start
0032 0144 DISK_END: .EQU 0F000h ;Pseudo disk buffer end
0033 0144 BLOCK_SIZE: .EQU 0200h ;Pseudo disk block size
...
3240 1559 B1 FF .WORD DISK_END/BLOCK_SIZE-DISK_START/BLOCK_SIZE
I changed it a bit, but even before I was getting different value:
31 DISK_START: .EQU 0A000h ;Pseudo disk buffer start
32 DISK_END: .EQU 0F000h ;Pseudo disk buffer end
33 BLOCK_SIZE: .EQU 0200h ;Pseudo disk block size
...
36 MAX_DISK_BLOCKS = (DISK_END-DISK_START)/BLOCK_SIZE
...
3565 1406 28 00 .WORD MAX_DISK_BLOCKS ;Max number of blocks
I think 40 blocks is the correct value (instead of some negative number...). The only explanation I can think of is that TASM doesn't have arithmetic operations precedence.
RAM image has base address 0x8000 and can be loaded by SCM for RC2014.
In this version SYSTEM table is moved to 0xFA00 not to collide with RAM memory reserved by SCM.
Original files are in the original directory. Original README follws.
I know NOTHING about FORTH but I just want something simple to mess around with.
I found a simple FORTH buried in a ZIP file on the Z80 info site at ...
http://www.z80.info/zip/z80asm.zip
I have no idea of it's origin but from snooping around it seems to be fig-forth-79 or based on that.
https://archive.org/details/FigFORTHGlossary
I combined it with Grant Searles initialisation and referenced the routines for the TXA and RXA where necessary.
Built with ... TASM -80 -b -c -dROM FORTH.Z80 FORTHROM.bin FORTHROM.lst ... For ROM binary
or
Built with ... TASM -80 -g0 FORTH.Z80 FORTHRAM.hex FORTHRAM.lst ... For RAM Hex file to run from Monitor
After some messing around and realising that I need to AND 7F to knock off the top bit of the characters as it uses this as an end of word marker.