On the Airflow UI if you have a task that fails is it possible to view logs specific to that task from within the UI?
Click on the task rectangle of custom_file_sensor_test_1_OmegaFileSensor with the red border within the graph -> View log.
The same is possible within Tree View. I find Tree View a bit more accessible, since you don't need to select the correct date as in comparison to the Graph View.
It is also possible to access the logs from the main screen. In the Links section, the hamburger-like icon to the right (v1.9.0) also leads to the logs.
Related
I could successfully compiled and run a program which has its icon in a system tray.
There is a good example explaining it here:
https://doc.qt.io/qt-5/qtwidgets-desktop-systray-example.html
Now is my problem: When I run my program, its icon is still visible also in the task bar.
How to get rid of this icon from the task bar?
Why it is important to me? My program will be an alarm-clock, so it should be visible on the Desktop, but it should not be occupying place on the task bar. This is why I decided to place it in the system tray. So, how to get rid of it from the task bar?
In the example you need to just uncheck the corresponding checkbox, which set the visible property to true or false:
Line in the example where the checkbox connects to the property:
connect(showIconCheckBox, &QAbstractButton::toggled,
trayIcon, &QSystemTrayIcon::setVisible);
Then you can close the window. The application will be still running because of the line
QApplication::setQuitOnLastWindowClosed(false);
Here is an answer
The information in the Task Bar about a particular running program is independent on the system tray information. So to make the program info (and its icon) 'not present' in the task bar - we have to work in a different place: We need to change the attribute of a main window of the program. In this case it was enough to add - the following line in a constructor of the program window:
setWindowFlags(Qt::Tool);
The tool dialog is a kind of window which information is not "placed" in a task Bar.
So now it works as I planed.
I am trying to figure out ways to authenticate a user and load the application (while showing a preloader) in Adobe Air [desktop application, and not web based flex app].
This is what I have been able to achieve so far:- A window asks for login details, once validated I make the login window invisible and open an instance of the application's main window. There are 2 problems with this:-
I have to open a new window (2 windows in total). It would rather look better if I was able to simply show the login form in one state and the main application view in another state. I tried that too, but the problem is that besides the view component, flash does execute all the actionscript and keeps the other state ready. There is some application view centric actionscript which starts throwing null reference errors etc. On top of that, the whole things takes a while to show up as both views are created, though only 1 of them is shown
I'd like to show a preloader once the user is authenticated, until the main application view is loaded
In a nutshell, this is what I am trying to do:-
Load the login window as fast as possible and stop there (actionscript in other parts of the application should not run)
Post authentication, load the main view of the application. Show a preloader until the loading is done
Load the main view in the same window (rather than opening a new window and making the login one invisible)
Need help and direction as to how this could be executed.
Thanks!
UPDATE 1
Ok, so now I've managed to consolidate the functionality in a single window. however, the transition between states is not a smooth one.
State 1 is the login screen. If the user enters the correct login credentials, State 2 of the view becomes active.
Now for State 2 to load (its a bunch of UI components and a grid with loads of data) it takes time. Until then, the application blanks out and then all of it is shown in a jerk. Can the transition be made much smoother? Just showing 'Loading...' would suffice. Because State 2 won't show up until all of it has been constructed and State 1 dies away as soon as I change the current state to State 2. Is there any way to monitor the progress and changing the state only when the next state has been loaded!
UPDATE 2
Ok, I got the transition animation to work between states. However, there still exists a problem with the transition switch. The problem is that the state I switch to after login has been verified has a lot of components and shows a lot of data.
Is there a way I can attach Listeners (if any), which I can fire when the state loading is complete and view has been generated! The current jerk like effect in the transition is because the state has changed but the view has not completed yet.
State Change to State 2 is not smooth as State 2 loads about 10000 rows of data from database. Is there a way I can change the state visibly for the user, after state 2 has been completely drawn out and has pulled in all data? creationComplete doesn't help much here.
In short, is there a way to start loading a state from an initial state and make it visible only when its complete loaded? i.e. can I fire an event from state 1 to load state 2, but to visibly transition to state 2 only when state 2 is completely loaded..
UPDATE 3
After a week of firefighting, posting a bounty and scavenging through the web I have still not been able to fix this! My application window becomes unresponsive for the time the UI is created and data is loaded. In Windows, it even shows 'Not Responding' at the window title bar for about 5 seconds. So its the UI getting stuck because the data is taking some time to be fetched and loaded - all of this happens in a single thread by default.
How do people who develop based on Adobe Air do this? I've mostly always seen a loading screen before the actual game is loaded - and when its loaded, its fully functional. There has to be a way!
The time consuming problem of loading 10000 rows in a grid can come in any web based language in any web application because web application needs to run in the environment of a web browser which has its own resource limitation.
So what I would suggest is that you don't load all 10000 records at loading the view. Instead load 1000 records first and then keep a link or button with label "Next" or "Show More" like and on that click bring the next bunch of 1000 records from the database. This way you can accomplish your task.
Thanks,
Jigar Oza
Functionnaly :
On one of my components of my application, I have an editing/lock system. When a user starts editing, he locks the file so other users cannot edit it.
Problem scenario : When the user activates "edition mode" and leaves screen, I would like to show a alert with two options : save changes, or discard changes.
There are different ways to exit screen :
There is a List on the left side containing other possible editabel data. A click changes the data in my component.
