[RFC] Another look at LFS vs filebuf::showmanyc

Paolo Carlini pcarlini@suse.de
Mon Sep 20 18:57:00 GMT 2004


Nathan Myers wrote:

>I agree.  I see little benefit to ever returning a value bigger 
>than (say) 1M.
>
Good, thanks. I'm already testing a patch.

>  Is this the only place where the streamsize proves 
>too small to yield the (strictly) correct answer?  That would be 
>uncommonly good luck.
>  
>
Eh! Almost... istream::ignore also required a special treatment, but
we are rather lucky, indeed ;)

Paolo.



More information about the Libstdc++ mailing list