MSVC support #8
MSVC support #8
Conversation
…`source_group`s [src/{daemon,main,options,utils}.c] MSVC support (WiP); [README.md] Use generator independent build command
| @@ -42,7 +44,10 @@ target_link_libraries(violet PRIVATE LibJuice::LibJuiceStatic) | |||
|
|
|||
| install(TARGETS violet RUNTIME DESTINATION bin) | |||
|
|
|||
| target_compile_options(violet PRIVATE -Wall -Wextra) | |||
| target_compile_options(violet PRIVATE -Wall) | |||
There was a problem hiding this comment.
Does -Wall work for MSVC? Is it translated by the CMake generator?
There was a problem hiding this comment.
Well it didn't error at least
| @@ -27,7 +27,12 @@ | |||
| #include <stdlib.h> | |||
| #include <string.h> | |||
| #include <time.h> | |||
| #ifdef _MSC_VER | |||
| #include <processthreadsapi.h> | |||
| #define localtime_r(timer, buf) localtime_s(buf, timer) | |||
There was a problem hiding this comment.
Since the return type is also different, I think defining a proper helper function like in libjuice would be better:
static int get_localtime(const time_t *t, struct tm *buf) {
#ifdef _WIN32
return localtime_s(buf, t) == 0 ? 0 : -1;
#else // POSIX
return localtime_r(t, buf) != NULL ? 0 : -1;
#endif
}
|
Thank you, I dropped Windows support for Violet because it was a bit cumbersome and I didn't think there would be a use case for it, even if the underlying library libjuice is compatible.
I could change to GPL but I'm not considering a change to a permissive license, see paullouisageneau/libjuice#115 (comment) |
Co-authored-by: Paul-Louis Ageneau <paul-louis@ageneau.org>
|
BTW: Two platforms I was thinking to port this to next are iOS and Android. For my TVconsole hobby project. |
| @@ -85,7 +90,11 @@ int main(int argc, char *argv[]) { | |||
| goto error; | |||
| } | |||
|
|
|||
| #ifdef _MSC_VER | |||
| SuspendThread(/*TODO*/); | |||
There was a problem hiding this comment.
This is still WIP, right?
There was a problem hiding this comment.
Yeah. My main question at the moment is whether running a STUN & TURN server on a node will even make a difference, i.e., if this use-case could be facilitated: https://stackoverflow.com/q/68600741
(my thinking is not which is why I've quasi-abandoned my contribution to the various open-source C STUN/TURN implementations)
There was a problem hiding this comment.
STUN est basically the standard way to implement NAT hole-punching (in particular in the context of ICE). The STUN server must be outside of the NAT you are trying to traverse (hole-punching always relies on a reachable third node outside the NAT), therefore, deploying a STUN server on one of the two connecting nodes won't make any difference.
There was a problem hiding this comment.
Yeah that's the same conclusion I came to (eventually).
😢
WiP
PS: Any chance for a license change? - I always use
(Apache-2.0 OR MIT)though also have a likening to zlib and CC0.