Objects are created by a creator of a proxy.
private static boolean isSupportedCreator(Object creator) {
return creator instanceof JustInTimeObjectCreator
|| creator instanceof
ReloadableServiceImplementationObjectCreator;
}
The creator is accessible by reading a priva
TapestryRunnerConstants was the class to vanish. I recreated it but the
page rendering is resulting in null response. Nothing is written so there
is something else to hunt for. I will extract the whole page tester related
classes (rewrote the PageTester class also) and the selenium is also
rewritte
On Wed, 16 Oct 2013 17:34:52 -0300, Martin Kersten
wrote:
Hi everyone,
Hi!
I just updated my pom today and removed all irritating dependencies
regarding to tapestry 5.3 and replaced them with their counterpart
versions of 5.4.
Don't forget that 5.4 is an alpha version. It's not fin
Hi everyone,
I just updated my pom today and removed all irritating dependencies
regarding to tapestry 5.3 and replaced them with their counterpart versions
of 5.4.
Now I run my page tester driven tests and it failed since the Page Tester
Constants class can be not found. It is just gone. The
Ah yes I am, I just created a generic service similar to the
selectModelFactory service to handle this for me. Now all I need is the
following code to get all my autocompletes up and running. It would be
really nice to have it as a part of the core.
@Inject
private ListLabelFactory listLab
On Wed, 16 Oct 2013 16:09:15 -0300, George Christman
wrote:
Thiago, do you have any more info on the ValueProvider? I'm just
wondering if what's already built into tap will accomplish the goal of
my above code snippet. If not, that code snippet would be a nice
addition to the
SelectModelF
Thiago, do you have any more info on the ValueProvider? I'm just wondering
if what's already built into tap will accomplish the goal of my above code
snippet. If not, that code snippet would be a nice addition to the
SelectModelFactory since it's so commonly used.
On Wed, Oct 16, 2013 at 2:56 PM,
I completely agree with you Bob, it broke my code and left me going in
circles trying to figure out why all of a sudden I was getting these select
exceptions when I was implementing the component no differently than I have
have.
On Wed, Oct 16, 2013 at 2:03 PM, Bob Harner wrote:
> The work-arou
It's not quite as elegant as I would have hoped. I only asked because I've
found over the years of using Tap you never know what these guys may have
already built, hoping not to reinvent the wheel. Anyhow, unless the tap
guys have something already built, this ended up being my reusable
solution.
ValueProvider?
On Wed, 16 Oct 2013 14:49:20 -0300, George Christman
wrote:
I'm wondering if there is already a tapestry service available for
creating
a string array from an object list similar to the
SelectModelFactory.createSelectModel.
I find myself creating similar code to below in m
What about these?
return F.flow(ldapCache.findAllLDAPUsers(keyword)).map(new
Mapper() {
public String map(LDAPProfile element) {
return element.getName();
}
}).toList();
On Wed, Oct 16, 2013 at 8:49 PM, George Christman
wrote:
> I'm wondering if there is already a tapestry servi
The work-around of setting the secure parameter to "false" works for me (if
I do it on *all* the selects in the form).
The issue was caused by the fix for TAP5-2179 ("The Select component can be
hacked to select a value not in the SelectModel").
I agree with Lance that the secure parameter's defa
I'm wondering if there is already a tapestry service available for creating
a string array from an object list similar to the
SelectModelFactory.createSelectModel.
I find myself creating similar code to below in my autocompletes.
public List onProvideCompletionsFromSupervisor(String keyword) {
On Wed, 16 Oct 2013 12:09:52 -0300, Alessio Gambi
wrote:
While ago I tried to patch directly RegistryImpl to get access to
modules information and various contributions inside the modules.
Unfortunately I did not managed to extract the actual contributions...
My plan was to capture the m
Oops, I meant to print out the HttpServletResponse class.
Once you know the response class, you'll need to put in some breakpoints to
find out why isComitted() is not true. A response should be committed as
soon as it's outputstream / writer is first written to or it is redirected.
I like the callback solution. Requires Tapestry support though :)
That wil serve my needs at least (BeanEditor that looks like the 'old' 5.3
BeanEditor :)
On Oct 16, 2013, at 7:30 AM, mailingl...@j-b-s.de wrote:
> Hi,
>
> just my two cents: In general coupling bean editor (and friends) tightl
Hi,
inside the troublesome method, RequestGlobals.getHttpServletRequest() returns
instance of
org.springframework.security.web.servletapi.SecurityContextHolderAwareRequestWrapper
this one wraps
org.springframework.security.web.firewall.RequestWrapper
and this wraps
org.mortbay.jetty.Request
While ago I tried to patch directly RegistryImpl to get access to modules
information and various contributions inside the modules. Unfortunately I did
not managed to extract the actual contributions... My plan was to capture the
moment when an object is instantiated by the registry and copy the
It's still failing with secure="false", the only way to get it to work is
to @Persist SelectModel computerModels.
btw, the onValueChanged method in my above example should be
onValueChangedFromComputer.
On Wed, Oct 16, 2013 at 9:38 AM, Lance Java wrote:
> It seems like a new 'secure' parameter
Can you @Inject RequestGlobals and log
RequestGlobals.getHttpServletRequest().getClass()?
I'm thinking that one of the spring filters is wrapping the request in an
faulty implementation. I'm starting to point the finger at
RequestCacheAwareFilter.
filed jira issue
https://issues.apache.org/jira/browse/TAP5-2204
On Wed, Oct 16, 2013 at 10:03 AM, George Christman
wrote:
> It's still failing with secure="false", the only way to get it to work is
> to @Persist SelectModel computerModels.
>
> btw, the onValueChanged method in my above example
It seems like a new 'secure' parameter has been added to select which
defaults to true (value must exist in the model).
Try setting secure="false"
I think this should be the default to maintain backwards compatibility.
Hi,
thanks for another hint. :) It was worth trying.
However, I am out of luck with this issue. RedirectFixRequestFilter from
gaeutils does not help either. :(
Cimlman
__
> Od: Lance Java
> Komu: Tapestry users
> Datum: 16.10.2013 1
Sample code and full stack. I'm at my day job and unable to recreate my
actual data model, so please keep in mind I use a multi tier selection
model, ie Category, Year, Make, Model, Trim. This worked without issue
prior to 5.4.21. The selectModel data is dependent upon the select model
value, so I'
I did truncate it but I think the interesting thing had already happened
which was the stack overflow.
My plan was to just override the services I wanted the configurations for,
save them into another service then just use the existing Tapestry service.
Seemed simple enough because I've overridden
Sorry, I missed your post where you'd isolated the problem to spring
security. I get the feeling that spring security is somehow introducing the
exact same bug as GAE.
Can you try adding the filter from the GAE fix to your module? I have a
hunch that it will fix this your issue too.
http://tinyur
Hi,
not sure if I understand your idea.
Spring Security is necessary to reproduce this issue. The problem occurs only
with Spring Security. The same event handling method works fine on page without
Spring Security.
Cimlman
__
> Od: L
It works in Safari on OS X!
It doesn't work in Chrome on OS X (as described earlier)!
Strange, since they're both based on webkit. Safari version is 6.0.5,
Chrome is 30.0.1599.101. Maybe it's something to do with the video codec
because I've found a video that has sound and Chrome plays the sound O
On Wed, 16 Oct 2013 09:21:41 -0300, Barry Books wrote:
I tried again with no success. The goal here is to gain access to the
configurations being passed to various services in order to build a
documentation pages listing the actual running configuration.
I don't think that's possible right no
Can you post a stack trace?
You can either persist your SelectModel, or recreate it in your
onPrepareForSubmit method (or in onPrepare, which gets called both for form
render and submit.
Yes, please create a JIRA issue.
On Oct 16, 2013 8:34 AM, "George Christman" wrote:
> Thanks Bob for correcting the subject line to 5.4,
Thanks Bob for correcting the subject line to 5.4, I forget T4 actually
existed.
Unless there is another work around, I was unable to get this to work
without the use of @Persist. Hopefully this is just a regression bug that
can be resolved. Should this be reported?
