Quote:The hp values shouldn't matter. Think about it. If you aren't in one fight, you're in another, and the computer doesn't care if it's one hp sack or another. So, there is always combat. So the only way that it should matter is if hp is overflowing the its type. Shorts are 0 - 32767
ints can go to +- 2147483647
I would be shocked if they were using shorts.
I wasn't referring to HP values, I was referring to procs.
Essentially they didn't plan ahead. If they were smart they would have sat down and looked at the game from the start "Ok, this level 1 sword does an extra 1d6 fire damage... so a good sword at end-game(currently lvl 10) might do 3d6." This is where their line of thinking stopped, so now you've got a sword at lvl 30 that does 15d6 + a shitload of other things.
Then you start getting into doubleshot/doublestrike, rapid fire, etc, etc, etc, and that's where all the overhead comes in: too damn many dice roles. If you took a typical endgame DDO character and tried to work out how many physical dice you needed per attack, i imagine it would be A LOT.
...that totally sounds like a fun project, actually.
Quote:They are however apparently running their server without using much multithreading apparently.
This is not surprising at all given the history. Remember: DDO owes much of it's server arcitecture to Asheron's Call, which predates multi-core CPUs by a lot. Waaaaaaaay back in the day, conventional wisdom was to use dual-socket servers, with 1 CPU to run the game code and a second to run the server OS(and of course you always had clusters of servers - so 1 for chat, 1 for logins, etc). At the time there was really no reason to go the extra mile and multi-thread your software because quad-socket servers were exponentially more expensive. It was easier to optimize your code or split elements out into additional hardware.
The first dual-core CPUs didn't hit the market until 2005(thus allowing for 4 CPUs on a dual-socket board). Development of DDO started, what, 2002, 2003ish? Unless you were a DEEP insider in the tech industry back then, it was pretty easy not to predict the rise of multi-core CPUs.