I would really like to hear the reasoning behind this move, and soon. XPath support was one of two ways to access data in v1 of the spec. I can't think of a good reason to drop the support entirely, and can in fact think of reasons why the second version of the JCR spec should go further than the first and include XQuery, which goes much further than xpath can.
I think the new Java object query is probably going to have some very nice features in certain use cases (ie. query by example). However, the choices for querying in the new version just became: 1) SQL2 2) JQOM SQL support is good, a lot of people know it, but its old and wasn't made with hierarchical data structures in mind and JQOM will *assumably* be good, but its specific to JCR. Why limit querying to a dated standard or something that is specific to JCR when XQuery is a new standard specifically made to support this use case and is supported by many of the vendors on the 283 committee? Not to mention the fact that everyone that used xpath in the first version just got dumped by the wayside... Furthermore, (leaving the dated and sometimes awkward SQL standard aside for now) why should querying data in JCR (ie. JQOM) be different than querying data from an XML database? Seriously, think about it...seems like we went through this before with accessing relational databases in the 90's and JDBC was defined so that we could access databases using a common standard. We all agree that the JDBC standard was 'a good thing' right? -XQuery has been specifically defined to handle querying of hierarchical data -Its a recommended standard from WC3 -Many committee members support it and helped define it Seems clear to me that it should be in the spec. Regards, Mark
