On Sat, Aug 27, 2005 at 03:50:20PM +0200, Wouter van Ooijen wrote:
> > When a programmer
> > becomes a BCD, the overriding factors become cost and ease of
> > construction.
> 
> Add: ease of getting it working. 
> 
> 
> I think a lot of time>>>money beginners do start with such a simple
> programmer, some of it get it to work, but I am afraid some never get it
> to work and drop the whole PIC idea.

That's why I debate the utility of programmers in general. My holy grail
for this segment is the PIC programmer equivalent of a Knoppix style
Live CD. You just pop it in and it just works.

> I have no idea of the relative size
> of the two groups. So yes, almost-zero-cost programmers are important
> for beginners (some might never start with PICs if it were not for these
> type of programmers), but I am afraise there are also beginners who
> maybe could have afforded a low-end intelligent programmer, but instead
> try an almost-ene-cost one, can't get it to work, and drop PICs
> altogether.

I'm with you there. The fact is that the pogrammer gets in the way of
a hobbyist getting started. It's one of the reasons that items such
as the Basic Stamp and the PicAxe have a following. Both of course
couple an HLL on top. I still debate the necessity of that.

The reason this is such a stick subject is that there are so so so many
variables. Just off the top of my head (in no particular order):

Cost
Reliability
Support
Culture (folks use programmers because every document says to need a programmer)
Interfaces (We need a solution to the USB problem)
Ease of Implementation

With so many vectors, I know that one solution will not satisfy everyone. However
I do think we need to give thought and discussion to appropriate options.

USB is a sticky subject. I've been thinking how far into the learning curve
would one get before implementing USB for the application becomes important?
It's one of those places where the programmer vs. bootloader debate becomes
murky. Is it possible or appropriate for the bootloader and the application
to share the USB interface? Mostly because of your thoughts on the subject,
I have grave issues about the transparancy problem there.

No answers yet, just questions.

> 
> > That means for ICSP
> > development that half your pins are being used as the interface
> > for programming durin development.
> 
> in most cases that is not needed, a little thinking can find a use for
> the ICSP pins. Or use a 14-pin chip as developer for an 8-pin chip. Or
> the uChip 'extra-pins' version of the chip.

So my proposed solution is along those same lines. I've built most of my projects
with 40 pin targets. Size wasn't a real concern.

> 
> > Anyway I wrap up with my tagline:
> > "PIC Programmers are overrated."
> 
> For a specific audience that might be true. But the question is of
> course how big and/or relevant that audience is.

Like everything else Wouter, statements are never considered until they
are made. 

BAJ
-- 
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist