Lpc2000

--- In lpc2000@yahoogroups.com, "Paul Curtis" <plc@r...> wrote:

I love this thread. I was also a fan of the P system, and I used
Borland Turbo Pascal and Delphi for years. I've also used C style
languages (most recently C#), and like many of you, I started with
BASIC (the PET, TRS-80, and Apple II). I couldn't afford to buy any of
these, but I was the guy everyone came to when they couldn't figure
something out, so I learned a lot about them.

I think I have got a document in PDF form that explains the p-code
instructions in case anyone wants it. I was thinking of using
something like that in a recent VM I was designing in my head, but I
am now hedging on doing something a little closer to the JVM. Support
for basic OOP contructs would be good, as would support for pointers
(something the JVM can't do). The interpretive approach of a VM gives
you some very cool debugging support, and allows you to integrate a
multi-tasking RTOS at the VM level, and like the P system, it's very
portable.

This was the original idea I had, and I was targetting 16 bit MCUs
such as the Freescale 9s12 (updated 68hc12; one of the strongest 16
bit controllers), and the MSP430 (the 16 bit chip that runs on the
mere hint that a few electrons might possibly consider switching from
one atom to the next):
http://www.ericengler.com/downloads/EmbeddedVM.txt

But. as I said, I think some of my ideas need to be re-examined, and
I'm also interested in a 32 bit version of it, now that I'm trying to
learn more about Arms. My 32 bit VM might be able to execute programs
compiled to target the 16 bit VM, since metadata and versioning will
be used in the generated code. But binary compatibility between device
families may not be needed, and I'm not sure of whether there is any
practical use for it.

> > Who is making the board and when might it be available?  It's not 
> > like this project is urgent.  My grandson is only in 2d grade.
> 
> Can't really say at present.
> 
> -- Paul.

I always enjoy the ability to speculate because sometimes I have
inside information, and I always try to keep quiet in that case. In
this case, however, I don't have any inside information, so I can try
my hand at guessing. This is especially appropriate now that New Years
Day is upon us. We're thinking of what will come in 2006, so here's my
speculation:

1) Rowley loves Olimex, and Olimex loves Rowley. Olimex will make a
new ARM9 board, and it will be based on a Philips LPC3000 series part.
Why Philips? 
  a) Olimex has more Philips boards than they have of any other ARM maker
  b) Paul likes Philips devices, and he posted this message here in
this forum (lpc2000 forum)

2) Why is Paul so fond of this idea? It's likely that Rowley's new
announcement (that might come within the next couple months), might
involve software tool support for this new device, and it's likely
they are already working with an alpha or beta version of it right now.

3) Rowley is not content with sitting still, and I haven't seen any
big moves from them in the past 2 or 3 months, so this means they are
working hard on their next toolset mentioned in #2. And, since the
Philips Arm9 is not yet common (is it even shipping yet?), it follows
that Rowley is probably not putting all their marbles in the Arm9 camp
just yet. This means their new tools (or maybe this should be
singular, as in "compiler") will target the Arm7 devices primarily.

Please understand - this is 100% idle speculation! Please don't push
Rowley or Olimex to "confirm or deny" these ideas. All companies need
to choose the right time to make announcements, and I may look quite
stupid when that time comes (which would be nothing new). I just got
carried away with this being New Years Day!

Happy New Year!

Eric