NPE on DefaultClientWindow.getWindowId after long period of inactivity
Hi,
I am currently using DeltaSpike 0.6 Snapshot some days old.. working on
Wildfly 8.0.0.Final Snapshot.
I get an NPE in DefaultClientWindow.getWindowId() after long period of
inactivity.
In this situation the application is _totally inaccessible_ from this
client instance, the only remedy is to delete the session cookies and
reset the URL:
2014-02-23 16:59:13,319 ERROR [io.undertow.request] (default task-36)
Servlet request failed HttpServerExchange{ GET
/mssms2/faces/csms/groupManagement.xhtml}: java.lang.NullPoi
nterException
at
org.jboss.weld.context.beanstore.http.AbstractSessionBeanStore.getLockStore(AbstractSessionBeanStore.java:113)
[weld-core-impl-2.1.0.CR1.jar:2013-09-26 16:53]
at
org.jboss.weld.context.beanstore.AttributeBeanStore.lock(AttributeBeanStore.java:210)
[weld-core-impl-2.1.0.CR1.jar:2013-09-26 16:53]
at
org.jboss.weld.context.AbstractContext.get(AbstractContext.java:88)
[weld-core-impl-2.1.0.CR1.jar:2013-09-26 16:53]
at
org.jboss.weld.context.PassivatingContextWrapper$AbstractPassivatingContextWrapper.get(PassivatingContextWrapper.java:76)
[weld-core-impl-2.1.0.CR1.jar:2013-09-26 16:
53]
at
org.jboss.weld.bean.proxy.ContextBeanInstance.getInstance(ContextBeanInstance.java:98)
[weld-core-impl-2.1.0.CR1.jar:2013-09-26 16:53]
at
org.jboss.weld.bean.proxy.ProxyMethodHandler.invoke(ProxyMethodHandler.java:78)
[weld-core-impl-2.1.0.CR1.jar:2013-09-26 16:53]
at
org.apache.deltaspike.jsf.spi.scope.window.DefaultClientWindowConfig$Proxy$_$$_WeldClientProxy.getClientWindowRenderMode(Unknown
Source) [deltaspike-jsf-module-api-0.
6-SNAPSHOT.jar:0.6-SNAPSHOT]
at
org.apache.deltaspike.jsf.impl.scope.window.DefaultClientWindow.getWindowId(DefaultClientWindow.java:119)
[deltaspike-jsf-module-impl-0.6-SNAPSHOT.jar:0.6-SNAPSHOT]
at
org.apache.deltaspike.jsf.impl.scope.window.DefaultClientWindow$Proxy$_$$_WeldClientProxy.getWindowId(Unknown
Source) [deltaspike-jsf-module-impl-0.6-SNAPSHOT.jar:0.6
-SNAPSHOT]
at
org.apache.deltaspike.jsf.impl.listener.request.DeltaSpikeLifecycleWrapper.execute(DeltaSpikeLifecycleWrapper.java:78)
[deltaspike-jsf-module-impl-0.6-SNAPSHOT.jar:0.
6-SNAPSHOT]
Re: NPE on DefaultClientWindow.getWindowId after long period of inactivity
Yes, definitely, in my case it has been about a day.
In my opinion the SessionScoped DefaultClientWindowConfig is
instantiated by the web layer, but the Wildfly session storage tries to
resurrect it from passivation, because an ID is provided.
Seems debatable, who's problem it is, but the effect is painful :-/
Am 23.02.2014 19:28, schrieb John D. Ament:
Is the long period of inactivity longer than your session timeout as
defined in your web.xml?
On Sun, Feb 23, 2014 at 11:49 AM, Thomas Frühbeck fruehb...@aon.at wrote:
Hi,
I am currently using DeltaSpike 0.6 Snapshot some days old.. working on
Wildfly 8.0.0.Final Snapshot.
I get an NPE in DefaultClientWindow.getWindowId() after long period of
inactivity.
In this situation the application is _totally inaccessible_ from this client
instance, the only remedy is to delete the session cookies and reset the
URL:
2014-02-23 16:59:13,319 ERROR [io.undertow.request] (default task-36)
Servlet request failed HttpServerExchange{ GET
/mssms2/faces/csms/groupManagement.xhtml}: java.lang.NullPoi
nterException
at
org.jboss.weld.context.beanstore.http.AbstractSessionBeanStore.getLockStore(AbstractSessionBeanStore.java:113)
[weld-core-impl-2.1.0.CR1.jar:2013-09-26 16:53]
at
org.jboss.weld.context.beanstore.AttributeBeanStore.lock(AttributeBeanStore.java:210)
[weld-core-impl-2.1.0.CR1.jar:2013-09-26 16:53]
at
org.jboss.weld.context.AbstractContext.get(AbstractContext.java:88)
[weld-core-impl-2.1.0.CR1.jar:2013-09-26 16:53]
at
org.jboss.weld.context.PassivatingContextWrapper$AbstractPassivatingContextWrapper.get(PassivatingContextWrapper.java:76)
[weld-core-impl-2.1.0.CR1.jar:2013-09-26 16:
53]
at
org.jboss.weld.bean.proxy.ContextBeanInstance.getInstance(ContextBeanInstance.java:98)
[weld-core-impl-2.1.0.CR1.jar:2013-09-26 16:53]
at
org.jboss.weld.bean.proxy.ProxyMethodHandler.invoke(ProxyMethodHandler.java:78)
[weld-core-impl-2.1.0.CR1.jar:2013-09-26 16:53]
at
org.apache.deltaspike.jsf.spi.scope.window.DefaultClientWindowConfig$Proxy$_$$_WeldClientProxy.getClientWindowRenderMode(Unknown
Source) [deltaspike-jsf-module-api-0.
6-SNAPSHOT.jar:0.6-SNAPSHOT]
at
org.apache.deltaspike.jsf.impl.scope.window.DefaultClientWindow.getWindowId(DefaultClientWindow.java:119)
[deltaspike-jsf-module-impl-0.6-SNAPSHOT.jar:0.6-SNAPSHOT]
at
org.apache.deltaspike.jsf.impl.scope.window.DefaultClientWindow$Proxy$_$$_WeldClientProxy.getWindowId(Unknown
Source) [deltaspike-jsf-module-impl-0.6-SNAPSHOT.jar:0.6
-SNAPSHOT]
at
org.apache.deltaspike.jsf.impl.listener.request.DeltaSpikeLifecycleWrapper.execute(DeltaSpikeLifecycleWrapper.java:78)
[deltaspike-jsf-module-impl-0.6-SNAPSHOT.jar:0.
6-SNAPSHOT]
Re: NPE on DefaultClientWindow.getWindowId after long period of inactivity
Well, from a CDI perspective this is pretty clear. We take whatever the Session
is. Regardless if this is an existing one or a new one. Seems to be a WildFly
issue. I'm sure the guys over there can fix this.
LieGrue,
strub
On Sunday, 23 February 2014, 19:37, Thomas Frühbeck fruehb...@aon.at wrote:
Yes, definitely, in my case it has been about a day.
In my opinion the SessionScoped DefaultClientWindowConfig is
instantiated by the web layer, but the Wildfly session storage tries to
resurrect it from passivation, because an ID is provided.
Seems debatable, who's problem it is, but the effect is painful :-/
Am 23.02.2014 19:28, schrieb John D. Ament:
Is the long period of inactivity longer than your session timeout as
defined in your web.xml?
On Sun, Feb 23, 2014 at 11:49 AM, Thomas Frühbeck fruehb...@aon.at wrote:
Hi,
I am currently using DeltaSpike 0.6 Snapshot some days old.. working on
Wildfly 8.0.0.Final Snapshot.
I get an NPE in DefaultClientWindow.getWindowId() after long period of
inactivity.
In this situation the application is _totally inaccessible_ from this client
instance, the only remedy is to delete the session cookies and reset the
URL:
2014-02-23 16:59:13,319 ERROR [io.undertow.request] (default task-36)
Servlet request failed HttpServerExchange{ GET
/mssms2/faces/csms/groupManagement.xhtml}: java.lang.NullPoi
nterException
at
org.jboss.weld.context.beanstore.http.AbstractSessionBeanStore.getLockStore(AbstractSessionBeanStore.java:113)
[weld-core-impl-2.1.0.CR1.jar:2013-09-26 16:53]
at
org.jboss.weld.context.beanstore.AttributeBeanStore.lock(AttributeBeanStore.java:210)
[weld-core-impl-2.1.0.CR1.jar:2013-09-26 16:53]
at
org.jboss.weld.context.AbstractContext.get(AbstractContext.java:88)
[weld-core-impl-2.1.0.CR1.jar:2013-09-26 16:53]
at
org.jboss.weld.context.PassivatingContextWrapper$AbstractPassivatingContextWrapper.get(PassivatingContextWrapper.java:76)
[weld-core-impl-2.1.0.CR1.jar:2013-09-26 16:
53]
at
org.jboss.weld.bean.proxy.ContextBeanInstance.getInstance(ContextBeanInstance.java:98)
[weld-core-impl-2.1.0.CR1.jar:2013-09-26 16:53]
at
org.jboss.weld.bean.proxy.ProxyMethodHandler.invoke(ProxyMethodHandler.java:78)
[weld-core-impl-2.1.0.CR1.jar:2013-09-26 16:53]
at
org.apache.deltaspike.jsf.spi.scope.window.DefaultClientWindowConfig$Proxy$_$$_WeldClientProxy.getClientWindowRenderMode(Unknown
Source) [deltaspike-jsf-module-api-0.
6-SNAPSHOT.jar:0.6-SNAPSHOT]
at
org.apache.deltaspike.jsf.impl.scope.window.DefaultClientWindow.getWindowId(DefaultClientWindow.java:119)
[deltaspike-jsf-module-impl-0.6-SNAPSHOT.jar:0.6-SNAPSHOT]
at
org.apache.deltaspike.jsf.impl.scope.window.DefaultClientWindow$Proxy$_$$_WeldClientProxy.getWindowId(Unknown
Source) [deltaspike-jsf-module-impl-0.6-SNAPSHOT.jar:0.6
-SNAPSHOT]
at
org.apache.deltaspike.jsf.impl.listener.request.DeltaSpikeLifecycleWrapper.execute(DeltaSpikeLifecycleWrapper.java:78)
[deltaspike-jsf-module-impl-0.6-SNAPSHOT.jar:0.
6-SNAPSHOT]
Re: Servlet Module - Do we really need @Web?
Another problem is that @web is a qualifier and @JsfPhaseListener is a stereotype and not a qualifier. So if we would change it, we must use different annotations. Right? IMO @Inject @DeltaSpike ServletContext context; would be great as it's just like a namespace. For the @JsfPhaseListener, something like @Advanced would be better. 2014-02-18 11:16 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com: IMO it's actually more expressive than any other names. Maybe @DeltaSpikeManaged would be more impressive... +1 @DeltaSpike then, before keeping the old redudant/ugly qualifiers. 2014-02-18 10:51 GMT+01:00 Romain Manni-Bucau rmannibu...@gmail.com: +1 kind of tradeoff Romain Manni-Bucau Twitter: @rmannibucau Blog: http://rmannibucau.wordpress.com/ LinkedIn: http://fr.linkedin.com/in/rmannibucau Github: https://github.com/rmannibucau 2014-02-18 10:46 GMT+01:00 Gerhard Petracek gerhard.petra...@gmail.com: we haven't seen a nice name so far - i would keep what we have right now (it's redundant, but at least a bit more expressive). regards, gerhard 2014-02-18 10:20 GMT+01:00 Mark Struberg strub...@yahoo.de: I do agree on @Managed not being very expressive. Think about the @ManagedBean disaster in JavaEE itself. 'Managed' is in the same ballpark like 'Class' or 'Object'. Managed by whom and what? In that case I'd rather go with @DeltaSpike or keep @Web... But I do not care that much about names... LieGrue, strub On Tuesday, 18 February 2014, 10:15, Romain Manni-Bucau rmannibu...@gmail.com wrote: @Gerhard: yeah but managed is not expressive, doesnt give the origin + is quite standard for jmxso i would avoid it. That said it doesnt bring any feature so i dont want to fight for a name. Romain Manni-Bucau Twitter: @rmannibucau Blog: http://rmannibucau.wordpress.com/ LinkedIn: http://fr.linkedin.com/in/rmannibucau Github: https://github.com/rmannibucau 2014-02-18 10:12 GMT+01:00 Gerhard Petracek gerhard.petra...@gmail.com: @romain: you will always have same/similar terms in different areas (see e.g. message). regards, gerhard
Re: NPE on DefaultClientWindow.getWindowId after long period of inactivity
H
You mention WildFly 8 Final Snapshot. Are you using one of their
prerelease snapshot builds or the final one? Can you try on the
final?
On Sun, Feb 23, 2014 at 1:44 PM, Mark Struberg strub...@yahoo.de wrote:
Well, from a CDI perspective this is pretty clear. We take whatever the
Session is. Regardless if this is an existing one or a new one. Seems to be a
WildFly issue. I'm sure the guys over there can fix this.
LieGrue,
strub
On Sunday, 23 February 2014, 19:37, Thomas Frühbeck fruehb...@aon.at wrote:
Yes, definitely, in my case it has been about a day.
In my opinion the SessionScoped DefaultClientWindowConfig is
instantiated by the web layer, but the Wildfly session storage tries to
resurrect it from passivation, because an ID is provided.
Seems debatable, who's problem it is, but the effect is painful :-/
Am 23.02.2014 19:28, schrieb John D. Ament:
Is the long period of inactivity longer than your session timeout as
defined in your web.xml?
On Sun, Feb 23, 2014 at 11:49 AM, Thomas Frühbeck fruehb...@aon.at wrote:
Hi,
I am currently using DeltaSpike 0.6 Snapshot some days old.. working on
Wildfly 8.0.0.Final Snapshot.
I get an NPE in DefaultClientWindow.getWindowId() after long period of
inactivity.
In this situation the application is _totally inaccessible_ from this
client
instance, the only remedy is to delete the session cookies and reset the
URL:
2014-02-23 16:59:13,319 ERROR [io.undertow.request] (default task-36)
Servlet request failed HttpServerExchange{ GET
/mssms2/faces/csms/groupManagement.xhtml}: java.lang.NullPoi
nterException
at
org.jboss.weld.context.beanstore.http.AbstractSessionBeanStore.getLockStore(AbstractSessionBeanStore.java:113)
[weld-core-impl-2.1.0.CR1.jar:2013-09-26 16:53]
at
org.jboss.weld.context.beanstore.AttributeBeanStore.lock(AttributeBeanStore.java:210)
[weld-core-impl-2.1.0.CR1.jar:2013-09-26 16:53]
at
org.jboss.weld.context.AbstractContext.get(AbstractContext.java:88)
[weld-core-impl-2.1.0.CR1.jar:2013-09-26 16:53]
at
org.jboss.weld.context.PassivatingContextWrapper$AbstractPassivatingContextWrapper.get(PassivatingContextWrapper.java:76)
[weld-core-impl-2.1.0.CR1.jar:2013-09-26 16:
53]
at
org.jboss.weld.bean.proxy.ContextBeanInstance.getInstance(ContextBeanInstance.java:98)
[weld-core-impl-2.1.0.CR1.jar:2013-09-26 16:53]
at
org.jboss.weld.bean.proxy.ProxyMethodHandler.invoke(ProxyMethodHandler.java:78)
[weld-core-impl-2.1.0.CR1.jar:2013-09-26 16:53]
at
org.apache.deltaspike.jsf.spi.scope.window.DefaultClientWindowConfig$Proxy$_$$_WeldClientProxy.getClientWindowRenderMode(Unknown
Source) [deltaspike-jsf-module-api-0.
6-SNAPSHOT.jar:0.6-SNAPSHOT]
at
org.apache.deltaspike.jsf.impl.scope.window.DefaultClientWindow.getWindowId(DefaultClientWindow.java:119)
[deltaspike-jsf-module-impl-0.6-SNAPSHOT.jar:0.6-SNAPSHOT]
at
org.apache.deltaspike.jsf.impl.scope.window.DefaultClientWindow$Proxy$_$$_WeldClientProxy.getWindowId(Unknown
Source) [deltaspike-jsf-module-impl-0.6-SNAPSHOT.jar:0.6
-SNAPSHOT]
at
org.apache.deltaspike.jsf.impl.listener.request.DeltaSpikeLifecycleWrapper.execute(DeltaSpikeLifecycleWrapper.java:78)
[deltaspike-jsf-module-impl-0.6-SNAPSHOT.jar:0.
6-SNAPSHOT]
Re: Servlet Module - Do we really need @Web?
it's just a stereotype to get a better tool-support... @Advanced is also a qualifier (at least in codi). however, we need ordinal or we add @InvocationOrder as it was in codi. once we do that, we break the consistency in that area. regards, gerhard http://www.irian.at Your JSF/JavaEE powerhouse - JavaEE Consulting, Development and Courses in English and German Professional Support for Apache MyFaces 2014-02-23 21:41 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com: Another problem is that @web is a qualifier and @JsfPhaseListener is a stereotype and not a qualifier. So if we would change it, we must use different annotations. Right? IMO @Inject @DeltaSpike ServletContext context; would be great as it's just like a namespace. For the @JsfPhaseListener, something like @Advanced would be better. 2014-02-18 11:16 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com : IMO it's actually more expressive than any other names. Maybe @DeltaSpikeManaged would be more impressive... +1 @DeltaSpike then, before keeping the old redudant/ugly qualifiers. 2014-02-18 10:51 GMT+01:00 Romain Manni-Bucau rmannibu...@gmail.com: +1 kind of tradeoff Romain Manni-Bucau Twitter: @rmannibucau Blog: http://rmannibucau.wordpress.com/ LinkedIn: http://fr.linkedin.com/in/rmannibucau Github: https://github.com/rmannibucau 2014-02-18 10:46 GMT+01:00 Gerhard Petracek gerhard.petra...@gmail.com : we haven't seen a nice name so far - i would keep what we have right now (it's redundant, but at least a bit more expressive). regards, gerhard 2014-02-18 10:20 GMT+01:00 Mark Struberg strub...@yahoo.de: I do agree on @Managed not being very expressive. Think about the @ManagedBean disaster in JavaEE itself. 'Managed' is in the same ballpark like 'Class' or 'Object'. Managed by whom and what? In that case I'd rather go with @DeltaSpike or keep @Web... But I do not care that much about names... LieGrue, strub On Tuesday, 18 February 2014, 10:15, Romain Manni-Bucau rmannibu...@gmail.com wrote: @Gerhard: yeah but managed is not expressive, doesnt give the origin + is quite standard for jmxso i would avoid it. That said it doesnt bring any feature so i dont want to fight for a name. Romain Manni-Bucau Twitter: @rmannibucau Blog: http://rmannibucau.wordpress.com/ LinkedIn: http://fr.linkedin.com/in/rmannibucau Github: https://github.com/rmannibucau 2014-02-18 10:12 GMT+01:00 Gerhard Petracek gerhard.petra...@gmail.com: @romain: you will always have same/similar terms in different areas (see e.g. message). regards, gerhard
Re: Servlet Module - Do we really need @Web?
