W3C home > Mailing lists > Public > public-html@w3.org > January 2010

Re: ISSUE-95 hidden - Chairs Solicit Proposals

From: Tab Atkins Jr. <jackalmage@gmail.com>
Date: Fri, 29 Jan 2010 14:03:54 -0600
Message-ID: <dd0fbad1001291203w4601f20bq96c9e37e36621b29@mail.gmail.com>
To: Leif Halvard Silli <xn--mlform-iua@xn--mlform-iua.no>
Cc: Joe D Williams <joedwil@earthlink.net>, Jonas Sicking <jonas@sicking.cc>, public-html@w3.org
On Fri, Jan 29, 2010 at 1:47 PM, Leif Halvard Silli
<xn--mlform-iua@xn--mlform-iua.no> wrote:
> Tab Atkins Jr., Fri, 29 Jan 2010 13:03:36 -0600:
>> I think @declare is a completely opaque name in this context.  I've
>> never heard of the @declare attribute before,
>
> If you haven't *heard* about it, then you haven't read whether HTML4
> nor the HTML5 spec draft.

The vast majority of authors have read neither.

I've read most of each spec, but not the sections about <object>
because I've never had to use an <object> in my entire career.  It's
entirely irrelevant to me, and so until it becomes relevant it's a
non-issue.  The tutorials or pieces of example code I've seen that
feature <object> have also never used the @declare attribute.

>> so I don't associate any
>> meaning with it, and it's about as useful as saying that we should
>> change @hidden to @foobar.
>
> We cannot build anything on the fact that you have not read HTML4 and
> HTML5.

You certainly can!  I've never heard of it.  I suspect that most
authors have never heard of it.  Thus, assuming that it holds some
relevant meaning to most authors that can be extended to cover other
situations is likely incorrect.  @declare may or may not be a good
name on its own, but its history on <object> is pretty much entirely
irrelevant to determining whether people will understand what it
means.

> (1) Regarding "legacy meaning", then regardless of whether an author
> has heard about it today, @declare exists in HTML4, and HTML4 is
> available online and described in thousands of tutorials - for example
> w3schools.com [1]. By replacing @hidden by @declare, we make HTML5
> smaller - since it removes @declare from the obsolete features sections.

That's basically zero benefit.  In the end it's probably negative
benefit, because <object> has a special (legacy) meaning for @declare
that other elements won't share.  If @declare on object is changed to
mean the same thing as @declare on other elements, then there will be
tutorials on the web talking about the old sense which may be
confusing.

> (2) Regarding "decipher what it does", then I claim that "the vast
> majority of authors are similar to me" in finding that "hidden" does
> make anyone decipher _correctly_ what it means. It is a _feature_ that
> authors has to make some effort in order to understand what it does and
> means.

I think that @hidden is a much better description than @declare.  What
are you declaring?  The possibility of confusion with @hidden has been
documented, but the error there is relatively small - some people may
think it hides the element only from visual UAs.  That's a problem,
but at least their thinking is on the right track.

> Thirdly, I think you should consider that we are debating a feature
> that is proposed to be deleted and that it is could be useful to
> consider if @declare would have a higher chance of becoming a amicable
> solution.

The issue concerning the removal of @hidden has nothing to do with the
name, though.  It's a philosophical objection.

~TJ
Received on Friday, 29 January 2010 20:04:48 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Wednesday, 9 May 2012 00:17:00 GMT