Tim Bray wrote:
On Nov 9, 2004, at 3:09 PM, Robert Sayre wrote:
There really aren't very many changes. Just removes all the stuff that says "feed" instead of "entry" and makes four changes.
A change this big, this late in the process, makes me nervous.
Let's see. The atom:head element appeared rather recently, and the format isn't supposed to go to Last Call until March, 2005.
There are other ways to shorten the spec (we were discussing some today). What makes me nervous is we're asserting that a <head> and an <entry> are the same, except that <head> has to have a generator... it smells like, when I'm building some software, and I catch myself thinking "gee, this thing and that thing are almost the same thing" and I try to unify them and find extra arguments and variables creeping into methods to make up for the fact that they're not *quite* the same. It seems to me that <head> and <entry> are two different bundles of metadata, in which a lot of the same elements can appear.
Well, I see your point, except that the protocol currently requires a generator element to appear in an Entry Document (not reflected in the format spec), so maybe the elements in the feed are the exception. [0]
> Finally, <headentry> is a bogus element name. Either it's a <head> or > it's an <entry class="head">.
It's a <head>, then.
Robert Sayre
[0] http://bitworking.org/projects/atom/draft-ietf-atompub-protocol-02.html#rfc.section.3.6.11
