Closed GoogleCodeExporter closed 9 years ago
Have you used any 1.3 snapshot before? This change was one of the first changes
for 1.3.
I will investigate it.
Original comment by balusc
on 17 Mar 2013 at 3:58
FYI: can't reproduce in Tomcat 7.0.37.
The exception indicates that ViewDeclarationLanguage can't be found. Can you
please let me know which URL pattern your FacesServlet is been mapped on? I
used *.xhtml.
Original comment by balusc
on 17 Mar 2013 at 4:17
Yes, I used OmniFaces 1.3 snapshot prior to using OmniFaces 1.3 release, but
OmniFaces 1.4.1 is the first time that I used 1.4 version.
My apologies. Myself and endusers access login.xhtml via login.jsf, and after
successful login, the app redirects to index.xhtml via userBean.loginUser()
returning "index.jsf?faces-redirect=true".
Below, is part of the web.xml.
<filter-mapping>
<filter-name>LoginFilter</filter-name>
<url-pattern>*.jsf</url-pattern>
</filter-mapping>
<servlet>
<servlet-name>Faces Servlet</servlet-name>
<servlet-class>javax.faces.webapp.FacesServlet</servlet-class>
<load-on-startup>1</load-on-startup>
</servlet>
<servlet-mapping>
<servlet-name>Faces Servlet</servlet-name>
<url-pattern>*.jsf</url-pattern>
</servlet-mapping>
FYI, I really was not motivated to migrate to 1.4 (at all), because I didn't
see anything that I really needed, even though someone else reported some issue
related to PrimeFaces 3.5, but my PrimeFaces 3.5 app has no issues with
OmniFaces 1.3 at all. I just decided to OmniFaces 1.4.1...just to see if
OmniFaces 1.4 would cause any issues with my app.
Original comment by smithh03...@gmail.com
on 17 Mar 2013 at 9:58
[deleted comment]
I couldn't reproduce the particular exception, but I could reproduce the
problem as described in comment 6 of issue 75. It's fixed in the below
snapshot, can you please let me know if it fixes your particular problem as
well?
Sorry for the regression. We really need a testing farm.
Original comment by balusc
on 18 Mar 2013 at 4:12
Attachments:
[deleted comment]
It is not necessary to for me to migrate from OmniFaces 1.3 to 1.4.1 at all,
but since I like to have the latest and greatest, and since I experienced this
issue (and assumed that my endusers would experience the same issue), I decided
to report this issue.
Agreed and I apologize for not providing a testing farm.
I repeated the steps that I provided in OP, and with that, I will mention one
requirement/prerequisite, which is OmniFaces 1.3 had to been running and user
logged into app prior to completing step #1.
omnifaces-1.5-SNAPSHOT-20130318.jar fixes this issue, but I wanted to report
the 'unexpected behavior' that I experienced.
1. Completed steps 1 thru 5 (as listed in OP)
2. App did not complete the 'loginUser' method, no exception displayed in
browser, no exception in localhost, so the login page displayed again, and the
following was in tomee's stderr and/or catalina log (which shows that the
server started, and then I attempted to login, but 'loginUser' was not
executed):
INFO: Server startup in 10440 ms
Mar 18, 2013 3:28:13 AM org.apache.myfaces.el.unified.ResolverBuilderBase
sortELResolvers
INFO: Chain of EL resolvers for Faces sorted with:
org.apache.myfaces.el.unified.OpenWebBeansELResolverComparator@25fa3dbb and the
result order is
[org.apache.myfaces.el.unified.resolver.implicitobject.ImplicitObjectResolver@4f
ead377,
org.apache.myfaces.el.unified.resolver.CompositeComponentELResolver@41e1529e,
org.apache.myfaces.el.FlashELResolver@2ff724dc,
org.apache.myfaces.el.unified.resolver.ManagedBeanResolver@a95001e,
org.apache.myfaces.el.unified.resolver.ResourceResolver@3c3ab6ff,
javax.el.ResourceBundleELResolver@32c43e3d,
org.apache.myfaces.el.unified.resolver.ResourceBundleResolver@71d230d4,
javax.el.MapELResolver@1e83e8ad, javax.el.ListELResolver@6620c581,
javax.el.ArrayELResolver@7186cc88, javax.el.BeanELResolver@48de2ef2,
org.apache.webbeans.el.WebBeansELResolver@72dc373d]
3. So, after 2 minutes, I decided to click Login button again, just to see what
happens, and the app logged me in, without issue/exception (as expected).
Below, is what was in tomee stderr/catalina log after completing this step #3.
