Fix for #14 - Buffer events always and check buffer always after socket polling

[holgerk]

No description provided.

[jakubkulhan]

Hi. Thank you for the PR! And sorry for such a delay :( I've looked through the code and it looks good, I've run current test suite and it passes. Could you add a regression test case, i.e. test that fails with the old version and gets fixed by this patch?

[holgerk]

Hi, you can checkout 96a5d5b to see the failing test for the DevtoolsClient.

A problem that i see with the current solution, is that the buffer is not flushed.
I think we could flush the buffer when a command is executed from stack level 0.
Commands issued from callbacks (stack level > 0) should not flush the buffer, because of possible awaiters from lower levels.
What do you think?

[oscx]

@jakubkulhan can the pull request be confirmed or any conflict from manual review pending?