cocoon-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From (Carlos)
Subject Re: [OT] Design Rant
Date Thu, 01 Jan 1970 00:00:00 GMT
on 12/19/01 6:19 AM, Stefano Mazzocchi at wrote:

> Copied to general@ since this is a general discussion.
> Ugo Cei wrote:
>> giacomo wrote:
>>> I know :) but many sites only use *one* table to achieve the above which
>>> (at least for older browsers) result in a need to have the hole page
>>> downloaded prior to have it displayed in the browser. This above layout
>>> can display the header as soon as it is available in the browser. This
>>> way you don't have to wait in front of a blank screen too long.
>> Many (well not that many, but they are starting to appear, see
>> for example, or
>> my own [C2 based]) sites use NO tables to achieve
>> layout, but instead rely completely on *standard* CSS positioning
>> properties to achieve layout.

As bad/ugly as tables are and as much as I want to promote standards we have
to accept the fact that, right now, we don't know what our users are using
and we don't have the luxury that an intranet has of dictating what the user
can use and can't use.

>> Let's face it: HTML <table>'s were never designed for laying out pages,
>> but for laying out tabular data. Unfortunately, since the support for
>> CSS was until recently very poor both in the browsers and in the design
>> tools, 99.9% of current web pages use tables for layout. This is IMHO an
>> ugly hack and we, as a community that strives to adhere to open
>> standards and to the concept of separation of style from content, should
>> avoid it like the plague. BTW, CSS-1 was published in 1996, so it's been
>> out for more than FIVE years, it's time that people start using it for
>> what it was meant for.
How about adoption rates? What even worse, you willing to write branches of
CSS that work the same with all browsers?

Don't misunderstand me, I agree with what you're saying here, but don't
punish the user. Most of the browser problems I deal with come from users
who have _no idea_ how to install any software.

>> Using a CSS-based layout also means that people using 4th generation
>> browsers (NS 4, IE 4, etc.) must be "protected" from such a stylesheet
>> or they will see utter garbage. Hiding the CSS from them means that they
>> won't be able to appreciate the layout, but will nonetheless be able to
>> read the full *content*, just not very well styled. But come on, this is
>> a site devoted to *developers* developing for the Web. Can you imagine a
>> web developer today using ONLY NS4 or IE4?
Again it's more work than it's worth until technology adoption for CSS 1 and
2, XML, Native DOM 1 and 2 support grows beyond the latest browser versions
into something more people will use

>> Incidentally, adopting a pure-CSS based solution for both layout AND
>> styling means that people using:
>> - text browsers
>> - screen readers for the sight impaired
>> - mobile devices
>> - anything you cannot conceive now but that will be make web
>>    access available from your washing machine or whatever :)
>> will be able to access the site contents without their "screen" or
>> reader being cluttered with spurious markup that is not in any way
>> related to the content they need.
And what about the people using Netscape 3 because their computer will not
run anything else? I have tu turn people away who are in that situation at
an alarming rate

Accessibility is not just about being standards compliant. It's about giving
the maximum number of users equal access to the material. I'm afraid that
you're moving away from that accessibility goal when you speak of using
technologies that are not about doing that.

Also, one of the goals for Forrest is simplicity. IMO it's a lot easier to
build an HTML template where people can plug things in, therefore not a lot
of time is invested and people can build the projects subsites relatively

And, FYI, unless you screw up the markup, JAWS will read the table but will
skip a lot of that is called "suprious markup" above I will run Jaws through
the current Forrest layout today and report back to the group what do I

>> Before you start mentioning Cocoon's ability to select a different
>> stylesheet based on the User-Agent request parameter, keep in mind that:
>> - we are talking about pregenerating a static version of the site
>>    for performance reasons
>> - as I wrote above, you cannot foresee what user agents will browse your
>>    site in the near future
>> In other words, what I am proposing is that we stop worrying about being
>> bacward compatible in order to accomodate old, buggy and non-compliant
>> user agents, but instead start to be FORWARD compatible in order to
>> accomodate FUTURE standard-compliant user agents.
I disagree. I consider any such radical position to be to the users'
detriment, not their benefit. I believe that any time you close a door in
the way suggested on this message you're alientaing people who are:

* Have to support older browsers because of requirements, legal or
* Use custom browsers such as Links that will display tables for both data
and layout but not stylesheets
* Learning disabled for which learning a new browsing interface would be too
hard to learn or it would take too long

>> Let me know what you think about it and sorry for being slightly OT.
> It's a strong position but, hey, I find resonating with what you're
> saying :)
> We have the *luxury* to know what our user base is and estimate their
> needs very precisely.
> Moreover, this is a site dedicated to new technologies for the web and a
> site dedicated to evangelize open standard compliance thru reference
> implementations and cooperation.
My concern is that the Forrest web site, whether it's built with Cocoon or
not, will be seen as a model, something like: This is what we can do with
this software, and one of the things people will not see there is "how do I
support older browser." Other people, and that was my knee jerk reaction
when I read this, will say Why do I have to download a new browser to view
this? Forget it

> If we page a page on the 'about' section that talks about our reasons, I
> think people might even appreciate our effort to both evangelize the
> technology and 'put in practice' what we say.
See above

In principle I agree with the sentiment, I do not agree with taking it too
far in the other direction.

> What do others think? (we must have a wide agreement to go forward on
> this)

---+ Carlos Araya
 P | WebCT Administrator/Trainer
 _ | California Virtual Campus, Region 1
 G | C/O De Anza College
---+ 21250 Stevens Creek Blvd.,
Cupertino, CA 95014

Phone:      408.257.0420
Fax:        408.255.4406
Web: (work)

"Paradoxically, a refusal to 'put a monetary value on life' means that
life is often undervalued."
-- Artificial Intelligence: A Modern Approach

To unsubscribe, e-mail:
For additional commands, email:

View raw message