W3C home > Mailing lists > Public > www-style@w3.org > December 2009

Re: [css3-2d-transforms] skew(<angle> [, <angle>])

From: Dr. Olaf Hoffmann <Dr.O.Hoffmann@gmx.de>
Date: Tue, 8 Dec 2009 12:48:55 +0200
To: www-style@w3.org
Message-Id: <200912081148.55645.Dr.O.Hoffmann@gmx.de>
Hello www-style,

due to the previous discussion of skew(<angle> [, <angle>])
it turned out, that skew(<angle> [, <angle>]) is currently
undefined and what was reported to be implemented is
not related to skewing and meaningless. 
Therefore I would like to propose to define it as follows:

skew(s,a) = rotate(a) skewX(s) rotate(-a).

If a is not provided a=0 is assumed.

What it means:
To skew in an arbitrary direction one can for example
rotate the skew direction into the x-direction, followed
by a skewX and followed by a rotation in the opposite
direction again.

If the skew angle is s and the angle from the x-direction to
the skew direction a, then the corresponding transformation
is rotate(a) skewX(s) rotate(-a).

The  attached graphics correspond to examples with s of 40 degree.
The filled samples on the left correspond to an rotation of
90 degree. Note, that for 90 degree this corresponds to
The stroked samples provide several skewing directions with
a difference angle of 45 degree.


(image/svg+xml attachment: SVG12skew01.svg)

Received on Tuesday, 8 December 2009 10:51:12 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 22:49:35 UTC