/FONT#1=BASL30/FONT#2=BASB30/FONT#3=NGR25/FONT#4=NGR20
␈↓↓␈↓α␈↓β␈↓∧␈↓α␈↓ ¬→STANFORD UNIVERSITY
␈↓ ↓H
␈↓β␈↓ ¬_STANFORD, CALIFORNIA 94305
␈↓ ↓H
␈↓∧COMPUTER SCIENCE DEPARTMENT␈↓ 
+   Telephone:
␈↓ ↓H
␈↓ 
+415-497-4971
␈↓ ↓H
␈↓↓␈↓ ε¬JUl 23, 1974
␈↓ ↓H

␈↓ ↓H

␈↓ ↓H

␈↓ ↓H

␈↓ ↓H
Dear Mr Schorer:
␈↓ ↓H

␈↓ ↓H
␈↓ ↓HDon't␈αgive␈α
up;␈α
I␈α
still␈α
do␈α
exist.␈α
The␈α
problem␈α
is␈α
to␈α
get␈α
manuscript␈α
finished␈α
as␈α
soon␈α
as␈α
possible.␈α
I'm␈α
glad
␈↓ ↓H
␈↓ ↓Hyou␈α⊃are␈α⊃getting␈α⊃some␈α⊃good␈α⊃from␈α⊃text,␈α⊃though␈α⊃the␈α⊃version␈α⊃you␈α⊃have␈α⊃is␈α⊂very␈α⊂much␈α⊂outdated.␈α⊂In
␈↓ ↓H
␈↓ ↓Hparticular␈αlarge␈α
revisions␈α
to␈α
incorporate␈α
Scott's␈α
semantics␈α
as␈α
applied␈α
to␈α
LISP␈α
have␈α
been␈α
incorporated.
␈↓ ↓H
␈↓ ↓HBy␈α
the␈α
way␈α
I␈α
would␈α
like␈α
to␈α
know␈α
where␈α
your␈α
copy␈α
came␈α
from;␈α
they␈α
copies␈α
seem␈α
to␈α
by␈α
multiplying.
␈↓ ↓H

␈↓ ↓H
␈↓ ↓HOn␈αfixed-points:␈αthey␈αare␈αbasically␈αsolutions␈αto␈α(functional)␈αequations.␈αThink␈αof␈αfact␈α<=␈αλ[x][x=0;␈αT
␈↓ ↓H
␈↓ ↓H→␈α⊂x*fact[x-1]]␈α⊂as␈α⊂an␈α∂equation␈α∂in␈α∂fact.␈α∂A␈α∂solution␈α∂(i.e.␈α∂fixed␈α∂point)␈α∂is␈α∂a␈α∂function.␈α∂There␈α∂are␈α∂good
␈↓ ↓H
␈↓ ↓Hreasons␈α
why␈α
you␈α
should␈α
think␈α
of␈α
<=␈α
in␈α
this␈α
fashion.␈α
Book␈α
will␈α
explain␈α
why.
␈↓ ↓H

␈↓ ↓H
␈↓ ↓HAs␈α⊂to␈α⊂proving␈α⊂assem.␈α⊂prog.:␈α⊂I␈α⊂think␈α⊂it␈α⊂is␈α⊂a␈α⊂step␈α⊂in␈α⊂to␈α⊂wrong␈α⊂direction.␈α⊂Being␈α∂able␈α∂to␈α∂SPECIFY
␈↓ ↓H
␈↓ ↓Hwithout␈αPROGRAM␈αis␈αmore␈αuseful␈αand␈αprobably␈αmore␈αfeasible.␈αThere␈αare␈αtoo␈αmany␈αdifficulties␈αin
␈↓ ↓H
␈↓ ↓Hproving␈α⊂properties␈α⊂of␈α⊂assemb.␈α⊂progs.␈α⊂since␈α⊂they␈α⊂have␈α⊂almost␈α⊂no␈α⊂inherent␈α∂structure;␈α∂c.f.␈α∂argument
␈↓ ↓H
␈↓ ↓Hagainst␈α
goto's.␈α
It␈α
applys␈α
"in␈α
spades"␈α
to␈α
assembly␈α
language.
␈↓ ↓H

␈↓ εW

␈↓ εW
Yours sincerely,
␈↓ εW

␈↓ εW

␈↓ εW

␈↓ εW
John R. Allen
␈↓ εW
Research Associate
␈↓ εW
Computer Science Dept
␈↓ εW
Artificial Intelligence Labs
␈↓ εW

␈↓ εW

␈↓ ↓H