W3C home > Mailing lists > Public > public-html@w3.org > January 2011

Re: hgroup: a possible alternative - or drop completely

From: Lars Gunther <gunther@keryx.se>
Date: Fri, 21 Jan 2011 01:20:11 +0100
Message-ID: <4D38D13B.5060409@keryx.se>
To: HTML WG LIST <public-html@w3.org>
2011-01-11 15:32, Toby Inkster skrev:
> On Mon, 2011-01-10 at 23:46 +0100, James Graham wrote:
>> Why not<h1>Main Heading<subhead>Subheading</subhead></h1>
> Yes, or that.
> Either way, the key point is to flip the<hgroup>  inside out, so rather
> than putting stuff outside the<h1>  element, you put stuff inside it.

Thus we seem to have 5 major proposals at least. I will rank them 
according to my estimation how easy they would be to teach, and thus 
where I would expect the least abuse or misunderstanding.

1. Drop hgroup, but have an sh child as a child within hn

2. Keep hgroup, but allow only one hn + sh
2b. Allow multiple sh

3. Drop hgroup. Have nothing instead. Await a microformat to emerge and 
prove its worth.
http://www.w3.org/Bugs/Public/show_bug.cgi?id=11828 (Filed by me)

4. Keep hgroup, using hn and hn+1 as children. The current proposal in 
the spec.

5. Drop hgroup, introduce sh to be put directly beneath an hn

I will now try to give a perspective from a *teachability* POV, since I 
teach this stuff for a living. Based on my experience, what concept 
would be easiest to learn?


Argument in favour of: sh is part of the heading, but plays a lesser 
role. That would be the normal understanding.

BUT: What if the tag line is relatively disconnected? In that case I 
would probably argue that putting it in a p-element below is just fine.

(BTW, I had a student show me some markup this very week that was a 
heading and a subtitle. She had marked them up h1+h2. I've not taught 
them hgroup, so she was reliying on HTML 4 semantics, and this seemed 
most intuitive for her, until I asked what role the subtitle really 
played. When asked about what such markup should look like in if it 
existed, she indicated that it would probably make most sense as a sub 
element and I did not even suggest such a structure.)

(Implementation perspective. Drop one element, add one element. No 
browser has yet implemented hgroup in any other way but add some defualt 
styling and bragging about it on their web sites. This is a low cost 

2. As I've said previously, having an element alter the semantics of 
another element is bad for learners. By not allowing more than one hn 
element per hgroup this is avoided.

By keeping hgroup we make the relation explicit and we will avoid the 
abuse of seeing disconnected sh elements sprinkled all over the 
structure. That's the reason I've put idea no 5 as the worst one.

2b. Allowing multiple sh-children OTOH will also be a nightmare. There 
simply are no valid use cases for sub-sub-titles worthy of adding this 
potential confusion. Let's cover the 90 % and forget about the other 10. 
Their cost will be way too high.

Eventually we will see hgroups containing long essays, if we do not 
clearly specify that its should be one succinct subtitle only.

Thus I emphatically say no to sh1, sh2, etc, or to multiple sh within 
one hgroup. KISS!

(+ That would really have a high implementation cost. How would you map 
that to AT?)

3. Delay the teaching to until one speaks about common patterns  and 
microformats. Students will by know be more knowledgeable.

Introducing point 2 above I've argued against 4.

Now taking off my teachers hat.

I am generally strong in my support of adding semantic capabilities to 
HTML, but since there are so many different suggestions and no real 
concensus, we must really ask ourselves this:

Is the lack of marking up subtitles/taglines in HTML 4 such a big 
problem, that we should at this stage introduce a solution - any 
solution - that clearly is not really liked?

The decision that will be taken now will live on for many, many years to 
come. Thus my suggestion really is 3. Drop hgroup altogether and await 
patterns to emerge, research those and revisit this for HTML5.1 (or 
whatever the W3C will call the next frozen version.)

Thus I've filed a bug for this.

If there indeed shall be a subtitling/tagline element or construct in 
HTML5 at this moment. I'd cast my vote for suggestion no 1. But we are 
not really voting, are we?

Lars Gunther
Received on Friday, 21 January 2011 00:20:46 UTC

This archive was generated by hypermail 2.4.0 : Saturday, 9 October 2021 18:45:30 UTC