Not long after the iTunes Music Store launched as a Mac only service in April, enterprising hackers started figuring out how to expose the service’s information in XML for the purpose of doing deep links. Not long after that, Apple changed the format of the information available by following those links so that the XML wasn’t exposed, apparently to foil deep linking. Much wringing of hands followed. Now, in iTunes 4.1, Apple has baked deep linking directly into the user experience.
There appear to be three ways to do deep links on the iTMS without rolling your own URLs: one is to drag any hyperlink (not individual music tracks, but underlined artist or album names) to an application that supports drag and drop. You can also right click on a hyperlink or an individual track and choose the Copy Music Store URL option. That’t how I constructed the links last night. Doing either one results in a hyperlink using HTTP that goes to phobos.apple.com.
The third is the iTMS Link Maker, a web app on phobos.apple.com that walks you through a wizard to build links and gives you a JavaScript-based URL to go to the link. It also incorporates an IE plug-in control called the iTunes Detector that is loaded as part of the JavaScript solution that makes clicked hyperlinks behave “intelligently”—that is, when clicked, the links either open the selection in the iTunes Music Store, or take the user to a download page where they can get iTunes if they don’t already have it installed.
I’m less enthralled with this for a number of reasons. Number one, it requires embedding the script detector in the head of each page that bears the links—not a big deal when all pages on your blog are generated dynamically, but still something of a hassle. Number two, the iTunes Detector may be a lightweight piece of code, but it is not a lightweight user experience. I think I would rather have a link fail than pop an installer dialog over my blog pages.
Article comments
1 - Eric Olsen
Thanks Tim, very interesting. I was wondering WHY I would want to do this, but you came through in the end and it makes perfect sense.
2 - Mac Diva
Tim, has anyone declared 'ownership' of deep linking protocols? I am wondering because of the legal battle over who gets credit for Linux. Will Amazon, as apparently the most effective pioneer of deep linking, take exception as other companies horn in, possibly depriving it of market share? For example, if Blogcritics' contract with Amazon is not exclusive, it could start using deep links to other sites.
3 - Tim Jarrett
To the best of my knowledge, most of the legal activity around deep linking has been sites that have fought to prevent it, who see bypassing portal pages, ads, and other navigation as a curse and not a blessing. I don't think anyone will "own" the protocol for deep linking--it's just a permanent hyperlink--but will someone try to patent automatic link building tools that support affiliate programs, for example? Probably...
4 - Tim Jarrett
Hmm, spoke too soon: Patent 6,029,141, "Internet-based customer referral system." The claims are fairly specific to Amazon's business method, but they do cover "generating hypertextual documents with item-specific links that, when selected by a customer, cause the user's associate identifier and an identifier of a recommended item to be transmitted to the Web site system in a request message," as well as a bunch of stuff about associate recruitment and payment. That's as far as I go, since IANAL (and there are 42 claims).
5 - Mac Diva
Thanks! This is the kind of topic that awakens my 'legal' brain. I encourage you to blog it since it may become significant down the road.