What about reusing javax.annotation.Priority? 2014-02-23 22:02 GMT+01:00 Gerhard Petracek gerhard.petra...@gmail.com: it's just a stereotype to get a better tool-support... @Advanced is also a qualifier (at least in codi). however, we need ordinal or we add @InvocationOrder as it was in codi. once we do that, we break the consistency in that area. regards, gerhard http://www.irian.at Your JSF/JavaEE powerhouse - JavaEE Consulting, Development and Courses in English and German Professional Support for Apache MyFaces 2014-02-23 21:41 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com : Another problem is that @web is a qualifier and @JsfPhaseListener is a stereotype and not a qualifier. So if we would change it, we must use different annotations. Right? IMO @Inject @DeltaSpike ServletContext context; would be great as it's just like a namespace. For the @JsfPhaseListener, something like @Advanced would be better. 2014-02-18 11:16 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com : IMO it's actually more expressive than any other names. Maybe @DeltaSpikeManaged would be more impressive... +1 @DeltaSpike then, before keeping the old redudant/ugly qualifiers. 2014-02-18 10:51 GMT+01:00 Romain Manni-Bucau rmannibu...@gmail.com: +1 kind of tradeoff Romain Manni-Bucau Twitter: @rmannibucau Blog: http://rmannibucau.wordpress.com/ LinkedIn: http://fr.linkedin.com/in/rmannibucau Github: https://github.com/rmannibucau 2014-02-18 10:46 GMT+01:00 Gerhard Petracek gerhard.petra...@gmail.com : we haven't seen a nice name so far - i would keep what we have right now (it's redundant, but at least a bit more expressive). regards, gerhard 2014-02-18 10:20 GMT+01:00 Mark Struberg strub...@yahoo.de: I do agree on @Managed not being very expressive. Think about the @ManagedBean disaster in JavaEE itself. 'Managed' is in the same ballpark like 'Class' or 'Object'. Managed by whom and what? In that case I'd rather go with @DeltaSpike or keep @Web... But I do not care that much about names... LieGrue, strub On Tuesday, 18 February 2014, 10:15, Romain Manni-Bucau rmannibu...@gmail.com wrote: @Gerhard: yeah but managed is not expressive, doesnt give the origin + is quite standard for jmxso i would avoid it. That said it doesnt bring any feature so i dont want to fight for a name. Romain Manni-Bucau Twitter: @rmannibucau Blog: http://rmannibucau.wordpress.com/ LinkedIn: http://fr.linkedin.com/in/rmannibucau Github: https://github.com/rmannibucau 2014-02-18 10:12 GMT+01:00 Gerhard Petracek gerhard.petra...@gmail.com: @romain: you will always have same/similar terms in different areas (see e.g. message). regards, gerhard
Re: Servlet Module - Do we really need @Web?
That's would be ambiguous with JavaEE 6 Romain Manni-Bucau Twitter: @rmannibucau Blog: http://rmannibucau.wordpress.com/ LinkedIn: http://fr.linkedin.com/in/rmannibucau Github: https://github.com/rmannibucau 2014-02-23 22:12 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com: What about reusing javax.annotation.Priority? 2014-02-23 22:02 GMT+01:00 Gerhard Petracek gerhard.petra...@gmail.com: it's just a stereotype to get a better tool-support... @Advanced is also a qualifier (at least in codi). however, we need ordinal or we add @InvocationOrder as it was in codi. once we do that, we break the consistency in that area. regards, gerhard http://www.irian.at Your JSF/JavaEE powerhouse - JavaEE Consulting, Development and Courses in English and German Professional Support for Apache MyFaces 2014-02-23 21:41 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com : Another problem is that @web is a qualifier and @JsfPhaseListener is a stereotype and not a qualifier. So if we would change it, we must use different annotations. Right? IMO @Inject @DeltaSpike ServletContext context; would be great as it's just like a namespace. For the @JsfPhaseListener, something like @Advanced would be better. 2014-02-18 11:16 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com : IMO it's actually more expressive than any other names. Maybe @DeltaSpikeManaged would be more impressive... +1 @DeltaSpike then, before keeping the old redudant/ugly qualifiers. 2014-02-18 10:51 GMT+01:00 Romain Manni-Bucau rmannibu...@gmail.com: +1 kind of tradeoff Romain Manni-Bucau Twitter: @rmannibucau Blog: http://rmannibucau.wordpress.com/ LinkedIn: http://fr.linkedin.com/in/rmannibucau Github: https://github.com/rmannibucau 2014-02-18 10:46 GMT+01:00 Gerhard Petracek gerhard.petra...@gmail.com : we haven't seen a nice name so far - i would keep what we have right now (it's redundant, but at least a bit more expressive). regards, gerhard 2014-02-18 10:20 GMT+01:00 Mark Struberg strub...@yahoo.de: I do agree on @Managed not being very expressive. Think about the @ManagedBean disaster in JavaEE itself. 'Managed' is in the same ballpark like 'Class' or 'Object'. Managed by whom and what? In that case I'd rather go with @DeltaSpike or keep @Web... But I do not care that much about names... LieGrue, strub On Tuesday, 18 February 2014, 10:15, Romain Manni-Bucau rmannibu...@gmail.com wrote: @Gerhard: yeah but managed is not expressive, doesnt give the origin + is quite standard for jmxso i would avoid it. That said it doesnt bring any feature so i dont want to fight for a name. Romain Manni-Bucau Twitter: @rmannibucau Blog: http://rmannibucau.wordpress.com/ LinkedIn: http://fr.linkedin.com/in/rmannibucau Github: https://github.com/rmannibucau 2014-02-18 10:12 GMT+01:00 Gerhard Petracek gerhard.petra...@gmail.com: @romain: you will always have same/similar terms in different areas (see e.g. message). regards, gerhard
Re: Servlet Module - Do we really need @Web?
Hmm... We could clone it for JEE6 support and handle both annotations for backward compatibility. Maybe we could also use it for other features (or later..) 2014-02-23 22:14 GMT+01:00 Romain Manni-Bucau rmannibu...@gmail.com: That's would be ambiguous with JavaEE 6 Romain Manni-Bucau Twitter: @rmannibucau Blog: http://rmannibucau.wordpress.com/ LinkedIn: http://fr.linkedin.com/in/rmannibucau Github: https://github.com/rmannibucau 2014-02-23 22:12 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com : What about reusing javax.annotation.Priority? 2014-02-23 22:02 GMT+01:00 Gerhard Petracek gerhard.petra...@gmail.com : it's just a stereotype to get a better tool-support... @Advanced is also a qualifier (at least in codi). however, we need ordinal or we add @InvocationOrder as it was in codi. once we do that, we break the consistency in that area. regards, gerhard http://www.irian.at Your JSF/JavaEE powerhouse - JavaEE Consulting, Development and Courses in English and German Professional Support for Apache MyFaces 2014-02-23 21:41 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com : Another problem is that @web is a qualifier and @JsfPhaseListener is a stereotype and not a qualifier. So if we would change it, we must use different annotations. Right? IMO @Inject @DeltaSpike ServletContext context; would be great as it's just like a namespace. For the @JsfPhaseListener, something like @Advanced would be better. 2014-02-18 11:16 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com : IMO it's actually more expressive than any other names. Maybe @DeltaSpikeManaged would be more impressive... +1 @DeltaSpike then, before keeping the old redudant/ugly qualifiers. 2014-02-18 10:51 GMT+01:00 Romain Manni-Bucau rmannibu...@gmail.com: +1 kind of tradeoff Romain Manni-Bucau Twitter: @rmannibucau Blog: http://rmannibucau.wordpress.com/ LinkedIn: http://fr.linkedin.com/in/rmannibucau Github: https://github.com/rmannibucau 2014-02-18 10:46 GMT+01:00 Gerhard Petracek gerhard.petra...@gmail.com : we haven't seen a nice name so far - i would keep what we have right now (it's redundant, but at least a bit more expressive). regards, gerhard 2014-02-18 10:20 GMT+01:00 Mark Struberg strub...@yahoo.de: I do agree on @Managed not being very expressive. Think about the @ManagedBean disaster in JavaEE itself. 'Managed' is in the same ballpark like 'Class' or 'Object'. Managed by whom and what? In that case I'd rather go with @DeltaSpike or keep @Web... But I do not care that much about names... LieGrue, strub On Tuesday, 18 February 2014, 10:15, Romain Manni-Bucau rmannibu...@gmail.com wrote: @Gerhard: yeah but managed is not expressive, doesnt give the origin + is quite standard for jmxso i would avoid it. That said it doesnt bring any feature so i dont want to fight for a name. Romain Manni-Bucau Twitter: @rmannibucau Blog: http://rmannibucau.wordpress.com/ LinkedIn: http://fr.linkedin.com/in/rmannibucau Github: https://github.com/rmannibucau 2014-02-18 10:12 GMT+01:00 Gerhard Petracek gerhard.petra...@gmail.com: @romain: you will always have same/similar terms in different areas (see e.g. message). regards, gerhard
Re: Servlet Module - Do we really need @Web?
