Re: Differences between the W3C and WHATWG specifications

On 06/18/2010 02:36 PM, Laura Carlson wrote:
> Hi Sam and Ian,
>
> Sam wrote:
>
>> I think we need to work more on working together.
>
> The 18 June 2010 editors' draft states, "Work on this specification is
> also done at the WHATWG. The W3C HTML working group actively pursues
> convergence with the WHATWG, as required by the W3C HTML working group
> charter."
>
> It is good to point this out in the spec and remind ourselves and
> others of this goal.
>
> Convergence does need to take place. Having one spec would be very
> beneficial. [1]
>
> Ian, what needs to happen from your perspective for convergence to occur?
>
>> The WHATWG
>> document removes a reference to WCAG for... what reason exactly?"  I'm
>> especially puzzled as that particular difference was not motivated by
>> any WG decision.
>
> In bug 9241 [2] I asked for the spec to say, "For guidance on
> accessibility requirements for text alternatives authors should
> consult WCAG 2.0" and link to WCAG 2.0. Ian did provide the link and
> relevant text. Thank you, Ian.

I guess I wasn't clear.

I understand that Ian fixed a bug.  What I don't understand (and the 
context here is the topic of convergence) is why Ian agreed to fix that 
bug, but only in the W3C copy of the spec.  Either there is a problem 
with it or there is not.  To fix it in just one spec is, in itself, an 
indication that convergence is felt to be an important criteria.

In short, I believe that if we are jointly pursuing convergence, each 
every difference merits having a fully documented rationale.

As to the other two examples: I understand that there are real customer 
requirements that the ping attribute attempts to address.  I also 
appreciate that there are deep technical issues with the solution as 
documented in the current WHATWG draft.  I believe that the current 
information in the "Is this HTML5?" section is deeply flawed in that it 
omits a crucial part of the story.  I will further note that the areas 
where this attribute is described later in the spec have absolutely no 
marking indicating that they are not part of HTML5.  To the contrary, 
they are each marked as being in "Last call" status.

As to Atom.  This happens to be an area of expertise of mine, and I was 
fully prepared to recuse myself from any decision relating to that 
topic.  I will say that I do believe that the algorithm as currently 
described is deeply problematic, to the point where I don't believe that 
the problems are solvable.  And to date, I have never seen anybody put 
forward a rationale for this feature, beyond a brief statement[3] that 
some unnamed person requested it, and Ian believes that the definition 
of this to be harmless.  A statement that Ian has made without providing 
any evidence.

I will further note that this item is also marked as in "Last call" status.

I truly believe that if convergence is desired, then differences such as 
these need to be minimized, and -- where the differences are truly 
necessary -- appropriately and fully documented.  I'll go further and 
say that if convergence is a priority (and I would suggest that it 
should be), the bar for differences should be considerably higher than 
"defining one seems harmless" as the divergence itself is actively harmful.

> Best Regards,
> Laura
>
> [1] http://lists.w3.org/Archives/Public/public-html/2010Jun/0320.html
> [2]  http://www.w3.org/Bugs/Public/show_bug.cgi?id=9241

[3] http://lists.w3.org/Archives/Public/public-html/2010Apr/0419.html

- Sam Ruby

Received on Friday, 18 June 2010 19:29:30 UTC