For my Social Networking Site, I would like to build a facebook, or twitter similar URL rewriting naming convention.
Using Twitter as an example, they have pages labeled twitter.com/about and another page labeled twitter.com/{$username}
However, how do you differentiate between say a user who has registers on to our site as "about" then. From this we are going to have a server conflict between the user "about" and the page about.
What is the best way to handle this?
I would restrict the ability for someone to create a username that would conflict with existing urls.
Usually, you'll see this implemented so that conflicts are not possible. For instance, you could camp all users inside a virtual /users directory, or a subdomain: mysite.com/users/msilvis, or users.mysite.com/msilvis.
I would not recommend that you make all users accessible via the root directory of your site, because this could potentially restrain you from adding pages. For instance, suppose you do not have yet an "about" page, but a user registers and calls itself "about"; you're screwed.
If you still choose to do so, before registration, try an HTTP request to your website to the page the user would have. If you don't get a 404, then something already has that name.
If you're still in the early stages of development, you could look at the Kohana PHP framework. It's routing features solve the problem highlighted by zneak and dd, you would simply define in your routes that http://yoursite.com/about would point to a particular controller/action, i presume you're using the MVC design pattern considering you want a twitter like url structure.
Hope that helps!
EDIT: I forgot to mention that this doesn't stop people from signing up as about, it would just prevent the rest of the world from being linked to that persons page, in order to prevent them you'd need to employ some kind of validation against your core pages aka about, contact ect.
What our final decision came down to, was upon requesting a page on our server mysite.com/user, it first checked to see if that was a page, if it IS NOT a page, it assumes that it is a user, in which case it checks to see if that user is an object, if it is not then it gets passed to our 404 page.
So ontop of this, we are going to use the HTTP request like you mentioned earlier, and then if some how user "about" still signs up which we now have a page for, essentially sucks to be him because he is not going to see his page.