Hacker News Reader

Hacker News

Comment by matt_morgan | Hacker News Reader

Full thread

matt_morgan

Or how about from a library.

https://www.worldcat.org/search?qt=worldcat_org_all&q=edward...

View on HN

Or the publisher:

https://us.macmillan.com/books/9781250237231

Or from an indie bookstore, using this site that will show you indie bookstores near you that have it available for preorder: https://www.indiebound.org/book/9781250237231

Given that there are like... seven copies in U.S. libraries, I think that's probably terrible advice.

Won't they order more on request? I thought that's how libraries work.

Also there are more than seven copies on order in the Brooklyn library system alone: https://borrow.bklynlibrary.org/r1s/iii/encore/record/C__Rb1...

Since Amazon says it's not available until September 17th and you can only pre-order it now, I think probably even those 7 are records inaccurately registered before the book is actually there, and there are probably 0! Or did some libraries get the book earlier than Amazon can? Possible, but my guess is not.

Fair enough! I was thrown off the fact that it hasn't been released by the Worldcat page saying that there were three editions. I assumed that it was not possible to have data quality issues of that magnitude.

Oh, Worldcat data quality is definitely very problematic.

At it's best, it definitely doesn't represent _every_ library copy.

It requires libraries to register their copies with the central database. All of these libraries are different organizations with different software, at varying stages of 'legacy', trying to interoperate with each other, usually without very well-resourced IT teams. Then WorldCat has got to figure out when copies at different libraries are copies of the "same" thing, and what "same" means.

The worldcat database is very useful, but it certainly has data quality problems.

Amazon also says "4 formats and editions": hardcover, paperback, kindle, audiobook. Technically I'd say these are "formats" not "editions" -- the text is the same in each, although you might consider the audiobook a separate "edition" maybe, as it's not text at all, it's a whole different sort of thing. (And the Amazon UI hides it behind a 'more' link, although still just referring to the whole list as "4 formats and editions").

I'd guess Worldcat knows 3 of those 4 -- maybe all except kindle. Worldcat, like Amazon, is not really capable of distinguishing "editions" from "formats".

I'm not sure I'd consider the "3 editions" a "data quality" error exactly, in this case. It does point to some of the complexities of figuring out what's out there in the bibliographic universe, and how to model it in a consistent way that makes sense to users. (What _is_ an edition vs a format anyway?). Amazon gets data and corrections from people trying to sell books there providing some data entry/correction labor for free. And Amazon's website and data are _core_ to their business. As well as from other DBs like ISBN. Worldcat has to try to piece things together from a bunch of disorganized under-funded non-IT-expert nonprofits, who may consider "getting good data to WorldCat" not the highest among competing priorities, along with other DBs like ISBN.

Neither WorldCat nor Amazon are _great_ at determining "what separate editions/formats exist of this thing, and how do they relate to each other" in the general case. Cause it's a hard problem. Amazon does well enough to sell books apparently. They each have strengths and weaknesses. For things published decades before Amazon existed and/or no longer in print, WorldCat will do better in some ways.

Maybe the long term solution is to put together an open source library database package that serves the needs of 90% of libraries without modification, and make it convenient to make high quality data available.

Worldcat / OCLC is pretty close to this. It's not fully open source, but it's widely used and pretty open.

(The fact that the API isn't publicly accessible without registration and payment is a major annoyance.)

Yeah, at one point OCLC was actually being kind of litigious with trying to make sure their db was _not_ open. I think they got over that, at about when they realized nobody actually really wanted it so much anymore. But it's debatable whether it's "pretty open" at present.

LibraryThing was sort of another attempt, that sort of still exists.

It's just a really hard thing to do, that takes a lot of resources to do well, and nobody's managed to figure out a funding model.

I don't think "Just create an open access database anyone can edit, like wikipedia but data, and books" will work (and actually, [that's](https://www.wikidata.org/wiki/Wikidata:WikiProject_Books) been tried too), but you can try to start another project if you want.

There are a handful of projects with various business models and degrees of openness of data that have tried or are still trying to do this. For whatever reasons (and we can debate em), they haven't really taken off or been succesful. shrug.

I think you are right. I called 3 libraries in my immediate ~15m drive area (university library, County, and City), all 3 confirmed they will have it, but none were listed on the World Cat site.

As noted:

1. The book is announced, but not yet released (17 Sept).

2. Worldcat tends to lag acquisitions, and whilst extensive, is not fully comprehensive. There are nonlisted collections, and lagged reports.

3. Further acquisitions will be added.

And most importantly: Libraries are highly responsive to patron requests. Ask for the book to be added, and in all likelihood it will be, and you'll be notified when it's in and/or added to the waitlist.

Or how about we all choose how we wish to get content, and stop implying judgment. When did reading become an us vs. them thing?

Is he implying anything at all or are the voices in your head?

The point, I believe, is to recommend a vendor other than Amazon, whose business model is datacenter management and datamining and being a front for cheap Chinese goods.

There is a bit of judgment, and that's okay.

Which added zero value about the book itself. But yet I’m downvoted for pointing out the stupidity. You win, Internet.

Considering Snowden sounded the alarm about privacy and surveillance by the U.S. government, as well as interactions with tech giants, I think the irony of buying the book on Amazon is relevant.

Reply on news.ycombinator.com