Re: How to listent for the Request Cycle End

2006-09-22 Thread Dobrin Ivanov
> > the only inelegant bit is this visitKey assembling > > part may be :) > > > > --- James Carman <[EMAIL PROTECTED]> > wrote: > > > > > >> Sorry. Try > >> tapestry.request.WebRequestServicerPipeline. As > I > >> said

Re: How to listent for the Request Cycle End

2006-09-22 Thread Bryan Lewis
request.WebRequestServicerPipeline. As I >> said, it was >> off the top of my head. Sorry for the typo. >> >> -----Original Message----- >> From: Dobrin Ivanov >> [mailto:[EMAIL PROTECTED] >> Sent: Friday, September 22, 2006 1:45 PM >> To: Tapestry

RE: How to listent for the Request Cycle End

2006-09-22 Thread Dobrin Ivanov
Message- > From: Dobrin Ivanov > [mailto:[EMAIL PROTECTED] > Sent: Friday, September 22, 2006 1:45 PM > To: Tapestry users > Subject: RE: How to listent for the Request Cycle > End > > 10x > > Now i get the error below. Any chance that there are > some d

RE: How to listent for the Request Cycle End

2006-09-22 Thread James Carman
AIL PROTECTED] > Sent: Friday, September 22, 2006 1:26 PM > To: Tapestry users > Subject: RE: How to listent for the Request Cycle > End > > And where should be located my hivemodule.xml? (I'm > using just Tapestry, I know it lies above hivemind > microkernel...) &g

RE: How to listent for the Request Cycle End

2006-09-22 Thread Dobrin Ivanov
ED] > Sent: Friday, September 22, 2006 1:26 PM > To: Tapestry users > Subject: RE: How to listent for the Request Cycle > End > > And where should be located my hivemodule.xml? (I'm > using just Tapestry, I know it lies above hivemind > microkernel...) > > >

RE: How to listent for the Request Cycle End

2006-09-22 Thread James Carman
ailto:[EMAIL PROTECTED] > Sent: Wednesday, September 20, 2006 1:30 PM > To: Tapestry users > Subject: Re: How to listent for the Request Cycle > End > > I do not know about this custom Engine classes > changes > (frowned)... > is there some information about this top

RE: How to listent for the Request Cycle End

2006-09-22 Thread Dobrin Ivanov
op of my head, but you get the idea. > This basically acts > like a servlet filter, but you can plug > hivemind-managed filters in (so you > can inject stuff into your implementation objects). > > > -Original Message- > From: Dobrin Ivanov > [mailto:[EMAIL PROTEC

Re: How to listent for the Request Cycle End

2006-09-20 Thread James Carman
his basically >> acts >> like a servlet filter, but you can plug hivemind-managed filters in (so >> you >> can inject stuff into your implementation objects). >> >> >> -----Original Message- >> From: Dobrin Ivanov [mailto:[EMAIL PROTECTED] >> Sent: We

Re: How to listent for the Request Cycle End

2006-09-20 Thread Hajaansh
nd-managed filters in (so you can inject stuff into your implementation objects). -Original Message- From: Dobrin Ivanov [mailto:[EMAIL PROTECTED] Sent: Wednesday, September 20, 2006 1:30 PM To: Tapestry users Subject: Re: How to listent for the Request Cycle End I do not know about this

RE: How to listent for the Request Cycle End

2006-09-20 Thread James Carman
stent for the Request Cycle End I do not know about this custom Engine classes changes (frowned)... is there some information about this topic? .. and also the other one with the pipelines/WebRequestServicerPipeline/interceptors? --- James Carman <[EMAIL PROTECTED]> wrote: > You can

Re: How to listent for the Request Cycle End

2006-09-20 Thread Dobrin Ivanov
I do not know about this custom Engine classes changes (frowned)... is there some information about this topic? .. and also the other one with the pipelines/WebRequestServicerPipeline/interceptors? --- James Carman <[EMAIL PROTECTED]> wrote: > You can plugin to the webrequest servicer pipeli

Re: How to listent for the Request Cycle End

2006-09-20 Thread James Carman
You can plugin to the webrequest servicer pipeline. > Ummm... you could do that too. You're correct that it would avoid the > visit-from-session inelegant bit. It would be conceptually similar to > the servlet filter approach. The downside would be that custom Engine > classes are frowned upon

Re: How to listent for the Request Cycle End

2006-09-20 Thread Bryan Lewis
Ummm... you could do that too. You're correct that it would avoid the visit-from-session inelegant bit. It would be conceptually similar to the servlet filter approach. The downside would be that custom Engine classes are frowned upon as Tapestry goes forward. I'm not sure there is an Engine.ge

Re: How to listent for the Request Cycle End

2006-09-20 Thread Dobrin Ivanov
Hi, Thanks Bryan, this looks like a hack:) What do you think if I override the Engine's method: public void service(WebRequest request, WebResponse response) throws IOException { super.service(request, response); // insert code here } Is this a bad approach? --- Bryan Lewis <[EMAIL PROT

Re: How to listent for the Request Cycle End

2006-09-19 Thread Bryan Lewis
It sounds like a servlet listener method could work for you. Or a servlet filter as in the previous suggestion. Both would give you a hook into the end-of-request, and you can get to the Visit via the session. Here's a listener approach. public class EventListener implements ServletRequestList

Re: How to listent for the Request Cycle End

2006-09-19 Thread Dobrin Ivanov
I have designed some small API in order to provide the session persistance of the presentation layer (Tapestry - Visit object/HttpSession) to the model layer (in order to be able to cache some session related stuff without being aware of how the above layer is doing it). So the data is attached to

Re: How to listent for the Request Cycle End

2006-09-19 Thread Martin Strand
Exactly what do you need this for? If you don't need any Tapestry logic, there might be other ways to do it - like a servlet filter or a threaded service that implements Discardable. On Tue, 19 Sep 2006 21:58:20 +0200, Jesse Kuhnert <[EMAIL PROTECTED]> wrote: It might not be super fun to l

Re: How to listent for the Request Cycle End

2006-09-19 Thread Jesse Kuhnert
It might not be super fun to learn, but I think the "tapestry" way of doing this would be to contribute something to the WebRequestServicerPipeline so that you know definitively when the cycle ends regardless of what services/engines are involved.. http://tapestry.apache.org/tapestry4/tapestry/hi