You can take a look at INullAcceptingValidator, it might do what you are after
-Matej On 11/1/07, Johan Compagner <[EMAIL PROTECTED]> wrote: > we had that and we changed it to a boolean because required is a thing > that must be validated before converters and other validators > > On 11/1/07, Bruno Borges <[EMAIL PROTECTED]> wrote: > > isn't this a reason to start with that talk again Igor, about replacing > > setRequired with an IValidator? > > > > regards, > > bruno > > > > Igor Vaynberg wrote: > > > On 8/13/07, Eelco Hillenius <[EMAIL PROTECTED]> wrote: > > > > > >>>> You are just nitpicking. 99.9% of people won't override setRequired to > > >>>> start with. > > >>>> > > >>> i dont know about that. seems to me that the majority of people who > > >>> implement a FormComponentPanel will be forced to overwrite it since that > > >>> > > >> is > > >> > > >>> currently the model you have put in place. > > >>> > > >> Yeah, they should, as they probably didn't support the required flag > > >> properly before. > > >> > > > > > > > > > they did just fine after my "hack" > > > > > > > > >> i had to do it for > > >> > > >>> formcomponentpanels in my projects, and i really dont like the code it > > >>> causes. > > >>> > > >> You had to? Well that's interesting. How did you fix this before this > > >> change then? > > >> > > > > > > > > > with my "hack" i had it in my own wicket branch. > > > > > > The situation is *exactly like it was before this whole discussion* > > > > > >> with the exception that setRequired isn't final anymore, so that I can > > >> fix the components I'm interested in in a way I think works good. I'm > > >> pretty sure the fact that setRequired isn't final anymore won't wreak > > >> havoc. > > >> > > > > > > > > > yes but pushing the required state is ugly, not to mention all the > > > noobishess that will come with opening it! > > > > > > -igor > > > > > > > > > Eelco > > > > > > > > > > > > > >
