If determining all possible object compositions doesn't work-- what
about piggy backing on JBoss's work and do something at the
ObjectOutputStream level?
Martin Haimberger wrote:
Hy Jacob,
you are right. This is an old artifact and i will remove.
Thanks for looking
Nice Regards
Martin Haimberger
On 2/25/07, Jacob Hookom <[EMAIL PROTECTED]> wrote:
This seams a bit odd:
private TreeStructComponent
internalBuildInitalTreeStructureToSave(UIComponent
component,FacesContext facesContext, Object state, int childIndex)
{
Object myState = null;
Map facetStateMap = null;
List childrenStateList = null;
if (state != null)
{
myState = ((Object[])state)[0];
facetStateMap = (Map)((Object[])state)[1];
childrenStateList = (List)((Object[])state)[2];
}
}
Under what assumptions is the state always going to be broken out
that way?
Martin Marinschek wrote:
> Hey *,
>
> I've finally found the time to go through the partial state-saving
> stuff Martin Haimberger did through his Google SoC assignment and we
> have together made sure the kinks are polished out pretty much. The
> rest will:
>
> a) enable much less state saved per each request (by a factor of 3,
> 4), at least with the MyFaces examples, we haven't tried with
> real-world applications that much so far (we'd be happy about
> additional data from you)
> b) allow tree-building just like in 1.2. You basically get most 1.2
> features with 1.1 like this.
>
> The jira issue with the patch is:
>
> https://issues.apache.org/jira/browse/MYFACES-1541
>
> I wanted to wait with the commit until the release is out, and I
> encourage you to do a lot of testing. Martin will be around next week
> to fix anything that comes up.
>
> Martin, please create a WIKI page where you describe how stuff can be
> configured.
>
> regards,
>
> Martin
>
