Blender 2.8 – the Workflow release

This is a proposal for work focus on for the coming year.

I’ve written this because we keep missing bigger development targets – we don’t have enough time for larger projects. Instead too much time goes to releases, bug fixing, reviews, maintenance and support topics. The bug and patch tracker duties are keeping the best of our developers away from their own targets.  As a result we then don’t have time for design docs, for planning, logs and in-depth sessions with the module teams, and have no time for the artists who are involved to make sure we’re well aligned and know what to do. I think everyone has noticed that we’re floating too much, things are not clear. Where are we heading? Who does what, and how do we decide on things?

So – it’s time to act and gather the troops to refocus and get back energy, to maximize involvement from everyone who’s active in and make sure Blender can survive for many more years.

—– Blender 2.8 – Workflow release —–

Just like for 2.5, the proposal would be to take a bigger leap to a bigger release by not releasing for a year. The 2.76 release then would be the last ‘real’ version we do until 2.80 somewhere in 2016.

Obviously, for the crucial fixes and smaller (stable) features we can do update releases 2.77, 2.78 and 2.79.

Topics to finish for 2.8 could be:

  • UI work: wrap up Python configurability project, make Workflow based configuring possible
    Proof of concept: the stripped “Blender 101” for high school kids.
  • Viewport project, including a PBR quality engine/editor that could replace BI and GE render.
  • A better designed integration of physics simulation in Blender
  • Invite the GE team to rethink game logic editing, to use viewport and new physics
  • Don’t add the half finished Gooseberry targets but take the time needed to code it well:
    Particle nodes, hair nodes, simulation nodes, modifier nodes…
  • Asset managing and browsing, linking, references, external files in general.
  • Integration in non Blender pipelines.

Practical considerations:

  • Move development to special 2.8 branch(es)
  • Module teams are empowered to cleanup quite radically and get rid of legacy code.
  • The 2.8 series is allowed to be not 100% compatible with 2.7x. (Physics, particles, games).
  • Spend time on organizing ourselves better, agreed designs should lead to more empowerment.

And some core principles to agree on:

  • We reconfirm and where needed update the 2.5 spec docs.
  • Stick to existing Blender data structures and code design for as much as possible.
  • Make Blender ready to survive until 2020, but…
    … start collecting the list of bigger redesign issues we need to for a 3.0 project
  • Bring back the fun in Blender coding! :)

The article for the roadmap of 2014-2015 is still valid in my opinion. We just need to take a break of 9-12 months now, to make it work for real.

Blender 2.8 Workflow Sprint (not yet scheduled)

In the coming months we can discuss and review the plans and make sure we’re 100% aligned on the 2.8 targets and for other work during the coming years. We should also meet and have good feedback sessions on it. This we can further discuss during the Blender Conference as well.

  • Four days of workshops and design sessions.
  • Travel and hotel covered for by BF (and Dev Fund, or a new fund raiser?)
  • We should try to get someone from every (active, involved) module team on board. Also key user/contributors have to be on board. But it’s also more efficient to keep it compact.
  • Proposal: we do this invitation-only: First we invite the 5 most active contributors of past years. Together they then invite persons more, until we have 12 (?) people.
  • Sprint sessions can be in parallel too – UI, Viewport, Physics, etc. Let’s make it public as good as possible.
  • The Sprint results get presented and reviewed during further on sessions during the Blender Conference.

Seven years ago, back in 2008, we also took a break for more than a year, to get the 2.5 project started up. It was a very exciting period where a lot of new things were possible and could happen, even though we didn’t finish everything… it gave us quite a solid foundation to build on, attracting a lot of new developers and great features.
I realize we have to be realistic now, not everything will be possible. But we also shouldn’t stop dreaming up a good future for Blender. Let’s take a break from our demanding release cycle, rethink it all, but not for too long. Let’s cherish what we agree on and enjoy the freedom of a configurable workflow that will enable you to do what you think is best… for making 3d art, games, film and animation!


  1. Greetings It lacks automatic opening of fbx and other popular formats by double clicking on the fbx file. I work in unity and every time I need to make changes to the model, I have to go through a long process of importing a file into a blender. If this is done, it would be nice to remember the import path and, when exporting, immediately go to the folder with the source file and the original file name. Now we have to use Cinema4d for this. There it is implemented conveniently. But I would like a blender.
    P.S. I know that this can be implemented using a script, but this is a very uncomfortable crutch.

  2. When is this going to be released? any ETA? is it going to be out in 2020 or something?

  3. I tried Blender 2.8. on my Mac (all WIN version fail in the start). Really don’t like constant Cycles Render in edit mode. Every movement made noise. After 5 min work my eyes were tired. Can’t find a way how to turn off it.

  4. Also, a request.
    Please, implement similar tool as Cinema4D Mograph.

  5. Could you keep Blender Render (BI) in future releases of 2.8 so it will continue it’s unique progress.

  6. Please don’t forget the architectural workflow in addition to film, animation, and games. Importing layers and materials from other programs are a top priority. Second most important is UV box modifier to assign materials with uv unwrapping everything, and third, better Nurbs. cheers!

  7. it is crucial that changes be made in the user interface, not only for advanced users who want to make trancision a blender from other modeling software, you need to make Blender a more optimized tool that facilitates learning to young people who are just entering the world of computer graphics, it is a difficult task because the user interface is like the soul of software and reform means a colossal task for programmers, but if this matter is not Incorrect, I fear that will remain seen as blender tool for amateurs …

  8. Nice!

  9. I never tried this before so imma see what can happen


    used it to understand me copy the text i wrote here. Paste it on the board press the speaker to listen if you don’t understand that translate the language to your country. thank you so much blender please fix them!?^_^

  11. blender you need to add the cycle rendering to the game logic.
    It’ll be great to put the cycle rendering with the game logic.
    so that way they can finally work together.
    blender you need to fix the clothes problem so they will not show their skins.
    blender you need to fix the disappearing and reappearing
    characters in objects can’t disappear and game logic. starting the game the object or character will not disappear and reappear when you’re pressing the f key. it does work animation but not in the game logic when you’re playing the game.

    Blender please fix the color so characters can change a differnt color when you assign one color on the body. assign 1 color to a character will not fix things in the game logic thegame logic the character will stay one color even if you assign it will work on animation but not on the game logic. using a double character will fix it in the game logic some parts.

    blender please fix the objects when you want to place an object in a body’s character when the objects are moving in the body that cannot show their appearace when they’re moving in the game logic. they don’t show an animation but game logic yes.

    Blender these are major critical bugs you need a fix.

    Please blender please please fix these major problems.

    it’ll also be great to improve the Sub Surf Modifier

  12. No. Way. I like Blender!

  13. I think the Blender’s GUI does need some enhancements. I do not know much what QT is but I do not think QT has anything to do with keeping what we have but switching them around. IMO, I find it still to easily to find myself jumping around too much. Most importantly, I think most users would love to see a bigger viewport area, without loosing or moving their second most important windows off to the side. To me it slows my work down when I have to either constantly toggle the maximize viewport area or shove the the second most important window to the side.

    I think a better method would be, first, renaming the Property Window to something else for less confusion with the viewport properties panel name. In fact, I really do not understand why there is a viewport properties panel and a properties window. In addition, it would be really nice if Blender could implement Tabs on the right-hand side. I think by moving the properties panel in its own tab along with a Properties Window tab, which should be renamed, and an outliner window tab, will allow users to easily swap between windows without having to close or resize other windows.

    Secondly, I think some of the stuff in the viewport properties panel has too much crap, leaving users to constantly scroll up and down to find what they are looking for. Therefore, it would make more sense keeping things like, the selected object, its transform data, and display mesh data, such as display seams, display normals ect. in the properties panel and move the other stuff accordingly.

