On 6 Jul 2006, at 20:22, Thomas Broyer wrote:
Conclusion: no, synchronization is not covered.
I once wrote a PaceOrderCollectionsByAppModified but it was reject.
http://www.intertwingly.net/wiki/pie/PaceOrderCollectionsByAppModified
+1 for me now, after listening once more to your arguments. Pleace
note this is a change of position
details below:
-----------------
Your examples are good ones. Synchronization it is true is only
covered if we treat :updated the way you are thinking of app:modified.
So the point is: if the spec is as we have it currently, people will
end up having to treat :updated as you wish app:modified to be
treated. This is because implementers of the protocol, not wishing
their users to fall into the traps you so well described, and having
to order the feed by :updated will make sure that even minor
corrections get a new updated time stamp. And so we will through
usage have
app:modified owl:sameAs :updated .
The problem with this though is that this may end up annoying casual
readers of the feed, who will keep finding these new posts pop back
up to the top of the stack, without there being anything they
consider significant. So readers may misinterpret the behavior of
perfectionists for that of spammers.
Ie. publishers will be in a bad spot. Either they loose their readers
or the loose their authors.
I think that therefore your case for app:modified is good. I have
CHANGED MY MIND ON THIS ONE. Editors note, and I think we reconsider
your PACE.
Henry
