Tim Bray wrote:
On Feb 6, 2005, at 3:16 AM, Henry Story wrote:
Thinking about it, I would be in favor of clarifying it to be "any change should result in a date update".
No, atom:update is subjective. How UIs support that subjectivity is best left to innovation. I like Graham's wording as it makes me think UI designers will need to give 'save != update' proper consideration.
If you will go back and review the correspondence, you will discover a lot of argument, the upshot of which was, different publishers have very different semantics for what they consider an update. I had a posting months and months ago that enumerated a bunch of corner cases (e.g., stylesheet update) that made the whole notion of "any change" rather problematic. There were many who, like you, liked the notion of "objective change" that always changed the date. The reason that we have "updated" with the current language is it reflects the reality that the *only* kind of change that can be cleanly defined is "The publisher thinks it changed."
Strong agreement and as I'm not constrained by being a chair, I'll go further than Tim - using this pace to reopen objective updating is out of order.
cheers Bill
