On Wed, Feb 13, 2002 at 12:20:51PM +0100, Martin Konold wrote:

> Sorry there is a missunderstanding. I wanted to say that we might even 
> consider to enable the new malloc unconditionally because I am very much in 
> favour of it ;-)

Please don't.

I have lots of confidence in PHK's ability to write a !#@! fast malloc,
and would rather not have KDE second guess these things.  If nothing else,
the fbsd malloc is *far* more configurable to aid in various debugging
stuff than the proposed replacement.

IMO the appropiate thing to do here is to push the glibc people to fix it
or write/use a better malloc.

I'm more than willing to run some test code to back up my beliefs.

- alex