On Fri, Mar 16, 2012 at 6:24 AM, Yonathan <yonathan@...> wrote:
Should the specification encourage onsuccess
for reads and oncomplete for writes?
I don't think any amount of documentation is going to stop this from being confusing. Committing transactions should really have been an explicit request with its own result, and returning to the event loop without committing should have rolled back the transaction.
(What does "when a transaction can no longer become active" mean? It doesn't seem to be defined.)