Lpc2000

> The p-System had the concept of a segment and even had a SEGMENT 
keyword
> for SEGMENT PROCEDURE or SEGMENT FUNCTION -- a procedure or 
function
> belonging to a program or unit, but whose code is not loaded with 
that
> program or unit's code until it is called.  This is demand loading 
of
> program code.

Yes, I remember that.  I first ran into Load On Call (LOCAL) with 
the IBM 1130, my first 'personal' computer, back in '69.  I really 
loved that machine and Fortran is still a favorite.  It 
was 'personal' in that the company paid me to do things with it and 
I had unlimited use during the off hours.  It was a great tool while 
stumbling through college.

> Then why not try Pascal-S?
> http://www.moorecad.com/standardpascal/pascals.html

I guess I am looking at the whole experience.  I'm trying to get 
past all of the OS related issues and just look at programming from 
the viewpoint of "Algorithms + Data Structures = Programs" by 
Niklaus Wirth.  This is my all time favorite book on how programming 
should be done.  But, opinions vary on that.

> 
> I still have this book which describes the P4 compiler
> http://homepages.cwi.nl/~steven/pascal/.

Got it.  And the original Manual and Report and a few other classic 
texts on Pascal.

But, I don't plan to take it that far.  This is going to be a lot 
like 16k Basic with structure.

> 
> > I had actually thought to do Lee Chen Wang's Tiny Basic but I 
don't 
> > think Basic promotes good program structure and it would be a 
waste 
> > of an ARM processor.  But it sure worked well on an 8080 and 
8085.
> 
> I don't think this is necessarily true--good program structure is
> learned, and it would not be a waste of an ARM processor given 
that RISC
> OS and Arthur ran a nice BASIC.

It certainly wouldn't take much effort to get it running.  And it 
would really RUN!  Maybe I'll do that next week.  I think the 
original fit in 1k bytes of assembly code.  I imagine I'll do it in 
C but it would be a good exercise to do it in assembly.  And, I 
really need to learn ARM assembly...

> 
> My personal introduction to microcomputers was with Commodore PETs 
and
> Acorn Atoms (yes, my first personal computer was an Acord Atom 
with a
> 6502, and THAT Acorn).  Both these had BASIC as an operating 
system and
> language combo.  I find it astounding that people denegrate BASIC 
yet
> BASIC kicked off the microcomputer revolution by letting people 
have
> *something* to write programs in that wasn't daunting and didn't 
need a
> compile-linkedit-run-crash-debug process.  It's astounding how
> resourceful you become when you have limited RAM to write programs 
in.

All true.  I don't mean to denegrate Basic and he will have to learn 
it sooner or later.  I couldn't wait to get enough RAM in my Altair 
8800 so I could run 8k Basic.  I remember paying $400 for a 4k 
board!  That was more than the machine cost.  I still have the 
Altair with a pair of 8" floppies and a home built floppy 
controller.  I also have a CompuPro industrial Z80 machine with dual 
8" floppies.

A year or so back I implemented the T80 core on an FPGA and I now 
have CP/M running silently using 2 CF devices to implement the 16 
logical drives.  It's fun to visit with old friends: ED, MAC, 
FORTRAN, PL/I, MicroShell and, yes, WordStar.

> 
> Good programming is learned over time--it's not easily taught.  You
> write good programs by looking at good programming style.  I find 
it
> rather like learning a natural language, there is no substitute for
> going out and using and talking to natives, it doesn't matter how 
long
> you spend in the classroom being "taught" good grammar and 
vocabulary.

Absolutely!  I had a side project of writing an 8080 assembler in 
PL/I in grad school.  In the end my advisor thought the code looked 
a lot like Fortran, with semicolons.

I also had to write a tiny Algol compiler in Basic for a compiler 
class.  It was ugly.  I wouldn't wish that kind of program on 
anyone.  I had a 110 baud acoustic coupler for a while and upgraded 
to a 300 baud modem.  But, it worked!

> 
> BASIC might not cut it in ivory towers, but hell, it's a great way 
to
> get started with programming and get some use out of a computer.  
I'd
> rather start my children off with BASIC on a micro they can 
directly
> control rather than with Logo or C or Pascal on a PC which they 
really
> can't.

Exactly my point.  Get past all the system related issues.  Keep it 
simple.  In fact, Pascal with only the main program block isn't a 
lot different than Basic except that variables have to be 
predefined.  Ignoring CONST and TYPE for a while and using only 
integers and strings makes Pascal look a lot like Basic, with 
semicolons.

I also have to change the P4 implementation of strings.  I think 
they were limited to 10 6-bit chars to fit the CDC 6400 60 bit word. 
Maybe multiples of 10, I'll have to look.

> I am *not* a fan of universities that teach highbrow stuff as a
> first language--ETH's approach is absolutely right, the staff know 
what
> they're talking about because they know how to construct custom 
hardware
> to run their custom software on, and they [used to] do it all
> themselves.  Bravo!

There's certainly nothing wrong with starting with Basic but it 
isn't what I have in mind.  We can move to Basic as implemented for 
Stamps in fairly short order.  In fact, we have already played with 
that a little and we'll be doing more as his interest in robotics 
grows.

That Tiny Basic is sounding better all the time.  I had added a MEM 
array and MEM function to implement POKE and PEEK.  That could be 
pretty neat with the ARM.  It would be possible to POKE to the IO 
devices by doing something like:

MEM(Port0) = SomeValue

There's nothing like flashing LEDs to add interest!

Richard