Re: spec review: ping attribute

On Sun, 28 Oct 2007 08:45:33 -0400, Julian Reschke <>  

> Charles McCathieNevile wrote:
>> ...
>> Indeed. What we are being asked to implement is a platform for people  
>> to make money or to keep a closer watch than ever on users.
>>  Fundamentally, the ping being sent is not a user request of any kind  
>> at all, ... I don't yet understand how it matters which method we  
>> decide to turn into a one-way message in the absence of a mechanism for  
>> such.
> It's bad because

You mean POST, right? As far as I am concerned, the HEAD request  
suggestion is the least departure from normal HTTP (since there is already  
llttle expectation that HEAD will pass a response to the user), but I  
still don't see

> - RFC2616 clearly states that the UA should not invoke an unsafe method  
> without the consent of the user, and
> - a unsafe method is not needed here.

We are effectively redefining a method, if we follow the current draft,to  
remove the expectation that it is a real transaction. While we are  
defining some magic handling of a method to make it "technically safe"  
(nothing bad should happen to the user in the UA as a result, although  
there is an argument that what we are doing is socially not reasonable -  
and that depends of course on your legal and moral framework and  
expectations about privay and other such difficult concepts) the argument  
that it should be one thing or another because of RFC 2616 seems to me  
specious, because I don't understand what is proposed as being what is  
offered in that spec - a way to perform direct transactions of information.

> So how is Opera going to do it?

Interesting question. "With a lot of careful thought about the industry,  
privacy and the implications" is all I can tell you at this point (except  
that while I may make recommendations on implementation, I won't be the  
one writing the code).



Charles McCathieNevile  Opera Software, Standards Group
     je parle français -- hablo español -- jeg lærer norsk              Try the Kestrel - Opera 9.5 alpha

Received on Sunday, 28 October 2007 13:18:27 UTC