Patrick Logan wrote:

# "Ben Tilly" <ben_tilly / hotmail.com> wrote
# in message news:LAW2-F74aJswFCHOzyf00004122 / hotmail.com...
# >
# > I would doubt it.  I have not looked closely, but
# > glancing at http://www.unipress.com/toolkit/cfront.html
# > it seems to be Unix specific (not even a Linux version!)
# > and proprietary.  Even if the output C is portable and
# > easy to link to, the development platform is now limited
# > to people with CFront.
# 
# I wonder if g++ is the most widely available C++.

(See below for reply.)

# One reason I ask is that the FOX GUI seems pretty good for
# Ruby, but it relies on C++.

Well, so does wxWindows. AFAIK, Python works well with both. 

As for g++, if by "widely available", you mean "cross platform", then
AFAIK, you are undoubtedly correct. If by "widely available", you mean
"most commonly used", I would guess that might still be MS VC++. 

I don't know how compatible g++ is with VC++, but I do know that there
seem to be some annoying differences between AIX C++ and g++, even
though both think that they are standards-compliant. See the following
URL for one issue on which they differ.

    (http://www.informatik.hu-berlin.de/~loewis/corer9.html#95)

Mike Miller wrote:

# The standard is schizophrenic on the issue of whether such friend
# declarations introduce names into the enclosing namespace.

Conrad Schneiker
(This note is unofficial and subject to improvement without notice.)