- From: <jens.meiert@erde3.com>
- Date: Mon, 2 Jun 2003 12:18:22 +0200 (MEST)
- To: "Ben Meadowcroft" <cee.plus@virgin.net>, www-html@w3.org
Well, I suppose either to create another element group instead or to solve this for your own via other elements (e.g., related to the type of the supposed visualization, with <dl />, looks nearly the same). In my opinion, you 'abuse' the list elements for such purposes. Regards, Jens Meiert. > Ernest Cline wrote: > > Ben Meadowcroft wrote: > > > >> In a recent post entitled "XHTML 2.0 List Module negates the semantic > >> usefulness of definition lists." I mentioned that the specification > >> lacked a "more general/neutral list element > >> that relates two items together (like a DT DD pair)". The motivation > >> for this was to reduce the need for authors to "abuse" the > >> definition list to model labelled dialogues for example. > >> > >> Having thought about it I would like to propose that the label > >> element introduced with navigation lists be introduced into ordered > >> and unordered lists. The motivation for this is that these more > >> content neutral elements will now be able to utilised instead of > >> wresting the semantics of a definition list. We will now have a > >> proper mechanism for associating labels with normal list items. > >> > >> Use Case: modelling a dialogue. > >> > >> <ol> > >> <label>Dave</label> > >> <li>Hi, how are you doing?</li> > >> <label>Kieran</label> > >> <li>Fine Thanks, and you?</li> > >> <label>Dave</label> > >> <li>Doing well, I got that job I applied for at CERN so I am really > >> pleased.</li> > >> </ol> > >> > >> produces (with suitable css of course) > >> > >> Dave > >> Hi, how are you doing? > >> Kieran > >> Fine Thanks, and you? > >> Dave > >> Doing well, I got that job I applied for at CERN so I am really > >> pleased. > >> > >> > >> So does this idea have legs then? Perhaps a different element name > >> than label(although I think it's quite neat)? > > > > Well first off, while I can see the desirability of having an element > > to indicate a semantically important marker that is associated with a > > list item, it shouldn't be placed outside of the list item or you are > > just getting a <dl> with different element names. > > We agree on the desirability then, a good start! > > I understand your point about placing it inside the list item, however I > disagree that this is the best method. > > To ensure I am understanding you correctly you are envisioning something > along the lines of: > > <ul> > <li><newlabelel>Label for item</newlabelel> Miscellaneous item text</li> > </ul> > > The reason I question this concept is that it seem too similar to what can > be achieved with a <h> to me. For example > > <ul> > <li><section><h>Label for item</h> Miscellaneous item text</section></li> > </ul> > > Here a heading or label is clearly related to a single item. > > What are the merrits of my inital proposal. Well as you stated "you are > just > getting a <dl> with different element names", this is true in a sense, we > would be getting the functionality provided by a <dl> but without > polluting > the semantics of the <dl> element. > > This functionality, the ability to associate a label to multiple items of > data, and vice versa, is a powerful one that should be extended to the > more > generic list elements <ol> and <ul>. An example of this is a list of > slides, > with attached labels. > > <ol> > <label>Introduction</label> > <li>What is RDF</li> > > <label>Practical Applications</label> > <li>RDF in RSS 1.0</li> > <li>RDF in FOAF 0.1</li> > > <label>Implementation Issues</label> > <li>Syntactic issues</li> > <li>Validation</li> > > <label>Conclusion</label> > <label>Future Directions</label> > <li>n-triples</li> > <li>rpv</li> > </ol> > > Determining which labels are associated with which items would follow the > same algorithm used to determine which <dt>s are associated with which > <dd>s. Nesting issues should not be an issue as this would be inherently > supplied by the use of nested lists. > > > In fact since your example is not a list, using an <ol> is just as bad > > as using a <dl>. > > I used an <ol> as the dialogue is a series of phrases given in a specific > ordered. > An <ol> is described as containing "Ordered information", a dialogue can > be > seen as an ordered series of information, wheras seeing it as a series of > terms and definitions is something more of a stretch. > > > Rather, I'd use: > > > > <div class="dialogue"> > > <div> > > <cite>Dave</cite> > > <blockquote>Hi, how are you doing?</blockquote> > > </div><div> > > <cite>Kieram</cite> > > <blockquote> > > Fine. > > Thanks. > > And you? > > </blockquote> > > </div><div> > > <cite>Dave</cite> > > <blockquote> > > Doing well. > > I got that job I applied for at CERN, so I am really pleased. > > </blockquote> > > </div> > > </div> > > Seems like a nice implementation of a dialogue to me. > > > Now if there were a Diologue Module with specific elements for this > > sort of thing, it would be even better, but do dialogues occur often > > enough to justify it? > > Probably not, however my argument was not based on the need to > specifically > model dialogues, it was motivated by the need to provide a more general > mechanism of providing some method of relating a label to a list element > (and thus stopping the abuse of the <dl> element). I propse extending <ol> > and <ul> elements as these are fairly content neutral. > > -- Jens Meiert Steubenstr. 28 D-26123 Oldenburg Telefon +49 (0)441 99 86 147 Telefax +49 (0)89 1488 2325 91 Mobil +49 (0)175 78 4146 5 eMail <jens@meiert.com> Internet <http://meiert.com>
Received on Monday, 2 June 2003 06:18:36 UTC