Re: <q>

2008/10/29 Justin James <j_james@mindspring.com>

> From: public-html-request@w3.org [mailto:public-html-request@w3.org] On
> Behalf Of Sam Kuper
> 2008/10/28 Justin James <j_james@mindspring.com>
> >> I like <q> as a tag to indicate that something is a quote, kind of like
> "blockquote is
> >> to div as q is to span". I don't like it as trying to plaster on
> automagical punctuation
> >> marks.
>
> > Then either don't use <q> in your own markup, or else specify a style
> rule to prevent <q>
> > from making presentational changes that include the insertion of
> punctuation.
>
> And lose the benefits of semantics within my pages?


If you follow the second suggestion, you won't lose the benefits of the
semantics <q> usage brings.


> HTML doesn't define a default style for many elements at all. Why are you
> insisting that one be defined for <q>?


Partly because I'm used to it from HTML 4.01; partly because section 9 of
HTML 5 is obviously incomplete, leading me to think that HTML 5 may yet
recommend default presentation for many elements.


> > Essentially, the proposals I've made are like "blockquote is to div as q
> is to span". It
> > is only because standard typographical style in English and many other
> widely used
> > languages is not to append/prepend punctuation to block quotes that I
> (and presumably
> > others) have not suggested that <blockquote> should do so by default.
> However, because
> > standard typographical style in many languages is to append/prepend
> quotation marks to
> > inline quotations, I think the <quotation> element should do so.
>
> OK, then for the sake of consistency, I insist that you insist that <p>
> require proper capitalization and punctuation.


I don't see how that would be consistent, given that I have been making
suggestions about <q> and not about <p>. In other words, the scope of my
suggestions has been the <q> element, not all HTML 5 elements, nor even all
HTML 5 elements containing phrasing
content<http://www.w3.org/html/wg/html5/#phrasing-content>
.

I think each element should be addressed on its own merits to begin with. If,
after that has been finished (and I don't think it has been finished yet),
proposals for a number of elements are similar/identical, then the
consistency between those elements be formalised.

Even if the community were to agree with some or all of my suggestions for
<q>, it would not necessarily reach consensus on similar proposals for <p>
(and, as far as I know, none have seriously been proposed). Therefore, the
issue of consistency between <q> and <p> in that regard is moot.


> >> What if the inner contents come from elsewhere, like an authoring tool
> or are pulled
> >> from XML and inserted via JavaScript?
> >
> > I can't see how this is a problem, as long as the inner contents are
> decently marked up
> > (which they should be anyway).
>
> You are kidding, right? It is *extraordinarily* unlikely that, say,
> user-generated content will be marked up with <q>.


I said, "decently marked-up", I didn't say, "marked up with <q>". So what's
the problem?


> Furthermore, to think that external content or user generated content will
> be well marked up is absolutely incorrect.


Depends who has authored it and how. I'm all for having authoring tools make
it easier for users/authors to author good HTML, and increasingly they are
doing so (cf. Fred Knabben's efforts and many others). Over the lifetime of
HTML 5, I hope the quality of authored HTML will increase as a result.

In fact, among forum software, [quote] is probably one of the most used
elements. It would, I think, be easier for phpBB et al to become HTML5
compliant wrt. the <q> element based on my proposals (and all equivalent
proposals made by others) than otherwise.


> > I think I've addressed all of your stated objections to <q>. Please
> reconsider whether
> > <q> is really such a bad idea after all.
>
> I think that <q> is a good idea. I think that demanding a particular
> styling of it which generates quotation marks "on the fly" is a very bad
> one. What about authors who put in quotation marks and then mark it up with
> <q>?


Those authors are duplicating their own efforts. Why mark up twice what you
can mark up once?

For authors with that much time on their hands, it's reasonable to require
them to specify in their stylesheets that <q> should not render quotation
marks.


> Pretend that they are using an authoring tool that lets them indicate
> quotes... say, something like Word which lets them indicate that something
> is a quote and make a citation for it.


Word is the worst HTML authoring tool I have ever used in my life, but I'll
answer your question...


> Do you want the authoring tool to walk them through a wizard or popup,
> interrogating them if the quotes are part of the quotation, or if they
> merely delimit the quotation?


Lots of text editors (Vim, Emacs, PSPad, ...) are capable of parsing
expressions delimited by quotation marks. I'm pretty sure Word can/could do
this too. This, combined with Word's ability to apply styles, and to store
settings, opens up a number of possible workflows:

   - Have styles for quotes, just as Word already has styles for lists, etc.
   So the user types a quote, highlights it, and selects the "Quote" style,
   which would apply, say, double quotes by default if the language is English.
   If the user later wants all quotes to be single-quoted and italic, they just
   have to change the style, and it's done.
   - Have an option (de-selectable via the settings menu, like smart quotes
   are already) for Word to automatically detect quotes. That is, when the user
   types a quotation mark, Word begins parsing the following text to see if it
   gets terminated with a quotation mark in a way that suggests, based on
   Word's grammatical knowledge, that it was a quote. If so, Word automatically
   applies the "Quote" style to it, so the quotation marks now become part of
   the style.
   - Permit users to modify single instances of "Quote" styled text, by
   manually changing double quotes to single quotes, etc. (just as they can
   already modify single instances of text with other styles to produce "Body
   text + indent" or "List + italic", etc.; the result would be "Quote + single
   quoted" or some such).



> I repeat:
> Automatically generation punctuation is fraught with danger.


I repeat:

I think I've addressed all your concerns :)

Regards,

Sam