Tweet · 4 May 2017, 20:44 UTC

Replying to @SciPubLab and @csarven, @CrossrefOrg

Because the data model is severely underconstrained and extremely print/visual oriented (to the detriment of useful semantics).