[Public WebGL] Re: retrograde webgl support levels

Kenneth Russell [email protected]
Wed Nov 5 23:59:35 PST 2014


Thanks Florian for continuing to analyze and publish these statistics. When
mistakes happen like Chrome's accidental blacklisting of certain older
hardware, it's very helpful to both see the effect of the bug as well as
the fix.

-Ken


On Tue, Nov 4, 2014 at 12:56 AM, Florian Bösch <[email protected]> wrote:

> The first few days of November stats are in (Saturday, Sunday and Monday).
> It's looking very good overall, with *webgl support standing at 80.4%*.
>
> Congratulations to the Microsoft IE team, WebGL support on desktops is up
> 8% from last month to *80%* this month, (at least with the limited
> demographic I see). This beats Firefox down 5% from last month to *76% *this
> month.
>
> OSX Safari which has been a troublesome spot for WebGL is also making good
> progress, up 16% from last month to *30%* this month.
>
> Chrome has fixed their blacklist issue and is in the process of recovering
> previously lost WebGL support level and with *92%* this month they're
> standing near their average support level again.
>
> Several extensions have seen implementations come online such as:
>
>    - *EXT_shader_texture_lod*: Chrome and Opera on desktops, not
>    supported by IE and firefox
>    - *EXT_frag_depth*: Chrome, Firefox and Opera on desktops, not
>    supported by IE and Safari
>    - *EXT_blend_minmax: *Chrome, Firefox and Opera on desktops, not
>    supported by IE and Safari
>
> Some points of concern:
>
>    - Browser fragmentation on Android is an issue with many of the custom
>    browsers rarely updating their engine, and replacement of the built-in
>    android browser with chrome mobile is slow.
>    - Neither Safari nor Firefox implement WEBGL_debug_renderer_info. It's
>    commendable that IE and Chrome do implement it, thank you guys!
>    - EXT_disjoint_timer_query isn't implemented so far. This would be a
>    useful tool among other things, to improve the conformance tests.
>    - IE mobile and Android browser (built in) doesn't implement
>    OES_standard_derivatives (nearly universally implemented by any other
>    platform).
>
> I've also added recording of WebGL2 and its parameters to
> http://webglstats.com/ . I'm looking forward to integrate this data on
> the site once implementations come online, and explore the ranges of the 26
> new parameters that'll come with WebGL 2.0.
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://khronos.org/pipermail/public_webgl_khronos.org/attachments/20141105/77473f7e/attachment.html>


More information about the public_webgl mailing list