On 2012-10-01 16:20, Julian Reschke wrote:
On 2012-10-01 15:57, Carsten Ziegeler wrote:
2012/10/1 Julian Reschke <[email protected]>:
I'm not sure that it's a good idea to call something "Jackrabbit"
that is so
different in what it supports compared to "current" Jackrabbit.
Just from the peanut gallery: we did the same years ago with Cocoon,
the 2.0 version was 100% different than every previous version, even
migration was not possible. And it worked out pretty well - in the
beginning everyone referenced the new version as "Cocoon 2" (like JR
3) in contrast to just "Cocoon" for the old one. But over time this
changed and wasn't a problem at all.
Regards
Carsten
Offlist: wenn das nächste CQ auf defaultmäßig auf Oak läuft bin ich
überzeugt :-)
Well, the off-list part didn't work so well. Sorry for that.
Explaining my concern: in order to achieve it's goals, Oak needs to make
compromises with respect to JCR features that have proven to be tricky
to implement in a performant way, for instance, ordering of child nodes
and support for same-name siblings. If you have an existing application
on top of JCR that uses nt:unstructured a lot, it will be hard to figure
out what parts of the content tree will actually rely on these features
and which do not.
Best regards, Julian
