>
>>The main problem and it has been mentioned here is that CAD wants to be 3D
>>and that gets into solid modeling and tolerancing. Somehow all this got
>>turned into a pannecia. A magical solution that works for others.
>
>I haven't made a 3D CAD drawing yet.  Most work to be done on a lathe or
>mill is easily expressed in 2D, and designing things that NEED 3D is
>probably a warning sign at our level.
>
Intersting, When I was learning CAD I had to rock the 3d model back and
forth to se the image. The professor was quite curious as to why I was
dooing this. I needed to see the perspective shift in order to visulise the
3Dness of the model.

Mostly I am interested in photogrmeretry. What is taking multiple images
and creating a virtual model of the scene. I have an old Nasa book that
talks about this. One day I was working on a color scanner on assignment in
New Hampshire and talking to the Old man who made this colorementrically
perfect scanner the size of a refrigerator. He brought in the book put
together for his team of the martian landscape, In the back were dozens of
images of the martian surface, But he wanted to show me a picture of some
sort of car he had scratch built. In looking at the picture of the camera I
asked "Who actually made the camera that sits on mars." His look told me I
was correct.
Recently I found this book in a list of books for sale, so have my own
copy, but that means one of the the origional owners is no more.

It really does seem there are gender based ways of looking at things such
as 3D models in space. Micheal Wallis can see topographic features in
single images, where as I need at least two. I remember when ERPS was at
the tv studio and we were lookinging at photos of Venus. That a few lines
can make a whole construction. Perhaps this is why there are only a few of
us on the distaff side here. Sometimes I think math is the natural language
of the male brain and latin that of the female brain.


>I am probably going to have to do my first 3D drawing when I get a quote
>for a hyperbolic gun drilled nozzle
>

The real issue of CAD is the tolerancing of the dimentions especially when
there are only 2. Take clothing the human body for example. Taylors tend to
print patterns, where as women tend to drape the fabric and cut it out. A
pattern only needs to be accurate to 1/8 inch. The trick to fitting clothes
is to know where to be accurate.

The old system of meausurments were accurate to say 1/32 of an inch, a 64th
if one was precise. How many floating point numbers fit into 16 or 32 bits?
they are only an aproximation. You have the same storage as the integers,
no more no less. If I want to represent one sixtyfourth of anything it
means I have to have a lot of bits of storage. When this is printed as
decimal the machinist said the tolerance is unobtainable. 1/2 inch is
precicely 1/2 inch. where as 0.503234 is what exactly?

Real numbers do not exist in the real world and imaginary numbers can give
you quite a shock if you touch a capacitor the wrong way. No wonder I
failed math.

An archetect told me that as long as an inch or a foot or a yard can be
divided into quarters and thirds with no remander they will be used. Sure
we have ten fingers, but where is the .333333333333333333333333 finger.
Accountants love decimal numbers becouse it is so easy to loose the
fraction of a cent here and there.

-julie :-)

>John Carmack
>


_______________________________________________
ERPS-list mailing list
[EMAIL PROTECTED]
http://lists.erps.org/mailman/listinfo/erps-list

Reply via email to