From: Jonathan Wakely (cow_at_[hidden])
Date: 2004-07-06 10:35:41
On Tue, Jul 06, 2004 at 02:50:55PM +0200, Joaqu?n M? L?pez Mu?oz wrote:
> A test of mine is failing in the gcc-2.95.3-linux toolset with
> the following:
> ostream: No such file or directory
> which, I guess, indicates that the std library used
> by this toolset does not have a <ostream> header. Digging
> through the regression logs, seems like the library being used
> is SGI's.
The IOStream parts are not from SGI, IIRC.
They are part of libio, written for G++, and are very pre-standard,
non-templatised narrow streams.
> Now, there is not defect macro in config/stdlib/sgi.hpp indicating
> that <ostream> is not available, so my question is:
> * Is this defect in gcc 2.95 known, and not an installation artifact?
Yes, GCC 2.95 does not provide <ostream>.
Both istream and ostream are declared by <iostream>.
> * If so, is there any way to figure it out from Boost.Config defect
> macros or some combination of these and SGI-specific macros?
Errrr, you got me there. I've not seen one.
-- "Most people with low self-esteem have earned it." - George Carlin
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk