Glenn Marcy wrote:
> I believe that the original issues related to dynamic
> validation being set and people wondering why nothing
> happened.  When we told them that they also needed to
> set the validation feature they said that was dumb.
> Of course, when they later turned off validation they
> asked why we still were validating and we said that
> they still had dynamic validation set, and again we
> were told that was dumb. 

I don't understand why Xerces1 required turning on both features
(dynamic and validation)?
The DOM L3, for example, defines that turning on dynamic validation
automatically turns off validation feature and vice versa: turning on
validation will disable dynamic validation [1]. In this case, user would
not have to wonder what happens if you turn both on and then turn one of
them off. 

Given that Xerces2 does not do what Xerces1 did, the users should have
been broken for at least a year. However, we did not receive any bugs
reports. Also we've received no replies to Sandy's posting about this
problem, thus it seems that Xerces users do not care which way we are
going. On top of it, our documentation claims that dynamic validation
does not depend on the state of validation feature. So I fail to see how
the change could possibly break anybody.

Thus, I would be in favor changing the implementation according to the
description of dynamic feature: if dynamic feature is set to true the
document will be validated iff it has a grammar. While the state of
dynamic feature is true, the state of validation feature is of no
importance.

[1]
http://www.w3.org/TR/2002/WD-DOM-Level-3-Core-20021022/core.html#DOMConfiguration
-"validate-if-schema"

Thanks,
-- 
Elena Litani / IBM Toronto

---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to