Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!

2 months after NGE and the only thing SOE fixed was lag...

...oh, my mistake, they didn't fix that either ::::18::

Comments

  • duncan_922duncan_922 Member Posts: 1,670
    The sad part is that the rubberbanding issue is not due to lag, but do to faulty coding.  That problem has existed since pre-cu and SOE hasn't been able to fix it.  Incompetence at it's best!

    SOE knows what you like... You don't!
    And don't forget... I am forcing you to read this!

  • Morat20Morat20 Member Posts: 89

      IIRC, there are three basic causes of rubberbanding:

    1. Server boundary transitions. When you pass from one server to another in a cluster, they communicate. Rubberbanding in this case is caused when there's a rejection between the servers, bouncing you back to the originating server. It's sad that this is a reoccuring problem.
    2. Speed sanity checks: The server -- following Koster's "The client is in the hands of the enemy" belief (which is a good idea) -- keeps tabs on your speed, and if you transition between two points too rapidly, it will claim the movement is "wrong" and bounce you back to your last allowable point. That THIS hasn't been fixed is bad enough, but that it was so badly coded that speedhacks in the NGE work just fine is even worse.
    3. Lag. Server gets confused as to where you are, bounces you back to your last position. Bad net code, sounds like.

      Frankly, the SWG engine appears to be shit from the bottom up. What wasn't shit to begin with got turned into shit over time. And I don't even WANT to hear their whining about DB issues. If their DB is causing that much lag, then they're incompetent. EVE-Online has MUCH higher DB requirements than SWG -- orders of magnitude higher -- and they manage through good coding, expert DB design (and hiring DB experts to sanity check their setup) and state of the art hardware.

       SWG has higher database requirements than most MMORPGs, but they're not even remotely straining the technology. It's bad design and worse implementation that's at fault -- and yes, I realize their database requirements for crafted objects.

       I suspect the real issue is implementation -- from all accounts, they had decent DBA's working there -- which means the problem is either hardware (refusal to upgrade), bad initial DB design (difficult to overhaul without using a LOT of resources), or simply badly sub-optimal DB utilization on the software side.

Sign In or Register to comment.