Benjamin Höglinger-Stelzer

Considering this done.

Yep, that race/timing condition was a huge problem with the previous design of how driver and the client SDK handled these packets, this has been fixed for good in 1.17...

@mika-n which binaries for dotnet do you use and from where exactly? Built from sources? Which branch, tag etc.? I haven't merged the changes to dotnet master branch yet so...

Thanks for the summary, sounds pretty much the same I did the testing with a simple tester app, but the native C/C++ lib. I'll cook up a simple .NET console...

Here's the test with a dead simple console app using v1.17.x of the **C library** (and v1.17.x driver):  Es expected, the last pair of zeros signalling both motors off...

Here's the same test working as expected with the latest (`master`, v1.17.x) .NET library:  The code: ```csharp using System; using Nefarius.ViGEm.Client; namespace ViGem_v17Test { internal class Program { private...

Another "extreme" example with a `Thread.Sleep(1000);` in the event handler to simulate slow feedback processing:  No event is lost, it just arrives later in the sequential order it was...

Might be, I provided the tests and results just as a token that I've not gone insane and it actually does what I claim 😅

Dumb question, are you sure that you test with the latest 1.17 dotnet lib **and** native 1.17 lib embedded into it? Not gonna discredit your efforts but it happened quite...

We could also have a chat on Discord about it.