Re: Caches

From: Jeremy Elson (jelson@CIRCLEMUD.ORG)
Date: 08/24/97


George writes:
>All tests were with an exit() statement immediately after the line log("No
>connections, going to sleep.");

Any profiling done should be done on an active MUD, really.  Boot time
makes little or no difference (IMHO) compared to long-term CPU
utilization while the MUD is actually being played.

The profiles in your last mail, for example, seemed to indicate that
the buffer routines were using relatively little time compared to,
say, fread_string().  fread_string(), however, is used a billion times
when the MUD boots, and then almost never again, so the profile would
have looked *very* different on a MUD that is being actively played by
30 people for 6 hours.

For a profile of a MUD to make sense, boot-time function calls really
should be taken out completely -- or the profile should be of a MUD
that has been played by a large number of people for long enough that
the boot-time usage is swamped by post-boot usage.

-Jeremy


     +------------------------------------------------------------+
     | 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