On Wed, Oct 16, 2013 at 7:01
I notice there's a whole lotta spring security stuff in your stack trace
before it gets to the TapestryFilter. Can you try removing spring security
from the mix to try to isolate the problem?
Hi,
thanks for these hints.
However, the exception is caused by something different. :(
I am not using GAE.
Setting "tapestry.gzip-compression-enabled" to "false" does not help either.
Cimlman
__
> Od: Lance Java
> Komu: Tapestry u
Has anyone got mp4 video outputting successfully within a page? In T5.4 or
T5.3.
What did you use in the template? I'd like to use HTML5 like the following,
but I'd settle for using the object or embed tag if I could get them to
work...
Your browser does not support the video tag.
What con
On Tue, 15 Oct 2013 20:18:39 -0300, Chris Mylonas
wrote:
Sweetness!
Updated,
[INFO] BUILD SUCCESS
Yay! :)
Thanks Thiago.
;)
My project website used to run on drupal for content/forum and trac for
pretty much code viewing. With the delinquencies (constant upgrades,
performance draggin
Using serviceBinder.bind(Service, ServiceImpl).preventReloading() solves
the issue. Now I can use impl =
(ServiceImpl)ServiceProxyUtil.getImplementation(registry.getService(Service.class)).
So I dont need to invoke methods using reflection and I have AOP and
Decoration for the rest of the service m
Hi,
just my two cents: In general coupling bean editor (and friends) tightly to
bootstrap is strange even I am aware its the de facto standard. Furthermore I
can not believe it will ever produce a layout which makes all of us happy. Is
it not possible to add a additional, optional parameter to
I checked the sources. After setting production mode to test my proxy util
for this it does not change reload behaviour. It seams to be only switched
off in case a service is bound with preventReload or a global property to
prevent reloading is set or the service is not an interface.
So in the end
Are you using GAE? You might be hitting this:
http://code.google.com/p/googleappengine/issues/detail?id=8201
http://apache-tapestry-mailing-list-archives.1045711.n5.nabble.com/T5-and-AppEngine-1-7-3-dev-server-IllegalStateException-td5717683.html
Fix here:
https://bitbucket.org/akochnev/tap5-gae-u
I noticed a couple of days ago in 5.4 alpha 22 that the SelectModel is now
being used on form render AND submission whereas it used to be used only on
render of the form. This is a big deal performance-wise because creation of
the SelectModel commonly involves a database query to populated a list o
Dont see this to work with AOP on the LegacyInterface or do I overlook
something.
Better might be the use of a legacy service and provide an extended service
using refelction to access the legacy service implementation (subclass).
Would also make it clear and nice.
2013/10/16 Lance Java
> Here
Hi,
after some time, I've encountered another instance of this exception. This
time, I am a little bit further in my investigation.
The exception occurs only with Spring Security.
Spring Security 3.1.2
Tapestry 5.3.6
Returning StreamResponse from onActivate is always fine.
public class MySecur
Here's a workaround
public interface LegacyService {
void method1();
}
public class LegacyServiceImpl implements LegacyService {
public void method1() { ... }
public void method2() { ... }
}
public interface ExtendedService {
void m
I noticed. I read the source and this return bind(class,class) is what I
missed: "So thanks you two". Maybe updating the documentation... well let
it be :).
Sadly I need AOP for that too, so having an undecorated service is not good
enough. Currently I just stick with
Reflection.invoke(ServiceProx
If you really really really want to do this thing that we are both
suggesting is a bad idea, you can do one of the following in AppModule
public static MyServiceImpl buildMyServiceImpl(…) {
return new MyServiceImpl(…);
}
OR
public static void bind(ServiceBinder binder) {
binder.bind(MyServ
>
> Also it makes everything feel stiff and decoupled in a not useful way.
>
> Stiff is the opposite of decoupled. And interfaces are quite useful, but
I'm not really in the mood of explaining why. OOP and IoC already do that
for me.
Nope. Completely wrong :). Stiff means not changeable in terms
47 matches
Mail list logo
