The Web has been busy trying to find a way out of URL shortner services. Kellan Elliott-McCrea has worked out a solution which lets the publisher gain control over the shortening. I think using rev to indicate the shortened version might get confusing, but it is also true that it seems to offer the best balance in the current situation.
Somehow I am still biased towards rethinking the idea of URL shortening. I rarely use URL shortening services because it not only hides the original URL but also adds a level of indirection, which I think goes against the basic nature of the Web. I like the question that Dave Winer asks:
Why should a user ever see the longer crappy url?
Why do we have the longer version at all? If the long URLs create a problem for the users, it talks of their unusability. We already have clean URLs, maybe it is time to force the original URLs themselves to be short and comprehensible. Or maybe we need a hand-held device specific solution.
Also, I feel that most of the problems with long URLs are more because of the interface through which we are using them, than the URLs themselves. So I just wonder if a better solution to this problem might lie elsewhere.
April 13th, 2009 at 7:57 pm
I think microblogging services are much more to blame for the proliferation of URL shorteners than anything. Even the most succinct clean URL can’t come close to it’s is.gd short URL. And with only 140 characters to work with, most people don’t want anything longer than that. I do think there could be some nicer services to expand short URLs for those of us browsing the web who would like to see what things are before we click them.
April 14th, 2009 at 3:47 pm
G’day,
The canonical URL for a resource should be clean, relatively concise and contain human-friendly information such as the title, category, date, etc. This can be tightened up using rel=”canonical” and I think we’ll start to see this taking off.
There is a place for a short URL outside of microblogging services. For a start there’s email (some clients still break URLs) and of course mobile access (small screen, SMS, etc.). Finally there’s any time you need to manually enter the URL – e.g. printed or spoken versions.
So we need both. rev=”canonical” is a notoriously bad solution though… and it’s downright dangerous given people are already confusing rev with rel (rev was deprecated with good reason and IMO should stay dead). My money’s on shortlink – there’s no confusion here and no risk of severe breakage.
Sam
May 4th, 2009 at 11:30 am
[...] I have always wondered about the reasons behind the length limitation of SMSs, that has been inherited by numerous applications that use it. Mark Milian has an interesting story of how the length came out to be 160 characters. The story highlights that the research conducted to deduce the number, that was sufficient to convey messages, was driven by usability and effectiveness than technology limitations. The research team used data from postcards and telegraphs to support their assumptions. The only data type that breaks the SMS today is Web URLs, which would have been difficult to consider at that time. And that, I think, deserves its own solution. [...]