That being said, technologies like massively parallel search engines (Google, etc.) have all infrastructure needed to reconstruct these link relationships. If a popular link shortener did go belly up, Google or Microsoft have all the data needed to reconstruct the data. Similarly, Twitter could record all the link relationships in their own logs and restore them if any should die. In fact, if I were twitter I would do that since a portion of their value prop is directly related to their role as a link archive. If any of these services die they would loose some portion, admittedly unknown, of their value.

http://www.sciencetext.com/shorten-your-urls.html

There are other plugins that just do the shortening using your domain. Unless you have a two letter domain and two letter TLD they’re not going to be as short as tr.im et al, but v useful nevertheless.

Besides, even all public URLs are subject to link-rot.
You have to accept that the web is not static, but its becoming a real time information river; URL shortening services fit in perfectly, as an additional service.

It has puzzled me why Google has not bought one of the shortening services or created their own. There is a tremendous amount information that can be captured about the relevancy of a site.

Although the shortening is needed for Twitter, I don’t think it should be used elsewhere. I see some site using it to expose the URL, but I think it is counterproductive. You see a short string, but have no idea where you are headed. That could lead to viruses and other bad sites. tinyurl.com/prettyflowers could lead to a porn site. Exposing the URL gives the reader some information to decide on whether to click on the link.