Hello,
as it has been mentioned in my posting in this thread at Tue Jun 24, 2008,
it is one option to use a h:commandButton or a h:commandLink
instead of a4j:commandLink.
If you decide to accept a page reload and use a h:commandLink
then the parameter
<context-param> <param-name>com.sun.faces.enableJSStyleHiding </param-name> <param-value>true</param-value> </context-param>
<context-param> <param-name>com.sun.faces.preferXHTML</param-name> <param-value>true</param-value> </context-param>
If you have any patches created feel free to submit them to jira. we'll revise them and apply if ok.
Hello Ilya,
the good news is:
After we reevaluated open source alternatives (Tomahawk, Tobago, Trinidad, ICEFaces, Woodstock) and comparing
http://jira.icefaces.org/secure/BrowseProject.jspa
with
https://jira.jboss.org/jira/browse/RF
we stay with RichFaces in our current project.
On the other side there is a decision not to use the ScrollableDataTable
in the present state our current project. At the moment we switched to use the DataScroller for a DataTable.
There are two reasons for this decision:
1. The observed progress with the known bugs in the ScrollableDataTable and especially of the bug
https://jira.jboss.org/jira/browse/RF-4038
2. Even if the component was robust we see problems with speed of the
rendering of the cells that are shown on the table. If the factor
visible rows * columnsof the table is larger than let's say 400 hundered the table tends to be far too slow.