Gili, All you would be doing is factoring out semantic incompatibilities into the strategy interface/impls. In a way the old pageabledataview was a navigation strategy for the regular dataview. Please try to understand the problem. Let me lay it out for you:
The dataview class has setFirstIndex() and setViewSize() methods. The pagabledataview extends the dataview and so inherits those, alongside introducing the setCurrentPage()/setRowsPerPage(). The problem is that setFirstIndex() and setViewSize() do not make sense in the pageable subclass. What does setViewSize(5) mean when you already called setRowsPerPage(10)??? What does setFirstIndex(10) mean when you called setCurrentPage(5)? That's the problem. It has nothing to do with design patterns and java 101 lessons. Currently when paging is enabled setFirstIndex() and setViewSize() will throw an illegalstateexception which makes perfect sense to me, you shouldn't call those if you enabled paging. Also setCurrentPage() will throw the same exception when called while paging is disabled. What we are discussing is simply removing setFirstIndex() and setViewSize() because with paging built in they are not that useful. You lose some flexibility, but not too much. Maybe now that all the research is laid out for you you can post some meaningful comments. -Igor > -----Original Message----- > From: [EMAIL PROTECTED] > [mailto:[EMAIL PROTECTED] On Behalf Of Gili > Sent: Sunday, September 04, 2005 11:27 AM > To: [email protected] > Subject: Re: [Wicket-user] PageableGridDataView > > > It sounds to me like you might be able to use the > Strategy design pattern here. Your design (mentioned in a > previous email) made a lot more sense to me than Igor's idea > of throwing "unsupported" exceptions. > In essense, you could say there are two available navigation > strategies so far: linear and page-oriented. Then you simply > allow users to plug them in and everything should (hopefully) > be clean from inheritance point of view. > > Gili > > Christian Essl wrote: > > On Fri, 02 Sep 2005 18:04:31 -0400, Gili > <[EMAIL PROTECTED]> wrote: > > > >> > >> What is the reason? The new method sounds very anti-OOP to me. > >> Usually one layers functionality using interfaces. > Stuffing them all > >> into the base class but having them throw exceptions when > "disabled" > >> sounds a lot like a flat object model to me. > > > > > > The reason for unifying them in one base class is also to make the > > inheritance hierarchy smaller, less deep and more > consistent. If you > > split DataView and PageableDataView than you also have to make a > > GridDataView and a PageableGridDataView. Two classes which > are from the > > hierarchy total unreleated but are otherwise more or less the same. > > > > Further if you look at the old code of DataView and > PagabeableDataView > > the PageableDataView is not easy to understand, because you have to > > consider two classes with all the protected and overriding > co-work. A > > general problem with inheritance, which gets worse the deeper the > > hierarchy is. > > > > And last but most important the user has to learn two classes where > > there always remains some uncertainty wheter the > child-class does indeed > > behave like the super class, because it could always override some > > methods (as it currently does). > > > > So I am realy for merging them in one class. > > > > Christian > > > >> > >> Gili > >> > >> Johan Compagner wrote: > >> > >>> that is something we had.. and want to get rid of i think.. > >>> > >>> johan > >>> > >>> > >>> Gili wrote: > >>> > >>>> Igor Vaynberg wrote: > >>>> > >>>>> Yes the support delegate is lazily created. > Setcurrentpage() throws > >>>>> illegalstateexception if paging is disabled. > >>>> > >>>> > >>>> > >>>> Sounds like an anti-pattern to me. Why don't you have a base > >>>> class without Pagable support and a subclass that is Pagable > >>>> (always) with these methods. > >>>> > >>>> Gili > >>>> > >>>> > >>>> ------------------------------------------------------- > >>>> SF.Net email is Sponsored by the Better Software > Conference & EXPO > >>>> September 19-22, 2005 * San Francisco, CA * Development > Lifecycle > >>>> Practices > >>>> Agile & Plan-Driven Development * Managing Projects & Teams * > >>>> Testing & QA > >>>> Security * Process Improvement & Measurement * > >>>> http://www.sqe.com/bsce5sf > >>>> _______________________________________________ > >>>> Wicket-user mailing list > >>>> [email protected] > >>>> https://lists.sourceforge.net/lists/listinfo/wicket-user > >>>> > >>> > >>> > >>> ------------------------------------------------------- > >>> SF.Net email is Sponsored by the Better Software Conference & EXPO > >>> September 19-22, 2005 * San Francisco, CA * Development Lifecycle > >>> Practices > >>> Agile & Plan-Driven Development * Managing Projects & > Teams * Testing > >>> & QA > >>> Security * Process Improvement & Measurement * > >>> http://www.sqe.com/bsce5sf > >>> _______________________________________________ > >>> Wicket-user mailing list > >>> [email protected] > >>> https://lists.sourceforge.net/lists/listinfo/wicket-user > >>> > >> > > > > > > > > -- > http://www.desktopbeautifier.com/ > > > ------------------------------------------------------- > SF.Net email is Sponsored by the Better Software Conference & EXPO > September 19-22, 2005 * San Francisco, CA * Development > Lifecycle Practices > Agile & Plan-Driven Development * Managing Projects & Teams * > Testing & QA > Security * Process Improvement & Measurement * > http://www.sqe.com/bsce5sf > _______________________________________________ > Wicket-user mailing list > [email protected] > https://lists.sourceforge.net/lists/listinfo/wicket-user > > > ------------------------------------------------------- SF.Net email is Sponsored by the Better Software Conference & EXPO September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf _______________________________________________ Wicket-user mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/wicket-user
