Hacker News new | ask | show | jobs
by greenie_beans 328 days ago
thank you!

you can create book products three ways:

1. automated file upload via FTP (this is what most POS providers can do)

2. manual inventory file upload (requires the same file format as FTP sync, but as a simple file upload in the admin UI)

3. manual user entry via a web form (this is what i do for my own inventory since i don't have a POS)

each way will search for the title's bib data and use that to create the foundation of the product. does not work for books without an isbn, unfortunately.

the bibliographic data is evolving and something i like to nerd about. right now, i'm only using isbndb's database. it gets the job done for what i need right now but it's not perfect. when a book product is created, i take the isbn and fetch the bib data for that `Edition`. this data includes author, title, description, publisher, cover image, etc. a bookseller can edit the `Edition` by add/removing whatever they want. they can add additional information to the `Copy`, like the book condition, jacket condition, signed, first edition, etc. they can also add custom photos to the `Copy`, which is useful for collectibles.

i have more details about the data model here: https://docs.bookhead.net/docs/inventory#about-bookheads-dat...

if isbndb data isn't available for a title, then the web form has a way to search the open library database to auto-populate the bibliographic data as best it can. this is not that great tbh and something i might remove. one day, in the further future, i would love to create my own bibliographic database and share it with open library.

i've prototyped a data pipeline that will be able to take onix feeds from ingram and any publisher. ingram's data is the best, but their subscription is expensive, so i'm waiting for a paying customer who requires the best data. it also requires a lot of storage space and compute for the data pipeline. that data source will be a premium feature.