"Controlled environments" in scope for HTML working group

In http://lists.w3.org/Archives/Public/public-html/2009Dec/0130.html
Henri Sivonen wrote:

> The language aspect of the deliverable is scoped to
>  documents and applications on the *World Wide Web*.
>  Clearly, the Web isn't a "controlled environment". 
> I think language features aimed solely at controlled
>  environments fall outside the charter of this WG.

>  As Hixie mentioned, anyone who wants to reuse HTML
>  in a controlled environment can add their own features.
>  After all, they control the environment.

First, and most importantly, the "controlled environments"
of the world are definitely part of the web. While you
might argue about who might define this in general, the
scope of W3C working groups is determined by W3C members,
and I can assert with some certainty that the W3C members
who fund the W3C are as concerned -- if not more concerned --
about "controlled environments" as they are about the
public web. (The WhatWG constituency is of course different,
but we're talking about the scope of W3C HTML WG, and not
the scope of WhatWG.)

There are numerous W3C standards (as well as IETF standards)
which are intended only for deployment in "controlled
environments" but whose proper interoperable functioning
is important for making the web "world wide". If
it were impossible to use the web in intranets, if
URIs didn't work for WINS resolution, they would not
have been as successful or widely deployed. Features that
only have benefit in controlled environments are definitely
still of benefit to the public.

Second, the "control" of a "controlled environment" is
typically operational control -- the ability to control
configurations, to insist that everyone in a group
run the latest browser, plugins; the ability to restrict
software deployment, or manage the security context,
or filter for viruses, or ensure that web servers are
properly configured to label content with correct
content-type. The "control" is generally not one of
being able to build and support individual, unique,
or organizational-specific software. It is part of the
responsibility of standards groups like W3C and IETF
to develop standards that aid users of standard
software in their own "controlled environments", and
the suggestion that they can "add their own features"
because they "control the environment" -- nonsense.

Third, even if only considering web content for 
delivery use on the public internet, the process of 
building, editing, deploying content intended for the
 public Internet is often within a "controlled environment";
although those processes are not themselves part of public
web, features that support construction and management
of content prior to delivery are applicable and
relevant. 

Larry
--
http://larry.masinter.net

Received on Saturday, 2 January 2010 23:58:42 UTC