[Cosmo-dev] prototyping cosmo with jdbc instead of jcr

[Vinubalaji Gopal]

On Fri, 2006-07-07 at 17:54 -0700, Elliot Lee wrote:
> hmm... i know little about jcr/jackrabbit. however, i like this
> proposal because i've been working quite a bit with MySQL, which i
> believe is a typical RDBMS.

JCR/jackrabbit is a new approach of storing data and I look forward for
wide deployment and better performance.

Its just that you need to think everything in terms of a tree with nodes
and properties in jackrabbit. 


> dbms) is very well-known with excellent documentation and references
> available.
In the API side, If Hibernate is used - you still will not really see
the complete Mysql queries using JDBC. But yes you can query the
database and see the result instantaneously.

JackRabbit can be configured with Mysql or any other database, but you
will be seeing nodes and properties in the database and that is barely
useful, unless a custom class is used to display the entire repository.


> and of course, mysql is also very fast, and many people have optimized
hm yes it is - but I like PostgreSQL because of it more open nature and
it completeness. If Cosmo is changed to use a traditional RDBMS, I guess
it will  surely not use any vendor specific SQL queries and will be
usable with any of the open or closed RDBMS system.