At 4:44 PM +0200 4/29/02, Peter Gibbs wrote:
>
>Herewith the first set of patches to the memory allocation routines.
>
>There is no new functionality here yet; basically I have been working on
>trying to remove some of the code that is duplicated between the various
>pools, before even more copies get made for the new stuff.
>
>The result is some fairly major changes, but I believe it will make adding
>the new features easier. I have tested the end product using the standard
>test suite, 5000 generations of life, loading Clint's basic wumpus, and
>reversing resources.c using Melvin's reverse.cola. We don't have much in the
>way of PMC-intensive test code at present.
Cool. We're getting closer to the 0% Dan-code limit that signals a
1.0 release. :)
Few questions:
1) Has the external interface changed, and are you planning on having
it change?
2) How's the performance relative to the old stuff?
3) Do you want the memory manager/GC hat?
--
Dan
--------------------------------------"it's like this"-------------------
Dan Sugalski even samurai
dan@sidhe.org have teddy bears and even
teddy bears get drunk