[omniORB] Sun Visual Workshop C++ 5.0

[Sai-Lai Lo]

>>>>> Tom Dube writes:

> Two months ago, there was a brief thread saying that
> omniORB was not yet ported to Sun Visual Workshop C++ 5.0.

A more accurate question is whether Sun's C++ 5.0 is ready for real use
yet!

To "port" omniORB to use C++ 5.0 is just a matter of setting in
CORBA_sysdep.h to enable the use of namespace and bool type as well as to
remove a workaround for older Sun compilers which is no longer necessary.

As I have posted earlier, C++ 5.0 we tested can't even do a straight
forward std::cerr print statement in a MT program. Its too frustrating to
get Sun's support to understand the problem.

Anyway, with the latest Sun C++ compiler patch (107311-03 of May 20), the
compler at least seem to produce correct code. *BUT* it still generate
incorrect exception handling code when using -O2 flag alone. Strangely,
"-O2 -g" produces correct code. I'm not wasting my time to report this to
Sun given my previous experience. In the meantime, we are still using the
compiler in testing only and not in production use.

Another anonying feature of the compiler is that it has a habit of taking
up all the CPU cycles and huge amount of memory (350M) doing who knows
what. Sometimes compiling a moderate size omniORB2 stub could take 20
minutes or so on a 300MHz UltraSparc.


Sai-Lai


-- 
Sai-Lai Lo                                   S.Lo@uk.research.att.com
AT&T Laboratories Cambridge           WWW:   http://www.uk.research.att.com 
24a Trumpington Street                Tel:   +44 1223 343000
Cambridge CB2 1QA                     Fax:   +44 1223 313542
ENGLAND