W3C home > Mailing lists > Public > www-style@w3.org > August 2014

Re: [css-transforms] suggestion: single transformation function by simply giving edge positions

From: Tab Atkins Jr. <jackalmage@gmail.com>
Date: Fri, 8 Aug 2014 13:34:38 -0700
Message-ID: <CAAWBYDBO227OP_T74bmkTeCR5A6U30+fegGgOeq4qEpqjUFkTA@mail.gmail.com>
To: Axel Dahmen <brille1@hotmail.com>
Cc: www-style list <www-style@w3.org>
On Fri, Aug 8, 2014 at 5:28 AM, Axel Dahmen <brille1@hotmail.com> wrote:
> Actually, in fact, my suggestion is not specific. On the contrary, it's the
> most generic form compared to all other transformation functions:

I *think* you're right that the other functions can be implemented in
terms of this, but that doesn't necessarily mean that it's possible to
implement.

Your desired transformation, for example, is definitely not affine, as
it can distort a square into a trapezoid (breaking the conservation of
parallel lines), so I'm now certain that it can't be reproduced with
the existing normal transformation functions.

However, it's probably possible to reproduce it with a matrix()
transform; the perspective() function is also not affine, after all,
and can be done with matrix transformations.

(Going further, as your movie suggested, by doing things like turning
each edge into a bezier curve, definitely isn't possible with matrix
transformations; you have to do per-point resampling, I think, which
is a far more expensive operation than applying matrix transforms.)

~TJ
Received on Friday, 8 August 2014 20:35:27 UTC

This archive was generated by hypermail 2.3.1 : Friday, 8 August 2014 20:35:28 UTC