Re: [MiNT] XaAES: menus in windows - window handle?

["Helmut Karlowski" <helmut.karlowski@ish.de>]

Peter Persson wrote:

> >> msg[3] =3D parent menu
> >> msg[4] =3D menu item
> >> msg[5] =3D menu tree (high word)
> >> msg[6] =3D menu tree (low word)
> >> msg[7] =3D parent object index
> >=20
> > Don't give [3] or [7] any information?
>
> I use [3] to find out which menu it is. Somehow I don't get what 7 is =
> for, I have the impression that if I use [5] + [6] to get the tree =
> address, [7] is redundant. Maybe I'm wrong, dunno.

Maybe this says something to you:

      titles = k->m.titles;
      about = k->m.about;
      ks = k->p.parent;
      kc = k->m.current;

ks is msg[7], kc msg[3]. I don't know what's what ...

>
> > It would not be very easy to make the message longer - it's a generic
> > funtion for all messages, better would be to misuse another parameter =
> -
> > trying to avoid the risk that some other app may get confused.
>
> Some other app shouldn't get confused, because that app shouldn't have =
> received the message in the first place?

No, but this would be done for every submenu in any app that has one.

>
> How many apps currently use this feature? (if any?) :)

There's only keyedit I know of, but this has just one window. I'm sure Jo Even can tell more
about this than I.


>> msg[1] =3D ap_id
>=20
> This should also be 0 according to the source.

> The source, yes :)

> Maybe the ap_id field only makes sense when using appl_write()? Let's =
> say we put the window handle in there - could this break compatibility =
> with any real applications?

I really do not think that it's a valuable information for a client to
know who sent an MN_SELECTED (who else could/would do this other than
the AES?)


-Helmut