- From: John Lewis <lewi0371@mrs.umn.edu>
- Date: Fri, 20 Dec 2002 16:34:19 -0600
- To: www-style@w3.org
Mikko wrote on Wednesday, December 18, 2002 at 7:24:32 AM:
> Shelby Moore wrote:
>
>> No I think we were referring to a new EOS character, e.g. &eos;
>
> If we start putting extra characters to point out where sentence
> borders are, why not just markup it correctly? For example: <s>This
> is a sentence.</s> <s>This is an another sentence.</s>
In my response, I assume you're talking about adding a sentence
element to XHTML.
1. There is very little (if any) need to mark up sentences in XHTML,
except for styling purposes. Should we add a firstline and
firstletter element to XHTML? I don't think so, because they have
very little value to a UA except for styling purposes.
If there was a great need of a sentence element for styling, we'd
see people marking up sentences with span elements and styling
them--and I'm not aware of that being at all common.
2. An end of sentence character is easier for an author to type. He
need not worry about nesting or anything related to markup. He just
needs to insert one at the end of every sentence, which isn't
easy--but is easier than marking up every sentence.
3. I think the end of a sentence should be content. This has
advantages.
1. A text editor could recognize the end of sentence character and
present it in the document, in a variety of ways, with a minimum
of fuss. A text editor cannot recognize XHTML elements, and thus
the sentence element is useless and cumbersome when editing
text.
2. With every sentence marked up, XHTML source becomes harder to
read. It's overkill for a generic language. I think the same
thing about marking up punctuation or verbs or nouns, for
example. XHTML just isn't the language for it.
4. I think authors aren't likely to utilize the end of sentence
character, but I think authors are many times *less* likely to
utilize a sentence element. On the other hand, I doubt authoring
software will utilize either anytime soon, unless entered by hand.
Besides, if authoring software can recognize sentences, why not
shift the burden to the UA (away from the author or authoring
software)?
I agree neither solution is good, but I think an end of sentence
character is better. The ideal solution would be a UA that recognizes
sentences without "extra" markup or content, but I'm not at all
confident of that becoming a reality in the short run. Let's hope I'm
wrong. :)
--
John
Received on Friday, 20 December 2002 17:34:37 UTC