IValidator extends IClusterable in trunk

[Matej Knopp <ma...@knopp.sk>]

Hi

As far as I remember, the idea was to have interfaces from 
wicket.validation in trunk independent of the rest of wicket. But with 
the last Eelco's change, replacing ISerializable with IClusterable, it's 
no longer independent. What do we do about that?

-Matej

Re: IValidator extends IClusterable in trunk

[Eelco Hillenius <ee...@gmail.com>]

Done.

Eelco


On 3/3/07, Matej Knopp <ma...@knopp.sk> wrote:
> IIRC the reason why it was separated was to allow implementing
> validators in model/service layer without causing "strong" dependencies
> on Wicket.
>
> I think this is a valid reason and I'd like to keep it as separae as
> possible. Therefore I suggest making IValidator extend Serializable, and
> make AbstractValidator extend IClusterable.
>
> -Matej
>
> Hillenius wrote:
> > Oh, ok. Sorry Matej, I didn't remember that. Is it still relevant
> > though, as the reason to do that was that we could lift it out from
> > Wicket in case we would work together with other frameworks, and it
> > doesn't look like that'll happen soon?
> >
> > If you have a compelling reason, we can roll it back though. Just
> > IValidator?
> >
> > Eelco
> >
> > On 3/3/07, Igor Vaynberg <ig...@gmail.com> wrote:
> >> id say we remove iclusterable, use serializable, and add an extra line of
> >> config to terracotta: it can do something like all serializable
> >> classes in
> >> package foo right?
> >>
> >> -igor
> >>
> >>
> >> On 3/3/07, Matej Knopp <ma...@knopp.sk> wrote:
> >> >
> >> > Hi
> >> >
> >> > As far as I remember, the idea was to have interfaces from
> >> > wicket.validation in trunk independent of the rest of wicket. But with
> >> > the last Eelco's change, replacing ISerializable with IClusterable,
> >> it's
> >> > no longer independent. What do we do about that?
> >> >
> >> > -Matej
> >> >
> >>
> >
>
>

Re: IValidator extends IClusterable in trunk

[Matej Knopp <ma...@knopp.sk>]

IIRC the reason why it was separated was to allow implementing 
validators in model/service layer without causing "strong" dependencies 
on Wicket.

I think this is a valid reason and I'd like to keep it as separae as 
possible. Therefore I suggest making IValidator extend Serializable, and 
make AbstractValidator extend IClusterable.

-Matej

Hillenius wrote:
> Oh, ok. Sorry Matej, I didn't remember that. Is it still relevant
> though, as the reason to do that was that we could lift it out from
> Wicket in case we would work together with other frameworks, and it
> doesn't look like that'll happen soon?
> 
> If you have a compelling reason, we can roll it back though. Just 
> IValidator?
> 
> Eelco
> 
> On 3/3/07, Igor Vaynberg <ig...@gmail.com> wrote:
>> id say we remove iclusterable, use serializable, and add an extra line of
>> config to terracotta: it can do something like all serializable 
>> classes in
>> package foo right?
>>
>> -igor
>>
>>
>> On 3/3/07, Matej Knopp <ma...@knopp.sk> wrote:
>> >
>> > Hi
>> >
>> > As far as I remember, the idea was to have interfaces from
>> > wicket.validation in trunk independent of the rest of wicket. But with
>> > the last Eelco's change, replacing ISerializable with IClusterable, 
>> it's
>> > no longer independent. What do we do about that?
>> >
>> > -Matej
>> >
>>
> 

Re: IValidator extends IClusterable in trunk

[Eelco Hillenius <ee...@gmail.com>]

Oh, ok. Sorry Matej, I didn't remember that. Is it still relevant
though, as the reason to do that was that we could lift it out from
Wicket in case we would work together with other frameworks, and it
doesn't look like that'll happen soon?

If you have a compelling reason, we can roll it back though. Just IValidator?

Eelco

On 3/3/07, Igor Vaynberg <ig...@gmail.com> wrote:
> id say we remove iclusterable, use serializable, and add an extra line of
> config to terracotta: it can do something like all serializable classes in
> package foo right?
>
> -igor
>
>
> On 3/3/07, Matej Knopp <ma...@knopp.sk> wrote:
> >
> > Hi
> >
> > As far as I remember, the idea was to have interfaces from
> > wicket.validation in trunk independent of the rest of wicket. But with
> > the last Eelco's change, replacing ISerializable with IClusterable, it's
> > no longer independent. What do we do about that?
> >
> > -Matej
> >
>

Re: IValidator extends IClusterable in trunk

[Igor Vaynberg <ig...@gmail.com>]

id say we remove iclusterable, use serializable, and add an extra line of
config to terracotta: it can do something like all serializable classes in
package foo right?

-igor


On 3/3/07, Matej Knopp <ma...@knopp.sk> wrote:
>
> Hi
>
> As far as I remember, the idea was to have interfaces from
> wicket.validation in trunk independent of the rest of wicket. But with
> the last Eelco's change, replacing ISerializable with IClusterable, it's
> no longer independent. What do we do about that?
>
> -Matej
>