LMI Forth (was: porting an MS-DOS Forth to a 386 Unix/Xenix) 
Author Message
 LMI Forth (was: porting an MS-DOS Forth to a 386 Unix/Xenix)

Actually the LMI Z80 Forth, along with the early PC-Forth, was a
considerable bargain.  You not only got a very clean FIG Forth, but
you also got lots of clear and well commented example programs and a
manual that was far and above better than anything out there I ever
saw.  Especially noteworthy was the explanations of Forth iteration
constructs.

Although I ended up using a version of Fig-Forth that I wrote myself
for my rather large commercial graphics application, the money I
spent on Ray Duncan's LMI offerings was *very* well spent for the
education it gave and for the utilities such as the very complete and
robust assembler that was included.

Alas, the current LMI Forth documentation, while complete, does not
include an upgraded version of that excellent overview documentation
& tutorial.  Can't say why.  

What I can say though is that UR/Forth currently offered by LMI is an
amazing product.  It uses a proprietary hashing technique for FIND,
which makes it compile even monster programs in a blink.  The 386
version now uses Phar Lap's VMM for state-of-the-art virtual memory
management: never again run out of dictionary space.  

My original intention was to again use LMI Forth, along with Dr.
Ting's updated book and the reams of PD Forth source to roll my own,
but the robustness and speed and compatibility between 32-bit 386 and
16 bit 8086 versions, not to mention LMI's long-term (10 years in
this industry is about as good as it gets) commitment to maintaining
and upgrading the product convinced me to put my time into the
application and let LMI to the Forth engine.

The question of why pay more than $50 for a commercial Forth is along
the lines of why pay for good tools.  Answer: if you are a
professional, the money you invest in your tools is trivial compared
to the value of your time spent using them.  This is so obvious it
got me to wondering why it was even asked.  I think it might have
something to do with the issue of why do people use Forth at all when
the professional programming community is now so well supported by C.

I personally have come full circle with Forth: a decade ago I wrote a
Forth application that grew into a company.  When it came time to
write Version 2 of the application, it was done to my specifications
by the company's programming staff.  And it was written in C, not
Forth.  Now I have left the company to once again write original
applications solo and - after a year of mushing about in various
Lisp's - I am back to writing in Forth.

Thinking is greatly affected by the language we think in, and to no
small extent is limited by that language's limitations.  Procedural
languages such as C and Pascal and (horrors) Ada are just the ticket
when you are following a blueprint, or when you are hiring other
people to follow your own design.  But if you are doing original,
creative work you need the malleability of a self modifing language
such as Forth (or Lisp or Smalltalk).  Forth ends up being the best
of these because you can easily strip away the language support and
have the application running on its own, lean and mean.

So my answer to "Why Forth?" is, if the project is well defined so
that there is little creative design work involved, Forth is *not*
the best language.  You need something that narrows your focus, that
works to keep you from coloring out of the lines.  Barring other
considerations (such as military job specs), C is probably the
language of choice for such a project.

But if you are doing creative programming, where the end result is
being defined in an iterative process as you write and rewrite, Forth
seems to be the best (although far from perfect) choice.  Perhaps this
is why Forth seems to appeal to the fringe crowd: creative people
have always been considered fringe.  It goes with the territory.



Wed, 19 May 1993 14:40:00 GMT  
 
 [ 1 post ] 

 Relevant Pages 

1. porting an MS-DOS Forth to a 386 Unix/Xenix

2. Porting an 8088/MS-DOS FORTH to 386 Xenix

3. forth for sco xenix 386

4. 386 MS-DOS port of SB-Prolog available for ftp

5. 386 Assemblers in Forth for Forth

6. Searching for Forth on 386/Unix

7. Anybody ported KCL to SCO XENIX 386?

8. UNIX vs. MS-DOS Forth

9. Changing from DOS to XENIX and from XENIX to DOS

10. IBM370 emulator for 386 running MS-DOS

11. IBM370 emulator for 386 running MS-DOS

12. looking for a PL/I compiler for 386+ in MSDOS or NT-DOS

 

 
Powered by phpBB® Forum Software