20-100 Oh yes... my SC/MP days were fun too... you remember the Elector series about it?
I miss "understandable" systems.
It isn't the complexity of our systems that turn them into black boxes. It is the lack of information about the inside and the impossibility to change stuff in these big chip blackboxes. That might change with FPGAs being like scriptable digital breadboards.
I hope to someday see NetBSD on FPGA(s) being capable enough to rebuild itself in all dimensions from Verilog/HDL descriptions upwards. Maybe we just would need a swarm of interconnected smaller FPGAs instead of waiting for affordable huge chips with the additional restriction of a FOSS toolchain.
For a while a "standard" PC as booster to do that faster than the FPGA based system alone may be ok, but even if a rebuild would need a week on that FPGA system itself, it would be a proof of autonomy and really feel good.
Too many stuff today relies on SDKs on (sometimes even Windows-only-)PCs instead of getting autonomous.
â:-/
I've still a long way to go. The most complex thing I seem to understand in Verilog currently is a simple (better lets say "dumb") serial interface.
Dreams...
