Implements DefaultTlsServer.getRSAEncryptionCredentials() which may be required by certain cipher suites.

Fixes the list of cipher suites supported by the DTLS server so that it reflects the capabilities of the actual Bouncy Castle implementations at this time.

Cleans up the background Threads associated with TURN candidates more explicitly/aggressively/thoroughly in an attempt to mitigate a leak of such reported by Enrico Marocco.

Lets RTP packets generated from mute data sources from time to time in order to fix an issue at the receiver which could cause it to get out of sync with respect to the SRTP index of the sender. Issue reported and fix suggested by Emil Ivov.

Always uses Jitsi's own gain control because the use of PulseAudio's stream-specific volume control is reported to undesireably maximize the volume (https://java.net/jira/browse/JITSI-1050).

Fixes an issue which could cause SRTP to start failing to authenticate RTP packets after a number of RTP packets indicated as generated from a muted audio source were dropped.

Updates fmj.jar to r48 (Runs the SSRCCacheCleaner every 5 seconds (and not only on reception of RTCP packets).

Mitigates an issue with sample rate conversion in Windows Audio Session API (WASAPI) which leads to audio glitches.

Mitigates an issue with sample rate conversion in Windows Audio Session API (WASAPI) which leads to audio glitches.

Mitigates an issue with sample rate conversion in Windows Audio Session API (WASAPI) which leads to audio glitches.

before being started.

Uses UTF-8 instead of ASCII in Core Audio on OS X in order to potentially resolve issues with non-ASCII characters. (The solution is unlikely to be complete though because of the mixes usage of original UTF-8, modified UTF-8 and strlen.)

Fixes an audio quality degradation observed with Core Audio on OS X when the Audio tab in Preferences is open during a call.