Better shutdown

[Silex]

Only the last commit is relevant here, as the others are from #13

I'll rebase this PR onces #13 is merged.

This basically removes all the foo.kill. It looks like "it works" but I didn't give it a lot of testing.

By the way, does https://github.com/Silex/mjpeg-relay/blob/217904d862cade494d031667d9e30884a9e1e34d/relay.py#L26 really make sense? I know it works, but theorically how can quit() access global variables here? I guess I don't know python enough 😝

I think it's simpler to move the quit() body at the bottom of the try block, and remove the quit() calls. There's no real need to have multiple exit paths.

I was thinking of refactoring a bit more and replacing the foo.connected thing with a function instead, that way we can get rid of the variable and the check is always accurate (simply return the socket state).

Also, each of these "servers" should have a saner start/stop mechanism. At the moment the thread thing for each of them is a bit weird... not the thread in itself but its permanence. I think in general, "start" should start listening/open the socket and spawn a thread, and stop should close the socket/server and kill the thread.

I'll see what I can do, and ping you when I have something. Give me your thoughts in the meantime 😉

p.s: you code on windows? I was a bit surprised by the windows line endings.

[OliverF]

Will this work correctly when we're listening for clients on line 103 in a different thread but with the same socket? If this does properly close the port on Windows and Linux then this is much cleaner. The reason it used the kill flag and the quitsock is to avoid acting on the socket from two different threads. Perhaps this works fine though, I haven't tested.

[OliverF]

Regarding foo.kill please see my comment on the code.

quit() works but I suppose it should mark the variables as global scope. A better solution is to just pass them as parameters!

Making foo.connected a function makes sense, that way we can get the true value of the socket like you say.

And I agree, the start/stop needs a re-think :)

P.S. yes I do - I love Linux and I'd switch in a heartbeat, but sadly a lot of my programs run on Windows only. One day :)