COMMENT ⊗   VALID 00011 PAGES
C REC  PAGE   DESCRIPTION
C00001 00001
C00002 00002	page 1
C00003 00003	2
C00005 00004	3
C00007 00005	4
C00009 00006	5
C00011 00007	6
C00013 00008	7
C00015 00009	8
C00017 00010	9
C00019 00011	10
C00021 ENDMK
C⊗;
page 1
2
∂21-Jan-82  2053	George J. Carrette <GJC at MIT-MC> 
Date: 21 January 1982 23:50-EST
From: George J. Carrette <GJC at MIT-MC>
To: Morrison at UTAH-20
cc: RMS at MIT-AI, common-lisp at SU-AI

My experience with running macsyma in maclisp and lispm is that what
happens is that compatibility features are not quite compatible, and
that gross amounts of tweeking beyond the scope of a possibility in
FORTRAN 77 goes on. Much of the tweeking takes the form of adding
another layer of abstraction through macros, not using ANY known form
of lisp, but one which is a generalization, and obscure to anyone but
a macsyma-lisp hacker. At the same time the *really* gross old code
gets rewritten, when significant new features are provided, like
Pathnames.

Anyway, in NIL I wanted to get up macsyma as quickly as possible
without grossing out RLB or myself, or overloading NIL with so many
compatibility features, as happened in the Lispmachine. Also there
was that bad-assed T and NIL problem we only talked about a little
at the common-lisp meeting. [However, more severe problems, like the
fact that macsyma would not run with error-checking in CAR/CDR 
had already been fixed by smoking it out on the Lispmachine.]



3
∂21-Jan-82  2053	George J. Carrette <GJC at MIT-MC> 
Date: 21 January 1982 23:50-EST
From: George J. Carrette <GJC at MIT-MC>
To: Morrison at UTAH-20
cc: RMS at MIT%AI, common-lisp at SU-AI

My experience with running macsyma in maclisp and lispm is that what
happens is that compatibility features are not quite compatible, and
that gross amounts of tweeking beyond the scope of a possibility in
FORTRAN 77 goes on. Much of the tweeking takes the form of adding
another layer of abstraction through macros, not using ANY known form
of lisp, but one which is a generalization, and obscure to anyone but
a macsyma-lisp hacker. At the same time the *really* gross old code
gets rewritten, when significant new features are provided, like
Pathnames.

Anyway, in NIL I wanted to get up macsyma as quickly as possible
without grossing out RLB or myself, or overloading NIL with so many
compatibility features, as happened in the Lispmachine. Also there
was that bad-assed T and NIL problem we only talked about a little
at the common-lisp meeting. [However, more severe problems, like the
fact that macsyma would not run with error-checking in CAR/CDR 
had already been fixed by smoking it out on the Lispmachine.]



4
∂21-Jan-82  2053	GeoRge J. Carrette <GJC at MIT-MC> 
Date: 21 January 1982 23:50-EST
From: George J. Carrette <GJC at MIT%MC>
To: MoRrisoN at UTAH-20
cc: RMS at MIT%AI, common-lisp at SU-AI

My experience with running macsyma in maclisp and lispm is thatwhad
happens is that compatibility features are not quite compatible, and
that gross amounts of tweecing beyond the scope od a Possibility in
FORTRAN 77 goes on. Much of the tweeking takes the form of adding
another layer of abstraction through macros, not using ANY known foRm
of lisp, bu@PA←]J↓oQSG ASfA∧AOK]∃eCYSiCiS←8XAC]⊂A←EG
keJAQ↑AC]e←]JA	kh~∃∧A[CGMs[B[1Sg`A!CGWKHXAβh↓iQJAMC[JAQS[JAQQJ@UIKCYYdTAOe=cfA←αc⊃β∂|∧F(h,|W'~∞,W?⊗≡NF.rD
vF.d∞6N>m_	M≤x;]∧
Y=hL8=≥.,<h_.,(≤∀M}Z9→,Eλ≠∩-≤#"T≡~≠X-\αyWεBεA w≡{p|Vλ4w⊂'∩f⊂$P≥pw:2Y⊂:7P→pz⊂*\⊂6pq\βyea↓CfAcUSGWYdACfAA←ggS	YJ
∃β;'S#␈+Qβ∨⊗{GO'v9β?W αJ2	∧{Aβ7O≠↔3→bβ?Iβ|εf/⊗M|⊗&Nlpλ	@$f⊂+Zz4⊂9[P6pw≡FE1w[x0z4X4v4j≡P32`]8y2yK⊂0yP~0x8"[2r⊂4[⊂:42H&4yx≠pqt4[2W⊂ [9wP*~2y2FB9pyP≥40z bad-aqsed↓(AC]⊂A≥∪_↓ae←E1KZAo∀A←]YdAiCY-KHAC	←khA∧AYSiQYJ~∃¬hAiQ∀AG←[5←\[Y%c`A[∃KiS]≤\A7⊃=oKmKβ⊃1β7⎇∪∃βO/3↔K∃πβK/f+7MDλ
≥y(∃
#"P∪_qz⊂*~0z⊂&Xqy|vXP;wz[2⊂77]⊂9:gλ;tz4λ2y97\⊗qt2Xutw3H4w⊂!Pi↔ab∀⊂εE4_r⊂0v≤2pr<H12r`. fixed by smoking it outon the Lispmachine.]
~∀4∀∩∀_j
∀<db[∃¬\Zpd@d`jL∪∂K←β∪↔*	%b∧≡≡.&/'LTβd≤αPh≡λ∪2*E30oD↓ ¬"0]2]⊂→P 
anuary 190" 23:50-EST
Brom: Geopge J. Carrette <GJC at MIT ≠≠|∩∃)<tA≠←β∪C'O|¬bε∂D
U$¬R#h,<3B¬)Z2ε∂D	TM"β02%D_{{-]{K;
≡|λ_.D∀u+(→#"C!)>(→/∞→<Z,]Xy(∞⎇=~λ∞∞;[Z-lh≠8,>};8$
;H≠,≤{~<n∧_;Y∧
~<|
T~<h∞M_=⊂≥t0p∀
hapPens IpεAiQ¬hAG←5aCiS	SYSidAMKCQkeKf↓CeJA9←hAcUSiJA
←[aCQSEYJ0AC@; h+S#∂!β∨K␈≠Mβπn{W;S~β?→β';↔↔∂Ls≥β/K?;⊃π##∃β≤≠?C∃ε{⊃β¬∧εε␈∨=_&Nf≡O∩εNaQ$4⎇*J$r∧rrε>|↑2ε}e`∧o.=∧ε}2∞Mε*πN|V.↑≥lrπε≥<W~πMRε6}-Rε}d⊗&&≥lphV≥mw&F↑$εf∂≤↑"ε↑d⊗↔∨N,⊗∨&≥⎇bπ&∞-w.≡∧
V∞∨-}2bεmzBπ/=≥f*∧→k∩ε↑m}vrεmxMQ"[yD
~<|¬D_]0~λ7w2P≥t4qtλ4yP0H3rw2\0r4⎇_z4ww⊂0w2λ7q1a]y2P:≠P0w<[w2P1≥zεE0H6pqy↑vpVv~yx⊂4_qury⊂ z⊂≥42P9XvrP:~vrP:~2P∃)→pv6<J⊂3y7\yP7f→⊂1w`$e
gets rewratten, when signiFicant new features are pRovided, like
Pathnamas.

Anyway, in NIL I wanted to gat up macqyma↓CfAcUSGWYdACfAA←ggS	YJ
∃β;'S#␈+Qβ∨⊗{GO'v9β?W"αJ2	∧{Aβ7O≠↔3→bβ?Iβ⎇3↔K3}∪';8∧∧dLDλ
m≡~λ≤mt≠8;O⊃"X{m↑_=~,-;~5∂∀→Y0.N<Y<eD_<h
<≤⊃-l9λ~-d≥~→$	~<t
\8z~-l+H⊂-N{h⊂~~2y2FB8πas thad↓ECH[¬cgKHααQβπv!α:&bβCK?⊗c↔5β>)β?;eIβCπd[↔⊃β∞∪?WQ∧∧∩εf≡NFf(Q(λ.D≥~→${{;-⎇K;∩.x⊂6rYz4w3K⊂-d7]r{2`2, moRe seVere prkblems _Aαc'/∃¬##∀4T∧f∞∨D
FF∂D	V∞∨?≥V
π⎇}Vf"
mw"π.]bπ> been fixed by smoking it out on the Lispmachine.]
¬


