The new server came and, after some rough hours, everything is running smoothly. Ongoing server and programming tweaks should win further gains. The site is unlikely to be much faster during off-peak hours, but I’m hoping not to have hourly crashes and twenty-second waits during peak times.

I’ve done something of a major overhaul of the design. If you’re on Firefox, Netscape or Safari you shouldn’t see much of a difference. Windows Internet Explorer users will suddenly see… what you’re supposed to see. Why didn’t anyone tell me it looked so terrible! Owing to the death of my Windows machine, I hadn’t done much cross-platform testing. Ouch! Those who weren’t turned off by strangely-large fonts and boxes extending 5% off the edge of the window, I salute you!

Comments encouraged. I’m frankly a bit snowed-under right now, so replies may take a few days. Enjoy!

I’m changing how tags work to speed things up a little before the new server arrives.

Selecting tags had become a chore. The list of other people using the tag was taking 1-2 seconds to load. The new one usually takes less than .001 seconds. Tonight and tomorrow I’ll be speeding other tag-based functions up similarly.

To achieve these effects I made a decision: From now on each tag cannot exceed thirty characters in length. This is enough for most purposes, even “used modern history textbooks,” but not for “things that are almost but not quite dictionaries” (a real tag!). Most processes now already use the “short'” version; the long versions will remain visible and editable until next Sunday. Then the axe falls! Actually, since most of the long tags are really comments or reviews, I will move them to the users’ comments field.

I’ve implemented a Del.icio.us-like “change tags” feature, available under the tags tab (also improved) and in the catalog under “power options.” You can also delete tags and change one tag to many. The latter is useful for disambiguation, ie., change “un” to “u.n.” and “united nations.”

A “mass edit” feature will come tomorrow, allowing you to add tags or make other changes by ticking off books from a basic but complete list.

QUESTION:

Currently the tags are unsorted. If you enter “zebras, apples” it stays that way, and you can sort by it, making the first tag a sort of “primary tag.”

Making it always alphabetical would be easier and in some way more elegant. What is the norm? What do people want?

85,000 books. I’m haven’t checked how many uniques that is (which is also somewhat definitional), but 1% of them are by J. K. Rowling…

LibraryThing was featured on Lifehacker and Metafilter, among others. My server is melting and service is a little slow. I’ll have a much faster one up tonight or tomorrow.

My email box is also filling with suggestions. For new-comers, here are some improvements I’m working on:

• User-defined fields
• British Library searches, then other library searches
• “Power” editing, including applying tags to multiple books in one sweep
• Bar-code/UPC/EAN numbers
• RSS feeds
• iPod export (so you can check if you have something)

At present I’m whispering soothing words to the server. Once that’s finished I’ll shift my attention back to features. Keep sending suggestions, comments and criticisms. The site has come a long way since it opened less than two weeks ago, and user suggestions have been absolutely essential.

I’ve added a “power user” button into the catalog (next to “search library”). I’m going to hang a lot on it later—I want to keep the interface simple for new users, but give people with hundreds of books new options.

At present it just adds two things: the delete icon , which people have been clamoring for, an a icon, which tells you if you’ve reviewed something ( just tells you it’s reviewed). When you’re in power-user mode you can also sort by whether or not you’ve reviewed it. It sorts your reviews first, then books reviewed by others, then unreviewed books.

Now my question: Clearly I need a page listing what books a given user has reviewed and another listing recently-reviewed books, books with the most reviews, etc. But how much stress should reviews have? How interested are people in it? Should I, for example, have a “Reviews” tab like the “Tags” tab? Would this convey the impression that LibraryThing was all about reviewing and discussing? I want to keep some focus. This site is not a universal book portal. It’s a book-cataloging service with some diverting social extensions.

Opinions pro and con solicited.

PS: Next up—”Power tagging.” Vroosh!

I originally designed the “comments” field to hold information like “slight scratching; given to me by Bunny.” But people started putting reviews in, some many reviews. Others, such as dhamell wanted a place to stick URLs to a blog review.

