Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!watmath!clyde!burl!ulysses!bellcore!decvax!decwrl!pyramid!amiga!jimm
From: jimm@amiga.UUCP
Newsgroups: net.micro.amiga
Subject: Re: Defualt actions for System Requesters
Message-ID: <1223@amiga.amiga.UUCP>
Date: Mon, 26-May-86 15:59:34 EDT
Article-I.D.: amiga.1223
Posted: Mon May 26 15:59:34 1986
Date-Received: Wed, 28-May-86 04:35:27 EDT
References: <431@elmgate.uucp> <288@cbmvax.cbmvax.cbm.UUCP>
Reply-To: jimm@homer.UUCP (Jim Mackraz)
Organization: Commodore-Amiga Inc., 983 University Ave #D, Los Gatos CA 95030
Lines: 46

In article <288@cbmvax.cbmvax.cbm.UUCP> randy@cbmvax.UUCP (Randy Weiner) writes:
>In article <431@elmgate.uucp> jdg@elmgate.uucp (Jeff Gortatowsky) writes:
>>
>>Does anyone know if Amiga Kernal 1.2 has changed the system requesters so
>>that, say, pressing return causes a defualt action?...
>>...a pain to have to 'mouse' on over to the cancel gadget. 
>
>	Well Jeff, there is good news and bad news. First the good news.
>There are Amiga menu short cuts in V1.2 that apply to Autorequest
>requesters. The bad news is that the key assignments are not
>obvious. The short cuts are:
>	retry ==  LeftAmigakey + B
>	cancel == LeftAmigakey + V
>

I wish Randy would have made this comment during alpha phase, when 
we requested input.  :-)

The rationale behind B and V is as follows:
    They are next to the M and N keys which (for some historical reason)
     bring the workbench to the front and back (with LeftAmiga qualifier).
    Not all autorequests make it clear which choice is "affirmative" and
     which "negative," so positional queueing has merit.
    Even ignoring the previous point, an initial/mnemonic choice of keys
     would need to be translated for different countries (and would change
     when a single machine owner used  different language set-ups).
    
So the good news is that there is no bad news!

I like the idea of "default choice" being available via carraige return,
but it's not always possible to retrofit consistent themes to something
out in the market.  BUT: there is a new feature which allows your application
to do such a thing.  Setting the NOISYREQ flag bit in the Requester structure
means that the requester being active will NOT inhibit normal input to
your window.  So if you set the RAWKEY IDCMPFlag in your window you can
hear all keystrokes while your requester is active, and act as you wish 
on Return or whatever.  Bearing in mind that you can end a Requester under
program control, your requesters can now serve as mini-menus or what
have you, and can be more useful in keyboard-oriented applications.

Beta sites: try the new Requester features and report back soon if 
they aren't behaving as advertised or there are gaps in functionality
that still prevent them from serving your needs.

			Thanks,
			    jimm