Adding certain books loses cover image and/or ISBN numbers

Here's the ISBNs for a couple of books I added today (I'm trying out LibraryThing and making sure it works properly before I add the bulk of my collection):

0578630028
9780062970701

Both of these show the proper image (via Amazon) when I scan them via the Android app. Both promptly *lose* that image in favor of the "default" blank leather cover once I add them.

Turns out the same thing happens if I just manually enter ISBN numbers on the website here.

When I review the entry for each, I see that the first (railroading) book shows the scanned ISBM (ISBN-10) number in the ISBN field... the latter doesn't even show the ISBN (ISBN-13) number at all in that field.

These seem like pretty fundamental problems - and while the first book is uncommon the latter certainly isn't. Are these known or common issues here?

This isn't a bug. Amazon images depend on Amazon and when they change or delete a cover for a given ISBN, it is changed or no longer available. I'm guessing search results still show them because they are cached. Either way, if you want covers to be stable, you need to chose member uploaded or upload your own as those are hosted on LT.

As for ISBN, the field itself will contain the ISBN provided by the source you use, so if they use 10-digit that one, but your book details will contain both and AFAIK your catalogue is searchable by both 10 and 13-digit one either way. Choosing a good source for importing book data is important if you want accurate data. In general, library sources are much better than Amazon.

You'd think the cache would expire eventually - perhaps that cache needs to be refreshed.

That's if caching would be related to the issue, which it evidently is not as we see the images on both Amazon pages:

https://www.amazon.com/Rails-Around-Indiana-Hoosier-State/dp/0578630028

https://www.amazon.com/Strange-Planet-Nathan-W-Pyle/dp/0062970704

This also doesn't explain why ISBN isn't being saved with the book (if it's somehow not on Amazon then it's certainly part of the barcode search term itself, so it's a known quantity).

>3 martymacgyver: Truth is the search is just that - a search. Nothing is saved from the search. Covers are not brought across, and items scanned are not maintained. All data that goes into your catalog come from the selected record. The covers are not cached as suggested, but simply part of the provided data from the Amazon feed, though those covers are not part of the agreement that Librarything has with Amazon on the record brought over. Those come as a secondary link which may or may not change, depending on the record with Amazon over the next months/years. Data that comes across is whatever is under the ? when you search. This, however, is editable at the time of entry, so you can correct any missing or incorrect data.

Amazon is actually the worst of the source to use, except maybe for very new books. It is suggested that you use Overcat first, as this contains searches of library sources which will have more complete data. Then by other sources available of the more than 4500 that Librarything offers.

And this is still not a bug.

And yet - there is a bug in there if the isbn of a scanned book disappears.

I've noticed over the past 2 or 3 years imported Amazon data no longer includes ISBN even if you've used the ISBN to search for your book. If you go to the Amazon page for that book, ISBN is usually no longer recorded.

It just means an extra step in my workflow: search book, add suitable record, change cover to own or member cover, add ISBN manually. I've got a catalogue view with everything I need to cross-reference my Calibre record (I rarely buy print these days) with LibraryThing.

If the search results in a user taking that match, it would be reasonable to store the ISBN number for it - maybe add an uncertainty boolean flag if it's not coming from the data source but still, it was entered and ought to be retained.

The answer is to avoid using Amazon.

>6 Maddz: For eBooks? Yes, they don’t include ISBNs anymore (since some time early last year - before that it depended on the book for a year or so) even if they have them associated for search purposes. But should not happen for paper books (or at least there should be a warning).

>4 gilroy: I appreciate the suggestion to switch to Overcat as the primary method. It was somewhat better at getting the images, and it actually got the ISBN field filled in correctly.

For 0578630028 I then used the "Grab" from URL function successfully to get the proper cover image.

>9 AnnieMod: This wasn't an ebook FWIW.

If you add 9780062970701 with Amazon as the primary source, then do it again with Overcat as the primary source, you'll notice the difference.

The first search result for this is at

https://www.amazon.com/s?k=9780062970701&ref=nb_sb_noss

which leads to

https://www.amazon.com/Strange-Planet-Nathan-W-Pyle/dp/0062970704/ref=sr_1_1?key...

which is the hardcover edition that curiously doesn't have the ISBN in the Amazon listing. Thus my suggestion is that, in the absence of ISBN info from the site searched, the ISBN used to search should be retained.

>10 martymacgyver: Yep. That’s why I said that there is a problem - ebooks behave that way, paper books don’t in my experience.

However - The lack of isbn on the record explains that. You may need a RSI for that - the way LT works is to pull the data form the source. If the source does not have the isbn, it does not add it.

For a long time LoC sent back their first ISBN, no matter which one you had searched on. You get the ISBN (or not) that your source sends back. If you are checking the data anyway, which I always do, making sure you have the ISBN, the right one, 13 or 10 if you have a preference, takes only a fraction of a second. Fixing it takes longer, of course. I try to remember to copy the ISBN if I find that I am going to have to use Amazon.

>11 AnnieMod: FWIW, kindle books behave that way, meaning again Amazon as a source. I've never had problems with ebook ISBNs searched and retained from library sources.

>13 SandraArdnas: as I explained in >9 AnnieMod:. :) Amazon changed their model around eBooks.

>10 martymacgyver: It was somewhat better at getting the images, and it actually got the ISBN field filled in correctly

Those two things are really the same. As was said above LT does not save the cover image from the search but instead uses the ISBN later to pull the image from Amazon. Unless you change it to use a user provided cover.

However, if you use the Amazon cover, it can change or disappear later if Amazon changes the cover it uses for the ISBN, or stops having a cover. The Amazon cover is not saved on LT, but is a link to Amazon.