Hi Morgen,

>> Grant and I discussed several ways of fixing this.  We could doing share
>> restore in parallel, and instead queue up all the collections to be
>> subscribed to.  This would slow down share restore, but would lead to
>> fewer merge errors.
> 
> I think this also means never having more than one sharing operation
> going at any given time, including background syncs, and subscribes, right?

Blocking subscribe workflows during a background sync doesn't seem like
a good outcome.  I couldn't remember the circumstances in which parallel
sharing threads are likely to happen, but that's it.  That makes me lean
away from avoiding parallel merges, instead we can think of parallel
restores as a good bug finding tool for those rare simultaneous
subscribe/background sync cases. :)

Sincerely,
Jeffrey
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

Open Source Applications Foundation "chandler-dev" mailing list
http://lists.osafoundation.org/mailman/listinfo/chandler-dev

Reply via email to