FarmerPotato

I'm much further along in the code disassembly. There is some brilliant 9995 code in there. Details on the other thread, but these are now certain: CRU map 0000 9901 0040 9902 serial port A 0080 9902 (not populated) 0140 something 0180 something 0FC0 something CPU Memory Map (top 2 address bits go through a PAL) 0000-7FFE EPROM 8000-83FE memory mapped area. Registers used to access 34010 memory. 8400-BFFE EPROM bank switched: 8400-BFFE or C400-FFFE. C000-FFFE Static RAM The EPROM mixes code, strings, data, and pointer tables. I've got most of the 34010 disassembler disassembled. (Learning the 34010 opcodes!) It uses a neat "sieve" to match instructions.

I put some cleaner source files into the first post.

If you could capture a demo's writes to VDP and SOUND, you might compress and stream that from ROM. Such a "recorded demo" could run faster--no one needs to know how long the calculations really took

I took my graph paper in the car. Erased and redrew sprites without ever seeing them on the screen.

My idea involves animating a lot of recognizable 4A sprites in hilarious ways. Recognizing the history, but re-mixing it.

Good thing: board has a TMX7000, which means no internal ROM. The jumpers are set for external ROM - 4K at F000. For more info on the hardware, read

OK there are a lot of typos! I will compare and fix. I confess I used OCR as a first step.

If you read the source, you'll notice that the Speech Editor has graphical output! It makes little bar charts on the terminal. It puts a "display list" into the DAC buffer. You connect the output to an oscilloscope!

I look forward to dumping the ROM! 2532s. Anything you can do to test on the 4A would be welcome. (Any typos are mine.) My thoughts: VSP! VSP@ and GETBITS are the only words missing? Should be straightforward to implement in 9900 assembly. If you want to mock it up, VSP! is a no-op. Mock up GETBITS and VSP@ to return (requested bits of?) a CALL SPGET string (XB). Maybe bit reversed. Is this all FIG-Forth? In my dad's Lubbock speech papers, I found a 1982 version of the TI Forth manual. It is bound with a nice report cover. Clearly there was interest in Forth at TI.

Forth source from the Speech Education Module User's Guide The SEM was a single-board kit produced in 1982 to demonstrate the TMS5220 speech chip. It had the VM60002 "Industrial Vocabulary" PHROM. (phrase ROM or TMS6100), two serial ports, and audio out jack. User software was a speech editor, written in Forth. The Forth source of the speech editor will be of interest to 4A users. So I'm posting in TI-99/4A Development. Most of the Forth should be compatible with the Speech Synthesizer. The 2532 EPROM in the SEM is the Forth interpreter, written in TMS7000 assembly or Forth, plus the compiled source listed here. This source is Appendix L of the manual. About 7 more pages of code are in TMS7000 assembly. This code sends bytes to the TMS5200 and reads the PHROM. I haven't typed it in yet. It's not clear to me how it's called from Forth. Each routine ends with a RTS instruction. I don't have the complete manual. Sadly, I don't see any source code for the Forth interpreter or kernel. It would be really sweet to have a Forth cartridge for the CC-40 or ExelVision. If I find that source, I'll post in the CC-40 forum. ( EDIT: comments in lower-case are mine. ) ( EDIT: comments at top are mine. ) ( the two serial ports are 6850s. the PIA are 8255s. COLDIO initializes them. ) EDIT: here are cleaner files. Syntax checked in ANS Forth, any dependencies (FIG) defined empty at top. Probably still some typos. SPEECH_EDU.FS LOADMEM.fs SPEECH_EDU.fs

John Philips' story fits with an anecdote I heard. Texas Instruments engineers would go to the arcade under the student center of Texas Tech University, to see what was "in". TTU had the most impressive arcade in Lubbock.

LOL same! I was trying to cheat by editing sectors when I stumbled across some familiar hex codes, like 0103070F. Then, monster parts! Starting using up graph paper. I replaced some monsters using DISKO too.

https://xkcd.com/2928/ Software Testing Day

Rolling around in my head for a long time: a big update a little demo of mine. What kind of timeframe are we looking at?

"Taking the Wraps Off the 34020" BYTE Magazine, September 1988 on www.ardent-tool.com