Dana’s user experience blog

Gordon gave me the idea for this post, while venting his frustrations about eBooks (someone needs to tell me whether that capitalization ought to be there — I never really know). His specific irritation was that he could not print more than four pages, thus meaning that the e-version of a real book one of his lecturers has set as required reading does not do the same job as a physical copy would (and the physical copy is on back order). What, asks Gordon, is the point of these things?

To me, it seems that eBooks are a bit like Wikipedia (only more authoritative): They’re good for getting short sharp bursts of information while you’re already online. My library’s subscription to Safari Techbooks saved me no small amount of time during the tail end of my masters; instead of having a recall war with someone over the only book our library had on the (then relatively new) DHTML, I was able to read about it, with code examples, online and just in time. EBooks are probably good for all sorts of things like that, from physics equations to Shakespeare’s 116th sonnet. If you want to read ‘King Lear’ or ‘A Brief History of Time’, though, forget it. Buy the book, if you can;t get it from your library.

The (apparent) reason why eBooks are so awful seems to be to me a triumph of copyright over common sense. Copyright concerns seem to be the reason why eBooks are neither fish nor fowl, neither electronic nor book. eBooks, at least the ones I have seen at Swinburne, are printed in a PDF-like format, making for worse on-screen reading, longer load times, and a distinct lack of the rich hyperlinking that adds value to online reference content. I can only think of 3 reasons why PDFs are being used here instead of natively online formats:

• Because you can lock a PDF and prevent someone copying the text
• Because the books are natively created in PDF-like form and the publisher sees no need to convert them
• To present the Greek symbols so often found in mathematics textbooks.

The only reason of those, that is really good enough, is the last one, and we can only hope that text-presentation technologies catch up with need soon enough that we won’t be dependent on preformatted text for too much longer (yes, theoretically unicode can handle it, but too often web browsers fail to adequately interpret unicode, resulting in either garbled nonsense, or that little square box thing). Not only do eBooks fail at being electronic, though, they fail at being books. They can’t be read without a web connection, the amount you can print is dictated by an online publisher and embedded in the technology, rather than reflecting copyright law, and all the wonderful affordances of a regular book — annotating, falling open at a frequently used page, coming back to where you left off and prolonged comfortable reading — are not available.

Despite the poor usability and poor readability of online book, though, I think it is important that we continue to have the available. The web statistics show that eBooks are quite heavily used, and a recent survey of our students has demonstrated that they like and expect to be able to access their textbooks online. Are our eBooks popular because the next generation is different? Possibly. My guess, though, is that most students are using eBooks for reference, to avoid purchasing (or carrying about) hard copy textbooks. As for me? I’ll read more eBooks when the usability of the electronic interface, and the complete unwillingness on the part of publishers to publish in an online readable format changes.

Addendum 18-2-2008: My colleague Tony has made some excellent points in the comments on this post that need raising here: eBooks have a significant advantage over traditional books in storage space and price, and are a hugely valuable resource for distance students.  Not only that, but our eBook provider (EBL) is very generous in terms of the printing users are allowed — 20%, considerably more than copyright in Australia; it seems Swinburne users have been facing some technical hitches at our end in this regard.  Tony’s most important point, though, was one I missed because I am used to libraries (and I should have caught this): It is not necessarily easy to find a book on the shelves of a library, or find the right information in that book, so eBooks may have the advantage in this regard.  All these points are reasons to continue to purchase eBooks, but also to manage expectations about what they are, so their users get the most out of them and not the least.

As a reference tool, Google Books is pretty good. You can do a normal search, and get as results any matching books that Google has indexed. With the recent burgeoning of Google deals with large and well known libraries (for example The NYPL, Oxford University Library, and Harvard Library), Google Books looks set to include the full text of a decent chunk of published works. This means it is now possible to effectively run a Google search on the content of a very large library, and have the results returned in a relevance ranked order with little snippets of text for context. It’s also possible to add the things you read to a “personal library”, assuming that you have a Google account, meaning that when you just have to find the poem you read in a book that includes the line ‘the stars carried the helpless one ribbed moon away’, you can search specifically in the books you have read.