I'd prefer something more explicit. Just changing the package would be enough for me Romain Manni-Bucau Twitter: @rmannibucau Blog: http://rmannibucau.wordpress.com/ LinkedIn: http://fr.linkedin.com/in/rmannibucau Github: https://github.com/rmannibucau 2014-02-23 22:17 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com: Hmm... We could clone it for JEE6 support and handle both annotations for backward compatibility. Maybe we could also use it for other features (or later..) 2014-02-23 22:14 GMT+01:00 Romain Manni-Bucau rmannibu...@gmail.com: That's would be ambiguous with JavaEE 6 Romain Manni-Bucau Twitter: @rmannibucau Blog: http://rmannibucau.wordpress.com/ LinkedIn: http://fr.linkedin.com/in/rmannibucau Github: https://github.com/rmannibucau 2014-02-23 22:12 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com : What about reusing javax.annotation.Priority? 2014-02-23 22:02 GMT+01:00 Gerhard Petracek gerhard.petra...@gmail.com : it's just a stereotype to get a better tool-support... @Advanced is also a qualifier (at least in codi). however, we need ordinal or we add @InvocationOrder as it was in codi. once we do that, we break the consistency in that area. regards, gerhard http://www.irian.at Your JSF/JavaEE powerhouse - JavaEE Consulting, Development and Courses in English and German Professional Support for Apache MyFaces 2014-02-23 21:41 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com : Another problem is that @web is a qualifier and @JsfPhaseListener is a stereotype and not a qualifier. So if we would change it, we must use different annotations. Right? IMO @Inject @DeltaSpike ServletContext context; would be great as it's just like a namespace. For the @JsfPhaseListener, something like @Advanced would be better. 2014-02-18 11:16 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com : IMO it's actually more expressive than any other names. Maybe @DeltaSpikeManaged would be more impressive... +1 @DeltaSpike then, before keeping the old redudant/ugly qualifiers. 2014-02-18 10:51 GMT+01:00 Romain Manni-Bucau rmannibu...@gmail.com: +1 kind of tradeoff Romain Manni-Bucau Twitter: @rmannibucau Blog: http://rmannibucau.wordpress.com/ LinkedIn: http://fr.linkedin.com/in/rmannibucau Github: https://github.com/rmannibucau 2014-02-18 10:46 GMT+01:00 Gerhard Petracek gerhard.petra...@gmail.com : we haven't seen a nice name so far - i would keep what we have right now (it's redundant, but at least a bit more expressive). regards, gerhard 2014-02-18 10:20 GMT+01:00 Mark Struberg strub...@yahoo.de: I do agree on @Managed not being very expressive. Think about the @ManagedBean disaster in JavaEE itself. 'Managed' is in the same ballpark like 'Class' or 'Object'. Managed by whom and what? In that case I'd rather go with @DeltaSpike or keep @Web... But I do not care that much about names... LieGrue, strub On Tuesday, 18 February 2014, 10:15, Romain Manni-Bucau rmannibu...@gmail.com wrote: @Gerhard: yeah but managed is not expressive, doesnt give the origin + is quite standard for jmxso i would avoid it. That said it doesnt bring any feature so i dont want to fight for a name. Romain Manni-Bucau Twitter: @rmannibucau Blog: http://rmannibucau.wordpress.com/ LinkedIn: http://fr.linkedin.com/in/rmannibucau Github: https://github.com/rmannibucau 2014-02-18 10:12 GMT+01:00 Gerhard Petracek gerhard.petra...@gmail.com: @romain: you will always have same/similar terms in different areas (see e.g. message). regards, gerhard
Re: Servlet Module - Do we really need @Web?
Could you explain this? I can't follow you. 2014-02-23 22:23 GMT+01:00 Romain Manni-Bucau rmannibu...@gmail.com: I'd prefer something more explicit. Just changing the package would be enough for me Romain Manni-Bucau Twitter: @rmannibucau Blog: http://rmannibucau.wordpress.com/ LinkedIn: http://fr.linkedin.com/in/rmannibucau Github: https://github.com/rmannibucau 2014-02-23 22:17 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com : Hmm... We could clone it for JEE6 support and handle both annotations for backward compatibility. Maybe we could also use it for other features (or later..) 2014-02-23 22:14 GMT+01:00 Romain Manni-Bucau rmannibu...@gmail.com: That's would be ambiguous with JavaEE 6 Romain Manni-Bucau Twitter: @rmannibucau Blog: http://rmannibucau.wordpress.com/ LinkedIn: http://fr.linkedin.com/in/rmannibucau Github: https://github.com/rmannibucau 2014-02-23 22:12 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com : What about reusing javax.annotation.Priority? 2014-02-23 22:02 GMT+01:00 Gerhard Petracek gerhard.petra...@gmail.com : it's just a stereotype to get a better tool-support... @Advanced is also a qualifier (at least in codi). however, we need ordinal or we add @InvocationOrder as it was in codi. once we do that, we break the consistency in that area. regards, gerhard http://www.irian.at Your JSF/JavaEE powerhouse - JavaEE Consulting, Development and Courses in English and German Professional Support for Apache MyFaces 2014-02-23 21:41 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com : Another problem is that @web is a qualifier and @JsfPhaseListener is a stereotype and not a qualifier. So if we would change it, we must use different annotations. Right? IMO @Inject @DeltaSpike ServletContext context; would be great as it's just like a namespace. For the @JsfPhaseListener, something like @Advanced would be better. 2014-02-18 11:16 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com : IMO it's actually more expressive than any other names. Maybe @DeltaSpikeManaged would be more impressive... +1 @DeltaSpike then, before keeping the old redudant/ugly qualifiers. 2014-02-18 10:51 GMT+01:00 Romain Manni-Bucau rmannibu...@gmail.com: +1 kind of tradeoff Romain Manni-Bucau Twitter: @rmannibucau Blog: http://rmannibucau.wordpress.com/ LinkedIn: http://fr.linkedin.com/in/rmannibucau Github: https://github.com/rmannibucau 2014-02-18 10:46 GMT+01:00 Gerhard Petracek gerhard.petra...@gmail.com : we haven't seen a nice name so far - i would keep what we have right now (it's redundant, but at least a bit more expressive). regards, gerhard 2014-02-18 10:20 GMT+01:00 Mark Struberg strub...@yahoo.de: I do agree on @Managed not being very expressive. Think about the @ManagedBean disaster in JavaEE itself. 'Managed' is in the same ballpark like 'Class' or 'Object'. Managed by whom and what? In that case I'd rather go with @DeltaSpike or keep @Web... But I do not care that much about names... LieGrue, strub On Tuesday, 18 February 2014, 10:15, Romain Manni-Bucau rmannibu...@gmail.com wrote: @Gerhard: yeah but managed is not expressive, doesnt give the origin + is quite standard for jmxso i would avoid it. That said it doesnt bring any feature so i dont want to fight for a name. Romain Manni-Bucau Twitter: @rmannibucau Blog: http://rmannibucau.wordpress.com/ LinkedIn: http://fr.linkedin.com/in/rmannibucau Github: https://github.com/rmannibucau 2014-02-18 10:12 GMT+01:00 Gerhard Petracek gerhard.petra...@gmail.com: @romain: you will always have same/similar terms in different areas (see e.g. message). regards, gerhard
Re: NPE on DefaultClientWindow.getWindowId after long period of inactivity
In fact I found the problem with an 8.0.0.CR1 version, and did not
succeed in reproducing the issue by manipulating the windowId cookie.
I have activated JSF 2.2 WindowId in web.xml to work around a bug in url
handling of DS 0.6 snapshot some time ago, but I cannot imagine, that
this has any relevance in this case, because the exception happens when
DS DefaultClientWindow tries to detect the WindowId handling config..
param-namejavax.faces.CLIENT_WINDOW_MODE/param-name
param-valueurl/param-value
Unfortunately I yet do not have any clue, how to reproduce the problem
else than waiting _very_(??) long :-/
Not really sure, how URL jfwid and cookies work together..
But anycase I will upgrade to 8.0.0.Final ASAP - or 8.0.1.Final, or... :-)
Thomas
Am 23.02.2014 22:02, schrieb John D. Ament:
H
You mention WildFly 8 Final Snapshot. Are you using one of their
prerelease snapshot builds or the final one? Can you try on the
final?
On Sun, Feb 23, 2014 at 1:44 PM, Mark Struberg strub...@yahoo.de wrote:
Well, from a CDI perspective this is pretty clear. We take whatever the Session
is. Regardless if this is an existing one or a new one. Seems to be a WildFly
issue. I'm sure the guys over there can fix this.
LieGrue,
strub
On Sunday, 23 February 2014, 19:37, Thomas Frühbeck fruehb...@aon.at wrote:
Yes, definitely, in my case it has been about a day.
In my opinion the SessionScoped DefaultClientWindowConfig is
instantiated by the web layer, but the Wildfly session storage tries to
resurrect it from passivation, because an ID is provided.
Seems debatable, who's problem it is, but the effect is painful :-/
Am 23.02.2014 19:28, schrieb John D. Ament:
Is the long period of inactivity longer than your session timeout as
defined in your web.xml?
On Sun, Feb 23, 2014 at 11:49 AM, Thomas Frühbeck fruehb...@aon.at wrote:
Hi,
I am currently using DeltaSpike 0.6 Snapshot some days old.. working on
Wildfly 8.0.0.Final Snapshot.
I get an NPE in DefaultClientWindow.getWindowId() after long period of
inactivity.
In this situation the application is _totally inaccessible_ from this client
instance, the only remedy is to delete the session cookies and reset the
URL:
2014-02-23 16:59:13,319 ERROR [io.undertow.request] (default task-36)
Servlet request failed HttpServerExchange{ GET
/mssms2/faces/csms/groupManagement.xhtml}: java.lang.NullPoi
nterException
at
org.jboss.weld.context.beanstore.http.AbstractSessionBeanStore.getLockStore(AbstractSessionBeanStore.java:113)
[weld-core-impl-2.1.0.CR1.jar:2013-09-26 16:53]
at
org.jboss.weld.context.beanstore.AttributeBeanStore.lock(AttributeBeanStore.java:210)
[weld-core-impl-2.1.0.CR1.jar:2013-09-26 16:53]
at
org.jboss.weld.context.AbstractContext.get(AbstractContext.java:88)
[weld-core-impl-2.1.0.CR1.jar:2013-09-26 16:53]
at
org.jboss.weld.context.PassivatingContextWrapper$AbstractPassivatingContextWrapper.get(PassivatingContextWrapper.java:76)
[weld-core-impl-2.1.0.CR1.jar:2013-09-26 16:
53]
at
org.jboss.weld.bean.proxy.ContextBeanInstance.getInstance(ContextBeanInstance.java:98)
[weld-core-impl-2.1.0.CR1.jar:2013-09-26 16:53]
at
org.jboss.weld.bean.proxy.ProxyMethodHandler.invoke(ProxyMethodHandler.java:78)
[weld-core-impl-2.1.0.CR1.jar:2013-09-26 16:53]
at
org.apache.deltaspike.jsf.spi.scope.window.DefaultClientWindowConfig$Proxy$_$$_WeldClientProxy.getClientWindowRenderMode(Unknown
Source) [deltaspike-jsf-module-api-0.
6-SNAPSHOT.jar:0.6-SNAPSHOT]
at
org.apache.deltaspike.jsf.impl.scope.window.DefaultClientWindow.getWindowId(DefaultClientWindow.java:119)
[deltaspike-jsf-module-impl-0.6-SNAPSHOT.jar:0.6-SNAPSHOT]
at
org.apache.deltaspike.jsf.impl.scope.window.DefaultClientWindow$Proxy$_$$_WeldClientProxy.getWindowId(Unknown
Source) [deltaspike-jsf-module-impl-0.6-SNAPSHOT.jar:0.6
-SNAPSHOT]
at
org.apache.deltaspike.jsf.impl.listener.request.DeltaSpikeLifecycleWrapper.execute(DeltaSpikeLifecycleWrapper.java:78)
[deltaspike-jsf-module-impl-0.6-SNAPSHOT.jar:0.
6-SNAPSHOT]
Re: Servlet Module - Do we really need @Web?
using javax.* from deltaspike is quite ambiguous: is it a DS behavior or a spec one? I'm no opposed Priority but in org.apache.deltaspile...Priority Romain Manni-Bucau Twitter: @rmannibucau Blog: http://rmannibucau.wordpress.com/ LinkedIn: http://fr.linkedin.com/in/rmannibucau Github: https://github.com/rmannibucau 2014-02-23 22:34 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com: Could you explain this? I can't follow you. 2014-02-23 22:23 GMT+01:00 Romain Manni-Bucau rmannibu...@gmail.com: I'd prefer something more explicit. Just changing the package would be enough for me Romain Manni-Bucau Twitter: @rmannibucau Blog: http://rmannibucau.wordpress.com/ LinkedIn: http://fr.linkedin.com/in/rmannibucau Github: https://github.com/rmannibucau 2014-02-23 22:17 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com : Hmm... We could clone it for JEE6 support and handle both annotations for backward compatibility. Maybe we could also use it for other features (or later..) 2014-02-23 22:14 GMT+01:00 Romain Manni-Bucau rmannibu...@gmail.com: That's would be ambiguous with JavaEE 6 Romain Manni-Bucau Twitter: @rmannibucau Blog: http://rmannibucau.wordpress.com/ LinkedIn: http://fr.linkedin.com/in/rmannibucau Github: https://github.com/rmannibucau 2014-02-23 22:12 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com : What about reusing javax.annotation.Priority? 2014-02-23 22:02 GMT+01:00 Gerhard Petracek gerhard.petra...@gmail.com : it's just a stereotype to get a better tool-support... @Advanced is also a qualifier (at least in codi). however, we need ordinal or we add @InvocationOrder as it was in codi. once we do that, we break the consistency in that area. regards, gerhard http://www.irian.at Your JSF/JavaEE powerhouse - JavaEE Consulting, Development and Courses in English and German Professional Support for Apache MyFaces 2014-02-23 21:41 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com : Another problem is that @web is a qualifier and @JsfPhaseListener is a stereotype and not a qualifier. So if we would change it, we must use different annotations. Right? IMO @Inject @DeltaSpike ServletContext context; would be great as it's just like a namespace. For the @JsfPhaseListener, something like @Advanced would be better. 2014-02-18 11:16 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com : IMO it's actually more expressive than any other names. Maybe @DeltaSpikeManaged would be more impressive... +1 @DeltaSpike then, before keeping the old redudant/ugly qualifiers. 2014-02-18 10:51 GMT+01:00 Romain Manni-Bucau rmannibu...@gmail.com: +1 kind of tradeoff Romain Manni-Bucau Twitter: @rmannibucau Blog: http://rmannibucau.wordpress.com/ LinkedIn: http://fr.linkedin.com/in/rmannibucau Github: https://github.com/rmannibucau 2014-02-18 10:46 GMT+01:00 Gerhard Petracek gerhard.petra...@gmail.com : we haven't seen a nice name so far - i would keep what we have right now (it's redundant, but at least a bit more expressive). regards, gerhard 2014-02-18 10:20 GMT+01:00 Mark Struberg strub...@yahoo.de: I do agree on @Managed not being very expressive. Think about the @ManagedBean disaster in JavaEE itself. 'Managed' is in the same ballpark like 'Class' or 'Object'. Managed by whom and what? In that case I'd rather go with @DeltaSpike or keep @Web... But I do not care that much about names... LieGrue, strub On Tuesday, 18 February 2014, 10:15, Romain Manni-Bucau rmannibu...@gmail.com wrote: @Gerhard: yeah but managed is not expressive, doesnt give the origin + is quite standard for jmxso i would avoid it. That said it doesnt bring any feature so i dont want to fight for a name. Romain Manni-Bucau Twitter: @rmannibucau Blog: http://rmannibucau.wordpress.com/ LinkedIn: http://fr.linkedin.com/in/rmannibucau Github: https://github.com/rmannibucau 2014-02-18 10:12 GMT+01:00 Gerhard Petracek gerhard.petra...@gmail.com: @romain: you will always have same/similar terms in different areas (see e.g. message). regards, gerhard
Re: Servlet Module - Do we really need @Web?
which consistency gerhard? 2014-02-23 22:45 GMT+01:00 Gerhard Petracek gerhard.petra...@gmail.com: independent of the name, we would break the consistency. regards, gerhard http://www.irian.at Your JSF/JavaEE powerhouse - JavaEE Consulting, Development and Courses in English and German Professional Support for Apache MyFaces 2014-02-23 22:40 GMT+01:00 Romain Manni-Bucau rmannibu...@gmail.com: using javax.* from deltaspike is quite ambiguous: is it a DS behavior or a spec one? I'm no opposed Priority but in org.apache.deltaspile...Priority Romain Manni-Bucau Twitter: @rmannibucau Blog: http://rmannibucau.wordpress.com/ LinkedIn: http://fr.linkedin.com/in/rmannibucau Github: https://github.com/rmannibucau 2014-02-23 22:34 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com : Could you explain this? I can't follow you. 2014-02-23 22:23 GMT+01:00 Romain Manni-Bucau rmannibu...@gmail.com: I'd prefer something more explicit. Just changing the package would be enough for me Romain Manni-Bucau Twitter: @rmannibucau Blog: http://rmannibucau.wordpress.com/ LinkedIn: http://fr.linkedin.com/in/rmannibucau Github: https://github.com/rmannibucau 2014-02-23 22:17 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com : Hmm... We could clone it for JEE6 support and handle both annotations for backward compatibility. Maybe we could also use it for other features (or later..) 2014-02-23 22:14 GMT+01:00 Romain Manni-Bucau rmannibu...@gmail.com : That's would be ambiguous with JavaEE 6 Romain Manni-Bucau Twitter: @rmannibucau Blog: http://rmannibucau.wordpress.com/ LinkedIn: http://fr.linkedin.com/in/rmannibucau Github: https://github.com/rmannibucau 2014-02-23 22:12 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com : What about reusing javax.annotation.Priority? 2014-02-23 22:02 GMT+01:00 Gerhard Petracek gerhard.petra...@gmail.com : it's just a stereotype to get a better tool-support... @Advanced is also a qualifier (at least in codi). however, we need ordinal or we add @InvocationOrder as it was in codi. once we do that, we break the consistency in that area. regards, gerhard http://www.irian.at Your JSF/JavaEE powerhouse - JavaEE Consulting, Development and Courses in English and German Professional Support for Apache MyFaces 2014-02-23 21:41 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com : Another problem is that @web is a qualifier and @JsfPhaseListener is a stereotype and not a qualifier. So if we would change it, we must use different annotations. Right? IMO @Inject @DeltaSpike ServletContext context; would be great as it's just like a namespace. For the @JsfPhaseListener, something like @Advanced would be better. 2014-02-18 11:16 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com : IMO it's actually more expressive than any other names. Maybe @DeltaSpikeManaged would be more impressive... +1 @DeltaSpike then, before keeping the old redudant/ugly qualifiers. 2014-02-18 10:51 GMT+01:00 Romain Manni-Bucau rmannibu...@gmail.com: +1 kind of tradeoff Romain Manni-Bucau Twitter: @rmannibucau Blog: http://rmannibucau.wordpress.com/ LinkedIn: http://fr.linkedin.com/in/rmannibucau Github: https://github.com/rmannibucau 2014-02-18 10:46 GMT+01:00 Gerhard Petracek gerhard.petra...@gmail.com : we haven't seen a nice name so far - i would keep what we have right now (it's redundant, but at least a bit more expressive). regards, gerhard 2014-02-18 10:20 GMT+01:00 Mark Struberg strub...@yahoo.de: I do agree on @Managed not being very expressive. Think about the @ManagedBean disaster in JavaEE itself. 'Managed' is in the same ballpark like 'Class' or 'Object'. Managed by whom and what? In that case I'd rather go with @DeltaSpike or keep @Web... But I do not care that much about names... LieGrue, strub On Tuesday, 18 February 2014, 10:15, Romain Manni-Bucau rmannibu...@gmail.com wrote: @Gerhard: yeah but managed is not expressive, doesnt give the origin + is quite standard for jmxso i would avoid it. That said it doesnt bring any feature so i dont want to fight for a name.
