On 5/11/05, Graham <[EMAIL PROTECTED]> wrote:
> The "Implementor's Guide" may or may not ever exist and may or may
> not be read by anyone, so that isn't an option. 

Would you like to rephrase that as "I am concerned that the
implementor's Guide will never happen."? My honest impression is that
it will be far more popular than the spec itself, since the spec is
minimal and abrasive. The style of the document makes sections that
don't have specification words stick out like a sore thumb. I think I
understand the conflict.

> Myself and Sam, and
> (I think) Tim and Antone, would all like the actual spec to make an

I am asking you to detail the operational result you are after. You
are telling me what you want the spec to say. This is not productive.
I also made two suggestions in the email I pointed you to, but you
responded to only one.

> actual recommendation that entries without enclosed content include a
> textual summary, if at all possible.

Careful rhetoric proposing a different, and very strained,
interpretation of our charter is not going to fool me. This is the
kind of thing that makes negotiation unworthwhile. The charter is on
my side. Live with it. I am not especially fond of making the
argument, but it becomes necessary if people make scope arguments.

> Note no one wants to ban title-
> only feeds if they come from title-only resources. 

This is the problem I have. Title-only feeds are perfectly ok from any
resource. They are probably not a good idea as the only possibility
available from a blog or news site, but it's really all semantics and
judgement calls. Judgement calls for the publisher to make, based on
the type of application and users they are targetting. The whole point
of HTTP is to allow servers to represent resources in the amount of
detail they see fit.

Unfortunately, RFC2119 is not perfect. It contains no specification
word that would allow a validator to emit a warning while
simultaneously requiring support for title-only feeds. SHOULD would
also tar title-only feeds as bad practice. They aren't.

If the Implementor's Guide existed and were widely read, would it be
the right place for this encouragement? I think it would be. In my
opinion, this is advocacy. I happen to agree with it, but I don't
agree with normative enforcement in the data format. I will do
whatever I possibly can to make sure the Guide gets the attention it
deserves. I will write drafts of the section in question, if that's
what it takes. Do you want it to receive the most prominent placement
on atomenabled.org and atompub.org? I'm sure that could be arranged.

Robert Sayre

Reply via email to