Hey, ISI team! Since we just recently ran the Bullrun 1000 (with driver swaps) and the problems are still fresh in my mind, I wanted to bring this up here with some suggestions (never bring up problems without suggestions to fix them my USAF commander used to say)... I understand that you all feel that you've looked at a lot of the problems with driver swaps from RF1. That's great! However, I know from experience (and I know you all know this) that many problems just don't crop up until server loads reach a critical mass. I would like to propose that when you all feel you are ready that you all stand up a server, and I would be glad to round up enough drivers to give it a good test. From what I've found over the past few years, I think most problems stem from two things: 1) server connection overload 2) driver connection issues Here's a comprehensive list of items that we've found, and what I *think* are the issues relating to them. 1) server dump (en-masse driver discos) - server connection overload. Most likely due to a large number of spectators in a race, especially close to a pit window. Proposal: separate drivers that are reconnecting from spectators as a different 'class' of connection, and get priority or that spectators can be banned but reconnecting drivers are allowed (password or file hash that allows them back on the track) 2) Car gets dumped in strange location around the track, and is out of commission - driver connection issue 3) Driver or relief driver sees "riding with _" (no driver name). driver connection issue 4) AI takes over car instead of relief driver. driver connection issue 5) Random discos of drivers, sometimes separately, sometimes a few at the same time. server connection overload OR driver connection issue? One of my proposals for driver connection issues would be this: the car sits in the pits until a driver gets in the car. So if a driver change was initiated but failed due to a poor connection, the car would just sit there until one of the drivers was able to get back into the car. I would also like to suggest to completely drop the 'hot swap' option, I can't think of a single good reason to allow a driver to swap somewhere on the track, just make it where swaps are only allowed in the pits. Finally, it would be soooo nice if you all could add driver reconnects to rF2 as a server option. I think the way N2003 did it was nice, it would allow a reconnect if you somehow discoed, but obviously if your car got destroyed, no reconnect. N2003 also made sure you got penalized, even if you managed to reconnect right away you were going to be 1 or more laps down. I know most drivers wouldn't mind discos so much if they were able to finish the race, even a few laps down. Anyhow, thanks for all you are doing, looking forward to the end product! Cheers!
One other thing while I'm thinking about it. I like the server hash idea, and here's my proposal for how it works. 1) All drivers sign into the server with the same car/team, and the server issues them a hash file that gets dropped onto their local machine 2) Now all the drivers for a specific car/team are locked in. At any time during the race, if one discos or for driver swaps, the server checks their hash and sees that they are allowed in that car. 3) This allows you to set the server to 'no spectators' but still allow the correct drivers to get back into the car/team they signed up with. Anyhow, just another suggestion... Cheers!
Rather server overload or track with limited pit box/grid quantity. If while swaping (doesn't know any another occasion) - definitely swap bug related issue.
Tracks without enough grid locations dumps drivers at the start of the race. Not enough pit boxes causes cars to pit in same stall. Poor driver connections have caused this type of incident to happen with a low server load. Cheers!
Another thing that would be great is for another connection 'class' called crew. This would let crew members join the server and just see telemetry and have the ability to make pit menu adjustments for the driver. It is so hard to crunch numbers in your head and then fumble through the pit menu while your trying to drive. Sure you can have TeamSpeak running so you can communicate with your team and do the number crunching for you but. But they still can't change the pit menu for you and it would be better if they had access to tire wear/temps, break wear/temps, fuel level and all that so you don't have to mess with it while driving. I guess a plugin could be easily developed that can relay this info to your team somehow. Hmm.. I may look into writing that. Hmm, one way to deal with spectator load would be to have a second server dedicated just for spectators. So the main server would only have one spectator connection from this spectator server. Then if the spectator server gets overloaded it may crash but not take the main server with it and the race can continue on.
Great ideas, Noel...I like them! Yeah, it's amazing how long it can take you to just calculate your next pit stop when you're trying to concentrate on your driving... Cheers!
Ho did you confirm this explanation? Of course it may be due to bad connection it happened only if dedicated server reached its limit to handle number of drivers. So I suspect that it is server which has desynced clients and does 'think' that there are more clients as actually are. In that reason it may send grid position for driver as he would be ie 60th one, which can be handled by track. Since rF is based on the same engine family as GTR1 was, I would like to point to GTR1 issue where server has been overloaded by too much connections. In result drivers has been split into groups being able to use chat only inside a group but not with other drivers. Strange but true. I guess that overloaded rf server (doesn't matter if it is CPU overload or data traffic one) causes mess in communication - not clients.
