After Blender's great scores in our CG upgrades survey, we interviewed Ton Roosendaal, Chairman of the Blender Foundation, to get his views on the results.
Confidence
Overall scores for user satisfaction with software upgrades
CGenie: Blender received overall the highest mark from its user base in the survey, despite it having a smaller share of the market than many of the bigger contenders, it certainly evokes a lot of passion from its customers, as can be seen from the comments posted both on the article and the survey itself.
Firstly, congratulations on the excellent results that you've received - if you didn't know it already, it's pretty clear you have quite a vote of confidence from your users. Was this expected?
Ton: We are certainly well aware of the passion of our artists and developers for Blender, so that's not a surprise! What I'm very happy with is that this survey was executed independently of Blender channels or websites, so the results didn't get polluted by orchestrated internet community floods. It gives us confirmation of successes of our strategy over the past five years -just working hard on getting the program improved and staying loyal to the original concepts behind Blender and to its user base.
And even though there are perception issues - free and open source programs typically are ‘made for the average user' and ‘horribly designed' - the results clearly show an increasing awareness that Blender is being made for artists with a serious interest in using 3D on a professional level. Which, like any program you've surveyed, just requires time and energy investment to get most of out of it.
Being free
CGenie: The whole foundation behind Blender and the product that you have created seems to have really strong support from your users. The usual challenge you must face is ‘that's because it's free', do you think Blender would receive the same scores if it was a commercial product like the rest of your competition?
Ton: Honestly, not such high scores as these.
On the other hand, marketing practices also taught us that free products or gifts are typically evaluated with much more criticism than the product you've paid two months' worth of salary for, which - of course - must be a hundred times better than this crap free stuff! So let's average out this aspect to be only slightly relevant.
The key issue is not the "free" aspect but the fact that Blender is open source. This makes it entirely different, both in development and in feature set, but especially in defining ‘customer' satisfaction.
Also, Blender's "free" aspect doesn't mean people don't spend money on it. Blender.org runs a popular e-shop with support products, and organizes community-funded film and game projects. Blender Foundation's spin-off Blender Institute currently employs four employees, and during projects here we hire six to twelve more people. Another good example is from book publishers we work with, who happily conclude Blender books do as well as their other 3D titles. We're very proud to see the rise of professional trainers embracing Blender, colleges and universities teaching it, artists getting income as integration consultants, or developers being hired for servicing and support.
The key issue is not the "free" aspect but the fact that Blender is open source. This makes it entirely different, both in development and in feature set, but especially in defining ‘customer' satisfaction. You can't take that aspect away from Blender and still evaluate it - it would be an entirely different product then.
Lastly, if you were to look at the license fee income for commercial 3D applications and check the actual amount spent on 3D development, you'd be quite upset! Licensing is typically not the best income source for companies either - they make greater profits with servicing, support, bug fixing, special versions and training. This is, surprisingly, the typical open source business model as well!
Features
CGenie: One really strong area where you shone above the competition was around developing the features that the users need first. Can you share a little detail on how you find out these needs and how you decide the features to be focused on in each upgrade?
How users rated Blender on the statement
"The company focuses on developing the features I need first"
Ton: This is what I usually call the "open source dynamics". No matter how much influence I might have, I can't make the software move in any particular direction without a very solid and well covered consensus by its active volunteers. A good example is the period shortly after Blender became open source, in 2003. Instead of picking up all the beautiful visions I had for exploring real-time interactive 3D creation, the developers just worked on fixing and improving the modelling tools first. This kind of dynamics really comes from the users themselves. Blender has always attracted people who start using the software first, and only later get involved with the coding side of it, just to add what they need or to improve what they don't like.
Around 2005, when development really paced up well and the Foundation's income allowed us to hire people, I decided to not attempt to set up a software company again but instead organize a project to hire the best artists in the community to make a short animation with Blender. This concept of artist-centric development, jointly creating an open content film, has helped Blender incredibly well. The logical next step was founding a permanent studio in Amsterdam to do this on a regular basis. In 2007 and 2008 we worked on another short animation movie and an open game project. For this year a new short film is on the agenda -this time exploring epic and action-based content, and hopefully in 4k stereo digital cinema!
