http://tinyurl.com/RHonJonU takes you back to this post.

Webcite/Webcitation do provide caching, but they also provide a single point of failure. I’d rather have a heterogeneous rotting collection of links than to depend on one domain, no matter how dependable.

My domain is vrypan.net. I also own pv0.info my “short domain”. I setup vrypan.net/urlborg.xml and let urlBorg know about it.

Now, anyone that goes to urlBorg and enter a link in my domain (ex http://vrypan.net/log/2008/05/14/urlborg-previews-mp3-maps/ ) will get a short link like http://pv0.info/j/k

Advantages:
1. I could make a backup of all the short links in my short domain and if urlBorg disapears, I could easily recreate them (after all I own the short domain)
2. Users know that all short links like pv0.info/…/ point to my domain so “blind-clicking” is reduced.

Regarding blind-clicking, any user who has an accoun in urlBorg can choose to see a “preview page” (a page showing the long url) whenever they click a short one.

Here’s one: http://tinyurl.com/garmin

David Berlind has a bunch of them here: http://twitter.com/dberlind

I’ve hunted around trying to find out how to generate the “prettier” and more intuitive ones but can’t find out how it’s done. Anybody know?

I agree that Twitter is an interesting case in point. While tweets are generally submitted in plaintext, they’re frequently viewed in HTML.

I would like to see twitter implement it’s own auto-linking service, which both provides shortlinks, and proof of the target domain. So this URL in a tweet:

http://some.really.long.com/with-a-really-long-url

Could become:

http://short.twitter.com/a4t6d2 [long.com]

Using third-party shortlinks in twitter.com seems to invite both link-rot and link fraud.

Thanks,
–kirby

the idea i have had regarding this issue is to simply nudge google search, and other search engine companies to implement new logic for short-url handling. if a short-url is no longer properly pointing to the real url, then i would like to add the short-url into a search engine like google and have the first ‘sticky’ result be the real url.

using search engine API services, this can be integrated as an anti-spam measure as well… such as calling the real url and web screenshot within an application.

so, a way to leverage the monster search engines with url-shortening services so that most concerns can be ommitted. Together with the url-sshortening services own measure, this should suffice.

The problem isn’t link-rot, the problem is spam – the inability to *see what you are clicking*.

A federation of short url services is a wonderful idea if it exists for reasons such as implementing blacklist’s to reduce spam and the hazards that come along with “blind-clicking”.

But link-rot? No.

As a blogger, webmaster, site-owner, url-renter, or whatever you want to call yourself – your primary focus should be influencing people to regularly visit *your site*, via word-of-mouth, bookmarks, bookmark-services, or random search. Emphasis on *your site* though, not short url’s. You don’t want people bookmarking short-url’s – which is what would happen if you gained any kind of control over them, such as “short-url mapping”.

They should be exactly what they are – throw-away, disposable, and insignificant.

Mapping would just allow spammers even more control over their spam is what my ultimate argument with this is.

Again, that’s not to dismiss the idea of a short-url federation. I think it’s a wonderful idea, just for different reasoning than you.

I’d like to see blacklists and bookmarking tools (del.icio.us) which automatically translate an “accidentally” bookmarked short-url to its proper url.

And the sites and services that are very reliant on short-url’s need to have some sort of accountability over the way they handle them. Twitter being the prime example. There’s so much blind-clicking going on on Twitter it’s ridiculous. They should either create their own short-url service, or fix the way they handle external short-url’s.

They would argue that it’s not their responsibility to do so… but when it affects nearly their entire user-base, I would suggest otherwise.