Mark Pilgrim just commented on my post from last month, Maybe RSS autodiscovery link usage should change (in Echo?) [Echo, at this point, is being called Atom].

updated: I added the link to Mark's website. I was packing last night when I originally wrote this, and totally spaced that!

What follows is an excerpt from an email I sent Mark trying to clarify my main concern / use suggestion, and asking for his feedback:

So, the summary use I am suggesting is:

a consistent interface to get the XML (e.g. Atom format) version of the complete content and metadata of each post, always available, beyond the scope of the "most recent X posts" feed

So, this is less than the RSS feed of a post and its subsequent comments, though the first item in such a feed would satisfy what I am suggesting.

Perhaps, with Atom, what I am suggesting can be accomplished via RSD pointing to interfaces that includes a getPost(URI) method, where the URI is the permalink of the post*. This would be sufficient autodiscovery from my point of view.

*are the web log APIs used to allow public access to getPost(), or are they only for the blogger's tools? I am definitely thinking about a most often publicly accessible interface.

I asked Mark to post any further comments / questions here, and am interested in other comments / questions as well.