Alan Coopersmith <[EMAIL PROTECTED]> wrote:

> Roland Mainz wrote:
> > Umpf. See my original posting. My intention was to improve performance
> > and usuability. |snprintf()| would make it slower. MUCH slower.
>
> Is snprintf() really the performance bottleneck in your application?

I currently only know of one single application where sprintf() would be 
(and in former times really was) a performance bottle neck:

star in a mode that creates POSIX.1-2001 extension headers.

My solution that made star in this mode even faster that GNU tar in historical 
tar mode was to use unrolled macros rather than using strcatl().

> > And: snprintf() may be evil - see Mozilla.org Bug 332006
> > (https://bugzilla.mozilla.org/show_bug.cgi?id=332006 - "Avoid raw calls
> > to snprintf").
>
> What broken OS doesn't guarantee null termination on snprintf()?
> Solaris and the Single Unix Spec v3 both do.

Thank you for verifying this, I was confuced from the message.

Jörg

-- 
 EMail:[EMAIL PROTECTED] (home) Jörg Schilling D-13353 Berlin
       [EMAIL PROTECTED]                (uni)  
       [EMAIL PROTECTED]     (work) Blog: http://schily.blogspot.com/
 URL:  http://cdrecord.berlios.de/old/private/ ftp://ftp.berlios.de/pub/schily
_______________________________________________
opensolaris-discuss mailing list
[email protected]

Reply via email to