[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [MiNT] MiNTLib for ColdFire



Vincent Rivière píše v St 29. 04. 2009 v 09:37 +0200:
> Petr Stehlik wrote:
> > I understand that uninformed people feel like ColdFire is almost m68k
> > compatible so they think it will be easy to run their existing Atari
> > binaries on it. Unfortunately the problem is in the word
> > "almost" (compatible)
> 
> I totally agree with you. But the ColdFire is the only "modern" CPU of the 
> m68k family, that's why it is seen as the only possible improvement by 
> Atari-hardware enthusiasts. No rational arguments can change that.

Well, the key thing is to realize that the ColdFire is actually NOT part
of our m68k family. If it were it would have compatible instruction
set... 

> people are frustrated if they can't feel the hardware under their fingers, 
> and they cannot imagine typing on a keyboard connected to little-endian 
> hardware... Things are like this.

I understand that (even though it took me quite some time to realize
that). I would just like to point out that the ColdFire isn't any better
"upgrade path" than say Intel or AMD.

When programmers are going to spend their time to adapt the existing
source code to run on a new processor, to adapt operating system, to add
m68k emulation to it, then why don't they do it the way Apple did and
move to a more mainstream platform (Apple did it even twice, both time
successfully - first PPC, then Intel)? Doing the same work for a
hardware that doesn't even exist yet is ... how to say that. Not smart?

Anyway, I understand it's fun project so it need not to make sense
commercially or in any other way. My only concern here is that while
binaries compiled for a totally different CPU than m68k wouldn't run at
all on Atari computers or clones the ColdFire binaries would run - they
would just misbehave slightly... And I'd like to prevent that
misbehavior, user frustration and disappointment by adding the CPU
detection to at least the MiNTlib.

Anyway, I think that if the ACP will be successful and some programmers
start writing software for that hardware without proper CPU checks then
it will just cause confusion and will split the remaining small user
base.

Petr