    1. 3D Cursor Location
    Starting with the 3D cursor location, I think of it more as a precision tool, when you want the cursor to be placed at an exact location then a property data. Therefore, I think it should be moved to the Tools tab.

    2. Scripts
    IMO, I think scripts that now are placed in the viewport properties panel should be moved into a tab called Scripts or Add-Ons.

    Graphic Properties
    Things that have to do with graphics, such as shading mode, backface culling, quick render, grid setup, and so forth can be placed in a render tab. In addition, you could then move the rendering properties window data into this tab.

    If anyone wants more information, please do not hesitate to ask.

  14. There are a lot of comments about the UI, so trying to determine the high level direction is going to be difficult. Blender is a very powerful application that spans multiple disciplines. This is going to cause a lot of differing opinions and make it difficult to make a decision. UI is a moving target that can always get better with iteration.

    I think the biggest struggle with workflow improvements will be how different areas of the application will interact with each other. The more specialized people become, the less they want to look at non-related functions. Someone that just does sculpturing does not want to see animation, compositing, or game engine functions.

    Someone that is more of a generalist will want that type of flexibility at their finger tips without changing their entire workflow.

    For determining what direction to choose, I would poll your existing artists and find out how they currently use the program. Are they more generalists or specialists? Understand that each person has their own needs. Most people only use a small subset of an application, so you can’t make a high level design decision based off any one person’s feedback.

    The goal isn’t to have a perfect solution but make improvements that will help artists create artwork faster. Ground your decisions in use cases. Blender probably has thousands of use cases, so there will be a lot to document during this upcoming sprint. It takes a lot of time and energy to ask these questions but it is important.

    I do a lot of design work for the open source application for Krita. For the 3.0 release, we had 7 UI revisions for the layers management system before we had something that we liked. There are still some negative parts to the final solution, but the negative sides were weighed against the postive sides.

    People expect the UI to “just work”, but doing good UI work takes a considerable amount of energy and research before a ‘good enough for now’ decision can be made. For this beginning UI research, I would highly recommend this book that discusses the process.

    Can’t wait to see what you guys come up with. I am sure it will be great.

  15. -alembic
    -“right click pie menu”
    -float windows on top option

    is my wish list

  16. I’m just worried about all the nodes that exists in BI and not in Cycles, just in case BI disapers, and exited about all the changes.

  17. Some are nice and some are mean

  18. Crucial for Blender development is improved fluid simulation, which for now is shi… really, really bad. Don’t get me wrong, I love Blender as whole, but better particle simulation should be priority now, otherwise Blender will be always behind in VFX area.

  19. Its good

  20. The new particle system could emit particles from other particles?, this is very cool for combining effects.

  21. if you can add network in logic BGE it is help us to make easy network games or online games and add publish game to mobile android

    we want add particle system in BGE

  22. Sincerely appreciate the information from Ton, and the mostly helpful comments by those included here. Fully support the ideas of the “deep” planning and release with some planned breakage, a necessary consideration. I know that BI is heading towards the exit but Cycles is far from useful in NPR work as noted in the thread spoken by B-NPR, far more implications for use than some other ideas noted. Anime, animation, illustration, game dev., arch-wiz, and remarkable hybrid cinema development, plus faster response creatively. Push for the GL or other responsive render choices as part of Blender’s standard is important, and I would hate to have it get lost in this sea of suggestions.

  23. Can’t wait for the 2.8 !
    I have a question, if anyone can give me answers or links to answers.
    Is it planned to add support for smoke rendering in Cycle on GPU ? And when ?

    Thanks a lot in advance

  24. Will there be an oficial android version of blender?

  25. Bug or ????

    I built a fire inside a fire bowl. When I parented the fire to the bowl and tried to move it all….all moved EXCEPT the fire simulation. So now when i render, the fire is not attached to the fire bowl.


  26. Can you possibly separate the Smoke and Fire Simulators please. I am using 2.76b and tried to follow the manual because I need FIRE ONLY…the manual does not seem to be current and bounding between smoke and fire seems to mess with the node system


  27. As a Student Teacher Assistant, I have always believed that something should be done to help beginners not get discouraged so quickly with overload. This Blender 2.76 is Great but FRIGHTENS beginners with so many buttons, and obviously the UI is NOT built for them nor for lab teachers who try to teach something like this in three course: Beginner, Intermediate, and Advanced.
    So if you want to make it easier for beginners and crate a nice leaning curve flow towards Advanced, I suggest you consider the following change to the USER PREFERENCE: A Experience Level selection (Beginner, Intermediate, and Advanced).
    When Beginner is selected, only things that help beginners learn to create and produce are enabled.
    When Intermediate is selected, all beginners are enabled and then things that add variety to their product without getting overwhelmed
    When Advanced is selected, EVERYTHING IS AVAILABLE

    For example, CG-Cookies has a nice FLOW path for people to move from Beginner to Intermediate.

    I would suggest if you have not thought along these lines, you have a conference with real front line teachers who teach beginners especially High School teachers.

    Thank you

  28. Do you think we could get rid of the blender game engine and maybe introduce a cycles game engine of sorts? It seems more as if this software isn’t really meant for game creation, and is more for video creation. I’d really like to be able to see some better games have the title of blender on it. Also I am fluent with blender but not many other 3d softwares’. I’m beginning to understand unreal engine 4 but it doesn’t really seem to be all that well at sculpting. These are the reasons why I’d like to be able to see blender progress into a better game engine than BGE.

  29. It is helpful

  30. I thought of another feature that could be extremely useful in sculpting. The ability to store masked selections would be awesome. So I can edit fingers easily and cycle through previous selections instead of having to select and invert each time. So in essence having the possibility of assigning a vertex group from the masking in sculpt mode. And when selecting a vertex group have it create a mask selection of those vertices.

  31. I think to blender grow faster, the top one priority is integrate a full working ALEMBIC Import as standalone.

    Because i wish to integrate into Houdini pipeline it is very, very important, vital.


  32. Hello all …i am in game development and i don’t can imagine the Blender without game engine…is a must have for ever feature.
    Focus on real time physics water,particles,fire olso it would be so nice.

  33. I come from using Maya, Zbrush, 3dcoat, and Mudbox. All 4 different applications that have different key functions that I have to remember. I switched to blender because I am a programmer and there is no way in the world I could juggle all those applications and do my duties. I found blender to be a wonderful fit being that it can do almost everything those 4 applications could and for free(compared to the thousands those programs amount to).

    With that said I have an issue regarding Texture Painting, and a request regarding Sculpting.

    1. The one application I have not been able to get rid of is Mudbox. And it is not because blender cannot do the duties of texture painting, but rather it has a number of issues. Painting at 4096 chugs, painting on faces stretches across adjacent faces, and the functionality is spread across so many tabs that it takes away a lot of time to where I’d rather just use another program.

    2. There are also some functionality of Zbrush that would be awesome in sculpting. The lasso trim/cut tool to easily shape meshes with smooth surfaces would be awesome. I can emulate this using the bool tool but I’d have to shape the object which is time consuming. In Zbrush you just lasso and it is cut on release. There is also a feature in the sculpt tool add-on which somewhat does this with the grease pencil but it is clunky to use. Lastly it would be a dream to get something equivalent to zRemesher where the remesh retopologizes with the flow of the object. I say dream because I know that is probably not feasible at the time.

  34. Also think about cycles rendering, is was not too much, i’ts something insterresting too don’t we ?

  35. I really can’t understand why the Blender team doesn’t even consider fixing the main issues that are around since 2.4 and still here as now.

