On Sun, 25 Aug 1996, jeremy elson wrote: > You'd be amazed at some of the strange optimizations > I'm putting in to get a mud to run on this beast :) > > I love your idea of buffering the most recently searched-for mob, etc.: I > will probably put that in the stock code. I have two questions for you, > though: 1) did you, perhaps, add some kind of counter to see what > percentage of the calls to real_xxx are hitting the buffer? That would > be interesting to know. 2) what other similar optimizations have you used? Sorry to take so long to reply. Busy busy busy.... 1) No but I've been meaning to :) Next time I'm coding and remember it I'll put it in and let you know how it does. 2) Mostly zone reset stuff. There were several zones that were taking a very long time to resets, so I made a change to the code where it resets all zones at boot time to skip certain zones by vnum, so they wouldn't reset till the first scheduled reset. THat shaved a minute or two off the boot time. I also changed most of the fatal errors. Every boot error produces a killscript so the mud doesn't keep booting. Some day when I have time I'll make it put the sign command up (btw I also put a sleep(3) in sign right before it disconnects for people using certain windoze clients). I've also made a lot of the fatal errors less fatal. If a board object doesn't exist, I set the board vnum to a generic board object and log a warning. I had an overeager imp delete a bunch of files in the world hierarchy without editing the index files a couple times which produces some annoying file not found fatal errors. I just made a little tweak to skip missing files, or maybe that's on my todo list. I don't remember :) I've made a lot of little changes that would probably be great for stock circle but I rarely write anything down so I keep forgetting what I've done. I'll check the mud boards and my rcs logs (thanks for the advice btw I love rcs) and let ya know what else I've done. Sam
This archive was generated by hypermail 2b30 : 12/18/00 PST