W3C home > Mailing lists > Public > public-digipub-ig@w3.org > April 2016

Re: Wiki summary of q element default styling issues

From: Asmus Freytag (c) <asmusf@ix.netcom.com>
Date: Mon, 25 Apr 2016 08:53:15 -0700
To: John Cowan <cowan@mercury.ccil.org>, ishida@w3.org
Cc: www International <www-international@w3.org>, W3C Digital Publishing IG <public-digipub-ig@w3.org>
Message-ID: <5664ad1d-99ff-a87d-a3b0-9b9c6a357870@ix.netcom.com>
I think John sums it up nicely.

You have a feature that

a) provides content, but (by default) not style
b) can't be controlled with CSS
c) doesn't (can't) do what people expect

The existing default "works" to some degree in monolingual environments, 
where there's limited nesting, but it is not robust.

In multilingual environments, or with deep nesting, it's difficult to 
impossible to specify with CSS what should happen, because the most 
common scenarios is to base the choice of quotation marks on the 
surrounding language, not the language of the mark itself. And, more 
precisely, on the outermost language (that of whatever unquoted text 
surrounds the outermost quote).

The use of quotation marks is not only language dependent, but some 
languages allow alternate styles of quotation mark usage (or alternate 
styles are used in common practice). Some material uses alternate styles 
of quoting for effect, including some use of "foreign" marks to make a 
quote look foreign.

Some quotations are indicated with leading marks (dash) only. For 
example in rendering dialog.

Some users may have the expectation that the <q> element handles not 
just quotation marks but somehow handles all aspects of the insertion 
(embedding) of the quoted element. While it might be possible to use it 
to isolate quoted material from surrounding bidirectional text, the <q> 
element cannot handle automatic suppression of redundant punctuation. (I 
think rather than removing the period from the examples, they deserve 
their own discussion).

The <q> element cannot reliably function as a container.

As a result, I wonder if it wouldn't be best to
a) suggest that the default behavior of the <q> element be limited to 
monolingual use, and limited nesting (most implementations seem to do that)
b) suggest that that limitation be explicitly documented
c) suggest that any robust multilingual or deeply nested use would use a 
setting of "none" for the automatic content, and
d) suggest that users (authors) supply their own marks

A./


On 4/25/2016 6:35 AM, John Cowan wrote:
> ishida@w3.org scripsit:
>
>> For example, thinking about
>> blockquotes or figures, I don't choose to use blockquote or figure
>> elements only when i think there's a chance that someone will try to
>> harvest blockquotes or figures, i use the elements because they
>> describe the content, and also because they typically come with some
>> minimally useful default rendering.
> Sure, a default *style*.  That's a different story from default *content*,
> which is what q elements try to provide.
>
>> I don't understand this.  Why would i need to add a class to every q
>> element?  Surely you'd only need to introduce a class for quotations
>> when you don't want to follow the default - and if you're using CSS
>> properly, often you don't even need class names then either, since
>> the selectors can be written to understand the context in which a q
>> element sits.
> Is CSS capable of specifying alternation of marks in nested quotations
> to arbitrary depths?  My understanding is that it is not.
>
>> The specific issue that concerns me in this thread is how to ensure
>> that any fallback default styling best represents what the majority
>> of people would expect to see.
> Is there in fact such a majority view?  It doesn't seem so.
>
Received on Monday, 25 April 2016 15:53:40 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 25 April 2017 10:44:42 UTC