I did some coding of a custom fragmentation system that works for my needs but in the process I am now getting this error on ‘Connect’. The error is quite vague and leaves me no clue as to what exactly is wrong. From what I have read it means mismatching configurations and channels. I have dumped both on client and host and they are identical, not to mention they share the same code.
What is quite odd about it too is, it always happens on the first build after changing code. If I build the client out enough, (usually 2-3x) it will eventually work. Once it begins working all code built after that will not have the issue. Once, however a single line of code is changed and recompilation begins, the same issue repeats.
Are there any other causes and or places someone could refer me to look to solve this issue? It would be nice if the error could give me more information about exactly what is mismatched so I know where to go to look to solve the issue. Right now I haven’t the slightest clue to what is causing this or how to go about fixing it.
crc mismatch means that you peers are used different configuration. It is not dependent how often do you compile your code. To avoid this error, put all configuration parameters in the separate class/file and reuse this class for all peers/ When you change this configuration do not forget recompile client and server…
in my project this error appears sometimes even equal builds (copied), so if this only should be affected by the configuration, there’s definitely a bug somewhere.
As @any_user has said, my project is equal builds and share identical code - I have even gone as far to dumping both configs on the client and host and they too are identical. I am feeding in a custom configuration though. If I leave the configuration as just a ‘new’ default one, the issue goes away. Something is going wrong when assigning custom parameters.
I reported it as a bug (Case 705744) and contacted support (#00182403) for assistance a week ago. It has halted my network development completely
I haven’t received this bug from qa yet… But, I will take a look on them today and will keep you informed. Sorry for delaying. To make response quicker in the next time please send me mail, forum is going to be huge and I just do not have time to monitor all records every day
What I found so far:
ping timeout is different on server and client (on server 75ms on client 50ms)
and in file Core.cs I found:
static void AdditionalConfiguration(RuntimeSettings runConfig)
{
runConfig.pingTimeout = (uint)(LocalSendRate * 1.5f * FrameDeltaTime * 1000f);
}
Looks like you calculate ping dynamically? So, if I understand correct it can be different on different machines? If yes - it is why you have crcmismatch error
That was indeed the issue. That code wasn’t even suppose to be active, but it seems I turned it on accidentally! Doh. It still doesn’t solve nventimiglia or any_user’s issue, but I would try to look back and see if there was anything you might have done to cause differing values on two built, identical copies. Overall sending back in my case ‘pingTimeout’ as being the problem child could have saved many frustrating hours.
How do you found out what’s different? I have this same issue, and whilst I have no doubt that it’s my fault, I’ve zero clue how to go about narrowing down the issue?
