Discovery Tools

Kindle for the iPhone

Yesterday, when I posted my review of the Kindle 2, I had no idea that Amazon was about to release the Kindle for iPhone application. But this morning I saw an article about it on my iPhone, while I was on the train to work. I downloaded and installed it before I got to the next station! When I did, I found my Kindle books listed in the archive, transferred the one I’ve been reading to my iPhone, and was whisked straight to the last page I read on my Kindle. Pretty cool.

Even better, the reading experience is better than any of my other ebook applications. The page is totally uncluttered with software interface controls. There are controls; to get to them, you tap the screen. Then they appear. To page forward or back, you swipe left or right.

Audio Rights and the Kindle

One other Kindle update, while I’m at it. One of the new features of the Kindle 2 will read your book out loud in a computer-generated voice. The quality is remarkably good. Words are occasionally strangely pronounced or given an odd accent or emphasis, but the text is quite understandable. The Authors Guild has complained that this functionality violates an author’s copyright–specifically the right to sell the audio rights to his or her book. I happen to disagree, both because the quality of the reproduction–good as it is–is not at equivalent to a recording of a human being reading the book, and because no new copy of the work is made or retained. You can make up your own mind. Roy Blount Jr., president of the Authors Guild, makes his argument in the New York Times. The Electronic Frontier Foundation argues, as I do, that the Kindle’s reading function does not violate copyright.

Amazon, in reaction to the Authors Guild has apparently agreed to allow publishers to decide whether the device will be able to read their books. But they continue to argue that the Kindle’s voice synthesis does not violate copyright.

I just received my Kindle 2 in the mail last Friday and spent a portion of the weekend experimenting with the device. A brief review follows.

Kindle 2 Design and User Interface

In some aspects, the new Kindle is an improvement over the old one. It is thinner, fits the hand better, and does not encourage unintended page turns like the old one did. However, Amazon has introduced two new problems that were not issues with the first version of the Kindle:

• The battery is not user-replaceable (although there are already posts on the web explaining how to do it, anyway.)
• Amazon removed the SD card slot, so there is no way to increase or swap out the storage.

Amazon says the lack of a user-replaceable battery is a result of the new, slimmer form factor. Fair enough. The second issue, however, appears to be a concession to publishers who do not want you to be able to share content with your friends, even if the restriction prevents you from sharing content that is copyright free, or of your own creation. It may also be an attempt to further tether users to Amazon’s online services–such as its online store of content you have purchased.  (Personally, I think publishers will eventually figure out new business models that do not require digital rights management (DRM) schemes and crippled hardware–or they will go out of business.) Whatever the motivation, the lack of expandable storage does not serve users well.

Two other minor issues: my Kindle already looks grungy, because the white plastic surrounding the screen picks up dirt and fingerprints easily, and it does not come clean with water. If, like me, you are inclined to do all the things with your Kindle that you do with your books (eat, sleep, throw them in a bag with other stuff of questionable cleanliness), the device is going to loose its tidy look fast. In addition, screen updates are slow and involve flashing that can be annoying at times.

Kindle 2 Software

The Kindle 2 software is, well, odd. It uses the physical hardware (the little, square button which Amazon calls the five-way controller and the delete button) in ways that will be unfamiliar to most users. Further, it is annoying and slow to use the controller move to the links in a page, or to select text for an clipping. A better selection method is needed. (I’m looking forward to a touch screen Kindle, but the current E-Ink technology cannot support touch input without a loss of sharpness. This problem is likely to be solved in the near future.) This problem is more likely to be an issue for most users when reading blogs and newspapers, which tend to be sprinkled with links, than it is likely to be an issue when reading books.

This brings me to another issue: the web browsing experience is poor. JavaScript, flash, and other plug-ins are not supported, breaking many pages. The screen flashing is annoying, screen updates are slow, and, of course, there is no color.

One feature I like very much is the ability to change the font size of any content. When my eyes are tired, I particularly appreciate this feature, and it is an advantage not offered by any books I own.