    – Linked objects are not available to the shape key editor. This makes the whole shape key editor unusable for animation and we are forced to use bones just to drive morphs.

    – We Can’t link multiple instances of an object into the scene. So to have two dogs in the scene we have to keep two copies of the same dog file on disk.

    – Morphs store the whole geometry of an object even if a few vertices are used. This is a major waste of memory nobody cares of (???). This also breaks compatibility with dae fbx i/o when morphs store only used vertices.

    – The CMS stills reading srgb values as linear values as far as import/export of legacy or game models is concerned. this means if we load a 2.4 model its colors are gamma-shifted in 2.5-2.7. This also affects obj 3ds dae fbx materials as well.

    Any chance to have these bugs fixed in a Blender “never” release may be ???

    Alessandro (Italy)

  36. just finished watching the conference talk about this.
    very interesting stuff and the wait for it may kill me.
    i’m especially excited for the PBR viewport/rendering stuff. i’m not really a game artist but i also don’t have really powerful hardware so even though i love working in cycles the render times can be slow. real time PBR rendering with the advantages of being in blender (render layers, smoke, particles, freestyle, etc) even if those advantages slow the full render times a bit is an amazing thought.

  37. Are there going to be updates on OpenGl, since now its using 1.4 version which is very old ?

  38. In August it was clear that a ‘sprint’ in October would be too soon. We will need more time to gather the topics, get the design issues clear and find out who will be able to work on it. It’s going to be a main topic of the coming Blender Conference in Amsterdam, end of October.

  39. I always wished the mirror modifier could have a vertex group selector.

    Vertex group — subgroup (which is for animations, and which are for modelling purposes, for modifiers etc)

    Just throwing some ideas here.

    2.8 hopefully is a break through.

  40. I want use Blender this way: divide work become five main steps: Model, Build Scene, Animate, Render, Composite. This follow natural for dividing work.
    1. Model: Model and character build people create props, characters, and backgrounds files. Model people add color, textures for the models.
    For characters models, also they have rigs, constraints, control points, basic animations, basic expressions. Model people no need know WHERE put object or ORIENTATIONS of objects in scene, job for scene builders construct scenes for animators.

    2. Build Scene: Build scene use model and object files and put in scene file. Scene builders orient, scale, position, duplicate objects and add lights, camera follow story boards. If need edit objects MUST edit in original object file, can’t edit models from scene file. After edit, models AUTOMATIC update, no need delete and reload model.

    3. Animate: Animators load characters and animate them in scene file. They also animate special effects, cameras, lights, etc. Often animators want test animation with temporary objects before have real models. When have real model, animator can select temporary object and replace it with real object and real object will follow animations of temporary object automatic.

    4. Render: Render all passes into image files from scene file. Generally, Cycles and OSL work great.

    5. Composite: Load images from render and composite them in composite file. Do final render. Blender already have automatic reload for image files if they change.

    Now, Blender have no way link model files but still allow change position, orientation, animation, etc. Must append models in scene file, make scene file large and no automatic update if edit model files. This is big problem if use one model in many scene files. I think this is BIGGEST work flow problem. Replace temporary objects with real model and keep temporary object’s, position, orientation, scale, animation is second biggest problem for me. May be Blender can do this now but I not know.

    One way do this is start with a simpler core Blender, and create proto type for testing this work flow.

  41. One improvement that would be very helpful would be a node editor for modifiers and operations, similar to the way Houdini works. This would be a powerful feature, allowing for more non-destructive editing and would make many more possibilities for procedural and general modeling/animation. For me personally, this is a feature I would like to see the most (along with the particle nodes).

  42. for me, i just need ONE improvement in the UI. add margines/lines/borders to separate settings in the properties panel. just draw a line to separate the two groups, like in unity. I think this is very important for a mainstream 3d software as it induces greater confidence among its users.

  43. I am Blender user since version 2.25 and I look forward to version 2.80, especially to see what the Blender development team is prepararando / idealizing the advertised Interactive, which aims to integrate the BGE (of which I am most frequent user) the main Blender code.
    Anyway, I thank all the devs to give us this fantastic tool called Blender. Of course it does not (and it is not humanly possible, taking into account that it is an open source application) meet all susgestões but trying to provide those who favor the majority of the community and breaking acquire new users.
    I agree with what a colleague said above when referring to poor documentation suite, it hinders and the adoption of the tool, which leads to many inicianteas users to migrate to other tools available in the market.

  44. Dear Blender developers,

    There is this post available on “” where some of the users of the Game Engine are trying to summarize this topic:

    “Invite the GE team to rethink game logic editing, to use viewport and new physics”

    And understand it both from a user’s point of view and also from a developer’s point of view.


  45. Many people go to bge for the nodes, but everytime there is a problem because the engine “is not designed like a game engine” and kill the project.
    I think, Bge is a great program but things need to help stop polishing the project; many tiny things like “loading” screens, an upgrade for the text (plane image mode?), on game change resolution, instancing, simple pause logic brick…

    And, what about replace logic brick by nodes like cycles materials?

    Atte. @Mautar55 desde Argentina :)

  46. There have been some great suggestions and comments here. The biggest problem for me, when starting out in Blender, has been the paucity of documentation (by that, I mean stuff you can *read*, not video. I don’t have internet on my own computer, and often, videos are not downloadable; also, they don’t have the “random access” that a readable document does. I *have* learned a number of things from watching videos, but finding again the place where the information was seen is not always easy). The manual has been improved greatly, but what’s needed is not just a list of options, but things like what situations I might want to change an option, what will be the effect on what I am modelling, why I will find a particular thing useful … etc. There are documents like noob2pro, of course, and the information in them is quite extensive, but I find they are extremely repetitive in a way that often makes them hard to follow.

    My learning has been (and still is) very slow as a result.

    In general, I quite like the UI; I like being able to see what options are available (even if I don’t know what they all mean!) rather than being hidden away in a ribbon structure – which in my opinion often makes finding options harder rather than easier. I like the way the windows layout can be customised, and I like the way that headings can scroll if the window is not wide enough.

    The file browser is a bit cumbersome, though, as has been pointed out. One very frustrating thing for me, as a Windows user (I don’t know if this applies to other OS) is that Blender doesn’t seem to recognise Windows shortcuts. I use these extensively on my desktop to quickly navigate a complex directory tree in a single click. It would be great if Blender could have the ability to display these shortcuts and navigate by means of them.

    I could probably say a lot more, but time, unfortunately, does not permit right now.

    Looking forward, as always, to new developments,


  47. “Invite the GE team to rethink game logic editing, to use viewport and new physics”


    – Wanna be “Steam Greenlight”/indie game dev:



    – Bizarre hippie/campi. Wanna be rockstar / artist / VJ inside contemporary art museum or event:



    – Robotic’s hero:


    – Eskil Steenberg & Verse + Logic Bricks binary.

  48. I am a Blender (modeling and animation) user that is interested in getting into simulations. I know some Unity and Unreal, but want to do prototyping in Blender since I dislike having to leave the program. Just like I do not enjoy going to After Effects from Maya. Having the compositor there in Blender is what I love more than anything else. Also, I am hoping that the logic is easy for individuals that do not have time for fulltime programming.

    Anyways, is it worth it to learn the current Blender GE? How much will it be changed once it becomes the Interaction mode? I do not want to relearn how to use the newest logic bricks and all of that. Is this a 100 percent remake?

