machine figure (re)

2008-07-24 Thread Jason mclaughlin
can't a machine always be seen where every part of the machine is one
way, for another part to be one way?

i mean... say you can take a bunch of parts and have it so each part
has only one way to be.
then parts can be together, where they fit the way that each part can
only be one way for the way another part is.
imaginarily ... can't they always fit together to say a machine?

like.. if I say a few parts mean a number, and a few other parts mean
another number, and parts between the numbers fit them together, with
another part that fits
like it's something about the two numbers.

so for everyway that a number part is, is for the other part like how
the other parts are, for another part that fits how the numbers are
together.

so each part fits exactly like how the two numbers fit together with
parts and another part that has something about the numbers, so that
the way the numbers are is the way
the other parts are, is the way there's something about the numbers...

so if i move a part, which way do i move another part? think about that...
what if a part is taken off and i try to fit another part? how do i
make it fit the other part?
but what happens to how the number parts fit if i have to make all
parts fitting?

can't parts always come together so for the way a few parts are, are
for the way a few other parts are?

can't parts together like this always say a machine?


machine figure

2008-07-23 Thread Jason mclaughlin
Say pieces on a board, make each a pair with another piece.

like...

|55|44|66|
|44|66|55|

so figure out how a piece can move.

pick any piece, try to move it somewhere.
when you move a piece you have to move it's pair at the same time.
when you move to a piece it's pair has to move at the same time too.
a piece always becomes a pair with the piece it moves to.
no matter how many pairs, there's only one answer to how a piece can move.

A common problem, I forget what it's called.

There's only one answer for how any piece can move, another piece has
to move where the piece left.

No piece can move to where a piece moves from.

No such thing as a free space, a piece always moves to another piece.

A pair never moves to a pair.


so try this...

draw for each piece a line from one piece to another that connects
each piece to move from the first piece until the last piece that goes
back where it starts.

see this as a machine diagram.

move a piece then figure the machine diagram again.


Re: machine figure

2008-07-23 Thread Mark Mielke

Jason mclaughlin wrote:

Say pieces on a board, make each a pair with another piece.

like...

|55|44|66|
|44|66|55|

so figure out how a piece can move.
  


This is the second post I've seen like this - and no responses.

What is the proposed value of this line of thinking? Is there some 
particular optimization that can be designed as a result? Do you have 
verifiable evidence that it will provide value? Perhaps sample C code 
and the theoretical assembler code you think should be generated 
alongside the current assembler code generated by GCC? Or is this 
rambling thoughts on a subject? I'm interested in reading about the 
former. I have no interest at all in reading about the latter. :-)


Cheers,
mark


--
Mark Mielke [EMAIL PROTECTED]