Object-oriented Forth, Neon/Yerk, etc.
ForthNet articles from GEnie
Message 144 Sat Mar 28, 1992
M.HORE [Mike] at 23:48 EST
Hi all -- or maybe hi: all would be better :-)
I'm Mike Hore, the author of Mops, which has been mentioned a few times on
this topic (see esp. msg 1. That's right, *one*. It was apparently my letter
to Bob Loewenstein in March 90 which via several levels of indirection led to
this topic being started!).
GEnie has just arrived in Australia, so after a long electronic isolation,
I've been able to get on the Forth RT at last. I'm happy to answer any
questions about Mops, and am likely to throw in my $0.02 at any time in
defense of OOP, the Neon/Yerk/Mops way of doing things, and various other
subjects as well :-)
Mops is definitely still alive and kicking. It's now System 7 savvy and has
support for AppleEvents. Although at present it comes with a documentation
file listing the differences to Yerk/Neon, I'm currently revising the Yerk
manual to produce a proper Mops manual. Maybe this will be done in a couple of
months??
OOP doesn't have to take a speed penalty. In Mops, early-bound method calls
are almost as fast as straight Forth word calls. And they can be in-lined, in
which case they are really equivalent to hand-coded assembly. I've sent an
article to Joerg Langowski which he may publish in MacTutor (when it
reappears, in April hopefully), in which I show that Mops can beat LS FORTRAN
really comes from the Forth approach to things plus native code compilation
plus peephole optimization. I'm sure any native code Forth system could do
something similar. Who says FORTRAN is best for number-crunching?
Mops is available for anonymous ftp at the same site as Yerk:
oddjob.uchicago.edu. I don't know if it's available on GEnie or via email.
(If Doug Philips is reading this, could you possibly make it available for
email, if not done already? I'd be very grateful.)
I now want to take up a point raised by Mitch Bradley some time ago, that we
should be concentrating on developing tools etc. for existing OOFs rather than
trying to reinvent the wheel. We have Yerk/Mops now (and I believe the JForth
ODE isn't much different). The last two issues of Forth Dimensions have
presented articles with source code to implement a version of Yerk on the PC.
These work, they're being used for real applications, and they can all evolve
to become better. If we try to start over from scratch, I think we'll throw
away years of time for a very doubtful advantage. Moreover, Forthers being
the individualists they are, I doubt that the process would ever converge to
anything looking like a consensus. OOP is just too nebulous. Or if it did
converge, how long would it take? 5 years? 10 years?? Can we wait that
long?
But these are just my opinions, and a discussion about this would be more than
timely. If we take 5 years contemplating our navels in the OO area, the world
will pass us by (if it hasn't already). Let's at least do something
constructive! Yes, I know, it is constructive to be sure you're doing things
right before pushing ahead. My point is precisely the same as Mitch's, that
the systems in existence now are right enough already, and it's time to move
forward.
Cheers, Mike. (I'll include my full signature this time): ---------------
Michael Hore PO Box 821 | Internet: mi...@peg.pegasus.oz.au (best)
Nhulunbuy, | CompuServe: 100033,3164
NT 0881 | GEnie: M.HORE AUSTRALIA | ----------------
-----
This message came from GEnie via willett. You *cannot* reply to the author
using e-mail. Please post a follow-up article, or use any instructions
the author may have included (USMail addresses, telephone #, etc.).
Report problems to: d...@willett.pgh.pa.us
ForthNet articles from GEnie
Message 145 Sun Mar 29, 1992
E.RATHER [Elizabeth] at 01:51 EST
Welcome abord, Mike.
Several people have commented lately on the lack of any articles in OOP
publications re OOPs work in Forth. You sound like an ideal person to remedy
this grievous deficiency. Hope you can help!
Cheers...
James P. Norris
I saw a show on PBS the other night about the history of the
computer and up to that time I thought it was better to keep
reaching for the best idea and the best way of doing something.
Then I learned that the person who first came up with the idea
for a computer never finished it. He ran out of funds and died
before it was completed. Only now over a century after it was
concieved can we even begin to appreciate his idea.
This man (Charles something or other) was working on a
difference adding machine and stopped working on it to build
a mechanical computer. An infinitely better device, granted.
However, his funders got impatient and neither useful device
was built.
Eniac however was built and the man most responsible for it's
completion had an interesting comment. He said that by the time
they had finished working on Eniac they had better ideas on how
to do everything but at some point they had to draw the line
and finish working on what they were doing so that they would
have something that would work. Otherwise Eniac would have never
been built.
My point is this. I am a Forth programmer programming in
Multi-Forth. i have just finished a really nice string handling
package for it. I wish I could have spent more time working
on finished applications. Why can't the existing Forth's be
expanded and the range of current Forth applications be expanded
to include just about everything a modern programmer would need
instead of trying to come up with a whole new standard? And once
that is done work on the new ideas? Or are the old Forth standards
so horrible no-one could possibly use them productively?
Up until this program I was always chasing the new idea and
not finishing anything. From now on though I think I will focus
my efforts a little bit more on short term goals so that
others will be able to recognize my talents before I am dead.
--
Doug Philips
M.HORE [Mike] writes:
+Mops is available for anonymous ftp at the same site as Yerk:
+oddjob.uchicago.edu. I don't know if it's available on GEnie or via email.
+(If Doug Philips is reading this, could you possibly make it available for
+email, if not done already? I'd be very grateful.)
Ok. It is now available for email access from FNEAS, ask for file: mops.cpt
As for GEnie, as you are both the author, and on GEnie, I'll leave it
to you to upload it there.
-Doug
If you would like a copy from FNEAS, send a message to:
and put the following in the _body_ of your message:
send file-name-not-number-just-name
path your-Internet-relative-email-path
Notes: You need one send command per file.
You need only one path comman per message sent to fneas.
Your email path _must_ be Internet relative: ...@host.doo.dah
us...@hostname.uucp and f...@hostname.bitnet are _not_ Internet
relative email paths. If you don't know how to construct your
Internet relative email path, please consult your system's
postmaster or administrator.
---
Preferred: d...@willett.pgh.pa.us Ok: {pitt,sei}!willett!dwp