On 9/30/2011 1:18 AM, Invisible wrote:
> On 30/09/2011 03:14 AM, Patrick Elliott wrote:
>
>> Depending on your definition of "computer"
>
> There is that too.
>
> Turing-completeness is a reasonable definition, until you consider that
> a sheet of paper and a pen is Turing-complete given a suitable human to
> operate it. So perhaps the significant thing is the sophistication of
> computations that the device can perform without human aid.
>
>> there is evidence of a one
>> of a kind "Roman" device that was capable of predicting eclipses, and
>> timing the correct date to start the Olympic games, among other
>> features.
>
> Last I heard, nobody had decided exactly what that device was for. It
> seems opinions have changed...
>
>> In other words, an electronic calender.
>
> I think you mean /automated/ calendar. It's only /electronic/ if it
> operates by moving electrons around. :-P
>
Uh, yeah.. lol

>> But, at that time,
>> such things where nearly impossible to replicate, so when the ship it
>> was one sank...
>
> We're talking about something from a /long/ time ago. The fact that no
> others have been found yet doesn't mean none existed.
>
This is true, but from a purely practical standpoint, the expense, 
precision, and engineering needed to get just "one" to work, using 
basically copper and bronze gears, and little if any fast and replicable 
casting methods (where talking hand making each gear here), the odds of 
an exact copy is pretty much nil. The odds of something similar, still 
close to zero. Something much simpler.. possibly, but its basically a 
clock, more or less, and I am pretty sure that water clocks where just 
about as close as they ever got, based on everything known/found/written 
about, other than this thing, to that level of precision. Even some 
mechanisms they used in temples, to open and close doors, where "one 
off" designs, which where never exactly replicated, and are only half 
known, half speculated, at this point (given the missing bits, and 
descriptions we have to go off of).

If you wanted something that could use water to time things to the hour, 
or measure an exact amount of "holy water" to do something with in a 
temple, you where in good shape, since the mechanisms didn't need to be 
*that* precise. Something like a mechanical calender... Hope you own 
stock in the Colosseum. lol

>> Arguably, its gearing system had to have some sort of "algorithm".
>
> By that description, the way that trees use the laws of physics to move
> exactly the right amount of water from their roots to their leaves could
> be considered an "algorithm". Which would mean that algorithms predate
> mankind by several billion years...
Umm, well.. I would argue that what makes it an algorithm is the 
"intentional" development of the steps and process needed to form a 
predictable result. Gearing matches this definition. Stuff that just 
kind of does that already, naturally... is a bit iffier. But, that isn't 
to say that, say genetics, couldn't contain something that could be 
"interpreted", if you had the right JIT compiler, so could be, to an 
extent, algorithmic. Water happening to get from roots to leaves.. not 
so much.

Post a reply to this message