Okay, found the bug, opened a JIRA, and submitted a patch: https://issues.apache.org/jira/browse/WICKET-2063
The patch attached to the issue at least lets me build my own version that I can ship right now that works. As I mention in the patch, there is probably actually more that needs to be done to fix this issue in a more general way. This patch simply fixes it for the Wizard component. But I suspect that this is a bigger issue - someone who knows about the isEnabledInHierarchy cache should look at this more in depth. The issue is that isEnabledInHierarchy is cached, and when you submit the wizard to move from one step to another, the cached result is first calculated BEFORE the step is advanced. So, it calculates the "enabled" flag for all the buttons based on, say, the first step, just before rendering the second step. -- Jeremy Thomerson http://www.wickettraining.com On Thu, Jan 29, 2009 at 3:13 PM, Jeremy Thomerson <[email protected] > wrote: > No - it works fine with RC1. Tested with all three wizard examples under > "abacadabra" in wicket-examples. > > > On Thu, Jan 29, 2009 at 3:02 PM, Jeremy Thomerson < > [email protected]> wrote: > >> I can't easily revert my codebase to RC1, so I am checking out the Wicket >> source tree at tag rc1 so that I can check the examples there. I'll let you >> know in a few minutes. >> >> >> >> On Thu, Jan 29, 2009 at 2:26 PM, Frank Bille Jensen < >> [email protected]> wrote: >> >>> On 29/01/2009, at 20.58, Jeremy Thomerson wrote: >>> >>> I just came across a bug in 1.4 trunk today that should delay the RC2 >>>> build >>>> this weekend if we can't get it fixed. >>>> >>> >>> Can it be reproduced with rc1? >>> >>> Frank >>> >> >> >> >> -- >> Jeremy Thomerson >> http://www.wickettraining.com >> > > > > -- > Jeremy Thomerson > http://www.wickettraining.com >
