Re: libstdc++ tests assumes all locale data are available

[Phil Edwards <phil at jaj dot com>]

On Mon, Dec 02, 2002 at 12:27:33PM -0800, H. J. Lu wrote:
> > the libstdc++ testsuite will not pass the named locale tests. If this 
> > isn't an issue, don't worry about it. If named locales are needed, the 
> > underlying locale information must be installed. Note that rebuilding 
> > libstdc++ after the "C" locales are installed is not necessary.
> 
> Those tests are meaningless.

Beg pardon?

If you're using locales, then you'll have locales installed.  And we
test them.  When the tests break, we know something is wrong.

> On RedHat, only selected locales are installed.

Same on most every platform I know about.

> Personally, I will simply ingore all libstdc++ test
> failures since most of them are bogus on my machines.

That's the idea.  Hence the "if this isn't an issue, don't worry about it"
in the documentation.


Phil

-- 
I would therefore like to posit that computing's central challenge, viz. "How
not to make a mess of it," has /not/ been met.
                                                 - Edsger Dijkstra, 1930-2002