[uf-discuss] hatom - multiple feed viability questions, now with examples

[Chris Casciano]

I brought up some questions about the definitions and consumption of 
html documents with multiple hatom feeds in them a week or two ago and 
after some discussion here and on IRC I was left with a feeling that we 
really needed some informal tests to outline some of the issues. I've 
tossed together a first attempt at a suite of examples now be viewed 
at:

http://placenamehere.com/mf/hatom_tests/

I /think/ all of these examples are valid to the hatom spec. I also 
think they all make sense from an html standpoint.

What I don't know, and the two main questions I'm looking to chase with 
these tests and the discussion around multiple feeds:

* Does the hatom spec have to change any to resolve some of the 
ambiguity that arises with multiple feeds? Do some of the current ideas 
like optional hfeed elements or optional id's on those elements fall 
down in practice.

* Are pages with multiple hfeeds consumable even if all the currently 
optional information is provided, and if so, what are the rules needed 
for feed detection, feed selection and ongoing feed consumption.

The later clearly a big question -- not one that I think we can answer 
easily -- a lot of the selection issues first come down to what 
applications you're dealing with -- but I think its worth keeping those 
issues in mind as the first question and the hatom spec gets flushed 
out more.


... more later, and I'll try and raise some of these issues in the 
wiki, but I'd appreciate first if anyone has any feedback on the tests 
themselves, any clarifications to the issue needed or if you think i'm 
totally out in left field with this stuff.

-- 
[ Chris Casciano ]
[ chris at placenamehere.com ] [ http://placenamehere.com ]