Re: Website ideas



On Fri, Nov 17, 2000 at 05:01:59PM -0500, Simon J. Hernandez wrote:
> On 18 Nov 2000, Ryan Muldoon wrote:
> [Content elided]

> Thank you Ryan for bringing this up.  It seems, before any real
> development takes place, it might be interesting to pursue a thread on
> what design philosophy the site seeks to pursue, i.e., will the site be
> HTML 4.0 (4.01) compliant?  Will the site be designed and coded to allow
> accessibility for those with disabilities?  This can be a model site
> without having to jump through too many hoops, and at the same time be
> easily manageable.  This is an exciting new beginning, and I believe if
> this considerations can be worked in early on, it could make a lot of
> thins easier to implement.  Using valid HTML makes focusing on particular
> browser much less of an issue, and really irrelevant.  All good HTML
> should be cleanly viewable through any compliant browser.

> I do not mean to slow things down, as I realize there is a desire for
> rapid development.  What does everyone think about these issues?

> References:

> 1.  http://www.w3.org/TR/html401/
> 2.  http://www.w3.org/WAI

This is an interesting question, and an important one. The current
www.gnome.org design that I did is mainly HTML4.01 compliant, but I do a few
things that causes it to fail validation. Why? To make it work better with
older browsers. What do I do? Use some attributes that aren't in the HTML
4.01 DTD, mainly things like marginwidth and marginheight to <body>, etc.

I'm of the opinion that we should continue doing this. We should probably
provide our own HTML4.01 variant DTD, so people can actually validate the
HTML on the site, and add the necessary attributes to that DTD. This is
actually a fairly common way of doing things if you want to follow the
standards but still need to use some so-called "vendor tags". The main idea
is to not clain the document is something it's not (in the DOCTYPE).

What's my rationale for this? Older browser don't do CSS. There are some
attributes that were added by browser manufacturers after the HTML 3.2 spec
(which was supposed to be a snapshot of HTML supported by major browsers at
the time), that provide equivalent mechanisms to parts of CSS, but these
ended up not being in any W3C standard, because HTML4 was designed mainly to
be a vehicle for layout using CSS. In my opinion, it's a shame these
attributes didn't make it into HTML 4.0 Transitional, since they would have
been very helpful for people who actually care about backwards compatibility.
It should also be added that user-agents encountering attributes they don't
know, discard them. This is the policy recommended by the W3C for forward
compatibility.

The result of all this: We can provide our own HTML4.01 variant DTD that will
validate our documents properly even if we use things like <body marginwith>,
and I think we should. There are good, sane usability reasons for it, since
screen realestate is at a premium.

As for the other questions you raise, yes, I think it should be easy to make
the sites accessible for people with disabilities and so on, and it should be
a goal for us to do so.

I disagree with your sentiment that if one only uses valid HTML 4.01,
focusing on particular browsers becomes irrelevant, though. Even though the
W3C (and I, too) might wish this to be true, that doesn't change reality. And
reality is a large pile of browser, probably the vast majority out there, not
supporting HTML 4.01, CSS1 and CSS2 properly (and not just because they're
not new enough, the majority attempt to support the standards, but do it
wrong). We *can't* afford to simply ignore those browsers, we will have to
test in all sorts of different browsers on different platforms, and devise
workarounds so pages are usable on as large a set of common user agents as
possible.

This is simply our duty.

-- 
Joakim Ziegler - Helix Code web monkey - joakim helixcode com - Radagast IRC
      FIX sysop - free software coder - FIDEL & Conglomerate developer
            http://www.avmaria.com/ - http://www.helixcode.com/




[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]