- From: Adam van den Hoven <adam.vandenhoven@gmail.com>
- Date: Thu, 8 Nov 2007 08:31:26 -0800
- To: HTML WG List <public-html@w3.org>, Mark Baker <distobj@acm.org>
> On 7-Nov-07, at 9:34 PM, Mark Baker wrote: > Consider that when my Web server receives GET requests, it logs > (appends) a record of that in a log file. Does that make the message > non-idempotent and non-safe? No, of course not, because GET messages > are safe by *definition*. It does make the implementation both those > things though. Mark, In this case, the whole point of the request is to log some record to a log file. That is the only thing the request does. Its not a coincidence its the thing that resource is meant to do. Regardless of what the user intends, requesting that resource ALWAYS has the side effect. Besides, just because all you do is trivially add log the request in a file, that doesn't mean that is the only thing someone will do. Commercial analytics systems do WAY more than simply log results. Or lets consider a more sophisticated example. I'm going to use the information about the links that a user has followed to gain intelligence about the sorts of things that they currently care about. I could encode that directly into the URLs that they click but then I'll have to rebuild the whole system. Instead, I'm going to apply some aspect oriented thinking and use ping to tell me what people are clicking. This is contextual content negotiation and its definately a side effect. Adam
Received on Thursday, 8 November 2007 16:35:20 UTC