Faster and more flexible - OpenSubdiv 3.0 is here!
OpenSubdiv 3.0 represents a landmark release, with profound changes to the core algorithms, simplified APIs, and streamlined GPU execution. Providing faster, more efficient, and more flexible subdivision code remains our principal goal. To achieve this, OpenSubdiv 3.0 introduces many improvements that constitute a fairly radical departure from previous versions.
I checked with developer Lukas Toenne if this means 3.0 will be part of Blender 2.76 (it was mentioned in his week's developer meeting notes). He says:
I only updated Sergey's branch recently and looked into the API, but i'm probably not going to be working on it. I don't know about concrete plans. Switching to 3.0 is on the agenda, but that's about all I know.
So, it looks like we'll have to wait a little longer.
17 Comments
This is eternal...
Can someone explain the differences? and what the best part of this will be?
Lol what a bunch of excuses...
Just implement that damn thing and put someone competent to do the job. Is something stopping to release the older version?
Otherwise when you are done adjusting to 3.0 version a 4.0 version will be out..
Nothing at all - you're welcome to help and do it yourself! Of course, if you don't know how, then maybe you also shouldn't suggest it's an easy thing to implement.
I encounter this "if you don't like it do it yourself" arrogant argument with open-source quite often. It's totally out of line even if the software comes for free and without any warranty:
The foundation gets money from community, if the software stinks, community will give less money. On the other hand if good actually useful features are being implemented and we see capable people rather doing things instead of postponing forever more funds will come. Also features can be funded beforehand from community and devs hired to do it - happened before nothing new.
So the foundation should listen for negative feedback and do what the community wants - not bash negative feedback with "do it yourself" - that's just douchbaggy. This is if the foundation cares about Blender..but lately it looks like it cares more about movies..
Now about the porting itself - I know what I am talking about. You would be amazed what a single person can implement in matter of weeks or months when being properly paid. I am quite confident I could do it myself and there is even a guide and good documentation how everything works:
http://graphics.pixar.com/opensubdiv/docs/porting.html
So why I won't do it? Because I already have an 8hour/day job and I want to have some life left. Otherwise believe me I wouldn't mind fork blender, put opensubdiv in and let it be merged into trunk - but all for money, no one works for free. I am sure feature like this would be funded from the community.
Yeah I guess it is a bit of a lame thing to comment, you're right. I guess I knee-jerked when I read "Just implement that damn thing" - not exactly subtle either ;-)
Anyway, mea culpa!
It's important to remember Blender is made by small team of dedicated developers and volunteers. It is not a monolithic mega-corporation like the commercial 3D software. The resources available are limited. I am continually amazed at what they manage to achieve! If you feel really strongly about it, why don't you organize a funding campaign or find some other way to support Blender and the blender foundation. Make a monthly donation? Join the cloud, buy a training dvd.
There are resources (thats money) but they are put in wrong direction - in making a movie instead of putting them to hire devs to implement features into blender.
I would make a bigger monthly donation or join the cloud if I knew what the money would be spend on!
Organizing a funding campaign(s) or changing leadership is very time consuming. That's not a job for me. Rather I can spend little time writing this comment so other people will realize they are being sheeps waiting again and again for blender to stop being mediocre. They are only making or listening to excuses (made by small team, open-source, change in API, blah-blah). If you don't change something you cannot expect different results. The leadership and where the money is spend needs to change.
It doesn't have to be me, I can only plant the idea.
There is enough money in this big community and enough freelance devs. Only a bad leadership that doesn't exploit this. Look what people are buying on blendermarket for how much dollars..and that even doesn't compare by far to opensubdiv.
Tell me I am wrong..
Unpleasant, but true. I agree with this 100%. I would even say that I feel kind of disrespected by Blender Foundation, they hear only what their local animators/designers say about features and other developments needed, although they already have so many professional users all over the world. Movie is a good thing, but don't tell me that it's the only way to find what developments are needed. Just ask the community! And I totally understand that they want to be sure they ask a professional, not just a fanboy, but it's really easy to check, you know.. portfolio says all. And I don't even want to start thinking about what could have been done with that money they invested in movies.. oh my, oh, my. And those movies are not even good, yes, they are a good technical achievment presentation, but stories and direction sucks.. I am not surprised at all they didn't get funding from State this year.
I really like the Krita approach to manage the funding of the features
https://krita.org/item/penultimate-kickstarter-voting-results-2/
Do you think that something like this can be helpful to know better the user's opinion?
Greetings
Though once they are making movies they expand blender in many ways. Gooseberry branch has done so much for blender. Yea, not everything was on one subject and maybe a little limiting. Another thing is people around the world help are making these adjustments and not just the core blender crew. Magnificent people will come to help in code. Sounds far fetched but it happens with every single thing know to man. Just because it sounds simple doesn't mean it is.
Also if they started hiring people of that stature why keep it open source and become just another renderman, autodesk, or paid software. Blender is quite powerful as it is and can be the leading 'top 5' best designing software easily. So, you would see it that it will loose it's open-source claim?
Making a movie is NOT a waste of resources. First, when you make a movie you find out first hand ( the movie maker and the programmers are in the same room) what needs to be done and why. It brings the users (top ones at that) together with the devs and lets them work together to improve the product. Second, a movie give Blender a HUGE marketing boost! This leads to more pros and newbies using it and this increases the budget that lets Blender hire more devs and artists to help develop the program.
I would gladly contribute to a funding campaign purely to pay somebody to implement OpenSubDiv in Blender, as I imagine would many others. Is there an existing method to propose such a course to the foundation?
Unfortunately there is not.
I just put this idea in the table
"I really like the Krita approach to manage the funding of the features
https://krita.org/item/penultimate-kickstarter-voting-results-2/
Do you think that something like this can be helpful to know better the user's opinion?"
Greetings
I like that approach very much. It plays well with the community aspect of open source as well as directed funding. We need to make THIS happen for Blender.
I agree there should be a way to fund crucial features like alembic support or opensubdiv implementation. Nobody works for free and i know that a great amount of people would gladly invest in the blender development.