How to show a popup dialogue box AFTER a long running operation completes

Dialogs like this have gotten me all fuddled up in the past, and I still don’t really know exactly how they work, but I don’t suppose it would be as simple as encapsulating your first dialog like:

bool didWriteToTrace = false;

if (!didWriteToTrace)
{
    CountParams.AskExt(...)
}

if (didWriteToTrace)
    Base.PIHeader.Ask(...)

return adapter.Get();

I’m not sure if, when the method is hit the second time, didWriteToTrace will still be true (since the code is setting it to false initially), but maybe be worth a shot.

If that doesn’t work, you could try pulling your didWriteToTrace variable out of that method in hopes that it would remain true through the process.

@darylbowman  ChatGPT actually suggested a similar thing, which I tried, but it doesn’t work.
I even tried creating a private bool property like “ParamsCollected” to the graphExt which I set to true inside the IF statement, but that still didn’t work. It’s almost as though the entire graph is re-instantiated when the dialogue box is closed, because when I debugged it, the value was set to true after clicking OK on the smart panel popup, but then after the process was finished and the second pop-up opened and I clicked “OK” on it, the value of the property was reset to it’s original state.

I wonder if you need to check the .Ask value of Base.PIHeader.View like you do for the .AskExt call above?
 

if (didWriteToTrace)
{
  if (Base.PIHeader.Ask(...) = WebDialogResult.Ok) {};
}

In another thread around dialogs Dmitrii Naumov said:

It is related to the way Web Dialogs work in the framework.

In fact, when you call a web dialog, it throws an exception and remembers the current method where the exception has been thrown from. It then sets ‘Dialog Answer’ variable to the view and executes the method again.

So, I think,  the second .Ask triggers the dialog to appear and then the code is run again and since you’re not testing for the answer that the user gave, that makes the dialog appear the second time. I stress I think because, based on my theory, your main loop should also run a second time and I think you’d notice that happening.

@Django Unfortunately checking the .Ask value doesn’t make a difference. I’m still at a loss.

I know that the method is executed again, when the dialogue box closes, so it makes sense that it’s triggering the first pop-up again. And you are right, it would go through the loop again, except for the fact that it’s interrupted by the first pop-up, which I then just always close without clicking okay. 

It just seems strange that there would be no good way to show the user a popup message at the end of a long running operation.

I did a very quick look in the source code for AskExt to see if there are any methods when there are two calls to .Ask. I stopped looking after finding about 10 .AskExt calls but I didn’t see any methods where it was called twice.

You might have better luck.

Now, another option might be to move your processing logic into a different method so that the “physical” mechanics of the dialog calls are segregated.

@Django I’m not sure I understand the first part of your reply.
As far as separating the dialogue calls into separate methods, I’ve tried that and it didn’t make a difference.

You have access to large portions of the Acumatica source code and if you have a text editor (or alternative tool like grep) that allows you to search for text within multiple folders and sub-folders you can search through the code base to see how the Acumatica programmers implement their logic. I was hoping to find somewhere in the code where a routine or process had two calls to .Ask/AskExt that you could look at and see how they handled that. My concern was that there might not be any code that makes two calls to .Ask/AskExt because it doesn’t work like that.

You mentioned trying to set the didWriteToTrace as a variable at the graph level - did you try making it Static?

The other thing I was thinking you might do was to use the Warning tab/button on the processing screen instead of using the Trace. Then you don’t have to worry about directing the user to look at the Trace because it will be right there on the screen.

Oh, yeah I use DnSpy to review source code. I just wasn’t sure what it was you were searching for, but this makes sense.

@andriitkachenko You’re my hero!! 
This is exactly what I needed. 

Thank you so much!