Re: Bug in libm or libstdc++.

[Joe Buck <jbuck@Synopsys.COM>]

> At 03:32 PM 2/24/99 PST, you wrote:
> >You have already been informed that libstdc++ is not complete.  Sorry.
> 
> The standard has been around in largely complete form for over a year, and
> finalized for many months. I am curious as to what is the cause for all of
> these omissions...

Limited resources.  Up to now most resources have been concentrated on
completing the compiler portion of the standard, and we are very close.
The only significant unimplemented feature is "export".

> In any case, to avoid any more unpleasant surprised I would like to see a
> list of all the omissions in the current libstdc++ and other
> non-conformances, and where possible ETAs for the missing components.

If you would like to volunteer to assemble such a list, and do most of
the research yourself, you might find that others are willing to help.
If you continue to demand that others do work, you will probably just
create more hostility.

As for ETAs, in the free software world the only answer you will get is
"when it's ready".

> It
> would be a good idea to post a periodic libstdc++ progress report and calls
> for volunteers to implement some of the stuff. I think that might attract
> more developers to fixing up and finishing implementing the standard C++
> library.

See http://sourceware.cygnus.com/libstdc++/

> An area where I may be able to help is with the standard valarray and its
> friends, once I get back a response regarding making the compiler eliminate
> temporaries and excess copies in argument passing and return.

Again, see above.