curt428 wrote:CBJ wrote:Yes, we've seen it, and yes we're aware. No need to keep bumping.

Ok... plan on doing anything about it? I haven't heard much of anything about actually addressing this issue... which is hard to understand as this is THE issue.
I don't think it's so simple.
First, it is really just a hypothesis at this stage (although arguably a relatively well-documented one). Until the hypothesis has been confirmed by further analysis on the developers' end, it is pointless to ask for a fix.
Second, the complexity of fixing a thread synchronization bottleneck varies greatly. It can go from a simple removal of mutex objects where they are not actually needed by the code logic, to a partial rewrite of code to reduce access to synchronized resources. In some cases, core design of an application may be such that design has focused heavily on (single-instanced) heavily-featured synchronized objects. In those cases, addressing the bottleneck without essentially rewriting the entire application is practically impossible.
Third, threaded application writing is quite different from single-thread application writing. As this is the first threaded application I know of that EgoSoft wrote, it is reasonable to assume that, as skilled as the developers may be, they will face a learning curve on that front, especially on the optimization side of things. They are human beings just as we are, and deserve a break from the criticism while reviewing their first attempt at a new (and complex) coding technique).
Fourth and last, I think Bernd might be surprised (in a good way) at the end, if all goes well. Multi-threaded applications (that take advantage of that mechanism through coding best-practices) have huge potential to deliver performance above and beyond what is achievable with single-thread coding. It may well turn out that X:Rebirth, once optimized, will be the first X game that is not CPU bound... and that would be a great achievement that I think we should compliment the devs for if they can pull it off.
Astyrrean