Le 05-05-23 à 18:04, Robert Sayre a écrit :
Hi Karl. Thanks for the review. Some thoughts inline.
my pleasure
On 5/23/05, Karl Dubost <[EMAIL PROTECTED]> wrote:
Requirement 01: Include a conformance clause.
I tend to agree with Paul wrt conformance levels:
http://www.imc.org/atom-syntax/mail-archive/msg10296.html
:))) A conformance clause is helping you to do a bit more than the
language used in the specification. RFC 2119 is a good choice to
express the requirements of the specification. Nothing wrong with
that. But it's a kind of hidden conformance layer and with a lot of
regular different interpretation from the community. :)
Let's go a bit further though. Let's say I'm a developer who's
creating an authoring tool for Atom, so more a producer than a
consumer application.
"Atom Feed Documents"
"Atom Entry Documents"
I read "An Atom Entry Document represents exactly one Atom entry,
outside of the context of an Atom feed. Its root is the atom:entry
element."
Nowhere it is said that
"the root element of a conformant Atom Entry Document MUST be
"atom:entry"."
What is a conformant Atom Authoring Tool?
The specification says at the start: "This specification describes
conformance in terms of two artifacts; Atom Feed Documents and Atom
Entry documents. Additionally, it places some requirements on Atom
Processors."
If I read through the text "Atom Processors" are seen as consumers
application, not producers. It means that all XSLTs, Web hosting
services, etc can NOT claim conformance to Atom. :) I know it sounds
strange. You can decide in the specification to forbid such a claim
too, but at least it's better to be explicit than "obvious hidden
meaning" which might be misinterpreted.
Requirement 06: Create conformance labels for each part of the
conformance model.
NO. The conformance model being not completely clear. It's very
hard to know what are the different type of conformance and then to
put labels on them.
See comment #1.
I haven't taken the time :) to put the RFC 2119 together. It would
help to understand
what are:
- conformant "Atom Entry Document"
- conformant "Atom Feed Document"
- conformant "Atom Processor"
(- conformant "Atom Producer")
An Atom feed s/validator/conformance checker/ is a processor, though
it doesn't have to display or render the "Atom Feed Document". Will
it be a problem with regards to the specification?
I see also
"Atom Processors handle URIs."
"handle" is not an RFC 2119. It means that it's not even optional,
and not mandatory. Is it ok? Maybe "Atom Processors MUST handle URIs."?
In the prose before it is said:
"If the "src" attribute is present, Atom Processors MAY use the IRI
to retrieve the content. "
What is the behavior of the Atom processor when the MAY is not
implemented because it's optional and because it's not mandatory for
"Atom Processors" to handle URIs or IRIs.
I'm not being picky, but just showing that a conformance model tries
to tackle all these questions that might arise. It helps to really
define how you implement the requirements of the specification.
No consensus to make the schema normative. Everyone had a different
reason for opposing it, as I recall. :)
:)))
no more feed validator then. :)
Hope it helps.
--
Karl Dubost - http://www.w3.org/People/karl/
W3C Conformance Manager
*** Be Strict To Be Cool ***
