[Bf-viewport] OpenGL ES compatibility

[Antony Riakiotakis]

Hi,

For me, supporting 2.1 was more to ensure that we had broader hardware
support for blender. My personal aspiration was to jump immediately to 3.3
for blender2.8. The plan was to do this incrementally so the community
could adapt to the change. 3.1 sounds OKish, but again, we're leaving some
people behind. Whether we do it or not is more a political decision, the
technical benefits are quite clear.

We could do the more radical forward looking changes in a branch, but on
the other hand we've seen what tends to happen to big branches like that
and we decided to better work directly in master. If others also agree too
then I would agree to raise the bar to 3.1.

On 6 December 2015 at 15:55, Martijn Berger <martijn.berger at gmail.com>
wrote:

>
>
> On Sun, Dec 6, 2015 at 3:31 PM, Brecht Van Lommel <
> brechtvanlommel at pandora.be> wrote:
>
>> Making a separate thread for this discussion..
>>
>> * OpenGL ES 2.0 I can kind of understand, but why OpenGL 2.1
>> compatibility?
>>
> Personally i think 2.1 could go and we should take sandy bridge opengl
> 3.1-ish as lowest common denominator for desktop but I do think we should
> decide explicitly.
>
>> * Why ANGLE? As far as I know this is for old cards that we already
>> decided to stop supporting.
>>
> ANGLE brings us a reasonable fallback on windows systems. It is the
> default backend for WebGL  for chrome  and firefox so it could bring us a
> more stable then vendor drivers fallback if things look wrong.
> Also it gives GLES 2.0 when no graphics drivers ( or remote sessions are
> run )
>
>> * Why not require OpenGL ES 3.0 as a minimum when someone decides to port
>> Blender to Android / iOS? Support for this seems pretty good.
>>
> I could live with that as a binding minimum also, I would think 2.0 might
> be to restrictive to be worth it. Newer devices support 3.0 or 3.1 and
> better anway.
>
>> * WebGL 2 will match ES 3.0. If no one is actively working on WebGL
>> support now, why bother with the backwards compatibility?
>>
> I agree.
>
>> I really think we should be able to assume support for FBOs, VAOs, float
>> / half-float textures, instanced draw, and not have to deal with GLSL
>> syntax differences between versions.
>>
> I think this is a good list of things that we really do want to have.
>
>> How is Blend4Web affected by Blender's minimum OpenGL version? What kind
>> of issues would there be if Blender assumed ES 3.0 minimum, while Blend4Web
>> still targets ES 2.0?
>> On Dec 6, 2015 12:54, "Martijn Berger" <martijn.berger at gmail.com> wrote:
>>
>>> Hi Mike, Brecht, Jason, Anthony and others
>>>
>>> On Sun, Dec 6, 2015 at 12:34 AM, Mike Erwin <significant.bit at gmail.com>
>>> wrote:
>>>
>>>> Re: Uniform Buffer Objects -- Compatibility with GL 2.1 is a temporary
>>>> concern, so anything we do there should be minimal and marked for future
>>>> removal.
>>>>
>>>> Is this the case? i would make the argument that we might want to try
>>> and keep OpenGL ES 2.0 and OpenGL 2.1 for as long as possible.
>>> Especially on windows keeping ES 2.0 + what ANGLE exposes could really
>>> help us have a good fallback.
>>> On android and IOS GL ES 2.0 / 3.0 could allow us to run blender there.
>>> And even compile blender for Google's NaCl stuf..
>>>
>>> We should not do this regardless of the cost but we should not discard
>>> this without seriously considering if this is feasible.
>>>
>>>
>>> _______________________________________________
>>> Bf-viewport mailing list
>>> Bf-viewport at blender.org
>>> http://lists.blender.org/mailman/listinfo/bf-viewport
>>>
>>>
>> _______________________________________________
>> Bf-viewport mailing list
>> Bf-viewport at blender.org
>> http://lists.blender.org/mailman/listinfo/bf-viewport
>>
>>
>
> _______________________________________________
> Bf-viewport mailing list
> Bf-viewport at blender.org
> http://lists.blender.org/mailman/listinfo/bf-viewport
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.blender.org/pipermail/bf-viewport/attachments/20151206/f41ca79d/attachment-0001.htm