A Whole Lotta Nothing Matt Haughey’s Personal Blog

outside.in

I’ve been playing with Outside.in for a few minutes (here is the Portland feed), and my first thought was that this is exactly what friends told me to do in 2000. MetaFilter was just taking off being barely a year old, and San Francisco friends urged me to make city-specific local news hubs that interested people could find links and news for and post to. Tom Coates had a great write up in fall of 2000 asking for precisely this. I was (and still am) too busy to take on such an effort, but it’s good to see someone trying to pick up the reins and try it.

I’m surprised they didn’t use CityName.outside.in URLs instead of outside.in/City_Name though.

Your last sentence reminded me of some trivia about how the web might have been.

One of the things Sir Tim Berners-Lee regrets is the way he set up URLs. In retrospect, he wishes he had reversed the domain and treated subdomains as folders (and skipped the //). For example, this post would be http:org/wholelottanothing/a/2006/10/24/outsidein/

Aside from preventing all of those clever subdomains (mine included), the format wouldn’t have made any distinction between directories in the root of a web server and roots on subdomains. What this means is that under the alternate-reality URLs CityName.outside.in and outside.in/CityName would be the same: http:in/outside/CityName

Who knows how he would have resolved conflicts when both existed, maybe that’s why we’ve got the URLs we do.

Or, more precisely, here is *a* Portland feed….this is also a Portland feed: http://outside.in/97214

Viva la Eastside…

i bet the folder method is used to be google friendly. Search engines have a habit of treating subdomains as different sites, so you can’t aggregate your page rank when splitting your site up with those. Also, most search engines tend to treat the folder parts of urls highly when ferreting out keywords, so they’d get a lot more juice for “portland” out of the folder name.