Well Ken I did fix it as we discussed but this needs more work. Right
now I am using the SipConfigBean to check if the sip backend is
available.
However I found one problem with this. The rendered property does not
seem to have an effect for the dynamicTreeNode element in the
peTree.jsf.
To elaborate: lets say you have a sun:treeNode as a parent and two
nodes of type dynamicTreeNode within it. Now if you set the rendered
property of the sunTreeNode as false (dynamically), the treeNode as
well as the two dynamicNodes are not shown. This is fine.
Now lets say you have a sunTreeNode which has say 4 dynamicNode
children within it. Now if I set the rendered property of two of these
to false, the dynamicNode is still rendered.
Is this a bug. It seems that the rendered property take effect only on
the sunTreeNode and not the dynamicTreeNode
Rgds
Irfan
Ken Paulsen wrote:
Irfan Ahmed wrote:
Hi Ken,
It seems that the "rendered" property of a treeNode is ignored while
creating the root tree. Hence even if the rendered property is false
the child components of a tree node will always be created.
Now once the child components are created, the rendered property is
used to display the node components on the page. I think the children
should not be traversed if the rendered property of a parent component
is false.
The problem is the value of the rendered property may not be correct
during tree creation time, so it cannot be used reliably at that time.
This may lead to portions of the page missing even when the rendered
property is true (at render time).
However if you want to achieve this functionality, I may be able to
accommodate this via a new property... I don't think we should do this
via the "rendered" property, though.
Ken
(I cc'd the jsft developer list in case others have comments.)
I made a
code change in LayoutViewHandler.buildUIComponentTree
function, that does exactly this. It will not recurse into building the
children components of the parent component's rendered property is
false. By default rendered property is true and hence for all other
cases the UI tree will be built as normal.
I have attached the diff.
Comments.
Rgds
Irfan.
Mark A. Basler wrote:
Hi
Anissa,
The Sailfin branch was merged into the FCS branch a couple of weeks
ago...
Hope this helps - Thanks - Mark
Anissa Lam wrote:
Hi Mike,
Thanks for clarifying this. I don't remember seeing any email that
tells us that SAILFIN_BRANCH has merged back to FCS_BRANCH and thats
the branch we should work on for the 9.1.1 release. But you are right,
i just checked the CVS log, and indeeed Irfan's checkin is to the
FCS_BRANCH.
thanks
Anissa.
Mike Wright wrote:
Hi Anissa,
We're working on JBI deliverables intended for 9.1.1/SailFin, which
we have been told is now in the FCS branch "SJSAS91_FCS_BRANCH" (not in
the tip). My understanding is that the private SailFin branch was
already merged into FCS and, later in November, the 9.1ur1 branch
"SJSAS91_UR1_BRANCH" will be merged into FCS (but we need to merge our
changes before feature freeze). I also took this to mean that no more
work is being done in the private SailFin branch.
The stack traces I'm seeing are after building the
SJSAS91_FCS_BRANCH, due to a missing resource bundle, used by the
SIP-related nodes in peTree.jsf.
Also, I believe that the V3 branch was just merged into the tip,
but my team is not working on V3 code yet (maybe early next year). My
team is currently working on only two GlassFish branches:
1. SJSAS91_UR1_BRANCH for a few pending bug (and findbugs) fixes going
into 9.1_01.
2. SJSAS91_FCS_BRANCH for Sierra features going into 9.1.1 (some
already in 9.1ur1, some ready to commit, some to be finished by feature
freeze, 11/19; these are all pending approval)
Regards,
Mike
---
Anissa Lam wrote:
Hi Mike,
Ken is out today and tomorrow.
I checked out the FCS
(9.1.1/SailFin)
branch
I am confused. Do you checkout the SAILFIN_BRANCH or the tip ?
I haven't built the sailfin branch for a while and not sure the status
of it.
I have cc'ed Irfan who has made changes to the console in the
SAILFIN_BRANCH lately to plugin SIP related tree nodes. He should be
able to tell you what you need to do to get this running.
Yes, please file issue against 9.1.1 for SailFin release.
thanks
Anissa.
Mike Wright wrote:
Hi Anissa,
I haven't received any response from Ken. Are you able to
build/use the console using the SailFin branch?
Should I create a GF IT issue against 9.1.1?
Regards,
Mike
---
------------------------------------------------------------------------
Subject:
SJSAS91_FCS_BRANCH console broken
From:
Mike Wright <Michael.A.Wright@sun.com>
Date:
Wed, 10 Oct 2007 14:53:07 -0700
To:
Ken Paulsen <Ken.Paulsen@Sun.COM>
To:
Ken Paulsen <Ken.Paulsen@Sun.COM>
Hi Ken,
I checked out the FCS (9.1.1/SailFin) branch and built/started a
developer-profile domain. When I log into the console, the navigation
tree has a stack trace due to a missing extension Strings properties
file for sip.
Is this a known problem? Do I need to do something to get this
file? (I've rebuilt the console with the sip-related tree nodes
commented-out for now).
Regards,
Mike
---
Index: com/sun/jsftemplating/layout/LayoutViewHandler.java
===================================================================
RCS file: /cvs/jsftemplating/src/java/com/sun/jsftemplating/layout/LayoutViewHandler.java,v
retrieving revision 1.33
diff -u -r1.33 LayoutViewHandler.java
--- com/sun/jsftemplating/layout/LayoutViewHandler.java 19 Sep 2007 18:44:14 -0000 1.33
+++ com/sun/jsftemplating/layout/LayoutViewHandler.java 12 Oct 2007 12:46:04 -0000
@@ -529,8 +529,10 @@
// NOTE: functionality here will simply do the same thing
// NOTE: automatically).
- // Recurse
- buildUIComponentTree(context, child, childElt);
+ // Recurse only if rendered is true
+ if (child.isRendered()) {
+ buildUIComponentTree(context, child, childElt);
+ }
} else {
buildUIComponentTree(context, parent, childElt);
}