Re: #238, was: User interface requirements for redirecting to unsafe methods

On Oct 23, 2012, at 12:17 AM, Julian Reschke <> wrote:

> On 2012-10-23 09:12, Mark Nottingham wrote:
>> On 23/10/2012, at 6:00 PM, Julian Reschke <> wrote:
>>> the text you complained about back then has been rewritten a few months ago. Please check the WGLC draft.
>>> With respect to tracking: I think this is issue <>, see also <>.
>> Yes.
>> I do see, however, that we have this language at the top of <>:
>>> If the required action involves a subsequent HTTP request, it MAY be carried
>>>    out by the user agent without interaction with the user if and only
>>>    if the method used in the second request is known to be "safe", as
>>>    defined in Section 5.2.1.
>> MAY... if and only if is a badly constructed requirement, and it also goes against the resolution of #238. I'll enter a WGLC issue to correct that.
> Indeed; when we resolved #238 we didn't fix all places where this was mentioned.

I only checked the first reference, which was apparently unfixed. I have now double-checked the others and they do seem to address the issue.

Thanks to Mark for offering to raise the overlooked spot as a WGLC issue.

As a matter of process, I wish I could have been notified in some more active way about the resolution when the ticket was resolved. Since it was resolved several years after I raised the issue, I was no longer monitoring the WG closely enough to notice. Had I been aware of the resolution, I likely would have spotted the error (i.e. that one of four mentions of confirmation was not addressed).


Received on Tuesday, 23 October 2012 08:57:17 UTC