On 2012-02-03 19:44, Michael Dürig wrote:
...
Sounds very interesting.
I have a comment that may not even apply to this work, but still:
recently a JCR issue was opened because Jackrabbit doesn't fire
compliant NODE_MOVE events for reorder operations. Turns out that
Jackrabbit doesn't keep sufficient information, and then tries to guess,
and sometimes guesses wrong.
We should make a conscious decision whether it's important to have this
information in observation events. Based on that we may want to tune the
JCR 2.1 spec.
Best regards, Julian
