Blender Feedback Survey Results 2024

[LudvikKoutny]

Creating a truly transparent and effective ‘vote with your money’ system can be challenging for developers and for the product vision. It is also clear that creating campaigns with clear messages yields more financial support from the community. We are looking into this topic.

This feels kind of good and bad at the same time. Good that it’s finally acknowledged many users simply want to vote with their wallet. Bad that the team’s interpretation (at least partially) is “People want more campaigns”

No we don’t. The only reason campaigns work better is that it’s just marginally better than throwing money at the wall and see what sticks, but it’s still not voting with the wallet.

But what I disagree most is that “can be challenging for developers and for the product vision”. No one has said that the feature/improvement bounties users can put money towards would not have to align with the team’s vision.

There are quite a few features and improvements team members have proposed themselves I’d start throwing money at in a second. The prime example being Brecht’s past proposal for the new texturing (and possibly also baking) system which would allow us to treat PBR attributes as single node connection (structures) and blend/mix them same way we can blend and mix BSDFs but without the limitations and performance BSDF mixing has. This would bring Blender on another level in terms of shading and texturing capabilities, but it never happened.

It would be totally fine if only the development team could choose proposals that can have bounties of them, and users would use their money only to vote for the priority of these features.

For example if you want to make $10k in 2 days, just put a progress bar going from $0 to $10k with a big “Pay” button next to it under “Bevel node for Geometry Nodes”

I guess my overall point is that there are quite a few improvement and feature proposals from the core dev team members themselves that got stuck in a limbo but many people would be more than willing to put money towards getting them unstuck.

[Howard_Trickey]

@LudvikKoutny

For example if you want to make $10k in 2 days, just put a progress bar going from $0 to $10k with a big “Pay” button next to it under “Bevel node for Geometry Nodes”

I know this was just an example, but I wanted to point out that this particular example would do nothing to speed up the development of that. I’m working on this as fast as I can, and it is really a one-person job, so throwing money at it won’t help.

[LudvikKoutny]

Could be, but 3 years for a single node seems to be a bit excessive in general, especially in the context how advanced GN is these days yet it still missing one of the few most essential polygonal modeling operators. So it could perhaps spawn interest from someone else to create a different branch and try different approach. Maybe even porting the current Bevel modifier code as a node without any improvements.

It would likely not be the perfect Bevel node you are seeking to build, but we’d no longer have this frustrating situation where one has to split GN setup between multiple GN node trees just to slap old Bevel modifier in between them in the old modifier stack.

I don’t want to divert this discussion offtopic, but that was one of my points of wallet voting. It could attract other developers to try approaches alternative to those that ended up in a limbo.

[thorn-neverwake]

Creating a truly transparent and effective ‘vote with your money’ system can be challenging for developers and for the product vision. It is also clear that creating campaigns with clear messages yields more financial support from the community. We are looking into this topic.

I can agree that having a “feature specific” funding campaign can create a challenge - however. In my NPO organization, a lot of our projects are funded with grants. None of them are open-ended with a scope of “accomplish anything”. We have a list of goals, deliverables, time tables, etc.

I don’t believe anyone expects Blender to move to a fully-targeted type of donation system similar to a standard GoFund campaign. But, i think the success of fundraising on sites such as that - prove that people are willing to participate more in donations, when they see a clear goal that they’re going to support. So I think it’s worth more consideration on doing this.

IE, “Please help us raise money for new fire equipment” is more appealing than “Please send money, we’ll do whatever we want with it.”

[sentharn]

Echoing what others have said, but I had no idea there was a formal survey, and neither did several artists I checked with. These are people who make their living with Blender and probably would have been good data points to have. (It would have likely been more dissatisfaction with Eevee-next for the most part, as well as hair/hair physics/development direction complaints.) Hopefully 2025 is more broadly advertised.

As far as I’m aware, none of them donate to blender. Personally, I don’t either and will not until there’s some assurance that features I need are on the roadmap and won’t languish for another 5+ years. I would absolutely donate healthily towards specific features if the option was available.

