[bf1942] Sorting out the lag-issue

[Andre Schackt]

This is great news Andreas :-D

For our part: we have now disabled PB on half of our servers, and we've asked our community to report laggy situations as best as they can (they dont know about the disabling of pb though ;) ). I hope we will see a pattern within a few days running!

I'll of course report back our experiences then :)

brgds
Sussidal @ Battlefield.no


----- Original Message -----
From: Andreas Andersson <andreas.andersson at dice.se>
Date: Wednesday, June 29, 2005 3:47 pm
Subject: RE: [bf1942] Sorting out the lag-issue

> Hello,
> 
> This is very interesting. We have not seen any relation between 
> VoIP and
> the rubberband-bug. But that's not saying that there is no relation.
> Could you please report if this removes the bug or maybe if it 
> doesn't? 
> I thought I should give an update on what we know about this bug.
> 
> We have been working hard with this here and have found several 
> bugs in
> the networking code that could explain things. Some of the fixes were
> included in the demo, which is a later build than the one we sent 
> out to
> production (2446.12 > 2442). This explains why people haven't seen the
> issue as much there. Other fixes was done after 2446, and will be
> included in the patch we will release "soon".
> 
> It wasn't really an issue with the speed of the internet-connection or
> the speed of the machine running the server. It was rather how the
> server prioritized which objects it should send updates about to which
> clients. If the client doesn't receive object-updates it will try to
> update their position based on the latest information it has (basic
> interpolation). This makes objects behave weird when updates are not
> frequent enough. 
> 
> Hopefully we have fixed all issues concerning this. If not we 
> should at
> least have made things a lot better. I guess we'll see the truth of it
> when we release the patch :)
> 
> /AndreasA
> 
>