The next dot-version of Blender, 2.8, is the most exciting update (in my opinion) since BMesh (n-gons) was added in 2.63. It is focused on workflow patterns and interface changes, but this involves several secondary changes. First, the minimum OpenGL version of the viewport renderer has been bumped to 3.3, which targets DirectX 10 hardware. (Previously, Blender required OpenGL 2.1.) The method for storing objects in the scene will also change. In short, it’s a huge update, both visibly and behind-the-scenes.
Here's Blender Today doing a live stream about Blender 2.8 Alpha
And now the Blender Foundation has published its expected release schedule.
First up is the alpha release – Today!
Blender 2.8’s alpha is still not useful for production content, and it’s even too early for add-on developers to update (because the Python API will change). After playing around with it for a few minutes, I’m admittedly confused with some parts. For instance, is OpenGL Render Image / Animation going to stick around? If so, what does it mean compared to clicking Render Image when Eevee is selected. For the moment, Blender’s bug tracker is only interested in crash bugs. You cannot report anything other than crashes.
After the alpha will be the beta – around August 12th.
Blender 2.8’s beta will still not be ready for add-on developers to update their plug-ins. Instead, this will align with merging the 2.8 branch into master, updating the release notes, and opening up the bug tracker to non-crash-related reports. All features are expected to be finished by then, and development will work on fixing bugs.
After the beta will be a Python API finalized – around September 19th (give or take).
When this date occurs, Blender will communicate to add-on developers that it is safe to update their plug-ins to Blender 2.8. It looks like this is to give them a little time to focus on API-breaking bugs that appear in the beta phase, fixing them first so add-on developers can get to work while the Blender Foundation focuses on bugs that do not affect the API. They will give add-on developers at least a month before RC is called.
Which brings us to release candidate – around October 20th.
The Blender Conference 2018 event will take place from the 25th, 26th, and 27th of October, so I’m guessing that they want to get the release candidate out in time for then. This should hopefully mean that the whole update arrives before the end of 2018.
As always, Blender completely free, even for commercial use, as licensed under the GPL.
The videos of the EEVEE
The videos of the EEVEE viewport rendering looks a lot better than the old versions of Blender’s OpenGL Render Image/Animation that was very feature limited compared to what can be done in Blender 2.80.
I’m very excited because I’m on older AMD GPU hardware(Radeon Discrete mobile 7650M/Terascale Rebrand) on my laptop that does not have the OpenCL support for GPU Cycles rendering(Cycles on AMD GPUs requires GCN 2.0/1.2 or later GPUs). The 7650m on my laptop supports OpenGL 4.4 and the Ivy Bridge CPU/Integrated Intel Graphics supports at least OpenGL 4.0.
EEVEE in the viewport rendering and that more complete set of features sounds great.
It looks like OpenGL 3.3 support for Sandy Bridge processors is not there for windows(OpenGL 3.1) according to wikipedia Intel’s Integrated Graphics listing but MacOS/OSX supports OpenGL 3.3 for any Sandy Bridge Based Appple products. Linux only goes to 3.1 for Sandy Bridge according to wikipedia but that may be outdated information.
Sandy Bridge on Linux is
Sandy Bridge on Linux is OpenGL 3.3, with some more advanced extensions also implemented.
In general Linux have better OpenGL quality drivers on Linux than MacOS, and better than Windows on Intel and AMD cases.
Indeed this is one good step
Indeed this is one good step for Blender, and I’m quite excited by it.
Using it for few weeks now, and even though unstable (however for what I used it so far no crashes) it is so.. different. Learning the new keymap is going to take a while especially to overwrite my muscle memory.
Still Eevee, full code re-write. The overall interface optimizations. Can’t wait to switch to it full time once it becomes much more stable.