hi ronald,
Ronald van Kuijk wrote:
First of all sorry to have brought this up and put you up with this additional
work that probably
keeps you from becomming fully spec compliant.
no, this is simply driven by real-world use cases, which are not covered
by the spec. sure, our engine should become as compliant as can be, but
it has to be usable in real-world scenarios.
<snip/>
Suppose we have
<node1>
<child11>
<child12>
<repeat>
<.....>
</repeat>
<node1>
And we want to add another repeat, it would be possible to prevent revalidation
of child11 and
child12 right? But if the node1 itself is also a repeat
<root>
<child01>
<child02>
<node1>
<child11>
<child12>
<repeat>
<.....>
</repeat>
<node1>
</root>
It would become more difficult to do this on chil101/child02 and
child11/child12 right?
i'm not sure if i get you right here, but why should this be more
difficult. just have a binding expression like "node1/repeat" for the
validation, and only the nodes contained in the nodeset will be revalited.
when you want to revalidate non-related nodes, just have a
revalidate-action for each.
regards, uli.
If this is the only limitation, I'm all for this solution
Ronald
--
Ulrich Nicolas Lissé
-------------------------------------------------------
This SF.Net email is sponsored by:
Sybase ASE Linux Express Edition - download now for FREE
LinuxWorld Reader's Choice Award Winner for best database on Linux.
http://ads.osdn.com/?ad_idU88&alloc_id065&op=click
|
