Dave,
Thanks for your reply, and for clarifying the file-name issue. That sounds like a great solution -- just to handle the case where the person uploading a brand-new file doesn't specify a title, if I understand correctly.
What you describe doing with the content-repository sounds excellent.
For my kludge job, I just created a new table, fs_checkouts, with four columns:
file_id (references object_id of the file),
checkout_date (a timestamp),
checkout_user (a user_id)
checkout_comment (a comment that the user makes when checking it out).
Then, in folder-chunk-postgresql.xql, I added a couple of joins (ugly!) to the contents query, in order to get all of this information, along with the "comment" field in the revisions table for the live revision, into the @contents@ variable. Then I could display all of the revision comments, or the checkout comments for the checked out files, in the folder view, which gives a nice snapshot of the status of the project.
There is no "history" in my implementation: When a file is checked out, a record is added to fs_checkouts. When the file is checked in, that record is deleted. In keeping with the versioning of the content-repository, it would make sense to keep history, too. If checkout information was added to the cr_revisions table, then each revision could hold the record of one checkout. (This would also allow locks to be taken on different versions of a document, which might actually be useful in some situations.)
Take care,
Shawn
