Change in processing of Accept/Quit trigger sequences between V8.4 D608 and V9.4

Author: gls.tools@free.fr (Gilles)

Until the In V8.4 D608, processing of <Accept>/<Quit> trigger sequences was:

As of UNIFACE V8.3.02 or M311, child components will not close until the moment
the parent has confirmed that all child forms can successfully close.

Since V8.4 D609 and V9.4,processing of <Accept>/<Quit> trigger sequences is :

Childs components can close although the parent has not confirmed that she can

it's not a bug....thus....

no one was warned ...it's normal.....

 

5 Comments

  1. I know :-(

    So I programmed my own SP_QUIT which will "ask" all depending instances (belonging to a TAB-widget) if or if not a quit will be accepted. If not, SP_QUIT will return $status=-1 and so the instance will stay alive.

    A little bit tricky and sometimes there are still "loses" of instances :-)

    Ingo


    Author: istiller (i2stiller@gmx.de)
  2. ... wouldn't it be a good idea for the CPWR technical staff to provide some example code

    "bringing all back the good times"?

    I think they have good reasons for changing the behaviour,
    on the other hand, they should have the best insider-knohow to provide the old functionality.

    Success, Uli


    Author: ulrich-merkel (ulrichmerkel@web.de)
  3. This means that we should stop using trigger QUIT / ACCEPT, to control
    if the users can or can not exit the component ?
    The trigger QUIT / ACCEPT triggered the foremost trigger QUIT / ACCEPT child components.
    If the first child component does not return an error it closes without waiting for other
     child components /or parent component are closed without error.

     


    Author: Gilles (gls.tools@free.fr)
  4. Hello,

    There seems to be a problem when calling the Quit / Accept operation (e.g. “activate $instancename.quit()”) when using version 8.4.06 + D611 (or higher), version 9.2.03 resp. Service Pack MO01 (or higher), and version 9.3(.01 or higher). This issue has been recognized as (internal) bug report #28999 and a solution is (currently) planned for the (version 9.4.01) patch R109 (scheduled release date February 4, 2011) – a fix will most likely also be back ported to version 9.3 (probably to patch P219, planned for February 18, 2011).

    If instead the Quit / Accept trigger is called (either by using e.g. 'macro "^QUIT"' / 'macro "^ACCEPT"' or with the structure editor function <GOLD>+Q / <GOLD>+A) then the deletion of the attached (child) instances is canceled in the described scenario (as expected).

    Hope this helps.

    Kind regards,
    Daniel

    *** Usual disclaimer ***


    Author: diseli (daniel.iseli@uniface.com)
  5. hello daniel,
     
    it's a damn good news ;)
    I did not find the bug report 28999 on Frontline's website. it's official or not yet ?
    Thanks a lot.
     

    Author: Gilles (gls.tools@free.fr)