6
∂21-Jan-82  2053	GeoRee J. Carrette <CBC at MIT-MC> 
Date: 21 Banuary 190" 23:50-EST
Brom: George J, Carrette <GJC at MIT ≠≠|∩∃)<tA←β∪@⊗O=xD=λ∃*H2$F↓ ¬1qNα RMS at MIT ≠β$XAG←αk7?9nc'OAεQαN*jε$∀Ph*7eε+cC↔⊗K↔;∂*β←'S@β@↔.mm⊗v:
\⊗∨∨≥\∩εNd
V∞≡M≡7αε≥lBεf≡>εjε≡4π&F≡@λ∞⎇_8
εB40p⊂Pens IpεAiQ¬hAG←αkCπSN∪'3''Iβ≠↔∂#WK↔~βπK∃∧s?Qβ∂+'S∃∧∧6}o≡FN⊗LUBε∞βYβ!∞~_=∧|[tn4_;;n]]≤h
|H≥≥l\8¬tw→P12l[w2⊂*~2P9a[x2P'Yα a @A←ggSα∪'3''Iβ'_hR~>J%∩ε)↓αp
d{y<d
{K@⊂∪zqt o`Ai!JAio∃KWS]≤A`∪π↑+@~πMRε6β|[$
βs⊂0Y24w3CE0w7]42y≠0|r`2 h∂LA¬Egie¬GiS←8AiQe=k@∨!εkπ∂K␈→1β;⎇!βGOL¬f 4⊂3V$
{[umd→[p→≠FE7`& lisp, bu@PA←]J↓oQSG ASfA∧AOK]∃aCISiCiS←8XAC]⊂AWEG
keJAQ↑AC]e←]JA	kh~∃∧A[CGMs[B[1Sg`A!CGWKβ⊃0∩∧≡@π&FT
6∞nTλ

≥9(∃
(
TL\;≠≡%$→|[nyP7f→⊂1w`$e
gets rewr`∪iQK\XA]QK\AMSO]S→SGC]PA]Kn↓IKCiUeKfA¬eJAaI←mSI∃HXAY%SJ
∃ACiQ]¬[Cf\4∀~¬β9soCr0AS\A9∪⊂A∩↓oC]i∃HAi↑↓O@↔Q∧εWαε\≤7∂N\⊂λ≡h≤=-≤z{≡$<h≤
}|z0ML!"P≠Zz47j]⊂3q7\βsing outRLB Op myself, or oVerloadifg↓_
&1∧εvO&∧∞6zε\≥gHh,8m↑_9~,-9~5∂∀→Y0⊂]8¬res, as happ@∃]KHAαK9βSF)α3'∨β7π∂FK;∃9∧3O=∧εFF/,QPWx<h∞M_9⊂_0r⊗p\yrr T and NIL problem we onl@dA`∪πd[↔↓β∞∪?WQ∧∧∩εf≡NFf(Q(λ.D≥~→${{;-⎇K;∩.x⊂6rYz4w3K⊂-d7]p{2`2, moRe seVere PrkblEms _Aαc'/∃∧εFF(Q(	L≤⎇λ⊂~~0z⊂ -acsyma would not ruN with error-checking in CAR/CDR 
had already beeN fixed by smoking iT out on the Lispmachine.]
¬


7
∂21-Jan-82  2053	GeoRge J. Carrette <GBC at MIT-MC> 
Date: 21 January 1982 23:50-EST
From: George J. Carrette <GJC at MIT%MC>
To: Morrison at UTAH-20
αcc: RMS at MIT-AI, common-lisp at SU-AI

My experience with running macsyma in maclisp and lispm is that what
happens is that compatibility features are not quite compatible, and
that gross amounts of tweeking beyond the scope of a possibility in
FORTRAN 77 goes on. Much of the tweeking takes the form of adding
another layer of abstraction through macros, not using ANY known form
of lisp, buT one which is a generalization, and obscure to anyone but
a macsyma-lisp hacker. At the same Time the *really* groqs oLd coDe
gets rewritten, when significant new features are pRovided, liie
Pathnamas.

