Why no namespaces: Difference between revisions
Revision as of 10:37, 18 October 2009
In most of computer science, namespaces are a wonderful invention and greatly simplify maintenance.
XML namespaces were intended to provide these advantages to markup languages, and they have had some success within XML, perhaps because its draconian error-handling encourages careful use.
Unfortunately, XML namespaces don't work so well in HTML -- in practice, the web is more like a bunch of piles than like a well-organized filing cabinet.
Maciej Stachowiak explained some of the problems very well in http://lists.w3.org/Archives/Public/public-html/2009Jul/0919.html
An unfairly short summary is that:
Short names like "rdf:" or "foaf:" often work OK, if they are treated as globally unique. But by definition, they aren't globally unique, and they really can't be. In some cases, HTML has opted to define prefixes itself (such as aria-*), but the general XML solution is to use a URL.
Long namespace names, including URLS, are a pain to type, make the code unreadable, and are magnets for cut-and-paste editing which sometimes ends up invalidating the use anyhow.
See also: Namespace confusion