Re: abstracting markup

carifio@black-widow.mro.dec.com
Mon, 12 Dec 94 14:34:26 -0500


Ben Franklin once said something to the effect that its
better to keep your mouth shut and be thought a fool than
to open it and remove all doubt. I guess I'm about to
demonstrate that I'm doubly foolish.

>
> Overloading is a mistake just as much as underloading. What you want is the
> correct mix that minmizes both the number of tags and the number of
> parameters.

Yes. A good insight.

>
> Of course, if you work this one out, you have a general theory of abstraction
> -- which is probably more than anyone else has.
>

I think you could probably make good abstraction decisions without
a full blown theory.

When I look at a list, I see an ordered sequence of items that differ
only in the decoration of the individual items. Perhaps I'm missing
something. Is it intended that <ol> ... </ol> really does differ
from <ul> .. </ul> in terms of how I can rely upon order? Is it just
an artifact of existing browsers that they happen to present the items
in the order of appearance and consequently I'm missing their distinction?

One could envision a situation where the browser may change
the order of the list for unordered list, perhaps alphabetizing by
first character (or something) or perhaps by estimated cost of painting
each list item. If that's what we're anticipating, great. This is
good design, one that anticipates a future need and I'm merely misinformed.

On the other hand, it feels like reaching to me. When you add DIR and MENU,
it feels even more problematic. When does it stop,
<CHRISTMAS_CARD_LIST>, </CHRISTMAS_CARD_LIST>?

Logical encoding is a good idea. So is orthogonality. If different tags
are needed, please clearly state their differences. This thread started
with the observation that there weren't any real distinctions. I
believe that was a good observation. I still do.