Eric over at eldergame.com has imho a realistic approach re: developing MMOs. From about us"Eric has over a decade of experience as a Senior and Lead Engineer on such products as Asheron’s Call 2 as well as being Producer for AC2. In addition, he has impressive credentials as an MMO systems designer, most recently for the upcoming MMO Star Trek Online. "
Here's one post by Eric over at eldergame.com. Among other things he explores the use of middleware etc. for an MMO. http://www.eldergame.com/2009/04/smartfoxserver-the-mmo-engine-for-indies/
"And if I plug realistic timelines into a realistic engineering schedule, it appears that two or three engineers
working half-time on an MMO will never ever finish it
if they have to develop the server tech themselves (or the client tech, or the tools pipeline, but those are a different post). "IMHO
Multiverse has plenty of inefficiencies, bugs, graphics challenges etc. It could use improvement.
Most of the systems and most of the tools are already there
It takes tons of time and resources
to develop this stuff and most of it is already done for you.
We could all spend from now until World of Warcraft 3 comes out developing and "improving" tech for Multiverse. We could add in megatexturing, port to IOS, Android, PS3, realtime raytracing, voxel terrain, the list is almost endless.
The next 3 or 4 years could go by and there will still be improvements "needed" or tech that could be added to Multiverse. It's (quite obviously) related to things like "feature creep" or "scope creep". Related in a tertiary sort of way to "design by committee".
Spending tons of development time constantly adding in the next great feature kind of defeats the purpose of using middleware to begin with.
If you're developing a game at some point you have to cut it off and actually make the game.
...or in Eric's words above you will "never ever finish it"
Just my two cents