h***@gmail.com
2019-12-05 05:49:25 UTC
This is a response to a post in this thread:
https://groups.google.com/forum/#!topic/comp.lang.forth/BDkprcLYfGg
I started a new thread because that thread
(discussing the concept of smudging CODE definitions)
is very novice-level and uninteresting.
long. fig-FORTH itself was published in mid-1979, and dave Kilbridge's
fix for FORGET was written at end '79 and published in FORTH
Dimensions March 1981. Maybe people didn't update for a long time.
Andrew.
Who knew. Responding to a letter in FD asking why Fig wasn't issuing
bug fixes they said it was now the responsibility of vendors. They
saw themselves in the new role of facilitating standards. The final
fig-6502 listing didn't even include fixes that had made it into the
Fig Installation Manual.
According to Jeff Fox, Elizabeth Rather wanted to sue Fig because
Fig had stolen MicroForth (proprietary to Forth Inc.) and made it
public --- this was blatant pirating --- the lawsuit would have worked.
This resulted in a major disagreement between Elizabeth Rather and
Charles Moore. He was willing to talk it over with Fig and make a deal.
He did this, and the deal was that Fig would just stick with FigForth
that they already had, but they wouldn't upgrade FigForth because
that would be competition against Forth Inc.. This deal is why FigForth
never got upgraded, not even with bug fixes --- it stayed static
forever and continued to be of educational value and nothing more.
Some bug-fixes and minor feature upgrades got published in
"Forth Dimensions" and/or passed around, but Fig themselves didn't
provide any bug-fixes or upgrades, nor endorse any PD ones available.
Charles Moore made this deal because he felt confident that
his future Forth systems would be superior enough to MicroForth/FigForth
that they would be worth commercially viable. He was okay the rise
of a Forth community that didn't necessarily buy Forth Inc. products
because he felt confident that he could always come out with new
Forth systems that would be commercially viable in the sense that
they were technically superior to any Forth system that anybody else had.
Elizabeth Rather was adamantly opposed to the rise of a Forth community
that didn't necessarily buy Forth Inc. products. She wanted the Forth
language to be proprietary to Forth Inc. and to require (with lawsuits)
every Forth programmer to buy a Forth system from Forth Inc., and
nobody could be a Forth programmer without paying her for the privilege.
This disagreement between Charles Moore and Elizabeth Rather was the
reason why Charles Moore left Forth Inc.. After Charles Moore left,
Elizabeth Rather didn't have any real programmers employed at Forth Inc.
so all she could do was employ maintenance programmers whose job was to
keep Charles Moore's old Forth systems going (like keeping a corpse warm
so that appears to still be alive), but there was no more innovation.
Multiple companies came out with Forth systems for sale.
Elizabeth Rather couldn't sue them for stealing Forth Inc.'s
"look and feel" because Charles Moore was the inventor of Forth
so he owned any "look and feel" that Forth had. This was the era
of Apple's "look and feel" lawsuit, so the inventor of Forth could
have realistically sued Forth system providers for stealing his ideas
(Forth was full of ideas that Charles Moore had thought up, that had
never been thought of previously). Charles Moore continued
to decline to hit Forthers with lawsuits --- he liked Forthers or,
at least, he didn't hate them the way Elizabeth Rather hated them.
This continues today --- Charles Moore doesn't hate Forthers, but
there is no evidence to indicate that he particularly likes Forthers
either --- he seems to consider us to be a gaggle of idiot wanna-bees
who continue to fail badly at innovation, just like in the 1970s.
LMI became the dominant Forth system provider of the late 1980s
and early 1990s. The quality level of UR/Forth was really superior
to anything that any other Forth provider offered. Forth Inc.'s
quality level was abysmal --- many public-domain Forth systems
were of higher quality. In those days, being a Forth programmer
mostly meant having familiarity with UR/Forth --- when I interviewed
for the job at Testra, the entire interview involved showing
UR/Forth code that I had written (my 65c02 cross-compiler) and
answering questions about the internal workings of UR/Forth.
Elizabeth Rather wasn't ready to surrender though!
Elizabeth Rather came up with the clever trick of developing the
ANS-Forth Standard primarily for the purpose of making UR/Forth
non-standard and forcing all Forthers to get on their knees to
beg her to allow them to call themselves Forthers. That worked!
Forth Inc.'s quality level was still abysmal --- and it still is!
It is not about quality --- it is about cutting throats.
This is why I describe Forth today as being similar to the Donner Party.
I can remember in 1984 that Forth was popular. Finding Forthers was easy;
I just rode my bike around Isla Vista near the college campus and
looked for cars with a bumper stick that said: "Forth [heart] if honk then."
I said at the time that Forth-83 had problems, with PAD being especially
dumb (I knew that passing data between functions in global variables was dumb
because this was the primary problem with line-number BASIC). I was told
that everybody knew Forth-83 was weak, but that a viable Forth standard
was in the works and would soon be available, and this would make Forth
respectable and competitive with C. After ANS-Forth came out one decade
later (very late!), ANS-Forth was found to be full of ambiguity and hence
worse than Forth-83 rather than better. This killed Forth.
Essentially all of the Forthers disappeared. Where did they go?
