[rescue] ASM programing and PDPs (was something about macs a LONG time ago)

[joshua d boyd]

On Fri, Aug 03, 2001 at 01:06:10PM -0400, Brian Hechinger wrote:
> > >   You should study the pdp11 architecture and its assembly language,
> > > Josh.  I think you'd really like it.
> > 
> > I'll have to do that some time.
> 
> i could probably scrape up an LSI-11/03 CPU and a 32K ram card for you, but
> you'd need to dig up some of the other parts.  and you would have to be serious
> about doing it.  or at least serious about getting it running and getting an
> OS on it.

Not that serious about it right now.  Too much else is happening in life
to take anything else on.  Need to get things arranged and working, and
need to get quite a bit of coding done for some people, then get quite a
bit done for myself.
 
> > I have to say that I really like RISC machines.  Yeah, I know that
> > they made a lot of sacrifices, but I like that they have so few
> 
> i've always liked them, but the one big downfall with them is they require a
> VERY large amount of memory compared to anything else.  but that's the price
> you pay when you make some choices.

Memory is cheap these days.  So, you trade memory (cheap) for man hours
(expensive).  Well, that's should only be within reason.  If you do it so
much that it makes the machines slow for the end users, well that is just
evil.
 
> > instructions.  I mean, who is going to remeber how to use the Vaxes
> > pointer advance instruction or the polynomial instructions without
> > looking it up everytime?
> 
> by using it.  if you do it all the time you will remember.  if you don't do it
> all the time you won't remember.  doesn't matter if there are 20 instructions
> or 2000 instructions.  maybe, i could be completely full of shit.

Err, you use the polynomial functions alot?  And the linked list
instructions alot?
 
> > But then again, since ASM should only be used for optimization, maybe ASM
> > shouldn't be easy to remeber.
> 
> it would require copious amounts of research to use, so you really think about
> your problem.  hmmmm, that has merit.

True.

-- 
Joshua D. Boyd