RE: Bad CSS implementations
> David Siegel [SMTP:email@example.com] wrote:
>I believe part of the problem is that the W3C never had an adoption
>program, never had any plans for verifying implementations,
Have you seen the CSS Test Suite? See the message attached at the end
of this one for Hakon's post to this list about this. It is not a full
validation suite - but it does test a fair bit of the fringes of CSS.
This is part of our regular test pass, so we know exactly how much of
this we pass and how much we fail.
During the end of the specification of CSS1, there was verbiage put into
the spec that basically said, "conformance is determined by running a
test suite provided by the W3C." This was resoundingly rejected by the
working group, as a test suite is a poor substitute for a specification.
I imagine their paws still smarting from this was one reason they have
not been too anxious to develop a real test suite.
>and generally pulled up short on insisting that the browser
>companies build something we can all use.
As someone else pointed out, "insisting" is hardly something they can
do. What can they do? Kick the browser developers out of the
Consortium? Aside from "holding meetings in a telephone booth", this
really wouldn't help the situation, as S.N Brodie described. I would
suggest that the W3C create a CSS-enabled logo program for user agents,
although this is a fairly sizeable task, and of course it may be
difficult to convince the W3C member companies that this will be a good
use of our money.
>I am disappointed to see that the browser companies act like
>children when it comes to incorporating basic typographic control into
I'm disappointed that that is your impression of what we do.
>and still we don't have PNG,
You do in IE4.
>or a widespread vector-graphic standard,
I actually haven't seen this be a huge item in comparison to others on
people's wish list. I would suggest that you, David, as a content
developer, push the W3C to agree on a vector graphics standard as a
baseline for support.
>or a widespread midi implementation,
I'm not sure what you mean. Don't both IE4 and Nav4 (not to mention the
3.0 browsers) implement a midi player? What is the issue here?
>and don't even get me started on the font situation.
Well, okay, since you told me not to. But we (W3C, Microsoft, and other
W3C members, together and separately) are working on fonts - have been
for some time. WD-font is a product of that.
>I think what we can learn from this thread is that the W3C and the
>companies have a lax attitude toward implementation standards.
I'm extremely sorry to hear that, because quite frankly, I do not. I
try to guide our implementation to match the specification exactly.
Does it? No, of course not - there are my own misinterpretations,
unclear portions of the various specifications, specification arguments
that go on too long for us to wait and die down and then rear their ugly
heads just at the last moment when it's too late for us to change, and
more than anything else, there's the reality of a looming shipping
deadline and only so much I can do without forsaking sleep and seeing my
wife altogether. Risking mimicking a much better person than myself, I
have a dream - to finish off implementation of the CSS specification at
some point. There are portions of it that I may never accomplish - for
example, I have yet to justify even to myself why I would waste the time
to implement the 'whitespace' property - but it is a personal goal, and
not one that I have to fight my superiors tooth and nail to receive
approval for, either. At the same time, we have many other goals, and
we're trying to accomplish a lot on those different fronts, too. I
would easily have made the decision to drop a few more CSS features to
in order to expose the CSS properties through the object model, so they
can be dynamically controlled. (In fact, hmm, I did.)
>waiting for press-announcement surprises and "cool" demos of floating
>logos, we should all agree that basic typographic control is a right,
>privelege, and that the browser companies have nothing to gain from
>implementations. The only way out is a reference standard. Only when
>attitudes change will we see fundamental improvement. It's too late for
>4.0 browsers. Let's see what we can do about the 5.0s.
How our marketing department spins what I do is neither under my
control, nor in my realm of concern. I agree - basic typographic
control should be a right, not a privilege. That's why I embarked on
evangelizing stylesheets when we first started design work on IE3, even
though I was new to the team, fairly new to Microsoft and few people on
the IE team (with a few notable exceptions) had heard of stylesheets.
Reference standard? Letters to Microsoft, describing the bits of CSS
you think we've dropped the ball on? Yes, please.
From: Hakon Lie [SMTP:firstname.lastname@example.org]
Sent: Friday, June 13, 1997 5:18 PM
Subject: CSS test pages
A set of CSS test pages  are now available from W3C. The pages are
primarily helpful for people who develop CSS formatters, and they are
not intended as a validation suite. The pages are influenced by
discussions on this list and several people here have contributed.
H å k o n W i u m L i e
email@example.com W o r l d Wide W e b Consortium
inria §°þ#¡ª FRANCE http://www.w3.org/people/howcome