CryptoExecutor and DatabaseExecutor now use bounded thread pools with
unbounded queues, since running too many tasks in parallel is likely to
harm performance; IncomingConnectionExecutor, PluginExecutor and
ReliabilityExecutor use unbounded thread pools with direct handoff,
since their tasks may run indefinitely. There are no longer any bounded
executors, and all executors discard tasks when shutting down, which
fixes issue #3612189.

Responsibility for starting and stopping services has been moved from
BriarService in briar-android to LifecycleManagerImpl in briar-core.
However, BriarService is still responsible for stopping the
Android-specific executors, which is ugly. It would be better if
executors registered themselves with LifecycleManager.

If the group's owner is spamming, unsubscribe.

Find the newest dead secret for each endpoint, erase any others and
derive from the newest.

(Although it isn't possible to create an identity yet...)

The invitation and private messaging UIs are currently broken. Some key
rotation bugs were fixed; others may have been created (unit tests
needed). An encoding for private keys was added. Pseudonyms were moved
out of the messaging package and ratings were moved in.

The Tor plugin tests are consistently failing with the latest versions
of Silvertunnel and Spongy Castle; we may need to ship native Tor
binaries for Windows/Mac/Linux and use Orbot on Android.

Fields in Android UI objects that are accessed from background threads
must be declared volatile. UI objects use data attached to DB events to
avoid DB lookups, which complicates the UI code but should improve
performance.

Generate a random UUID instead of using a fixed UUID. Close sockets when
exceptions are thrown (not doing so can cause problems with subsequent
sockets on Android). Use a semaphore with tryAcquire() instead of a lock
when making alien calls, to avoid possible deadlocks.

Only the bundle contents created by Briar classes are encrypted.