Software divide
CGenie: Reading through the reader comments of the survey article, you get a real impression of a divide between Blender and the rest of the CG software. Some users seem to consider Blender as an ‘amateur' application and some are fiercely passionate about it. Why do you think this is? Is it more perception or is there some extra features needed before Blender can be considered ‘production-ready' for use in major studios?
Ton: Of course that's mostly perception. From people who know both Blender and commercial programs well you'll get a more balanced view on issues. Blender stands out as a good and fast subdivision surface modeller, has excellent tools for UV unwrapping, top quality fluid dynamics and cloth, a fully integrated render-compositing pipeline, and character animation support that's up to par with what you would expect in any of the top CG tools. There's even advanced options in Blender you wouldn't find in others until recently, like quaternion-based bone deformation and harmonic-coordinate mesh deformers.
I'd rather focus on delivering the best 3D tools possible for individuals, independent teams or small/medium sized studios. For them an open source program like Blender can work as their own in-house software, similar to what the major studios mostly use anyway.
Blender has plenty of weak points as well. Our rendering system (shading, lighting) has fallen behind a lot, our event/tool system got stuck in the nineties (no custom key maps, no history stack), and our Nurbs tools are only laughable! All these and other topics are just inspiring and fun development targets, which we'll work on and for which we welcome new developers.
Whether Blender is "production ready for major studios" is not really our interest though. We don't make Blender for Hollywood studios - those people can take care of themselves pretty well. I'd rather focus on delivering the best 3D tools possible for individuals, independent teams or small/medium sized studios. For them an open source program like Blender can work as their own in-house software, similar to what the major studios mostly use anyway.
I would also like to express that I humbly respect and admire each of the other tools in the survey. These programs all have great developers and designers working on it, features and tools that are excellently designed, innovative, and in many areas more powerful than what Blender offers now. I'd be the first to admit we're not among the top 5, but do consider us to have firmly entered the top 10 now, with flying colours!
User support
CGenie: A strong result for you was around support for users, where again you were some way ahead of many of your competitors. This is obviously a perceived downside of an open-source application, how do users get support for their problems/technical issues?
Ton: On the contrary, if open source projects do one thing well it's user support. Around Blender there's really hundreds of user websites world-wide. Everyone with a bit of technical background can help fixing bugs or provide improvements in areas. You don't have to purchase exclusive support contracts, nor have to hide in public forums that you're actually using the software illegally.
What we especially need to improve in further is professional-oriented artist/development support. These would be great business opportunities for everyone to grab, and by nature the most viable business model for open source programs.
Frequency of upgrades
CGenie: The frequency of upgrades was very popular with your users - it seems like you've pretty much got it spot on. How do you know what constitutes an upgrade? When do you say ‘this is a new release'.
How users rated Blender on the statement
"I feel the frequency of upgrades/new versions is about right"
Ton: This is actually getting a bigger problem nowadays, with more developers and projects running simultaneously. Ideally we should try to release every 3-4 months, but in practice releases come in 6-8 month intervals now. What I think users mostly appreciate is the fact we really release true value - packed with new features - and that we spend a lot of energy in tackling every bug report that comes in. There's an unwritten rule to only release when the bug tracker is below 50 open reports, which we managed to achieve many times.
The actual decision to make releases is another example of a dynamical process. Developers just want to get their work published officially, users start begging for it, and then sooner or later these signals are too strong to ignore really. In our weekly IRC (internet chat) meeting we then just freeze the functional targets for a release, kick everyone to finish those, add time to test it all well, freeze the code for a month to only do bug fixing, and then there's a release! I personally love to coordinate such efforts, deadline rushes make everyone inspired and add the adrenaline to move on to new projects.
Value
CGenie: A very interesting result was around value, unsurprisingly you came out top - but it wasn't quite as huge a margin as you might expect - Houdini ran you very close. Could you ever see yourselves offering optional commercial support in the future for customers who need it?
Ton: It's not the Blender Foundation's goal to conduct business; the projects we do are to coordinate online projects around Blender, stimulate development, and organize presentations and conferences.
