On 10/3/2011 1:03, Invisible wrote:
> I thought the problem with reversible computing is that it's just as likely
> to run backwards (which is no use at all) as to run forwards?

Yes. Basically, it takes as much energy as you need to keep it moving 
forward. The less energy you give it, the slower it goes. If you give it no 
energy, it stops (on average) making forward progress.

OK, it doesn't take *no* energy. It takes arbitrarily little energy.

> By that definition, nothing that can exist in the physical universe will
> ever by Turing-complete. That's not a useful definition.

Yes it is, if what you're talking about is what's possible to even 
theoretically calculate.

> You said "it's always absurdly easy to prove that something is
> Turing-complete".

No I didn't. I said it's "usually not too hard."

> I hold up an example of something that took ages to prove,
> and even now some people argue the proof isn't very sound.

Sure. But that's because he's pushing the boundary. He's not interested in 
whether a CA can be Turing complete. He's interested in the absolute 
simplest CA that can be proven to be Turing complete. I'd argue that that's 
intentionally on the very boundary of what's most difficult to prove, rather 
than the "usually" case.

-- 
Darren New, San Diego CA, USA (PST)
   How come I never get only one kudo?

Post a reply to this message