RE: [CSS3] Generated content: environment variables

Hi Boris,

Boris said:
> Melinda said:
> > Well, we certainly look forward to your detailed proposal. ;-)
>
> The mail you replied to explicitly said that I don't see a 
> good solution to this, right?  Other than not implementing 
> this module of CSS, of course...  Or rather making the 
> user-specified headers/footers !imporant, which has the same effect.

Sorry, I misunderstood. I thought you were advocating standardizing some
kind of complicated layering or shuffling of ua/author/user content in
margin boxes.

I think it's great that we all care so deeply that the user gets what he
or she wants. And I think the passion expressed in this thread comes
from that caring...

As a user, my desired use model is:

1. For the usual case, let the author control what goes in these regions
of the page; this provides me with information not available through the
usual browser UI for headers/footers (e.g., section numbers, the name of
the collection I'm printing or the model airplane shown on the page, the
date of publication... Hopefully the information most pertinent to a
particular document.)
2. If I'm unhappy with the result, and for pages for which certain
running content is important to me but not provided by the author, I
will override the author and hope I can find a way to put what I want in
the margin boxes through the browser UI. In that case, it would be great
to be able to use whichever of the 16 possible margin boxes I choose. I
don't want those overrides to be sticky, because my selections are
document-specific.
 
But your needs as a user are obviously different.

Seems like a good time to do some UI mock-ups and real usability testing
with representative users.

I'm guessing the answer that would emerge is to allow the full cascade
model: provide ua defaults which are replaced by any author content,
which can in turn be overridden by !important user selections.  Provide
a switch to either remember the overrides or not, and present the
opportunity to change along with a print preview. This would enable both
my preferred model of use and yours, if I understand it correctly,
pretty simply.

But that's just my guess, and I think our individual opinions shouldn't
count for much here: we need systematically collected user data to get
to the best answer. ;-)

Best wishes,

Melinda

Received on Thursday, 15 November 2007 20:06:42 UTC