« hard drives | Main | Trackback sucks »
March 23, 2003
Web URLs
While listening to Rasmus Lerdorf talk this past week, he began discussing his ideas on how web URLs should work. Essentially he believes that there is no reason for URLs to be as complicated and confusing as those you can find being used today. The idea being that for example you could have a url like http://www.ibm.com/t70/processor where each sub-directory is a smart script to find the information you're requesting. This is a rather interesting idea, in that it makes the idea of the web usable again. There really isn't a reason why I should have a URL with a line like show.cgi?adlk20wljkhwso928eweohh2 for me to find information about a laptop (in this case).
I've heard him talk about this before, and never really paid attention to it for some reason. Only catch is this time I realized that I actually believe it (witness my recent blog archive file change). I'm starting to wonder why it's not really implemented in every area, considering the usefulness that this would provide not only to the common web surfing, but also to the corporate world and cutting down on costs for support.
Posted by Dan at March 23, 2003 07:39 PM
Comments
I don't know. I go back and forth between the usefulness of making URLs friendly. I agree that creating a URL of archives/2003/03/23/... is a simple way of showing when the current page was created, but shouldn't that info already be on the page? Do many of your visitors browse your site by changing the URL?
On one hand, I like it because it's simple and the same across many websites. I don't need to figure out the format. On the other hand, I don't see the practical benefits of making my URLs that long.
So, yes it shows the date this document was created, but that's about it. I'll still continue to consider it though.
Posted by: Joshua Kaufman at March 26, 2003 05:56 AM