I just took a few minutes and built a simple MySQL table to store the 51 Dublin Core metadata elements that make up the “core” and “other elements”. Except for elements that are dates, I used “text” fields for everything so that it can hold anything you can throw at it, while still being indexable and searchable.
I haven’t updated the Lightweight Asset Manager to use the full schema yet, but here’s the raw SQL to generate your own copy of the table if you want it.
There is a small chance that there are some typos etc (since I haven’t tested it yet, and I built it while stretched out on the couch watching the finale of the Race).
{ 3 comments… read them below or add one }
[...] I finally made time to install Rails, planning on taking it for a test drive by making a Ruby on Rails version of the lightweight asset manager database we put together to store assets for Pachyderm. [...]
When reading your article I was exited…at first. But I expected a lot more ;-(.
Dublin Core allows you to apply any metadata element multiple times to one object of data it describes. This is a strengh not a weakness. For instance you can apply a subject or author or contributor multiple times. I did not (yet) dig too deep into Dublin Core but futhermore you can cascade metadata descriptors like author.nickname (nickname is a more exact definition of the author description i believe).
Anyway… whats the use of that simple table ;p.
Lets discuess – what are you going to do with dc, why are you using it. I am really messing around with metadata in my mind lately and I cant get to a clear concept in my mind, yet.
(i left my email, i hope i ll be notified if there is any futher discussion, i found this blog entry by feeding google with )
Jonas, thanks for the feedback.
The simple table is quite useful when you’re needing to manage a straightforward set of records. I’m using it on a few projects to manage assets (images, videos, URLs, etc…) where the multiplicity would have been more difficult to implement, and wasn’t needed by the metadata.
I’m also coming from the perspective of having previously implemented IMS/IEEE LOM a couple of different ways, so having a barebones simple “library card” record database has been very refreshing…