Please help

[Nitrox]

Yep crashed. I could enter new track but did see only 3 drivers, no one else on track + no timing. When I left, it wasn't in server list anymore... Track change too fast to see results.

BTW did me or NOP finish first? XD

[Schwoni]

NOP u was the guy on gras right?
i was standin there i have seen

[NOP]

I think i spinned before finish line. That was great finish shame that i didn't save replay.

[Nitrox]

I was on the outside of the curve..
I had no time to save replay however, it was a good finish XD

EDIT: BTW Stuttering wasn't that bad this time!

[PiBoSo]

Yes. Crashed again.
Enough data to debug the problem has been collected.
Thanks everyone for helping.
Server closed.

[Aritz]

I wish to add one more thing:

In prefinal, a little touch between Schwoni and NOP sent Schwoni to the grass and was disqualified because of the speed limit.

In final, I got stocked on a kerb and someone hitted me, both disqualified with the same message.

the problem was surely the lag, collision was magnified and the game has not "lag detector" for avoid the disqualification 

So, two problems; LAG and the "speed limit" limitation, it doesn't works correctly with these kind of bugs

[PiBoSo]

There seem to be 3 main problems:
1) crash when a server is restarted
2) connection impossible or very long when there already several clients connected
3) performance problems with several karts on track
4) stuttering when there are several karts on track ( this is probably linked to n.3 )

[dibu]

Something else to keep in mind.
The Friday test race and your debug server were both running on quite good lines.

I observed the ping of both lines from an independent third line. Also with 10 clients connected, both ping changes were <5ms.
When I watch other lines, they change ping times in a range of 70 to 200 ms.

If the ping times of the server line changes so massive, the results are even worse.

[PiBoSo]

Something else to keep in mind.
The Friday test race and your debug server were both running on quite good lines.

I observed the ping of both lines from an independent third line. Also with 10 clients connected, both ping changes were <5ms.
When I watch other lines, they change ping times in a range of 70 to 200 ms.

If the ping times of the server line changes so massive, the results are even worse.

The wide ping variations are probably due to a "clogged" upload line.
Not much can be done to improve netcode in that situation.
The only solution is to lower bandwidth usage.