PIC Assembler.

[john jardine]

[...]

What were you planning to shoot?

The scrote that spawned the "tutorial".

 

[john jardine]

[...]

Are any of the older versions of MPLAB any less "bloated"?

No. Even the 10 year old original was bloated and from there they have
laboured on each iteration to increase the bloat by a factor of about
sqrt(2).

 

[Eeyore]

You're just being anti-PIC.

Not at all. I'm just poking fun at the oft-seen throwaway "use a PIC" comment.

Do you really think there are more people using
805x parts than PICs?

I believe they're used in comparable numbers.

Considering that you have little experience in programming micros

Meow !

and aparently none with PICs, I'm at a complete loss to
understand why you want to ram PL/M and 805x down everyones throat every
time this stuff comes up.

You sound a bit over-sensitive to me. I'm not ramming either down anyone's
throat. I was however posing a reasonable question to Ian.

What is it exactly that is so great about PL/M
and so "Bleh" about C?

'C' was never intended as a programming language for uCs. It's massive overkill.
PL/M was. Check the efficiency of the code they produce.

Graham

 

[Eeyore]

The scrote that spawned the "tutorial".

I agree. WTF is a "Manifest constant" anyway ?

Graham

 

[ian field]

[...]

Are any of the older versions of MPLAB any less "bloated"?

No. Even the 10 year old original was bloated and from there they have
laboured on each iteration to increase the bloat by a factor of about
sqrt(2).

Ever heard of the EPE magazine TK3? - I think that contains an assembler
but I'd have to dig out my collection of old magazines to find out.

 

[David L. Jones]

Not at all. I'm just poking fun at the oft-seen throwaway "use a PIC" comment.

Like it or not PIC is now a ubiquitous term that referes to a
microcontrollers in general.
In reality people pick the microcontroller that suits them and their
project.

PIC did not get to be the #1 selling microcontroller in the world by
being a "joke" of any sort. That must mean that more than hobbyists
are using them to flash LEDs.
I don't think it's #1 now (reliable figures are hard to get), but it's
still probably in the top 3.

Dave.

 

[David L. Jones]

Not at all. I'm just poking fun at the oft-seen throwaway "use a PIC" comment.


I believe they're used in comparable numbers.


Meow !


You sound a bit over-sensitive to me. I'm not ramming either down anyone's
throat. I was however posing a reasonable question to Ian.


'C' was never intended as a programming language for uCs. It's massive overkill.
PL/M was. Check the efficiency of the code they produce.

The top quality C compilers can produce superbly tight and efficient
code, even on the 14bit PICs. With C compilers you do often get what
you pay for. All HHL language compilers are not created equal, there
is massive variability between vendors. You can't just generalise and
say PL/M is better than C or whatever.

And if the efficiency of your C (or any HLL) compiler is that
important to your project then you are too close to pushing the limits
of your design, and should probably re-think your choice of processor
for that project. Sometimes it's necessary for various reasons of
course, but on your average project the efficiency of your HLL won't
matter a rats.

Dave.

 

[Anthony Fremont]

Not at all. I'm just poking fun at the oft-seen throwaway "use a PIC"
comment.

At least "some" of the people proposing PICs have supplied code, have you
ever posted any code?

I believe they're used in comparable numbers.

PICs took over the 8-bit market about 5 years ago.

Meow !



You sound a bit over-sensitive to me. I'm not ramming either down
anyone's throat. I was however posing a reasonable question to Ian.



'C' was never intended as a programming language for uCs. It's
massive overkill. PL/M was. Check the efficiency of the code they
produce.

AIUI, PL/M is a mish-mash of PL/1, ALGOL and some other odd stuff. It looks
like some bastardization of COBOL and C to me. In my experience, decent
quality C compilers usually produce tighter assembler code than I do.

 

[Nobody]

I agree. WTF is a "Manifest constant" anyway ?

Roughly, a constant which is given a name rather than being embedded in
the code as a literal value. E.g. in:

#define PI 3.14159
....
v = vp * sin(2 * PI * f * t + ph);

"PI" is a manifest constant.

 

[Jamie]

Like it or not PIC is now a ubiquitous term that referes to a
microcontrollers in general.
In reality people pick the microcontroller that suits them and their
project.

