27 Oct
2006
27 Oct
'06
4:24 p.m.
On 27 Oct, 2006, at 07:41, ard at p850ug1.demon.co.uk (Tony Duell) wrote:
You think I don't realise that...
It's not that hard to trace out schematics of
a classic computer [1],
Not all classic computers are micros. Fancy tracing the schematic for my 500 square foot second generation mainframe?
Fortunately I
have most of them, along with the 'address book' which lists every
logic signal, its sources (think wire-OR) and destinations. Its on
roughly A3 paper and is in two volumes four inches thick.
Ah, right... Actually, given enough time ) would _love_ to trace out
something that complicared (in my experience, time taken to trace the
schematic goes roughly as the square of the number of component-pins [1]).
[1] Although there are many other factors. A known 'big chip' like a
microprocessor speeds things up a lot. A big board of TTL, or worse still
just discrete transistors, slows things down. The reason is that a micro
can be used in essentially one way, it will indentify the data and
address buses (OK, maybe then buffered, but you can find those buffers
and thus the buses inthe rest of the machine). A NAND gate can be used
for dozens of different things. And a transistor even more. That's why
sorting out the HP9100B took so long.
-tony