debugging modal dlg boxes

Charles Hartman charles.hartman at conncoll.edu
Mon Oct 17 08:07:43 EDT 2005


On Oct 17, 2005, at 12:47 AM, J. Landman Gay wrote:

> Charles Hartman wrote:
>
>> I know this is a dumb question -- but I'm having a hard time   
>> debugging a modal dialog box. What am I missing?
>> When the box is full, the "Done" button processes its data (and  
>> sends  it off to a MySQL database). But debugging the mouseUp  
>> script for  that button isn't working -- because when I try to  
>> stop at a  breakpoint in it, though the script does stop, I can't  
>> click in it,  or in anything else at all: menus, message box, the  
>> Done button  itself . . . so there's nothing to do but Force Quit,  
>> which doesn't  exactly give me the information I need for debugging.
>> Is there some method for debugging modals that I don't know  
>> about?  Can't find anything in the docs . . .
>>
>
> Unless you have put in some kind of escape route for debugging, you  
> are stuck. Modals just don't allow anything else to happen until  
> they are dismissed.

Ah, OK. The cost of the app (stack) and IDE not being distinct, for  
the sake of avoiding compilation. Hm.

> Fortunately, Rev has a built-in escape from these. Hold down  
> Command-Control-Shift (on Mac, anyway) and click on the stack;  
> you'll get the contextual menu that allows you to change the stack  
> style to something other than modal. Then you won't have to force- 
> quit.

Excellent; that's what I need.

> While debugging, I usually call the stack in some other mode until  
> I know it works; toplevel, or modeless, or anything besides modal.  
> Once it is working correctly, then I change the code back to modal.  
> Note that modal dialogs are
> universally hated and shouldn't be used unless absolutely  
> necessary. ;)

I got into a situation in which a dialog had to be modal, and what it  
did *as* modal was what I needed to debug . . . Which isn't exactly  
*why* I've abandoned that trial design for this app, but it does make  
me glad I did.

Charles Hartman




More information about the use-livecode mailing list