SemiOT: Mourning for Classic Computing

From: Greg Lindsey <glindsey_at_ssinc.com>
Date: Wed Aug 15 12:19:54 2001

"Douglas Quebbeman" <dhquebbeman_at_theestopinalgroup.com> wrote:


> > I'd tend to agree, in general. However, I think something like MIPS
> > assembly (such as that taught in CS courses at UIUC) might work well as
a
> > "learners" language, and doesn't require knowing TOO much of the
hardware
> > (aside from the registers / memory distinction, which could be taught
using
> > a "file cabinet / cubbyholes" analogy).
>
> Was the MIPS a true 3-address machine (or do I mean 3-operand?)? Or
> am I thinking of the NS32000 family?

Yes, I'm almost positive it was a three-operand machine -- I still remember
long nights writing "add $r5, $r3, $r12" and the like. I don't know if what
we did was totally real, or more stylized -- all of our machine problems ran
through a simulator. But it was elegant: our implementation had a fixed
32-bit instruction size, thirty-one static 16-bit general purpose registers
(as $r0 was hard-coded to zero), some 32-bit floating point registers (don't
recall how many), plus stack and instruction pointers. Memory was 16-bit
addressable flat space... beyond that I'd have to pull out my old textbooks,
unless I sold them (which I think was the case).

> > But, yeah, I'd say BASIC is still a pretty good language to see if
someone
> > can "get" programming -- provided that someone moves to a structured
> > language quickly if he or she wishes, instead of getting into bad
> > programming habits (as I did for a while).
>
> I grudgingly agree.
>
> Niklaus Wirth thought that BASIC hopelessly polluted a mind from
> every being a good programmer. I, too, think that's harsh, probably
> because I also started with (a superset of) BASIC, and wrote journal
> articles for two years for The Cobb Group's Inside Microsoft BASIC
> and Inside Quickbasic. And every time I think I've finally put it
> behind me, it creeps back into my life. Most recently, TOPS-10 BASIC
> running on a simulated DECSYSTEM-2020 (KS-10) (fixing MONPLY.BAS).

I adopted Quickbasic pretty easily, so learning structured programming
wasn't too bad; however, QB made me lazy in terms of things like syntax
checking (since it was on-the-fly) and variable declaration. It took me a
while before I could program in C without getting hopelessly frustrated.
(Remembering semicolons alone was enough to drive me nuts...)

But BASIC holds many fond memories for me, between learning it on Commodore
PETs, Apple II's, and GWBasic on my old 1186... and as far as getting people
to understand variables, conditional statements, loops and the like, it's
still very useful.

GSL, who now remembers that that LOGO simulator was written in QBX, not
GWBasic...
Received on Wed Aug 15 2001 - 12:19:54 BST