Ok then we have a bug here as it seems
well one alternative to saveState probably would
be to use the Orchestra conversation and flash scopes!
But Savestate does nothing fancy, it just serializes its bean
and deseralizes it, according to the request cycle it has to go through.
I wonder if we have a serious bug in the myfaces 1.2.2 save and restore
cycle.
Anyone of the 1.2.x guys care to comment?
Werner
Martin Marinschek schrieb:
Hi Werner,
nothing has been enabled by default.
regards,
Martin
On Feb 12, 2008 12:02 PM, Werner Punz <[EMAIL PROTECTED]> wrote:
<context-param>
<description>Define partial state saving as true/false.</description>
<param-name>javax.faces.PARTIAL_STATE_SAVING_METHOD</param-name>
<param-value>false</param-value>
</context-param>
<context-param>
<description>Only applicable if state saving method is "server" (=
default) and if
org.apache.myfaces.SERIALIZE_STATE_IN_SESSION is true (=
default) If true (default) the serialized state will be
compressed before it is written to the session. If
false the
state will not be compressed.</description>
<param-name>org.apache.myfaces.COMPRESS_STATE_IN_SESSION</param-name>
<param-value>true</param-value>
</context-param>
Ognjen Blagojevic schrieb:
Hi Werner,
I didn't change anything in the webapp configuration. I just replaced
MyFaces 1.2.0 jars (and dependencies) with 1.2.2. Where are savestate
optimizations configured, and how do I turn them off?
Regards,
Ognjen
Werner Punz wrote:
Werner Punz schrieb:
Ognjen Blagojevic schrieb:
Hi all,
I have MyFaces 1.2.0 + Tomahawk 1.1.6 webapp that works fine. I use
mostly request scoped beans, in order to keep the server memory
free. I'm also using t:saveState tag, when there is the need to
preserve the data for h:dataTable (as described on [1]). Parameters
are passed using
<f:attribute name="isSomething" value="#{row.idSomething}"/>
However, when I try to switch to MyFaces 1.2.2, t:saveState shows
unexpected behaviour. It seems that when MyFaces evaluates
expression row.idSomething, it doesn't consult the data that is
saved for the last time with the t:saveState, but the one that is
older, and therfore invalid.
Any ideas on this one?
Sounds like a bug to me, do you have the savestating optimizations
turned on?
To clarify things, to my knowledge some savestate optimizations went
in 1.2.2, maybe there still is a bug there, you might try to turn them
off if you have turned it on.
Werner
