[
https://issues.apache.org/jira/browse/JCR-3776?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13982818#comment-13982818
]
Chris Poulsen commented on JCR-3776:
------------------------------------
We use the JCA stuff at work and it works for us (we use it for XA
transactions).
I agree that the the docs are not great on anything related to jcr /
jackrabbit, but I do not think you are right in everything being utterly
broken...
Obtaining the correct session during xa transactions can be a bit tricky, but
that is related to JTA and not a jackrabbit issue.
There are a couple of weird issues related to transactions in jackrabbit, but
no show stoppers so far (and to some extent, probably corner cases). Most of it
seems to be related to the fact that various indexes (for queries) are not
updated before a transaction is committed.
Your stackoverflow post looks somewhat suspicious, the code looks way more
complicated than needed. For example we use SimpleCredentials and it is located
in the jcr-2.0.jar ( http://mvnrepository.com/artifact/javax.jcr/jcr/2.0 ).
Its use is simply: Credentials credentials = new SimpleCredentials( "admin",
"admin".toCharArray() );
And to obtain a session we lookup the repository from JNDI (for integration
testing with full transaction support we just use a transient repository wired
up outside a web/jca container).
> The JCA is completely broken
> ----------------------------
>
> Key: JCR-3776
> URL: https://issues.apache.org/jira/browse/JCR-3776
> Project: Jackrabbit Content Repository
> Issue Type: Bug
> Components: jackrabbit-jca
> Environment: Linux/Geronimo
> Reporter: Stuart Wilson
> Priority: Critical
>
> Over sixty methods of the JCA Session Handle do not work, and cannot as the
> JCA exists today. I do not think that the JCA should be an Apache project.
> It's existence at this point amounts to a stumbling block for developers. It
> reflects very poorly on the project and the Apache Foundation as a whole. The
> website has an unacceptable number of dead links - mailing lists et al. The
> project seems to be still developing, but there is no indication of what they
> are working on. I will go so far as to say it was sabotaged. (See my post on
> Stack Overflow,
> http://stackoverflow.com/questions/23186713/has-the-jca-ever-worked-it-cant-according-to-its-documentation)
> No one would have developed all of the code in the JCA just to see it lined
> out. And no one would have associated this with JSR-283 if it could not, as
> it can't access a repository node.
> I am submitting this because there is no place else that I can, short of
> trying to complain to the Apache Foundation itself, which I hope to
> eventually be able to do.
> I wasted a considerable amount of time working on this and am left with
> nothing to show for it. For me it is a Critical issue, although I have no
> hope of seeing it resolved.
--
This message was sent by Atlassian JIRA
(v6.2#6252)
