Weening myself off amazon data

Due to the increasing issues that it is causing LT, I have been weening myself off of Amazon and its data. I no longer use Amazon to source my imports, and am (slowly) working on replacing my covers with non-Amazon covers.

My question: Is there anything else I need to do to remove my piece of the Amazon burden on LT? What about my older entries that were originally sourced from Amazon....should I remove and re-catalog from another source, or is their no continued impact to LT if I let these entries remain as-is?

Since this thread

http://www.librarything.com/topic/78213

was posted, I've worked more aggressively to wean myself from using amazon's covers. I'm not sure about the originally sourced data ... I've edited much of my data to an extent that I'm reluctant to re-catalog it. Perhaps an export - import would do the trick?

Not really - the import will search several sources and you may well end up with Amazon all over again. I don't believe having something listed as sourced from Amazon is any direct burden on LT. Aside from the covers, it's only a field in a database, as far as I know.

This is a long-running concern for us. It's why we're working on back-up options for every book in our records, and ways of moving to them, as necessary. The "as necessary" may mean either if Amazon kicks us out—unlikely—or if we want to do things with the data that their terms don't allow, like putting us on a mobile app.

4>

So, if I added a book ages ago from Amazon (before I knew any better) and have since extensively changed many fields (including the cover), that still counts as "Amazon data" that could no longer be used if you part ways with Amazon? To be honest the only reason I haven't been deleting and re-adding those is that I don't want to mess up my date-added statistics.

I am interested in this issue too, because I have a lot of books sourced from Amazon (and then edited by me): most of the books I entered initially, before I knew about the Amazon problems and other options, as well as a lot of paperbacks that I can't find from any other source, particularly library sources. I would go through and gradually change from Amazon sources, despite the work involved for re-editing, but I just don't know where to look, particularly for paperbacks. Some are on LOC, but very few. Any recommendations?

So Tim, does this mean that the source is a problem?

Just to clarify: Do the Amazon TOS make a distinction between (a) data that has been pulled from Amazon at one point sometime in the past and is now stored in LT's database and (b) data that is actively being pulled from Amazon every time someone accesses a page with that data?

I can understand the TOS from the latter (what with LT hitting up Amazon's image servers like whoa), but not from the former, as all that data is either factual, and thus, not subject to copyright protections, or copyright someone other than Amazon (book covers, etc.). It seems strange that Amazon could regulate your usage of public, factual data, just because you found it at Amazon first. Under that reasoning, the Houston Chronicle could limit my discussions of new stories, just 'cause I read about them there first. The difference is the existence of a provable "paper trail", but the idea is the same...

I understand, also, that TOSes in general don't have much to do with copyright law, and that you're using Amazon as a service from which you pull data, and blah, blah, blah, but I wonder what the difference is, to Amazon, between, say, LT automatically pulling and storing data, and a user opening Amazon and LT in separate tabs and copying/pasting all that data by hand.

I'd also like to get rid of Amazon-sourced books but am reluctant to do that for the same reason as lorax ("I don't want to mess up my date-added statistics").

Tim, please note that this question comes from a person who knows zilch about how LT is coded etc. so I have no clue if this is absolutely outside the realms of technical possibility:
would it be possible to create a feature which would enable "source-change"? Sort of making the source just one more editable (is that a word?) field in our catalogues? What I'm thinking is that I'd first use Add Books to check where I could find title X. Ok, found it in non-Amazon source. I'd then go to my collections, find the title X (amazon-sourced) and "switch" the source to the one I found.

I'm guessing LT would get rid of a huge amount of Amazon-sourced books if that were possible...
But would it? And even if it were, would it be such an enourmous amount of code-work that it wouldn't be worth the trouble?

I have a hard time believing the 'source' is going to cause problems, just the serving of covers from amazon.

>10 readafew:

That's why I asked.:)

10>

That's what I had always assumed, but Tim's #3 ("back-up options for every book in our records") sure makes it seem otherwise.

Well then it would be nice when he has the time, to clarify this for everyone, though I still lean toward being right in #10.

