Re: [xml-dev] Browser innovation efforts -- where's W3C in this picture?

* Michael Champion wrote:
>I've noticed a couple of things that seem a bit odd, and would like to 
>get a better idea of the context.  First, there was the Mozilla/Opera 
>collaboration on Web Forms 2.0 [...] Uhh, why this rather than just
>supporting XForms?

Extending HTML's form features has lower implementation cost, a lower
learning curve for content providers, and can be - to some extend -
designed to degrade gracefully in clients that do not support those
new features, so such new functionality would be relevant for content
providers today rather than in several years when support is stable
and available to the relevant audience. So the question rather seems
to be why should they support XForms?

A good reason would be that users demand support because they want to
use XForms content. It does not seem that there is a significant class
of users who demand that, probably because they are not aware of XForms
content they want to use but cannot and/or they are not aware why they
would prefer XForms content over HTML forms content.

Another good reason would be that content providers demand support
because they want to provide XForms content. Such content providers
are rare, too, probably because they are not aware of XForms at all,
or they are not convinced of the benefits XForms offers over HTML
forms, or they are convinced but do not speak up loud enough or not
at all.

W3C's only answer to "better" forms on web sites so far is XHTML 2.0
for which the latest Working Draft is fourteen months old (and the
first planned date for Last Call that I know of was 7 Dec 2001...),
so, as XHTML 2.0 is going to be incompatible with all previous
HTML/XHTML versions, what would be the migration path for content
providers? There is none, the best we can currently expect is that
XHTML 2.0 will use a different MIME type than XHTML 1.0 which was
at least controversial last time it was mentioned on www-html.

Even if there will be a new MIME type, it would mean that content
providers will have to provide at least two versions of their content
until user agents that support XHTML 2.0 will be sufficiently deployed.
Unless there is a sufficiently cheap tool that auto-generates HTML
content from XHTML 2.0 content of similar quality, this will probably
be seen as too expensive. So they would wait until it is sufficiently
deployed. Considering that there are still content providers who care
about supporting Netscape Navigator 4.x it seems unlikely that this
will happen before 2008.

That's a bit long for content providers to wait, isn't it? After all,
the need for better forms was present even during the development of
HTML 3.0 about ten years ago, which proposed e.g. a range control and
scribble pads...

If there is no explicit demand, it might also be a good reason for a
browser vendor to implment a specific technology if they are convinced
there is such demand but it just isn't expressed for some reason, so
that they can reasonably assume that the investment in an implementation
will pay off. For this they would need to be able to reasonably assume
that content providers will adopt the technology. For XForms it is
however argued that they would not as it is very different from what
they already know and that it is far more complicated. Maybe that is a
false perception so that this is basically a marketing issue, maybe it
is true and thus a technical problem.

And even if the browser vendor is convinced that it should implement
XForms, they apparently struggle into implementation issues. My
understanding is that for example Opera wants to ship its browser
with equal features for desktops and small devices of which the latter
have constraints in terms of storage/memory capacities and that
according to them an XForms Basic implementation does not fit into
these constrained facilities, not to mention the entire XForms
specification.

I have however good faith that they would sort this out. To me it seems
that they do not think that XForms is clearly a superior solution for
forms on web sites. Is it?

And if it is, can we tell content providers anything else than to wait
three to six years to use XForms on their web site? And if not, what is
exactly wrong with creating something that can be used in the meantime?

Received on Tuesday, 6 July 2004 22:23:30 UTC