Implementing BIBFRAME seems a long way off. This needn’t be the case.
In reality, BIBFRAME’s backwards compatibility with MARC + cataloguing rules means there’s nothing in the way for potential implementers to produce data compliant to the BIBFRAME profile. Of course, BIBFRAME itself is far from finished, but any developer worth their salt should have flexibility enough to change their implementation on-the-fly.
Mapping MARC21 to BIBFRAME is an easy job; ingesting and using BIBFRAME data is slightly more complex as it involves actually doing something with other people’s data (there are also workflows and other non-technology irritations in the way). It should also be pointed out that it’s possible to do BIBFRAME without any real reference to a “living” linked data infrastructure; RDF makes no claims about the veracity of statements, so it’s possible to provide dummy URIs or internal references which can be improved upon later.
More problematic, however, is creating a true tool that uses BIBFRAME as its internal format; actually, that’s disingenuous — good support for the format is there (including the metadata management that is needed behind any triplestore*), but the workflow isn’t. It is rather easy to imagine an workflows for data entry for BIBFRAME, even if form-round tripping isn’t easy in HTML. Here, though, I guess that the road ahead is less clear — there’s heavy investment in architectures geared towards the existing MARC infrastructure.
BIBFRAME tools are being created apace, but seemingly only by LoC and partner. I expected to see results from serious research from the big players coming out in production systems last year, but this hasn’t happened. This inertia is worrying for the whole of the BIBFRAME effort; without dynamic commercial partners, BIBFRAME threatens to become an academic exercise much as the linked data resources are that we developed four years ago.
* It’s obvious that some researchers have rather bogged themselves down in this particular issue without reaching any particularly insightful resolution. Suffice to say, thinking about this a little less hard makes for a simpler solution.