W3C home > Mailing lists > Public > public-fx@w3.org > October to December 2012

Re: [css-compositing] blending in canvas

From: Tab Atkins Jr. <jackalmage@gmail.com>
Date: Thu, 15 Nov 2012 14:46:29 -0800
Message-ID: <CAAWBYDDRS6jbuVOiWFGPAALUDvLc_akmFApzt9hx3MX71FMGQQ@mail.gmail.com>
To: Rik Cabanier <cabanier@gmail.com>
Cc: "public-fx@w3.org" <public-fx@w3.org>
On Thu, Nov 15, 2012 at 2:36 PM, Rik Cabanier <cabanier@gmail.com> wrote:
> Maciej still objects because he feels that this is a substantial difference
> between CSS and Canvas.
> I think we have a couple of choices:
> 1. Keep globalCompositeOperator for blending and compositing but collapses
> the 2 CSS properties into one that takes the same arguments as
> globalCompositeOperator
> 2. Keep the 2 CSS properties but split the Canvas properties into
> globalCompositeOperator and globalBlendOperator
> 3. Don't change anything and live with them being different.
> 4. Don't change anything but also define a new CSS shorthand that combines
> blending and compositing. Canvas is compatible with this shorthand.
> I'm unsure what approach we should take.
> option 2 has the issue that we can't implement this correctly in the near
> term.
> option 1 has the issue that transitionable blending will be more confusing
> in the future.
> option 4 should cover all concerns but introduces yet another keyword.

I like option 4, actually.  The existing blending/compositing
properties have pretty long names (and the names use inconsistent
conjugation, which is a bug that should be fixed).  They'll usually be
set together, but I can see use-cases for animating them separately.
This suggests a shorthand property.

The email Dirk just sent, where he proposes an at-rule for defining
custom filters, uses the name "mix" for a descriptor that sets both
blending and compositing at the same time.  I don't know if it's too
short, but I like it.

Received on Thursday, 15 November 2012 22:47:16 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 19:49:43 UTC