What will Havok-2 change?

Apologies if this was discussed in older threads (just post a link to them if this is the case, thanks)... and yes, I do know that Havok-2 has been described as "in development" for ages now.

Please don't hijack this thread to complain about SL software development process, customer dis-satisfaction, poor quality assurance policies etc.

I have tried to find some external sources comparing Havok-1 to Havok-2 to help me understand what we can expect from the switch, whenever it will happen, but I wasn't able to find anything useful (I checked, among others, wikipedia Havok entry and a review on Gamasutra).

So, can someone offer some explanation and possibly reasoned speculation on what Havok-2 would give us?

There is a thread where it's mentioned, but I'm not gonna go searching it out right now. I think it depends on how LL decides to impliment it, but IIRC the things possible are:

1. Better performance
2. More accurate collisions
3. Better joints
4. Ragdolls
5. Things floating in water
6. Ropes

Hopefully someone who's in the know will say if LL have said what they are/aren't implimenting or I've missed anything.

Thanks.

Does "better performance" includes "more prims x sim (and ultimately x area)"?
Also, what do you mean by "ropes"? We have currently "ropes" made of particles. What do ropes do in Havok-2?
I've seen hints about Havok-2 being a panacea for the current vehicle problems. Is this the case?

The current "ropes" are just visual particle effect combined with a sensor and push. A true physical rope could just be placed between two objects and all the movement behaviour for the objects being tied together would be generated by the physics system.

Unlikely, the main problem with vehicles is handoff between sims when crossing borders. They'll be more reliable in a single sim, but I don't think it's going to fix the 'problems'.

Oh, I think Havok 2 has dynamics objects like springs and dampers as well. Might be that Havok 1 had that too though...