On Wed Sep 27 22:34:37 2000, tamstorf@xxxxxxxxxxxxx wrote:
>
> Yes, gcc 2.95.2 is much more standard C++ compliant, which means more
> features but also being more picky :-/ Are there any efforts to make the
There has been a number of posts about standard C++ compliance, from the
trivial, e.g.
g++ issues a warning because it found a variable declaration in a for-loop
expression, i.e. 'for ( int i; ...'. This isn't illegal; it just means 'i'
exists only within the scope of that loop. The -fno-for-scope option will
take care of the warning.
to the insanely picky, e.g.
the KAI C++ compiler returns an error because it can't decide whether to cast
the '2.0' below to a float or a double:
float f = 10.0;
float x = pow( f, 2.0 );
plus numerous warnings about functions not being declared.
> inventor code "standard C++" compliant ? Would a set of patches to make it
> compliant be accepted or are there compelling reasons not to move forward ?
We'll have to review the patches beforehand. I don't have a problem with moving
forward, at the same time I want to avoid "doing too much". There's another
issue with the variations on math functions, e.g. fsin(), sinf(), sin(), that I
want to look into resolving first.
> On a related note, I believe that the problem I'm having with my
> application might be because I'm using a newer version of libstdc++. Could
> you add the "correct" version of libstdc++ to the list of dependencies in
> FAQ.misc ?
Running ldd on libInventor.so doesn't return libstdc++.so as a requirement. In
any case, I have libstdc++-2.9.0-12 on my system.
> Rasmus
Jonathan Lim _ Silicon Graphics _ Mountain View
GPS Graphics Computer Systems CA, USA
|