[
https://issues.apache.org/jira/browse/MYFACES-4048?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Leonardo Uribe reopened MYFACES-4048:
-------------------------------------
Some tck test fail with patch, because PhaseId could not be set in that case.
> TransientStateHolder values must be stored in the state if current phase is
> before render response
> --------------------------------------------------------------------------------------------------
>
> Key: MYFACES-4048
> URL: https://issues.apache.org/jira/browse/MYFACES-4048
> Project: MyFaces Core
> Issue Type: Bug
> Reporter: Leonardo Uribe
> Assignee: Leonardo Uribe
> Priority: Minor
> Fix For: 2.2.11
>
>
> In JSF 2.1, it was introduced an interface called TransientStateHolder.
> According to the javadoc, the idea of this API is:
> "... This interface is implemented by classes that need to save state that is
> expected to be available only within the scope of the current request. ..."
> This API is used in very, very few cases (only used once in JSF 2.2 core to
> hold FaceletState temporally), but in portlet case, it is expected that when
> you call saveState(...) over the component on a phase different than render
> response phase, this transient state should be saved with the component state
> too.
> This behavior was not originally thought (because in the JSF EG we never went
> that far to consider portlets) but it has a lot of sense, and in the "spirit"
> to make things just work, I think we need a new issue for this behavior too.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
