On 5/21/05, Eric Scheid <[EMAIL PROTECTED]> wrote:
> > Here are some simple questions, which
> > you can answer by reading the example I gave, and reading the draft.
> >
> > http://www.imc.org/atom-syntax/mail-archive/msg15380.html
> >
> > Who is the author of that entry?
> > Who are the contributors?
> 
> The problem with the example you gave is that it suggests that even entries
> with just the one author/contributor would need two person constructs in the
> entry, or maybe just the one ... either way it's confusing.

No, it doesn't. Why are you saying it suggests that? There are two
questions and two answers. I don't see a problem.

> 
> Also, more importantly, how do you then indicate which individuals listed as
> <contributor> have an <author> credit and which individuals were only
> <contributor>s.

Why are you mapping elements to types of credit? There is nothing in
the spec that suggests they map to types of credit. As I said, please
read the draft.

> 
> Consider this example (and please note that the second line isn't the
> literal bits on the wire):
> 
> <entry>
>     <author><name>##### a list of THREE names #####</name></author>
>     <contributor><name>Bob Bellows</name><uri>B.html</uri></contributor>
>     <contributor><name>Fred Fellows</name><uri>F.html</uri></contributor>
>     <contributor><name>Jon Jello</name><uri>J.html</uri></contributor>
>     <contributor><name>Ada Aiello</name><uri>A.html</uri></contributor>
>     [...]
> </entry>
> 
> Now, this *is* a valid format-08 document, right? BUT can you tell me who
> the THREE authors are, and who the fourth person is who is only a
> contributor (and not an author)?

No. Why do you think that's a requirement? Couldn't you add an
extension to differentiate them?

> 
> So: valid format-08, but junk data.
> 
> > There is no mention of 'byline'.
> 
> There most certainly is a mention of the thing I have referred to elsewhere
> as a "byline". It happens to be serialised within <author><name>. Instead of
> the literal element name "byline" I could just as easily use the literal
> element name of "authorship".
> 
> > format-08 works.
> 
> for some definition of "works" ... like "passes the validator", but not
> "provides meaningful semantics".
> 
> > I fully agree that other ways of arranging authors and contributors
> > are possible and reasonable, but no one has demonstrated a document
> > that format-08 can't cover.
> 
> can we shoe-horn data into elements? sure.
> would that document then pass the validator? sure
> can we extract that data in a meaningful sense? no.
> 
> > At this stage, changing the spec to suit
> > religious preferences would be extremely arrogant.
> 
> Oh, please, stop trolling.

Seriously, 2 more months of this crap to solve a 'problem' which you
can't give an example of. You might have a perfect spec, but no one
will care if it is never done.

Robert Sayre

Reply via email to