I'm curious what you think of trying to keep the hibernate tagnames
matching the mapping DTD one for one.
For example, <join> has a <key> subelement in the mapping DTD, so I
would prefer to have hibernate tags named @hibernate.join (which we
have) and @hibernate.key (which is currently join-key instead).
For cases where a element-tag is ambiguous and may be placed in more
than one parent element, then the parent element could have an
"element-id" parameter and the child could have a "parent-element-ref"
parameter, to explicitly specify how the mapping should be formed.
I think this strategy ultimately makes life a lot easier for the user,
as they only need to reference the hibernate documentation and mapping
DTD in order to write their java class (with very few exceptions,
anyways). Any mapping element naturally translates to a @hibernate tag
of the same name so there's no need to go hunting through xdoclet
documentation to find the tag that does what you need.
I'm hoping to implement a strategy like this for the xdoclet 1.3
release, but I won't if it makes migration to xdoclet 2 by changing
basic tag names...
Any thoughts?
Paul