Re: Servlet Module - Do we really need @Web?
If it was agreed, we should leave @JsfPhaseListener as it is... Sorry if i'm a bit annoying but... I still like the idea of a global qualifier instead @Web. As Romain said, it would see it as an namespace - @Inject @DeltaSpike ServletContext context looks good. 2014-02-23 23:01 GMT+01:00 Gerhard Petracek gerhard.petra...@gmail.com: at other parts you specify an artifact with an interface, annotation,... and at the same point (if possible) you are able to specify the ordinal (without an extra annotation). that's easier for users and therefore we agreed on it in the beginning. regards, gerhard http://www.irian.at Your JSF/JavaEE powerhouse - JavaEE Consulting, Development and Courses in English and German Professional Support for Apache MyFaces 2014-02-23 22:48 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com : which consistency gerhard? 2014-02-23 22:45 GMT+01:00 Gerhard Petracek gerhard.petra...@gmail.com : independent of the name, we would break the consistency. regards, gerhard http://www.irian.at Your JSF/JavaEE powerhouse - JavaEE Consulting, Development and Courses in English and German Professional Support for Apache MyFaces 2014-02-23 22:40 GMT+01:00 Romain Manni-Bucau rmannibu...@gmail.com: using javax.* from deltaspike is quite ambiguous: is it a DS behavior or a spec one? I'm no opposed Priority but in org.apache.deltaspile...Priority Romain Manni-Bucau Twitter: @rmannibucau Blog: http://rmannibucau.wordpress.com/ LinkedIn: http://fr.linkedin.com/in/rmannibucau Github: https://github.com/rmannibucau 2014-02-23 22:34 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com : Could you explain this? I can't follow you. 2014-02-23 22:23 GMT+01:00 Romain Manni-Bucau rmannibu...@gmail.com : I'd prefer something more explicit. Just changing the package would be enough for me Romain Manni-Bucau Twitter: @rmannibucau Blog: http://rmannibucau.wordpress.com/ LinkedIn: http://fr.linkedin.com/in/rmannibucau Github: https://github.com/rmannibucau 2014-02-23 22:17 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com : Hmm... We could clone it for JEE6 support and handle both annotations for backward compatibility. Maybe we could also use it for other features (or later..) 2014-02-23 22:14 GMT+01:00 Romain Manni-Bucau rmannibu...@gmail.com : That's would be ambiguous with JavaEE 6 Romain Manni-Bucau Twitter: @rmannibucau Blog: http://rmannibucau.wordpress.com/ LinkedIn: http://fr.linkedin.com/in/rmannibucau Github: https://github.com/rmannibucau 2014-02-23 22:12 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com : What about reusing javax.annotation.Priority? 2014-02-23 22:02 GMT+01:00 Gerhard Petracek gerhard.petra...@gmail.com : it's just a stereotype to get a better tool-support... @Advanced is also a qualifier (at least in codi). however, we need ordinal or we add @InvocationOrder as it was in codi. once we do that, we break the consistency in that area. regards, gerhard http://www.irian.at Your JSF/JavaEE powerhouse - JavaEE Consulting, Development and Courses in English and German Professional Support for Apache MyFaces 2014-02-23 21:41 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com : Another problem is that @web is a qualifier and @JsfPhaseListener is a stereotype and not a qualifier. So if we would change it, we must use different annotations. Right? IMO @Inject @DeltaSpike ServletContext context; would be great as it's just like a namespace. For the @JsfPhaseListener, something like @Advanced would be better. 2014-02-18 11:16 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com : IMO it's actually more expressive than any other names. Maybe @DeltaSpikeManaged would be more impressive... +1 @DeltaSpike then, before keeping the old redudant/ugly qualifiers. 2014-02-18 10:51 GMT+01:00 Romain Manni-Bucau rmannibu...@gmail.com: +1 kind of tradeoff Romain Manni-Bucau Twitter: @rmannibucau Blog: http://rmannibucau.wordpress.com/ LinkedIn: http://fr.linkedin.com/in/rmannibucau Github: https://github.com/rmannibucau 2014-02-18 10:46 GMT+01:00 Gerhard
Re: Servlet Module - Do we really need @Web?
wonder if 2 is really a *public* question once 1 is decided Romain Manni-Bucau Twitter: @rmannibucau Blog: http://rmannibucau.wordpress.com/ LinkedIn: http://fr.linkedin.com/in/rmannibucau Github: https://github.com/rmannibucau 2014-02-23 23:26 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com: Shoudln't we do 2 polls? 1) Should we introduce a global qualifier instead @web? If yes, which name? 2) Alternative for JsfPhaseListener? - keep it as it is - just rename it as it redudant - use the global qualifier + introduce a second annotation for setting the ordinal 2014-02-23 23:21 GMT+01:00 Romain Manni-Bucau rmannibu...@gmail.com: maybe we can do a public poll? 1) Keep it like it 2) @DeltaSpike 3) ordinal() 4) other? Romain Manni-Bucau Twitter: @rmannibucau Blog: http://rmannibucau.wordpress.com/ LinkedIn: http://fr.linkedin.com/in/rmannibucau Github: https://github.com/rmannibucau 2014-02-23 23:18 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com : If it was agreed, we should leave @JsfPhaseListener as it is... Sorry if i'm a bit annoying but... I still like the idea of a global qualifier instead @Web. As Romain said, it would see it as an namespace - @Inject @DeltaSpike ServletContext context looks good. 2014-02-23 23:01 GMT+01:00 Gerhard Petracek gerhard.petra...@gmail.com : at other parts you specify an artifact with an interface, annotation,... and at the same point (if possible) you are able to specify the ordinal (without an extra annotation). that's easier for users and therefore we agreed on it in the beginning. regards, gerhard http://www.irian.at Your JSF/JavaEE powerhouse - JavaEE Consulting, Development and Courses in English and German Professional Support for Apache MyFaces 2014-02-23 22:48 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com : which consistency gerhard? 2014-02-23 22:45 GMT+01:00 Gerhard Petracek gerhard.petra...@gmail.com : independent of the name, we would break the consistency. regards, gerhard http://www.irian.at Your JSF/JavaEE powerhouse - JavaEE Consulting, Development and Courses in English and German Professional Support for Apache MyFaces 2014-02-23 22:40 GMT+01:00 Romain Manni-Bucau rmannibu...@gmail.com: using javax.* from deltaspike is quite ambiguous: is it a DS behavior or a spec one? I'm no opposed Priority but in org.apache.deltaspile...Priority Romain Manni-Bucau Twitter: @rmannibucau Blog: http://rmannibucau.wordpress.com/ LinkedIn: http://fr.linkedin.com/in/rmannibucau Github: https://github.com/rmannibucau 2014-02-23 22:34 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com : Could you explain this? I can't follow you. 2014-02-23 22:23 GMT+01:00 Romain Manni-Bucau rmannibu...@gmail.com : I'd prefer something more explicit. Just changing the package would be enough for me Romain Manni-Bucau Twitter: @rmannibucau Blog: http://rmannibucau.wordpress.com/ LinkedIn: http://fr.linkedin.com/in/rmannibucau Github: https://github.com/rmannibucau 2014-02-23 22:17 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com : Hmm... We could clone it for JEE6 support and handle both annotations for backward compatibility. Maybe we could also use it for other features (or later..) 2014-02-23 22:14 GMT+01:00 Romain Manni-Bucau rmannibu...@gmail.com : That's would be ambiguous with JavaEE 6 Romain Manni-Bucau Twitter: @rmannibucau Blog: http://rmannibucau.wordpress.com/ LinkedIn: http://fr.linkedin.com/in/rmannibucau Github: https://github.com/rmannibucau 2014-02-23 22:12 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com : What about reusing javax.annotation.Priority? 2014-02-23 22:02 GMT+01:00 Gerhard Petracek gerhard.petra...@gmail.com : it's just a stereotype to get a better tool-support... @Advanced is also a qualifier (at least in codi). however, we need ordinal or we add @InvocationOrder as it was in codi. once we do that, we break the consistency in that area. regards, gerhard http://www.irian.at Your JSF/JavaEE powerhouse - JavaEE Consulting, Development and Courses in English and German Professional Support for Apache MyFaces 2014-02-23 21:41 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com : Another problem is that @web is a
Re: Servlet Module - Do we really need @Web?
