Path: utzoo!attcan!uunet!samsung!uakari.primate.wisc.edu!dali.cs.montana.edu!ogicse!oregon!jmeissen
From: jmeissen@oregon.oacis.org (John Meissen)
Newsgroups: comp.sys.amiga
Subject: Re: Difficulty in programming
Message-ID: <515@oregon.oacis.org>
Date: 5 Jun 90 16:47:11 GMT
References: <56964@bbn.BBN.COM>
Organization: Oregon Advanced Computing Institute (OACIS), Beaverton, OR
Lines: 46

People think the Amiga is difficult to program for a variety of 
reasons, but so far I haven't seen any that make sense. For
instance,...

In article <56964@bbn.BBN.COM> cosell@bbn.com (Bernie Cosell) writes:
>I think you people are missing part of the point.  Granted that
>programming for a really flexible GUI on a multitasking machine has
>some inherent difficulties.  NONETHELESS, AmigaDos has a bunch of what
>I think are bad design decisions that make it a *lot* harder than it
	[etc. etc.]
>1) the absence of resource tracking.  This decision pollutes the Amiga
>at every level.  It makes msot routine programming tasks a (relative)
>nightmare, since you have to be anal-retentive to a degree I find a
>constant nuisance.  Freeing locks, closing open files, freeing memory,
>etc, etc.  It makes interrupt-handling code a mess, it means that it is
>hard to get rid of a runaway program...  altogether a loser.

Oh, give me a break. This is just lazy programming. Any programmer
that wasn't careful about closing open files and freeing memory, etc.,
wouldn't be working for me for very long. And I don't care WHAT
system he is writing programs for. It can be nice to have the
system clean up after you when you make mistakes, but it's not
commercial quality software when your program depends on it.

>2) Having the sender be the 'owner' of messages.  Having a
>message-passing paradigm within the operating system is a lovely and
>clever idea.  Having messages be the 'property' of the *sender* was a
>serious mistake: again, making programs more complicated, making it
	[long dissertation deleted]

This is another misconception. Remember, there isn't any resource
tracking on the Amiga. No process is the "owner" of any memory. It's
merely convention that a process return a message when it is through
with it. This has more to do with synchronization than ownership.
When the message comes back, you know the recipient is finished with
it. There's nothing at all preventing you from implementing a
different scheme within the framework of your own programs.

>   /Bernie\


-- 
John Meissen .............................. Oregon Advanced Computing Institute
jmeissen@oacis.org        (Internet) | "That's the remarkable thing about life;
..!sequent!oacis!jmeissen (UUCP)     |  things are never so bad that they can't
jmeissen                  (BIX)      |  get worse." - Calvin & Hobbes