And yeah, searching needs some work! That's on my task list for this month. Just this Friday I spent most of my day working on updating our search engine, Solr, from 3.6 to 8.7 (wip!). But search is a _BIG_ pain point. We're a small team with a big long list of things to do, but we are making progress! This year we updated to Python 3, switched most of our production environments to docker-based for easier deploys and to give open source contributors more control of production infra, added reading history stats for users, added a new interface for exploring books, worked on a novel recommendation system, added text selection to the online BookReader for public domain books, added GoodReads importing, grew our community, added the ability to search by classification, and much, much more (you can see highlights from our year here: https://github.com/internetarchive/openlibrary/issues/3891 ).
There is still _definitely_ a lot to do, but I think the biggest reason worth using/contributing to Open Library is likely its open source community. Anyone can jump in and help make improvements to the system (as they very often do!). Personally, I think it's more likely that a system with a community will survive/flourish than one maintained by a single person (I also wondered whether I should just create my own before contributing to and now working on Open Library!). And there are also loads of different tasks associated with a site like OL, which would be impossible for me to do if I was going it alone.
If you would be interested, checkout the GitHub repo: https://github.com/internetarchive/openlibrary . It's very active, and you can get an idea of how we work :)
How do you even get an ISBN for a candle?
ImportBot scrapes Amazon.com, matches product id with ISBN10 (which can be converted into ISBN13 without anything else IIRC), imports product as book.
Yes. There clearly only needs to be one way to identify things, so commonly larger systems just incorporate the smaller ones wholesale. The 13-digit system just incorporates the prior 10-digit system for International Standard Book Numbers with a 978 prefix called "bookland" (most prefixes in this system are geographic, but of course books aren't really from one single geographic region, so they're from "bookland") and adds a new set of possible codes. It also incorporates the entire 12-digit American "UPC" system.
Several other (less well known) systems were gobbled up the same way as "bookland", just allocating them imaginary geographic regions in the 13-digit system.
There's actually a fourteen digit system, but the lead digit tells you about how many of something are bundled e.g. so a distributor can distinguish a truck full of Pepsi cans from just one case or a single can in terms of things you can order. Lead digit 0 means "single" so if you know the 10 digit ISBN you can not only make a 13-digit EAN for that, you can make the 14-digit GTIN that means "just one of this book" which in most cases would be what you want.
There's very little control. E.g I just published a novel,and Amazon did not in any way validate that the ISBN I have them actually belonged to me - I had not yet registered the book data, so what I told them would not have matched anything they might have looked up.
In this case, based on the reviews on Amazon, it looks like someone changed the description of an existing product, and that the ISBN probably actually belongs to a book.
There is an explanation under "About the author" section :
This "author" was created to segregate those items which have ISBNs but are not actually books. For more information, see the manual and/or start a thread in the Librarians Group.
When an item which is not a book is imported via ISBN into Goodreads, it does no good to delete it: the item will only be re-imported as long as it remains on the feeder site. (Often these are book-related items which are assigned ISBNs by book publishers so that they can be tracked through their book systems.)
“Why not? What’s the big deal?” said someone who has never worked with garbage data.
"paying $2-10k+ to add an International Standard Book Accessory Number field to their software will never get approved for a bookmark that works as a compass, or a promotional Harry Potter bookend set."
Tough to disagree.