I know, personally, of a chap in Germany who is, at
this moment,
putting the finishing touches on a modern-day vacuum-tube computer.
I'm looking forward to seeing the designs and writing a simulator for
it so we may all have fun.
I would certainly like to find out how he is doing this - his life depends
on this...
While it may be a very interesting experiment, finding out what could have
been done at some point in history (like that Thor Heyerdal (sp?) guy), it
could also be a bit damaging to other fields. Some vintage parts are
really drying up and make accurate restorations VERY difficult and/or
expensive. This is very evident in the old tube (valve, bottle, etc.)
field. Some of the tubes I need for restorations are next to impossible to
find these days, unless I pay large amounts from the audio people (I have
little love for them). While some of the tube types are not suitable for
computers, some are. I certainly hope he is using some of the more common
types that can be purchased by the bushell.
Computers aren't all _that_ hard to design and
build! Just remember
that you don't _need_ all the instruction-set bloat that's so common
nowadays. Think RISC. Of course, before there was CISC/RISC there
was... well "RISC". The first time I heard of RISC the though that
popped into my mind was: "Finally! Back to basics!".
Of course, once you start playing with tubes - lots of them - problems
come up that are not usually found when using TTL.
William Donzelli
william(a)ans.net