Recently Dave Winer cited Brent Simmons' Law of CMS URLs : " The more expensive the CMS, the crappier the URLs." Around the same time I read Jesse James Garrett's essay, User-Centered URL Design . He notes:
...many of the most successful commercial CMS vendors require their customers to go to extraordinary lengths to implement an alternative to the cumbersome strings of commas, dashes, and digits their systems generate by default.
and goes on to stress the importance of URLs that can be easily read, written, spoken, and guessed by humans.
Naming is a deep, mysterious, powerful, and problematic subject. Of course it stinks if your CMS prevents you from assigning names, but even when you exercise full control over a namespace, it's hard to make best use of that freedom. In fact, the freedom to assign names can be a terrible burden.
In programming circles, from time to time, there's discussion of the so-called Hungarian notation pioneered by Microsoft's Charles Simonyi, and widely practiced both inside and outside Microsoft. (Simonyi has recently spun off a new company to explore his ongoing research in intentional programming .) Here's an MSDN article on Hungarian notation. It makes a point that's often missed: " the decision on the name will be mechanical, thus speedy." The same point was made in a 1991 BYTE article 1 by Charles Simonyi and Martin Heller:
[This example] demonstrates the simplicity of name giving once you decide on the basic type names. Name finding is easy because name giving was so simple in the first place.
It's hard to think up useful names for things. Rules and conventions to make naming more automatic and mechanical can be incredibly valuable. And not just to programmers. Every kind of collaboration creates and uses a shared namespace. Everyone benefits from a framework that brings clarity to the naming problem. The rules will vary by domain, but can be articulated in any domain.
A final point: the notion of user-oriented URL design is, of course, implicitly RESTian. It presumes that a web service wears its name on its sleeve, so to speak, where it can be seen.
1 That article unfortunately isn't online. When I created the original BYTE archive, we arbitrarily decided to start with 1994 content and make everything thereafter fully available. As it happens, I've got an electronic copy of the Simonyi/Heller article. It was on the BYTE CD-ROM, which covered the period 1990-1998. Could I reprint the piece on my website? Dunno. Maybe. It's hard to find out. Too bad there aren't licensing schemes in place that would help me decide.
Former URL: http://weblog.infoworld.com/udell/2002/10/02.html#a431