  49. First of all: blender is the most complex and most impressive free software avalible. I hardly can imagine how many thoughts and work it took to create it. Living from 3D for many years i of corse know blender well and installed it many times. But to learn it i had to get forced. It took ages and robed my nerves. I normaly am quite fast in learning new software, so what is the reason that even though i always felt atracted by blender never realy learned it, and why fo i ansver this thread. Well i think taking a break and thinking in big terms is fantastic and not just smart but also really needed for the further positive developement. But judging the topics that are suposed to get addressed in the time i think, that it is not the big concept stepbthat blender needs. All the relevant people that contribute to b
    Ender are part of the system for a long time and i have the feeling that besides the implementation of new features that have been seen somewere else they lack the view from oitside on the whole product. I know that there are people that are fond of the UI and can use it. Of this spftware is just for them please ignore my post. If blender wants to be atractive and fun to use for as many artiats aa possible…. There is just one thing to do in this year. 1. Apply standard UI principles to the whole interface (the rest of the industry, and their users are no idiots that just dont know what is good for them). 2. the outliner is the masterlist of objects… It meeds love, structure and propper rules what things gget handed down and what not. 3. it should always be possible to access the real local, globel and frozen coordinates and edit them. 4. a outliner for materials is needed.

    If you can find a global userfriendly concept for blender this is more than enough for a year.

    Best regards

  50. I am excited to hear about the plans for 2.8 It was the re-design of the interface for 2.5 that really got me interested in Blender. Since then I have learned about the amazing array of tools it has to offer. I am sure that the workflow oriented release will make blender more accessible to new users and even more useful to dedicated users like myself. Good luck in the next stage of blenders evolution,many thanks to Ton and blender institute.

  51. New Blender 2.8 get some new file extension??? Something like .b3d or .blender??

  52. I have a few suggestions have someone fix the carve library so the Boolean modifiers work every time not just sometimes

    also if have herd of the snap-utilities add on i think with a bit of work it would be a great thing to add because precision modeling in blender is nonexistent see

    also blender needs a way to use guides and snaps like sketch-up

    and yes this would benefit animation studios

    and no there is no good cad out there that is free

    so please add

  53. Blender could be easily expanded into a VJ tool, only if some basic option like ui midi mapping and live microphone input in viewport was available. Personally I don’t think it’s that hard to implement.

  54. ” … for making 3d art, games, film and animation! ”

    Please add some focus on VR as well for 2.8 … !

  55. First I want to thank the developers and other contributors to the greatest free software on the planet, and one of the most powerful and interesting softwares even in the commercial world.
    I have been using it since the early 90’s. (2.2 or 2.3?)
    Anyway, because I am not a professional CGI user, I don’t have a workflow per se; instead Blender is sort of my swiss-army knife for all things CGI, sometimes NOT CGI..3d printing, mechanical simulations, image-format and CAD model conversions.

    So from my perspective here are my long-term desires and frustrations:

    1) A ‘Blender App’ repository/store/interface directly within Blender, maintained by BF!
    With the internet, content is king! Blender adroitly addresses the design side but is almost entirely negligent of the asset side. The way Blender now addresses assets is still based on the old way of using the web.
    Like the Android, Windows, and Apple App stores, it would be a one-stop-shop for everything Blender, and access would be integrated directly into Blender; BF would only manage the infrastructure and official assets; everything else would be user contributed.
    It would probably require a dedicated Blender window and download/uploads would be from within Blender, not a web-browser. In fact, the goal would be to eliminate an OS-based web-browser as much as practical; video tutorials for example will probably need to link to the outside world like YouTube. Even then, there are thousands of Blender tuts on YouTube but you have to search from within Youtube, which is not suitable at all considering the billions of other things that get in the way, especially the extreme level of visual noise on their site. A nice little pop-up window in Blender that JUST displays the help video would be the ticket. Canva, a recent cloud-based graphics design startup, uses that approach and I instantly loved it. Instantly!

    I know this might bring up huge server issues though…which is not the fun side of CG.

    This would address the complaint, especially from beginners, about having to search endlessly for help or resources that are Blender-specific.
    It would contain any kind of Blender asset, tutorials, third-party commercial products, wikis. It would have database properties (tags?), not just a folder-tree approach. Otherwise searching for stuff could be a real chore once the resource base gets huge.
    There have been many third-party attempts to provide a consolidated one-stop-shop for Blender stuff (Blendswap is best example I know of), but >> ONLY BF << has the weight of authority to establish itself as the one defacto source for all things Blender.
    To be clear, this would not be a fractured list of URLS pointed to other URLS; it would attempt to provide a consistent structure and interface organizing a huge amount of resources into one location with immediate access directly into Blender.

    I think most people are happy that we don't have to try 5 different internet search engines anymore; let's do the same for searching for Blender resources.

    And it wouldn't be too hard to allow the integration of local resources into the same interface, thus addressing the need to manage local libraries. Saving a model from your current Blender into a library Blend is horribly tedious right now.

    The Cloud is too important today for Blender to ignore. Even applications are moving to browser-based interfaces. Since Blender is too sophisticated for even HTML5, let's bring the Cloud INTO Blender.

    BTW, this is not about asset-management for projects; that is a separate task.

    2) Fake user
    I agree that the fake user needs to go. Too easy to lose your work and not consistent with experience with other softwares. I liked the new recent feature to list unused elements with the 'purge' button, which I started using right away!
    At the very least, set the Fake user to default on..although I would much prefer to actively control deletions.

    3) Give me a customization a whole special window just for me to customize a tool-set. Nothing is faster to learn that a system you build yourself. Find something you like just once and never have to search again. Hotkeys are only good to a degree for me and for little used items I ALWAYS end up looking in the menus and panels.
    For this window only, the user could chose to use icons/text/icons+text display. Shortcuts may be fast but there are hundreds now, some that require me to use two hands! But to keep my hand on the mouse I need a one-click solution.

    4) Micro-tesselation
    I adored it in Kai Kostock's custom build way back in 2.49ish times. And I'm fairly sure it is default by now in every other similar software. It really helps with computing and memory limitation when trying to make deeply detailed models. It is a default approach when designing games (LOD), why not for off-line rendering?

    5) Tabs!
    There are many, many Tab-based designs and I think it is one of the most effective ways to keep an interface uncluttered and still provide visibility to all of a software's details. And keep them horizontal; they are hard to read when sideways, and yes, I can tilt my head, at least a few times. After a while my neck gets sore though and my eyes DO NOT adapt after a while, words don't work that way in our brain. IMHO it is one of the few bad design decisions, considering how recently it was implemented.

    6) Library management
    I don't know where the direction is going for library management. The current scheme is very flexible but VERY uninformative and has no higher-level structure or management tools, at all!

    7) Panoramic set design
    VR and other panoramic designs are becoming increasingly popular. Building a 360-degree stage around a camera, ESPECIALLY outdoors, requires a whole new level of resource optimization. There are almost no tutorials about this; ie tutorials about making a grass field have been useless to me. This might need a solution at the renderer level ie see how Vue and Terragen manage huge stage renders. In those cases you can't even export anything useful to Blender without losing all the rendering advantages of those programs.
    I don't have answers, but the situation won't go away and will need solutions. The game engine can address this but is no help with sophisticated off-line renders.
    -A LOD approach for offline rendering is one idea that I keep wishing for and attempt to make by hand, with some luck.
    -Bigger reliance on procedurals that are not tied to any mesh. Terragen is virtually all procedural and the rendering method is intrinsically LOD, which is why I can't export its advantages into Blender. Terragen has no issue with rendering details from 1mm to 100km in the same render!

