V. good.  Thank you.

I last thought about this problem in 1st year college, many decades ago.
At that time a fellow freshman Norman (who went on to get a Ph.D. at MIT)
argued, for x%y, that there must be a 1 and then did the construction for
x%y using 1.  I recall he said "there must be a 1" in the sense of "it has
to exist" rather than in the sense that "you have to use a 1 in the
construction" or "you can not construct x%y without using 1".  I don't
remember what we did with x*y; there was some doubt in my mind earlier
today that we _were_ able to construct x*y without using 1 all those years
ago.



On Thu, Jul 14, 2016 at 10:46 AM, Raul Miller <[email protected]> wrote:

> Proof that x*y requires reference length:
>
> if x is reference length then x*y is y
>
> if y is reference length then x*y is x
>
> If we do not know the reference length and if x and y are different
> then we do not know if x*y is x or y but we know that it could be
> either of those.
>
> --
> Raul
>
>
> On Thu, Jul 14, 2016 at 1:19 PM, Roger Hui <[email protected]>
> wrote:
> > Given line segments x and y, construct (using compass and straight edge)
> > line segments having the following values:
> >
> > x+y
> > x-y
> > x*y
> > x%y
> >
> > The first two are immediate.  I have proven that x%y is impossible if you
> > are not given a reference length 1 (or some other reference length from
> > which to construct 1).
> >
> > The problem is, prove or disprove that you construct x*y without using 1.
> >  (Constructing x*y and x%y _with_ 1 are pretty easy.)
> > ----------------------------------------------------------------------
> > For information about J forums see http://www.jsoftware.com/forums.htm
> ----------------------------------------------------------------------
> For information about J forums see http://www.jsoftware.com/forums.htm
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm

Reply via email to