« Crawling for Sitemaps | Main | Document Design Matters »

Wednesday, October 01, 2008

Comments

Feed You can follow this conversation by subscribing to the comment feed for this post.

yiming

That's good to know. Amazing that there's no firm answer to this ambiguity from specs. Must be a headache for search engine indexes and caching servers.

There are things like ASP .NET and various Microsoft servers, which explicitly state in their docs that "query string values or form post values with identical key/value pairs as identical for caching purposes, regardless of the order in which the parameters are passed". Pragmatically, sometimes all you get is hash-like access to the parameters. Your app dev stack made the philosophical distinction (or lack thereof) for you.

You'd think that the standards bodies here would want to lay the down a ruling on the matter. Why should something as important as equality of resources as represented by URI be left to the application implementer? Not being able to assume equality one way or another seems to me to be inviting inconsistency at every level (the app decides to take one definition of URI equality; the web framework is designed for another; the programming language's HTTP libraries yet another; and the caching servers in the middle all potentially configured differently ... )

The comments to this entry are closed.

Flickr