Mar 18, 2013 3:30:32 AM jsf.users.pf_UsersController loginUser
INFO: pf_UsersController.loginUser(): administrator logged in at Mon Mar 18
03:30:31 EDT 2013
Mar 18, 2013 3:31:48 AM jsf.users.pf_UsersController logout
INFO: pf_UsersController.logout(): administrator logged out at Mon Mar 18
03:31:48 EDT 2013
Conclusion: OmniFaces 1.3 would have logged me in successfully after completing
steps 1 thru 5 in OP, but OmniFaces 1.4.1 and 1.5 introduces
abnormal/unexpected app behavior.
I wanted to try to repeat original test steps but clear tomee's work folder as
step '3a', prior to starting TomEE. will let you know the test results. thanks.
Original comment by smithh03...@gmail.com
on 18 Mar 2013 at 7:49
[deleted comment]
I had to delete comment #8 above, because I made a mistake in my
build-deploy-WAR step. :)
I repeated test steps (without deleting/clearing tomee/temp or tomee/work
folders), but test step #3 (in OP), I did the following instead:
3. Replace OmniFaces 1.5 SNAPSHOT JAR with OmniFaces 1.3 JAR in web app,
generate and deploy WAR to TomEE
And still, with OmniFaces 1.3, I am not able to duplicate exception in OP and
not able to duplicate the behavior noted in comment #7 above.
Next, will clear tomee/temp and tomee/work folders and retest with Omnifaces
1.4.1 and 1.5 SNAPSHOT JARs.
Original comment by smithh03...@gmail.com
on 18 Mar 2013 at 8:08
Deleted files/folders in tomee/temp and tomee/work (but not
tomee/temp/safeToDelete.tmp, since that is included in tomee ZIP file), and
repeated steps listed in OP, and duplicated the issue reported in OP.
Step #3 was left as-is:
3. Replace OmniFaces 1.3 JAR with OmniFaces 1.4.1 JAR in web app, generate and
deploy WAR to TomEE
Original comment by smithh03...@gmail.com
on 18 Mar 2013 at 8:17
BalusC, I would assume that OmniFaces 1.3 handles the situation, best.
Please note, I have a login servlet-filter in place (one of many that I saw you
provided to people on stackoverflow.com), which will initiate http session and
test if 'user' bean is valid/instantiated.
I am assuming that http session has been instantiated after completing step #1
in OP, and OmniFaces 1.3 restores view more cleanly and 'successfully' better
than OmniFaces 1.4.1 and 1.5 SNAPSHOT.
Since I don't have a test farm to provide, please feel free to mark this as
Fixed. :)
Original comment by smithh03...@gmail.com
on 18 Mar 2013 at 8:23
I assume that myself and my endusers will continue to see the behavior recorded
in comment #7 above, since I have already deployed OmniFaces 1.5 SNAPSHOT to
production server and will see how it performs. So far, I don't see any other
defects/issues.
Recently, I have been in stable-version mode, and OmniFaces 1.3 is very stable
with my app-and-environment. :)
Original comment by smithh03...@gmail.com
on 18 Mar 2013 at 8:41
Assumption in Comment #12, because i stop/shutdown TomEE, when I need to
build-n-deploy new WAR, and i usually always repeat steps in OP (logout,
build-n-deploy new WAR, restart TomEE, and login).
I am sure that endusers will duplicate the logout-and-login (after I/Howard
updates web app software on production server), and endusers will experience
behavior noted in comment #7 above.
At most, they may say, i clicked Login, but I had to click Login button one
more time and it logged me in. :)
Original comment by smithh03...@gmail.com
on 18 Mar 2013 at 8:46
OmniFaces 1.5 SNAPSHOT has been running on production server, and I have
heard/seen no issues in tomee logs and endusers have NOT reported any issues.
As requested earlier, please mark this issue as fix. I can be okay with
behavior/regression reported in comment #7 above...if that continues to happen
when i update web app software on production server.
Original comment by smithh03...@gmail.com
on 18 Mar 2013 at 6:28
Thank you.
Fix was already checked in
https://code.google.com/p/omnifaces/source/detail?r=c21197b442b81657a23c5b28ae67
eb27c94b7267
Original comment by balusc
on 18 Mar 2013 at 10:56
You're welcome. Is there an explanation about the difference in behavior or
performance between OmniFaces 1.3 restore view and OmniFaces 1.4.1/1.5-snapshot
restore view?
Original comment by smithh03...@gmail.com
on 18 Mar 2013 at 11:21
Downloaded omnifaces-1.5-SNAPSHOT-20130430.jar and now I am 'not' able to
duplicate the unwanted/unexpected behavior reported in comment #7 above. :)
Original comment by smithh03...@gmail.com
on 11 May 2013 at 4:12
Original issue reported on code.google.com by
smithh03...@gmail.com
on 17 Mar 2013 at 2:04Attachments: