I have a CMFCMenuButton on a CDialogBar that has a very strange behaviour: it shows its dropdown list in the Debug build but does not show it in the Release build.
Debug build:
Release build:
It seems that there is happening something wrong in the call of __super::OnShowMenu() of the OnShowMenu() method of my derived class.
To make thing worse, there is some configuration on my machine that is not stepping into the MFC feature pack methods ... I made everything to put this working and I could not get there.
Please help.
Sorry, it was my mistake:
On my derived class OnShowMenu function, I had to replace:
pMyMenu= new CMenu;
ASSERT(pMyMenu->CreatePopupMenu());
this->m_hMenu= pMyMenu->GetSafeHmenu();
with this:
pMyMenu= new CMenu;
pMyMenu->CreatePopupMenu();
this->m_hMenu= pMyMenu->GetSafeHmenu();
Yes, the bug was caused by the ASSERT.
Related
I would like to try the new pipeline feature in MediatR: https://github.com/jbogard/MediatR/wiki/Behaviors
I tried the following, but it does not get executed
services.AddMediatR();
services.AddTransient(typeof(IRequestPostProcessor<,>), typeof(PostHandler<,>));
What am I missing?
You need to register the behavior associated with post-processors, like this unit test shows.
Your registration code would look like:
services.AddMediatR();
services.AddTransient(typeof(IRequestPostProcessor<,>), typeof(PostHandler<,>));
services.AddTransient(typeof(IPipelineBehavior<,>), typeof(RequestPostProcessorBehavior<,>));
That behavior will get all the post-processors you registered and execute them.
Edit
After a comment about the post-processor running twice, I had a look at the code that registers MediatR in the ASP.NET Core built-in DI container, and it turns out instances of IRequestPreProcessor<TRequest, TResponse> and IRequestPostProcessor<TRequest, TResponse> are automatically registered as you can see here. What's left to do to get them running in the pipeline is just register the associated behavior. So the necessary registration is then:
services.AddMediatR();
services.AddTransient(typeof(IPipelineBehavior<,>), typeof(RequestPostProcessorBehavior<,>));
I encountered the same issue as Larsi in the comment above. My code looked like this
services.AddScoped<IPipelineBehavior<AddMessageRequest, MessageResponse>, RequestPostProcessorBehavior<AddMessageRequest, MessageResponse>>();
but the behaviour still executed twice. My solution was to simply not register it manually, seems like the registration is handled otherwise now.
In my case services.AddMediatR(Assembly.GetExecutingAssembly()); was enough.
Would be nice if someone could expand on why this is.
In roboelectric when you write your test class you have declare the #RunWith annotation of which there is RobolectricGradleTestRunner and RobolectricTestRunner.class. What is the difference and which one should we use ? Why is there 2 in the first place. If I use RobolectricTestRunner then it does not work for me it says some weird error like :
"java.lang.IllegalStateException: You need to use a Theme.AppCompat theme (or descendant) with this activity."
However this seems to go away if I use RobolectricGradleTestRunner.
RobolectricTestRunner was first and is for maven based projects.
RobolectricGradleTestRunner was for gradle based projects because some paths have changed.
Since robolectric 3.1.1 this is not more necessary and is now deprecated.
For more details see also http://robolectric.org/getting-started/
I have an Xcode project I've been working on for months. I've never had a problem with NSLog, but after upgrading to Xcode 4.2 nothing will log to the console. I even tried throwing this in viewDidLoad:
- (void)viewDidLoad
{
[super viewDidLoad];
NSLog(#"Can anyone hear me?");
And nothing. Is anyone else having this problem or know a solution?
Well, this is embarrassing. The console got deactivated somehow and I was actually watching the variables window. Pressing Shift + + C did the trick.
Many thanks to Robert King on this thread:
https://devforums.apple.com/message/565880#565880
This is a bug of Xcode8 + iOS10, we can solve it in this way:
When on simulator, add the Name OS_ACTIVITY_MODE and the Value
Variables disable and check it (Product -> Scheme -> Edit Scheme -> Run -> Arguments -> Environment).
When on device, only add OS_ACTIVITY_MODE and check it（don't add the Value). You'll see the NSLog in the Xcode8 Console.
In iOS10, a lot of system logging is displayed on the console. This can be disabled by including the OS_ACTIVITY_MODE = disabled in the Run Arguments for your scheme
However for iOS10, NSLog messages will not be displayed anymore.
In lower iOS versions the messages will still be displayed. So maybe for most test cases you can use a lower iOS version.
Otherwise you can include a Swift function which prints (print function) your text, this is working okay in iOS10.
