Lists Home |
Date Index |
Joe English wrote:
> I think omissible *end* tags are very useful; those and
> NETs are the two features I miss most when writing XML.
> It's omissible *start* tags that I find objectionable.
> They can save a little bit of work when marking up
> preexisting text, but cause headaches later on in the
> The main problem is that you have to make sure that contextually
> required elements *stay* contextually required when the DTD is
> modified. Everytime I've used start-tag omission in a DTD it's
> come back to bite me.
I use it mainly during conversion; in that context it comes down to a
data management issue. Conversion is often best served by running the
data through a couple of stages, validating and normalising as soon as
possible. There may be a couple of DTDs, each only used for one stage
and only subtly different. Once you normalise, the tool has served its
> For me, the first form isn't _that_ much better than
> <title>Laundry basket
Yet you say you would like to omit end tags, providing the same saving
of inserting one tag. Every unecessary tag is an overhead, and I've yet
to meet anyone who doesn't find it frustrating that over and over again,
they're telling the processor something that it can determine without help.
> to be worth the trouble anymore. (I have used this kind of
> thing in the past; IIRC what came back to bite me in this case
> had something to do with short references, and the inscrutable
> error messages you get if you leave the title out...)
Usability should be a factor in determining tag omission, so under the
circumstances that you describe, I'd likely make the start tags required
as well. The fact that they're not ideal everywhere doesn't mean that
they're not ideal anywhere though - for conversion, they often are ideal.
Marcus Carr email: email@example.com
Allette Systems (Australia) www: http://www.allette.com.au
"Everything should be made as simple as possible, but not simpler."