Description
This is loosely related to Qt RHI, in the extended sense that it is about stream-lining the graphics foundations in Qt 6.
On Windows, you currently have the possibility of running OpenGL directly on an OpenGL driver, ANGLE or a software implementation that we ship with Qt.
In Qt 6 we are removing ANGLE, since Qt RHI makes it possible to run Qt itself on systems that do not have an OpenGL driver by going directly to Direct 3D. The users who depend on OpenGL in their own applications and have to run on these systems, can still include ANGLE and run Qt on top of this with the OpenGL backend to RHI.
Would it make sense to take the same approach for the software implementation? It is unsure how much overhead the updates and building of this actually is, but if it is significant, the users who need to run on systems that are not supported by RHI and which are not already covered by the software backend in Qt Quick (which should be very rare with Qt 6?) could still ship this with their apps.
Note that the task here is just to talk to the people who might now the impact of this and make a decision. If the decision is to remove it, make a follow-up task for this.