Tested on iOS10.0.2 and iOS9.3
My issue is that I have Debugger Output selected. Once I changed it to either All output or target output NSLogs appear fine.
In case this confuses you its on the left of the Debug area.
I had the same issue.
The trick is to search and remove the below mentioned line from the project.
#define NSLog(...)
Search the entire project and remove it.
I just experienced this, so here's another thing to check.
-> Make sure you don't have anything typed in the Filter field below the Log Output. So, in my case I was searching for a term in the logger and forgot to delete the searched terms out of the Filter text field. DOH!
My issue was that I'd accidentally severed the referencing outlet from App Delegate (delegate - file's owner).
I'm looking for a good tutorial on how to use modeltest to test models based on QAbstractItemModel. I don't know how to interpret debug messages that are displayed.
Also I'm having trouble configuring modeltest project to work with my app in QtCreator. Including the .pri/.pro doesn't work. I get an error saying "No rule to mage target ..". After fixing paths in modeltest/modeltest.pro file it starts to compile. But i get this wierd assertion
ASSERT: "QTest::testLogger" in file c:\ndk_buildrepos\qt-desktop\src\testlib\qtestlog.cpp, line 232
Any ideas why this happens ?
My modeltest folder is located inside my project. I added following line at the end of my *.pro file
include(modeltest/modeltest.pri)
The modeltest.pri file contains the following
load(qttest_p4)
SOURCES += modeltest/modeltest.cpp modeltest/dynamictreemodel.cpp
HEADERS += modeltest/modeltest.h modeltest/dynamictreemodel.h
I modified my code to use modeltest this way
model = new TasksModel(this);
new ModelTest(model, this);
ui->treeView->setModel(model);
TasksModel is my implementation of QAbstractItemModel model.
ui->treeView is the widget that displays data.
No other modifications where made while integrating modeltest with my app.
Qt version is 4.7.
This will sound a little overgeeky - but it is, in fact, what the ModelChecker dev intended for you to do ;) When you hit one of the asserts, go to the point in the code where it is hit and read the comments which are written along with it. The entire thing is extremely heavily commented, and describes what is breaking and likely reasons why. This is by far easiest to do if you run your app through a GUI debugger, such as that included in for example KDevelop, Qt Creator or Visual Studio.
The reason you are getting this error is because you aren't actually using the ModelTest inside a proper QTestLib test case. If you take a look at /tests/auto/modeltest (where you presumably got the modeltest class in the first place), you can see how to properly construct a test case using the ModelTest.
If the planets are not correctly aligned when I run a Flex 3.5 build, the .swf produced immidiatly throws a null-object error from the _MyAppWatercherSetupUtil:
TypeError: Error #1009: Cannot access a property or method of a null object reference.
at _MyAppWatcherSetupUtil/setup()
at MyApp/initialize()[/Users/wolever/my_app/MyApp.mxml:0]
at mx.managers::SystemManager/http://www.adobe.com/2006/flex/mx/internal::childAdded()[C:\autobuild\3.5.0\frameworks\projects\framework\src\mx\managers\SystemManager.as:2131]
at mx.managers::SystemManager/initializeTopLevelWindow()[C:\autobuild\3.5.0\frameworks\projects\framework\src\mx\managers\SystemManager.as:3400]
at mx.managers::SystemManager/http://www.adobe.com/2006/flex/mx/internal::docFrameHandler()[C:\autobuild\3.5.0\frameworks\projects\framework\src\mx\managers\SystemManager.as:3223]
at mx.managers::SystemManager/docFrameListener()[C:\autobuild\3.5.0\frameworks\projects\framework\src\mx\managers\SystemManager.as:3069]
Is there anything I can do about this?
I've tried debugging it, but it seems like the Flex framework goes out of its way to make sure it's impossible to debug into the generated code…
I just upgraded an old Flex 3 app to Flex 4.6 and had the same problem. What worked for me was a project clean. So far so good.
--Keith
I was struggeling today with a WatcherSetupUil/setup() TypeError: Error #1009: Cannot access a property or method of a null object reference.
For me it looks there is a strong connection to databinding. For me it looks this happens if you do not take care of the rules that Flex asks for. I found a resource online I upload here: "Flex data binding pitfalls: 10 common misuses and mistakes." After changing - in my case especially error 6 - Using the wrong bindable event name - took the WatcherSetupUtil away.
Perhaps this is helpful for you too :-).
http://www.adobe.com/devnet/flex/articles/databinding_pitfalls.html
It would be interesting to have a glimpse at generated code for "setup" method for your MyAppWatcherSetupUtil class
Just an idea - did you try to set the compiler flag -keep? Like this you may have a look how your class looks like.