So, in addition to “comments” there is now a field for “reviews.” These reviews can be seen on the catalog pages, but there’s a more efficient way.

Before the change books were designated either or , indicating whether other people owned it or not. These are now replaced by and if the book is reviewed by at least one person. Click the button to see both the “social information” (who has it) and the reviews. I should probably also have a special icon indicating when you’ve reviewed it (ie., a hot pink bubble?). Icons make me happy.

There’s one complication. Before I distinguished between comments and reviews LibraryThing users added over 3,000 comments. I briefly considered trying to separate them myself. Instead, in a day or so there will be a page where you can select the comments that are reviews and move them over.

Let me know what you think. I’m all ears.

PS: We’re over 60,000. From now on I’ll only announce multiples of 25,000.

We’ve hit 45,000 books—well more than 5,000/day now. This is remarkable in that “Add Books” was broken for an hour. It’s fixed now, and I’m moving on to pressing bugs and expansions.

By popular request, I have tentatively installed “autocompleting” tags on the Add Books screen (click “autocomplete tags.”) Autocompleting means that when you type part of a tag, LibraryThing will suggest the rest of it based on previous tags. Some browsers try to do this, but field-by-field not tag-by-tag. Play around with it; it’s nifty. (It uses the amazing WICK.)

If people like it, I’ll install it on the card edit page. I’ll also make it use recent tags without using “update.” But I myself find it a little distracting. Should it continue to be merely an option?

LibraryThing was written up by Andrew Brown in the Guardian‘s daily email digest “The Wrap.” The Wrap is subscription only, but apparently reaches some 40,000 people, and not a few have since stopped by.

The article is long and, as stated, not free, but I can share the opening paragraph:

“Almost every day someone comes up with a use for the internet that
makes you wonder why no one has thought of it before. Once a year,
maybe, one of these good ideas gets built. Last week, one of these
simple, brilliant ideas got built, and it’s good news for anyone who
owns the books they read.”

Brown’s weblog notes Bibliophil and Reader2, who also had similar ideas (and Bibliophil had it first). Brown’s opinion that Bibliophil has “rotted” is a bit strong—the webmaster wrote to me that he’s in the process of a major upgrade.

Brown is optimistic about the business possibilities. I think he’s wrong: the internet produces much more value than it captures. But I intend to make some money (2-4 memberships/day?) and that money will pay for the new server I’m getting this week, and some of the time. At the current rate, LibraryThing is adding 150,000 books/mo. That kind of database isn’t free.

Comments suggest that $10 isn’t a big barrier for people. I may raise this a little in the future, or switch to a yearly rate, but “free or cheap” will always be the deal, and, of course, current paid members are locked in for life.

Meanwhile, I’m still working on features. The “Add Books” page changed a little. If the LC has just one match it no longer automatically adds it to your library, but shows you the cover and lets you try Amazon. There are a few other such tweaks.

Keep the suggestions coming. I have a backlog, but expect to work continuously on this for some weeks to come.

We’ve hit 30,000 books! That’s 5,000/day for four days in a row.

New features slowed today as made technical changes to ensure LibraryThing can handle many times that number. (But you’ll notice profiles now have watchlists and a user search function.) But don’t worry, a major upgrade (hint: more libraries) is just around the corner.

Keep sending me suggestions. If I don’t respond right away, it’s probably because I’m acting on it.

Some new features.

• Formerly the icon just gave you the Library of Congress record. If you wanted to know a given book’s ISBN you had to switch display styles. This was clunky, and not everyone got it (eg., my sister complained that LibraryThing only recorded author and title!). Now you can click on the card icon to get all the fields, a larger version of the cover AND the LC card.
• The screen also gives you Amazon links. I need to include links back to Amazon if I’m going to use their API, and I think non-flashy links are a service of sorts. Yes, I get a 5% commission. I plan to take than commission and buy a cup of coffee every week.
• In addition to the icon, which adds someone else’s book to your library, books you already have are marked with a plus icon . I’ve kept the icon, which tells you others have the item (but not necessarily you).
• When you click to someone’s catalog, it now defaults to sorting by “sharedness” (thing you also have first, followed by things others have).

