Danny Ayers wrote:
I get the impression that the assumption is being made that by
reducing the number of elements, the amount of work needed for
producers and consumers will be reduced. But in cases like this the
more information the publisher can supply (discriminating between
trivial and significant changes), the less work is needed at the
client.
If there is an impasse, I believe it resides in the bracketed text above. I claim we can't usefully specify a distinction between trivial and significant - irrespective of whether we believe such a distinction exists or not.
I'm opposed to specifying that anyone coming into contact with Aom will have to think about change in terms of the format, i.e. as a matter of specification - they have enough to do as it is. Imagine for a moment a format that had "atom:trivial-change" and "atom:significant-change" elements. Why trouble people with this?
cheers Bill
