Re: [geometry] Discrepancy in calculations, Firefox vs my JS implementation. from Simon Pieters on 2016-01-04 (public-fx@w3.org from January to March 2016)

From: Simon Pieters <simonp@opera.com>
Date: Mon, 04 Jan 2016 09:02:42 +0100
To: public-fx@w3.org, /#!/JoePea <trusktr@gmail.com>
Message-ID: <op.yapfaslfidj3kv@simons-mbp>

On Sat, 02 Jan 2016 07:18:21 +0100, /#!/JoePea <trusktr@gmail.com> wrote:

> I'm working on polyfills for the Geomtry Interfaces at
> github.com/trusktr/dommetry.
>
> In my implementation, the `matrix.rotate(45).a` shows the same result as
> the Firefox implementation in the console: `0.7071067811865475`. However,
> the result of `matrix.rotateSelf(45).a` differs between my implementation
> and Firefox. Mine shows `0.7071067811865475` and Firefox shows `
> 0.7071067690849304`.
>
> Any idea why this is? Is this expected? Or might there be a bug  
> somewhere?

I suppose Firefox does an extra cast to double precision in the rotateSelf  
case. I don't see that there should be any difference per spec. Can you  
file a Gecko bug?

-- 
Simon Pieters
Opera Software

Received on Monday, 4 January 2016 08:03:14 UTC