The Foundation's strategy is to invite and help as many third parties as possible to set up businesses themselves. Most of this is still quite in its infancy, we still need to improve the way we communicate how potential clients can find reliable user or development support.
We have started an official certified trainer program on blender.org, and some websites offer commercial opportunities already, like the independent Blender news portal BlenderNation. For the rest, people can just email me and I'll forward them to the right person to contact (ton at blender.org).
The user interface
CGenie: If there is one area that users commented negatively around it was the user interface of Blender being quite different to other applications and users have found it difficult to pick up - is this something you're looking to improve?
Ton: Blender's much disputed UI hasn't been an unlucky accident coded by programmers but was developed almost fifteen years ago in-house by artists and designers to serve as their daily production tool. This explains why it's non-standard, using concepts based on advanced control over data (integrated linkable database) and the fastest workflow possible (subdivision based UI layouts, non-blocking and non-modal workflow).
Blender uses concepts similar to one of the world's leading UI designers Jef Raskin. When I read his book "Humane Interface" five years ago, it felt like coming home. His son Aza Raskin has continued his work with great results. He's now lead UI design at Mozilla and heads up his own startup. If you check on his work on humanized.com you can see concepts we're much aligned with.
Making good user interfaces is not easy, and people don't always have time to work on that.
To be clear, it's not that "Blender has been based on Raskin's ideas" but that these concepts are quite obvious from a UI designer perspective, more people have developed similar concepts and use it with a lot of success. If you look at the evolution of other CG tools you can see similar concepts being adopted as well, cleaning up the cluttered desktop with a subdivision system, and presenting a parallel and non-modal workflow.
What Blender especially failed in was mostly a result of evolutionary dilution, coming from some laziness and a pragmatic ‘feature over functionality' approach. Making good user interfaces is not easy, and people don't always have time to work on that. Moreover, because of our fast development, it's not always clear to new users whether they are trying to use something that's brilliant or is still half working, or simply broken by design! For me this mostly explains the tough initial learning curve for Blender, or why people just give up on it.
The fact that Blender's core was written so long ago was also a reason why it wasn't easy to be redesigned or replaced. Over two years ago we started working on this: collecting design proposals, evaluating the good and the bad, and setting up a good target for a new UI system that can bring us into the next decade. This is the Blender 2.5 project - development for it started up again last year in October and is expected to show tangible and usable results this summer. For further reading on this please check our 2.5 wiki.
Blender 2.5
CGenie: Blender 2.5. Your users seem to be hyping it quite a lot! It's also being watched with interest from users of other software - so what's the latest news on it? Can you share some of the major improvements users will notice?
Ton: Yeah lots of excitement here. The 2.5 project even gets abused a bit to tackle any criticism ("wait for 2.5 and your problem gets solved!"). Although this project promises radical improvements, I prefer to keep everyone focused a bit on the simple basics we should achieve first. It should still be a project with a short-term deliverable too. I would really love to show the first working version on Siggraph this year!
The results will make it much easier for new users to become expert users and configure Blender to match their own preferred workflow and creation targets
In short, the project first solves our arcane event/tools system - which internally was still based on the original 1995 IrisGL conventions. We now have centralized dynamic event handlers which can be registered and fully customized. All tools have a uniform API to enable repeat, redo, macros, history stacks and scripting access.
In addition to that we added a uniform method to approach and extend all internal data, not only to assist the above but to allow semi-automatic UI generation (Python scripted) and provide all-over access for the animation system and scripting API.
These technical targets will then enable designers to solve a lot of communication issues in the UI as well. We spent a great week together on this in a workshop in Amsterdam in early March. The results will make it much easier for new users to become expert users and configure Blender to match their own preferred workflow and creation targets as well.
For the full picture I'd like to forward the readers to the Blender 2.5 web page.
The future
CGenie: We'll be running the same survey next year, what change would you like to see?
Ton: With similar good results I'd be more than happy!
CGenie: Thanks very much for your time Ton, and keep up the great work!
Related interviews
- CGenie interview LightWave
- CGenie interview Luxology
- CGenie interview Side Effects
- CGenie interview Autodesk
"