Why does "Add to your library" (on a book page) not do what it says?

Talk about LibraryThing

Join LibraryThing to post.

Why does "Add to your library" (on a book page) not do what it says?

This topic is currently marked as "dormant"—the last message is more than 90 days old. You can revive it by posting a reply.

1fjordaan
Apr 7, 2009, 4:14pm

For example, on this page:

http://www.librarything.com/work/151100

I have this book, and I want to add it to my library. When I click the "Add to your library" link, it instead launches a search on the title -- *and then doesn't find it!* (Using the default Amazon.com search.) Sometimes it finds it, but different versions from the one I was looking at, with inferior metadata.

How do I add that book to my library?

francois

2kevmalone
Apr 7, 2009, 4:24pm

There are a number of threads about this.
Here's a recent one
http://www.librarything.com/topic/59180

Welcome to LT

3lorax
Apr 7, 2009, 4:35pm

The first step is to change your default to Library of Congress.

The second is to actually add books via "Add Books".

4fjordaan
Apr 7, 2009, 6:41pm

Thank you very much -- that answers my question.

Despite this measure, LT seems to be riddled with bad data (i.e. duplicate & inconsistent). Is there development underway to gradually improve this? (As Last.fm have been doing quite successfully with a corpus of even dirtier data.)

Also curious as to why I should change the default to LoC -- it hardly ever finds anything. (Perhaps because my library consists mainly of comics.)

Sorry for what are probably naive questions; I'm a relatively new and irregular user here.

5staffordcastle
Apr 7, 2009, 6:45pm

The recommendation to use LoC is because their data is much cleaner than Amazon's - but your point about your libarary being mostly comics is very cogent - LoC isn't going to do it for you.

As to improving the data, that is pretty much a grass-roots effort on the part of energetic members, with the assistance of the management providing tools, such as Canonical Title, which can override an erroneous title from a sloppy source like Amazon.

Welcome aboard!

6lorax
Apr 7, 2009, 6:53pm

4>

LT's data is only as good as the source data; that's why people who have been here a while, and who care about data quality, suggest the use of library sources, which are generally much better than Amazon. (If your books aren't in LoC, of course, that's another matter.)

7kevmalone
Edited: Apr 7, 2009, 6:55pm

> 4 Data-entry in LT is user driven. Sure it's possible to grab data from other sources (and many do) but it's not mandatory and there's no guarantee that those sources are correct. For example I've seen on average at least one correction per day sent from here to Amazon.

"Everyone's a librarian" is one of the site philosophies.

Sooo the "development underway to gradually improve this" is what we all (LT users) do. You, me, lorax, and the 661,443 other site users.
Most people don't care to: many do, and they're darned good at it.

There are places where you can post problems you are having with the site
here
http://www.librarything.com/groups/bugcollectors
http://www.librarything.com/groups/combiners
and people will fix stuff for you, or tell you how you can fix it yourself.

Finally - welcome to LT Francois - always good to see another Chris Ware fan here.

ETA: lorax and staffordcastle beat me to the punch and said it more succinctly!

8Esta1923
Apr 7, 2009, 7:00pm

Perhaps because most of my books are old ones I always add manually. That way I can put in as much info as I have and it will apply to my copy (not a later version, paperback or otherwise).

9staffordcastle
Apr 7, 2009, 7:07pm

Something none of us has said yet is actually the answer to the original question - at the beginning, that green plus sign really did what you expect it to: add *that book* to your library. The functionality was changed in order to prevent the propagation of bad data on the site.

10jjmcgaffey
Apr 7, 2009, 7:09pm

Tim said that he specifically decided NOT to bring out the new, improved green plus along with collections, because it's not ready and it would delay collections (more). However, that does mean that he's been working on it and has made some progress. So there's hope for the green plus yet!

but until then, yeah, choose your best source (best data that will have your books) and have fun on the Add Books page.

11lorax
Apr 7, 2009, 7:34pm

9, 10>

That's a good point.

To get back to the original question, part of the reason why the green plus was eliminated is the issue of books vs. works.

