[vbox-dev] OpenGL 3.2 support

[Klaus Espenlaub]

Hi Socratis,

my preference would be keeping this as low key (and limited audience) as 
possible until we have more accurate planning data - currently there 
isn't a line of new code in the tree (i.e. we've not reached the "proof 
of concept" or "prototype" stages), and it's totally unknown how long 
it'll take to get the code working. Even if I don't make the most 
pessimistic assumptions this could take until 2018 to go beyond the 
break-even point compared to the existing 3D code. Don't want to create 
false hope at this early stage which leads to more frustration.

We're still doing small improvements to the existing 3D code if there's 
genuine customer need, but that's exclusively bug fixing, not adding any 
new features.

Klaus

On 18.01.2017 08:46, Kalogrianitis Socratis wrote:
>
> On 16/Ιαν/2017, at 17:32, Klaus Espenlaub <klaus.espenlaub at oracle.com> wrote:
>
>> Since we're already planning to completely throw out Chromium (since it's effectively dead for a long time) and redoing all our 3D support based on Gallium (part of the Mesa project) which should give us OpenGL 4.x coverage I don't see a good reason for teaching the really old dog some new tricks...
>>
>> Switching to Gallium/Mesa based code would also make it far easier to improve the Direct3D coverage for Windows VMs, where with the old style 3D we would first need to improve Wine's D3D code, usually pushing Chromium beyond its comfort zone, needing more work.
>>
>> Regards,
>> Klaus
>
> Hey Klaus,
>
> Would it be OK if I use the above quotes (anonymously and slightly paraphrased) in the forums, if there's a need to? Of course with no time lines or promises. Or do you think that if I reply to someone with "VBox devs are evaluating phasing out the old 3D code" might open Pandora's box?
>
> TIA,
>
> Socratis