W3C home > Mailing lists > Public > public-html@w3.org > February 2008

Re: Emphasizing STRIKE

From: Leif Halvard Silli <lhs@malform.no>
Date: Thu, 07 Feb 2008 04:10:24 +0100
Message-ID: <47AA76A0.7020605@malform.no>
To: gonchuki <gonchuki@gmail.com>
CC: HTMLWG <public-html@w3.org>

<em>Gonzalo</em>,

At 08-02-06 19.47 you replied:   ­
> The point is exactly that. If you just need a visual representation of
> stroked-through text then use CSS.

The point is exactly that the readers, including nonsighted, should be 
informed that the text has been striked over. A STRIKE or S element can 
do exactly that. CSS can't. DEL gives wrong information, becuase it says 
that the text has been deleted - which is not what STRIKE says.

Strike should not be used when the stroken text represents a removal 
from the document. On the contrary, the document might not be valid 
without the stroken text. The stroken text might contain information 
necessary to make the rest of the document valid - for its holder. For 
instance, a typical example could be  that the holder of a document 
changed name or something, and that the old name was striked over.

> Your "that is for the reader to judge" is exactly the contrary to a semantic meaning,
>   

The semantic meaning is: this text has been striked over. How hard is that?

You want to know if the stroke _also_ means that the text has been 
deleted. Well, I said above that that it doesn't. But - of course - it 
could also have that meaning. This requires analysis by the reader. If 
the document is a HTML version of a paper document, and the reader is 
aware of that, then he/she will know how to process that information.

Actually, there is not difference between STRIKE and EM, STRONG, I or B 
in in this way: I am afraid you have to leave the understanding process 
to the reader. He/she must make sense of how the document is formatted.

> the markup must infer meaning to the text and not just visual
> representation.

The mark-up must mark up the text. The text is - virtually - 
«preexisting». You mark it up. If the text contains striked out text, 
then you mark it up with <strike>, so the readers know where striked out 
text begins and ends.

>  If it's a "strike joke" or text in which you don't
> exactly know the reason for being striken, then a span with
> text-decoration: line-through is enough.
>   

Very unhelpful for unsigted users, I am afraid.

> You should treat that "if you have no information about such a thing"
> the same as you would transcribe red and blue text: If there is no
> information on whether blue or red is for emphasizing then leave both
> with default markup and don't force a <strong> or <em>.
>   

The author/translator/mark-up person is obligued to understand the text. 
Texts are littered with e.g italisized and boldend phrasees. You as web 
author must judge _why_  it is emphasized that way. If you should use 
EM, STRONG, B, SPAN or whatever. In fact, often you will be _adding_ 
meaning - the meanng you yourself get from the text - through that process.

HTML has no way of documenting how your mark-up of the text developed. 
It can only document how your _text_ developed.

What I said about "have information about why something is striked out", 
was relating to «document history». If you have a piece of paper, a 
sertificate e.g., then that document is often signed. Should you then 
insert <INS> around the signature, to demonstrate the fact that the 
signature was probably added at a later stadium? Why? This could give 
wrong information. The point is to present the document as a final 
product. (And paper documents are creates through a long proces: 
printing, siging, stamping ..) Not as something to which something was 
added, as if it still is a work in progress, or as if someone added 
things that doesn't belong there.

> > Of course it will. But as explained, the point with STRIKE is not to
> > «clarify the edition process», but to accuratly mark up the phrase
> > structure of a certain text. Without regard to the historical process
> > that text might have gone through.
>
> Your example had an "historical process" behind:
>   

Everything has a historical process behind.  So what? We write documents 
in order to present a final product. We translate documents, from 
language to language, or from paper to web, for the same purpose - the 
final product. DEL and INS are help elements on the path to the final 
product. Nothing more. Nothing less. It is uninteresting _noise_ if you 
deliver it to the final reader, though.

> > > > For instance, to insert a striked out text - that you forgot to notice
> > > > the first time. Without the STRIKE element, we would have to use a
> > > > meaningless DEL inside INS.
>
> it's not a "meaningless <del> inside <ins>" if datetime attributes are
> correctly set, or if you are just inserting stroked text with no real
> knowledge on why, then it's the same red/blue situation as above, use
> CSS since you don't know why the text has been stroked.
>   

I did not say that DEL inside INS was meaningsless in other contexts 
than the one I gave.

It is often very much possible to have an opinion about why something 
has been deemphasized with a stroke, and when, in history, it happened. 
However, the point is - for the kind of documents I have in mind - to 
leave that judgement to the reader. The point is also to separate 
editing notes from the final text. The fact that you cannot - or are not 
allowed to - _date_ the time something was added, is a bad reason 
instead preferring SPAN with CSS over a semantic element such as STRIKE.
-- 
leif halvard silli
Received on Thursday, 7 February 2008 03:10:44 UTC

This archive was generated by hypermail 2.3.1 : Monday, 29 September 2014 09:38:52 UTC