3 Replies Latest reply on Jun 7, 2007 1:17 PM by iradix

    Pages Hierarchy

    iradix

      First off, Gavin I wanted to say thanks for all the improvements to pages.xml since 1.0. Even though my patch didn't make it in (JBSEAM-397), it's great to see most of the functionality there. I do have a couple of questions about how it's all been implemented though.

      In my experience most pages within a specific conversation seem to match a wildcard (say /cart/*) where one of the pages in that group (say /cart/view.jsf) is responsible for starting the conversation. In my head this would translate into:

      <page view-id="/cart/*" conversation-required="true" no-conversation-view-id="/cart/view.xhtml"/>
      
      <page view-id="/cart/view.xhtml" conversation-required="false">
       <begin-conversation join="true"/>
      </page>


      The way Pages is implemented however, since pages are processed individually starting with the worst match, the conversation-required attribute on the wildcard forces a redirect before the conversation can be begun by the more specific match. When I put my patch together I had laid out pages in a parent child relationship so that single parameters like conversation-required could be overridden by more specific matches and multiple parameters like action could still be processed worst match first. It makes more sense to me that way and I noticed at least one other poster with a similar problem regarding login-required. I'm not saying that my way was better, but could we find a compromise between the two? It seems like the ability to override wildcards would be useful in mutliple circumstances.

      I am willing to submit another patch, but I wanted to see what others thought and gauge how well it would be accepted before I do.

      Thanks.

        • 1. Re: Pages Hierarchy
          iradix

          I remember hearing that Gavin wasn't around when I posted this initially, I don't know if he's back now but either way, does no one else have any interest?

          • 2. Re: Pages Hierarchy
            gavin.king

            I guess I could make it three-valued, ie. true|false|inherit, which would give you the behavior you prefer.

            Add a feature request to JIRA if this is something you really *have* to have.

            • 3. Re: Pages Hierarchy
              iradix

              To be honest, if I'm really the only person interested in the feature I can live without it. Unless someone else feels like chiming in I can't see having you devote the time to changing it.