After putting in a last-released-buffer-cache variable, I would like to
vote against it. It seems that inevitablly the 32k buffer created during
load_help() becomes the last released buffer and fulfills requests in the
range of 256 bytes. I'll leave it in for testing but at this point it
either needs more intelligent caching or to not be used.
btw, the buf_switches in 'show stats' is still a count of how many large
buffers are allocated by the BBS for player output but it is not the number
currently in memory because they will eventually time out. Also found the
reason for the ought to be null message, malloc_buffer should null it out
too.
Any criticism?
--
greerga@muohio.edu me@null.net | Genius may have its limitations, but stupidity
http://www.muohio.edu/~greerga | is not thus handicapped. -- Elbert Hubbard
+------------------------------------------------------------+
| Ensure that you have read the CircleMUD Mailing List FAQ: |
| http://democracy.queensu.ca/~fletcher/Circle/list-faq.html |
+------------------------------------------------------------+
This archive was generated by hypermail 2b30 : 12/08/00 PST