Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!think.com!spool.mu.edu!rex!ames!ncar!hsdndev!rice!uupsi!sunic!fuug!clinet!dix
From: dix@clinet.fi (Risto Kaivola)
Newsgroups: comp.sys.amiga.hardware
Subject: Re: HELP! (please!) on OS...
Message-ID: <1991May26.185945.24811@clinet.fi>
Date: 26 May 91 18:59:45 GMT
Article-I.D.: clinet.1991May26.185945.24811
References: <1991May24.212123.11386@lynx.CS.ORST.EDU>
Organization: City Lines Oy, Helsinki, Finland
Lines: 38

rudolpe@urubu.CS.ORST.EDU (RUDOLPh ERIC) writes:

>HELP! I'm trying to do a little program in assembly that sets up a viewport and
>displays it. However, I can't seem to restore the original Workbench display. 
>I know this:
>I must call LoadView() while passing a pointer to the View Structure in a1.
>However, I can't seem to find the original View Structure. My manual says to
>find it in the offset ActiView in the GfxBase structure. My book says the
>offset is 2. I don't believe it. It looks wrong, the the program immediately
>Gurus, with no Guru number, because the display is frozen.
According to "Amiga ROM Kernel Reference Manual: Includes & Autodocs
(revised and updated)", i.e., the book that contains listings of all the
standard include files provided by Commodore, struct GfxBase is as follows:
[p D-38]
struct GfxBase
{
  struct Library LibNode;
  struct View *ActiView;
...
 
That is, the pointer to the active view cannot be in the offset 2 for the
first *member* in the structure is a structure, the size of which is 34
bytes. The second element in GfxBase is a pointer to a View. In other
words, the view pointer is in the offset 34, calculated from the beginning
of the GfxBase structure. You seem to be using assembly language, and
this C'ish explanation is not perhaps a very clarifying one, but I hope
this helps.
  When I had a problem of the same sort, I solved it using either
RemakeDisplay() or WBenchToFron() [can't remember it off hand, sorry].
BTW, if you are programming the WB environment, I recommend using some
high-level language, like C or C++. It can save you a lot of hair :-).

>Anyone who could shed some light on this would have my infinite gratitude, and
>many profuse thanks!

>rudolpe@jacobs.cs.orst.edu
-- 
Risto Kaivola (dix@clinet.fi) or (Risto.Kaivola@f123.n220.z2.FIDONET.ORG)