W3C home > Mailing lists > Public > public-respimg@w3.org > October 2012

Re: WebP, anyone using it?

From: David Demaree <ddemaree@adobe.com>
Date: Tue, 16 Oct 2012 08:11:09 -0700
To: Marcos Caceres <w3c@marcosc.com>
CC: Matthew Wilcox <mail@matthewwilcox.com>, Tom Lane <tom@tomlane.me>, Peter Gasston <pgasston@gmail.com>, David Newton <david@davidnewton.ca>, François REMY <fremycompany_pub@yahoo.fr>, "public-respimg@w3.org" <public-respimg@w3.org>
Message-ID: <6BE1D2AF-ABE3-4D00-8349-982A141BA906@adobe.com>
I'm coming in late to this discussion and am new to the group, so forgive me if I'm off base. In my opinion, while it's silly to try to solve specific problems that don't exist yet, if there could be an elegant way for <picture> to allow developers to choose newer image formats (raster or vector) without sacrificing backward compatibility, that can only be a win. I'd love to experiment with WebP, or to have a more reliable way to start using SVG without breaking IE 8/Android < 3.0 compatibility. A type attribute could be a good vehicle for this kind of forward-looking experimentation.

For me, the SVG use case is even more compelling than WebP. Many developers, myself included, are forced to move images that rightly should be part of their HTML content into CSS background images because CSS's cascading fallback support is robust, while HTML's is nonexistent.

WebP vs. PNG vs. Microsoft's JPEG-whatever is interesting, but ultimately could result in unnecessary fragmentation in an area of web technology that's been stable for years. (And I will allow, one fear I have about a type specifier is that then we'll all be expected to produce "bulletproof" responsive images markup. I certainly don't want images to turn into Web Fonts 2: Responsive Boogaloo.) But the ability to start using vector formats, or for vendors to introduce new ones without a huge adoption curve, could be transformative.


On Oct 16, 2012, at 9:46 AM, Marcos Caceres <w3c@marcosc.com> wrote:

> On Tuesday, October 16, 2012 at 3:34 PM, Matthew Wilcox wrote:
> 
>> The point being this: looking for evidence of WebP is not the right thing to look for. That's not the problem I want to address, because it's not the problem we have. WebP's poor support is a symptom of the problem, and the problem is that our image-format support is binary. Which creates an environment that's hostile to development of new, more efficient, better quality file formats. Now and ever after. Because they can never get a foot in the door.
> 
> I don't agree with this (again, citing PNG). Any browser has the ability to indicate to a web app that is supports WebP through the Accepts: HTTP header. It's possible to set up content negotiation on the server side to deal with this.  
> 
> Also, the  problem cannot be isolated or blamed on one single point of failure: in your case, you are blaming the markup. Even if all browsers supported WebP, there might not be any tools that allow the creation of content (think SVG's failure to take off because no one can author it). 
> 
> There are lots of things that can paralyze a standard from reaching critical mass - markup issues is just one tiny part of the problem. 
> 
> 
> -- 
> Marcos Caceres
> 
> 
> 
> 
Received on Tuesday, 16 October 2012 21:33:39 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 16 October 2012 21:33:48 GMT