Sorry i leave at 22:30 CET as it was way too late to finish the race at 23:30 but it seems you guys enjoyed this race! Well, about this bug, i've never experienced this in any simracing games i played. But we have to see the good side for yesterday's test race, and the good point is that there were good racing and many drivers compared to previous races! Will try to do the race, if you do another next week ! Grats to all drivers!
The same was for me - a certain other half gets annoyed enough by me being up 1,5h later than ordinary
Fun to hear there were some fun racing after all! Impressive stamina
@David Turnbull !
I think we need another test race and quickly to iron out the issues. I don't think it's a build bug as I jumped into the vec server shortly after dropping connection here and it was fine. 20 players, same build and no issues. Veeerrry strange.
Thanks for sticking with it and giving it a bash tho
As said, I dwelled on this during sleep
From what I've learnt in my computer science courses and as IT-technician, I can see some reasons:
1. Possible build issue - we were doing fine until the several practice restarts if I got the picture? Logically that says it then less likely matter if 20 players worked at other servers. Maybe we have found a bug that not many will. Maybe a difference in how the application handles connections being made/shut during session switches. Like with memory leakage, the same can be with connections, triggering thresholds for buffer overrun/maximum connection capacity at different application levels (rf being one, windows, router, ...)
2. Router issue - would be interesting to view router logs. Some explanations could be gotten about whenever a single TCP/UDP connection drops. Was the server in DMZ mode at some point?
3. Ive got the not so well informed picture that rf2 is rather sensitive when it comes to correctly set up bandwiths.
@Gijs van Elderen seems to have lots of insights?
4. Certain drivers livery or something whole else causing this
We seem to have people with connections to ISI's staff (through
@Gijs van Elderen perhaps?) that could help us get a solution/answer at least? Maybe in a less time consuming way.
We could try to run a test race at my dedicated box if that would be interesting.
Im eager to find a solution to this. Finally some real life practice of years of college studies. Could be my masters degree project!