There are a few implications of this technology, though, that are problematic. The first is that under the current law, Google is being sued for copyright infringement because they have to make a copy of the works they make searchable to create the search index. Normally I would think this was a reasonable use (even though technically it’s legal), but there is a loophole that I discovered yesterday that does make me slightly uneasy on behalf of all poets: The context that Google provides around the search terms in the results allows you to search for the next line of the poem, and for a short poem, it is relatively easy to read the whole thing. Admittedly this is a somewhat cumbersome process, and admittedly it is not likely that any poet will lose a sale out of it, but you see these snippets without direct attribution to the poet, if your search results come from an anthology, and this is a sad loss of a moral right for the poet involved.

The second problem is that this knowledge is tied up in a commercial corporation who by law has first responsibility to their shareholders, but by popular cachet is the source of information on the internet. Libraries are nervous about a monopoly on information, and while some may view this as just one more twist in the historical antipathy between libraries and Google, I think it is in line with the freedom of information principle that it should be available from more than one source, if possible.

The third issue is one that is close to my heart, and one that Sara and some of these comments got me thinking about. Google books are great if you already know what you are looking for, but if you don’t have some search terms already, it’s hopeless. More than that, though, there is no serendipity: you go, you type in some words, you find the book and either read it online, buy it, or reserve it at your local library, and you leave. You never get to see the book on the shelf next to it might also have been useful, or just walked past a display that might have had something interesting for other reasons. Now, chances are that some people wouldn’t have bothered to go find a book if they didn’t have Google books, but some of them would have. Improving serendipitous information encounters (i.e. online browsing of information sources) is something that attracts a lot of research attention (including my own, for a year), and some novel approaches. And to me it is this that is the real user experience failing of Google books — not that I don’t want to actually read online, not the copyright issues, but that their browsing experience is boring and cumbersome and smacks of an afterthought. Until Google can provide me the same rich browsing experience that an actual library or bookstore does, it will only be a reference tool.

I’ve used LibraryThing for a while now, but today I used it to set up a public catalogue of books about user experience that I like (I will keep updating this catalogue as I read more interesting books).

LibraryThing is a piece of software that I find quite compelling — one of the most compelling out of all the ones we are using for 23 Things, in fact. There are other pieces of software that do similar things, for example Shelfari*, but there is something about LibraryThing that I really like.

I don’t believe anyone I know will upload a list of every book they own (though it has a bulk upload function, if you do keep a list for insurance purposes), and I don’t believe anyone in the world will upload every book they have ever read. There are a couple of things about LibraryThing that are just bad usability, like not allowing users to rate a book when they add it in the simple add interface, and not being able to see reviews in the catalogue interface. I have no desire to talk to strangers about books, so the social functions are completely wasted on me.

So why do I like it so much? Well, part of the reason is because it is fun. The personal stats that let you know how obscure your reading taste is are interesting, and the unsuggester is pretty hilarious (and reasonably accurate). Terms like “special sauce” recommender might technically be poorly usable (because they require a specific cultural context to be understood), but they make me smile when I read them. Now, having fun can make for a very positive user experience, but it is dangerous to rely on fun (not everyone would consider it “fun” to try and push up their obscurity rating, for example), and fun has a limited lifespan — to really engage people, it pays to be fun and useful**.

If you have a good reading community, and if you have no reason (such as providing a reading list) to share a catalogue of books then LibraryThing might not be that useful. Having said that, when you have uploaded a few books (and most people will think of their favourite books first) LibraryThing gives pretty good recommendations — for example when I was creating my reading list, it suggested a couple of books I had been meaning to add and forgotten about. The Suggester is really worth a try — and the unsuggester could potentially broaden my horizons.

I don’t know that LibraryThing makes life any simpler, but it is pretty usable, somewhat fun, and it certainly doesn’t make life any worse.

*Shefari put me off because when I tried it it was slow and only worked properly in Internet Explorer. These things have changed, but I haven’t really been back. This is a classic example of how one bad experience can put a user off for life.

**Not everything can be or should be fun, but for a service designed around a recreational activity, for example recreational reading, a little bit of fun can be a good thing, if it is done right.

Update 03 October 2007: A couple of my commenters have pointed out their frustrations with LibraryThing, notably that there is a 200 book limit on a free account, and that certain profile attributes that would be nice (like hand picking favourite authors, rather than relying on catalogued books) are quite difficult. These limitations are not things that would have occurred to me, but I am a casual user who is just looking for recommendations (and a researcher who is interested in recommender systems), and my commenters are librarians. The moral of this story is that if a system has many usable feature but some features let the side down, or if the system doesn’t offer what a regular user might consider to be a reasonable feature set, or if it depends on being “fun” to create a good user experience, it is doomed to create a poor user experience for at least some, if not many of its users.