[makizar]

@LudvikKoutny

This feels kind of good and bad at the same time. Good that it’s finally acknowledged many users simply want to vote with their wallet. Bad that the team’s interpretation (at least partially) is “People want more campaigns”

I do think more focused donations could be good, but doing a bounty system is a terrible idea

It’s a door that is easy to open and hard to close, if it happens once users will expect it to continue and get frustrated if it doesn’t.
Doing funding through as an accumulation of small feature requests would inevitably produce bloat, incoherent architecture and degraded performance. Developpers are already stretched thin enough as it is, they shouldn’t carry the mental burden of delivering some random features to secure their funding.
Its a well known adage that users don’t know what they want. Making decisions on where to focus development efforts is best left to the Blender team.

That being said, I agree focused campaigns would be a great way to increase donations. People donate because it makes them feel like they have an impact, and with the huge amount of ground Blender covers that can be difficult.

Maybe in could take the form of “Community Grants” for specific modules (i.e Physics, Modelling, Sculpt, …). Part of the funding goal would go towards the salary of a module developper for 1/2 years (only a part so other areas can still be funded). The current projects of the module would be explained but the funding wouldn’t be tied to a specific goal. So donors could know more or less what to expect but the development team would be free to organize and prioritize its work.

As a concept for a campaign it could work really well and keep things fresh. A lot of the Blender communication is around animation (splash screens, Blender Studio films, website visuals, …), which in the survey covers 20% of users. Doing focused campaigns showcasing other areas might be a great way to show the richness of the community (it’s - truly - crazy - what - people - are - doing - out - there). There could be showreels, interviews with creators, BlenderHeads episodes with the module developers, Nodevember-like challenges, etc.

[LudvikKoutny]

@makizar

It’s a door that is easy to open and hard to close, if it happens once users will expect it to continue and get frustrated if it doesn’t.

This argument could be used against any form of improvement:

• Once Blender development speed ramps up, users will expect it to continue and get frustrated if it doesn’t.
• Once Blender development gets more focused on existing UX issues, users will expect it to continue and get frustrated if it doesn’t.
• Once Blender development pushes for radical reduction of existing bugs, users will expect it to continue and get frustrated if it doesn’t.

“We should not do good things because people get used to good things and expect them to continue…”

@makizar

Doing funding through as an accumulation of small feature requests would inevitably produce bloat, incoherent architecture and degraded performance. Developpers are already stretched thin enough as it is, they shouldn’t carry the mental burden of delivering some random features to secure their funding.
Its a well known adage that users don’t know what they want. Making decisions on where to focus development efforts is best left to the Blender team.

You haven’t read my post:

@LudvikKoutny

There are quite a few features and improvements team members have proposed themselves I’d start throwing money at in a second. The prime example being Brecht’s past proposal for the new texturing (and possibly also baking) system which would allow us to treat PBR attributes as single node connection (structures) and blend/mix them same way we can blend and mix BSDFs but without the limitations and performance BSDF mixing has. This would bring Blender on another level in terms of shading and texturing capabilities, but it never happened.

It would be totally fine if only the development team could choose proposals that can have bounties of them, and users would use their money only to vote for the priority of these features.

I proposed developers themselves would decide which features they want to put the bounty on. It would be a tool to let the users vote for priorization of the proposed features (by the developers) and bring additional income to temporarily ramp up efforts on them. I did not say anything about bounties being on small feature requests or users being the ones who decide which features or improvements they’re put onto.

@makizar

Maybe in could take the form of “Community Grants” for specific modules (i.e Physics, Modelling, Sculpt, …). Part of the funding goal would go towards the salary of a module developper for 1/2 years (only a part so other areas can still be funded). The current projects of the module would be explained but the funding wouldn’t be tied to a specific goal. So donors could know more or less what to expect but the development team would be free to organize and prioritize its work.

People probably would not want to donate towards “Physics, Modeling or Sculpting” but rather specific large scale features and improvements. With your idea, the same issue remains: People have no idea what they get for their money.

@makizar

As a concept for a campaign it could work really well and keep things fresh. A lot of the Blender communication is around animation (splash screens, Blender Studio films, website visuals, …), which in the survey covers 20% of users. Doing focused campaigns showcasing other areas might be a great way to show the richness of the community (it’s - truly - crazy - what - people - are - doing - out - there ). There could be showreels, interviews with creators, BlenderHeads episodes with the module developers, Nodevember-like challenges, etc.

