The issue is that the patches can only be thoroughly reviewed by devs. who know the code they touch.
However, the BF could have people whose job is to check consistency in code style and in variable names, test them to make sure they work as advertised, and prioritize them to determine what gets a more complete review first.
With that said, I wait to see what is going to come out of the next sculpt and paint meeting.
All of Blenderās problems can be solved with more money. I donāt know recent numbers but I think every release in 2020 and 2021 had about 3.5 million downloads. If we could get everyone who really uses Blender more than a few times a week to make an annual one time donation of one dollar, that would be somewhere between 1 million and 3 million dollars. --edit lol where did i get those numbers from ā
We need some kind of viral marketing strategy, Fund Blender Day, Blender Birthday, Blender December, Blendcembler, Decemblerā¦ something like that. If we could make it part of the Blender Community Culture to pick 1 day a year where as many people as possible donate just one dollar, we could get enough money together to get a dedicated assistant/apprentice for all of the core devs. This would allow the devs to focus on new development while having someone who knows their thought process regarding accepting patches to streamline the patch review process before getting the core dev involved.
We damn need that and also money to train future wannabe blender devs to learn the blender code so that they can ease burden in the development of complex code and features, for example the patch reviewal which seems that only the head devs only know better due to more time and experience knowing itā¦
It looks like it based on what Ton said (at least to the point of getting the refactor committed). This also means a decent shot of getting more buildbot builds soon. This is good news for that specific part of the sculpt core rework anyway, the plan for Multires meanwhile may depend on what happens after the final commit.
EDIT: This is rough, and not really for high resolution artwork or otherwise it would be much higher res and much more detailed, but this is a small sampling of what I dream of will be possible with the refactoring work.
Vertex painted and vertex weighted for animation with no need of the retopology step. The performance is already good enough because there is no subsurf. The Corrective Smooth modifier ameliorates any concern with the topology when moving bones. Sculpting can continue if needed because attributes are now propagated and preserved.
I think that companies should step up whose employees use Blenser and where managemwnt says, ānice, free softwareā as well as seek to obtain governmental funding. I mean FOSS is also very much a way to ensure open data and can be used within software development education by sending in patches for real worls projects.
Not sure this was suggested here or elsewhere but I couldnāt find a task or suggestion so. An idea of supporting manual value typing and detail sampler during interactive voxel size adjustment.
But man, this interactive voxel size operator should set the voxel size AND run the remesh operation automatically when you click to confirmā¦ the fact that I still need to hit ctrl+r to remesh after that is a no goā¦ thatās why I donāt use itā¦
You refuse to use something because it takes one more button press?
I donāt want it to automatically remesh after setting the size, because sometimes my hand slips and I set it to 0.0001 by accident, which makes Blender crash and I lose all my work.
Iām not a hotkey junkieā¦ I avoid them whenever possibleā¦
Yeah, probably itās not for everyone, but I know a few people who expected that same behavior as wellā¦ it would be faster and more to the point, imoā¦
But of course, it should be optional, preferably a checkbox in the operatorsās keymap settings in the preferencesā¦ or done via a modifier key like alt+click for exampleā¦
Perhaps all CPU intensive operations must run inside a thread and the user screen be presented with a lock-screen modal message box. Thus the CPU intensive operation runs in the background and user might press ESC at anytime to cancel.
These techniques are fundamental to software development now, unfortunately there wasnāt there such a strategy in Blender at that time, so it would be integrated in the architecture. Now we are kinda stuck with the obsolete paradigm (single-threaded-execution) and things as such happen.
Can you recommend some reading or videos about SDF modeling? I donāt think I have full grasp on them. Iām familiar with Voxel Modeling, is that the same? What are itās usecases, etc.