Timex Sinclair User Group Meeting with European Friends — June 26, 2022

David Anderson, Lon Hildreth, Carl Miles, Adam Trionfo, João, Daniel, René, Tim H, Tim Swenson, Willie Jones

Former Timex software engineer Lon Hildreth joined the group as the featured guest, walking through documents from a box he had sealed since 1987 and only just reopened. He showed internal memos and an org chart tracing how he, Dave Boyle, Scott McCourt and Lou Gailey came to Timex from Burroughs in early 1983, his service-call record for the Zilog development system, the team's later move to a VAX, and his own performance review. The review confirmed several long-standing community guesses: he wrote the initialization and channels chapters of the TS2000 system-software design, a Rev5 ROM test spec was prepared but management decided not to reissue the ROM (explaining why the 2068's ROMs are socketed but a corrected version never shipped), and bugs found while writing that spec led directly to the bank-switching controller chip being dropped from the design — the missing chip that has puzzled board reverse-engineers ever since. Dave Boyle, who Lon called the best engineer he ever worked with, did most of the bank-switching code; he died in 2011.

The bulk of the technical discussion was Lon's CP/M-on-2068 project. He started in September 1983, flew to Digital Research in Monterey in November with a 2068 and a couple of microdrives, and had a project completion date of March 30, 1984 that he never reached because Timex shut the computer business down first. Notes in his binder confirm the plan was an L-ROS boot cartridge that would take over the machine if no boot ROM was present, with microdrives as the eventual mass-storage target — Timex was fully committed to the Sinclair microdrive, contrary to what Portuguese sources had suggested. Lon also produced pre-production cassettes (the Timex free-tape with keyboard tutor, turtle graphics and home accounting; 64-column display code; a Spectrum-only Hobbit copy) and EPROM cartridges including Ant Attack, Cycle Path, Casino and an unreleased Scrabble. He confirmed the States-and-Capitals random-seed bug story (the cartridge always picked the same state at power-on because the RNG seed was the tick count) and Dave Boyle's one-day fix for a feature already printed on the U.S. box that had been cut from the spec.

Tim H demonstrated his extension to ZMakeBas, the command-line tool that compiles a plain-text BASIC listing into a tape image. Spectrum Next and Spectranet support had already been added at opposite ends of the Sinclair keyword table; the four or five TS2068-specific keywords slot in just above the Spectrum Next range, so the patch was only about ten to twenty lines, with the one trap being an overlap on the ON keyword (TS2068 ON ERR vs Spectrum Next ON ERROR). He has tested a SOUND example in an emulator and is asking for a representative 2068 program to bundle as a sample.

David reported that Jack Boatwright sent two large boxes of Timex newsletter material, of which he has now scanned roughly two and a half linear feet, and is buying a Czur CZ24 book scanner via Kickstarter to handle bound material non-destructively. Carl showed Egyptian eBay finds — a genuine Sinclair SJS1 joystick and a 128K Spectrum +2 for about $110 — and mentioned that the same seller had Arabic-ROM Spectrums with a switch between Latin and right-to-left Arabic modes. Daniel briefly displayed a Czechoslovak Didaktik Spectrum clone, tying into David's recent read of "Behind the Iron Curtain (When It Was a Country)" whose author has agreed to join a future meeting. The next meeting was scheduled for Monday July 11 (skipping July 4), and João from the Load ZX Spectrum Museum flagged that the Timex Portugal telex archive he is cataloguing may eventually need a formal release blessing from Timex Corp before public sharing.

Highlights

Lon Hildreth opened a Timex box sealed since 1987 — internal memos, org chart, performance review, CP/M notebooks and pre-production cartridges/tapes
Performance review confirms a Rev5 ROM test spec was written but management dropped the reissue, and the bank-switching controller chip was cut after bugs surfaced
CP/M-on-2068 plan was an L-ROS boot cartridge targeting Sinclair microdrives; Lon visited Digital Research with a 2068 in November 1983 but the project died with the business
Tim H extended ZMakeBas to support TS2068 BASIC keywords, slotting them in next to existing Spectrum Next and Spectranet ranges (~15 lines)
Daniel showed a Czechoslovak Didaktik Spectrum clone; the "Behind the Iron Curtain" author has agreed to join a future meeting

Topics: TS2068 / TC2068 · CP/M · Cartridges, EPROM & dumping · Tape & library archiving (TAP/TZX) · BASIC programming · ZX Spectrum

Summary generated from the meeting recording; the full transcript is kept in the group's private archive.