Kindle 2 as a Service

The Whispernet service that allows you to find, purchase, and download content directly to your Kindle works well. To my mind, this service is the primary reason to purchase a Kindle–especially if you travel and do not like to be caught without a book. This service will only get better if Amazon follows through on its plans to make your purchased content available on other devices, such as the iPhone. The idea is that you will be able to read a book on the Kindle, put it down, pick up your phone, and continue to read the same book at the place where you left off. That would be a high value service for me.

Data Formats and DRM

You can put new, personal content on the Kindle by e-mailing a file of a supported format (HTML, Word, PDF, GIF, or JPEG) to the Kindle at its individual e-mail address. (What exactly, besides the need to guess my e-mail address, prevents my Kindle from being spammed with unwanted files, I wonder?) Personal files, however, are not backed up by Amazon, so back them up before you put them on the Kindle. Text files are converted to Amazon’s proprietary format, sometimes incorrectly. More work is needed on that process. A large but simple word file I tried didn’t convert cleanly. I would rather that my content were not converted at all, but supported in its original format. Worse, some of the content I’ve downloaded from Amazon has suffered from the same conversion problem I saw with my own file: indented paragraphs result in lines that extend beyond the right edge of the screen and cannot be read. This is unacceptable in purchased content.

Of course, all the copyrighted material downloaded from Amazon is in a proprietary format with DRM. Thus, it is not usable on any device not supported by Amazon. Interestingly, you can download a copies of your purchased files to your PC, from Amazon’s “Manage Your Kindle” web page. But you cannot choose a non-proprietary format. I’m guessing this means there will eventually be an Amazon reader for your PC. But it could be intended just for back-up.

Tim O’Reilly, of O’Reilly Media, has pointed out another problem with the Kindle’s proprietary format. It does not support the complex formatting required by many technical books, including tables and monospace fonts for code examples. As a result, many of O’Reilly’s books are not available on the Kindle. All O’Reilly books are available in the open, ePub format on the iPhone and other devices. Presumably, this formatting restriction applies to other technical publishers as well, whether or not they are inclined to support an open format.

Conclusions

The Kindle is a very useful device for reading non-technical content such as novels, particularly when traveling in the United States. The screen is sharp and the reading experience is comfortable. It is also a means of acquiring the latest bestseller at a reduced cost. It is not a good device for reading technical material, because of the restrictions imposed by its proprietary format. Nor does it support blogs, newspapers, or other highly linked content well, because of the limitatations of its browser and its awkward page navigation.

The second product I saw at ALA Midwinter 2009 that struck me as significant was Ex Libris Rosetta. I hesitate to try to describe it because it really falls outside my primary area of expertise; Rosetta is a digital preservation system. According to Ex Libris it

supports the acquisition, validation, ingest, storage, management, preservation and dissemination of different types of digital objects while enforcing the relevant policies that can vary from one institution to another.

What strikes me as different about Rosetta as compared to the digital responsitories or digital asset management systems that I’ve seen before is the product’s emphasis on roles and workflows. It seems to me that much of the work of setting up a digital preservation program in any library or cultural institution centers around extablishing appropriate policies, procedures, and workflows. This software could really help in that work. In addition, it was designed to a very specific and detailed specification out of New Zealand and a set of related standards: the Open Archival Information System (OAIS) model and the Trustworthy Repositories Audit and Certification (TRAC) checklist. Like most modern software applications, the product comes with a set of APIs so that organizations may integrate it with their existing systems. If you are interested in digital preservation, you should check out this software.

At ALA Midwinter I saw two products that seemed particularly forward thinking and important for libraries. The first of these was Serials Solutions Summon, a new unified discovery tool. This service will make it easy for libraries to provide a single unified search interface for their catalog, local digital assets, and licensed database content. A couple of features make this service important: first, the service does not use a federated search architecture. Serials Solutions has acquired access to an impressive meta-data collection from content providers and publishers. This should allow them to index that meta-data using more consistent practices and to provide a more consistent and effective searching experience for users than is possible with federated searching. Another important feature is the API (Application Programming Interface), which will allow libraries that want to create their own next-generation discovery interface to incorporate search results from Summon. This means that libraries that are using open source discovery tools like VuFind, Blacklight, or SOPAC, will have the means to integrate Summon with their local system. (There is no surprise here. Andrew Nagy, the original developer of VuFind, has been hired by Serials Solutions to work on Summon.) Imagine for a moment, your own discovery tool that provides a seamless search of your own local data, results from Summon, and results from the WorldCat API. As a sometime developer, the idea has got my imagination roiling. Of course, for libraries that prefer it, Serials Solutions will provide a search interface as well.

