8 views

### martin_gnu

Oct 18, 2007, 1:53:00 PM10/18/07

I have added several Franz primitives. At the end we have the new
especification.

there are a web with this kind of things in STEP???

Regards.

Oct 19, 2007, 7:13:35 AM10/19/07

### Gertwin

Oct 25, 2007, 3:52:13 PM10/25/07
Hi,

The point definitions are not clear to me.

I understand the simple constructions like P(x, y, z) but the
construction P2=P1, D1 is unclear.
D1 is a vector so it must be something as P2=P1+D1, the vector D1
starts at P1 and P2 is the point where the vector ends.
What is the meaning of the ',' in the expression P2=P1, D1?
Also P2=P1-D1 can exist, how is this expressed?
The same for P2=P1, D1, distance, i think this is something like
P2=P1+s.D1, s is a scalar.
And sometimes the ',' is omitted, how must the space between the
operands be interpreted?

I have a feeling some information is missing.
In the expression P11=R1 P10 i think a transformation of p10 from a
ucs to wcs is ment, i am missing the information how transformations
are constructed, does P11=R1 P10 mean the same as P11=P10 R1?

Do we need some kind of "formal" description of our format or is
something like that available?

Greetings,
Gertwin

### Franz Reiter

Oct 27, 2007, 3:16:49 AM10/27/07
Hallo,

> construction P2=P1, D1 is unclear.
> D1 is a vector so it must be something as P2=P1+D1, the vector D1
> starts at P1 and P2 is the point where the vector ends.

correct

> What is the meaning of the ',' in the expression P2=P1, D1?

> Do we need some kind of "formal" description of our format or is
> something like that available?

the ',' as well as one blank as well as ',' plus one ore several blanks serves
as a word-delimiter, nothing else.
We should clearify that in the general format description.
Lowercase characters are translated into uppercase (exception: filenames).

> Also P2=P1-D1 can exist, how is this expressed?

For objects there is no sign possible till now; you could use:
P2=P1 D(D20 REV)
# this creates a dynamic vector opposite (REVers) to D20.

> I have a feeling some information is missing.
> In the expression P11=R1 P10 i think a transformation of p10 from a
> ucs to wcs is ment, i am missing the information how transformations
> are constructed, does P11=R1 P10 mean the same as P11=P10 R1?

usually you cannot change the sequence of the parameters;
the scanner analyzes: outputObj=inputObj's and must find the correct operation
for this given parameters.

Transformations have 'T'
# translate with vector:
T1=D(10 0 0)
P2=TRA P1 T1
# same as P2=P1 D(10 0 0)

# define rotation: around axis(point vector) angle(deg.)
T2=ROT L(P1 DZ) ANG(30)
# rotate point p2
P3=TRA P2 T2

# translate between 2 coordinate systems:
# from wcs to ucs:
P1=P(10 0 0)
R20=PERP P(25 0 0) DZ DX
P2=TRA P1 R20
# absolut 35,0,0

# from ucs to wcs:
# create ucs (interact: PLN PT Z-Axis X-Axis)
R20=PERP P(25 0 0) DZ DX
# activate it
R20
# create point 10,0,0 in ucs; absolut in 35,0,0
P20=X(10)
# copy P20 > wcs (interact: Transl.obj(PLN))
P21=TRA P20 R20 REV
# this creates the inverse-matrix of R20 and transforms P20.
# the result is in wcs (10,0,0)

not easy to explain for me in english ..

Franz