We’ve hit 20,000 books this evening, going from 10,00 in less than two days. Hooray.

You can now leave a comment on someone’s profile page (eg., “hey, I have lots of books on Maori art too!”). You can shut off comments on your profile page. By default I shut it off on everyone with a private library (about 20 of 680).

Do people want the ability to leave comments? This is super easy to add. I’m picturing each profile has a section at the bottom that works like a guestbook. You can leave either a public or a private comment. The recipient can delete them. The sender can delete or edit his own. A user can turn off commenting, with all private libraries defaulted to comments-off.

I’m thinking it would be nice. But I don’t want to turn this thing into “Friendster for books” as someone (wrongly) dubbed it. I may add the ability to “bookmark” other people’s libraries, but the bookmarks will not be called “friends,” with the inevitable “you’re so-and-so’s friend but they’re not yours” dynamic. Besides, I hold to the traditional view that friends need to have been drunk together.

I’ve added support for Amazon.co.uk, Amazon.ca, Amazon.fr and Amazon.de. It took a while, but it works. (Don’t mention “character sets” to me today. I might start crying.) There’s a little menu on the Add books page.

I’ve also added the ability to search “the Amazons” first. Whether this makes sense depends upon the depth of the LC’s collections in a language and what you’re looking for. I have a feeling that non-US scholarly works are well-represented, but for something like French Manga it’s pointless to try. If you’re entering a lot of non-US books, let me know which you end up using.

Next up: Sleeping.

10,000 books reached at 6:23:17 EST.

Well, I think I’ve come up with the right solution, using Amazon and the LC together. It’s a little complicated, but the complexity is hidden from the user.

It works something like this. First it looks in the LC. If it can’t find it there—either because it’s not there or because the search didn’t follow LC rules—it goes to Amazon. If it finds it on Amazon, it makes one last heroic and generally successful effort to find it at the LC, this time using Amazon’s data in a LC lookup.

People who don’t care about LC data can structure their search as loosely as they want and will still end up with LC data most of the time. People who must have LC results can make sure they get them. If you have any doubt, the Add books screen tells you where the data came from for each book.

When there’s LC data, it tends to prefer it over Amazon data. This is because Amazon plays a bit loose with authors and titles. Authors are first-last sometimes, last-first others. Titles often include the name of the series the book belongs too. The LC is more careful. At the same time, it always uses Amazon date and publication info. This ensures that, although the LC may have an older edition, your info will match the book you clicked.

Inevitably the multiple sources hamper attempts to “match up” equivalent books. Right now it tends to match books up by LC control number (which can embrace two ISBNs) or by ISBN. In the future I’ll be doing a more sophisticated sameness test, involving titles, authors and other data. The same/different issue can never be solved fully, but I’ll try to strike a reasonable balance.

Confused? Don’t be. I think it works pretty well. Feel free to differ.

No response from Amazon yet. If they insist on freezing data and requiring constant refreshes, I will have to make some changes.

I’ve added a few more requested social features:

• Profiles now show users with overlapping libraries, so you can spot who has similar tastes.
• When you look at someone else’s profile you can now see exactly what your “shared” books are.
• When you browse someone’s library you can click on to add a book to your collection.

On the Amazon front I’m waiting for word whether LibraryThing can bend their terms of service.

The problem: Amazon’s terms seem to preclude users modifying data. If Amazon says a book is by C. S. Lewis, LibraryThing users ought to be able to change that to Clive Staples Lewis, but this seems forbidden. Amazon also requires that most data be “refreshed” every 24 hours, which would wipe out any changes anyway.

Refreshing has its own problem: you can only make one request a second. In less than a week LibraryThing has acquired 8,100 books(!). That’s a lot of back and forth with the Amazon servers, and once LibraryThing hits 86,401 there won’t be enough seconds in the day to refresh everything, let alone add new books.

LibraryThing’s “competitors” (about which more in a later post) all use Amazon and seem—correct me if I’m wrong—to allow changes. So maybe there’s an exception for sites like ours. Or maybe they just turn a blind eye.

