On 7/5/06, Timo Schoeler <timo.schoeler@riscworks.net> wrote:
> hi,
>
> i just came across reading this
>
> http://blogs.23.nu/ilja/stories/12323/
>
> and am curious now what developers would like to say in this case... is
> it just FUD?
>

Well, I'm certainly no expert on low-level stuff like this, but if you look at:
http://cvsweb.netbsd.org/bsdweb.cgi/src/lib/libc/stdlib/malloc.c?rev=1.46&content-type=text/x-cvsweb-markup

You will see that NetBSD's malloc.c only mentions "unsigned"
Here:
/* Set when initialization has been done */
static unsigned malloc_started;	

Here:
/* Number of free pages we cache */
static unsigned malloc_cache = 16;

And here:
/* How many slots in the page directory */
static unsigned	malloc_ninfo;

And none of those are unsigned on size_t.  Darwin's malloc.c has
-many- more references to the word "unsigned".  Also, it's plain to
see that the malloc.c's from the two operating systems are -very-
different.

As such, I would go with FUD, which can also be gleaned from the tone
of the article referenced.  Example:

"I can't believe they fucked it up
ilja, 03:40h
A while ago I spend some time in the osx heap allocator. Man there is
so much crap in there, unbelieable. "