    8) Texture Painting
    I was a promoter of PTEX until I realized that it can only work with texture painting and can't use outside art programs for serious artist work. If so, the Texture Paint engine MUST have some high-level paint-program abilities like layers.
    Also, fix the code to allow real-time preview rendering WHILE painting. It occasionally works on my PC but usually does not update the model being rendered for several minutes. When it does work, it is wonderful! Like the first time we used cycles preview window. What seems like an extravagance now will become an expectation later..

    thanks for your consideration

  56. Love this idea!!! I love blender, and think it’s by far the prettiest to look at 3D program.

    I’m hoping that the year off gives the opportunity to really polish everything and make the workflow more intuitive and automatic (fewer clicks for obvious stuff).

    Super excited for this :)

  57. Blender needs to consider its user workflow. Hotkeys are very efficient, but the general layout needs a overhaul. I highly recommend looking to Modo’s latest version to see what innovation in UI can do for development.

    While taking on Blender 2.8, please consider what the most efficient workflow might be. Consider how you can reduce the number of needed screens to work.
    For example, if I want to move keys along the timeline, I need to go to a separate screen to select and move them, when many programs allow that directly through the almost always present timeline. These types of efficiencies can greatly enhance the user experience and speed workflow.

    Also consider accessibility. Is the program laid out and designed well enough that disabled users might be able to use it with just a cursor. Is its layout intuitive enough that people with mental issues or onset of senility could still navigate if given time. These are issues that may affect us all at some point in our lives, so please be considerate of making the program usable for everyone’s, including your own, sake.

  58. Hi, I just wanted to let you know that there’s no need for ‘Blender 101’. I taught a class as an 8th grader in my middle school about using Blender for video game creation and there was a large, interested audience that wasn’t clueless.

  59. Lack of a very good “solid” DWG importer…

  60. Please be compatible with rigify to unity mecanim

  61. i’m sure there’s no end of things people will say blender needs to have. i may even agree with a few of those things but i’m also sure that you have all been doing a fantastic job and will continue to do so.
    keep it up and i’m looking forward to the 2.8 release

  62. Just wanted to say that I really love the way the GUI works. I export to my own xml file and render/write textures automatically and it is really awesome how easy it is to set up your own panels and set properties on scenes, objects etc. No to mention that it automatically sets them when selecting scene/objects etc.

    I, for one, hope they do not change that part of the GUI (personally I am pretty happy the way, the whole GUI works, now).

  63. I wonder if some of the ideas of game jams might not be applied to the development process. The idea is to get groups of people to work on the same project together, and get something working in a short period of time, right? Then to follow up with polishing. This post seems to be saying the polishing stage is where the problems lie, and that stage can only really be done by those implementing the feature.

    So, instead of having a developer go off an write a first implementation, then come back and have a code review, why not try out something akin to a code jam. The idea is to solicit devs that are interested in getting started with Blender, and get them together with someone that has Blender experience.

    They all work on the project, in a short time frame, and each of them now have the responsibility to see it through. The experienced dev then shows them the ropes on the polishing phase, but you now have 2 devs with the knowledge to finish the polishing of the project, not just one. It also means those same devs can help polish a few bugs off the list, if they wish.

    Do this on a monthly basis. you don’t have to start with a new feature, finishing up a feature should be fine. Yes investing the time to pass on that knowledge takes up valuable time, but that is how you grow the developer base.

  64. Major enhancements are always welcomed, but I’d like to see a few development cycles dedicated to improving the current blender.

    There are literally dozens of annoying time consuming workflows that could be improved. There are standard workflows that require me to remember to jump to different tabs, panels, etc. A materials node library that I could create and manage would be a time saver instead of appending from different blend files. Once I create and object that I will reuse in the future, I’d like to save it to an object library in Blender so all I need to do is click on it to add it to my scene.

    Texture painting has improved but it is still lacking in streamlined work flow. I shouldn’t have to jump around so much to tweak brushes. I shouldn’t have to create a blank image in the UV editor to start painting in Cycles–it should all be set up automatically in the background. How about adding layers like GIMP/Photoshop/3D Coat/Kitra, etc.

    Tabs were a good idea, but I have no control over how they are organized. Add on developers are creating their own tabs and my list of tabs is growing to the point I can’t remember what tools is in which tab and some tools show up in every tab. I’d like to create a tab for say “Modelling” and move the tools I use most into that tab. Pinning is nice, but that tool then shows up in every tab. Some control for the user, please.

    I’d like the Blender developers to take some time and evaluate what is already in Blender and make it better, easier, more streamlined, etc.

    Thanks for all the good work you do

  65. Who is working on the Blender 101 project? I do a lot of work with high school students through the project, we’d be very interested in contributing our findings and running some test for you

  66. It’s a good app

  67. Hi friends, please ask the attention of Lords for a very useful tool that is very much needed for through advertising … the “TEXT” tool because I know Blender since version 2:46 … when the updates already started in 2.5 I noticed that this “TEXT” tool never suffered an update.
    I work with production of vignettes for television, I take a lot of use texts … but I have to resort to another program … I mean the fabric of the text, for any application curves to deform or just create a bend in the text, since suffers flaws … to have the use of a modifier … which in my opinion is much ride to just text … I would suggest an improvement in the fabric of the text … to deformation among other applications.
    please .

  68. BLENDER lacks foundamental tools for cad modelling, they are absolutley mandatory to make BLENDER a real multiporpouse software.

    Building simple item we use everyday like chair, table or architectural structures, in a fast and quite accurately way the main software is Sketchup.

    These items needs measures and precision, so BLENDER needs tools to work in this way.

  69. Something I’d really like to see is giving the user the ability to write their own GLSL shaders and have them be editable in the properties panel and render in the 3D View. I use Blender for video game assets, and it would be really helpful to be able to use the game’s shaders in blender.

    It would also allow me to write special purpose shaders, such as an all-in-one stencil map. Blender’s current stencil map functionality doesn’t export well.

  70. I think there are two solutions that would solve 90% of peoples issues with blender.
    1 – If we could use a node chart to script our own add on’s , all the silly little things that some people love , others don’t , wouldn’t be an issue anymore , I can’t code but I understand the logic , so I could put a bunch of predefined nodes in an order that would enable me to create an add on that adds a function I need.

    2 – A node chart that enables you to choose what parts of blender you want to use that can execute a self build of it afterwards – Uber Ui custom design tool. If I wanted to teach you how to sculpt I need only access about 15% of blenders user interface , and any other specialised area is usually the same , leaving the rest just get in the way. Also in terms of The Redcurrant project this could be useful as , If departments were only using a portion of the blender tool at a time they could get merged into the Directors “Master build” in a non destructive manor.

    I know its easy to say something stupid like “solve the problem with nodes” and there is a huge amount of work that would have to go into it. But the future would be so much more adaptable and development would happen in and of itself because people could work on specific areas as a lot more manageable sized chunks of data.

    I love blender and appreciate all the hard work you few guys (& gals?) do to serve the huge end users you have. Unfortunately I don’t have the skill set to help you in the ways I would love to be able to , If you empower me with simplified tools that can benefit both of us its a win win situation.

    P.s thanks for the snap to absolute grid alignment , that doubled my work flow speed on hard surface modelling. Awesome. Good Luck with The Gooseberry release , Its lookin good.

  71. Long term user, first time poster here.

    I think this is a great idea. You guys have done an outstanding job of bugfixes, blender is very stable compared to most 3d suites – I havn’t tried maya in a while but i think blender could be the MOST stable.

    I think we can handle a bumpy ride for a while in order to focus on the bigger picture – I’m excited for 2.8.

    Where’s the first port of call for people that want to get involved and help out beyond money donations?

  72. Moving blender to Bullet Physics 3.0 based on OpenCL should be done early on.

    Getting that flushed out early would help across all aspects of physics work.

    • A critical look at CPU muti-threading would also be helpful for simulation features. Especially given processor direction in 2016 and beyond for Intel with up to 22 cores and 44 threads.

  73. Hi guys!
    I hope that developers will not be angry at me, I want make blender better only.

    Fake user
    Fake user is very fake idea. Blender is deleting our work if we forget to press “F” button.

    Change fake user system. We want the possibility to delete our actions, materials… by own hand.
    Make automatic fake user without user assistance and join possibility to delete files by easy way.
    ( x buton on each action or material, or right click menu – delete)

    10 000 buttons ( NRK) and we are hunting where my helish button is. Very large buttons for everything

    icons for everything like” Icon_Tools” addon

    10 ( or 11?) icons for everything

    more icons or… possibility to bring own icons (svg format).


    everything above is under development,
    we must be patient. ( it is not irony)

    2D planar manipulator like MAYA, Max…


    rotate manipulator is out of industry standards

    rotate manipulator like Maya, Max… please look to Maya and try. It is like track ball.
    Blender novices are frustrating from it. Character animators also…
    addon with “normal” (like maya) rotation functionality

    right click select

    PLyczkowski addon – rSelection

    “use nodes” check box.
    Developers are afraid that users want use compositing window for party sex? :-)

    bye bye “use nodes” check box

    “Save” with ( frustrating) confirmation.
    We want it without confirmation.
    It is like : are you sure? Are you really sure? Are you really really sure?
    YES! We want save our file! Not answer for Blender silly questions!

    Stop this or check box in Preferences. ( ask or not ask)

    addon – Simple Carnival Save File No Confirmation
    But it works with ctrl+S only. If you give save button to viewport header or toolbar, blender ask you again.

    After EXIT must Blender save everythig what we are using in file – our material, textures, actions ( every 3d programs are doing

    But if we forget save our textures separately, where is big warning notes? ( Our non fake user actions are going to hell also).

    Maybe developers will be thinking about it.

    save preset ( render, cloth) save only several basic parameters, not all.

    addon “multi-preset” ( put it to Blender core addons)
    but it is not work with freestyle.

    something is twice.
    For instance node name ( File Output) has name in the” strack of subpaths” and under stack in “File Subpath” also.
    If you can rename it, you can do it in the stack and in the subpath also.

    bye, bye …
    Layer management – 20 layers only.
    Bigger projects needs more layers. Bone Layer Management has 32 layers,
    Rigify Layer Names has 28 layers…

    more layers ( developers)
    SuperGrouper (Magic layer also?)

    Super Grouper helps you divide layers to small pieces, ( sublayers)
    But it not works in render. Only base 20 layers are respected for RenderLayer.

    “normal” animation layers like Maya, Motion Builder, Max….


    Open – Save Browser is non – standard

    addon or you must get used. We have not another.
    It is not very big problem, it is very big Browser only.

    Outline – only alphabetize.
    But we want to arrange our objects by custom way.

    Look to Maya…

    modifiers at the bottom of the stack are really at the top.
    Textures at the bottom of the stack are really at the top.
    Nodes – top layer is backgroud, bottom layer is foreground.
    sound is above of video layer.

    stop the collaboration with australian developers. :-)
    or handstand on the head

    coordinate system – Y is depth.
    But every world is saying about z-depth ( All industry and Blender also)
    z- depth pass, z for depth of field, z – composition
    2D image has X, Y coordinate
    3D image has X, Y, Z coordinate, Z is depth ( but not in Blender)

    change coordinate system to Z as depth.

    Grouping in Blender is more like a tagging. Real grouping of objects as known from other software is still missing.
    Yes Group system in Blender is TAG system.

    Normal group system via “Parent to empty” addon ( industry standard system)

    After creating rig we get a lot of widgets files to Outiner. But Outliner is not trashcan.
    Pitchipoy Human Meta Rig is doing the same result.

    addon “Parent to empty” and put all WGTs under empty by hand.
    or put to empty automatically ( corrected Rigify script)

    Rigify Rig Layers do not work preperly. If you attach new rig from another file, Rig Layers are missing.
    Or when the character is linked into a new file, this script just got lost.

    Correct Rigify script ( they know about it some years)
    new addon for Rig Layers

    The timeline is pretty much useless (simplecarnival)

    nudge keys
    move keys
    delete keys
    please look to Maya

    lock buttons at Location Rotation and Scale (Transform panel)
    It close widgets in viewport only !
    Loc – rot – scale is possible via numbers, keyframes are possible, you can move keyframes in Graph editor.

    “lock” must lock everything

    Automatically save the rendered image

    addon “render_auto_save”

    Export presets as txt file. Actions or material or compositor presets.
    It is not very practical to save to blend file. You must close opened file. Open new file, append what you want ( material) and save

    and close. It is easy and quick way?

    industry standard – save and load ( export, import ) to TXT file ( or another extension).

    Mominou [ADDON] Shape Keys Presets has export import presets to txt file. It is very handy and quick, bettes as very rigid Blender


    Oscurart ( Oscurart Tools) has export to his own file. But not to versions. Only export Groups and Import Groups. ( It is very helpful for

    weight painting). If you want another version, you must go to project folder and rename your file by hand because next export resave

    your base Group file.


  74. Yes this was very helpful

  75. a more refined vision… (excuse the extra post)

    a feature/fix/wish list page, with a tied in $$$ amount set by the devs after suggestion moderator approved. People pledge money, held in escrow (refundable at any time so long as the pledge has not yet been met). Upon reaching the stated $$$ goal, all escrow held money is secured as a non-refundable asset and the feature goes into development.

  76. I’m thinking getting the entire Blender community together, to give their realistic wishlist for tweaking existing stuffs (fixing/improving/altering interface/speed of use -of things that already exist (tweaks and edits and slight but intelligent re-designs) and if peoples personal wants were on that list, as stretch goals, they would donate to make it happen.

    People might get somewhat excited about the idea of new physics / BGE / Rendering etc., but they aren’t going to donate in advance for that before it’s done, and probably not after either.

  77. TON! Dude, this is all you need to do. Turn this user created thread, into a massive fund-raising kickstarter.

    Lots of money to be had for the foundation, and this is the stuff people actually want. Love you guys!

  78. I agree with the pause button. Blender has been on a sprint release mode for a few years now and it is time to take a breath.

    My suggestions:
    Move to to Python 3.X

    Break backwards compatibility.

    Refocus Game Engine to be less game and more focused on allowing users to create real time simulations for use in sales and demonstration situations.

    Create two additional interfaces: K-3 and 4-8 rough grade levels. Strip things down, disable functionality appropriately. This will greatly help with teachers being able to integrate Blender as learning opportunities and curriculum. This also is why I suggest refocus the game engine on simulation so teachers can use it easier for teaching physics, math etc.

    Move this to an eighteen month project. Make 2.77, 2.78, 2.79 exclusively bug fixing. No new functionality or features. Put four months between each of those releases so that puts you a year out and by then, 2.79 should be the best release of Blender.

    While the above is being done, continue working on the 2.8 release and relevant bug fixes done on those pulled into the 2.8 branch.

    I would vote that the 2.8 branch should have at maximum two major new additions. What those are is beyond my skill level to contribute much to but keep it obtainable. The rest of the 2.8 focus should be on enhancements to existing features and polish and bug fixes.

  79. Thanks for the positive feedback and enthusiasm. Keep the ideas coming, but don’t forget about suggestions for how to organize and realize it all. You don’t get workflow topics solved by a feature request list. It’s an active design process that goes along with development.

