perm filename MAR.MSG[2,TES] blob sn#032110
filedate 1973-03-28 generic text, type T, neo UTF8
What is the current state of PUB?
28-MAR-73 0751 XGP,TVR
I WAS RECOMPILING COMP TO INCLUDE THE NEW DECS/RECS AND GOT AS
FAR AS COMP.DO(7) -- COMPILING INTRIN -- WHEN IT HIT AN ERROR:
"*** ERROR ***, WRONG NUMBER OF ARGUMENTS TO IDENTIFIER: ((PNAME PRIVATE
(STRING)) (PROPERTIES PRIVATE (LIST))) ≠ (0 0 0 NIL)"
IT'S IN THE FUNCTION AFTER TTYUUO, AND MAY BE YOUR OR MY BUG
(BUT I'M TOO TIRED TO TRACK IT DOWN). AT ANY RATE, COMP.DMP[L70,TES]
IS UPDATED, BUT COMP2.DMP[L70,DAV] IS NOT!!
28-MAR-73 0008 L70,DAV
I'M FIXING UP THE DECISION POINT ROUTINES TO BE THE REAL THINGS, BUT
I'M CONFUSED ABOUT PUBLIC VARIABLES. HOW AND WHERE ARE THEY STORED? THEY
SEEM TO BE STORED AS ORDINARY RECORDS, WHICH CAN'T BE RIGHT.
THE CODE FOR REP'S IN REWRITES IS GRADUALLY GETTING COMPLETED,
BUT THERE SURE IS A LOT OF IT!
27-MAR-73 2323 L70,DAV
Any thoughts on text/graphics formats. We're working on our
"new"xgp system and are using that as a forcing function to work
on that problem. We'd very much like to hear from you.
How's life at PARC?
27-MAR-73 1112 LDE,DCS
Consider the following PUB macro:
.MACRO PAR (TEXT)
.⊂ SOME CONTROL STATEMENTS;
.CONTINUE ETC. ⊃
If I want underlines to go all the way across underneath TEXT, I
have to put "_" chars in for " " throughout TEXT. This means changing
all calling sequences if I change my mind about the underlining later.
If you agree, praps we should put a feature like this in the list of
good PUB things to do.
23-MAR-73 1957 SLS,DCS
LOOK AT MLISP.L70 IF YOU WANT TO SEE A (NEARLY) COMPLETE
VERSION OF THE MLISP→LISP TRANSLATOR. THERE ARE STILL A FEW FEATURES
I'M OFF TO SKI THIS WEEKEND; SEE YOU MONDAY. HAVE A NICE WEEKEND.
23-MAR-73 1249 L70,DAV
LARRY: THIS IS RICH JOHNSSON. I MANAGED TO LOSE YOUR LAST MESSAGE,
BUT I'LL TRY TO ANSWER THE QUESTIONS I REMEMBER.
OUR NEWEST VERSION OF PUB DOES XGP JUSTIFICATION BY THE SUBSTITUTION OF
VARIABLE LENGTH BLANKS (XGP ESCAPE CHARACTERS) FOR THE SPACES AT WORD
BREAKS. THIS WORKS FINE FOR SINGLE COLUMN AREAS, I'M NOT SURE IT WILL
WORK FOR MULTIPLE COLUMNS(I HAVEN'T TRIED). IT ALSO USES
VARIABLE LENGTH BLANKS FOR CENTERING AND FLUSH RIGHT. ∀TABS ARE
CONVERTED TO "TAB TO RASTER POSITION".
THE PROBLEM WITH FORWARD REFERENCES IS RELATED TO VARIABLE WIDTH CHARACTERS.
CURRENTLY PASS TWO DOES NOT KNOW THE WIDTHS OF THE HARACTERS, SO KNOWING
THE LENGTH OF THE STRING IS OF LITTLE USE WHEN TRYING TO JUSTIFY. I HAVE
MODIFIED PASS ONE TO PUT OUT THE SHORTM IN RASTER POINTS INSTEAD OF SPACES.
I THINK THE PROBLEM IS FIXED FOR FORWARD REFERENCES IN THE TEXT, BUT
THERE IS STILL A PROBLEM WITH FORWARD REFERENCES TO BE CENTERED OR
FLUSHED RIGHT. I HOPE TO FIX THAT THIS WEEK. ALSO THE ?)X FACILITY IS
NOT SUPPORTED BECAUSE OF THE DIFFICULTY WITH VAR. WIDTH CHARS.
THANKS FOR THE COMMENTS ABOUT APARTMENTS, ETC. I WILL WANT A PLACE
BY MYSELF. AN EFFICIENCY OR ONE BR. RENT OF $150-$175 SHOULD NOT BE A
PROBLEM. IS THE SITUATION SUCH THAT I CAN PROBABLY FIND SOMETHING
IN A COUPLE OF DAYS, OR SHOULD I ARRANGE FOR SOMETHING BEFORE I GET THERE?
I'LL GET BACK TO YOU IN A DAY OR TWO ABOUT A CONVERSATION.
12-MAR-73 1938 NET,GUE
THOUGHTS BEFORE ANSWERING RICH:
It seems that PUB2 can do the final justification even
for forward-references; I don't understand the problem.
Pass 1 could provide with each label value its raster width;
this combined with the knowledge of SHORTM in raster
units is enough to compute the amount of space needed. No?
This is Rich Johnsson. A standard version sounds like a good
idea, but I have very little time to work on it. I'm trying to get
the final changes in to let PUB handle all the XGP justification
and related things. Big problem at the moment is with forward references.
When I get that done I don't think I'll be able to do much more.
I'll be coming to XEROX about 15-May. By the way, I'll need
some place to stay 15-May to 15-Aug. Any suggestions you have will
01-MAR-73 1332 NET,GUE
REWRITES WORK COMPLETELY!! (AT LEAST THOSE TESTED IN RTEST.L70)
WHY DON'T WE GET TOGETHER AND LOOK AT WHAT HAS TO BE DONE TO THE REWRITE
VERSION OF THE LISP COMPILER. (I UPDATED ITS FORMAT SOMEWHAT -- IT'S ON
LISP.L70.) ALSO IT SHOULDN'T BE TOO HARD, AND IT WOULD CERTAINLY BE NICE,
IF WE COULD AT LEAST READ FROM THE TELETYPE. ARE YOU STILL GOING TO
IMPLEMENT THE FULL READ?
28-FEB-73 2310 L70,DAV