PIC did not get to be the #1 selling microcontroller in the world by
being a "joke" of any sort. That must mean that more than hobbyists
are using them to flash LEDs.
I don't think it's #1 now (reliable figures are hard to get), but it's
still probably in the top 3.

Dave.

100% correct, We use them along with AVR's in industrial applications..
Easy to modify as you work out the bugs. Replaces countless of logic
chips that otherwise would make the circuit un practical.

 

[Jonathan Kirwan]

I can think of no reason whatever to use a language (assembler) that requires
more than one line of code to even add two numbers, never mind anything more
complex.

Then your imagination is unusually restricted.

Three choices are obvious at this juncture. Feel free to suggest
others. (1) You can retract your sweeping, godlike statement, tell me
you really aren't interested in knowing any more about it, and we can
disengage with a notched-down statement from you that allows for other
possibilities you cannot and have not been able to imagine. (2) You
can re-affirm your sweeping and comprehensive knowledge on this
subject and say you willing to defend that position against specific
challenges I will present you. (3) You retreat to a less sweeping
position, simply saying you cannot imagine a case on your own, but
allow me to present one or two for your enjoyment and we leave it
there.

Keep in mind that, for the most part, I am not expecting to convince
you about "assembly-only" projects -- though I can actually draw from
actual projects that I still support (in fact, today, I'm currently
working on precisely one of these), which cannot possibly have been
done except as entirely assembly projects. But your claim goes so
much further than this and was quite broad, "Assembly language makes
NO sense whatever to my mind in the modern world." In other words,
you make no room whatsoever. Period. Not even as an adjunct with
other languages in use. That is so completely and obviously false, to
anyone with the slightest consideration and experience, that I truly
marvel you feel able to make it at all.

Your choice. But you are wrong. There is no question of that.

Jon

 

[john jardine]

[...]

Are any of the older versions of MPLAB any less "bloated"?

No. Even the 10 year old original was bloated and from there they have
laboured on each iteration to increase the bloat by a factor of about
sqrt(2).

Ever heard of the EPE magazine TK3? - I think that contains an assembler
but I'd have to dig out my collection of old magazines to find out.

Yep. Over the years they've given excellent coverage and support to many PIC
(starter or otherwise) projects. I think I came across their 'house'
assembler in a listing for a nice "L-C meter" they did. Had to modify the
text in a couple of lines, to allow a nicked lump of their code (Peter
Helmsley's? 32 bit maths collection), to be assembled on a 'standard' PIC
assembler but it was straightforward otherwise.

If you perchance you ever find yourself with an unshackled £18, I'd suggest
a worthwhile spend as http://www.oshonsoft.com/pic.html. It's much easier
and more fun than MPLAB.
I'm a lazy sod, hate typing and look for quick results, so as a cop-out I
now regularly use his BASIC compiler. No problems over the past 18 months
and his code's as good as mine to prog's about 1kbyte in length.
But ... I can only make this quality comparison as I've done/do the
'assembler' bit.

 

[Jonathan Kirwan]

Not at all. I'm just poking fun at the oft-seen throwaway "use a PIC" comment.

I think you raise a valid point, that not everyone uses PICs, though I
think your choice of words is very poor after that point.

Microchip is no joke to anyone. I've been using their parts as part
of my consulting with many companies, since around 1987 or so, when
they first started making them available for smaller businesses to use
(they used to be only in multi-million use products like rice
cookers.) They support their tools, from what I can see in practice,
literally forever. Even well after they stop selling an old tool,
such as the Pro Mate II, they will continue to support it as well as
supply new plug in modules for it to this very day. I can't say the
same for most of their competition. Business-wise, they are very good
to mid- and smaller companies as well as large customers, so they are
good business partners for most of us.

Sure, the PIC cpu is like looking at a transparent-man anatomy lesson
instead of a mannequin. You get to "see" pretty much every tiny,
silly detail of how a cpu works inside, with them. I can cobble up a
nand-only digital circuit to emulate the instruction set in no time
flat, because they are so transparent, in fact. But that's the only
joke about them, the rest is quite serious business. Microchip is a
good company to partner with, for a great many of us. And that is
more important than some argument about cpu internals.

