m0p@k.cc.purdue.edu (S. Kulikowski) (12/09/87)
Comparative Pascal Compilers on Mac2
Stan Kulikowski II
if YOU_HAVE_USED_2_OR_MORE_PASCAL_COMPILERS_ON_A_MAC
then
REPLY_TO_ME;
here are the Pascal compilers i have found references and reviews to for any
mac (i am interested in mac2 only):
Lightspeed Pascal MacUser Feb 87
MacAdvantage: UCSD Pascal
Macintosh Pascal v. 2.1
TML Pascal MacUser Jun 86
Turbo Pascal
MPW Pascal
i spent most of this afternoon (on hold) on the phone getting through to
publishers of Pascals for macs. i have taught i intro courses in programming
for 10 semesters (CS-1 with Pascal) and have developed ~2000-line user
interfaces with Pascal for six years in MSDOS and CPM machines. when i got to
this research project in a small subdepartment, i chose the mac2 for
acquisition because it would have the performance capacity to serve the people
here long after i leave, but the color software will be coming sometime down
the line. i have been acquiring the general software they will use later, but
for my research i need a few high-level compilers (Pascal, C, and Prolog)
right away to get the research data acquisition rolling quickly.
ok, the mac2 arrives two weeks ago. i bought MacDraw, MacPaint, MacTerminal
and MS Word as the fundamental compliment... Hypercard came free. real
disappointing to install everything and the only color is the little DA apple
on the command line. (yea, i know, the fragment mac in 'Welcome to the
Machintosh' has some color too.) hardly what i expected from the mother
company... have they lost all contact with color graphics since Apple II or
is the operating system really so sluggish that it is so diffiuclt to provide
color demos? from what i have heard, i guess it is the former.
i have read over comp.sys.mac messages for the last month, and everybody seems
to prefer Lightspeed C, and reading the reviews in various magazines, it seems
that Lightspeed is the way to go there. i will get to Prologs later, this
message is a request for comparative Pascal evaluations.
i have some specific requirements for a Pascal compiler:
A.) color graphics-- i am familiar with GKS and PHIGS primitives in other
systems. i could get by with turtlegraphics, but
don't want to be required to write my own bitmapping
routines. i hope to make my name here in graphics
programs.
(* i had always refused to acquire dinky-screen macs
because of the unusual monitor proportions and the
the lack of color... a real step backwards from the
old Apple IIs. i guess the Apple mac design team
has lost touch with color programming? they have
had since April and still not any color in the
delivery package beyond the DA apple??? *)
B.) flexible edit-- i have been working on a text in style and
composition in high level languages. having had
2700+ intro students i have read a lot of bad
code, so that poor style is like a sty in the eye.
having learned 15+ editors, i want to customize
my keyboard usage to what other 4th-generation
editors allow (ie, visible block transfers, multiple
RAM buffers, several windows, redefineable numeric
keypad use)
(* to learn another 2nd-generation Wordstar-like editor
is a sty in the mind... much worse than output style
insensitivity. if a syntax-insensitive editor forces
a painful pretty-print, i can force my eye to look
at it and write batch-level processors to convert
their styles... at least i could in MSDOS. i suspect
i will be able to in Apple mac opsys, given time. *)
C.) runtime error-- how much time have i spent chasing flakey errors in
Turbo Pascal? i usually don't write data-intensive
programs, but i do a lot of natural language string
processing, and i don't expect Pascal extensions to
handle much beyond position-of-substring (POS func)
and real-number-equivalence-values (VAL function).
i have maybe six 3000-line projects on the scrap
heap because debugged procedures and functions stop
operating when the compiler loses track of its
pointers, rendering garbage as complexity rises.
(* am real tired of this crap. when you debug a part
of a program it should stay debugged, and if runtime
threatens it, then visible evidence should come up
BEFORE kludgey rewrite is necessary. *)
D.) compilers only-- i think we can forget the interpretive Machintosh
Pascal which apple made native to the macs. i like
the edit-compile-debug sequence. it has cognitive
coherence as human minds interact in high-level
conversation. interpretors interrupt if any
sentence is ill-formed... no matter what train of
thought: if typo then interpretor comes in with an
error. to me, i would read sentence errors after
train-of-thought is entered. compile-time is best
minimalized, but i like the debug after i have
finished procedural thought. no way around it, i
guess.
well, i started this diatribe with the intention to tell you what the
compiler publisher techies said to me about these issues. guess, i will enter
this and see what you got say pre-priori. i know what some their technical
support said.
PLEASE, PLEASE, PLEASE, if you reply... e-mail me a copy personally.
comp.sys.mac has TOO, TOO, TOO much redundant message. if you follow-up
there to this message, also post an e-mail reply to me personally as i may not
be able to sort it out of 375 other messages waiting there for me in the local
system which i am still new to.
if there is interesting reply to me in my e-mail, i promise i will put it
back on this net. let's develop some sub-post connections so this net becomes
readable. there is too much painful redundancy here. net addicts like me
feel convulsed in reading in the same >-quotes 40 or 50 times...
stan
BITNET : XM0P @ PURCCVM (* note, zero, not Oh *)
SnailMail : Special Education; Purdue University; W. Lafayette, IN 47907
USENET : k.cc.purdue.edu!m0p