He can try until the lights go out, I can build any version of Blender that I want. :evilgrin:
:eek: You are a lot more brave than I could ever be. I would never mess with the russian guy.
Just commit a whole lot of code that uses spaces instead of tabs, thatāll keep him distracted for a while.
I get Sergeyās point but thereās already stuff like Animation Nodes and Sverchok where you can make things that are kind of like modifiers but still fast enough. Any kind of modifier creation system within Blender (maybe nodes) would be nice to have. Same thing for addons/scripts really.
I found out that the level of distraction increases proportionally with the different amounts of spaces being used for the indentation.
The new dependency graph is made in a way which allows highly parallel evaluations in Blender. Right now, if it was possible to have Python modifiers, you would throw a lot of that away even if there is just a single modifier that is being used. This would certainly not scale well if it was used extensively in the scene. In general, this is clearly not a good idea.
This is not related to nodes. It is possible to implement modifier nodes in Blender which work nicely with the dependency graph to allow parallel evaluations. Only if the computations within a node were handled in Python, you would castrate the parallelism.
Best comment of the thread.
At any rate Iād have to say compared to other DCC applications out there Blender does lack in 2 areas: 1. Studio adoption and contribution of feature drive or commital. 2. Support from other commercial apps/plugins.
Iām not sure 100% of why that is, but I do come to the conclusion that those things are somewhat important to the quality and sustainability of the project. And that some % of the reason behind them has to be the licensing issues. Studios and large companies are very wary of adapting software to the GPL (industry insider knowledge here).
I get the point about wanting to keep everything free and the point from Ton about licensing your pencil is well taken. But What if App X comes out next year that is free to hobbbyists and is more attractive for people who want to sell their plugins licensing wise? Suddenly App X is a competitor and has Arnold, realflow, shave and a haircut etc plugins. Or maybe the conclusion here is that there is no real competitor to blender to incentive the move.
This point is continuously being improved. The number of commits from developers who are working for studios is clearly increasing and include some quite incredible contributions.
Absolutely. But Blender has been around for 20+ years, why only now?
Because it is established in the industry and there is enough trust in Blender that complete studios with lots of employees are using it as their primary tool. There have been many small studios using it. Usually, they are not in a position to hire developers and contribute to Blender. Nevertheless, those studios were able to show, that Blender is a serious solution.
Proprietary solutions are quite often pushed with expensive marketing campaigns in order to being recognized and seen as a serious competitor. Something that was not possible for Blender for obvious reasons.
And the BF is indeed going in the direction of bringing in features that professionals have been wanting or needing for years (and itās seeing results without the need to re-license or turn Blender into just another hugely expensive 3D app. if you want everything).
I donāt know if itās the language barrier, but everything mentioned here about the efforts of the BF (to increase their resources and number of developers), does not seem to be registering at all (the only response is the same narrative about the almost magical effects of removing the GPL from Blender). Do you now understand why I posted a link to the Modo store?
As you pointed out, Blender has been around for over 20 years. Each year, there seems to be an increased awareness and recognition of what Blenderās capabilities are. The ecosystem of companies (including studios) that use Blender as part (or all) of their business continues to grow. Why would we be doubting how sustainable the project is at this point? Itās not like the first release was yesterday with only one part-time developer.
I think he asked why studios uses blender only now and not before, 20 years itās a loooong time.
If blender was more open, everyone will use it for personnal and professionnal projects.
More than a language barrier I see a parallel universe barrierā¦ what features are BF adding that professionals have been wanting? in blender 2.8 the only big features added by BF directly is eevee and UDIM. But people have asked a lot of years to fracture, fluids, hair, particles, more modifiersā¦ features that max and maya have since Āæten years? and without take account of other things like sculpting, meshfusion, better UVs and pack systemā¦
THe problem with efforts of BF is that this efforts are for BF targets. I meantā¦ For example, a modeling artist donāt see a lot of improvements in the last years, but like blender have a really powerfull modeling tools before, is not a big problem. But if a studio want some of this features, they needs to buy a propietary software and donāt use blender, they automatically use maya or modoā¦ I think that itās better to pay for it in blender or to a company that make it in blender.
The default Python interpreter is not the best choice for this application, indeed. One could easily imagine a different language interpreter being used here (IMHO for applications like this, functional languages are a natural fit), or even leveraging LLVM for JIT compilation. Then the scripting language could even be a subset of Python, executed at native speed.
its only work, all we need is someone to do itā¦
That would at least be a good reason to switch to a different programming language for the scripting in Blender. However, I donāt think it justifies that all the existing scripts would need to be ported. Of course, you could have more than one supported language, but for that you would also need someone to maintain it.
Overall, I donāt think it is worth the effort.
- Object collections (which essentially allows unlimited layers and nested ones at that), limited layering abilities has been a common complaint for years
- Fast and modern viewport (not just Eevee), lackluster performance has been another major complaint
- A completely new hair and fur system in progress from Lukas Tonne (which will address weaknesses in that area)
- ID overrides (the lack of was a major limitation and point of soreness among professionals)
- Workflow improvements such as operator widgets (professionals have long been wanting to see improvements in that area as well).
- A new depsgraph that allows for powerful animation setups that were not possible before
- (Possible); Scorpionās fracture modifier work as part of the planned overhaul of the physics system (bringing much wanted improvements to the area of VFX).
- Improved asset management (was also considered a sore point among pros)
- Other things I may have missed
In short, a lot of things professionals have wanted to see is either in Blender 2.8 now, in progress, or planned (but postponed due to lack of resources and due to needing to make 2.8 more manageable).
I think that is better donāt reply you because you are not able to see more than points in a list.
I really donāt see a point in discussing things further then, I pointed out the progress being made in many areas and it just gets dismissed.
Thereās no point is discussing further. If the GPL even being there makes everything totally moot as-per development progress and the like, purchase a copy of Modo then and stop bothering with Blender.
Like I told in the first messages, if you donāt want to read the thread, donāt enter. You only enter to try to stop any other message diferent to your opinion. I didnāt see any problem with devs messages.