Be Careful What You Wish For..
- Standards conformance by web sites improves accessibility.
- Far too many sites are unnecessarily non-compliant.
- Web developers also have an obligation to support the standards.
- The extra efforts can be justified to management and clients.
- Great sites can be built in HTML4 within current limitations.
- Better sites will be built when the browsers support HTML4 etc.
- Most present sites may break when browsers improve.
- Building compliant sites gets easier with practice!
|
Some Popular Myths
- The browsers don't support the standards so I can't either.
- Anyway, if I try to it'll take too much time and effort.
- All that really matters is how a site looks under Microscape..
Given the budget and time constraints we all work under and the range
of expertise we're all expected to have (HTML, client-side scripting,
the CGI, typography, image optimization, copywriting, usability
engineering, user interface design, information architecture,
salesmanship, customer service, etc. etc), it's unrealistic to expect
developers to write to a third browser (which is what standards in
effect are) that doesn't even exist.
As not only a web developer, but also a software developer for a rather
long time - I'm very familiar with the "real world, clients, and
practical constraints" argument. And, on the surface, it carries a lot
of weight and "management appeal". But, I think that it needs a deeper
examination to ensure that it doesn't miss some factor(s) which cost us
dear down the road. Let me give you an example from the software world:
if you have to write a program, there's basically two ways - as quickly
as possible, or, spend some time on the requirements, design, and code
quality issues. In the early days #2 wasn't popular with management -
until software systems started getting more complex, bugs more frequent,
and maintenance became a nightmare (e.g. Y2K). I'm not saying that
history will repeat itself exactly the same for web developers, but,
that there might be some useful parallels, and lessons to learn.
Improvements happened in the software world when developers started
"front-loading" their development cycles, i.e. spending more time on the
early phases. "Structured Analysis, Design and Programming" were
invented and caught on - not only because management began to understand
the downstream consequences of software that was written too quickly -
but also because no-one wanted to be accused of "unstructured
programming". So, if
valid HTML is a Good Thing, perhaps we could coin
a similar phrase that carries the connotation of "good vs. sloppy"..
Now, I claim that valid HTML is a Good Thing, and I imagine few
professional developers would passionately disagree. But, I suspect that
many developers feel that the "real world" doesn't allow time for the
niceties of writing it - or, that it's just not possible, because the
browsers don't support it. And my point is, yes,
you will have to invest some up-front time, to study the W3C specs and
to write valid code - but, you'll minimise the downstream problems, and
find that it gets easier with practice (SP was hard at first, but after
some experience, it became hard to not do it).
Let's recall the main reason for writing to the standards: pages should
be viewable across the widest range of user agents, versions, and
platforms. That includes not only the most popular GUI browsers and OSs
but also text-to-speech synthesizers for the visually impaired; search
engine spiders and other robots; tools, esp SGML-based ones; text-only
browsers; devices other than PCs and Macs; and devices not yet
available. For example, if you use UL or BlockQuote or other such tricks
just to indent text then the page will break, sometime,
for someone or something.
"A conforming user agent for HTML 4.0
is one that observes the mandatory conditions
("must") set forth in this specification.." This implies, e.g. that
certain popular tricks such as indenting text with itemless lists will
break; since
"All lists must contain one or more list elements.".
The W3C HTML specs are carefully designed to maximise
accessibility, which most clients probably welcome - more potential
customers! The legacy we're creating now, with non-compliant websites,
could come back to haunt us if we're succesful in our efforts to get
standards-supporting browsers.
Looking at the source code of many web sites, I noticed that many of
the validation errors were easily avoidable - such as supplying the
missing ALT attributes (NOT 'tags'!). If it became habit for
people to include ALT (or to use generators that do) and do the other
things that the specs require, then the extra time to "do it right"
needn't be substantial, in most cases.
I'd like to propose a methodology for web site development:-
- Write the first (possibly prototype) version in standard HTML.
- Validate it and all subsequent versions.
- If/when it meets requirements, attach the W3C HTML
logo and link it to
http://validator.w3.org/check/referer
- If full compliance isn't possible with your first design, look for
other ways to code the page with valid HTML (e.g. workarounds; alternate
design; etc).
- If you just can't do it, let the code become invalid, and write to
the browser makers, this list, and/or me, with a description.
The most important step here is the first - it's harder to retrofit
conformance than to start with it in the first place.
It's a 2-way street. You can't demand the browser makers observe the
standards while you ignore them. What would Joe User think? And what
will Jane think when she visits your site in Microscape 5 or 6 and it's
broken because Netsoft decided to be stricter and less forgiving (as
they've done before). It's easier to do it right than to do it over.
Before someone turns my argument around, squeezes out all the juice,
and pushes it past the end-stop - I'm not saying that developers
creating only compliant sites will force the browser makers to support
the standards - directly. I *am* saying there might be some perceived
hypocrisy in demanding standards while not observing them; *and* that it
can be done more often and easily than many might realise.
Perhaps the worst thing that could happen is for the browser makers to
suddenly "see the light" and fully and
faithfully implement the standards, especially HTML 4.0 (e.g. by SGML
parsing against the DTD). Sure, you'll get lots of nice new tags, and
the old ones finally behaving right. But your spiffy sites will be spit
out! Do you really want the standards, lock, stock, and barrel? If so,
you may need to check your sites now. If not, state your wish more
carefully..
BroWWWsers
HTML Standards Compliance - Why Bother ?
HTML Checkers
|