Am 19. Juli 1999, Montag Morgen in der Zoka unweit des Honig Baers.
Today after I do this Grauschrift here, I need to return to Vaierre
and try to finish a complete re-write of Amiga Mind.Forth-28 in F-
PC Forth.
Unable to sleep, I logged on early this morning and I discovered
that Name Withheld of Enterprise in Country had sent me a really
neat translation of my 15jul99A version of F-PC Mind.Forth into
32-bit iForth.
Today I will print out the version of Name Withheld and use it as
a guide when I work on F-PC Mind.Forth.
Am 20. Juli 1969+30, Dienstag Morgen in der Zoka unweit des Honig
Baers. Oh, what a grueling, stark-crazy-madman ordeal I went
through yesterday at the computer keyboard!
Now I have sat here for about fifteen minutes just reading over
my print-out of the code for Mind.Forth that Name Withheld sent
me by e-mail two days ago. Name's code looks really nice, really
beautiful, much more professional than my own amateurish efforts.
I did go back to Vaierre yesterday morning and I did use the Amiga
1000 with dot-matrix printer to print out the beautiful code from
Name Withheld.
Then I fired up the IBM-clone Compaq Presario and I started to
port the entire Amiga Mind.Forth into F-PC Forth, but things
started to go wrong from the very start.
Oh, first I studied and learned how to move around within F-PC
Forth, because I knew that my Mind program was going to get large
and unwieldy. The F-PC search function works like a charm.
The first disaster occurred when I had typed in the complete list
of Amiga Mind.Forth variables. The program would not run, and I
could not figure out which variables the system was objecting to.
I had to start all over again, type in each variable, and then run
the program to see if the system would accept each incremental
variable.
Then I tried to change the name "MEMCHAN" of the memory channel
array code to simply "CHANNEL" and my entire system went dead.
I thought, oh no, "CHANNEL" must be some special word deep within
Windows 98. The computer told me that my system disk was now
corrupt, and I began searching in vain for where I stashed the
back-up CD-ROM. A deep gloom settled over me as I realized that
I had promised my AI to the world and now I had ruined my
irreplaceable computer (irreplaceable because I would never
actually go out and buy a Microsoft computer; I had to get it
gratis from Free-PC.com).
So I took my floppy diskette out of drive A and I poked around on
the keyboard in a miserable hopeless try for a miracle. To my
surprise, it had been the floppy "files" diskette in drive A that
the computer had been complaining about. The Presario sprang
back to life in the respected Compaq tradition (ever since Compaq
beat IBM to market with a 386 machine) and it never dropped below
a simple reboot process for me -- except two times yesterday when
the Forth screen froze up so bad I had to turn the power off.
The Amiga Mind.Forth code went on for so many pages after pages
that first I coded it in from the front, then from the back, then
in the middle where the bootstrap sequences are.
When the IBM-clone source code had enough meat in it, I began
running the program to see where it would crash. I quickly put
in message-displays to show me exactly what Mind.Forth was doing
just before it crashed. Thus I had to chase the bugs (they are
BUGS, Microsoft -- not "issues"!) down deep into the subroutines.
Somehow it was deeply satisfying to force the program to reveal
its troubles to me.
The beautiful source code from Name Withheld lay untouched on the
table, because I was far too busy to stop and modify my idiosyn-
crasies.
After four or five hours of ceaseless programming, I began to
feel fatigued but I did not dare stop. I kept thinking, every
day is crucial, because if I can get this AI source code out on
the Web today instead of tomorrow, that's one more day of oppor-
tunity for the creative geniuses to work with it. And how can I
possibly stop to rest if I am not done yet?
So I made a really stupid mistake -- no, two of them. I had my
two most recent printouts of Amiga Mind.Forth at hand, and I
accidentally picked up the outdated printout and started coding
from it. Suddenly I realized that I was looking at some code
that had been eliminated a month ago. I had to look very
carefully to make sure that nothing obsolete had gotten into
the new F-PC code.
At each point yesterday when I achieved a major advance in
functionality, I moved through the letters A, B, C and D in
designating the versions that I saved to disk. My second
colossal blunder was to forget to keep coding version 19jul99D
and to slip back into coding version C. I was so annoyed at
myself that I erased the offending version C after I had
brought version D back up to date.
When I could see that I would finish the port by the end of the
day, I turned my Vaierre telephone (bought with gold and silver
found in Green Lake) back on because I no longer feared that
anybody would call up and prevent me from porting Mind.Forth.
The next thing I knew, one of my family members was on his way
over and whistling "Deutschland ueber alles" outside my window
so I could buzz him in. (Or should I say, "Glorious Things of
Thee Are Spoken," the hymn by Haydn?) Anyway, I kept him busy
reading The New York Times while I worked on the Mind.Forth
http://www.geocities.com/Athens/Agora/7256/mind-fpc.html Web page.
Finally on Usenet I posted message #98 in the Deja.Com thread,
"It's primitive; it's dumb; it's brittle--but it's AI."