By the way, aren't you the one saying that assembly isn't used
anymore? If so, what exactly do you mean when you say "joke," then?
Can't be about the machine language, since obviously you don't care
about that, anyway. So it must be something else. And on all the
other facets, Microchip is a strong and supportive partner. So I am
unable to see what else you might be addressing that comment towards.

But yes, I think Microchip is only an important part of the total
business and everyone does NOT use them. No question. So you are at
least right on that point.

Jon

 

[John Fields]

Probably, it's their counterpart to the Microchip PIC. Atmel also makes ARM
processors. Of all the processors I've played with, I think the ARM has the
most beautiful architecture.

---
I don't know if they still use it, because I'm not in that game any
more, but I hated Intel for their segmented gangster-like addressing
scheme and I fell in love with Motorola's lovely flat address space
from the getgo.

I've never played with Microchips' stuff because it always seemed to
me that talking to their chips was like having to use a really fast,
power wasting clock to get some really simple shit done quickly.

Also, I've never had occasion to use Atmel's stuff because I've
always been able to do everything I needed to with Motorola's stuff.

Plus, I'm tooled up for Motorola.

If you've had a chance to look at Motorola's (Freescale's)
instruction set and addressing modes for 8 bitters and have compared
them against Atmel's, I'd be really, really, interested in your
opinions.

 

[John Fields]

At least "some" of the people proposing PICs have supplied code, have you
ever posted any code?


PICs took over the 8-bit market about 5 years ago.


AIUI, PL/M is a mish-mash of PL/1, ALGOL and some other odd stuff.

---
Check this out:

http://en.wikipedia.org/wiki/PL/M

 

[Anthony Fremont]

See.


You haven't used a good C compiler. ;-) Seriously, a good optimizing
compiler will generate code that is incomprehensible to a human, but it will
(almost) always work.

 

[Anthony Fremont]

---
I don't know if they still use it, because I'm not in that game any
more, but I hated Intel for their segmented gangster-like addressing
scheme and I fell in love with Motorola's lovely flat address space
from the getgo.

I've never played with Microchips' stuff because it always seemed to
me that talking to their chips was like having to use a really fast,
power wasting clock to get some really simple shit done quickly.

Also, I've never had occasion to use Atmel's stuff because I've
always been able to do everything I needed to with Motorola's stuff.

Plus, I'm tooled up for Motorola.

If you've had a chance to look at Motorola's (Freescale's)
instruction set and addressing modes for 8 bitters and have compared
them against Atmel's, I'd be really, really, interested in your
opinions.

I can't really comment as I haven't really written anything for a motorolla.
I've seen some code listings before and it looked ok. Haven't written any
code for an AVR either, but I can appreciate why people like them with the
register set.

 

[Eeyore]

The top quality C compilers can produce superbly tight and efficient
code, even on the 14bit PICs.

I dare say that must be possible, yet the anitpathy to HLLs in this thread is quite
extraordinary and must have some basis. Possibly as a result of poorly implemented 'C'
compilers.

With C compilers you do often get what
you pay for. All HHL language compilers are not created equal, there
is massive variability between vendors. You can't just generalise and
say PL/M is better than C or whatever.

I didn't say 'better' though did I ? I pointed out that PL/M is purpose written for
the target uP / uC.

And if the efficiency of your C (or any HLL) compiler is that
important to your project then you are too close to pushing the limits
of your design, and should probably re-think your choice of processor
for that project.

I agree with you 100%.

Sometimes it's necessary for various reasons of
course, but on your average project the efficiency of your HLL won't
matter a rats.

I think people fret too much and quite possibly erroneously assume that assembler will
provide a superior result when it's far from clear to me that's the case because it'll
depend highly on the individual programmer's skill. As I previously said, all the
worst code I've ever seen is in assembler.

Graham

 

[Eeyore]

At least "some" of the people proposing PICs have supplied code,

Where ?

Graham

 

[Eeyore]

AIUI, PL/M is a mish-mash of PL/1, ALGOL and some other odd stuff. It looks
like some bastardization of COBOL and C to me.

Shows how little you know about it.

Graham