| Age | Commit message (Collapse) | Author |
|
There's a bit of an assumption here that handlers don't need to be told multiple
times about the same change (to different pieces of content) but that seems
reasonable.
|
|
It was possible (and I believe it happend) for new signals to be queued
up in _pending while some old ones were being processed. They would
then be cleared at the end of resume() without ever being sent.
|
|
I think possibly only the first call to ::instance() would be locked.
|
|
of content changes taken from an examiner are handled at the same
time.
Should fix DoM #2523.
|
|
|
|
|
|
|
|
|
|
changes, since we setup_pieces() in response to at least one of these and
hence we must know before it happens so we can suspend the butler and player.
|