Markem wrote:
On Sat, 23 Nov 2013 17:20:24 -0500, Bill
wrote:

Leon wrote:
On 11/23/2013 1:49 PM, Bill wrote:
Leon wrote:
On 11/23/2013 4:00 AM, Gunner Asch wrote:
On Thu, 21 Nov 2013 11:33:32 -0600, Swingman wrote:

On 11/21/2013 11:22 AM, Richard wrote:

SketchUp actually has the same internal precision as AutoCad,
1/1000th
of an inch.

Look it up ...

1/1000th inch? Is that all???
LOL

Good enough for AutoCAD, and certainly well within the
specifications of
the OP's original request.
Some of us have to work to .00005" or smaller.
Well even fewer of us work to .0000000000000000000000000000000000005"
or smaller but that still does not mean you need a program to do that
when .0001 is way more than enough.



What would the value of the integral over (-infinity, infinity) of
1/(1+x^2) dx look like if you didn't express it as the Greek letter
PI????
Even being in error by the amount above would make it that you bought it
from Kmart--apologies to Kmart (and/or the Sears Holding Co.).


LOL, I was just making a point that in woodworking you don't need to
work in tolerances that the human can't see. And The OP had on top of
his list, drawing a board, not friggin atomic particles.
I wasn't replying to your post specifically. I just wanted to get in my
2-cents about accuracy. As far as computers go, integers can match
exactly, but not numbers with decimal points, in general. You can ask
and get exactly 3 twobyfours, but not of any exact dimension! ; ) Of
course, the value of PI can be matched exactly--just not by a typical
computer. If one is willing to express numbers with base PI instead of
base 10 or base 2, then all bets are off.
In Windows 7 I noted, in a solitary game 7 out of 10 score is 69%.
Garbage in garbge out.

..7 can't be stored exactly as such as a floating point number on a
typical modern computer.
Someone "casted" the number to an integer, losing what what stored as a
fraction.

They used: (int)(average)
when they should have used: (int)(average+.5).

The latter would have rounded.
Mark