[
https://issues.apache.org/jira/browse/WICKET-5387?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13791476#comment-13791476
]
Sven Meier commented on WICKET-5387:
------------------------------------
Hm, maybe with your patch the following "dirty hack" in Page#init() is obsolete?
// this is a bit of a dirty hack, but calling dirty(true) results in
isStateless called
// which is bound to set the stateless cache to true as there are no
components yet
stateless = null;
BTW can anybody make sense of the comment a few lines earlier? Does it still
hold then?
// All Pages are born dirty so they get clustered right away
dirty(true);
And why is #dirty(boolean) public anyway?
> Page#onInitialize called after an exception in the constructor of Page
> ----------------------------------------------------------------------
>
> Key: WICKET-5387
> URL: https://issues.apache.org/jira/browse/WICKET-5387
> Project: Wicket
> Issue Type: Bug
> Components: wicket
> Affects Versions: 6.11.0
> Environment: Linux
> Reporter: Walter B. Rasmann
> Attachments: 5387.1.tar.gz, 5387.tar.gz, WICKET-5387.patch
>
>
> Page#onInitialize is called when the constructor of Page throws an exception,
> i.e. when the Page is not initialized correctly. This can cause additional
> exceptions which are usually added to an error log even in cases in which the
> exception in the constructor is handled (by Wicket). This issue is possibly
> related to WICKET-5083.
> Another case of the problem occurs when setResponsePage(...) is used in the
> constructor to navigate away from a page that can't be initialized correctly.
> I'm attaching a quickstart.
--
This message was sent by Atlassian JIRA
(v6.1#6144)
