On Mar 15, 9:22 pm, John Doty <j...@[EMAIL PROTECTED]
> wrote:
> Well, in this role I see Forth as more like an RTL with sup****t for
> interactive, incremental development. But you mean Forth 94? There's an
> awful lot of baggage and ambiguity there. Anyone who isn't a really
> expert language lawyer likely would only be able to make it work on
> their development implementation. But, if it *could* be done ****tably
> that would automatically plug it into all of the work that's been done
> on optimizing Forth compilers, a very nice advantage.
I don't understand the term "****tably" as used in the segue
"would only be able to make it work on their development
implementation. But, if it *could* be done ****tably"
First you get something that works for one implementation. Then you
****t it to another implementation, while keeping it working on the
first, which is a mix of finding more ****table ways to say things,
moving system dependencies into distinct preface filed, and
conditional compilation. Repeat until you have all the many systems
covered that you have targeted.
If one of the systems have a difference to your original development
system that is too extreme for you to work out how to code ****tably
across the gap, move it from the first tier list to the second tier
list and make a remark in the release noting where you had problems in
getting it to work on that implementation. Maybe someone who prefers
that system ... or even the implementor ... can see how to bridge the
gap.
If its a big project, you do the ****ting at periodic intervals as you
hit specific signposts on your roadmap.
IOW, its like ****tability in most other general purpose languages
found on a variety of operating systems.
And the less ambitious your target range, the fewer hurdles to cross.
And, of course, most of the hurdles will be at the lower levels, since
as you approach the factors that constitute the main operating
processes of your target language, you will be working with more and
more words that are in terms of the problem area.
|
