TCS Daily

Deep Links? Yay!

By Arnold Kling - May 13, 2002 12:00 AM

Editor's Note: For a different perspective on deep linking, click here.

There is an old joke about how Hell was created. It seems that God and the Devil were walking side by side, and they passed by Heaven. "It's beautiful, " the Devil exclaimed. "Here, let me organize it for you."

These days, it seems as though the lawyers have discovered the beauty of digital innovation, and now they want to organize it for us. They want to be able to monitor and control the ability of people to use digital TV recorders to skip commercials. They want to control the design specification of every digital device in order to preclude copying.

And lawyers for the Dallas Morning News want to prevent other web sites from linking to pages within the DMN site, other than its home page.

Deep linking (the term for a link to a page other than a site's home page), is not a bug in the design of the Web. It is the Web's central feature. The Web was conceived in an academic setting (CERN, a European particle physics research organization). The idea of a hyperlink was to improve the annotation process, so that when as a reader you encounter a citation, you can go immediately to the page that has been cited.

Most Sites Want Deep Links

Even as the Web has evolved into a commercial medium, site managers overwhelmingly treat deep linking as a feature rather than as a bug. A typical site manager would prefer that you place a link directly to an article or service on the site, rather than link to the home page and force the user to search to find the article or service. For example, if I want to point to my TCS article on how music distributors could work with technology instead of against it, I say that the article is here. If I sent you instead to the TCS home page, that would waste your time. Also, it would put pressure on the TCS webmaster to design the home page as a gateway to the TCS archives, at the expense of other features. In fact, most web site managers anticipate and encourage deep linking. For example, sites where the content changes often, such as web logs, encourage others to "deep link" to the archives at which the content will be stored permanently.

Unlike 99.9 percent of sites on the Web, the Dallas Morning News does not want deep linking. I disagree completely with that philosophy, and I believe that it is indicative of a mindset that is likely to lead to a failed Web site. Nonetheless, it turns out that there is a straightforward technical fix. If you want to force new visitors to come through your home page, you can do that, by programming your web site to redirect outside users to your home page. (Part of the information that is passed back and forth between the browser and the web server is the page from which the browser came -- called the "referrer." Your web team could put a script on each page that reads the "referrer" information. If the "referrer" is an external web site, you could automatically redirect users to your home page, thereby overriding the deep link.)

Instead of designing its own site to work around the Web's deep linking feature, the Dallas Morning News sent out its lawyers to force everyone else with a Web page to change their behavior. They are making the argument that deep linking is a copyright violation.

As an ordinary civilian, I cannot see what the lawyers are talking about in making this a copyright issue. Suppose that in the print world you were to write an article that refers to something on page 27 of my book, Under the Radar. Could I claim a copyright violation because you referred the reader to page 27 instead of the book's cover? I assume not, but I lack any training in the law. From my amateur perspective, deep linking is analogous to specific page-citing.

Trickier Issues

I believe that the deep linking issue is a no-brainer. The Dallas Morning News is in the wrong technically, morally, and almost surely legally.

However, there are other issues involving the use of one site's content by another site that are more complex. These examples involve framing, screen capturing, and database access.

In framing, I create a split screen, where my site controls the main screen but your content appears in a frame on the page. Framing can be abusive. For example, it allows me to sell advertising for your content without your receiving any compensation.

As with deep linking, there is a technical work-around. If I do not want my pages to be framed, I can put up scripts that "unframe" them. However, I believe that the responsibility to avoid abuse of frames belongs with the framing site, not the site that is framed. Deep linking is normal and expected. Framing is not the norm, and a site manager should not be expected to design pages with the anticipation that they will be framed by other sites.

Screen capturing is even more problematic. As a technical matter, I can go to a page on another site, click "file/save" and then upload the page to my own site. I've seen this happen. In fact, when I was in charge of Homefair, I often would check the position of our content in search engines. On a number of occasions, several of the links that I found in the search took me to sites that had copied our content without our permission. They were so brazen that they did not even remove the sentence where we claimed copyright and said that the page could not be copied without our permission!

On the other hand, web databases such as Google and the Internet archive retain copies of pages, in part to maintain a record in case the pages disappear. This may be a fair use of screen capturing.

Finally, there is database access. Suppose that you have a site with real estate listings. I can write a "front end" program that allows users to run queries against your database from my site. As a technical matter, I can do this without your permission, although legally I would be on shaky ground.

The best solution for the database access issue might be open interfaces. That is, the database owner would make interfaces available to third parties. The database owner could capture revenue from third parties who use such interfaces. For an early example of this see Google's API.

Don't Fight Common Law

Screen capturing and database access raise valid legal issues. I can understand sites wanting to resist unrestricted screen captures and database access. And I do not believe that anyone would claim that screen captures and database access from third-party sites are intrinsic to the Web. However, I think it is fair to say that deep linking is intrinsic to the Web.

In fact, the Web has been around long enough that certain practices have become too entrenched to be uprooted by lawyers. Deep linking is one of those practices. Trying to use written law to fight against Internet common law is a loser's game.

TCS Daily Archives