zhouliuya

@jam-le Hi～ I encountered the same problem as you, have you solved it now? In other words, is there a solution?

@kuaashish Thanks for your reply, In fact, the day before yesterday, I had simply implemented it myself, but the writing was really ugly. It was just enough to satisfy my...

> Please remove the punctuation (`.`) from the commit message title. (You can edit it by `git commit --amend` and force push to GitHub.) done.

> Wouldn't doing this in `RTMPSockBuf_Close` be a more appropriate place? You are right.

> Out of curiosity, where exactly is that stack from? I'm unable to find any references to a liblive-push anywhere. Is this something based on librtmp? ```liblive-push``` is a library...

> Is this something that you are distributing? Is the source available? I'm reluctant to accept a fix based on a project that might be violating the librtmp license without...

> That should be fine then, no need to delay for your release if it is just internal for the moment. Thanks for the context, we were just curious about...

Ok, thank you very much for your reply, but I have a question, if I don't use ```registerAudioDeviceCallback()```, but only use ```onErrorCallback()```, will oboe select the device for playback internally?...

> After you get an onErrorCallback(), you should open a new stream. > > You should get an onErrorCallback() when a headset is connected/disconnected. emmm, Whether I switch playback devices...

> Ok, very good, just please @lilinxiong do the fixes: > > 1. Remove ANDROID from checks. The existence of `pthread_condattr_setclock` is ALWAYS required, if you set monotonic clock, regardless...