W3C home > Mailing lists > Public > www-forms@w3.org > October 2000

Re: XForms Input Validation

From: Dave Raggett <dsr@w3.org>
Date: Mon, 16 Oct 2000 13:55:35 +0100 (GMT Daylight Time)
To: Neil Groves <Neil.Groves@ait.co.uk>
cc: "'www-forms@w3.org'" <www-forms@w3.org>
Message-ID: <Pine.WNT.4.10.10010161347170.-278235@hazel>
On Mon, 16 Oct 2000, Neil Groves wrote:

> The abstraction of the validation logic with min/max values and
> filters is very welcome indeed.  I was wondering if there was an
> opportunity to cope with even more complex validation issues
> without resort to script code on the client side.  My beef with
> the client side script code performing the validation is simply
> that in many circumstances what is being validated are not
> simple min/max ranges, but algorithms that should be defined and
> implemented in the middle-tier.
> I was thinking that it might be an idea to be able to have the
> middle-tier define/generate the script and pass this in an XML
> block to the client-side where it is unwrapped and hooked into a
> control.  This clearly would involve more than just XForms, but
> wouldn't it be marvelous? A total end to validation algorithmic
> duplicaiton without unnecessary round-trips to the server.
> I'm probably missing the point, and am ready for a barrage of
> insults!  I love what is being proposed it looks a far cry from
> the terrible HTML/ASP nonsense I am currently working with.

Can you expand on what you mean by "middle tier"?

Can you provide use cases for the more complex kinds of
validation you have in mind?

The data type facets are fine for simple constraints, but we
envisage the need for more complex constraints expressed
declaratively via some kind of expression language. Sometimes even
this will be insufficient, and we envisage a means to call out to
locally executed scripts.

In some cases it will always be impractical to validate fields on
the client, e.g. where this involves a database that is too large to
download, for example, the US white pages. The working group expects
to handle this via some kind of RPC mechanism probably invoked via a

Any use-case information you can give us will be helpful for
checking out the emerging proposals.


-- Dave Raggett <dsr@w3.org> http://www.w3.org/People/Raggett
tel/fax: +44 122 578 3011 (or 2521) +44 778 532 0444 (mobile)
World Wide Web Consortium (on assignment from Phone.com)
Received on Monday, 16 October 2000 08:55:47 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:36:03 UTC