If Amazon won’t do it, I’ll take the data from somewhere else. Booksense would be my first choice. Although they don’t have an API I can find, I could screen-scrape. Better, I could not only give users the info but also hook it up to their nearest independent bookseller.

I’ve added “blog widgets.” Paste the widget into your blog and tell your visitors what you’re reading, who your favorite authors are, and what tags your using. You can even display random books from your collection (“LibraryThing Shuffle”?), or keep track of the whole community’s books.

Comments welcome. Next up: RSS feeds, Amazon integration and tag-completion.

I’ve added some new social features, and the “infrastructure” for many more.

First, all books now show either or , meaning either that others have the same book or don’t. Click the icon to find out who else has it, what they’ve tagged it, and what are the author’s other books. For example, see Rowling’s Harry Potter and the Goblet of Fire.

Second, LibraryThing now sports an oh-so-trendy “tag cloud,” a graphic representation of tag frequency. I also made an “author cloud.” I find the latter particularly interesting. What’s going on with Agatha Christie? ssquier is a big fan—a really big fan.

Next up, more social features, including library-to-library matching (“whose library is most like mine?”) and the ability to, when browsing someone else’s library, add the book to your collection.

Comments, criticisms, concerns and bugs most welcome. Thanks to all for your suggestions so far.

Here are my current priorities:

1. RSS feeds (with HTML version and a blog-widget)
2. User-defined fields
3. Add Dewey, author clicking
4. Book-level view, showing who else has it and what they’re tagging it
5. Improve some graphics

Thanks to Steven Cohen of LibraryStuff for blogging about LibraryThing (“Finally, a site with a name that is more generic than mine.”*) He suggests:

“I’d love a feed for someone’s catalog or specific tag. Let RSS technology take this tool to the next step.”

Interesting idea, and not too hard to program. It’s on my list. I’m thinking each library should have a feed. Should each library’s tags have feeds, or should the tag feeds cover all books? Hmmm..

*I originally planned to reify the “thing” graphically, either as my dog or as a Victorian monster illustration. Still pondering which…

I added a filter to import from Delicious Library, an elegant OS X application. It currently cross-checks everything with the Library of Congress, picking up LC call numbers and so forth. This takes time, and I think it’s been tripping people up. So I may cut it down to just picking up the information Delicious Library already has.

If you have problems, email me your file. I only had two files to work on—mine and some random guy who posted his music and video game collection—so I may not have caught all the format wrinkles…

I just added a CSV export feature, so you can throw the data into Excel, Access, Filemaker, etc. So, what now?

Some ideas:

• Import filters for popular book-cataloging software. I’d start with Delicious Library, an elegant OS X application.
• User-defined fields. I can add user-defined fields. For my sake, I prefer to use tags to keep track of where a book is, or who gave it to me. But others may prefer flexible user-defined fields.
• Use of Amazon. How are people finding the LC data? It works for me, but I see some problems. The LC doesn’t always have copies of paperback reprints. So, you have to choose between the original edition or entering the paperback manually. I was thinking I could make it “fail” to Amazon—check Amazon if the LC doesn’t work. Or Amazon could always be an option.

What do you think?

After a few weeks of development, LibraryThing is ready for public “beta” release.

When I worked in educational software, betas were suspect. Giving one to a customer would have probably been a firing offense. Google and Flickr have broken that taboo, and even given the term bleeding-edge chic. This is particularly amusing as “beta” means essentially the same thing as “under construction,” now as repellent thirty-somethings with Razor scooters.

What LibraryThing’s “beta” means:

• Features are still being added. The user-to-user features aren’t fully developed. It makes no sense to find your closest “library match” when there are three libraries in the system.
• I’m listening. Tell me what you like, what you don’t like, what you find confusing, etc. Post them here as comments, or email me at editor@isidore-of-seville.com.
• I’m giving away memberships. I’m giving away memberships to anyone I think will use it well, and might tell others about it. Mostly these are library blogers. Are you one? Let me know.

That’s it for now. Enjoy!