Thanks for the feedback. I've fixed the typo identified by Matthias in the source.
why the overall score does not matter.
The short version is that we really only care about the license data. Theoretically, incomplete metadata ("described data") could impact the license data, but our requirement that there actually be license data has--in my experience--mitigated this risk. We're still learning our way through this, so it's possible that we may discover that we want to have more assurances that, for example, the source pointer is accurate.
I'll see what I can do to tweak the text. Your link suggestion is a good one.
I think the EDP should be a bit more explicit
FWIW, this is more about our implementation of the IP Policy than the EDP.
I've been putting off getting our content into the repository. In fact, I wasn't aware that they'd harvested any of JDT.
So... here's the really cool thing about ClearlyDefined: you can assist in the curation directly in the interface. I've updated the metadata (I gave them a pointer to the commit that contains the actual source code for that version) and fixed the license information. By fixing the source code pointer, the harvester will take another crack at it, so we should have better and more complete data before too long.
The curation manifests as a
pull request. They have a group of curators who validate and apply the changes (at some point, our IP Team will put some efforts into getting these privileges). This sort of curation works well on a case-by-case basis, but is a bit of a PITA for wholesale changes of multiple libraries and versions.
Getting all of our content in there has been on my list for a while, but low priority.
Wayne