Probably I should be using the content repository instead of defining my own content item and revision tables. There are features in the CR data model I know I <i>don't</i> want, such as folders and templates, but that doesn't mean I can't go with the flow and use the cr_items and cr_revisions tables.
Michael Slater's post today over <a href="http://openacs.org/bboard/q-and-a-fetch-msg.tcl?msg_id=0001jz&topic_id=13&topic=OpenACS%20CMS
">here</a> mentions a technique for creating a categorization taxonomy for content items, and that alone is enough to make me want to use those tables.
I also need to store images in the database - what is the general opinion on using the cr_revisions table for this as opposed to rolling your own? Again, the possibilities for categorization are compelling to me. Has anyone had experience with creating a huge image catalog using the CR datamodel?
Originally I was trying to keep the ETP package as simple as possible, so naturally I avoided relying on the 28 files and 10000 lines of code in the acs-content-repository/sql directory. Of course I also recognize the benefits to be had from code reuse, and perhaps using the CR doesn't necessarily mean more complexity. If anyone can list the other benefits it brings here, that would be useful.