[
https://issues.apache.org/jira/browse/OAK-775?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13641574#comment-13641574
]
Jukka Zitting commented on OAK-775:
-----------------------------------
bq. But then we are back to queueing (although way fewer items than in my
proposal).
Yes, that's the price we have to pay to maintain the association of (local)
commit information with resulting observation events (i.e. we can't just squash
multiple commits to one content diff). The good part here is that, as you
mention, we only need to queue successive root states, not individual
item-level events, so the queue is orders of magnitude smaller.
bq. Of course we'd need to devise a way for clients to specify such
requirements when registering event listeners.
That's the tricky part, especially since we'd need to dynamically adapt to new
listeners that get registered. My intuition is to keep things simple at the
NodeStore level and only apply custom filters at a higher level, at least until
we have a good benchmark that shows the need for a more complex solution.
> Implement backward compatible observation
> -----------------------------------------
>
> Key: OAK-775
> URL: https://issues.apache.org/jira/browse/OAK-775
> Project: Jackrabbit Oak
> Issue Type: Sub-task
> Components: core, jcr
> Reporter: Michael Dürig
> Assignee: Michael Dürig
> Attachments: OAK-775-isExternal.patch, OAK-775.patch
>
>
> As [discussed | http://markmail.org/message/6bqycmx6vbq7m25c] we might want
> look into implementing an alternative approach to observation, which trades
> some scalability for improved backward compatibility.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
