I see... you may have already streamed content --> Maybe same interface serves to mark a component as "do not stream anything till the end", or stream the contents to an intermediate buffer... but what I way above might not make sense as well.
On Mon, Oct 13, 2014 at 4:33 PM, Martin Grigorov <[email protected]> wrote: > On Mon, Oct 13, 2014 at 5:09 PM, Ernesto Reinaldo Barreiro < > [email protected]> wrote: > > > @Martin, > > > > Just a (maybe stupid) idea: > > > > 1- Mark some component as ISafeFail > > 2- If rendering fails take Component_onfail.html and render it > > > > As I said earlier this is not so trivial. > The failure may happen in the middle of the rendering and the content > collected so far in RequestCycle#getResponse() could be invalid HTML. > Appending anything (loaded from a file or generated on the fly) is not > really a solution. > > > > > > On Mon, Oct 13, 2014 at 4:02 PM, Martin Grigorov <[email protected]> > > wrote: > > > > > https://issues.apache.org/jira/browse/WICKET-4321 > > > this is the ticker I meant > > > it suggests to restart the rendering completely for the whole page and > > this > > > is not enough > > > I'll see what kind of changes would be needed to accomplish this. > > > > > > Martin Grigorov > > > Wicket Training and Consulting > > > https://twitter.com/mtgrigorov > > > > > > On Fri, Oct 10, 2014 at 2:29 PM, Martin Grigorov <[email protected] > > > > > wrote: > > > > > > > Hi, > > > > > > > > There is no support for this, even in 7.x. > > > > I remember Carl-Eric Menzel asking for the same functionality before > > ... > > > > Behavior#onException() sounds like something similar but there is no > > way > > > > to suppress the bubbling of the exception at the moment. > > > > The bigger problem is that the rendering can fail in the middle, i.e. > > the > > > > component can have written some response already and then fail. If > the > > > > written response is proper HTML then it is OKish. But if some tag is > > not > > > > closed then the rendering of the complete page may fail. > > > > So if we try to add this functionality we will have to use temporary > > > > Response objects for the rendering of each component to be able to > > throw > > > > away whatever it has produced before failing. > > > > > > > > Usually the problem is related to the component's model. A workaround > > for > > > > you could be to use a wrapper Model that returns "empty data" when > the > > > > underlying model throws an exception. > > > > > > > > Martin Grigorov > > > > Wicket Training and Consulting > > > > https://twitter.com/mtgrigorov > > > > > > > > On Fri, Oct 10, 2014 at 12:59 PM, Ernesto Reinaldo Barreiro < > > > > [email protected]> wrote: > > > > > > > >> Hi, > > > >> > > > >> On Fri, Oct 10, 2014 at 11:50 AM, Tobias Gierke < > > > >> [email protected]> wrote: > > > >> > > > >> > Hi, > > > >> > > > > >> >> Wouldn't it be possible to "embed" the failing prone porlets > > inside > > > >> >> iframes so that each one is a Wicket page? > > > >> >> > > > >> > I already thought about this but the page uses quite a lot of > fancy > > > >> > CSS/Ajax/Javascript (portlets are rendered in a grid with > > configurable > > > >> > row/column count, drag'n'drop to move them around etc.) and I'd > > rather > > > >> not > > > >> > touch the existing code if there's a Java-side only solution ;-) > > > >> > > > > >> > > > >> I do not know of any :-( > > > >> > > > >> Another possibility is build each client entirely on JavaScript and > > use > > > >> Wicket just as a service layer... not very Wicket like but you would > > not > > > >> have this problem. > > > >> > > > >> > > > >> > > > > >> > Cheers, > > > >> > Tobias > > > >> > > > > >> > > > > >> >> On Fri, Oct 10, 2014 at 11:12 AM, Tobias Gierke < > > > >> >> [email protected]> wrote: > > > >> >> > > > >> >> Hi, > > > >> >>> > > > >> >>> In our web application we have a dashboard-like homepage that > > > >> displays a > > > >> >>> number of user-configurable 'portlets' (which are really just > > > ordinary > > > >> >>> Wicket components and have nothing to do with the Portlet spec). > > I'm > > > >> >>> looking for a way of preventing the application from becoming > > > >> unusable in > > > >> >>> case one or more of these portlets continuously fail to render > > > >> because of > > > >> >>> some internal error/bug. > > > >> >>> > > > >> >>> We're currently using a custom RequestCycleListener with the > > > >> >>> onException() > > > >> >>> method redirecting to a generic error page, thus when rendering > > of a > > > >> >>> 'portlet' fails the user will never get to see the homepage and > > > always > > > >> >>> end > > > >> >>> up on the error page - which is obviously not really desirable. > > > >> >>> > > > >> >>> Is there a way to to hook into Wicket's rendering cycle so that > I > > > can > > > >> >>> provide some default markup in case rendering of a component > > > (subtree) > > > >> >>> fails with a RuntimeException ? > > > >> >>> > > > >> >>> I understand that this maybe be very tricky since the component > > > >> subtree > > > >> >>> might've rendered partially and thus internal state will be > > > >> inconsistent. > > > >> >>> It would probably require serializing the initial state of the > > > >> component > > > >> >>> (subtree) before rendering starts and reverting the wholle > subtree > > > to > > > >> its > > > >> >>> initial state once a RuntimeException is thrown. > > > >> >>> > > > >> >>> We're running Wicket 1.5.12. > > > >> >>> > > > >> >>> Thanks in advance, > > > >> >>> Tobias > > > >> >>> > > > >> >>> -- > > > >> >>> Tobias Gierke > > > >> >>> Development > > > >> >>> > > > >> >>> VOIPFUTURE GmbH Wendenstraße 4 20097 Hamburg, Germany > > > >> >>> Phone +49 40 688 900 164 Fax +49 40 688 900 199 > > > >> >>> Email [email protected] Web > > http://www.voipfuture.com > > > >> >>> CEO Jan Bastian > > > >> >>> > > > >> >>> Commercial Court AG Hamburg HRB 109896, VAT ID DE263738086 > > > >> >>> > > > >> >>> > > > >> >>> > > > >> >>> > > > --------------------------------------------------------------------- > > > >> >>> To unsubscribe, e-mail: [email protected] > > > >> >>> For additional commands, e-mail: [email protected] > > > >> >>> > > > >> >>> > > > >> >>> > > > >> >> > > > >> > > > > >> > > > --------------------------------------------------------------------- > > > >> > To unsubscribe, e-mail: [email protected] > > > >> > For additional commands, e-mail: [email protected] > > > >> > > > > >> > > > > >> > > > >> > > > >> -- > > > >> Regards - Ernesto Reinaldo Barreiro > > > >> > > > > > > > > > > > > > > > > > > > -- > > Regards - Ernesto Reinaldo Barreiro > > > -- Regards - Ernesto Reinaldo Barreiro