maybe we can do a public poll? 1) Keep it like it 2) @DeltaSpike 3) ordinal() 4) other? Romain Manni-Bucau Twitter: @rmannibucau Blog: http://rmannibucau.wordpress.com/ LinkedIn: http://fr.linkedin.com/in/rmannibucau Github: https://github.com/rmannibucau 2014-02-23 23:18 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com: If it was agreed, we should leave @JsfPhaseListener as it is... Sorry if i'm a bit annoying but... I still like the idea of a global qualifier instead @Web. As Romain said, it would see it as an namespace - @Inject @DeltaSpike ServletContext context looks good. 2014-02-23 23:01 GMT+01:00 Gerhard Petracek gerhard.petra...@gmail.com: at other parts you specify an artifact with an interface, annotation,... and at the same point (if possible) you are able to specify the ordinal (without an extra annotation). that's easier for users and therefore we agreed on it in the beginning. regards, gerhard http://www.irian.at Your JSF/JavaEE powerhouse - JavaEE Consulting, Development and Courses in English and German Professional Support for Apache MyFaces 2014-02-23 22:48 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com : which consistency gerhard? 2014-02-23 22:45 GMT+01:00 Gerhard Petracek gerhard.petra...@gmail.com : independent of the name, we would break the consistency. regards, gerhard http://www.irian.at Your JSF/JavaEE powerhouse - JavaEE Consulting, Development and Courses in English and German Professional Support for Apache MyFaces 2014-02-23 22:40 GMT+01:00 Romain Manni-Bucau rmannibu...@gmail.com: using javax.* from deltaspike is quite ambiguous: is it a DS behavior or a spec one? I'm no opposed Priority but in org.apache.deltaspile...Priority Romain Manni-Bucau Twitter: @rmannibucau Blog: http://rmannibucau.wordpress.com/ LinkedIn: http://fr.linkedin.com/in/rmannibucau Github: https://github.com/rmannibucau 2014-02-23 22:34 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com : Could you explain this? I can't follow you. 2014-02-23 22:23 GMT+01:00 Romain Manni-Bucau rmannibu...@gmail.com : I'd prefer something more explicit. Just changing the package would be enough for me Romain Manni-Bucau Twitter: @rmannibucau Blog: http://rmannibucau.wordpress.com/ LinkedIn: http://fr.linkedin.com/in/rmannibucau Github: https://github.com/rmannibucau 2014-02-23 22:17 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com : Hmm... We could clone it for JEE6 support and handle both annotations for backward compatibility. Maybe we could also use it for other features (or later..) 2014-02-23 22:14 GMT+01:00 Romain Manni-Bucau rmannibu...@gmail.com : That's would be ambiguous with JavaEE 6 Romain Manni-Bucau Twitter: @rmannibucau Blog: http://rmannibucau.wordpress.com/ LinkedIn: http://fr.linkedin.com/in/rmannibucau Github: https://github.com/rmannibucau 2014-02-23 22:12 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com : What about reusing javax.annotation.Priority? 2014-02-23 22:02 GMT+01:00 Gerhard Petracek gerhard.petra...@gmail.com : it's just a stereotype to get a better tool-support... @Advanced is also a qualifier (at least in codi). however, we need ordinal or we add @InvocationOrder as it was in codi. once we do that, we break the consistency in that area. regards, gerhard http://www.irian.at Your JSF/JavaEE powerhouse - JavaEE Consulting, Development and Courses in English and German Professional Support for Apache MyFaces 2014-02-23 21:41 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com : Another problem is that @web is a qualifier and @JsfPhaseListener is a stereotype and not a qualifier. So if we would change it, we must use different annotations. Right? IMO @Inject @DeltaSpike ServletContext context; would be great as it's just like a namespace. For the @JsfPhaseListener, something like @Advanced would be better. 2014-02-18 11:16 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com : IMO it's actually more expressive than any other names. Maybe @DeltaSpikeManaged would be more impressive... +1 @DeltaSpike then, before keeping the old redudant/ugly qualifiers. 2014-02-18 10:51 GMT+01:00 Romain Manni-Bucau
Re: Servlet Module - Do we really need @Web?
Hmm. So i would just start with 1) and we can continue our discussion after the poll. Ok? 2014-02-23 23:28 GMT+01:00 Romain Manni-Bucau rmannibu...@gmail.com: wonder if 2 is really a *public* question once 1 is decided Romain Manni-Bucau Twitter: @rmannibucau Blog: http://rmannibucau.wordpress.com/ LinkedIn: http://fr.linkedin.com/in/rmannibucau Github: https://github.com/rmannibucau 2014-02-23 23:26 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com : Shoudln't we do 2 polls? 1) Should we introduce a global qualifier instead @web? If yes, which name? 2) Alternative for JsfPhaseListener? - keep it as it is - just rename it as it redudant - use the global qualifier + introduce a second annotation for setting the ordinal 2014-02-23 23:21 GMT+01:00 Romain Manni-Bucau rmannibu...@gmail.com: maybe we can do a public poll? 1) Keep it like it 2) @DeltaSpike 3) ordinal() 4) other? Romain Manni-Bucau Twitter: @rmannibucau Blog: http://rmannibucau.wordpress.com/ LinkedIn: http://fr.linkedin.com/in/rmannibucau Github: https://github.com/rmannibucau 2014-02-23 23:18 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com : If it was agreed, we should leave @JsfPhaseListener as it is... Sorry if i'm a bit annoying but... I still like the idea of a global qualifier instead @Web. As Romain said, it would see it as an namespace - @Inject @DeltaSpike ServletContext context looks good. 2014-02-23 23:01 GMT+01:00 Gerhard Petracek gerhard.petra...@gmail.com : at other parts you specify an artifact with an interface, annotation,... and at the same point (if possible) you are able to specify the ordinal (without an extra annotation). that's easier for users and therefore we agreed on it in the beginning. regards, gerhard http://www.irian.at Your JSF/JavaEE powerhouse - JavaEE Consulting, Development and Courses in English and German Professional Support for Apache MyFaces 2014-02-23 22:48 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com : which consistency gerhard? 2014-02-23 22:45 GMT+01:00 Gerhard Petracek gerhard.petra...@gmail.com : independent of the name, we would break the consistency. regards, gerhard http://www.irian.at Your JSF/JavaEE powerhouse - JavaEE Consulting, Development and Courses in English and German Professional Support for Apache MyFaces 2014-02-23 22:40 GMT+01:00 Romain Manni-Bucau rmannibu...@gmail.com: using javax.* from deltaspike is quite ambiguous: is it a DS behavior or a spec one? I'm no opposed Priority but in org.apache.deltaspile...Priority Romain Manni-Bucau Twitter: @rmannibucau Blog: http://rmannibucau.wordpress.com/ LinkedIn: http://fr.linkedin.com/in/rmannibucau Github: https://github.com/rmannibucau 2014-02-23 22:34 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com : Could you explain this? I can't follow you. 2014-02-23 22:23 GMT+01:00 Romain Manni-Bucau rmannibu...@gmail.com : I'd prefer something more explicit. Just changing the package would be enough for me Romain Manni-Bucau Twitter: @rmannibucau Blog: http://rmannibucau.wordpress.com/ LinkedIn: http://fr.linkedin.com/in/rmannibucau Github: https://github.com/rmannibucau 2014-02-23 22:17 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com : Hmm... We could clone it for JEE6 support and handle both annotations for backward compatibility. Maybe we could also use it for other features (or later..) 2014-02-23 22:14 GMT+01:00 Romain Manni-Bucau rmannibu...@gmail.com : That's would be ambiguous with JavaEE 6 Romain Manni-Bucau Twitter: @rmannibucau Blog: http://rmannibucau.wordpress.com/ LinkedIn: http://fr.linkedin.com/in/rmannibucau Github: https://github.com/rmannibucau 2014-02-23 22:12 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com : What about reusing javax.annotation.Priority? 2014-02-23 22:02 GMT+01:00 Gerhard Petracek gerhard.petra...@gmail.com : it's just a stereotype to get a better tool-support... @Advanced is also a qualifier (at least in codi). however, we need ordinal or we add @InvocationOrder as it was in codi. once we do that, we break the consistency in that area. regards, gerhard http://www.irian.at
Re: Servlet Module - Do we really need @Web?
