Do Tiny URL Services Weaken Net Architecture?
Indus Khaitan writes "Thanks to twitter, SMS, and mobile web, a lot of people are using the url minimizers like tinyurl.com, urltea.com. However, now I see a lot of people using it on their regular webpages. This could be a big problem if billions of different links are unreachable at a given time. What if a service starts sending a pop-up ad along with the redirect. What if the masked target links to a page with an exploit instead of linking to the new photos of Jessica Alba. Are services like tinyurl, urltea etc. taking the WWW towards a single point of failure? Is it a huge step backward? Or I'm just crying wolf here?"
Nobody knows how long exactly the service is made available. Please do some long-term thinking before using this, esp. in public forums. More than once, I couldn't follow those stupid mini URLs for whatever reason. They're just bad. More criticism: http://en.wikipedia.org/wiki/TinyURL#Criticism.
A Firefox plugin that recognises a TinyURL (etc) and then uses a popup to identify in a tooltip the actual URL and title of the webpage. - ~~~~
XML is like violence. If it doesn't solve the problem, use more.
Do URL shrinkers make matters worse? Maybe. But on the other side the web has always been a single-point-of-failure architecture. If the webserver hosting your content is down, your content is no longer reachable on the net. Things get worse when you only have only a few webserver/provider that are hosting stuff, youtube, facebook, myspace and friends host a ton of content, if they ever go down, you lose a whole bunch of content. Sure, they have plenty of redundancy and are pretty stable so its unlikely to happen for longer periods of time. But you still hand over a hell of a lot of control to a tiny few companies.
Solution? Turn the web into something where you refer to content instead of servers. Request documents by their MD5/SHA1/whatever checksum and whatever server has that piece of content sends it to you. You no longer have a single point of failure. Freenet, Bittorrent and a bunch of other P2P tools are already doing it in one way or another, because it is simply a more failsafe and faster way to handle content distribution. The days where everybody had his own little webserver are long over and it might be time to start addressing this issue on a big scale.
Camping on quad since 1996.
I think that's exactly the point the submitter is raising: Say you post a link on slashdot to some random website. When I stumble over your post, coming from a search engine, in five years, the chance for this link to still work is p(x), the probability of that random website remaining live for 5 years.
Now, if you used tinyurl for your link, the chance for the link to not be broken by then is p(x)*p(y), where p(y) is the chance of tinyurl surviving the next 5 years. Since p(y) is less than 1, this lowers your chance to send me this little piece of information forward to in five years time.
The internet is built on dense connectivity, with no single node being able to uniquely control access to a large part of the whole net. Tinyurl works against this principle. If someone switched off tinyurl now, 54 Million links would break in an instant, all over the web, with no chance to correct them all automatically.
In other words, to return your ad hominem attack: If you expect Tinyurl to stay exactly where and what it is for the next 5 years, you have misunderstood the web.