First off, huge credit to @omg.jacky.wtf for actually getting this moving.
A bunch of us have been talking about books, bibliographic data and shared lexicons for a while, but mostly with our heads down building our own projects. Jacky got people into a room and started turning that into something concrete. That’s important.
I think the next step is getting the builders together and treating this a little like what happened with Standard.site.
We already have working code and different approaches in the wild. Bibliograph is now doing some really interesting work on the bibliographic side. BookHive, @xardex.dev, has a working book and reader implementation (CAVU is working on a reader - @chris.pardy.family). Byline.pub, me, has been working on author and book records. There are other projects coming at this from reading, publishing, metadata and discovery as well.
None of that work should be thrown away. Quite the opposite. That’s the material we should be bringing to the table.
I also don’t think we should assume any existing book.*, *.book or project namespace automatically becomes the canonical book lexicon. We aren’t really at that point yet. The useful question is: what is common between the things we’re already building, and what should apps across the ATmosphere be able to rely on?
I’d suggest we establish a small Book/Bibliography Lexicon working group.
Not an IETF bureaucracy. Nobody needs another committee eating their evenings. 
A group of the people actually building this stuff.
Maybe:
-
gather the existing schemas and implementations
-
identify the common primitives: works, editions, identifiers, contributors, authorship, reading state, reviews, relationships, etc.
-
run a few workshops across different time zones
-
document the places where our implementations disagree
-
get something resembling a v0 proposal
-
put it out publicly for comment
-
implement it in several real applications
-
learn where it’s wrong
-
then publish the shared lexicon
Basically: rough consensus and running code.
I’d also really like to invite some of the folks involved with @standard.site into that discussion — @aka.dad, @brookie.blog, @awarm.space and anyone else I’ve missed. They’ve already gone through the exercise of getting several independent AT Protocol applications to converge on common lexicons, and there are probably lessons there that would save us some bruises.
Jacky’s thread is a damn good starting point. Bibliograph gives us something concrete to experiment against. The various book apps give us real-world requirements.
Now we need to get those people around the same virtual table.
Different projects. Different time zones. Different needs. But we all have a stake in getting this right.
I’d be very happy to help organise it.