"Replace source" would be great. Or, editable "date added". I have been getting rid of amazon data but I hate mucking with my "date added" stats too.

Frankly, this whole topic gets into things I don't want to talk about. I can't say what's been told me. What's been told me is not nearly enough. I don't want to reveal too much about what we're going to do. I don't want to say we consider X or Y to be okay or not, as that might come back to haunt us.

I can say:

1. We have non-Amazon alternate records—generally multiple alternate records—for more than 90% of books already on disk and mapped to ours, with no extra "looking." We will soon be at 99%. We plan to allow source switching.
2. We do not expect to be forced to remove Amazon records, particularly if we don't force it ourselves, to gain some advantage their data denies us.
3. Amazon doesn't want to kill us. If they did, and if we couldn't make good on library data alone, I suspect that having Bowker, the second-largest source of commercial data bibliographic data (after Amazon) as a minority investor, would come in handy.
4. I can absolutely guarantee you that LibraryThing is in a better situation here than anyone else—by far.

Tim

Thanks Tim for commenting on this :)

I have no problems with Amazon-data per se -- my only worry is that IF something were to go terribly wrong with that particular source, it might damage LT permanently. I've had plans to be able to track my readings and library in LT for decades to come and I'd really hate that plan to go awry ;)

But it seems that everything is in order so I might just as well stop worrying :)

Thank you, Tim.

> 15 We plan to allow source switching.

... yaay! any time soon? in other words, ought I hold off on my own source switching, that upsets my date-added stats?

>19 timspalding:

I'll get back to you soon about the status of it.

T

I don't really care about the date added statistics, but I do have some 300 odd Amazon sources in my library. Currently there's no way to determine which are Amazon derived sources except by looking at each book's record.

With over 2400 books (and still adding), that's quite a bit of looking, even knowing that the vast majority of the Amazon data probably dates from my first days here.

I look forward to hearing that source switching is 'good to go'!

>20 MsMixte:

You can show the "source" feed, and sort by it. That'll show you.

Sort by source and then by date of entry. I'll try that next time, and should be able to really clean up a storm. But now I'd better go make dinner. My husband already came to announce that he has cleaned up the kitchen for me (hint, hint).

> 22: Very subtle ;-)

> 21: Thanks, I hadn't realized I could sort by that too.

Every time I read one of these Amazon threads I feel guilty and think I should be doing more to avoid the dreaded Amazon data. But then when I go to add books, I often search several sources with no success until I finally resort back to Amazon.co.uk, which almost invariably has my book, in my edition - or at least with my cover. So then I just think, why should I waste my time? Is the Amazon data really so damaging to LT?

Am I the only one that has a problem finding useful non-Amazon sources? Or are there useful sources out there that I haven’t found? I am in New Zealand, and we tend to get UK or Australian editions of most books, generally trade paperbacks instead of hard covers for new releases, then regular paperbacks.

I understand about the issues with Amazon covers, and I know that Amazon data can contain errors, but assuming I change my covers and correct any errors, is there any point in going out of my way to look for non-Amazon sources?

I second #24. I'm changing my covers but I do still add books from Amazon if I can't find the data easily elsewhere. Should I stop? If I have changed some of that data, is is still a problem?

24 & 25>
given the proposed source-switching - depending on how it works, it may be that the easiest thing will be to search on amazon, let it connect to a work, then source-switch it to a library afterwards.

I prefer entering manually to fixing someone else's messes. The things I have the hardest time finding are in German or French, and Amazon.fr and Amazon.de have certain fields that always have to be fixed, as well as the normal amazon problems.

Also as a combiner I have seen a lot of really bad author information coming in from Amazon. If you fix the authors, and anything else you care about, and are happy doing that, I don't see a big problem.

For what it's worth, I use British Library and Talis Union Catalog most, but they aren't great for new paperbacks. But then, although I buy a lot of British paperbacks, they are rarely new. The entering after being in the UK last summer was difficult, and might have been easier if I had used Amazon.

> 21, 22

Thanks! I didn't realise that I could sort by source. I will start weeding out the Amazon info now.