comp.lang.idl-pvwave archive: archive » Re: GUI interface update issues

Home » Public Forums » archive » Re: GUI interface update issues

Show: Today's Messages :: Show Polls :: Message Navigator
E-mail to friend

Re: GUI interface update issues [message #44499]

Wed, 22 June 2005 08:57

Rick Towler
Messages: 821
Registered: August 1998

Senior Member

Clearly I am learning this as I go and I probably posted a little too
early... I never have had issues with GUI interface drawing and I never
really cared about interrupting program execution. Let the users wait.
My code is pure genius at work! :) lol

> Reading the IDL_BailOut docs leads me to believe that IDL faithfully
> captures interrupts and will break before the next IDL command is
> executed and that the reason for IDL_BailOut is to provide an exit
> from external routines.
>
> But it seems that this is overly simplified. If I try to break
> my (pure IDL) program's execution once it has entered the model
> procedure it will continue to run thru the loop and thru 50 or
> so more lines of IDL code after the loop before it stops at a
> call to SAVE.
>
> So IDL is capturing the interrupt signal, but there seems to be
> limits on when it will stop execution based on where IDL is in
> the call stack. I would guess that SAVE has a call to
> IDL_BailOut to which is why my program breaks on SAVE. If I
> didn't have that call to SAVE execution would continue back up
> the call stack to some point where the interrupt would be handled?
>
> What's the logic? Is this an issue of how I have structured
> my program or is this how IDL always handles interrupts and I
> haven't noticed until now?

I'll add that the GUI gets updated when program execution returns to the
main program event handler. For example, I batched 3 model runs where
the 3 runs are executed via a loop in the main event handler. GUI
drawing is suspended while within the model procedure, but when control
returns to the main event handler the GUI is updated right before the
program enters another model procedure (where GUI updating is again
suspended).

The thing that confounds this issue is that when the application enters
a model procedure the GUI is updated for a few seconds, maybe 20
seconds. *Then* the GUI fails to update. The reason why I never
noticed this until now is that all of our models ran in less than 20
seconds. Now with models that take 2 minutes to 4 hours the lack of
updating is apparent.

So now I am really curious how IDL handles this on other OSes. And it
is starting to smell fishy (that's funny. These are acoustic
backscattering models of fish. Get it? Fishy? <nudge> <nudge> <wink>
<wink>)

<sigh> I need to make a pot of coffee.

-Rick

Report message to a moderator

[Message index]

		Re: GUI interface update issues By: Rick Towler on Wed, 22 June 2005 08:57
		Re: GUI interface update issues By: Rick Towler on Tue, 21 June 2005 15:16
		Re: GUI interface update issues By: Karl Schultz on Tue, 21 June 2005 13:04
		Re: GUI interface update issues By: Haje Korth on Tue, 21 June 2005 11:23
		Re: GUI interface update issues By: Rick Towler on Tue, 21 June 2005 11:07
		Re: GUI interface update issues By: Benjamin Hornberger on Tue, 21 June 2005 11:01
		Re: GUI interface update issues By: Haje Korth on Tue, 21 June 2005 10:05
		Re: GUI interface update issues By: Benjamin Luethi on Tue, 21 June 2005 09:59
		Re: GUI interface update issues By: Haje Korth on Wed, 22 June 2005 05:16
		Re: GUI interface update issues By: Karl Schultz on Wed, 22 June 2005 09:22
		Re: GUI interface update issues By: Rick Towler on Wed, 22 June 2005 09:37

Previous Topic:	MPFIT2DPEAK
Next Topic:	Re: iSurface lighting questions

-=] Back to Top [=-

[ Syndicate this forum (XML) ] [

] [

]

Current Time: Sun Oct 12 04:45:01 PDT 2025

Total time taken to generate the page: 0.24184 seconds