Huh? "GET would be a lot more useful if it didn't depend on URL semantics"

Article:
When to Use Get?

Subject: Huh? "GET would be a lot more useful if it didn't depend on URL semantics"

Date: 2002-05-09 21:55:08

From: Mike Dierken

This is in response to a comment containing "I think GET would be a lot more useful if it didn't depend on URL semantics."

Doing a GET - or any retrieval - requires an identifier of the information being retrieved. Semantically you can't avoid that.
The syntax of URIs are pretty much worked out, and the HTTP doesn't need to be replaced or changed.
If Java URL encoding is broken, then someone should fix it and provide the newer more compliant code for all to use. If I had the time and the understanding of the i18n issues, I'd do it, but time is a scarce commodity for me.

Also, if using verbose 'query' syntax to retrieve information is tough with URLs - and if you control the server - you could always create (via POST) a resource that represents the query and then do multiple GETs on that resource (which could have a much smaller/concise URL). Kind of like creating a view in RDBMS land.

Previous Message Next Message