Elizabeth Rather killed them and ate them --- a modern-day Lewis Keseburg!
https://groups.google.com/forum/#!topic/comp.lang.forth/BDkprcLYfGg
I started a new thread because that thread
(discussing the concept of smudging CODE definitions)
is very novice-level and uninteresting.
I think most apps used EMPTY to clean out an aborted compile. It
was more reliable than FORGET which was buggy in Fig-Forth.
I'm surprised by that. fig-FORTH FORGET wasn't buggy for verywas more reliable than FORGET which was buggy in Fig-Forth.
long. fig-FORTH itself was published in mid-1979, and dave Kilbridge's
fix for FORGET was written at end '79 and published in FORTH
Dimensions March 1981. Maybe people didn't update for a long time.
Andrew.
bug fixes they said it was now the responsibility of vendors. They
saw themselves in the new role of facilitating standards. The final
fig-6502 listing didn't even include fixes that had made it into the
Fig Installation Manual.
Fig had stolen MicroForth (proprietary to Forth Inc.) and made it
public --- this was blatant pirating --- the lawsuit would have worked.
This resulted in a major disagreement between Elizabeth Rather and
Charles Moore. He was willing to talk it over with Fig and make a deal.
He did this, and the deal was that Fig would just stick with FigForth
that they already had, but they wouldn't upgrade FigForth because
that would be competition against Forth Inc.. This deal is why FigForth
never got upgraded, not even with bug fixes --- it stayed static
forever and continued to be of educational value and nothing more.
Some bug-fixes and minor feature upgrades got published in
"Forth Dimensions" and/or passed around, but Fig themselves didn't
provide any bug-fixes or upgrades, nor endorse any PD ones available.
Charles Moore made this deal because he felt confident that
his future Forth systems would be superior enough to MicroForth/FigForth
that they would be worth commercially viable. He was okay the rise
of a Forth community that didn't necessarily buy Forth Inc. products
because he felt confident that he could always come out with new
Forth systems that would be commercially viable in the sense that
they were technically superior to any Forth system that anybody else had.
Elizabeth Rather was adamantly opposed to the rise of a Forth community
that didn't necessarily buy Forth Inc. products. She wanted the Forth
language to be proprietary to Forth Inc. and to require (with lawsuits)
every Forth programmer to buy a Forth system from Forth Inc., and
nobody could be a Forth programmer without paying her for the privilege.
This disagreement between Charles Moore and Elizabeth Rather was the
reason why Charles Moore left Forth Inc.. After Charles Moore left,
Elizabeth Rather didn't have any real programmers employed at Forth Inc.
so all she could do was employ maintenance programmers whose job was to
keep Charles Moore's old Forth systems going (like keeping a corpse warm
so that appears to still be alive), but there was no more innovation.
Multiple companies came out with Forth systems for sale.
Elizabeth Rather couldn't sue them for stealing Forth Inc.'s
"look and feel" because Charles Moore was the inventor of Forth
so he owned any "look and feel" that Forth had. This was the era
of Apple's "look and feel" lawsuit, so the inventor of Forth could
have realistically sued Forth system providers for stealing his ideas
(Forth was full of ideas that Charles Moore had thought up, that had
never been thought of previously). Charles Moore continued
to decline to hit Forthers with lawsuits --- he liked Forthers or,
at least, he didn't hate them the way Elizabeth Rather hated them.
This continues today --- Charles Moore doesn't hate Forthers, but
there is no evidence to indicate that he particularly likes Forthers
either --- he seems to consider us to be a gaggle of idiot wanna-bees
who continue to fail badly at innovation, just like in the 1970s.
LMI became the dominant Forth system provider of the late 1980s
and early 1990s. The quality level of UR/Forth was really superior
to anything that any other Forth provider offered. Forth Inc.'s
quality level was abysmal --- many public-domain Forth systems
were of higher quality. In those days, being a Forth programmer
mostly meant having familiarity with UR/Forth --- when I interviewed
for the job at Testra, the entire interview involved showing
UR/Forth code that I had written (my 65c02 cross-compiler) and
answering questions about the internal workings of UR/Forth.
Elizabeth Rather wasn't ready to surrender though!
Elizabeth Rather came up with the clever trick of developing the
ANS-Forth Standard primarily for the purpose of making UR/Forth
non-standard and forcing all Forthers to get on their knees to
beg her to allow them to call themselves Forthers. That worked!
Forth Inc.'s quality level was still abysmal --- and it still is!
It is not about quality --- it is about cutting throats.
This is why I describe Forth today as being similar to the Donner Party.
I can remember in 1984 that Forth was popular. Finding Forthers was easy;
I just rode my bike around Isla Vista near the college campus and
looked for cars with a bumper stick that said: "Forth [heart] if honk then."
I said at the time that Forth-83 had problems, with PAD being especially
dumb (I knew that passing data between functions in global variables was dumb
because this was the primary problem with line-number BASIC). I was told
that everybody knew Forth-83 was weak, but that a viable Forth standard
was in the works and would soon be available, and this would make Forth
respectable and competitive with C. After ANS-Forth came out one decade
later (very late!), ANS-Forth was found to be full of ambiguity and hence
worse than Forth-83 rather than better. This killed Forth.
Essentially all of the Forthers disappeared. Where did they go?
Elizabeth Rather killed them and ate them --- a modern-day Lewis Keseburg!