Re: Arcs for Xr

On Dec 5, Owen Taylor wrote:
 > > Would XrArcTangent be a better name than XrArcTo?
 > 
 > _Something_ would definitely be better than XrArcTo :-). 
 > 
 > (XrArcTo would be better as XrArcTO since I think the O is 
 > from 'Out', but that would still be a pretty bad name for
 > anybody but someone familiar with postscript)

I realize that in PostScript the "t" and "o" certainly function as
"tangent" and "output", but I think there is a case to be made for
ArcTo as well. Compared with CurveTo there are some similarities:

	Both specify curves that fit within a control polygon.

	ArcTo is a lower-order curve than CurveTo's cubic, and its
	control polygon has one fewer point.

	In both cases, the control polygon extends from the current
	point "to" a point specified as an argument.

Of course, there are differences. ArcTo requires a radius in addition
to the polygon. It also doesn't necessarily draw an arc all the way
"to" the final point in the control polygon. That's the biggest
failure of the name, but in common cases such as the rounded
rectangle, ArcTo draws exactly "to" the specified point and feels
quite natural I think.

 > XrArcTangent isn't bad. I might suggest XrArcTangents 
 > just so it doesn't sound like the trig function.

Ugh. That's a pretty bad name clash. Maybe XrArcSlopes would be more
palatable?

I'm liking XrArcTo more all the time. I know I'm bad at inventing
names, so at least with ArcTo I can claim innocence.

-Carl

-- 
Carl Worth                                        
USC Information Sciences Institute                      [email protected]