Wilkinson@HI-MULTICS.ARPA (06/25/84)
Does anyone know of any reference that lays out a method whereby a user
can select the best micro processor language for developing a given
application. In particular, we are considering writing an application
that has to run on both an Apple IIe CPM board and an MS-DOS (2.0)
machine. We think that either Pascal or C is the best language. We are
looking for some method that would lead us through an evaluation process
of the application in terms of each language with a bottom line
conclusion pointing to which one to use.
The application is reasonably sophisticated in that it gathers data
from a remote micro based data collector device via a unattended dialup
polling process in the middle of the night. It then stores this data in
appropriate files for user reference and reporting the next day. It is
a menu driven system for user friendliness. It can also handle dial-ins
from the remote devices when there is a data alarm situation. Also, it
can potentially handle 100's of these devices and gather data over
several years (retained on a winchester of course). So there is lots to
be concerned with and I think the choice of language is critical. Any
info would be appreciated.
Sincerely,
Richard
(Wilkinson @ HI-MULTICS 612-542-7788)young@uci-750a.ARPA (06/27/84)
From: Michal Young <young@uci-750a.ARPA> For a comparison of C and Pascal, with recommendations about the kinds of applications each is suited to, see ACM Computing Surveys volume 14 number 1, March 1982, pages 73-92. It may be worth your while to compare software development `environments' (a buzzword in the making) rather than languages alone. The language is one part of the environment, other parts are the compiler, editor, etc. The greatest language around is not a big help if the compiler you use is too slow or has bugs, or if has cryptic error messages. A lot of time is lost, too, in jumping back and forth from a compiler to an editor if they are not well integrated. If you find Pascal adequate to your project, I highly recommend Turbo Pascal from Borland. Compilation is very fast, the editor is fully integrated, and all syntax errors and most run time errors send you back to the editor, with the cursor at the line in error. This can easily cut programming time in half. Also Turbo is quite portable, and has tolerable interfaces to the operating system and machine language routines (not quite as good in this department as typical C compilers). --Michal Young, UC Irvine, young@uci
peterr@utcsrgv.UUCP (Peter Rowley) (07/01/84)
Have a look at "Comparing and Assessing Programming Languages: Ada, C,
Pascal", ed. by Alan Feuer and Narain Gehani (Bell Labs), 1984, pub. by
Prentice-Hall. ISBN 0-13-154857-3 (hard) or 0-13-154840-9 (paper). The
paper edition cost me $22 in Canada.
The book is a useful collection of 15 papers on the subject, starting with
the March 82 ACM Computing Surveys article "A Comparison of ... C & Pascal"
by Feuer & Gehani. This paper addresses the specific question of which of
C and Pascal to use for particular applications.
Other papers ("PL" = "Programming Language"):
Pascal vs. C: A Subjective Comparison (P. Mateti)
A Comparison of Pascal and Ada (B. Wichmann)
A Comparison of ... Ada, Pascal, C (A. Evans)
An Assessment of ... Pascal (N. Wirth)
UNIX(tm) Time-Sharing System: The C PL (Ritchie et al)
An Early Assessment of Ada... (Gehani)
Critical Comments on ... Pascal (A. Habermann)
More Comments on ... Pascal (Lecarme and Desjardins)
Why Pascal is not my Favourite PL (Kernighan)
Type Syntax in C... (B. Anderson)
A Methodology for Comparing PL's (Feuer & Gehani)
A Comparison of PL's for Software Engineering (Shaw et al)
A Critical Comparison of Several PL Implementations (Boom & De Jong)
PL's: What to Demand and How to Assess Them (Wirth)
The papers show the beginnings of solid methodologies for comparing languages,
and have a lot of good ideas, but I think it's safe to say that the state-
of-the-art is to use "educated intuition" in making choices.
p. rowley, U. Toronto