undefined | Better HN

0 pointscdrini5y ago0 comments

Open Library has a notion of "Works" which group together editions across languages: https://openlibrary.org/works/OL63073W

(Note the work id is in the url: OL6307W)

If you want to get the work id for a given ISBN: https://openlibrary.org/isbn/2070360075.json will redirect to the edition page, and there you can get "works[0].key".

Or, you can search by the isbn: https://openlibrary.org/search.json?q=isbn:2070360075

0 comments

loosetypes5y ago

Thanks. So Open Library implemented that because there's no interoperable, universal concept for literary "works"?

And then each of Goodreads, Amazon Retail, Ingram, et al. would likely have their own internal system for isbn identification and grouping accordingly?

And I guess that means similarly, there's no isbn-analog for a series or interrelated set of "works"?

cdriniOP5y ago

Yep! They do. But the book-universe has largely landed on similar definitions for what a "work" is.

E.g. here's GoodReads work id (in the url): https://www.goodreads.com/work/editions/69741-the-old-man-an...

Here's works/editions in schema.org: https://schema.org/CreativeWork (work) ; https://schema.org/Book (edition)

Here's works/editions in Wikidata: https://www.wikidata.org/wiki/Q47461344 (work) ; https://www.wikidata.org/wiki/Q3331189 (edition) (e.g.: https://www.wikidata.org/wiki/Q19320911 )

All of these are loosely based on the FRBR model in librarian science, where a work is a work, and an edition is called an "Expression": https://en.wikipedia.org/wiki/Functional_Requirements_for_Bi...

And yeah, there's no isbn-analog for series; each site has its own identifier.

Also as a side-note, ISBNs are very inadequate as a unique identifier for an edition; there are frequent cases where an ISBN will be re-used for books with different covers (Dover classics does this a bunch); some modern books published in odd ways are not given an ISBN (e.g. all of these books: https://sorted.club/books/ ); and, most importantly, books published before ISBNs went into effect (~1970) have _no way_ of being accessible when using ISBNs as a unique key.

loosetypes5y ago

Wow, most of that is new to me. I appreciate you taking the time to reply.

j / k navigate · click thread line to collapse

0 pointscdrini5y ago0 comments

Open Library has a notion of "Works" which group together editions across languages: https://openlibrary.org/works/OL63073W

(Note the work id is in the url: OL6307W)

If you want to get the work id for a given ISBN: https://openlibrary.org/isbn/2070360075.json will redirect to the edition page, and there you can get "works[0].key".

Or, you can search by the isbn: https://openlibrary.org/search.json?q=isbn:2070360075

0 comments

loosetypes5y ago

Thanks. So Open Library implemented that because there's no interoperable, universal concept for literary "works"?

And then each of Goodreads, Amazon Retail, Ingram, et al. would likely have their own internal system for isbn identification and grouping accordingly?

And I guess that means similarly, there's no isbn-analog for a series or interrelated set of "works"?

cdriniOP5y ago

Yep! They do. But the book-universe has largely landed on similar definitions for what a "work" is.

E.g. here's GoodReads work id (in the url): https://www.goodreads.com/work/editions/69741-the-old-man-an...

Here's works/editions in schema.org: https://schema.org/CreativeWork (work) ; https://schema.org/Book (edition)

Here's works/editions in Wikidata: https://www.wikidata.org/wiki/Q47461344 (work) ; https://www.wikidata.org/wiki/Q3331189 (edition) (e.g.: https://www.wikidata.org/wiki/Q19320911 )

And yeah, there's no isbn-analog for series; each site has its own identifier.

loosetypes5y ago

Wow, most of that is new to me. I appreciate you taking the time to reply.

j / k navigate · click thread line to collapse