On 13 Oct 2004, at 15:17, Ugo Cei wrote:
Il giorno 13/ott/04, alle 11:35, Luca Garulli ha scritto:
Why don't use JDO for the persistence? JDO allows to Daisy to be datastore independent. So you can run Daisy on top of an ODBMS or RDBMS or again the file system.
Probably because everyone has his own likes and dislikes when it comes to accessing data stores, so while JDO might be fine with you, others may prefer OJB, Hibernate, Cayenne, iBatis, straight JDBC
^^^^^^^^^^^^^
... and that's what Daisy query statements eventually get translated into: straight SQL-statements and JDBC calls. The Document object model might somehow fit into something an O/R tool could map between objects and tables, but we wanted to provide an easy, SQL-like query language for searches. And in order to be able to tune execution performance of the query language, we figured we'd have to do our own mapping as well, if only so that we know which performance issues we might encounter, and how we should optimize the database scheme in that respect.
Don't be too scared about MySQL dependency of Daisy, BTW:
"Daisy doesn't really depend on any special MySQL features (though a database with transaction support and row-level locking is required), but it simply wasn't a priority for us to support other databases as well."
(http://new.cocoondev.org/daisy/13)
HTH,
</Steven> -- Steven Noels http://outerthought.org/ Outerthought - Open Source Java & XML An Orixo Member Read my weblog at http://blogs.cocoondev.org/stevenn/ stevenn at outerthought.org stevenn at apache.org
