On Tuesday, November 2, 2004, at 10:28 AM, Bill de h�ra wrote:
I think the level of discrimination in intent some people are looking for with these dates isn't best achieved with the Atom spec. I suggest leaving this to people's own writing/publishing processes or UI > design.I think atom:updated defined that way would be useless. And I don't think that "the qualifying fluff" as you call it imposes an unwarranted decision making burden. I think its usefulness is worth the trivial effort required to remember to check or uncheck a checkbox, or whatever trivial task would be needed to tell one's publishing software whether or not to change atom:updated. As long as the default is set intelligently (letting the user decide what the default should be would be ideal), the effort required is going to be trivial. Anyway, if changing atom:updated is optional, whether the basis for changing it is loosely defined or totally undefined, the option will have equal need to be presented to users.
My preferred text for this element is this:
[[[
The "atom:updated" element is a Date construct indicating the time when the [feed/entry] was changed. Not all modifications to an entry require a changed atom:updated value.
]]]
ie, Robert hasn't gone far enough. My strong opinion is that with qualifying fluff, we're imposing on users to decide about things they needn't be bothered about much if any of the time.