I recently purchased an Asus Eee PC 1000. The netbook is 10.5 × 7.5 × 1.5 inches, which gives it a footprint smaller than a piece of US letter paper. It comes with a 10 inch screen, 1.6 GHz processor, 1 GB of RAM, 40 GB of solid state storage, and a 6 cell Li-Ion battery. It even has a 1.3 megapixel webcam. Being a netbook, it has strong networking capabilities (wired, wireless, and bluetooth). When I bought the PC, my idea was that it would be small enough to carry just about anywhere and yet powerful enough that I could do just about any of my regular computing tasks on it. My previous experience was that all my other laptops had been either too big or too heavy to carry around (partly, I’ll admit, due to the limitations of a bad back). I was particularly interested in being able to write anywhere–on the plane, in the coffee shop, or while on a library site visit. I also wanted enough storage space to be able to bring my current document with me, as well as reference files and a solid MP3 music collection. The Eee PC 1000 particularly interested me because of the relatively large solid state storage. I thought this would likely reduce power consumption, heat generation, and make for a very reliable system. There would be no hard drive heads to bounce and no spinning platters.

So how has it worked out? Very well. I wasn’t entirely satisfied with the distribution of Linux that came on the Eee PC. It was based on Xandros Linux distribution, but with a really simplified menuing system which hides most of the OS from view. I don’t need or want anything hidden. Fortunately, there is an Ubuntu-based Linux distribution which works very well: Ubuntu Eee 8.04.1. (It has now been renamed Easy Peasy for its lastest release, which is based on Ubuntu 8.10.)  Once I had that installed, it was easy to fall in love. The Ubuntu distribution provides all the applications and tools I need. The screen is large enough. The machine is quieter than any of my other laptops, although there is a cooling fan. I’ve never worked in one place long enough without a power outlet to run out of juice. The keyboard does have one fault. The shift key is small and not quite where my left pinky expects to find it. It has taken some practice, but I am getting used to that. (It helps that I’ve written about 40,000 words on the Eee PC in the last month.) If you’re looking for a netbook, I recommend this one. The cost? Under $400 from some vendors.

Google’s new mobile application for the iPhone is now available for download from the Apple App Store. This application is both impressive in its own right, and significant as a harbinger of what is to come. I was able to load it on my iPhone at about 8:30 AM EDT, about four days after Google led us to believe it would be available. Apparently, even Google has to wait for Apple’s ridiculous application approval process for iPhone applications.

The big news, of course, is voice search. I can now touch the “g” icon, lift the phone to my ear, wait for a beep, and say some search terms. The system beeps again to indicate that it is done listening and starting to search. In the search box, I can see the words for which the system thought I wanted to search. “Library information technology” worked perfectly in my limited testing, but the system mistook “flu” for “blue”. There is a little down arrow next to the search terms that can be used to pull up a list of alternate interpretations. In my experience so far, either the system is right, or it is wrong, and the alternates are also wrong. (Your experience may vary.) I’ve also found that more words seems to help the system. “States that have the flu” worked fine. The longer and more unique term, “Influenza”, also worked fine. The GPS is also used to provide location-specific information when available. “Movie showtimes” brought up movies in my zip code, without my having to specify my location.

Wow! This technology, when combined with the already impressive iPhone, is as close to magic as I’ve ever experienced.

Google is also planning to release the application for the T-Mobile G1 sometime soon.

Today, I spoke at the PaLA Annual Conference in Valley Forge, Pennsylvania. My goal was to suggest a step by step process for learning about open source. I wanted to suggest specific set of projects that were appropriate to a public library that would bring the staff involved to a point were they were comfortable utilizing open source applications. Some of the applications that I mentioned were Firefox, OpenOffice.org, Audacity, WordPress, and Drupal. The slides are here: Open Source 201.

