Sebastian Rahtz wrote:
> Martin Holmes wrote:
>> There are some things which couldn't easily be handled in a WYSIWYG
>> manner. For instance, this structure:
>> <expan>United Nations</expan>
>> shows up with both UN and United Nations visible, with tag start and
>> end symbols in the right places. That's probably what you want when
>> editing, but it's not WYSIWYG in the sense that it probably won't be
>> what you want in any kind of output.
> I assume you are using the CSS supplied by oXygen for TEI; that's
> as part of the TEI Sourceforge project. If you can come up a decent CSS
> for <choice>, please send it in!
My point is that if there was a "WYSIWYG" rendering that made sense, it
would make one or the other element invisible. I'm not complaining about
the CSS; I'm just making the point that "WYSIWYG" is not WYSYWIM (what
you say is what you mean).
>> I suspect that people coming to XML editing for the first time might
>> find this seductive, but that they'll get into trouble fairly quickly
>> (not bothering to distinguish between <emph>, <hi rend="italic"> or
>> <foreign>, because they "look the same").
> we can use colors or whatever to make them look different...
Certainly, but again, it's not WYSIWYG. I think when novice editors
crave WYSIWYG, what they crave is a view of the document as it will
appear when it's published, completed, rendered or whatever. That's a
forlorn hope, because if we're doing anything useful with our encoded
docs at all, we're hiding some bits of them (and showing them
scriptically when you click on other bits), we're suppressing some
completely (e.g. elements of the teiHeader), and we're rendering many
diverse elements in the same manner (<hi rend="italic">, <emph>, etc.).
Thus if we did have a true "WYSIWYG" view, much would be invisible, some
items would be indistinguishable, and some items would require script to
get at them at all. This is the basis of my deep suspicion of the whole
notion of WYSIWYG.
On the other hand, if the objective is to provide a rich CSS
implementation that renders a document in such a way that all these
elements are simultaneously visible and distinguishable, and yet the
interface and our interaction with the document are simpler and more
efficient than tag-editing, then great -- I'd love to see that. But it's
University of Victoria Humanities Computing and Media Centre
([log in to unmask])
Half-Baked Software, Inc.
([log in to unmask])
[log in to unmask]