Hey!! I'm also very eager to know if there are going to be any improvements on general stability and driver swaps for rfactor 2. The endurance racing community is constantly growing and the biggest events on rfactor are infact endurance events and leagues (f.e. sim-racing.de 24h races, up to 80 teams signed up with a total of about 400 drivers). Therefore a little statement on these issues would be great !! A possiblity to rejoin the race after a disconnect is probably too much to ask and most likely way down on your to-do list. I understand that with all the new features rfactor 2 is going to introduce. But I sincerely hope there are going to be improvements on stability ( Network,game and driver swap), so we will have less problems with rfactor 2 ( f.e. missmatches during the race, AI taking over when swapping drivers) I would prefer news on these issues over WIP pictures and features, as do many others. Regards Markus
Well actually an ISI representative have mentioned a few times on the forum that this feature is planned and is already undergoing testing.
Indeed, I hope that with rF2 the Driver swaps are more reliable, too many DNFs occur with the Driver Swaps in rF2 which is unfortunate.
Hi Jan! This is more for multi-class racing-but since endurance racing often goes hand-in-hand with same, I figured it belongs here. In other words, I hope ISI makes their sim more multi-class friendly: 1. Points earned for finish in class not overall finish 2. Lap times/sector splits visible in game of competitors in your class only-I don't care what the LMP1 leader is doing if I'm in GT2. 3. Way to tell on the in-race leaderboard who is and isn't in your class. I've had to edit all the AI drivers I run with in Enduracers with a single numerical code in front of their first name (e.g. "3Ron Fellows", where 3=GT1, 1=P1, 2=P2, and 4=GT2) so that I know on the track where I am running in class.
Tim (and team), Anything new you can report on the status and feasibility of fully supported/debugged driver swaps? TIA. - Doug
Let me cite from the developer q & a: That's no good news for many rfactor-users. Endurance racing is what kind of keeps rfactor still alive. But as the driver swap feature doesn't work properly it has become really frustrating when you practise for a +8h race a lot and in the actual race the software will prevent you from finishing. This should have been fixed for a long time I think. And yes, I know this feature is not documented but it would have helped rfactor find its unique niche aside the Simbin games (and aside Iracing, which will soon have a WORKING driver swap feature). But I'm still hoping for some improvements especially in this area with rfactor2. There have to be made compromises with such a small developer team and other (more important?) projects. :-(
Why do you feel the need to compare iR to rF? Ask any member of iR what they have paid for that service over the last year or two and IF you do get an honest answer it will be alot more than the $50 I paid for rF and probably alot more than what I will pay for rF2.
CdnRacer, I don't get you. Do you suggest that 'suckiness' of the software should/may be linear function to its prize? If some thing is cheaper means that may contains more bugs? I believe that anything is produced (doesn't matter for personal or commercial use) should be done as best as it can. If there is no enough resources, it will be just smaller, contain less features. But there should be no acceptance for screwed features left without support/fixes. Indeed, I meet such apporach each day: portal admins f***k a work up saying that it is done for free so they are free to do that. Open source software has unbelievable bugs, not fixed for years, just because it is open source distributed for free. rF1 has driver swap feature, after 2 patches still not fixed - and you are saying it is OK because the software is cheap? yeah.. I remember Max said, it is not supported feature. If not supported - shouldn't be available for public use. Otherwise it turns into just buggy feature. So, if soft must be limited just because low target price, it should have less features but done with perfection. Bunch of bugs shouldn't be an explanation. about iR connection: I personally hate the business model of iRacing. They are disabling even forum access after you stop to pay for the service. Additionally gfx engine is 10 yo, there are missing many features available in other sims and even night racing has 'fixed time' like in GTR1 or NR2003 etc etc. But this sim grows month by month. New features are added and it slowly starts to be real competitor to rF-line sims. Especially that rF2 is still a big unknown. From this point of view, business model doesn't matter. When time goes on, more and more simracers will be moving to platform which is just better. For example, if driver swaps will be working perfectly in iR while rF fails, endure drivers will change platform without any sentiment.
There are very high expectations in almost any aspect of racing which can be simulated. And you are right, that's why I mentioned there have to be made compromises regarding rfactor2. I payed 40€ for rfactor and in relation to the amount of time spent on it that was the cheapest game I ever purchased. Perhaps ISI should think about releasing some quality track/car-packs afterwards similar to Simbin right now.
Excellant point on the car and track packs that simbin is selling right now. That would be a great idea in my opinion for ISI to do the same.
I agree. That would be a great idea. I could buy in the future high-quality track pack made by ISI, not even one.
Yeah. It would be great deal. but as far as we know, ISI has not enough resources to make tracks. I suggested once to begin ISI cooperation with modder teams and pay them for work, selling their products with ISI certification. But there was quite negative response from readers. I think a lot of people think that all community related products should be for free. On the other hand, I often hear moaning about SBTs add-ons containing "only" cars and tracks. People takes it like a mods considering it should be for free or so.
