Blog

Counterpoint: The State of Search on the Kindle

(Over on the Intel Software Network blog, software ninja Clay Breshears put up a post today about why he will “Never Own an Electronic Book”. Turns out his frustration stems from how hard it is to search for something, if you don’t know exactly what you’re searching for. This is a problem not just for ebooks and readers like the Kindle, but for the web and our text-oriented world in general.

I typed up a response to Clay in a comment, which I’m reproducing here, because I think it illustrates that things aren’t as bad as they seem, and not yet as good as they should be.)

Clay, you make a very good point, but it seems to be aimed at the effectiveness of search in general, rather than just on ebook readers (though they suffer in the same way). Just like prices can only come down, I hope that natural language search can only get better from here.

That said, as a rabid Kindle fan, the situation there might not be as bad as you fear. ;-)

When you search on a Kindle, it searches across all the books you have on the device (it indexes them during idle time while you’re reading, and new books show up in the index quickly). And here’s what the search results page looks like:


Image credit Robert Mohns via Flickr

It shows you how many results were found in each book, and by selecting that book, you can quickly skip back and forth between the hits.

Notice that there are also options to find results from the web, Wikipedia, the Kindle Store, and the built-in dictionary on the Kindle, if what you’re looking for isn’t in one of your books.

Search DOES need to get better, but I’m pretty happy with how well it works on the Kindle today. Come by next time you’re in the neighborhood, and I’ll give you a demo! :-)

Standard

2 thoughts on “Counterpoint: The State of Search on the Kindle

  1. Josh, I’m interested in ebook textbooks (having a couple of kids in college will do that to you…). Being able to highlight, etc., is very important, and then being able to search within highlighted text only would be a cool feature.

    Textbooks typically have indices which, if done well, can be more useful than a search function. Do you have any experience with an ebook having an index? Do they dynamically rebuild the index as the book gets repaginated? Or am I just expecting too much?

    Looks like I’m going to have to take you up on your offer for (another) Kindle demonstration.

  2. Brent, come by. I’ll show you. :-)

    Short answer is, the search index is continually updated invisibly in the background. The only time you see any evidence at all of that activity is if you do a search right after adding a brand new book to the device, and don’t get any search results from that book.

    As far as being able to highlight/clip sections, and search within them, yes, you can do that on a Kindle. See the “My Clippings” in the search results screenshot above.

    As far as traditional indices or tables of contents, that depends on the book publisher. Some books I’ve seen have them, some don’t.

    Pagination is handled in an interesting way. Books on the Kindle are “pre-paginated”, with “locations” instead of page numbers. These are approximate – you might be reading, and see that you’re on “location” 120-135. This is so that the entire book doesn’t have to be repaginated every time you change the text size, like on other ebook readers (eReader in particular has to do this, and it’s a pain to wait while that repagination happens). And you can jump directly to a specific location number. So that’s a cool little innovation.

    Like I said, come by, and I’ll show you and hopefully answer your questions. :-)

Comments are closed.