W3C home > Mailing lists > Public > www-international@w3.org > January to March 2009

Re: Armenian numbering: findings, recommendations and request to CSS

From: Leif Halvard Silli <lhs@malform.no>
Date: Thu, 19 Feb 2009 11:46:33 +0100
Message-ID: <499D3889.7090006@malform.no>
To: Christophe Strobbe <christophe.strobbe@esat.kuleuven.be>
CC: www-international@w3.org
Christophe Strobbe 2009-02-19 10.19:
> Hi Martin,
> At 02:33 18/02/2009, Martin Duerst wrote:
>> At 20:05 09/02/17, Christophe Strobbe wrote:
>> >Is it really a good idea to generate heading numbers through CSS? 
>> It's very convenient for authors, but heading numbers are not really a 
>> styling aspect (as opposed to content).
>> What kind of numbering you use is very much a styling aspect.
> *What kind* of numbering may be a styling aspect, but the presence or 
> absence of numbering isn't. That's what I was really after.

Well, the kind of numbering can be really crucial. The *most* 
important thing is in theory that the numbering (re)starts at the 
point the author specifies. However, if a book has page 1 twice, 
then it got a problem ...

If we look at law texts published on the web, they will usually (I 
think) not rely on HTML or CSS list numbering, but rather type 
them directly. Why? Because the numbering is crucial for the 
identification of sections and law paragraphs. On cannot leave 
this over to what possibly could go wrong in a UA.

But perhaps more important: When a web page relies on HTML or CSS 
list numbers, the numbers typically disappaer if one copies and 
paste them (e.g. into e-mail)  Also, one cannot, in most UAs, 
search for a HTML or CSS number either, regardless of whether one 
use <ol type="some-list-style-type"> or ol{list-style-type:..}.  A 
law text that is impossible to navigate is useless.

So it is far from only a screen reader problem. It isn't even only 
a CSS problem, since <ol type="some-value"> got the same problems.
leif halvard silli
Received on Thursday, 19 February 2009 10:47:14 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 21 September 2016 22:37:30 UTC