As far as LT is concerned, there are two levels for each book -- the entry in your library, known as the "book" -- and the combined set of all entries for that work, known as the "work". So I have two copies of Ringworld in my library -- two books, with different publication information, which I can give different ratings, different reviews, and have show up as two lines in my catalog -- but they're the same "work", so it shows up as a duplicate, the ratings and reviews show up on the same page for anyone doing a search, etc.

It used to be that the green plus would just clone a "book" to your library -- add all the exact details (not just author and title but ISBN, publication date, etc.) The trouble is that there are people here who are very sloppy with data -- they'd add Harry Potter and the Philosopher's Stone, then just add it six more times and change the title to show that they had the whole set. Someone would then come along and say "Oh, Harry Potter and the Half-Blood Prince, I have that" and add it with the green plus -- but the ISBN and so forth would still be that for Book 1, not Book 6, and it led to bad data proliferating like mad. So Tim turned off that functionality -- not a popular move, but I for one agree with it.

There are two common use-cases for the green plus, which eventually may be re-introduced:

1. Someone who doesn't care about data quality or who is adding to a wishlist and just wants a "generic" book of author and title. They don't want or care about the rest of the information.

2. Someone who shares a rare book that had to be manually entered, and doesn't want to duplicate all the effort.

99% of the time, though, people are just trying to do something that could more easily be done through Add Books.

12jjwilson61
Apr 7, 2009, 11:38pm

The question was asked why LT doesn't have a project to improve its data, as some other site does. Some people have suggested that such work is underway, ultimately there is nothing anyone can do about bad data in someone else's catalog...by design. The philosophy on this site is the data in each person's library belongs to that person, which is contrary to the idea of completing one set of perfect data.

13fjordaan
Apr 8, 2009, 3:32am

Thanks, everyone, for the patient and detailed answers. I've learned a lot in this one thread. I'd love to spend more time on LT, but I currently just steal an hour now and then to add a dozen books or so. At this rate it'll be years before I finish entering my library.

@jjwilson: You describe a fundamental difference from Last.fm (my example). In their case they aim to maintain *only* canonical data, and will change your submissions if they have sufficient confidence in recognising it. I can see how this probably won't work in the world of books.

I've also come to realise that most of the terrible inconsistency I see in works' metadata is Amazon's fault. I don't have the time to enter works manually, unfortunately. What is the best way to leverage the effort of careful users who have the same books I do? Clearly I can't find their copies and duplicate it to my library, for reasons outlined above. How do I even know which users have the best versions? (I've been looking for those versions which are shared by the largest number of users.)

Chris Ware is an excellent case in point: it is extremely difficult to decide on a book's canonical title. Can you recommend forums or threads where people discuss / recommend the best way of describing works? Particularly comics, which could have series names and issue numbers.

I appreciate all the help.

14MarthaJeanne
Apr 8, 2009, 4:19am

I suggest looking through the libraries that seem to have the best data and see what sources they use. There probably are library sources with comic collections. Someone has probably found them.

15fjordaan
Apr 8, 2009, 8:44am

Where does it say what source they used?

16Alixtii
Apr 8, 2009, 9:15am

>14 MarthaJeanne:

I suspect most of the best data, especially for works like comics which might be less likely to appear in libraries, comes either from manual entry or from an entry from a source (probably Amazon) which is then re-edited to fix bad data (that's what I do for my comics). So there it is a case of user data being superior to the sources.

17jjwilson61
Apr 8, 2009, 9:49am

You can change the columns which display in your library by clicking on the letters, A through E, that appear at the top of your library page. You can also edit which columns appear in these views by clicking on the pencil icon next to the letters.

18readafew
Apr 8, 2009, 12:09pm

for comics? I looked them up in comicbooksdb and manually added almost all 1100. Generally it was easier and less frustrating to just do it manually, and if you are working on a series it is a little easier. Graphic novels or the collections are generally available but single issues are fairly uncommon.

