Date: Mon, 22 Sep 1997 14:21:37 +0200 (MET DST) From: =?iso-8859-1?Q?Martin_J=2E_D=FCrst?= <email@example.com> To: Jukka Korpela <firstname.lastname@example.org> cc: email@example.com In-Reply-To: <Pine.OSF.3.96.970922142832.15416Afirstname.lastname@example.org> Message-ID: <Pine.SUN.3.96.970922141538.361o-100000@enoshima> Subject: Re: Foreign Words and Phrases On Mon, 22 Sep 1997, Jukka Korpela wrote: > On Mon, 22 Sep 1997, Markku Savela wrote: > > Can 'class' imply attributes? Might be nice if 'species' could imply > > lang=latin so that I wouldn't need to repeat that for each element? > > (e.g. language inherited from the class too?). > > No. At least hopefully not. :-) It would be grossly illogical for > something constructed for affecting presentation to imply attributes > which are not presentational only. The lang attribute has many > potential uses. Only a few of them are presentational. I don't expect > the lang attribute have any effect on _visible_ presentation in most > cases. In _audible_ presentation it would be crucial, of course. There is some kind of misunderstanding here. CLASS in no way is "constructed for affecting presentation". Ideally, as above with "SPECIES" or "TAXON", it is purely semantical. It is a way to make finer semantic distinctions when the elements of HTML are not enough. Of course, CLASS can be misused for purely presentational purposes in the same way this can be done for element names. > Naturally, if defined as an element, TAXON could have any properties > that correspond to its meaning. It could have lang="la" as default > attribute (or perhaps lang="la-neo"), or even as an unoverridable > attribute (although you would not be able to say it SGML). This is not very feasible. It would mean that every browser has to know about TAXON and this default value. And I don't know how many other kinds of things exist where the connection between semantics and language is that tight on a rather worldwide scale. Regards, Martin.