+1 Romain Manni-Bucau Twitter: @rmannibucau Blog: http://rmannibucau.wordpress.com/ LinkedIn: http://fr.linkedin.com/in/rmannibucau Github: https://github.com/rmannibucau 2014-02-23 23:31 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com: Hmm. So i would just start with 1) and we can continue our discussion after the poll. Ok? 2014-02-23 23:28 GMT+01:00 Romain Manni-Bucau rmannibu...@gmail.com: wonder if 2 is really a *public* question once 1 is decided Romain Manni-Bucau Twitter: @rmannibucau Blog: http://rmannibucau.wordpress.com/ LinkedIn: http://fr.linkedin.com/in/rmannibucau Github: https://github.com/rmannibucau 2014-02-23 23:26 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com : Shoudln't we do 2 polls? 1) Should we introduce a global qualifier instead @web? If yes, which name? 2) Alternative for JsfPhaseListener? - keep it as it is - just rename it as it redudant - use the global qualifier + introduce a second annotation for setting the ordinal 2014-02-23 23:21 GMT+01:00 Romain Manni-Bucau rmannibu...@gmail.com: maybe we can do a public poll? 1) Keep it like it 2) @DeltaSpike 3) ordinal() 4) other? Romain Manni-Bucau Twitter: @rmannibucau Blog: http://rmannibucau.wordpress.com/ LinkedIn: http://fr.linkedin.com/in/rmannibucau Github: https://github.com/rmannibucau 2014-02-23 23:18 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com : If it was agreed, we should leave @JsfPhaseListener as it is... Sorry if i'm a bit annoying but... I still like the idea of a global qualifier instead @Web. As Romain said, it would see it as an namespace - @Inject @DeltaSpike ServletContext context looks good. 2014-02-23 23:01 GMT+01:00 Gerhard Petracek gerhard.petra...@gmail.com : at other parts you specify an artifact with an interface, annotation,... and at the same point (if possible) you are able to specify the ordinal (without an extra annotation). that's easier for users and therefore we agreed on it in the beginning. regards, gerhard http://www.irian.at Your JSF/JavaEE powerhouse - JavaEE Consulting, Development and Courses in English and German Professional Support for Apache MyFaces 2014-02-23 22:48 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com : which consistency gerhard? 2014-02-23 22:45 GMT+01:00 Gerhard Petracek gerhard.petra...@gmail.com : independent of the name, we would break the consistency. regards, gerhard http://www.irian.at Your JSF/JavaEE powerhouse - JavaEE Consulting, Development and Courses in English and German Professional Support for Apache MyFaces 2014-02-23 22:40 GMT+01:00 Romain Manni-Bucau rmannibu...@gmail.com: using javax.* from deltaspike is quite ambiguous: is it a DS behavior or a spec one? I'm no opposed Priority but in org.apache.deltaspile...Priority Romain Manni-Bucau Twitter: @rmannibucau Blog: http://rmannibucau.wordpress.com/ LinkedIn: http://fr.linkedin.com/in/rmannibucau Github: https://github.com/rmannibucau 2014-02-23 22:34 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com : Could you explain this? I can't follow you. 2014-02-23 22:23 GMT+01:00 Romain Manni-Bucau rmannibu...@gmail.com : I'd prefer something more explicit. Just changing the package would be enough for me Romain Manni-Bucau Twitter: @rmannibucau Blog: http://rmannibucau.wordpress.com/ LinkedIn: http://fr.linkedin.com/in/rmannibucau Github: https://github.com/rmannibucau 2014-02-23 22:17 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com : Hmm... We could clone it for JEE6 support and handle both annotations for backward compatibility. Maybe we could also use it for other features (or later..) 2014-02-23 22:14 GMT+01:00 Romain Manni-Bucau rmannibu...@gmail.com : That's would be ambiguous with JavaEE 6 Romain Manni-Bucau Twitter: @rmannibucau Blog: http://rmannibucau.wordpress.com/ LinkedIn: http://fr.linkedin.com/in/rmannibucau Github: https://github.com/rmannibucau 2014-02-23 22:12 GMT+01:00 Thomas Andraschko andraschko.tho...@gmail.com : What about reusing javax.annotation.Priority? 2014-02-23 22:02 GMT+01:00 Gerhard Petracek gerhard.petra...@gmail.com : it's just a stereotype to get a better tool-support... @Advanced is also a qualifier (at least in codi).
Re: NPE on DefaultClientWindow.getWindowId after long period of inactivity
hi thomas,
if you have any issue with the url-handling, please provide details about
it.
regards,
gerhard
http://www.irian.at
Your JSF/JavaEE powerhouse -
JavaEE Consulting, Development and
Courses in English and German
Professional Support for Apache MyFaces
2014-02-23 22:39 GMT+01:00 Thomas Frühbeck fruehb...@aon.at:
In fact I found the problem with an 8.0.0.CR1 version, and did not succeed
in reproducing the issue by manipulating the windowId cookie.
I have activated JSF 2.2 WindowId in web.xml to work around a bug in url
handling of DS 0.6 snapshot some time ago, but I cannot imagine, that this
has any relevance in this case, because the exception happens when DS
DefaultClientWindow tries to detect the WindowId handling config..
param-namejavax.faces.CLIENT_WINDOW_MODE/param-name
param-valueurl/param-value
Unfortunately I yet do not have any clue, how to reproduce the problem
else than waiting _very_(??) long :-/
Not really sure, how URL jfwid and cookies work together..
But anycase I will upgrade to 8.0.0.Final ASAP - or 8.0.1.Final, or... :-)
Thomas
Am 23.02.2014 22:02, schrieb John D. Ament:
H
You mention WildFly 8 Final Snapshot. Are you using one of their
prerelease snapshot builds or the final one? Can you try on the
final?
On Sun, Feb 23, 2014 at 1:44 PM, Mark Struberg strub...@yahoo.de wrote:
Well, from a CDI perspective this is pretty clear. We take whatever the
Session is. Regardless if this is an existing one or a new one. Seems to be
a WildFly issue. I'm sure the guys over there can fix this.
LieGrue,
strub
On Sunday, 23 February 2014, 19:37, Thomas Frühbeck fruehb...@aon.at
wrote:
Yes, definitely, in my case it has been about a day.
In my opinion the SessionScoped DefaultClientWindowConfig is
instantiated by the web layer, but the Wildfly session storage tries to
resurrect it from passivation, because an ID is provided.
Seems debatable, who's problem it is, but the effect is painful :-/
Am 23.02.2014 19:28, schrieb John D. Ament:
Is the long period of inactivity longer than your session timeout as
defined in your web.xml?
On Sun, Feb 23, 2014 at 11:49 AM, Thomas Frühbeck fruehb...@aon.at
wrote:
Hi,
I am currently using DeltaSpike 0.6 Snapshot some days old.. working
on
Wildfly 8.0.0.Final Snapshot.
I get an NPE in DefaultClientWindow.getWindowId() after long period
of
inactivity.
In this situation the application is _totally inaccessible_ from this
client
instance, the only remedy is to delete the session cookies and reset
the
URL:
2014-02-23 16:59:13,319 ERROR [io.undertow.request] (default task-36)
Servlet request failed HttpServerExchange{ GET
/mssms2/faces/csms/groupManagement.xhtml}: java.lang.NullPoi
nterException
at
org.jboss.weld.context.beanstore.http.AbstractSessionBeanStore.
getLockStore(AbstractSessionBeanStore.java:113)
[weld-core-impl-2.1.0.CR1.jar:2013-09-26 16:53]
at
org.jboss.weld.context.beanstore.AttributeBeanStore.
lock(AttributeBeanStore.java:210)
[weld-core-impl-2.1.0.CR1.jar:2013-09-26 16:53]
at
org.jboss.weld.context.AbstractContext.get(AbstractContext.java:88)
[weld-core-impl-2.1.0.CR1.jar:2013-09-26 16:53]
at
org.jboss.weld.context.PassivatingContextWrapper$
AbstractPassivatingContextWrapper.get(PassivatingContextWrapper.
java:76)
[weld-core-impl-2.1.0.CR1.jar:2013-09-26 16:
53]
at
org.jboss.weld.bean.proxy.ContextBeanInstance.getInstance(
ContextBeanInstance.java:98)
[weld-core-impl-2.1.0.CR1.jar:2013-09-26 16:53]
at
org.jboss.weld.bean.proxy.ProxyMethodHandler.invoke(
ProxyMethodHandler.java:78)
[weld-core-impl-2.1.0.CR1.jar:2013-09-26 16:53]
at
org.apache.deltaspike.jsf.spi.scope.window.DefaultClientWindowConfig$
Proxy$_$$_WeldClientProxy.getClientWindowRenderMode(Unknown
Source) [deltaspike-jsf-module-api-0.
6-SNAPSHOT.jar:0.6-SNAPSHOT]
at
org.apache.deltaspike.jsf.impl.scope.window.DefaultClientWindow.
getWindowId(DefaultClientWindow.java:119)
[deltaspike-jsf-module-impl-0.6-SNAPSHOT.jar:0.6-SNAPSHOT]
at
org.apache.deltaspike.jsf.impl.scope.window.
DefaultClientWindow$Proxy$_$$_WeldClientProxy.getWindowId(Unknown
Source) [deltaspike-jsf-module-impl-0.6-SNAPSHOT.jar:0.6
-SNAPSHOT]
at
org.apache.deltaspike.jsf.impl.listener.request.
DeltaSpikeLifecycleWrapper.execute(DeltaSpikeLifecycleWrapper.
java:78)
[deltaspike-jsf-module-impl-0.6-SNAPSHOT.jar:0.
6-SNAPSHOT]
