404221 - SVGMatrix multiply method (in JavaScript) has its arguments swapped.

Status: RESOLVED INVALID

(Core :: SVG, defect)

Description

[David Jones]

User-Agent:       Mozilla/5.0 (Macintosh; U; Intel Mac OS X 10.4; en-US; rv:1.9b2pre) Gecko/2007111704 Minefield/3.0b2pre
Build Identifier: Mozilla/5.0 (Macintosh; U; Intel Mac OS X 10.4; en-US; rv:1.9b2pre) Gecko/2007111704 Minefield/3.0b2pre

SVGMatrix from SVG, http://www.w3.org/TR/SVG11/coords.html#InterfaceSVGMatrix , has a multiply method.

A.multiply(B) is supposed to computer A _post_-multiplied by B.  In other words it is supposed to compute BA in conventional mathematical notation.  Instead AB is computed.

Reproducible: Always

Steps to Reproduce:
Here's some JavaScript that detects the bug:

  // truthy iff SVGMatrix.multiply is wrong (swopped).  For example in
  // Firefox.
  matmulwrong = function() {
    var g = document.createElementNS('http://www.w3.org/2000/svg', 'g')
    g.setAttribute('transform', 'matrix(0 1 1 0 0 0) translate(1 0)')
    var l = g.transform.baseVal.getItem(0).matrix
    var r = g.transform.baseVal.getItem(1).matrix
    return r.multiply(l).e
  }

Actual Results:  
matmulwrong() returns 1

Expected Results:  
matmulwrong() should return 0

Safari gets it right, Opera gets it wrong.

Comment 1

[Robert Longson [:longsonr]]

Various searching suggests that AB means is B pre-multiplied by A (or A is post-multiplied by B).

e.g. http://mrl.nyu.edu/~perlin/courses/fall2003/hw/4.html or http://mrw.interscience.wiley.com/emrw/9780470011812/eob/article/b2a13039/current/html#b2a13039-sec1-0002

Suggesting to me that the gecko implementation is correct.

Comment 2

[David Jones]

Not only do those comments suggest that, so does my friendly Cambridge Maths PhD.

I defer and agree. Close the bug.  I'll be opening one with Safari / WebKit.

Sorry.