Heads up on Test

[rogues]

Sklug from test patch notes on 6/1/09:


This patch contains the first phase of client memory optimization I've been working on for a while. This first pass should show some fairly significant savings in memory usage. A later pass (in a week or two) will have selectable memory modes so that you can choose between pre-loading (to minimize any on-demand loading artifacts) and saving the most memory.Guessing this going to have a pretty severe impact on ShowEQ. Knowing my Linix is pretty weak, is there anything at all of value we can be gathering data-wise to prepare for this? I'm at work now, but I'll log onto Test tonight. It's probably not worth going too far, I suspect they'll be making a lot of changes to this still before it goes Live, and who knows, it could end up dying like much of the proximity code did (probably some wishful thinking, although I sure wouldn't mind EQ being less of a memory hog).

[ieatacid]

Logs from the live servers as they are now would be most helpful.

[rogues]

Ok, I'll try to figure out how to get those :)

On test now, Showeq is working remarkably well. It's detecting just about everything, although it lists "You" as a player along with your real self and leaves your square there wherever you zoned in. Otherwise it seems to be working.

I'm not seeing a dramatic reduction in memory utilization (about 600MB on both my Live client and Test client).

[sammie]

playerProfileStruct grew by 120.

everquest.h:

/*08052*/ uint8_t unknown08052[280]; // changed from 160 for 6/1 patch


I made that change and recompiled and everything seems to be working fine. However, I did not examine the structs very closely, since I was in a hurry to get spawn colors working.

Later when I have more time I may try to confirm exactly which part of the stuct changed. I'm pretty sure it is somewhere between buffs and potion belt.