[
https://issues.apache.org/jira/browse/JCR-1232?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12545383
]
Thomas Mueller commented on JCR-1232:
-------------------------------------
Both patches have their own disadvantages:
- Jukka's patch is slower (more temporary objects), and copies source code.
- My patch is faster and avoids copying source code, but it may not be as
architecturally sound.
I can live with Jukkas patch as long as there it is fully backward compatible
(as long as the UUID class is still available, and as long as NodeId.getUUID()
is still supported).
So I take back my '-1', and vote '0' ;-)
Thomas
> Merge UUID to NodeId
> --------------------
>
> Key: JCR-1232
> URL: https://issues.apache.org/jira/browse/JCR-1232
> Project: Jackrabbit
> Issue Type: Improvement
> Components: jackrabbit-core
> Reporter: Jukka Zitting
> Priority: Minor
> Attachments: nodeId.patch, nodeid.patch
>
>
> The current NodeId class is mostly just a wrapper around UUID, which causes
> two objects to be instantiated for each node identifier that the system uses.
> The memory and processing overhead is quite small, but given that there are
> tons of NodeId instances it would be good to eliminate that overhead.
> There is also lots of code that just converts UUIDs to NodeIds and vice
> versa. We could simplify such code if we just used NodeId everywhere.
> Also, we might want to open up the possibility of using non-UUID node
> identifiers at some point in future, so it would make a lot of sense to
> remove the NodeId.getUUID method and rely directly on NodeId and it's
> equals(), hashCode(), and toString() methods in many places where we
> currently use UUIDs.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
