I'm moving this thread from ojb-dev to commons-dev, as this list seems to be more appropriate for this discussion.

Axel Hohaus wrote:

>How about joining forces with Jakarta Commons SQL project?
>(http://jakarta.apache.org/commons/sandbox/sql/)

it's an cool idea. We didn't know about Commons SQL until now.

>The easiest point of integration would be to make a template for
>generating the Commons SQL database schema XML file (should be easy).

That's correct. We can replace the ddl-template by an
commons-sql-template and integrate commons-sql in our ant build file.

>Taking it further, I think two projects could be merged, removing the
>need for overlap:
>* AXgen is used to generate database schema
>* database schema is used to generate entity classes, xml
>repository and
>DDL (at least DDL is already supported by Commons SQL)

Could you clarify this point.

Right now, if I understand everything correctly, then the two projects have the following features.

Commons SQL:
* transform schema.xml -> database specific DDL
As this code is originally from Torque, then it should be fairly easy to implement also:
* transform schema.xml -> object model (java classes)
* transform schema.xml -> OJB repository.xml
because these features are implemented in torque...
BTW, maybe someone from Commons SQL team can comment, why only DDL generation was extraced from torque and not the other generation features?

AXgen:
* transform UML -> entity classes (specialized object model?)
* transform UML -> OJB repository.xml
* transform UML -> DDL

So, if we would combine these two, we could have:
* transform UML -> schema.xml -> object model (entity or whatever)
* transform UML -> schema.xml -> repository.xml
* transform UML -> schema.xml -> DDL
+ additional features you mention below.

>These two projects are just too similar in my mind and should
>join the
>efforts...(if authors can agree on terms, of course)

Is it possible to read structure information from database with
common-sql? This could be an interesting point to use Axgen with
exisitng database structures and enabling a way back to XMI.
Is it possible to analyse differences from commons-sql XML-Input and an
existing database structure to creating ddl output for enhancing the
database sctructure?

Cooperation of these projects seems to be very interesting.
As far as I've understood (also mentioned above), Commons SQL is doing only one thing (transform schema.xml -> database specific DDL). Not sure about the features you are asking for:
1. transform live database -> schema.xml
2. diff live database and schema.xml

AFAIK, torque has the ability to perform #1, but not #2.
(http://jakarta.apache.org/turbine/torque)

Rgds,
Neeme


--
To unsubscribe, e-mail: <mailto:[EMAIL PROTECTED]>
For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>

Reply via email to