19kevmalone
Apr 8, 2009, 12:49pm

18> Single issues are killingly difficult - I gave up. AFAIK there is not a unique ISBN for each issue so I can see that automatic combination is going to be problematic.

20leahbird
Apr 9, 2009, 4:03am

>fjordaan

here is a thread (currently unactive) where we were discussing the ways to use librarything to catalogue comics. it has some good ideas to at least get you started. welcome to our little patch of heaven!

http://www.librarything.com/talktopic.php?topic=6994

21leahbird
Apr 9, 2009, 4:07am

it also strikes me as strange that searching through LoC wouldn't locate your comics (if they were fairly mainstream) since it's home to one of the largest comic book collections in the country... last i read, they housed over 6,000 issues.

22readafew
Apr 9, 2009, 10:10am

21 > did you mean 60,000? cause 6000 really isn't that many. I have 1100 and half of that are X-Men titles. Superman alone probably has 3000 issues, maybe more.

23leahbird
Edited: Apr 9, 2009, 4:38pm

> readafew

from the LoC website: For more than sixty years, the Library of Congress has acquired comic books through copyright deposit. The current collection of 6,000 titles, probably the largest in the United States, contains approximately one hundred thousand pieces and is growing by about two hundred issues each month.

i just grabbed the wrong number. 6,000 titles, but 100,000 pieces (i'm assuming that pieces means individual issues).

24staffordcastle
Apr 9, 2009, 4:41pm

I'm betting LoC is cataloging comics as serials, not monographs, which is why an LT search doesn't catch them - the system is built for books, not serials.

25fjordaan
Edited: Apr 9, 2009, 4:51pm

My comics tend to be from the indie/eclectic end of the spectrum, and include many in foreign languages, factors which add problems of their own. Few long serials, many one-offs, and often no ISBNs. But thanks for all the tips - it's great to see other people who are really passionate about cataloging.

Sounds like a real flaw in LoC's system -- surely it's very relevant to their mission to allow services like these to use their data?

26Helcura
Apr 9, 2009, 5:56pm

I've found that if you've got any paperback collections of comics you can often find them in the public libraries. Not so much the single issues, but sometimes I've been surprised.

27kevmalone
Apr 9, 2009, 5:59pm

Echoing #26, I've been successful with bound collections of issues too.

28saltmanz
Apr 9, 2009, 6:08pm

Yeah, most trade paperbacks have an ISBN, so they're easy enough to add via Amazon, even.

29Noisy
Apr 9, 2009, 6:15pm

Alternative library sources for UK-based Thingamabrarians are the Talis Union catalogue and the British Library. I'm not saying these are necessarily better.

I remember coming across a massive comic book library collection somewhere, but I've not added it to my watchlist. One thing to try is to use search and put in 'comic' as the search term for members. moccany looks interesting, but it isn't the one I remember.

30readafew
Apr 10, 2009, 10:00am

23 > yes 100,000 issues is a much more believable number for 'largest collection'.

31Alixtii
Apr 10, 2009, 2:51pm

>28 saltmanz:

The fact that "they're easy enough to add via Amazon" is sort of the point. The question is how easy they are to add not using Amazon.

>30 readafew:

Does anyone know what the total number of issues that, say, Marvel has published? DC? I don't know if a tenth of a million issues is a huge collection, or a drop in the bucket.

Of course, the fact that we can't effectively access those records through LT renders the question moot, but I'm interested.

32saltmanz
Apr 10, 2009, 2:58pm

I merely meant that TPBs having an ISBN should make them easier to add in general.

33readafew
Apr 13, 2009, 11:31am

31 > well http://www.comicbookdb.com has almost 22,000 titles listed and it is a user contributed db like library thing. I have no clue how many different issues are listed.

34Alixtii
Apr 13, 2009, 11:35am

32>

The claim is, I think, that for any given comic TPB with an ISBN, the ease of adding them from a non-Amazon source is less than the ease of adding a similarly popular book with an ISBN which is not a comic. I don't have the data to know if this is true or not, but it sounds plausible and fits my experience.