W3C home > Mailing lists > Public > public-html@w3.org > February 2011

Re: An inherent flaw in <canvas>?

From: Maciej Stachowiak <mjs@apple.com>
Date: Mon, 31 Jan 2011 20:07:37 -0800
Cc: Aryeh Gregor <Simetrical+w3c@gmail.com>, "Tab Atkins Jr." <jackalmage@gmail.com>, Justin Anthony Knapp <justinkoavf@gmail.com>, public-html@w3.org
Message-id: <78161F6F-5288-4BB6-8EF3-C6BF4BCEB568@apple.com>
To: Boris Zbarsky <bzbarsky@mit.edu>

On Jan 31, 2011, at 7:44 PM, Boris Zbarsky wrote:

> On 1/31/11 7:45 PM, Aryeh Gregor wrote:
>> But it apparently doesn't exhibit the problem.  Surely there's some
>> trick you could use to avoid the problem a lot of the time in
>> practice.
> 
> There are various tricks.  For one, you could supersample, and then downsample when rendering, at some possibly noticeable memory and CPU cost.  These may both be ameliorated if the GPU can be used, of course.

My understanding of Flash (vague as it is) is that it creates a display list even when you make what appear to be immediate-mode calls, which in theory could allow seamless edges with AA and without supersampling. I'm not enough of an expert to say what it is doing for sure though.

Side note: turning off AA and supersampling would be something that Web apps could do for themselves if canvas had a way to disable AA.

 - Maciej
Received on Tuesday, 1 February 2011 04:09:02 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Wednesday, 9 May 2012 00:17:21 GMT