Euro F3: Mosport

Wow, that was tough. I feel fairly drained after that!

Had the closest of battles with Simon all the way through. Probably the hardest race I've done yet. He beat me in the end, but up until the last 5 laps or so we were fairly evenly matched. F1 drivers should watch the replay so they can learn how to overtake :)

Good, clean race, punt rule helps at this track. Led the race up until making a minor mistake where John Bayes was able to pass me. I was then right behind him when he lost it going into T2 and I retook the lead from there.

Finished the race about 5 seconds ahead Dmitry.

Two races in a day and both had the car notably skatey online versus all the offline practice done. Or rather, when more cars are on the track.
It took nearly 20 minutes messing around with the setup to get it back to offline lap times, but you can't truly 'get grip back' and the same speed, you can only balance the car out again and make it driveable again.
It ended up pretty much my offline lap times, but notably tougher with it being quite loose to do that.

But then the big problem was the race car! There was no time to do anything to test that so I had to just use the Qual car and add a bit of wing at each end. And you don't just add one front wing for one rear wing to end up with the right balance and best setup, and with no time to even do one lap to test it out I had to guess the settings. And it ended up 1 or 2 steps too low in balance to the rear.
Lap one, which is generally a little bit off pace thanks to the no punting rule, was somewhat helpful to get a feel for the car and I could tell it was a bit loose on the rear but it felt ok.
Too bad Lap 2 into the last fast corner complex off the back straight I found out how much too loose it really was....and it suddenly broke free by surprise.

It is such an overly grippy track you could race around almost flat out all day very safely, but not if the car setup is amiss. You almost can't crash in most areas even if you wanted, in these cars at Mosport. So that was a very annoying result!

This "Online oddity" is becoming a real nuisance. It seems fine for ages, many many days and races, but just turns up now and then. There seems to be a pattern of 'time', like it is there for many hours or more likely days, not just a one off thing. But there is no way to test it all out because whatever it is it is only rFactor that cares about it as far as something you can see a result in. eg downloading, web browsing, don't care about data flow smoothness or whatever it really is caused by.

Peter, hile no one has ever proven that online vs offline use any different physics, this has been a complaint of sim-racers since 1997 when online racing started with Nascar Racing 2. Most people do all their setup testing online to solve this problem. The R2P servers are operational 24/7 for this very reason.

On a side note; While the actual physics are probably the same between on and off line, the fact that your PC is having to translate and use the online code when racing over the internet probably adds some latency to the physics code causing it to act slightly differently. This has never been confirmed by a developer, but chances are it is at least half-true. The more cars on a server, the more your PC has to work on graphics and such causing the physics to get skewed even more.

Moral... test online man... there just is not any way around it.

Yeah, I try to get online...and mostly do it on the R2P servers. I always get online and run laps, and they match offline perfectly really. But it is alone almost always and the changes come when more cars are on track. Tied in with the ISI bandwidth throttling I expect - which alters what data you get relative to how many cars are 'near' you.

But odder is how sometimes it can be perfect even in a 20 to 30 car field. And other times have big changes in even 10 cars.

It also must be 'Internet path'/distance related, or something, because I don't see too many other people making note of it in forums or Leagues etc. Thus likely to be far more negligible effect to them.
I have also proven having a router alters it, detrimentally, but I always remove that for online racing now. That also supports a theory it is 'delay' related. Not necessarily latency, because my pings are the same in the bad cases or good cases, with or without router, but something to do with 'linearity'/consistency of data flow.

For Mosport I would have lost close to 0.5sec if I altered the car to match the offline grip levels - via wings needed.
And I suspect that almost no one else is having to do that dramatic a change, due to less effect (undetectable to them really) in what seems to be most cases.
Worse is that is changes the car from situation to situation...eg practice with few cars around you and it is good, get a swarm nearby and it goes off balance not just equal loss of grip front and rear. So you are in big trouble when it hits big time.

It all seems very odd to me that they can't write net-coding that allows for all such cases. There should be no need for any tie in with data flow and physics cycles really. You could run asynchronous physics cycles to the data flow cycles perfectly fine. But it does seem to have been some common programming theory/system used over many sims for a long time.
I am not sure of they program to allow for it at all or not. Seemingly not.

Well, I think the fact that you're in Australia and the server is in America doesn't help. I seem to remember you having similar problems with GTR back at TM.net, I think you ran one race at Watkins Glen with us, discoed on 1st lap.