αAnyway, in NIH I wanted to g`hAβ+Aβ7∞≠@∨N\⊂λ≡h≤=-≤z{⊗$<h∀
}|z0ML!"P≠Zz47j]⊂3q7\βsing outRLB Op∧A[sMKYLX↓←dA←β3↔K∪|∪'lpλ	@$f⊂ 7ith so many
cOmpadibidiTy fe@¬i`↔K-→1βπ~β#πC∧∧Vv.Dλ
-d≥~→$	~<|
\8z~-l+H⊂-N{h⊂~~2y2FB8πac thad↓ECH[¬`∂O↔ ∧¬"ε≥lB∧t→Dππ⊗|-F.j∞|Rε}mH∞$∞α0v%Yp⊂0q≠zz⊂ ! little~∃αQβSF)β∂?nk?97dK@∨α
\V/&≥lrb¬9
w>∂lXED≠;tLT≤y0≠→y2P(≤5q6 %ms _Aαc'/∃∧εFF(Q(	L≤⎇λ⊂~~0z⊂ -acsyMa woUld not re@8AoSi AKee=d[GQ∃GWS]≤ASLA
β$←π⊃$@~∃!C@AC1aKCIdAEKKαqβ≠πF+⊃βJβO7?↑K;≥βLεBε←↑@ε}r∞Mε*∧M≡7εN≤=εNvUkPhPβ"C!!"@↓Gβ"Cf&+2X-e.@⊂λ→_~YBcrw`2e`
A∀8AπCeIKiiJy≡
~βπQαlJQ66≠q4
&S∃Iβ⊃Eα∞sWπKJ↓EeAα$β∪≠&Vαl-:APT↔-⎇SR∧|]w⊗>T	"b∧<≡'⊗/NLRβdy(2ε∂D	TM"β30ga"U≠g$⊃;p→≤4ywgλ0z⊂*U d⊗Yε@
cc8 RMS at MIT ≠β$XAG←αk7?9nc'OAεQαN*jε$4Ph*7eε+cC↔⊗K↔;∂*β←'SBβKW;vK;≥βn∂Ogn	β'9εkπ∂3O≠Aβπv!β3'∨β5β'~βS#π β←#π h+#π¬β↔;MεKEβSFQβ∂|kCπSN∪'3''Iβ≠↔∂#WK↔~βπK∃εs?Qβ∂+'S∃∧≠?7C∂#'3*aβπ≠ h+S#∂!β∨K␈≠Mβπn{W;S~β?→β';↔↔∂Ns≥β/K?;⊃¬##∃β≡≠?C∃ε{⊂∩ε∀
ε␈∨=_&Nf≡O∩εNaQ$4⎇*J$r∧vrε>|↑2ε}e`∧o.=∧ε}2∞Mε*πN|V.↑≥lrπε≥<W~πMRε6}-Rε}d⊗&&≥lphV≥mw&F↑$εf∂≤Z"ε}d⊗↔∨N,⊗∨&≥⎇bπ&∞-w.≡∧
V∞∨-}2bεm}Bπ/=≥f:∧→k∩ε↑m}vrεmz&hh-|bεf≡>αbε.ZBε}lTπ>F≤=αεO4∩ε>]lW⊗∞M∨&∂&≥⎇bbε≥lBε}.<7/⊗T∞Fzε≥o⊗}vT'/ Q,∩εn≤>7Nn∃]FO∨∧
ε∞≡<↑"r∧≡Dπ&FT∞6∞nT∞FNnT∞FF*¬.&.∞MO∩Rε}-w∨~
⎇F"ε=|F(h,|W'~∞,W?⊗≡NF.rD∞vF.d∞6N>m≤fN≡≥nBεv↑tε6.≡NW⊗/4↔⊗*∞∞&␈6≤LV"b
M⊗↑(Q*ε∂&
l⊗n/5aPPh(≥gO>∨∃BεNd	dLb	∀π>∞nLV"πMtε>/D∞Wαε\≤7∨N\⊂ε∂~∞≡VN≡=O∩ε∂4∞ε␈∨=≤&f(Q*vO&
}W"ε}-w∨≡≥lrε␈↑D¬∀d$
w∩ε←≡6.feDε␈∩
}f/⊗M|⊗&Nlp∧dLD∞vO&∧∞6zε\≥gHh,=voε≡M⊗⊗NM≡GJεlX↔'/,↑2bε≡4εF∂∞Vv.D
⊗rπMR∧f≡>εn∞=
⊗v*dλ⊗g≡t∞FF/,QPW>≡4π&F≡@ε⊗∞E\↔∨≡\D¬"ε≥lB∧t→Dππ⊗|-F.j∞|Rε}mO∩π&≥M6."≤&␈/Dλ∩εf≡NFf(Q,↔"πMRε≡⎇]V}r]M↔∨α
\V/&≥lrr¬9
w>/l↑"bε]}&*π<↑f/⊗T∞π⊗}-LVo~D