There is a menubar on top leading to other screens.
The edition component is embedded in a Tab navigator. When changing tabs, the alert has to show.
Closing browser.
Do I have to catch all of these events and plug at all those places?
Is there any kind of focusout mecanism?
The answer to the first question is: YES.
You need to watch all possible exit events that could harm the currently edited data.
Well, the problem is now how to manage this properly. Using an MVC framework you would trigger the appropriate commands from your components:
CHANGE_LIST_ITEM (new item)
CHANGE_TAB (new tab)
CHANGE_SCREEN (new screen)
Each command then checks if the currently edited tab has been saved or not. If not, it displays the Alert. Else, if there are no changes, it allows the list, the screen chooser and the tab bar to continue.
So your components (list, screens, tabs) need to implement some kind of rollback or preventDefault mechanism. Generally, changing their state must be allowed by a central validator (in MVC the command).
In the case of the list: I would suggest that the list is not selectable by mouse click but only programmatically. You set a listener on the list item click event. If the command allows setting of a new item it will notify the list. In MVC usually by sending an async message that gets received by the list's mediator. [[And even more correct: The command would set some model properties (e.g. currentListItem) and the model than sends an async message.]]
Edit: For the browser close event, you need to call a JavaScript expert.
I don't really know how to formulate my question this time...
I have my application with a QDialog as a main window.
The application is getting different values like temperature, humidity and so on from a remote machine.
For development I added a group box with different widgets to simulate these values. I have different limits for throwing warnings and alarms to the user.
For example if temperature rises over 30°C then I open a QMessageBox with the request time (the application does polling at the remote machine) and the current temperature. So this is updated each request cycle.
I use the show() method to bring up the message box which keeps my application running in background. The problem now is: the focus is at the message box and nothing in my main window/ QDialog can be clicked until the message box is not accepted/ has finished.
And that's my problem: in simulation mode I want to play around with different temperature values which I can adjust by slider in the main window. How can I access these widgets/ make the message box somehow "not-blocking"?
Best regards,
Matthias
What you're experiencing is called "modality" of a window. By default, a QMessageBox is "application modal". This means that input to all other application windows is blocked.
To change the modality, use setWindowModality() with a value from Qt::WindowModality just before you call show(). In your case:
box->setWindowModality(Qt::NonModal);
box->show();
Indeed you have a modal message box which is the way QMessageBox is intended to work - ie the user is to be alerted and the ui is protected from further interaction until the user has registered the message, closed the message box and (if necessary) taken any action required in response to the message.
Now if you set the message box modality to Qt::NonModal, (remember to assign it to a variable that won't go out of scope when your application continues after popping up the messagebox) you'll be able to interact with the ui even while the message box is displayed., which I have to say is 'unusual'. If I understand your requirement you will already have the message box up - then while this is still up, you'll then want to play around with different temp values in the main window - to what effect? Until another message dialog box is produced? The message box is going to have to be discarded at some point.
Not only that, but if you show a non modal message box and then interact with the main window, you're quite likely to just have the message box disappear behind the main window, out of sight anyway.
I'd suggest that message boxes are generally treated as intended - transient, temporary modal alert boxes only and that perhaps you require a clearly visible live report/status area in your main window rather than utilizing a non-modal QMessageBox.
Hope this helps
Regards
Roger
I'm using VS 2010. And in VS 2010 it's even more painful now to attach to the worker process when you want to test your local IIS site via IIS, not the VS built-in web server.
It's more painful because now when you select the worker process () via the Debug Menu | Attach to Process | w3wp.exe it brings up an additional prompt asking if you're sure you want to do this.
Even before this, attaching to the process in previous versions of VS was just outright painful. I must do this 200 times a day which wastes a lot of my time when it takes 5-10 seconds to click through and do this.
Now if you've got 2 different worker processes going, then I can see a need to choose. But before that, just getting to that attach screen is simply painful.
Anyone know a shortcut or just a quicker way other than the attach to process dialog in order to attach to the worker process in VS?
I don't understand why Microsoft doesn't add a nice toolbar icon or something that jumps you right to that Attach to Dialog screen and even better you could be able to specify only to show worker processes as an option which would save me a hell of a log of time sifting through that list even though it's already alphabetized. I still have to scroll to get to it.
Anyway, it's becoming a huge annoyance now..I'm really tired of the long way to get the attachment going to that process every single time.
Hm, not a great leap over and above what you are doing, but it may ease your workflow. have you taken a look at command customization?
Click Tools -> Customize ... (or Right-click tool bar and select Customize ...)
From Customize dialog, click Commands
Select your target, (I suggest Context menu)
Select your sub target (if Context menu, select Editor Context Menus | Code Window)
Click Add Command
From Add Command dialog, select Debug category
Select Attach to Process... from command list (near top)
Click OK
Click Close
Use your new shortcut!
Hope this helps! :)
ps: if you added Attach to Process... command to Code Window context menu, you should be able to launch the Attach to Process dialog from right clicking anywhere in a code window.
Record a macro while attaching and assign a shortcut for that macro and/or assign a menu item for it.
Steps:
Tools -> Macros ->Record Temporary macro
Attach to w3wp, Click on Ignore, etc
Stop recording the macro
Save it
To add a menu item (say under Debug) Right click on VS's Menu strip -> Customize... ->Commands -> ... Add Command -> Macros.YourMacroNAme(DebugIIS for eg)