Path: utzoo!censor!geac!torsqnt!news-server.csri.toronto.edu!clyde.concordia.ca!uunet!snorkelwacker!apple!uokmax!munnari.oz.au!goanna!minyos!monu6!edp367s
From: edp367s@monu6.cc.monash.edu.au (Rik Harris)
Newsgroups: comp.sys.amiga.tech
Subject: Re: cia-b
Message-ID: <1990Sep7.090650.2931@monu6.cc.monash.edu.au>
Date: 7 Sep 90 09:06:50 GMT
References: <5172@uafhp.uark.edu> <1990Aug31.000652.27385@monu6.cc.monash.edu.au> <14242@cbmvax.commodore.com>
Organization: Chisholm Institute of Technology, Melb., Australia
Lines: 51

daveh@cbmvax.commodore.com (Dave Haynie) writes:

>In article <1990Aug31.000652.27385@monu6.cc.monash.edu.au> edp367s@monu6.cc.monash.edu.au (Rik Harris) writes:
>>bbs00049@uafcseg.uucp (Red Roger) writes:

>>>Hi i've seen some code that uses the ciab serial port set register. 
>>>what is the real address of this register ??
>>>bfd000 or bfd0fe - the reason i ask is that in the RKM h/w manual it says 
>>>bfd000 yet this code works with it at bfd0fe 

>>From the hardware point of view, it's quite possible that it works with
>>any even address from $bfd000 to $bfd0fe (including $bfd002, $bfd004, etc).
>>This could also be true for the other hardware registers in the $bf[de]xxx
>>range.

>However, you must always use the documented addresses for a register, if you
>MUST directly use the register in the first place.  The register may appear
>all over the place on an A1000, but only be in the specified location on an
>A3000, for example.  There's no advantage to using the replicated image
>addresses, unless you specifically want the program to fail on some systems.
>If you use the documented addresses, they'll work as long as those particular
>registers exist in a system.  Anything that can be done at the device rather
>than register level should work forever, even if a future system replaces
>those registers with something better and different.

Yes, I didn't really make that very clear.  I have had any number of problems
with programs that were designed by sloppy programmers that are happy to
jump to whatever o/s call they feel like, and use whatever hardware register
works.

It is _guaranteed_ to make this software break on future models of the same
machine.  People whinge that Commodore, or Atari, or whoever has made their
machine incompatible, when it is generally the software producers who can't
be bothered making their software use the documented addresses and routines.

If I didn't make it clear the first time, I hope this helps.

>Dave Haynie Commodore-Amiga (Amiga 3000) "The Crew That Never Rests"

-------------------------------------------------------------------------------
Rik Harris - Faculty of Computing and Information Technology,
       Monash University, Caulfield Campus (was Chisholm Institute), Australia
edp367s@monu6.cc.monash.edu.au     <-- what the computer says it is
edp367s@monu6.cc.monash.oz[.au]    <-- SEEMS to work more reliably
Build a system that even a fool can use, and only a fool will want to use it.
-------------------------------------------------------------------------------
-- 
-------------------------------------------------------------------------------
Rik Harris - Faculty of Computing and Information Technology,
       Monash University, Caulfield Campus (was Chisholm Institute), Australia
edp367s@monu6.cc.monash.edu.au     <-- what the computer says it is