    • this might be not concrete enough, but I’d like to suggest a perspective. I’ve noticed beloved programming languages usually are the ones that the user claims to use as a tool for thought, and only then to create. Lisp would be the main example.
      Maybe a good guiding principle for a 3D application would be to enable the user to think about what he/she’s doing with the application as a mental extension, starting from a nebulous idea.
      I’m talking about seeing creation software as mental software too, even primarily. It’s an angle to approach tool development and workflow.

      I also want to suggest focusing the development in very few subjects. I’d say workflow and modeling. It would be nice if people could answer about why they use blender: “It’s the best modeler in the planet”. I think Modo is a little bit better. Workflow and modelling is the best starting point to improve and persist, sacrificing other subjects if necessary, and proceeding from that base. Maybe with a Krita-like style campaign.

  80. nice plans!
    personaly i think the release cycles are to short. every 6 Month would be a better choice. maybe. :)
    chears, Tobi

  81. For me Blender is a dream software. You can do so many things. Again I would like to thanks those who work so hard on this.

    Like many others above I agree that the idea to halt releases and focus on a greater picture is the right thing to do. The actual release is pretty awesome.

    I would like to speak for the Blender game engine. Again many in a way “despised”it. But many others are really using it fully, even overcoming the license limits.
    I do not understand fully technical stuff but the above plan by “Ton” seems really promising. But I would really point out a feature thats missing. Proper export to Mac, Android and if possible IOS. I do not know how difficult it is to implement these but BGE users really need this.

  82. Wow, this turned out to a massive wishlist very quickly.
    I would love to see:
    Personally, I dont need a new release every ~3 months, so I say take all time you need to make big things happen. But post more code blogs in this time – I love these insights!

  83. I am an avid user of Blender. One of my ‘hobby’ is teach Blender to young kids (age 8-16). The non-standard Blender UI is frustrating for beginners. As an example, often the kids have to *fight* with the unusable layout because of accidental creation of new views.

    If you think about it, it is a shame that a fraction of Blender related comments on Internet have something to do with the UI! The fact that some comments start with a brief statement like ‘Blender UI is not bad’ and continue with real issues says something about the ‘problem’ of Blender UI. The less-than-friendly UI is a distraction to the healthy growth of Blender software and its community.

    Just like many other comments sidetracked by UI, what I REALLY want to say is an improved (fundamental change?) game engine design/workflow would help kids motivated about modeling objects and putting them in action!

    Thanks to all Blender developers. Please make Blender a fun sandbox for kids to learn modeling and (visual and textual) coding! A build-in 2-way hook to external Python IDE’s (or at least notepad++) is highly desirable! An ambiguous goal would be to allow Blender to communicate with visual coding IDE such as Scratch (which understands Python).

    I believe Blender has a potential to be a fun and educational platform for kids to learn computer modeling, animation and coding! An approach is to provide a kid-friendly version of menu (simplify the menu, shortcut common operations) and tabs of panels of pre-build shapes, materials, textures, sounds and other assets.

  84. A more general observation:

    Comments in this section seem to be of one of several streams:
    1) documentation; 2) interface; 3) extension completions of
    physical exotics.

    Documentation needs to be in several streams: tutor, animator, special
    purpose — as well as both within blender and hypertext in-depth
    manuals also shipping as part of the package (i.e. built as the project

    Interface is a big one. I’ve been using Blender since 1.5x and it has
    always been a country unto itself… so many areas of (non)default
    settings, unusual scaling quantities etc. I liken it to being its own
    OS sitting on top of Kwin or Windows (while working =smoothly= between
    the two ports). Why not leave hooks for making it more so? Like File
    Menu links to a configurable editor (e.g. joe or gimp) for outside
    Blender graphics use. Or a few animation specific utilities like file
    renaming/renumbering or version condensation.

    Most of such operations could be handled by simple calls to the shell — the point being to make Blender more like its own windowing
    system and less like an add-on to some other windows system. And while
    doing this an /etc/blender.config-like file to allow users to set up
    the way a particular session of Blender was to behave through a simple
    (thus complete, backupable) customization in the character editor of
    their choice.

    Physical “exotics” like fur could also be left aside to devote more
    resources to the session actually being developed. To the exotics I would add my own pseudo-random lights for strobelight/fireworks/LED/
    candlelight sessions (compensated for the 24/25 fps) or for the
    wave exotic to enable deepsea/oceanside/stream/pond wave simulation.
    Or plants with a biologically accurate stem/leaf/flower system to be
    generated once or twice as near-primatives.

    4) gaming… not forgotten — just outside my own field of direct Blender experience ;)

  85. Libraries of properly packaged simple stock primatives would be a plus.

    Even if only one or two were to ship w/ 2.8 it would clean up their eventual use or reuse — especially by new users wanting to try out animation without getting caught up right away with model-making.

    “Properly packaged” is key here: objects descriptively named;
    preferably with likely named parent empties and scaled to real-world
    component object physical dimensions.

    Libraries of simple thematic primatives (like kitchen-lib.blend or
    carpenter-toolbox-lib.blend) would both teach how to construct
    reusable libraries and enable early animation. Not all would have to
    ship with b2.8… pointers to online stock libraries in each files
    documentation would cover that.

  86. A great idea. Take some time to refocus and especially grant the developers some time to get re-engaged with the project. After all, they are pretty much giving up all their free time for this project, so they might as well have some fun doing so.
    Concerning model reuseablility: Though SketchUp is only good, if you want to post stuff on GoogleEarth and though its Render Engine is nearly non-existant, it does have one thing I really cherished and later missed, when I moved to blender: Whenever you have created something you want to reuse in a different project, you can easily save it to a library of reuseables. After about two years of frequent SU-use, I had quite a decent library of my own and could work on projects with a lot more speed (and a lot more motivation as well, as I knew that every detail I added to my model components now, I would benefit from later). Such a feature would be great in blender, as that would save the time which would be consumed otherwise by creating a decent collection of primitives. Just a thought.

  87. Tabs are awesome, probably one of Blender’s best features. However tab legibility on the left hand side of the UI really sucks unless you are the one person in a billion born with a sideways head. Any chance amongst the more complicated stuff that you could do something as simple as turn the tab headings the correct way, ie we read left to right in Western culture and our heads are upright with two lateral eyes. In terms of useability it’s the equivalent of putting a car steering wheel facing the passenger instead of straight on how a driver needs it. Even if you limit the tab shortcodes to four or six letters kind of like they do it in softimage or fusion, it would still be better thsn the inevitable one letter to find a tab situation that happens now. How many Blender users enable a bunch of addons only to then have to hunt through multiple single letter tabs just to find the Misc tab. And good luck with anything involving a lot of subcategories like Animation nodes or Sverchock. Please if you fix on thing only – make tabs legible as in how human anatomy works. You could also revisit Andrew Price’s suggestions. He was spot on about useability and Blender does need a redesign regarding using the lateral, not horizontal screen space. And he was right that Blender needs a ribbon.

  88. Hello,
    I am not an expert in 3D, but I use Blender for hobby.
    I must confess that I am lost by the current development strategy.
    I see every two months little steps but I cannot understand what is the way which is followed on the long term. The roadmap, once described by Ton was a great, but I cannot where we are versus it.
    I also think that Blender is becoming quite complex with new features that are sometimes difficult to understand.

    What can be done for the future ?
    I think that Blender 101 is a good suggestion to simplify Blender approach.
    I Would appreciate more tutorials in Blender Cloud (i.e Manga type movie of 2mn with complete workflow, new animation movies, sculpting … with lot of details…)
    I think that the Blender fondation must focus more on trainings, by delivering every 2 or 3 months exciting tutorials. The explanations given are mainly for experts.
    It would be interesting to have more standard tools (Ex a basic library of Material with easy selection when downloading the new version, a standard HDR image,…).
    As hobbyist, I cannot understand all the internal debates of display that are of no use for the average user (Why do we have by default the Python info in the help when going with the mouse over the tools.
    Could the Blender fondation release a reference manual wi9th version 2.8?

