> Rob Raisch writes:
> >I would like to make a strong request for the ability of the
> publisher to specify that the stylesheet they provide should
> not/must not be overridden in the browser.
>
> Absolutely. But the inverse may obtain: some clients will give
...
> So would it make sense to allow the publisher also to specify that
> the doc is to be rendered with the published style sheet or not
> at all?
OK, then give me the receiver the ability to specity that any such
documents NEVER be presented to me. Just like I exercise the choice
to not subscribe I would want some electronic filters to bypass material
I can't deal with.
Lets face it folks ....
1. No matter what the HTML standard might specify, there is NO WAY to
prevent someone from writing a browser which ignores the 'rules'.
2. There is just a small chance that I may be able to improve the
rendering in my environment -- in the publishing world there isn't
too much that can go wrong after the material leaves the printer
while in our world much that can go wrong happens in that phase.
3. Most users won't bother to tamper with the publisher's efforts so
from a mass media perspective the masses will get it however it
turns out.
The best results are not likely to occur if either the publisher or
the user is given the heavy hand. As a standards group the challenge
is to anticipate problem areas and formulate approaches to allow the
author and reader to each contribute their best efforts to the end
result.
The earlier proposal (sorry I forget whose) seems to have some merit though
the complexity concerns me. The underlying notion of establishing some
orthogonal dimensions to the problem which can be merged may also
provide the mental paradigm to help tame the complexity.
In the HTTP domain there is much discussion of negotiation between
client and server. I think good solutions in the rendering arena
(and other browsers too <-:) will focus on defining the negotiation
process. I think inheritance is too unidirectional to satisfy anyone.
>From both the authoring and the reading perspective, I would be concerned
about organizational styles providing anything but base line defaults.
(It should be easy for individual users to share their style efforts
but I thinks that is a proactive adoption process.)
It would seem reasonable for a publisher (and reader) to each provide
recommended and mandatory choices. Weighting on the recommended might
be good. A conflict on the mandatory choices is conceptually no
different than any other compatibility issue which makes information
useless.
The next issue then could well be how to conduct this style negotiation
without transfering the whole document. In today's world, it is rather
annoying that some browsers will retrieve a file (audio, video)
without warning me first that I have no way specified to use the data.
It could get tedious for people with unusual rendering restrictions if
they frequently encounter documents then can't cope with.
Dave Morris