Re: WebRequest question

2008-04-11 Thread Maurice Marrink
> It is the developers responsibility that there are no conflicts. Exactly. But if you want A and B where both A and B override for instance the same method, then you are screwed. an automated process would pick just one or simply crash horribly but either way it does not matter because you need t

Re: WebRequest question

2008-04-11 Thread lars vonk
> > Even when using proxies this could get tricky if both try to implement > the same method. whichever implementation wins your app will not work > properly. What do you mean by wins? Isn't this the same as how for instance Spring interception works. There you can add multiple interceptors to on

Re: WebRequest question

2008-04-11 Thread Martijn Lindhout
I would let all the parts in a chain implement the same WebRequest interface for example, and than walk through this chain to let all parts do their work. I agree that there may be some combinations that migth break the chain, for example a security interceptor that will block further handling. 20

Re: WebRequest question

2008-04-11 Thread Maurice Marrink
On Fri, Apr 11, 2008 at 9:46 AM, Martijn Lindhout <[EMAIL PROTECTED]> wrote: > Can't we implement 'something' using an interceptor chain? Extensions can > register themselves in the Application#init() method. There might be a chain > for the webrequest cycle, for example. Each part in the chain m

Re: WebRequest question

2008-04-11 Thread Martijn Lindhout
yeah, you're right, Spring is not a good example 2008/4/11, Igor Vaynberg <[EMAIL PROTECTED]>: > > On Fri, Apr 11, 2008 at 12:30 AM, Martijn Lindhout > <[EMAIL PROTECTED]> wrote: > > > when using Spring you must extends Spring application > > > not if you are using annots, all you need to do is in

Re: WebRequest question

2008-04-11 Thread Martijn Lindhout
Can't we implement 'something' using an interceptor chain? Extensions can register themselves in the Application#init() method. There might be a chain for the webrequest cycle, for example. Each part in the chain may add some behavior to the cycle then. 2008/4/11, Maurice Marrink <[EMAIL PROTECTED

Re: WebRequest question

2008-04-11 Thread lars vonk
> > when using Spring you must extends Spring application You don't need to. You could extend WebApplication and in the init() add : void init() { addComponentInstantionListener(new SpringComponentInjector(this)); } But coming back to your real point: It would be nice if Application and Reque

Re: WebRequest question

2008-04-11 Thread Igor Vaynberg
On Fri, Apr 11, 2008 at 12:30 AM, Martijn Lindhout <[EMAIL PROTECTED]> wrote: > when using Spring you must extends Spring application not if you are using annots, all you need to do is install the spring component injector. -igor > 2008/4/11, Maurice Marrink <[EMAIL PROTECTED]>: > > > > > >

Re: WebRequest question

2008-04-11 Thread Maurice Marrink
You are correct in saying this may cause problems. So it is up to the developers of the extensions to provide workarounds for this. For instance wasp and swarm also provide an interface you can implement on top of the other application you must extend. And if i recall correctly an application can b

Re: WebRequest question

2008-04-11 Thread Martijn Lindhout
I don't exactly where I saw it also, but James has a point. I like to see the things also more pluggable. I don't know if Wicket needs to be modified or if 'wicket-extenders' should take another approach, but I find myself to many times having to extend the same thing (my Application class) from tw

Re: WebRequest question

2008-04-11 Thread Maurice Marrink
Wasp requires a custom WebSession, not a custom WebRequest, to handle some trivial authentication stuff. Maurice On Thu, Apr 10, 2008 at 11:56 PM, James Carman <[EMAIL PROTECTED]> wrote: > I've never really cared for this whole idea of requiring subclassing > to get your work done. Is there no

Re: WebRequest question

2008-04-10 Thread James Carman
I've never really cared for this whole idea of requiring subclassing to get your work done. Is there no way to make things more pluggable? Perhaps use the decorator design pattern? On Thu, Apr 10, 2008 at 4:10 PM, Martijn Lindhout <[EMAIL PROTECTED]> wrote: > If I remember well, Wasp/Swarm needs

Re: WebRequest question

2008-04-10 Thread Martijn Lindhout
If I remember well, Wasp/Swarm needs its own WebRequest subclass? 2008/4/10, Igor Vaynberg <[EMAIL PROTECTED]>: > > On Thu, Apr 10, 2008 at 12:34 PM, Martijn Lindhout > <[EMAIL PROTECTED]> wrote: > > My two questions: > > > > 1. is this necessary / bad? > > > no, it just adds some extra capabili

Re: WebRequest question

2008-04-10 Thread Igor Vaynberg
On Thu, Apr 10, 2008 at 12:34 PM, Martijn Lindhout <[EMAIL PROTECTED]> wrote: > My two questions: > > 1. is this necessary / bad? no, it just adds some extra capability to the request > 2. what if I need different types of requests in my application? i doubt you will. what usecases do you hav

WebRequest question

2008-04-10 Thread Martijn Lindhout
Hi all, In de sample UploadApplication (1.3.3) I saw this: protected WebRequest newWebRequest(HttpServletRequest servletRequest) { return new UploadWebRequest(servletRequest); } and I've seen this several other times for other request types. Putting this code here makes EVERY