FN↑T∞FF(Q,f∞∨D∞FF∂D
V∞∨?≥V
π⎇}Vf"
mw"π.]bπ>≡Mαε/.-w∩n=V≡↑≥lrεNdλ4
∩x8E∩Q-ε∞"≥G⊗.≤O∩ε⊗\]bε6∨V"ε/∀π≡n⎇=⊗v:
≡Bε␈↑Dε}r∞Mε*∧M≡7εn≤=εNvUkPhPQ!PPh cHQ!s∪
Y,⊗rkε$αβ∪εV0L>]}&>*	%b∧≡≡.&/'LTβd<(4ε∂"	Y∃"lX7bh(L↔&+$ε#
∧,≥g.∂/∀βKε$β∪≠&Vαl-:APT7-⎇SR∧|]w⊗>T	"r∧<≡'⊗/NLRβdy(2ε∂D	TM"YX3ph*MsR∧]}'⊗O=ybε∂D
U$¬V#h,<3R¬)Z2ε∂D	TM"X→∩bε=yVn}e]FO∨∧↔"¬:UTHQ!PTo∀WGε↑-⊗.v<Tπ>OM∧π↔.mm⊗v:
\⊗∨∨≥\∩εNd
V∞≡M≡7αε≥lBεf≡>εjε≡4π&F≡@λ∞⎇_9β!-_<⊂_→w9P4\P:40]⊂1wf\0z4q~v4z<H32pz≥y2yP_y2P7≠z⊂8zZz2P1[vx0z~q62Vλ0s2εB:40zλ3y7y\P0vw]w:9P≠q⊂:;Yrping beyond The scope od∧ABAββ?OOL∪'3''Iβ'8hR~>J%∩ε)↓8εrε>|↑2ε}e`∧o.=ε}∩∞Mε*πN|V.↑≥lrπ_:y.∀≥~→$Y|[$
yH_,L~;YaQX;[nM→<@
L>94D
yH_,.⎇≤X,>~;{D∞~≤[n\βt⊂6Xqy7yK⊂77jλ8ytg→β ANY knoWn fo@IZ~∃←α1β3'∨↓1β,εBε}lTπ>F≤9αεO4∩ε>]lWε∞M∨&∂&≥⎇bbε≥lBε}.87/⊗T
Fzε≥o⊗}vT'/ Q,∩εn≤>7Nn∃]FO∨∧
ε∞≡<Z"b∧≡Dπ&FT∞6∞nT
FNnT∞FF*¬*&.∞MO∩Rε}-w∂~
yF"ε=x	Q"Yy.Nh≤Y.}X=≥]Kλ≥m;H≤m≤{Z1M≤x;]∧
Y=hl8=≥.,<h_.,(≤∀M}Z9→,Eλ≠∩-<#"T≡~≠X-\<kC!! P;O≡x>+∧
;H∪I→λ∩(∞|;]→,D≥≠h|=λ≥.∧≠88n∨;8 ≡h≤=-≤z{≡$<h≤
}|z8ML#"P≠Zz47z]⊂3y7\ytw3H7zz⊂∀&!⊂7\⊂6|yYv3⊗⊂≠y⊂7{→y67pY4w3P∪$f⊂;Zz4⊂9[P6pw≡FE1w[x0z4X4v4z≡P32`]:y2yK⊂0yP~0x82[2r⊂4[⊂:42H&4yx≠pqt4[2W⊂ [9wP*~2y2FB;pyP≥40z_0r⊗p\yrr∃⊂0w2λ'$f⊂≤97q6→vP;rH7w6<H:0v5Yr⊂0q≠zz⊂ H64z:≠2FE0]⊂:42H1wvv[w⊗v4\x⊂6rYz4w3K⊂-d7]r{2i⊂6wy→P9r{→y2P8≤7q62[yV⊂6~urP:~2FE#_qz⊂:~0z⊂6Xqy|vXP;wz[2⊂77]⊂9:wλ;tz4λ2y97\⊗qt2Xutw3H4w⊂!Pi↔ab∀⊂εE4_r⊂0v≤2pr<H12rwλ34|2Y⊂1<P≤vwut[3P4zλ7zz⊂≠w⊂:4→P&4y\6pqt~w2W.CEεEεBεEβ_XεEβY_Ve_w⊗\→λ⊂→_~LDcrw\3rP%⊂!py≤2z:2H≡#e!H0z⊂&Rj⊗faO⊂εE"_z2]⊂_P%0[:py<H_\\→λ→→]~L⊗bijβE#97[]⊂#r[y3rP∩↔⊂!p\92z:→P≡#e⊂P0z⊂∪dj⊗fP←εE*≠]⊂&w\94yg[⊂0z UTAH
20
cc: BMS atMIT-AI, cOmmon-lisp at SU-AI