This is actually a great example of why better targeted funding would be so motivating. If I were to spend money and I would know for sure it will go into actual “hard development”, like new features, improvements and bugfixes, instead of this “soft” fluff, like interviews, youtube episodes, challenges, etc…, I’d be way more motivated.

The bottom line here is that the Blender foundation and developers are free to choose whatever works for them. It’s their project, their foundation, their decisions. It’s only that in the survey, they have explicitly asked about what would make people give them more money, so they got a honest answer. They don’t have to act on that answer, but it’s there.

I originally posted in this thread just to warn about misinterpretation of that answer - that campaigns are very different concept to direct funding of specific development answers, and if they misinterpret it that way, then the resulting funding boost from such effort will likely be much smaller than expected.

[fsiddi]

Let’s please widen the focus of the conversation. Yes, funding is important, but I’d like to hear comments on how the survey was set up, if questions are clear, etc.

To answer some of the comments: this year (in a few months) there will be more efforts to make a new survey visible and for a longer period of time.

[LudvikKoutny]

Well, ok, this will be my last post in this thread, but the original post just ambiguously says:
“This topic is available to share feedback or report issues on the feedback survey published at”

It does not specify you are not interested in feedback about the questions asked and answers received, and you are just interested in feedback on the technicalities of the survey form and the results page.

[Nubnubbud]

my comments regarding survey visibility are also applicable here. The goal is to nag the users, without being obtrusive, or while being in obtrusive as possible.

here’s one that I really liked recently.

in the software Fan Control (it’s a simple app that lets you visually script how your fans work), which subscribes to the same sort of ideals as blender (simple, lightweight, limited internet connectivity, open ecosystem, open source, etc) There are options to check for a new version, and most notably…
a permalink to the donate page in an inobtrusive, but seen spot at the bottom of the menu.
We have places like “developer community”, and “user communities”, “release notes”, etc, where the donate thing could go, and I had no clue until I looked for it right now, there’s a little link on “about blender” and the splash screen…
hmm.
perhaps it should get a larger section. I don’t think many people would complain about the donate button getting a little banner area of its own, in a primarily donation-funded project.
perhaps a little “ways to support blender” banner? just throwing ideas out there. I can make mockups.

anyway, back to the focus of the survey

@fsiddi

To answer some of the comments: this year (in a few months) there will be more efforts to make a new survey visible and for a longer period of time.

what methods are you thinking of using? just as a sneak peek, if any. It’s important to vet these things with some of the community, saying “would you notice this?” or “is this compelling?”

like, if you don’t say “help decide the direction blender takes!” or something to that effect, it’s a lot less interesting. you guys got that one spot on.

[fsiddi]

You made your points pretty clear. Perhaps others can provide input as well.

[And-Woo]

I think that funding per project it’s extremely effective; I thought about it in the past, I’d like to speed up the development in the areas I prefer.
Though I’m afraid that funding side projects might take away money from the the main fund; I mean, many people might decide to fund a side project and unsubscribe from the main fund

[Anthony_Gibbs]

@fsiddi

but I’d like to hear comments on how the survey was set up, if questions are clear, etc.

For the most part, unless there’s a very valid or specific reason, I don’t think the questions should really be changed. Because if you do so, then you can’t compare one year to the next and generally, it’s identifying trends that can be more important then single isolated results.

That said and based on the survey results page which I assume gives a true indication of the actual question asked and the options one could select (given that it was a few months ago and I don’t actually remember the exact original survey questions and layout), then I do have a couple of issues.

4.3.1 “Do you use Blender LTS?”
If that was the exact question, then its too open to interpretation.
Do I use LTS, well yes, when the Blender release happens to be LTS as well, then that’s the version I’m using. Come the next stable release and I’ll be using that.
If the survey happens to be asked at the time that the current release is also a LTS release, then not only do I use LTS, I’m using at that time.

If the question is read that way, then the survey result of 20% of the people that don’t use LTS, must mean those people just skip a LTS release??? Unlikely.

Given that the previous questions ask which Blender version people use, I largely see the results of the LTS question to be meaningless and invalid.
Add to the fact that 14% of the people don’t even know what LTS is, is somewhat concerning.

