Path: utzoo!attcan!uunet!world!decwrl!sdd.hp.com!elroy.jpl.nasa.gov!jarthur!ucivax!ucla-cs!math.ucla.edu!jimc
From: jimc@julia.math.ucla.edu
Newsgroups: comp.lang.c++
Subject: C++ for student lab
Keywords: c++,teaching
Message-ID: <490@kaos.MATH.UCLA.EDU>
Date: 4 Oct 90 22:14:17 GMT
Sender: news@MATH.UCLA.EDU
Reply-To: jimc@MATH.UCLA.EDU (Jim Carter)
Distribution: na
Organization: UCLA Mathematics Department
Lines: 29

We have an instructional lab with several Sun 3/280's.  We want to
obtain a good C++ compiler.  The main dimension of ``good'' is that the
students will not be demoralized by compiler bugs.  C++ v2.0 support is
preferred; v2.1 is nice but not a demanded feature.  

We are investigating Oregon C++, Comeau C++ and Green Hills C++.  Does
anyone have insights into these products that would help us choose?  Are
there other products that we ought to be looking at?  

We are also seriously considering Gnu g++ (presently in beta test).  One
of the faculty is using it; he has found one bug (reported and confirmed)
and wonders if other problems he is having might be more than just
programmer error. How does g++ compare in reliability with the commercial
products?  In particular, I came upon an article: 

   D. Saks, ``MS-DOS C++ Compilers'', Computer Language, Oct. 1990, p. 91. 

It includes results of five MS-DOS compilers on the Eckel-Saks test
suite.  If that test suite has been run on g++ I would be very interested
to see the results so I (and the faculty) could get a feel for how g++
compares with products we have some familiarity with.

Please mail direct since I don't normally keep current in this group. 
I will summarize about 15 Oct.

James F. Carter        (213) 825-2897
UCLA-Mathnet;  6221 MSA; 405 Hilgard Ave.; Los Angeles, CA, USA  90024-1555
Internet: jimc@math.ucla.edu            BITNET: jimc%math.ucla.edu@INTERBIT
UUCP:...!{ucsd,ames,ncar,gatech,purdue,rutgers,decvax,uunet}!math.ucla.edu!jimc