[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: PaceRepeatIdInDocument posted

On 6/2/05 8:41 AM, "Bob Wyman" <bob@xxxxxxxx> wrote:

> If
> anything we should address the lack of a standard method for creating
> atom:id's and we should create a requirement that atom:updated must be
> changed on *every* update -- not just on the whim of the entry author...


atom:updated was proposed and accepted because it provided the
noise-reduction feature of "author whim".

atom:modified was shouted down because (I suspect) people were sick of
discussing dates, and because it was argued that if two entries though
technically different in some content/metadata was not "significantly"
different enough in the publishers opinion then it was acceptable that the
newer version could be safely dropped on the floor.

At the time I argued that we need both, and that without atom:modified that
atom:updated would get polluted with every nitpicky modification under the
sun, thus destroying the original motivation for an <updated> date over a
<modified> date. And now you are proposing to enshrine that pollution into
the spec. Like I said: arrgh!

If you want atom:modified, you need to write a pace, and you need to really
bolster the rationale with all the use cases that require it, and be
prepared for a lot of resistance based not on the logic of the element, but
instead on a false logic of it having been discussed before (completely
ignoring any new and presumably stronger rationale in the pace). Religion is
like that.