TSU Meeting MONDAY — May 6, 2024
David, Paul, Joe, Carl Miles, Ryan, Adam, Ricardo, Gustavo, Keith, Johnny Red
The Monday meeting opened with Paul offering his last spare SCLD chip (with socket, date-coded March 1985) — Joe took it for a board he suspects has a bad SCLD. David fielded a question about Mira Pascal for the 48K Spectrum (a manual is being sought) and noted Fortran also exists for the machine. The bulk of the early discussion was David's read-out of an email from Richard Burt in Ontario, who still uses 'The Wiz' — a 1984 small-business accounting suite by Steve Cermak (payroll, inventory, AR/AP, general ledger) — on his TS2068 with a Larken disk system, having modified it for Canadian PST/GST. Rich is now porting it to run on the TS-Pico (which doesn't need Larken's PRINT #4 prefix). Gustavo in Argentina told a parallel story of a small Buenos Aires building-supply store that ran TS2068 invoicing on cassette — one tape per customer, ~30 customers — with an Epson printer for about five years, until invoice-format law changed.
Much of the meeting was a TS-Pico firmware status update. Ricardo, Gustavo, Ryan, and Joe are converging on a mid-summer (June-ish) release. Highlights: Ricardo wrote a tool that loads a .DCK file from SD card into the Pico's flash or SRAM (auto-handling the header bytes) so you can boot 2068 cartridge images like Flight Simulator; Jeff's Pico video board is essentially nailed down (with scan-line / CRT-shader discussion); SD-card load is being made more efficient; and the perennial 'J IO error' (a generic CRC/IO fallback Gustavo deliberately uses to avoid corrupt loads) is getting more granular reporting, plus a new TPI VERBOSE command that prints a status line after each load. Joe is making 3D-printed cases and Ryan is starting on a Norton Commander–style file browser for the Pico.
Carl Miles gave an update on his TS1510 clone: units have shipped to David, Adam, Ryan, and Johnny Red in Portugal (routed via Switzerland to dodge Portuguese customs, a month in transit post-Brexit). Ryan walked through a subtlety with cartridge ROM images: some loaders (e.g. ViewCalc, ViewFile) work on a TS1000 — which initializes BASIC system variables before handing off — but fail on a TS1500, which jumps straight to the cartridge loader. Ryan's 'tape-like' alternate loader wipes memory and rebuilds the sysvars to fix this, and he is planning a hybrid. Carl plans to send the EasyEDA design files to Joe (for 3D-printed cases) and David, to fix a couple of trace issues (A14/A15 swapped on the interface board, missing key-way notches) for the next PCB run. Carl also showed pictures of the rare K2 'Cradle' expansion case for the TS1000 — a heavy metal enclosure with 64K RAM, two expansion slots, joystick port, and full chiclet keyboard — into which he plans to install a 1000. A separate ZX81/1000 project on Facebook is building a back-of-machine board that selects between alternate ROMs (including ASMIC) without opening the case.
Other items: Jeff noted in chat that Zilog (now owned by Littelfuse) has put out a last-call for Z80 orders — end of Silicon-base production — prompting speculation about Rochester Electronics picking it up vs cannibalizing parts from existing TS1000s. Paul has been mailing EPROMs around for dumping: several JLO DOS revisions, QL ROMs including a Minerva-with-Toolkit-2, and Tom Bent's modified TS1000 ROM (kills the memory-check delay, with a documented manual on archive.org). Keith shared the assembly listing for Jack Doheny's old TS2068 keyword-entry program (lets you type 'PRINT' instead of using the P keyword), which hooks stream −3; Keith later corrected the group that the code lives at 23885 (low memory), is not relocatable as written, and conflicts with HOT-Z arrows. David read a passage from William Gibson's 'Pattern Recognition' about the ZX81/TS1000's US reception, which the group cheerfully debunked (no three-month RAM-pack delay; ZX81 was 1981, not contemporaneous with Nintendo). The meeting closed with a digression on TS2068 manufacturing — Korean-made PCBs and assembly, NEC SCLD chips packaged in the Philippines, and how US semiconductor fabs of that era shipped wafers overseas for dicing and packaging. Next meeting: Sunday, May 19, 2pm Eastern.
Highlights
- Paul gives away his last spare TS2068 SCLD chip (with socket) to Joe for a board repair.
- Richard Burt has been running 'The Wiz' (Steve Cermak, 1984) accounting suite on a TS2068 + Larken for 40 years; now porting it to the TS-Pico.
- TS-Pico firmware update targeted for mid-summer: DCK-to-flash loader, TPI VERBOSE command, and finer-grained error reporting beyond the catch-all 'J IO error'.
- Carl Miles' TS1510 clones are out to multiple members internationally; Ryan adds a 'tape-like' alternate loader to handle TS1500 vs TS1000 cartridge-boot differences.
- Zilog (Littelfuse) issues last-call for Z80 production orders — end of life for new silicon.
Projects discussed: TS-Pico · TS1510 Cartridge Interface Clone Topics: TS2068 / TC2068 · TS1000 / ZX81 / TS1500 · TS1510 cartridge interface · Pico / modern interfaces (UnoDos, etc.) · Cartridges, EPROM & dumping · Hardware projects & new boards
Summary generated from the meeting recording; the full transcript is kept in the group's private archive.