Short links (or tiny URLs) are the zip files of Twitter: they allow you to compress a large amount of information into a fraction of the characters needed to find the original object — be it an article, video, photo, or any other web-linkable resource.

While this is extremely useful — perhaps essential given Twitter's 140 character limit — there are some negative consequences to short links:

• hard to read
• ugly
• lack useful contextual information about what's being linked to
• can be a vector for malware or exploits
• many short links can point to the same resource
• they interrupt one's ability to scan through considerable amounts of tweets efficiently (again, because they provide no human-readable information themselves)
• they can expire, be redirected, and are otherwise unreliable for long term use

Short links, though, are not going away — as the medium defines the contours of the message.

Given this, there is still an opportunity to provide a better experience for consuming tweets with short links, and this page aims to capture current practice, approaches, and help explore new possibilities.

Expanding short links inline

Perhaps the most obvious approach to dealing with short links — like zip files — is to expand them in place, overwriting the original short URL.

Tweetie and Socialite support this behavior, as I'm sure many Twitter clients do. This works, but can lead to even uglier results (as seen in Socialite):

Tweetie takes an alternative and somewhat more elegant approach by previewing the result of an expansion in a separate window:

[Need to document other approaches: TweetDeck, Seesmic, etc]

Alternative proposals

I originally proposed a delayed-hover approach to this problem, substituting the ugly short urls and long urls with the text "[link]", and pulling the title of the target URL into the hover-bubble:

The expanded URL would be available in a user-selectable text input to make it easy to copy and paste the destination URL into another tweet, email, or other context.

Remy Rakic proposed a series of additional alternatives and described them on his Flickr page:

I've also taken one more stab at this problem with an approach that borrows from several of these ideas. Here you can see Brizzly's current auto-expansion approach:

And here you can see how I've changed their design to (hopefully) be more readable, user-friendly, and less noisy with the addition of arrows and an indication of the target domain:

I'm not sure the ellipses add anything, but the idea is that you could click on the ellipses to reveal the rest of the URL in place. To visit the destination, you'd click on the domain or arrow; in the case where someone linked to the root of a domain, there would be no ellipses for expansion.

Providing feedback

If this problem interests you, or you have ideas that you'd like to contribute — feel free to post them to Flickr, your blog, or elsewhere and tweet about them using the #shortlinks hashtag.

You can also contribute to this wiki page, but you'll need an account first, which you can request here. When you sign up, just tell me that you want to contribute to the short links page.