Michael Francis

Confirmed that your above suggestion works but I had to use the `mingw-w64-i686-gcc` variant of the `gcc` package instead of just `gcc`. Tested it with `mingw-w64-x86_64-` packages as well and...

Even if there wasn't a major new feature there have definitely been enough emulation and QoL improvements to justify a new version.

Typically we try to gracefully handle the freeze/crash before it would actually happen in Choco and print out a "(vanilla crashes here)" message or something similar. @MistUnky can you try...

I tried a `NET_SV_SendConsoleMessage` followed by a `NET_SV_DisconnectClient`. When testing with a dedicated server and one client, the client will get disconnected but never receive the message. When using one...

Thoughts on this? > I'm wondering if one way to deal with this would be to create a new packet type that is sent when the game settings are invalid.

I just ran through the demo in the latest version of Chocolate in Ubuntu as of my writing this and haven't noticed a desync. Can anyone else confirm?

In the latest commit on Ubuntu 19.04 this doesn't seem to be an issue. Can anyone verify on other platforms?

gdb output: ``` Thread 1 "chocolate-doom" received signal SIGSEGV, Segmentation fault. 0x000055555559ee2d in P_GroupLines () at p_setup.c:593 593 li->frontsector->linecount++; ``` `li->frontsector` and `li->backsector` are null.

``` Thread 1 "chocolate-doom" received signal SIGSEGV, Segmentation fault. 0x000055555559dabd in PTR_SightTraverse (in=0x5555556a64c0 ) at p_sight.c:60 60 if (li->frontsector->floorheight != li->backsector->floorheight) ``` Inspecting `li->backsector` shows that it is null.

@scfinniss any updates? Otherwise I don't think it's good to keep a nearly 3 year-old PR up.