Canonical schema naming

[Jörg von Frantzius <jo...@artnology.com>]

Hello JDO developers,

so far the spec doesn't prescribe a canonical naming of schema elements, 
as far as I know. As a consequence, it is up to the JDO implementation 
to generate names for elements the user didn't provide any orm 
information for, such as column names for FK columns.

For systems being written from scratch, i.e. without any need to 
reverse-map an existing schema, it is the easiest and therefore probably 
the most common way to not provide orm information for naming of 
columns, as the JDO implementation will generate them and why bother any 
further. The resulting schema being created then is tied to the JDO 
implementation in use, making it hard to swap out the JDO implementation 
later on after there is some data to keep in the database.

Of course it is possible to provide a. .orm file for making the JDO 
implementation B work with the schema generated by implementation A, but 
that can be very cumbersome, depending on the size of the schema.

My suggestion would be that the spec prescribes a canonical naming, 
maybe by just explaining in prose what the reference implementation JPOX 
is doing?

Thanks for listening,
Jörg

-- 
__________________________________________________________
Dipl.-Inf. Jörg von Frantzius  |            artnology GmbH
                               |                Milastr. 4
Tel +49 (0)30 4435 099 26      |              10437 Berlin
Fax +49 (0)30 4435 099 99      |  http://www.artnology.com
_______________________________|__________________________