In theory the purpose of the question is to help determine how useful the LTS release is. As such the question likely should be more like; “Do you primarily only use a Blender LTS release?”
With three options, Yes, No and Don’t Know.

A high yes count would then mean the time spent back-porting fixes for 2 years is well worth it. While a high Don’t Know count means you likely have a communication problem around what LTS is, etc.

4.4 Add-ons
Having add-ons data I think is useful, even if collecting the data accurately maybe a bit of a problem.
That aside, right now I have more an issue with the results format.
It lists Cycles as 9.97%, which in theory means 9.97 out of 100. Are you telling me that only 1 in 10 Blender users have Cycles installed and enabled!!! (keeping in mind it’s installed and enabled by default).

I’m assuming, that what it’s trying to say, is that 9.97 out of 10 people have Cycles enabled. In other words, there are very few (that ran the script) that actually disable it on purpose.
Is that correct?

4.5 “What do you do with Blender?”
Not sure that the options are totally the best (don’t remember if it was multi-choice or how many could be picked).
For example, I don’t earn a cent from Blender, so I guess that makes me a 'Hobbyist". But so what, does that mean I’m not an Artist?
Or, if I create educational Youtube videos around Blender (which I do), does that make me a Teacher/Trainer, but could I be a Hobbyist as well?

This question, along with the next one (4.5.1 Work) seems to be mostly tied around ‘professional’ usage of Blender, ie one is getting paid.

Maybe it all needs to be more well defined and split up. Along the lines of what you create with Blender (models, 2D and or 3D animation, motion graphics, data visualizations, etc) vs where or why one creates it (hobby, animation studio, game studio, etc).

Yes, that all needs a lot more work, given the possible combinations, etc. But either way, I think you need to much better define what exactly it is you want to know from those questions.

4.8 Do you use version control software when working on Blender projects?
Specific software, No, which would have been my answer.
Am I aware of both version control as a concept and/or some of the software used, Yes.
Do I implement ‘version control’ within my own projects, yes I do (using set naming conventions, etc).
If the goal is just find out about version control software, then you don’t even need question 4.8.
You can just have question 4.8.1 and one of the options is ‘None/Don’t Know’.

5.1 Do you make money with Blender?
OK, here we have a bit of a data problem and I feel some question double up.
If you make money with Blender, then in theory you are a ‘professional’.
We covered professionals in Question 4.5.1, asking what work they did.
Now you are asking again. But wait, it gets messy.
Of the three ‘Yes’ answers to 5.1, we have a total response of 2858 that make money with Blender.
Yet, in Q 4.5.1 we have 3624 people that are working in ‘Film and Animation’ alone (making money using Blender). How have we already lost over 700 people in Q 5.1 that are now no longer making money with Blender?
I know job layoffs in VFX, etc have been pretty brutal lately, but that’s some serious firing in the space of less then 10 survey questions…

I think that’s about it for now, much of the rest around donations/development, etc has largely been covered already.

[1D_Inc]

Nice example.

From my experience such kind of a development level indeed is dependent from a given dev skills and abilities.
(I remember at the time you put lots of efforts in order to resolve UV behaviour of bevel)
At the time I used to work with different developers to prototype different features, and indeed each time it was quite personal and very dependent from devs skills, attribution and background. Software development is like solving math - it depends on personal capabilities to solve the problem way more than on anything else.

I’ve been thinking about campaigns and voting with money for a while, and personally (as a diamond blenderfund donator) I found it rather doubtable.
The reason is that artistic userbase will mostly vote for making features while core development is far away from being limited to it.
Everybody will vote for frontend-oriented development (since this is part of a program they contact with), nobody will vote for backend-oriented development like refactoring, which will bring the gap, desmorale and tension between frontend and backend-oriented development.

