frieder.ferlemann at web.de
Thu Jul 19 16:52:03 EDT 2007
James Cameron schrieb:
> What I've done so far ...
> 1. reviewed an A-Test PCB to identify parts, and read the data sheets
> for the KB3700 chip being used on the B2 and B4,
> 2. briefly tested SDCC's assembler and simulator to verify that it can
> be used at least to produce emulated serial port output,
> 3. briefly tested SDCC's compiler, looking at the assembly output,
> 4. code-reviewed Camel Forth for 8051, mapping the functionality that
> it provides,
> 5. code-reviewed Mitch's rewrite of Camel Forth into OpenFirmware
> metacompiler semantics, providing feedback to Mitch,
> Mitch's most recent post to devel@ summarises the status of that effort
Very impressive list!)
I'm surprised there also is an effort which uses Forth underway.
If that path should be seriously considered we should adapt
the directory structure within openec git to leave room for a
Forth implementation as well.
My personal (biased, not based on facts and definitely objectional)
view is that Forth has advantages during prototyping, then looses
with respect to C for "small to medium" sized projects and then might
win again in terms of code density for "large" sized projects.
(I'd position the EC within the "small to medium" sized projects)
We "just" have to choose tools. And if there's manpower behind
Forth then it can be done in Forth (and a few pieces of assembler:)
While it can be done I'm not sure whether it should be done.
> What I've got to play with ...
> 1. two A-Test PCBs,
> 2. three B1 systems,
> 3. three B2 systems,
> 4. two B4 systems (due within the week),
> 5. an SPI reflashing serial dongle.
I'm slightly jealous but surely won't admit I am!
PS: There's little traffic on this list so I claim (to avoid duplicate work:)
I "own" the C-implementation for the 3x3 matrix for 7 days from now on.
More information about the Openec