In <1992Jan9.065251.18046@enea.se> lhc@enea.se (Lars H}kan Carlsson) writes:


| "What C++ compilers for ISC SysV.4 is there?
| "
| "I've tried to use HCR C++ (2.1 for V.3.2) and Glockenspiel C++ (2.0 for V.3.2)
| "on ISC V.4 and it does not work very well. The compiler does it thing BUT I
| "can't link it. Apperently the library elements is of wrong size (that's what
| "ld(1) says). Is there a way around this? Also the patch/mxx programs can't
| "find the .data section anymore because it's not COFF anymore!
| "

We are currently evaluating Oregon C++ and LPI (Liant) C++ on Dell
SysV4. Although it's too early to give definite results:

Oregon C++: fast, C++ 2.0 compliant, can not handle many #includes
(according to my colleague which found this out when trying to compile
X stuff), primitive debugger, SysV4 compliant (ELF format), doubt
about support (delivery of the compiler took 5 weeks - there seems to
be some change. Minimal documentation. We're not going to buy this one.

LPI C++: slow, C++ 2.1 compliant, SysV3 compliant (SysV4 expected in
1-2 weeks), non-standard flags, K&R, ANSI-C and C++, debugger seems
excellent, customer service as well as support by e-mail *excellent*.
We found some small bugs in the compiler which will be fixed in the
upcoming SysVr4 release. Good documentation (seperate manuals for
C-library, C++ library, C++ manual, debugger manual, quick reference
guide, Stroustrup's 2nd edition).

If LPI's SysVr4 indeed shows up during the coming weeks, we are fairly
sure that this will be an excellent, but slow compiler. They say they
are working on speeding up the compiler.

Karel

--
Karel Zuiderveld                            E-mail: karel@cv.ruu.nl
3D Computer Vision - Room E.02.222          Tel:    (31-30) 506682/507772
Academisch Ziekenhuis Utrecht               Fax:    (31-30) 513399
Heidelberglaan 100, 3584 CX Utrecht, The Netherlands

lhc@enea.se (Lars H}kan Carlsson) writes:

>What C++ compilers for ISC SysV.4 is there?

There are patches available for g++ 1.40.3 under SVR4.  You can get them from
me, or by ftp from physics.su.edu.au.

Andrew

In article <1992Jan9.221458.3048@cv.ruu.nl> karel@cv.ruu.nl (Karel Zuiderveld) writes:
>In <1992Jan9.065251.18046@enea.se> lhc@enea.se (Lars H}kan Carlsson) writes:
>
>
>| "What C++ compilers for ISC SysV.4 is there?
>| "
>| "I've tried to use HCR C++ (2.1 for V.3.2) and Glockenspiel C++ (2.0 for V.3.2)
>| "on ISC V.4 and it does not work very well. The compiler does it thing BUT I
>| "can't link it. Apperently the library elements is of wrong size (that's what
>| "ld(1) says). Is there a way around this? Also the patch/mxx programs can't
>| "find the .data section anymore because it's not COFF anymore!
>| "
>
>We are currently evaluating Oregon C++ and LPI (Liant) C++ on Dell
>SysV4. Although it's too early to give definite results:
>

I would steer clear of Oregon if I were you.  I have had awful experiences
with support from them (Pascal and ISC).  They have the ability to
port to many platforms and their sales staff will promise that everyone
gets support but we have not been so fortunate.  They may be stretching
their resources too thin.

1. They refuse to discuss known bugs until you slam headfirst into
   one and waste a week tracking down a bug that turns out to
   be a code generation error.

2. We paid $400+ for support.  By the time I got workarounds for
   bugs I had long since found my own.  We have not receive any sort
   of updated compiler despite bugs like:
   A.  The compiler crashes during a compile.
   B.  Runtime support for a standard function was not in the distribution.
   C.  Numerous other bugs remain

One might say that C++ would be better supported than Pascal but I paid
with the same dollars that a C++ buyer uses.  Also they claim that the
C++ compiler is written in Pascal,  the same one that has all these
bugs.

harry levinson

In article <1992Jan11.155436.19524@werple.pub.uu.oz.au> andrew@werple.pub.uu.oz.au (Andrew Herbert) writes:
>lhc@enea.se (Lars H}kan Carlsson) writes:
>
>>What C++ compilers for ISC SysV.4 is there?
>
>There are patches available for g++ 1.40.3 under SVR4.  You can get them from
>me, or by ftp from physics.su.edu.au.
                               ^^^

We haven't conformed yet -- and are still physics.su.oz.au.

David
--
------------------------------------------------------------------------------
 David Dawes (dawes@physics.su.oz.au)    DoD#210      | Phone: +61 2 692 2639
 School of Physics, University of Sydney, Australia   | Fax:   +61 2 660 2903
------------------------------------------------------------------------------