Re: HTML5's Q element

2009/9/4 Ian Hickson <ian@hixie.ch>
>
> On Wed, 2 Sep 2009, Seth Call wrote:
> >
> > For IE8, FF 3.5, and Opera 10 (although not Safari 4), this behavior of
> > adding quotes is done with the default CSS style sheet of the browser;
> > not in the HTML rendering itself.
> >
> > If you style q like so:
> >
> > q:before {
> > >   content: no-open-quote;
> > > }
> > >
> > > q:after {
> > >   content: no-close-quote;
> > > }
> > >
> >
> >
> > Then the quotes go away in those 3 browsers.
>
> On Wed, 2 Sep 2009, T.J. Crowder wrote:
> >
> > Having the UA add the quotes is not a good idea, and will almost
> > certainly lead to hacks like UAs looking to see if the quotes are
> > already there and only adding them if they're not.
> >
> > Even if IE8, FF3.5, et. al. and such are already doing this, they're
> > implementing the bleeding edge of this spec and should be agile enough
> > to cope with changes to it as it is finalised.
>
> On Wed, 2 Sep 2009, Ryan Roberts wrote:
> >
> > Whether it's no common or not it's a bad idea that needs changing sooner
> > rather than later.
>
> On Thu, 3 Sep 2009, Henri Sivonen wrote:
> >
> > It's a bad idea, sure. But it doesn't follow that <q> in itself is a
> > good idea and needs changing.
> >
> > My conclusion is that adding quotes is a bad idea *and* <q> in itself
> > isn't particularly useful. Thus, the solution is not to use <q> in
> > newly-authored documents.
> >
> > Hixie, maybe <q> should be obsolete but conforming...
>
> On Thu, 3 Sep 2009, Arthur Clifford wrote:
> >
> > I think this comes down to how HTML/HTML5 is going to be utilized, is it
> > describing a document object model; in which case a quote (or quoted
> > text) is an object and proper display of a quote object should be up to
> > the user agent, as informed by style sheets, thus keeping content and
> > its face separate. Or is HTML merely a markup language, in which case
> > what is the philosophical reason for having a q-tag? When is it used and
> > why?
> >
> > Back when there were proof-readers they would mark-up something that
> > should be a quote; isn't that what you're doing in html too? Quoted
> > content needs to be distinguishable from the content around it, which is
> > done with quotation-marks. Shouldn't the marks used in a document for
> > quoted content be identifiable through styles and/or along with a
> > section of quoted text? If so, isn't the quote tag the logical place to
> > define those styles, either in a global style or as a style attribute on
> > the quote tag?
> >
> > The localization argument I thought was a compelling argument for the
> > user agent to handle putting in quotes. It is an example of what is
> > possible when you treat a document as a DOM versus just marking up text.
> > If quoted content is treated like an object then things like
> > localization are easier to facilitate.
> >
> > I would hope that content-editors such as DreamWeaver would help out by
> > indicating usage of " and q together and would somehow notify the user
> > if that is what they want to do. The specification should recommend
> > against an html renderer doing such checks though. I think people would
> > figure out quickly that things are double-quoted and fixing that
> > shouldn't be hard.
>
> If you want quote marks in the source, use quote marks in the source, and
> don't use <q>.
>
> If you want quote marks added automatically, use <q>.
>
> It would be stupid of us to try to change this now given that all four
> major browsers ship with a <q> that inserts quote marks

Strongly disagree.  The spec is *not final* yet.  If vendors implement
in-progress specs, they must expect them to change.  Fix it now and
within a year, it's done, implementations and content will be updated
to match.  To fail to correct this now just sets up ongoing,
unnecessary silliness for years to come.

> This was
> discussed in depth last year, and the spec was changed (from not inserting
> quotes to inserting quotes) after it was concluded that swimming against
> the browser vendors here was futile.
>
> At this point, the <q> element's purpose is to enable CSS-based quotation
> mark injection. If you don't want that, then don't use <q>.
>
> --
> Ian Hickson               U+1047E                )\._.,--....,'``.    fL
> http://ln.hixie.ch/       U+263A                /,   _.. \   _\  ;`._ ,.
> Things that are impossible just take longer.   `._.-(,_..'--(,_..'`-.;.'

--
T.J. Crowder
tj / crowder software / com
www.crowdersoftware.com

Received on Friday, 4 September 2009 21:14:18 UTC