Twitter just lit up with people talking about Bundl.it, a service that lets you build one single link that points to multiple pages. Tweet one short link to, say, your five favorite music videos. Sounds great, right? If only there was a way to have one single link that lets people find out about more than one page at once.
Folks, whenever you create a link, you’re leaving a bread crumb trail of your history. You’re building a list of things that you might want to go back later and reread.
What happens when someone wants to go back and hit that link 24 hours later? Like if they bookmark your excellent list of programming tools that you built in Bundl.it, and they just bookmarked that page?
Are you sure Bundl.it is going to be around?
This is why I’m really leery of any web tool that pops up without a business model. If they can’t figure out a way to be profitable before they run out of money, then any data I’ve got saved with them will vanish. Maybe it’s the DBA in me, but I know that data and backups cost money.
Bit.ly had an interesting blog post about permanence back in August around the time Tr.im was about to shutdown. 301works on the surface is supposed to provide disaster recovery for participating URL shortening services to archive the original URLS that were shortened. It seems 301works is still in the “works”.
URL shortening needs to be distributed and cached like DNS. They never change, so you don’t have any TTL or redirection issues, and it should be easy to provide archiving of the links. If every site provided their own shortening (although obviously their domain name will not be as short as tr.im), then if the site is up, the shortened versions are up.
In fact, you could use the technique here (https://dgl.cx/wikipedia-dns) to do URL shortening over DNS.