The closing of tr.im has, for a day at least, shone a little bit of light on the URL shortening industry. It has also shed light on one of several problems with URL shortening - an idea with good intentions, but potentially bad results.

When tr.im shuts down for good on December 31, 2009, any links that were created using it will cease to work. Actually, that is the best-case scenario - the worst-case would be the domain falling into the hands of folks with malicious intent and old links suddenly becoming a trail to malware. One would hope that the people behind these URL shortening services have the foresight to register the domain for several years and allow it to just sit there dormant. However, if there are creditors involved, it won’t be that simple - a domain with thousands or millions of links pointing to it has a fair amount of economic value and would be sold off to help satisfy debts.

This is a scene that is likely to be replayed many times in the future - there are a plethora of URL shortening services, and no apparent ways to make money off the concept. This in itself could be another entire post, but let’s just leave it at “Hmm… Where have I seen this pattern before?”

This isn’t the only problem with URL shortening services - if it were, they wouldn’t be so bad. A lot of these services see the most use on Twitter, where the average link has a shelf life of about 38 seconds (I made that up), which mitigates the problems that occur when they die out to a degree. Lots of the links do hang around in other places, however, so we can’t ignore this problem, but let’s look at a few more:

1. They obscure what you are clicking on

This is probably the worst problem. For any given hot Twitter topic, there may be hundreds of tweets flying by per minute - many of them containing links. It would be easy for someone to slip in a link to that says “Click here to see pics from the Iranian riots”, but is really a link to porn, a rickroll, or some malware. When you see the actual URL of the website, there is a small amount of security in being able to see where you are heading. By every link looking exactly the same (ie, bit.ly/ followed by a few random characters), this first line of defense is circumvented completely.

2. They are an extra link in the chain

There are no shortage of things that can go wrong and make your website inaccessible to end-users. Power outages, network outages, hardware problems, DNS issues - just to name a few. Adding a URL shortening service into the mix is just one more link in the chain where a user can be frustrated by a website that doesn’t load.

An important lesson for webmasters and web framework developers

If anything, the fact that an entire industry has sprung up around shortening URLs so they aren’t impossible to even copy & paste effectively should serve as giant clue to web developers everywhere - your URLs are too damn long. We should be mindful of this when designing websites, web frameworks, and web applications (ie, CMS apps). A lot of new frameworks already are - Rails, Grails, an Lift all spring to mind as having nice facilities for making reasonably short URLs.

Some websites have necessarily long URLs because that is how they maintain state - this is unavoidable in some cases, but you can still provide a nice button or something in a prominent location on every page that has a shorter version of the link for people to use.

Small measures like these can help alleviate the need for URL shortening services and the problems that accompany them.