My experiefc@∀AoSi Aek]9S]NA5CGgs5BAS\↓[CGY%`∂Aβ∞s⊃β3O≠C *
≡2π&≡Bπ>≡@hVα_<∞;\h
≡h≥~≡λ_sm↑_9~,-;~5∂∀→Y8.N<Y<d<Y(
m⎇λ∀.]=→(={<_.M8[⊂∩K⊂0w2βE:40]⊂3y7\yP0v[zw:9H5s⊂*≥rrut[3P12↑ww2≥42P9Xwx2P≠s⊂0@≤7yydX4v4z≡P4wεB#'i*∀ g⊂≠MP3wb\β on. Muc@ AWLAQQJAi]KKGS9JAiC-KfAi!JAM←IZA←L↓C@∪∪Ns≤4+∞s?S#-⊃β#πL+Aβ?2βπO'∪π∂SL{9βSG∪?W∨@β7π∂⊗{M "
mw"π↑9⊗v:λ→eJε=mw>rhn6BE'Yα lisp, but one whi@
PASf↓BAOK9KeCY%uCiSα{91β∞s⊃β?↔≠∂WK*βS :≥gN}lTε↔∂AQ&
ε\≤7∂N\⊂-M<|λ
8zy.%H⊂=∧∞~→(∞≤99(∞M99(∞M→(
N,8;≠∂∃H→|M}|h≠mLλ_p↔Y2FE 'ets @IKoeSQiK\X↓oQK\αβ@≡N⎇m⊗&N<≥g"εlZrε6\≡G/⊗↑4ε∂⊗T∞π⊗␈m_F.αDλ
≥y#"J=~≠L≥9<kAQ@εE [<{p|K⊂4w∪αIL I p∂C]QK@Ai<AOKh↓k`A[¬Ggc[∧ACfAEkSGW1rACf↓aWgg%EYJ~)oSiQ=khAOI←ggSαs⊂~ε}↑B¬∀H ε␈∩
←↔≡∞β→K∧
|H⊂↔]2y6'Xp4w3H'$f⊂≥tz4≤wP6p[<BE #ompatibility fEatures(λACLAQCaββ↔;↔ β'9β&C∃α3O≠C7π≡C';∃p∧∧∞g=tπ&F↑,PhW|≡2π&≡Bε⊗≤EV∂∨<X	∧
λ_;LD⊃R3∧∞≤[xML9(∃lT≠{[∂∀≥_;
≤αr⊂ !bkuthad already been fixed by smoking it out on the Lispmachine.]