Ok, now I have a quick question for everyone. Several people mention a
desire to minimize the # of tags. In fact in the latest HTML3.0 stuff Dave
Raggett seems to be specifically coming up with as many ways as possible to
do things without additional tags. For example the stuff involving word
wrapping controls which I believe Netscape has implemented using <NOBR> and
<WBR>.
Anyway, there seems to be a general assumption that adding tags or
attributes has some cost? For example, sure, with <C> character formatting,
you need new attributes if you invent some new type of character attributes
you want to let the document creator control. But it seems to me that the
cost is in the features at all rather than whether it adds a tag or an
attribute, or simply a new construct like:
<STYLE>
font.family = hevetica
font.color = #FF0000
font.newthing = something
</STYLE>
In fact, most programs I have seen are really good at ignoring tags and
attributes that they don't recognize. But if we are going to create some way
to do style's that don't involve putting all the data in the header, we
should make sure it gets ignored by programs that don't recognize
it(Something that doesn't usually apply to text thats not inside of a tag).
To sum it up: the added cost appears to me to be in improving HTML at all,
not necessarily whether we let people control text attributes through style
sheets, character formatting, or both. At the risk of making assumptions
based on how my own products support HTML, once I have dealt with the
formatting issues of styles (IE: Something that draws text in colors and
sizes, etc), I can add support for another mechanism to access the same
stuff (For example the C tag) trivially...
Alex Hopmann
ResNova Software, Inc.
hopmann@holonet.net