Louis van Dompselaar

52 Followers
97 Following
98 Posts

Sysadmin and web programmer @ Brodgar Technology. History. Archeology. Whisky. Photography. Old (unix) computers. Synths (build and repair).

Migrated from @lvdompselaar

No, I don’t want an AI assisted experience

I want clean air, forests, and a future for the next generation

I may regret this at some point, but I felt the need to put down in writing how I feel about this moment in the tech industry.

It is not kind. You may well be insulted by it. If you are... then you really should question yourself.

https://www.garfieldtech.com/blog/selfish-ai

#AI #LLM #Programming

Selfish AI | GarfieldTech

This has now been sorted by using a PLD to adapt the bus timing, so the C012 now only requires 4 I/O-ports (as it should) and not 8.

That does require the additional logic and delay I figured it would, but it's neatly hidden in the PLD.

@ionica eentje voor de verzameling?
It's alive! "Flash Q LED" code from the book working as it should #cosmicos #cosmac #cdp1802
In the early 80s I came across the 1802-based Cosmicos in a book at the local library. I hand copied the schematics from that, which I recently came across. I decided to finally build it. I probably could have recreated the pcb, although that would have been expensive. So hand wiring it is. The data and address busses will be erm… interesting to do. #cosmac #8bit

I'm rather pleased with My First Transputer module (and my first 4-layer pcb, and first manufactured one after hand-etching for nearly 40 years). Module passes the Hello World test. Will have to wait to test the RAM, which will be easiest with a B004(-compatible) board and ispy/mtest.

This will connect to a Z80-based system later, as a small network (PCBWay sent me more than I ordered, so no choice there).

#transputer #inmos #pcb #pcbway

Finally got a stable interface from the #z80 to the T222 #Transputer. The timing of /IORQ and /WR are at odds with what the C012 interface IC expects, making it unreliable, and adding delays in the address decoding logic just made it messy. In the end just using A2 as RnotW and it’s rock solid now. Also testing differential line amps for the interconnects.