bci2000.org BBS

I'm building BCI2000 under Linux, but my compiled program fails to get the Operator connected to the other modules, so after it starts up, and I press the Config button and set the configuration, it never advances to the next state and turns on the "Set Config" button. I've been rooting around somewhat futilely trying to find where this connection is supposed to get made and the button turned on, but I can't find it - can anyone tell me where to find this, what cpp file it's in? I suspect there may be some Windows-specific socket stuff, but maybe the problem is simpler than that, I'm just having trouble finding it.
Also, Linux can't build the "3DAPI" stuff because it uses Windows DirectX, but it still gets invoked in the "Make Unix Makefile.sh". Is there a straightforward way to remove this? I'm not familiar enough with cmake to figure this out, so I've been slashing at the generated makefiles when it probably just requires commenting a line or two in the right place....
Thanks!
Dave Britton

I'm building BCI2000 under Linux, but my compiled program fails to get the Operator connected to the other modules

This is a known problem under Linux. The source file in question is src/shared/utils/SockStream.cpp which provides a cross-platform stream interface to sockets. It uses the generic socket API which is available on both Windows and Unix-based systems, and it has been working under Linux until rather recently. I suspect a concurrency issue but it must be rather subtle because BCI2000 uses mutexes to protect socket streams that are accessed from multiple threads.

I'm sorry about this issue but I have currently no time to fix it.

Also, Linux can't build the "3DAPI" stuff because it uses Windows DirectX, but it still gets invoked in the "Make Unix Makefile.sh".

The 3DAPI does not use DirectX but is OpenGL based. You need to install the MESA library in order to compile it under Linux.

Best regards,
Juergen

Hello, sorry to post on a 3 year old thread, but I am having the exact same issue. I have no problem with getting into the code and patching it up, but could you clarify a little more what the issue is? Or has the nature of the problem changed since it was first noticed?

I noticed it also had significant problems in compiling some of the contribution files, but I removed the ones that were breaking it for the time being.

jmarple,
I'd suggest reading/posting in this thread.

I thought about it, but this isn't a compilation problem. The program compiles fine, it just won't connect.

Unless you were referring to the second half of my post? In that case, yes, I will add to it if I don't find a way to fix them soon.