| Re: killing non-modal widgets [message #49018] |
Fri, 09 June 2006 09:35 |
btt
Messages: 345
Registered: December 2000
|
Senior Member |
|
|
David Fanning wrote:
> Ben Tupper writes:
>
>> Yes, I just double checked that each of the widgets have TLBs with the
>> parent widget argument but *not* the GROUP_LEADER keyword set to valid
>> widgets.
>
> OK, are these KILL_NOTIFY callbacks or CLEANUP callbacks?
> It *used* to be the case that you couldn't use KILL_NOTIFY
> with TLBs that were directly managed by XMANAGER. I see
> RSI code that violates this concept all the time, but I've
> always stuck to it and never had any problems. The intermittent
> nature of the problem suggests to me a problem with the
> various event loops that get set up to manage these autonomous
> widgets. It has never been clear to me in which order callbacks
> are executed, but I know it is different on different operating
> systems. I wonder if switching to CLEANUP callbacks would help.
>
Ah! That's the lead I am looking for. I'll switch it all over to
CLEANUP specified for XMANAGER rather than the KILL_NOTIFIES.
The documentation is pretty clear that the CLEANUP callback routine
supercedes the KILL_NOTIFY routine.
From XMANAGER online help...
CLEANUP
Set this keyword to a string that contains the name of the routine to be
called when the widget program dies. If this keyword is not specified,
the routine (if any) specified for the program's top-level base by the
KILL_NOTIFY keyword to WIDGET_BASE or WIDGET_CONTROL is used.
The routine specified by CLEANUP becomes the KILL_NOTIFY routine for the
widget application, overriding any cleanup routines that have been set
previously via the KILL_NOTIFY keyword to WIDGET_BASE or WIDGET_CONTROL.
Thanks for the help. I post something if that works.
Cheers,
Ben
|
|
|
|
| Re: killing non-modal widgets [message #49021 is a reply to message #49018] |
Fri, 09 June 2006 09:26  |
David Fanning
Messages: 11724
Registered: August 2001
|
Senior Member |
|
|
Ben Tupper writes:
> Yes, I just double checked that each of the widgets have TLBs with the
> parent widget argument but *not* the GROUP_LEADER keyword set to valid
> widgets.
OK, are these KILL_NOTIFY callbacks or CLEANUP callbacks?
It *used* to be the case that you couldn't use KILL_NOTIFY
with TLBs that were directly managed by XMANAGER. I see
RSI code that violates this concept all the time, but I've
always stuck to it and never had any problems. The intermittent
nature of the problem suggests to me a problem with the
various event loops that get set up to manage these autonomous
widgets. It has never been clear to me in which order callbacks
are executed, but I know it is different on different operating
systems. I wonder if switching to CLEANUP callbacks would help.
Cheers,
David
--
David Fanning, Ph.D.
Fanning Software Consulting, Inc.
Coyote's Guide to IDL Programming: http://www.dfanning.com/
|
|
|
|
| Re: killing non-modal widgets [message #49023 is a reply to message #49021] |
Fri, 09 June 2006 09:16 |
btt
Messages: 345
Registered: December 2000
|
Senior Member |
|
|
David Fanning wrote:
> Ben Tupper writes:
>
>> Yah!
>>
>> IDL> print,!Version
>> { ppc darwin unix Mac OS X 6.3 Mar 23 2006 32 64}
>>
>> But I thought you were going to say, "Roses?"
>>
>> I was thinking about that, too, but I am not sure I believe it is the
>> source of the problem, because it is just this widget application out of
>> the bazillions I fell like I have done just this way before. I am pretty
>> sure that I have frigged things up all on my own.
>
> No, I meant, MacOS, look there first! :-)
>
> Whenever I encounter something strange and intermittent
> I am in a class and some yo-ho has a Mac there. Have you
> assigned group leaders to any of these widgets?
>
Yes, I just double checked that each of the widgets have TLBs with the
parent widget argument but *not* the GROUP_LEADER keyword set to valid
widgets.
Thanks,
Yo-Ho
|
|
|
|
| Re: killing non-modal widgets [message #49025 is a reply to message #49023] |
Fri, 09 June 2006 08:43 |
David Fanning
Messages: 11724
Registered: August 2001
|
Senior Member |
|
|
Ben Tupper writes:
> Yah!
>
> IDL> print,!Version
> { ppc darwin unix Mac OS X 6.3 Mar 23 2006 32 64}
>
> But I thought you were going to say, "Roses?"
>
> I was thinking about that, too, but I am not sure I believe it is the
> source of the problem, because it is just this widget application out of
> the bazillions I fell like I have done just this way before. I am pretty
> sure that I have frigged things up all on my own.
No, I meant, MacOS, look there first! :-)
Whenever I encounter something strange and intermittent
I am in a class and some yo-ho has a Mac there. Have you
assigned group leaders to any of these widgets?
Cheers,
David
--
David Fanning, Ph.D.
Fanning Software Consulting, Inc.
Coyote's Guide to IDL Programming: http://www.dfanning.com/
|
|
|
|
| Re: killing non-modal widgets [message #49026 is a reply to message #49025] |
Fri, 09 June 2006 08:35 |
btt
Messages: 345
Registered: December 2000
|
Senior Member |
|
|
David Fanning wrote:
> Ben Tupper writes:
>
>> So, I guess I am fishing for an ideas of where to look first. Thoughts?
>
> MacOS?
David,
Yah!
IDL> print,!Version
{ ppc darwin unix Mac OS X 6.3 Mar 23 2006 32 64}
But I thought you were going to say, "Roses?"
I was thinking about that, too, but I am not sure I believe it is the
source of the problem, because it is just this widget application out of
the bazillions I fell like I have done just this way before. I am pretty
sure that I have frigged things up all on my own.
Ben
|
|
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
