W3C home > Mailing lists > Public > public-html@w3.org > June 2009

Re: Why Design Principles?

From: Jonas Sicking <jonas@sicking.cc>
Date: Wed, 3 Jun 2009 20:48:42 -0700
Message-ID: <63df84f0906032048g397c58c7q5dc6021554d12599@mail.gmail.com>
To: Maciej Stachowiak <mjs@apple.com>
Cc: HTML WG <public-html@w3.org>
On Wed, Jun 3, 2009 at 7:38 PM, Jonas Sicking <jonas@sicking.cc> wrote:
> On Wed, Jun 3, 2009 at 7:34 PM, Jonas Sicking <jonas@sicking.cc> wrote:
>> On Tue, Jun 2, 2009 at 2:54 PM, Maciej Stachowiak <mjs@apple.com> wrote:
>>>
>>> Ian and Anne both suggested that I should add most of this justification to
>>> the Design Principles document itself. I will likely replace the current
>>> abstract and introduction with something based on this email. I suggest that
>>> those with an interest in the Design Principles should voice their
>>> objections to this plan.
>>>
>>> There are also some suggested additions and removals of principles at the
>>> end.
>>
>> I would love it if we had a design principle that described the fact
>> that we're writing the spec to support the authors that tests their
>> pages by running them in a browser, rather than by checking that they
>> comply with the spec.
>>
>> See a more detailed description in [1], but the short of it is that we
>> are generally avoiding leaving things explicitly or implicitly
>> undefined and hoping that authors won't rely on it. Instead we are
>> trying to define things in great detail to increase the chances that a
>> HTML document that works in one complying HTML UA, will also work in
>> another.
>
> Doh! Just realized that this is exactly what principle 4.1 is. I'll
> see if I can come up with wording to make it stronger.

How about:

Prefer to clearly define behavior that content authors could rely on,
in preference to vague or implementation-defined behavior. This way,
even authors who validate their content by only running it in one or
more conforming User-Agents is likely to produce content that will run
in other conforming User-Agents.

In particular, it is impossible to write an automatic validator that
validates that a given script only relies on behavior defined in the
specification. Thus we cannot create a validator for scripts, which
would leave authors to manually verify their scripts against the
specification. If authors were forced to do this, this would
significantly raise the bar for who is able to author HTML content,
and reduce the complexity at which scripts could be authored.

However, implementations should still be free to make improvements in
areas such as user interface, quality of rendering and performance.

Feel free to improve the english, or anything else really. An example
would be great too.

/ Jonas
Received on Thursday, 4 June 2009 03:49:39 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Wednesday, 9 May 2012 00:16:38 GMT