Proposal for the 'long long' problems

Gabriel Dos Reis gdr@codesourcery.com
Tue Oct 30 21:37:00 GMT 2001


Phil Edwards <pedwards@disaster.jaj.com> writes:

| On Wed, Oct 31, 2001 at 01:44:49AM +0100, Gabriel Dos Reis wrote:
| > | b)  Same for GLIBCPP_ENABLE_C99.
| > 
| > This needs more thoughts.
| 
| Well, it's on by default already, yes?  A user wouldn't be giving
| --enable-c99 in the first place.
| 
| Or, are you saying that perhaps it should not be on by default?

All I'm saying is that we need to be careful about what we dump into
std:: -- those who are reading the LWG reflector know why I'm suddenly
being so overly cautious.  Currently we have put most the of C99
facilities in __gnu_cxx and before we currently commit ourselves to
making them available in std::, I think we might to carefully consider
them.  I am not at all against making them available by default in
__gnu_cxx::. 

[...]

| > Wasn't the current behaviour driven an unacceptable configure time
| > performance reasons?
| 
| I hadn't planned on doing different tests, just storing their results in a
| few more variables.

Aha, OK.

| G'night,

Thanks! (Actually I wnet to bed before your mail reached me :-)

-- Gaby



More information about the Libstdc++ mailing list