- From: Michael Champion <mc@xegesis.org>
- Date: Thu, 4 Dec 2003 08:58:03 -0500
- To: www-tag@w3.org
Mark Baker says: > These statements seem inconsistent to me. The latter seems to be saying > that it's ok to have a safe operation which isn't GET as long as it's > marked as safe, while the former says that in general (modulo the > described practical considerations), the only safe operation should be > GET. I don't see an inconsistency. There are many devils in the "practical considerations" details: Since not all safe operations can be practically bound to GET, and *empircally* not all GETs are safe (e.g. in a "pay by the megabyte" wireless environment), there is a use case for marking application-level operations as safe or not. Also, WSDL is designed to be independent of HTTP, so there is a use case for marking operations as safe so that this information can be preserved if there is no concept of safeness in some other protocol binding.
Received on Thursday, 4 December 2003 09:06:27 UTC