On Mon, 21 Jul 2008, Elizabeth D Rather wrote:

Elizabeth, with no offence intended, you seem to tend to reflexively
defend the commercial embedded marketplace against upstart open source
projects, but quite frankly, I don't see the conflict. What Mr. Davis is
proposing is large, and bigger than "HIM", and/or his skills (currently),
but sometimes the destination is not as important as the journey itself.

Is there a place for an open source Forth based OS in the embedded hobby
market (Palm/IPAQ/PDA/CELLPPHONE/legacy PC/ARM dev board etc.)? Yes, I
believe that there is. Can such an OS be useful? I think we have seen
both the BSD and Linux "hobby" market can indeed be useful in the PC
space, and given the emergence of "open" cell phones, and hacked PDA's and
such, the embedded market offers similar opportunities (only moreso).

For lack of such an alternative, Linux has IMHO, been forced into the
embedded marketplace in spite of being difficult to port, overly complex
and resource intensive. Why has Forth not dominated in the market space
where it is uniquely suited, and indeed shines like no other?

I suspect that the best approach that commerical Forth houses can take to
Forth based open source projects is to toss them a bone ... how about
"open sourcing" an ANS 94 IP stack? Perhaps recommending an interrupt
architecture? How about tossing a device driver or two over the "wall"?
Publishing some boot specs for various chips? Rather than admonish the
newbie hobbiest of the difficulties that they face, why not encourage (by
contributing to) a disruptive movement which can only put Forth under the
spotlight, and increase the commercial exposure?

Mitch's OpenFirmware was in the right place at the right time, but was
quickly supplanted in the commercial space by more run of the mill
language based boot/BIOS'es ... simply because Forth skills are hard to
find, whereas C skills are relatively common. Pity that!

It occurs that of all people, Elizabeth Rather (the Grande Dame of Forth)
should be doing whatever it takes to move Forth technologies out of the
shadows and into the spotlight. Sure, Forth offers a competitive
advantage in the embedded space, but it has never been a *secret*, and
what do either Forth Inc or MPE have to lose from an open source Forth
based OS (precious little, I'd hazard, and much to gain, I suspect).

Thus far, the RPN and stack based stuff has tended to scare away the great
unwashed of mainstream programmers, but perhaps now the time is night to
have the mainstream join the Forth "movement", at least in the embedded
space ... if not the hosted OS space, and would be enthusiasts and
hobbiests like Charles Davis are to be encouraged ... and assisted
wherever possible, not dissuaded from even starting!

Cheers,
Rob Sciuk