Hi Peter,

Peter Boot wrote:
> Hi Martin,
> Martin Holmes schreef:
>> I'm just thinking of rationalizing this so that my markup folks can
>> be told that any textual feature they need to describe can be 
>> described using CSS in @rend (no need to learn or look up 
>> project-specific usages such as rend="italic"). CSS is a broadly
>> useful language that they'll benefit from learning anyway.
> I'm not sure about the wisdom of the CSS-only approach, for a number of 
> reasons:
> - CSS is, it would seem to me, too complicated. This will depend on who
> does the XML encoding. If that is done by technically oriented persons,
> learning CSS does them no harm, as you say. But if the encoding is done
> by persons with a scholarly background, CSS may be one hurdle too much.
> - Even though in practice at the moment I only create HTML output, we
> use XML because, among other things, we want to maintain a sharp
> distinction between describing the source and defining possible output
> formats. CSS seems targeted at (HTML) output.
> - CSS may be too precise in some cases: if for instance a paragraph
> needs some extra white around it, I can invent a keyword to use with
> rend and write <p rend="extrapadding">; I wouldn't want to commit to a
> specific amount of padding in the XML source.

Actually, I would -- either in the form of ems (which are relatively 
easy to measure on a digital image, even if you don't know the exact 
scale of it), or cm (if you know the scale of your image), or percent. 
In this project, though, it's mostly typographical stuff -- 
superscripts, subscripts, italics, drop-caps, etc. -- and it's as easy 
for them to learn "font-style: italic" as it is for them to remember 
whether it's "italic" or "italics" or whatever.

CSS seems to me to be one of the greatest successes of the last few 
years in computing. It's easy to understand, easy to write (even for 
relative beginners), highly precise when necessary, and very flexible 
and detailed. You never hear anyone complaining that it lacks features 
(although we all complain that browsers haven't implemented enough of it 
yet), and it has much broader application than HTML -- think XSL:FO, for 
one thing.

I'm not talking here about the implications of cascading, on particular 
document structures; I'm talking about the syntax, keywords, units and 
so on. I don't really know what rend="extrapadding" means; but I know 
exactly what rend="padding: 0.2cm 0.5cm 0.3cm 0.1cm" means. CSS seems to 
me to be excellent for "describing the source" when the source in 
question is a printed book (which is the case in my project).


Martin Holmes
University of Victoria Humanities Computing and Media Centre
([log in to unmask])
Half-Baked Software, Inc.
([log in to unmask])
[log in to unmask]