Consensus and Legacy Compatibility (was: Support Existing Content)

Hi, Maciej-

Maciej Stachowiak wrote:
> 
> It seems to me that if there is disagreement on fundamental issues (such 
> as whether to be compatible with past versions of the language), and our 
> goal is consensus, there are only a few ways to proceed:
> 
> 1) Advocates of one side of the argument change their minds.
> 2) Advocates of one side of the argument leave.
> 3) We stop work.
> 4) We proceed over the objections of some, given strong arguments and a 
> rough consensus of a large part of the group.
> 
> Right now I think we're taking approach #3. I hope at some point we will 
> move on to #4, if we see no signs of #1 or #2 occurring.

In the interest of moving forward and establishing consensus (which I 
value highly), I am changing my mind on the issue of "compatibility". 
By compatibility (a loaded term if there ever was one), I take it to 
mean error handling for validity and well-formedness errors (found in a 
lot of existing content), and allowance for deprecated/obsoleted elements.

I have been reading the WHAT WG specs and the more lucid posts in this 
list, and chatting on the #html-wg IRC channel, and while I personally 
don't like the approach, I can't find serious technical flaws with it. 
Moreover, it has the support of some very smart people and most major 
browser vendors (and Microsoft seems neutral on the matter), so I'm open 
to the idea that my perspective is wrong.

I am sympathetic to the view that would like to encourage better 
authoring practices going forward (particularly for content from 
authoring tools), and I hope there can still be some way to do so.  But 
perhaps there are better ways of encouraging good authoring.

So, I encourage others like me who have been skeptical of the WHATWG's 
approach on this issue to reconsider your position without 
preconceptions.  I think we should concede on this point, and see where 
it takes us.  Frankly, if it doesn't work, we will still be better off 
than we are now, since browser interoperability is not great today.

Ultimately, it is more important that we get all the browser vendors on 
board with the same agenda and behavior than that we try to change the 
behavior of authors.

Regards-
-Doug

Received on Friday, 4 May 2007 08:35:33 UTC