Path: utzoo!utgpu!news-server.csri.toronto.edu!bonnie.concordia.ca!uunet!elroy.jpl.nasa.gov!usc!ucla-cs!oahu.cs.ucla.edu!stephen
From: stephen@oahu.cs.ucla.edu (Steve Whitney)
Newsgroups: comp.sys.atari.st.tech
Subject: Re: How to use the VDI from the Autofolder
Message-ID: <1991May27.104903.10283@cs.ucla.edu>
Date: 27 May 91 10:49:03 GMT
References: <1991May24.184249.21414@imada.ou.dk> <1991May25.152208.2925@actrix.gen.nz> <3491@quepasa.cs.tu-berlin.de>
Sender: usenet@cs.ucla.edu (Mr. News Himself)
Organization: UCLA Computer Science Department
Lines: 51
Nntp-Posting-Host: oahu.cs.ucla.edu

Karsten, I'm not attacking you.  You're quite correct in the things you say,
but in defense of Mr. Johnson, I'm the "Steve" he was replying to on GEnie,
and that was _years_ ago.  Since then, he has written the "11 Commandments
of Trap Stealing" (or something like that) which addresses your concerns.

In article <3491@quepasa.cs.tu-berlin.de> ki@opal.cs.tu-berlin.de (Karsten Isakovic) writes:
>
>In article <1991May25.152208.2925@actrix.gen.nz> Roger.Sheppard@actrix.gen.nz
>(Roger Sheppard) writes:
>>This came from a File that I found some time back on Genie, it could
>>be of help to otheres, with the Trap #2 problems.
...
>The code does not use the XBRA method, which _is_ standard...

XBRA didn't exist yet when this snippet was captured.

>And it does not work with RAM-TOS versions, or even with the new STE or
>TT machines...  (Please use the os_start entry in the _sysbase structure
>pointed to by the system variable $4f2.)

Quite correct.

>This method is not good. The 'unused' function call might be used in a 
>future TOS version.
>You simply have to install a Cookie interface for the TSR. The value of
>the cookie points to an internal TSR structure, so that a DA, CPX or PRG
>can configure the TSR.

Again, the cookie jar didn't exist back then, and installing a cookie
isn't really that simple.  Unfortunately, any program that wished to install
a cookie must be prepared to creat a cookie jar if one does not yet exist.
This wouldn't be a problem in itself, but if it _does_ install a cookie jar,
it has to install a reset routine to clear the cookie jar location on a
reset.  Any program installing a cookie must also be prepared to install
a new jar if the current one is full.  In that case, the program must copy
all the cookies to the new jar, install its own cookie, and install a
routine at the reset vector.  Fun, huh?

By the way, I imagine that you know this, but some of our gentle readers
may not have played with cookies yet...

>Sincerly, Karsten   (ki@opal.cs.tu-berlin.de)

	--Steve


-- 
  Steve Whitney - UCLA CS Grad Student                       (())_-_(())
 Soon to be working at Silicon Graphics                       | (* *) | 
     Internet: stephen@cs.ucla.edu          UCLA Bruin-->    {  \_@_/  }
          GEnie:    S.WHITNEY                                  `-----'