W3C home > Mailing lists > Public > public-html@w3.org > August 2016

Re: CFC on referencing the Image Description (longdesc) extension

From: Chris Wilson <cwilso@google.com>
Date: Fri, 12 Aug 2016 11:31:44 -0700
Message-ID: <CAJK2wqWxMjzAe+s5X1DZv53FkrZN1itQ3=ODvXyUrbJ0-R=jXw@mail.gmail.com>
To: Katie Haritos-Shea GMAIL <ryladog@gmail.com>
Cc: David Singer <singer@mac.com>, Léonie Watson <tink@tink.uk>, "Edward O'Connor" <eoconnor@apple.com>, HTML WG <public-html@w3.org>
See the Github issue for my response to that.  I understand the value of
compromise, and appreciated Léonie's work in crafting one; but I think
there is an architectural principle at issue here, and I don't think
architectural principles should be sacrificed for political reasons.
Modules should be modules; architectural layering is important.  Inserting
any examples of non-core modules gives them "some animals are more equal
than others" status, which sets a bad precedent.

On Fri, Aug 12, 2016 at 11:02 AM, Katie Haritos-Shea GMAIL <
ryladog@gmail.com> wrote:

> It sounded like a very good compromise to me, and I thought it had been
> agreed to and that is why it was going to CfC, and why I supported it.
>
>
>
> Though I wasn’t *happy* about the agreement (but was willing to accept
> it), compromise means "you don't get everything you want", right?
>
>
>
> ​​​​​
>
>
>
>
>
>
>
> ** katie **
>
>
>
> *Katie Haritos-Shea*
> *Principal ICT Accessibility Architect (WCAG/Section 508/ADA/AODA)*
>
>
>
> *Cell: 703-371-5545 <703-371-5545> **|* *ryladog@gmail.com*
> <ryladog@gmail.com> *|* *Oakton, VA **|* *LinkedIn Profile*
> <http://www.linkedin.com/in/katieharitosshea/> *|* *Office: 703-371-5545
> <703-371-5545> **|* *@ryladog* <https://twitter.com/Ryladog>
>
>
>
> *From:* Chris Wilson [mailto:cwilso@google.com]
> *Sent:* Friday, August 12, 2016 12:16 PM
> *To:* David Singer <singer@mac.com>
> *Cc:* Léonie Watson <tink@tink.uk>; Edward O'Connor <eoconnor@apple.com>;
> HTML WG (public-html@w3.org) <public-html@w3.org>
> *Subject:* Re: CFC on referencing the Image Description (longdesc)
> extension
>
>
>
> In the interest of clarifying my contribution to the discussion in the
> issue in Github, I believe we should:
>
> 1. Remove the longdesc attribute from the table of attributes in HTML core.
>
> 2. Remove the IDL information for the longdesc attribute from HTML core.
>
> 3. Remove the longdesc examples in HTML core.
>
> 4. Create a WG Note listing known extension specifications, and giving
> examples of how those modules work.
>
> 5. Optionally include a non-normative link to that Note from HTML core
> (probably in the index).
>
>
>
> IOW, longdesc (or any other module) should not be directly referenced in
> the core.  There should be informative examples, but they should not be in
> the Core, or it's not a Core document, it's comprehensive (e.g. every
> future module will want to be listed in the Core, or it will believe it is
> "second-class").
>
>
>
>
>
> On Fri, Aug 12, 2016 at 9:02 AM, David Singer <singer@mac.com> wrote:
>
> Hi
>
> I think we’re fine with the other actions in the CfC, but like Ted, I
> strongly feel examples belong with the specification of the thing they
> exemplify (or, if an example of integration of several specs is needed, in
> separate informative and tutorial material such as those that the
> accessibility groups in W3C are chartered to produce).
>
> cheers
>
>
> > On Aug 12, 2016, at 3:18 , Léonie Watson <tink@tink.uk> wrote:
> >
> > On 11/08/2016 21:31, Edward O'Connor wrote:
> >> Hi Léonie,
> >
> >
> > Hello Ted. Good to have your input on this, thanks.
> >
> >>
> >> You wrote:
> >>
> >>> 3. Keep the longdesc examples in HTML core **.
> >>
> >> This doesn't make sense to me. When we moved Microdata into its own spec
> >> because people objected to it being in HTML 5, we also removed Microdata
> >> from the examples in the HTML spec (e.g.
> >> <https://www.w3.org/Bugs/Public/show_bug.cgi?id=18726>). Also, in
> >> general, including features defined in extensions in examples in the
> >> core spec strikes me as bad layering.
> >>
> >
> > At the moment there is no real consistency in the way we reference specs
> (applicable or otherwise) from HTML. My hope is that over the next few
> months the WG will be able to discuss and agree on definitions for modules
> (as opposed to extensions), and a method for referencing either/both of
> those things. The idea being that we'd then be able to start putting that
> plan into action across HTML.
> >
> > In the meantime the hope is that by removing all the normative
> references and leaving only informative examples, we can find a consensus
> that will see us through until the WG has consensus around
> modularisation/extension handling etc. (and enough active contributors to
> make it happen).
> >
> > Léonie.
> >
> >
> > --
> > @LeonieWatson tink.uk Carpe diem
> >
>
> Dave Singer
>
> singer@mac.com
>
>
>
Received on Friday, 12 August 2016 18:32:13 UTC

This archive was generated by hypermail 2.3.1 : Friday, 12 August 2016 18:32:14 UTC