IN the environment menu, indexing post process

   using K (key function) and {ESC}

   or a subroutine with esc=char(27)

   & calling catalyst("k" , esc)

                  the screen stays up.. until a key is actually                     pressed on the

                  keyboard.. 

   and testing from another session... the window does not

   release the record until a key on the keyboard has been pressed

   Is there a way to "prime the pump" and get the window to 

   close (keeping in mind that we do not know if we are in a 

   window or at a menu)?

                  Any help is appreciated

Well you could check @tutor (I think - but I'm at home and can't check FKB) and react differently accordingly…

[email protected]

Sprezzatura Ltd

World Leaders in all things RevSoft

I recently had a very similar problem. How to kick an application on when the user walked away from it. Because of the way the post-indexing process works, a catalyst command stays at the level of the process. When the post-index process stops, and control returns to the previous level, the catalyst keystroke was lost. The only way I found to get around this was to load the @DATA variable, which is the keyboard buffer.

I believe the ansi sequence for ESC is char(27). I used the following code to achieve the desired results:

ESC=CHAR(27) ; *I think this is the escape sequence, but check first

IF TIME.TO.ESCAPE THEN

  @DATA=ESC

END ELSE

  NULL

END

STOP

My logic is slightly different, because of some other considerations, but that should do the trick. When your post-index process stops, the keyboard buffer will have an ESC keystroke stored in it. When control passes back to the window, the keystroke(s) stored in the buffer will be executed.

Hope this helps.

Matt Sorrell

MGA, Inc.

Matt… it worked great..

all I did was use a subroutine that essentually put

the escape into the variable @DATA

   @data=char(27)

                   Thanks loads

I recently had a very similar problem. How to kick an application on when the user walked away from it.

This sounds like the urinal application at the airport.

Guys,

What happened to PEEK?  Did it die?

If you want to know if you're in a window, RTI announced a LONG time ago that @TUTOR was verboten and unreliable.  I forget who came up with this (Steve Smith maybe?), but I use an INCLUDE called 

InWindow$:

 declare function SSPEEK

 InWindow$=(SSPEEK(1160)=135)

Don't worry, the Include name and the variable don't clash, but you can call them anything you want. Only if there are 135 system variable open are you in a window.

It seems to be the only TRULY reliable thing around. BTW, it seems I remember AMcA having it in his newsletter.

Larry Wilson

View this thread on the forum...