Learning something new

(PHOTO: SOOPAHTOE)

Unless you do all of your surfing by clicking on shortcuts and links, odds are at some point you’ll have typed in a URL. Sometimes these are stuffed to the gills with backslashes, these weird things ~, and often plenty of ? and & and a whole slew of numbers. Nightmarish to remember. But even if they’re reasonably straightforward it’s easy to make a mistake. Usually, such a mistake takes you to an error page, but sometimes, you hit on something nifty. I was trying to check my gmail the other day and by accident I typed http://www.gmai.com, which took me to GMA Industries Inc., a small scientific research company in Annapolis, MD. Once I was at their site I clicked around a bit, and discovered that they are also responsible for a site called pseudoscents.com, which they define thus:

As the developer of the worlds first non-energetic explosive pseudoscents, we design and develop scent products and services to industries related to defense, homeland security, fragrances, and the enhancement of the environment.

This is much more interesting than the misprint of hotmail. Typing in hotnail.com took me to a different page every time I clicked it - first a page about rheumatoid arthritis, then one called Blemish-be-gone, so it’s clearly an advertising site. It’s a great idea, if a bit more prosaic and markety.

I often feel like I’m going in circles on the web, clicking on the same old sites day after day. It was kind of nice to see something new, and sort of sciency too. 


Posted by Katie on June 09, 2008 at 2:31 PM in
Comments 3 Comments   Learning something new   Digg

Comments

Perhaps you could start a collection of links to web pages that take people out of their ordinary surfing circles.  Here are a few starting points:

Flickrvision: http://flickrvision.com/maps/show_3d

Random Website launcher: http://www.randomwebsite.com/

Random Website Design: http://www.strangebanana.com/generator.aspx

I wonder if anyone’s combined the idea of Markov chains and the clustering that typically comes in web links: build a spider that gathers information along the lines of “web pages that normally link to this page have such-and-such a probability of linking to this other page next” and so forth, and then generate random web pages from some seed URL.  In fact, it might be interesting to feed HTML to a Markov chain generator and see what comes out.


For really tangled URLs that you have to type (ie, can’t just cut and paste), I recommend TinyURL most highly. Its easy to use, free, and it does the job. Hyperdense URLs get ‘translated’ into something about fifteen characters long.

And for just finding stray sites of interest, the StumbleUpon Firefox add-on’s not bad.


Thanks for these, I will have a look. And in case any of you haven’t seen it before, the Digg Swarm is always entertaining! http://labs.digg.com/swarm/


Commenting is not available in this section entry.