Logic error: A device that answer 2 minutes after a call trigger the bad timeout
Since we removed the eventloop, this code doesn't make any sense:
Related to https://git.jami.net/savoirfairelinux/ring-daemon/issues/227
bool
JamiAccount::handlePendingCall(PendingCall& pc, bool incoming)
{
auto call = pc.call.lock();
// Cleanup pending call if call is over (cancelled by user or any other reason)
if (not call || call->getState() == Call::CallState::OVER)
return true;
if ((std::chrono::steady_clock::now() - pc.start) >= ICE_NEGOTIATION_TIMEOUT) {
JAMI_WARN("[call:%s] Timeout on ICE negotiation", call->getCallId().c_str());
call->onFailure();
return true;
}
Because handlePendingCall() is only done on negotiation success.