In my opinion Blender is a versatile platform (ARTwork media production software based on CADwork engineering software paradigm) - such kind of a paradigm is known to be versatile, and therefore could be technically expanded in any direction (this is why it was possible to combine sculting, mesh modeling and videoediting applications in one software).
So core development level is more about protecting infrastructural functioning, stability and API design rather than features making.
Since production demands required more intense features making than core devs can afford physically, for features making we preferred to hire local python devs that utilize that API (Such an approach works well until core devs decide to shift software paradigm from versatile to local, and in that case all bets are off and existing votes are reset)

[1D_Inc]

Speaking of a survey, it is sad to see that half of a responders do not use Blender for profitable deity (at the time we tried to figure out possible solutions, for example tried to investigate strong parts of a program in order to understand what we can hire blender users for - what they can do that is hard or impossible in other software), and it is probably interesting which part of it succeed to reach higher production conditions at DataManagenent Conditions scale.

DMC level 0 - enthusiastic use (no strong production requirements to satisfy)
DMC 1 - personal use + vanilla data
DMC 2 - personal use + imported data
DMC 3 - collaboration + vanilla data
DMC 4 - collaboration + imported data (heaviest production conditions to satisfy)

As it turned out, the higher level assumes different requirements to the software abilities and design ( for example while DMC 0 stands more about flexibility and features, DMC 4 stands for quite opposite manageability, predictability and workflow speed to withstand production deadlines )

Originally Blender put lots of efforts in order to satisfy DMC 4 level in my opinion.

[txo]

Regarding all that voting with your wallet thing…( I’m also rather skeptic about that model working well for blender development)
I feel there’s one aspect not being discussed in that front, and that is… What happens if the amount collected for that project is not enough? Or if it is, but veeery slowly and the campaign lags for months/years? Or what if it barely met the funding goal but the cost was underestimate due whatever reason?

I don’t think it would work out as in a GoFundMe fashion because one big big part of development is design discussions and iterations, and that work is mostly already done when a GoFundMe campaign starts.

So it would be like asking devs to work in something that’s not a priority before the funding effort even begins. OR to estimate the amount of time a design is going to take… And we all know of a bunch of projects that didn’t get developed, or were stalled for a long time because of an acceptable design (maybe I’m mistaken, if so, correct me please). Or they get sidetracked because other modules/tasks suddenly get more involved along the way in order to fix an unforeseen roadblocks.

And then what happens with the donations? Maybe the devs can’t deliver, maybe the amount is suddenly not enough, but then the money is already spent and can’t be refunded… Maybe the campaign never collected enough to begin with…

I think it creates very shaky grounds for the funding process, where a lot of people wouldn’t donate to core dev., and there would be a lot of fragmentation of the funding, making it less effective.

[txo]

Maybe there could be a dialog somewhere during the donation process that asked to the donor what feature would they liked to see developed the most, WITH A CLEAR DISCLAIMER THAT IS JUST FOR POLLING PURPOSES, without making any promises or compromising anything by the Blender team.

(Although it would create the issue of recurring donations not being able to change its answer each payment period?)

[1D_Inc]

I actually faced this dilemma, which is typical for venture capital investments.

Since such kind of a development is very personal (is very dependent from determination and skills of a developer who tries to challenge the task) and the result can not be predicted, to support such a development you basically have to donate in form of a bet - realizing that the task could fail any moment.

I used such kind of a funding when we tried to prototype and propose nodebased procedural modeling approach - it was a nice idea compatible with Blender infrastructure and paradigm, but there were absolutely no guarantees that we will get something like geonodes as a result.
So financially it was quite a leap of faith, where you can only provide available support and believe that one day everything will work out.

Sometimes I track and support different people who try to produce interesting development (for example, participated NURBS development fundraising recently)
Such kind of a deals are already available but they are definitely not for everyone, since there is no direct connection between investment and outcome)

[thorn-neverwake]

Any day of the week, I can contact a freelance blender python developer, and propose a function that I would like created.

They’re able to tell me if it can be done, a time frame, and a cost. Certainly time and budget can vary due to unexpected things happening, but those two factors are not simply removed from the conversation.

There is a goal, a time frame, and a budget.

This is sort of how the rest of the world works. I tell you I want a pizza, you tell me it will be 20 minutes, and when I receive it I give you $20.

I tell you I want a company website that does various things. You tell me it will take a year and a half, and cost $19,000.

How are those things different?

[ThinkingPolygons]

i think bf should just have a good feature request system, and tackle the most requested features in order, and thats it. pretty sure most people would be happy with that.
no dev time wasted on stuff no one asked for.