I don't care that much that the implementations differ if there is a
good reason for it. So if there is, discard my complaint :)
What I am concerned about right now is that several thread tests fail
horribly for 2.0. I hope that's easy to fix, but didn't investigate
that.
Eelco
On 10/21/06, Johan Compagner <[EMAIL PROTECTED]> wrote:
You mean that sync block that had to be a bit smaller so that resolve page
wasn't touched there?
Yes that couldn't be fixed in 2.0 because that sync block wasn't there at
all.
It is because the "requestParameters.isOnlyProcessIfPathActive()" call
That isn't there completely in 2.0 i asked igor and matej about that already
and we need to move
that code to 2.0
Because i guess an ajax request can now come in and do stuff on a page that
isn't the active anymore
in 1.2 this was bad because it would expire the real page. In 2.0 this
doesn't happen in the new session store ofcourse
so the error will not popup but it would be better to synch that behaviour.
So this constant:
/**
* Parameter name that tells decode to ignore this request if the
* page+version encoded in the url is not on top of the stack. The value
of
* this parameter is not important, it simply has to be present to
enable
* the behavior
*/
public static final String IGNORE_IF_NOT_ACTIVE_PARAMETER_NAME =
NAME_SPACE
+ "ignoreIfNotActive";
must be introduced in 2.0 i guess.
johan
On 10/20/06, Eelco Hillenius <[EMAIL PROTECTED]> wrote:
>
> Hi,
>
> DefaultRequestTargetResolverStrategy in the 1.2 branch and in trunk
> differ quite a lot. Kind of scarry, as I'm afraid we might be missing
> fixes in one of them. Could someone please track the changes and see
> whether the versions are in sync? Also, there was a patch submit for
> our syncing problems yesterday, but only in 1.2. Which probably means,
> 2.0 is still a problem? I'll update the threadtest project for 2.0
> this afternoon. If someone else could please make sure 2.0 is without
> syncing problems, that'd be great.
>
> Eelco
>