Yesterday, I spoke at the NFAIS Humanities Round Table in New York. I was asked to talk about PALINET’s podcasting program, for which I interview innovators in the library field. I talked about fulfilling our organizational goal of providing technology leadership to our members: podcasts allow us to target working professionals who may find it difficult to read professional materials during the work day, but who can listen on the way to work, or while they take a walk at lunch. You can get the slides here: Podcasting at PALINET.

Today, I gave a presentation at the SOLINET Web 2.0 Conference in New Orleans, LA. I talked about the new discovery tools that many libraries are installing on top of their integrated library systems. These new tools include such features as relevance ranked search results, search facets, spell checking, and enchanced content. Most can import content from Amazon, Google, LibraryThing, OCLC, Baker & Taylor, or Syndetics Solutions. They will display book jackets and reviews, allow users to tag records, enter comments, or make lists. Proprietary solutions include Aquabrowser, Ex Libris’ Primo, Innovative Interfaces’ Encore, and OCLC’s WorldCat Local. Open source options include University of Virginia’s Blacklight, Oregon State’s LibraryFind, and Villanova University’s VuFind. Here’s a link to the slides.

In the last month, I’ve been consulting on open source software and Koha. In preparation for that work, I put together a Koha 3.0 demonstration system that I could run on my laptop. The key components were a Lenovo T61, a free VMware server, Debian Etch Linux, and Koha 3.0.

The VMware server is a straightforward install that you can download from VMware website. I installed the version for Windows, since Windows XP is the base operating system on my laptop. The Debian Etch release is available in the form of a netinst ISO image. This is a small CD image that contains a minimal Linux distribution and an installer that will download the rest of the software over the network. This works well if you have a high bandwidth connection. I selected the 32-bit i386 image because I know from experience that the 64-bit versions tend to be harder to manage; Many applications don’t install or compile easily on 64-bit machines.

Once I had downloaded the ISO image, the next step was to run the VMware Server Console and create a new virtual machine. The wizard for creating a new virtual machine lists a number of specific Linux versions but doesn’t list Debian 4.0 (Etch). That’s OK, you can just chose “Other Linux 2.6.x kernel” and proceed. VMware allows you to configure the virtual machine’s CD drive to use an ISO image from your Windows filesystem. This is cool because you don’t even have to have a CD drive or burn a CD to do the install. I created a 16 GB hard drive to use with the VM. (This was a mistake and I had to add another 16 GB drive later and move one of my partitions to it, which was a pain. So my advice is to create a bigger virtual drive than you think you need. They can’t be increased in size later.) I also selected the NAT option for networking. This option allowed me to run the Koha system successfully, whether or not my laptop was connected to the Internet. If you choose the default bridged networking, Koha will only run properly if you are connected to the net.

Now, I turned on the virtual machine, which booted from the ISO image, and ran the installer. It can take a little while to download all the files, depending on what packages you use. Once Linux was running in the virtual machine, it was time to install Koha. Fortunately, there is a good set of installation instructions for that in the INSTALL.debian file that comes with the distribution.

The new Koha 3.0 installer comes much closer to creating a useful test system than the 2.x one did. It offers to load sample item types, locations, and other data necessary to set up a functioning Koha. However, you still have to load bibliographic records, and create item records before you can actually use the system. You can download large sets of clean ISO2709 MARC records from the Internet Archive, if you don’t have any of your own. But creating items, if you don’t have any to migrate, is more of a problem. I ended up writing my own script for this purpose. It goes through a file of MARC records and adds the data for an item record into the 952 field of each record. It randomly chooses a location from among those created by the installer, or you can specify a different set of locations on the command line. Koha comes with tools that will create items from data embedded in the 952 field. So once the item data is in your MARC records, you need only load the MARC file to get both your bibliographic and item records. The script will probably be included in a future release of Koha, but until it is, you can get it from me, if you are setting up a test system and would find it useful.

The final result of all this work was a working Koha system with 250,000 bibliographic records, 250,000 items. I also loaded about 40,000 authority records, but that’s a story for another time.