Re: @quotation and @indentedblock in html and DocBook

[Gavin Smith]

On Thu, Dec 18, 2014 at 11:58 PM, Karl Berry <address@hidden> wrote:
>     for example using semantic
>     output could allow better accessibility
>
> No question.  But that's not what's going on here.
> Neither <blockquote> nor css indent is especially semantic.
> If anything, <blockquote> is more so.
>
> Aside: Indeed, one of my recurring problems with CSS is that it is
> routinely used to (try to) force particular fonts and font sizes, which
> only decreases accessibiity for those of us with
> lousy-but-not-nonexistent vision.  Just based on the population at
> large, I venture to say there are considerably more people with
> deficient vision squinting at tiny screens trying to make out tiny fonts
> because of web site CSS than there are people using screen readers.  We
> want to do the best we can for *everyone*, of course ... in my
> experience, overspecification is worse than underspecification.  Anyway,
> this is off the topic at hand.
>
>     screen reading software does in the case of <blockquote>,
>
> In the absence of evidence to the contrary, I presume that screen
> reading software (I've never used such either) does the best that can be
> done with the standard HTML tags, and it would only make things harder
> to use CSS.  Saying "indent 3.2em" is not as meaningful as explicitly
> saying "start indented block".
>
>     It could also allow customization by the
>     user of the browser to add their own stylings to all webpages.
>
> I think/hope that could be done either way.  We want to allow people to
> do whatever they are crazy enough to want to do with CSS.  We just don't
> want to do it ourselves.  IMHO.
>
>     What @example, @display, etc., have in common with @indentedblock is
>     that they should be indented,
>
> Fair enough, but what they don't have in common is that they don't do
> filling.  Right now, the non-filled environments use <pre>+css for
> indent.  That somehow seems better (more semantic!) to me than
> <blockquote>+<br> for every line.
>
> @quotation and @indentedblock somehow seem to naturally map to
> <blockquote>, but not @example.  An example is not, prima facie, a block
> quote at all.
>
> Thanks,
> K
>

I see there is a difference with the preformatted commands like
@display in that if CSS not available, it will still be evident to the
reader that there is something different about the text included,
because of the text not being filled. So I agree that <blockquote> is
probably better than <div>+CSS for @indentedblock. Lynx may give the
text a gaudy colour in addition to indenting it to mark it as a
quotation, but that is better than it not being marked as special at
all.

So in this case slightly breaking the rules of having semantic output
(i.e., using <blockquote> for something that might not strictly be a
"quotation"), makes the output "accessible" to more users.

I will see about implementing this in the HTML output. I haven't
thought about DocBook yet.

Note that although the Texinfo manual says that @indentedblock is just
like @quotation, that shouldn't mean that they have to be expressed in
analogous ways in the output formats, because the semantics are
different.