The server's /bibliography endpoint currently serves bibtex information, but does so rather strictly. If a bib-tex field is DOI, then it is not read, because we expect doi. This should change, and we should eventually just allow case-insensitive fields, just in case.
it's not clear to me whether this should only happen on the server side for this particular API, or whether it should happen generally in the BibTex-reading. I'm leaning towards the former (only on the server), because I don't know whether stripping all case-sensitivity from the bibtex-parser might have other consequences.