- From: Jonathan Borden <jonathan@openhealth.org>
- Date: Mon, 19 Aug 2002 14:54:56 -0400
- To: <kynn@idyllmtn.com>
- Cc: <www-tag@w3.org>
<kynn@idyllmtn.com> wrote: > Jonathan wrote: > > <statement><quote><!CDATA![[ > > <kynn@idyllmtn.com> wrote: > > > > > > Why? Because arbitrary XML is, to the UA, worse than simply plain text. > > > It provides an unknown structure but the user -- especially the user with > > > special needs -- likely has no way of making sense of that structure in > > > any rational way. > > ]]></quote> > > <there/><is/><no/><reason/><to/><be/><so/><dogmatic/> > > <nor/><can/><you/><so/><easily/><predict/><the/><intelligence/><or/><lack/>< > > thereof/><of/><your/><readers/> > > Jonathan > > </statement> > > That's very cute, Jonathan, but you can't assume that people with > disabilities are going to rely on reading source code, plus it > reflects a basic misunderstanding of XML concepts if you expect all > meaning to be inferrable purely from the names of elements. > Sorry, you have me entirely unconvinced that: a) so-called 'generic XML' has any particular issues w.r.t. human readers with special needs. b) global decisions regarding serving generic XML ought only defer to any special requirements of humans (of any type). Moreover, I've never claimed that one can infer _all meaning_ purely from the names of elements. Rather, I was suggesting, or my example was meant to be an example of the fact that one might _sometimes_ infer _something_ from some arbitrary XML. You claim that 'arbitrary XML' is 'worse than plain text'. I take exception to this. Certainly garbage in-garbage out, is true for plain text as well as XML, but for well written XML, particularly when XML Namespaces are properly used, and the URIrefs that are the namespace names properly reference machine and human readable documents, UAs _might actually_ make good sense of so-called 'arbitrary XML'. My view of what we ought evolve the Web into includes more precise semantics and discoverable documentation, not something that HTML or any subsequent variant, _itself_ is likely to provide. If it is true that generic XML poses problems for humans with specific needs, then what we need are better UAs, perhaps not those primarily designed for HTML etc. -- but I really think that's not at all the point, because, for example, my mother would likely have equal success trying to make sense of some arbitrary XML as she would trying to parse HTML in a text editor -- at the end of the day, I suspect that your entire point boils down to the fact that HTML browsers have been designed as special UAs which interface humans of all sorts to HTML. Similarly I expect that specialized software programs will act as their own UAs, or share specialized UAs. Jonathan
Received on Monday, 19 August 2002 14:50:26 UTC