i havent look at the code but the fix could be something like that you don't
push
the page after construction on the request stack because if you do that then
you will overwrite
the page that was pushed in the constructor of that page..
So are there really landing 2 pages on the request stack?
the fix is already in.
There are no problems, everything works. This thread can be ended.
Eelco
On 2/20/07, Johan Compagner [EMAIL PROTECTED] wrote:
i havent look at the code but the fix could be something like that you don't
push
the page after construction on the
and you made sure everything is being detached properly? :)
-igor
On 2/20/07, Eelco Hillenius [EMAIL PROTECTED] wrote:
the fix is already in.
There are no problems, everything works. This thread can be ended.
Eelco
On 2/20/07, Johan Compagner [EMAIL PROTECTED]
On 2/20/07, Igor Vaynberg [EMAIL PROTECTED] wrote:
and you made sure everything is being detached properly? :)
Yes. That was the sole idea about keeping a stack with request targets
to start with.
Eelco
-
Take Surveys.
umm...that was a joke :)
-igor
On 2/20/07, Eelco Hillenius [EMAIL PROTECTED] wrote:
On 2/20/07, Igor Vaynberg [EMAIL PROTECTED] wrote:
and you made sure everything is being detached properly? :)
Yes. That was the sole idea about keeping a stack with request targets
to start with.
Eelco
On 2/20/07, Igor Vaynberg [EMAIL PROTECTED] wrote:
umm...that was a joke :)
Bastard! :)
-
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
you shouldnt be calling setresponsepage inside constructors
try RestartResponseException
Yes, RestartResponseException forced it to response fine.
Anyway, it's a shame setResponse sometimes fails in constructors, I
already have other web-flows working fine with it. That's because I
have some
available, and I cannot think of a good end-user reason why this
couldn't be called. The getMarkupId method can in fact be called in
if it is possible to make every method work at constructor, of course
this is not necessary.
It's a good idea to shield people from errors that aren't
i tried to do that, but the sentiment at that time was that throwing an
exception internally was not explicit enough. maybe times have changed.
-igor
On 2/19/07, manu [EMAIL PROTECTED] wrote:
available, and I cannot think of a good end-user reason why this
couldn't be called. The
sure, but does it abort the page or do you end up with an extra page in the
pagemap?
-igor
On 2/19/07, Eelco Hillenius [EMAIL PROTECTED] wrote:
Guys, guys, did you read my reply? I fixed this and you can now use
setResponsePage in your constructor or anywhere you want!. I just made
this
On 2/19/07, Igor Vaynberg [EMAIL PROTECTED] wrote:
sure, but does it abort the page or do you end up with an extra page in the
pagemap?
No, it doesn't abort the request. If you want to do that you need to
use that exception. So while it typically doesn't make sense to call
setResponsePage in a
at least if in devel mode you should log a warning if you detect this. or do
you not detect its in a constructor...you just push a new target?
-igor
On 2/19/07, Eelco Hillenius [EMAIL PROTECTED] wrote:
On 2/19/07, Igor Vaynberg [EMAIL PROTECTED] wrote:
sure, but does it abort the page or do
On 2/19/07, Igor Vaynberg [EMAIL PROTECTED] wrote:
at least if in devel mode you should log a warning if you detect this. or do
you not detect its in a constructor...you just push a new target?
Why? If users want to do that, let them. Though it's probably not the
smartest thing to do, there's
On 2/15/07, Carfield Yim [EMAIL PROTECTED] wrote:
On 2/16/07, Igor Vaynberg [EMAIL PROTECTED] wrote:
you shouldnt be calling setresponsepage inside constructors
I think I fixed this for 1.3 and 2.0. Could you please test this and
let us know whether that worked for you? It *should* work, but
available, and I cannot think of a good end-user reason why this
couldn't be called. The getMarkupId method can in fact be called in
if it is possible to make every method work at constructor, of course
this is not necessary. It just my personal suggestion if this is not
possible
Do we really need a bit? (haven't we already have a 'bit')?
If getmarkupid is called but it isn't set and we want to look it up
then we can already see that the markup is not attached right?
And if that is not the case throw an exception
On 2/17/07, Eelco Hillenius [EMAIL PROTECTED] wrote:
On
On 2/17/07, Carfield Yim [EMAIL PROTECTED] wrote:
available, and I cannot think of a good end-user reason why this
couldn't be called. The getMarkupId method can in fact be called in
if it is possible to make every method work at constructor, of course
this is not necessary. It just my
Hallo,
I have this web-flow sequences (two branches to the same leaf):
[web-flow 1] WebPage1 WebPage2 [+Applet] WebPage3
[web-flow 2] WebPage1B [+Applet] WebPage2 [+Applet] WebPage3
The Applet is the same in all cases: is an Applet that digitally signs
(windows keystore for certs) and
you shouldnt be calling setresponsepage inside constructors
try RestartResponseException
-igor
On 2/15/07, Manu [EMAIL PROTECTED] wrote:
Hallo,
I have this web-flow sequences (two branches to the same leaf):
[web-flow 1] WebPage1 WebPage2 [+Applet] WebPage3
[web-flow 2] WebPage1B
On 2/16/07, Igor Vaynberg [EMAIL PROTECTED] wrote:
you shouldnt be calling setresponsepage inside constructors
try RestartResponseException
-igor
Other than setresponsepage(), there are also methods like
getMarkupId() , which are not work properly in constructor. May be we
can add flag for
20 matches
Mail list logo
