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

Re: Social Contract GR's Affect on sarge

Don Armstrong wrote:
> > The human brain is way too limited for that task. Just explain me
> > how your brain interprets a bitstream which represents a JPEG of
> > non-trivial size. :-)
> Not particularly well, but I'd imagine that there are people out there
> who wouldn't have that much trouble with it. Moreover, while
> difficult, it's not something that's inherently impossible.

It is, simply because the human brain can't handle more than 7+-2 items
at the same time ("The magical number seven, plus or minus two", may
even show up some google results for it). So a human is rather limited
WRT algorithmic complexity. You can, of course, construct an improved
interpreter which involves a pencil and a stack of paper, but that's
already a different thing.

> Furthermore, there's no reason to require that the data not be
> preprocessed into a form more suitable for the interpreter doing the
> interpretation.

This preprocessing would mean a conversion to a different format,
with different properties. Like from EPS to PNG, for example.

> [Or do you want to require that all interpreters talk
> only in machine code?]

I don't know of even a good definition for "machine code". E.g. a x86
CPU has a microcode interpreter for some of its "machine codes".

> Finally, there's no requirement in your definition for there to be
> reasonable (or correct) output

The requirement of an algorithm is to produce the same results for the
same input.

> or for execution to take a reasonable
> amount of time. [You can't really add these two requirements, because
> you'll eliminate buggy programs and infinte loops respectively, both
> of which are quite obviously programs.]

The latter is a classic. :-)  If it doesn't terminate, it can't be an
algorithm by definition, and thus no program.


Reply to: