A. Pagaltzis wrote:
> * Sam Ruby <rubys@...> [2006-10-27 18:55]:
>> A. Pagaltzis wrote:
>>> No, it *is* a strange thing and it *is* obviously a mistake,
>>> no doubt about it. I was just saying that in my reading, the
>>> language in the spec leaves no room for any other
>>> interpretation, however nonsensical the prescribed behaviour
>>> may be. I certainly won't fault any aggregator for not
>>> complying with this aspect of the spec.
>> One could make a similar argument about adding attributes in
>> namespaces to existing core elements. On a number of elements,
>> the existing spec gives a list of attributes that are
>> permitted. At no point does it say "and you can add any other
>> ones you wish". The original RSS 2.0 spec also provides
>> a section on extensibility, and permits exactly one form.
>>
>> From my perspective, it seems to me that this advisory board
>> has taken a bizarre turn recently, from clarifying and
>> providing helpful observations on the state of practice (e.g.,
>> "produces should be aware that not all clients respect the
>> default"), to declaring portion of the spec that are perceived
>> to be inconvenient to be "sloppy" and engaging in the writing
>> of a new spec; in the process also ignoring the RoadMap
>> contained therein.
>
> You wrote this in reply to my mail, but I'm not sure whether it
> was specifically addressed at me, at least partially, or only at
> the group in general. So to clarify my stance, in case the above
> was aimed at me to any extent (also in case ayone else is
> unexpectedly paying attention to me):
>
> The above quote was meant purely as an observation about the
> original spec and implementations thereof, but in no way meant to
> suggest a direction for the profile.
>
> I have made such suggestions in the past and may make them again
> in the future, but they are merely for the sake of discussion;
> insofar as I have a voice at all (I have had no incentive to look
> at the RSS Board formalia), I am +/-0 on all decisions. I am
> interested in RSS insofar as it is to me a legacy format I have
> to support, but I have no stake of any kind in it otherwise.
> I also do not wish to affect those who have more vested interest
> in the format.
>
> Feel free to ignore everything I say on this list.

Aristotle, I was agreeing with you.

I recognize that this advisory board has a hard job. For example, in
the recent skipDays discussion -- as openly admitted by Rogers -- there
is only "thin justification" for stating GMT; but at least there is
*some* justification -- and this for an element which there is likely as
much disagreement over its meaning and support is likely at least as
sparse as ttl.

By contrast, I have seen no one bring forth any justification that the
base RSS 2.0 specification permits one to place namespaced attributes on
core elements, beyond Randy's assertion that "The schema attribute is
harmless".

And this was shortly followed up by a request that one overlook the
meaning of the word "default", which as Rogers correctly pointed out has
been in place since Netscape was actively stewarding this specification.

Based on this rather disturbing trend, I have gone from +0 to firmly -1
on this whole effort; accordingly I wish to distance myself from any and
all association with it, and I fully intent to start referring to this
as the fork of RSS 2.0 which does not respect the roadmap; and
furthermore I intend to advocate that validators like Validome provide
separate validation selections for "RSS 2.0 Harvard" and "RSS 2.0
Advisory Board", much like they currently do for "RSS 0.91 Netscape" and
"RSS 0.91 Userland".

As a point of reference, the Validome validator marks as in error any
namespaced attributes that may appear on core RSS 2.0 elements. As far
as I can tell, this was entirely based on an independent and thorough
reading of the specification; it most assuredly was not based on
anything I have said.

- Sam Ruby