    So I would agree with a stop in the frequency of the release, but regular info (monthly ?) about progress versus the roadmap decided.I would like to see a focus much more on the users, more trainings.

    I would like to thanks all the developpers for their involvement and the Blender fondation for this great software

  89. I’m really interested in seeing what kind of changes happen to the game engine. For me this is the funnest feature of blender. I’m really excited to see it become an integrated part of blender’s development.

  90. Rigging system needs overhaul. I can model a character fast and get it exactly how I want it to look/work, but rigging is a nightmare. We need a system that is more physically based vertex mapping deforms doesn’t work in any kind of realistic way. Bones effect surface normals of real bodies much differently during twisting and bending. It doesn’t have to be perfect, but it should be more based in physics and less based in puppetry. I’m very happy with the direction of blender other than my rigging issues, and it should be noted that every other rigging system I have seen is also fairly cruddy in the same ways.

  91. This one year interval of concerted, eyes-to-the-horizon development is a great idea. The incremental updates are important for feature implementation and bug fixes, but Blender will always require long-term, self-reflective retrenchments to philosophical ambition, and Ton needs space to work in that mindset now. Go captain!

    I am totally on board with the backward incompatibility consideration. It’s not a huge deal.

  92. I deffinetley think they need to work on the character physics for the game engine, make it work, along with adding to it.

  93. Working in Blender a lot. The goal is great, but execution is key.

    Two main concerns:

    1. – Potential to become a never-ending story.

    2. – Blender’s UI is not bad. Awful for a beginner, but once you memorize everything – in some places it becomes one of fastest in existence. However, comparing this perfection to holes there are, it becomes obvious, that perfection is there not because of stunning UI expertise, but rather because this extremely tedious task was partially offloaded to long natural selection, those in desperate need, passion, plus a bit of luck.
    UI is all the tiny details that don’t get noticed. And if what we see at first (marketing?) is the driver, there is big chance to squish the invisible.

    Invention is great, but take care :)

  94. love qt, but it is not relevant for make a good software or bad software.

  95. Hey guys, there are some stuff i miss in blender:

    1º Virtual Frame Buffer for cycles.
    2º Displacement. need it a lot! Microsurface displacement would be great
    3º Cycles GPU out-of-core features
    4º White Balance/ Highlight compression

  96. This is great news. i would love it so much if these improvements to come improve cycles render engine and make the UI much more easier for beginners.

  97. could exist in the future similar technology to pixols in Zbrush?

  98. Hey Ton, pixar lately released update for opensubdiv, is there is any chance to see it in blender before you guys disappear?

    • OpenSubdiv is a target for the next release.

  99. I use Blender for architectural renderings (mainly interiors) so I’m not expert in coding, game engine, animations, NPR, etc … and all I need is a fast, versatile and photorealistic rendering engine.
    As for modeling Blender is extraordinary, but for rendering I am forced to use Maxwell Render or LuxRender. I really like Cycles, but unfortunately it misses some fundamental features without which I believe it is unsuitable for professional use:
    – a “multilight system” (so that illumination can be modified without rendering again)
    – the possibility of pausing, saving and resuming rendering.
    I understand that those features are probably very complicated and need a lot of time to be implemented, but I never read anything about it (neither among developer projects for the future nor among users requests). Blender 2.8 could be the right occasion! Bye

  100. and by the way please give it some love to the render speed of the vse, yea not a cycles request ^^, but on long movies(not so long max length supported its 4 hours) it takes ages to render a simple edit

  101. Go blendaaa
    a good dependency graph, a nice viewport, and some ui simplification. . . actually i dont care that much about the ui, i feel it nice as it is, i mainly use keboard shortcuts and spacebar menu operator so my viewport its almost always clean, but many of the ui discussions can be settle by adding on the experimental tab menu under the render button a two more options, 101 or basic that will show a simplified version of the ui with many presets and nothing more and a second advanced option, that will show blender as is it now with all the buttons and options to twick and change.
    well at least i know have time to do some tutorials knowing that its not going to change or improve in two months ^- ^
    thanks coders !!

  102. What about cycles network tile rendering, sharing CPU and/or GPU over loka LAN. It’s getting quite essential!

  103. Way off topic, and I think it makes sense to stage minor/major build cycles on 6 month/18 month increments but not having much hands-on with Blender and rigging I just had to make an observation:

    Is the Computer Graphics/Modeling world claustrophobic to teaching Kinematics from a Mechanical Engineering perspective and then creating “riggings” to show how 3 degrees of Freedom (3DOF) from 1->2->3->4 bar linkages work in Robotics and the physical world should match in the CGI world.

    Rigging/Rigifying really should just be called Applied Kinematics.

    Kinematics has unfortunately been removed from a lot of ABET accredited ME Programs and that’s a disaster. Computer Graphics Modeling can go a long way in showing its importance working with FEA/CFD and Machine Design simulations.

    If you can drag n’ drop applied Kinematic physical objects that preset DOF to your models you could make Blender a very strong product in a lot of industries.

  104. When I try to imagine scenarios for what would happen if the Blender core devs, freed from such concerns as backwards compatibility, all went into seclusion for a year and focused on making a new, sleeker, more powerful blender: pretty much all the possibilities are amazing.

    I think Ton and crew will always do right by blender. :)

  105. First of all, we must to be thankful to all developers that dedicate hours and hours to improve Blender.

    About strategy, I think that we must to have to “battle fronts”. One, for conquer an army of new Blender users. For that, the UI must be improved, in order to become very, very friendly.

    The second front must be to keep experient users with us, with Blender. For this, a good UI is important too. And also improve Blender even more in several aspects:
    1) massive fluid simulation
    2) atmospheric tools (to create clouds easily, for example)
    3) muscular rigging, to get realism on characters

    I totally agree with MOONCAINE about the UI. And I add some more questions: why Blender don’t adopt by default the LMB selection? Someone that install Blender for the first time and can not do a simple action like select the initial cube probably will give up and try another software.

    For a software with so many resources, so powerful, tabs are very important to improve the UI.

    Blender has, nowadays, some context menus that appears with the RMB. For example, in the outliner. I love it. It’s practical. It’s friendly. We need a lot more context menus.

    And last but not least, a professional software MUST have a warning about “Save your file before closing?”. Ok, we can activate something in the User Preferences. But It must be there by default. And it doesn’t work if you open a new file before save your work. All software has this feature since the eighties.

    Indeed, to reach these goals, the developers must to work without the pressure of time.

  106. I use blender for my 3D printer and it works great. For the 3D printer world things we need, mesh closer. No holes allowed. and a pattern fill feature. most slicers do this but you are lucky to have two choices of design, and designing one of you own would be revolutionary, maybe not necessary, but if I want skulls and ghost over square blocks so be it. Basically solidify to make the hollow and the walls a printable thickness and a repeatable fill pattern of desirable thickness. That would do nicely, Thanks. So far love your work, and the YouTube overviews of the new features. If you what to learn this program its all on YouTube in almost all languages. A few credible links on your sight might be helpful to some. Unbelievable job so far, Thanks again.

  107. This is indeed great news.

    I remember being hugely excited by all the developments hinted at in — it’s already July, so let’s get cracking!

    Personally I would like to see a massively improved sculpting and texturing workflow, especially the touted Ptex support. Let’s make things much more artist-friendly, allowing us to intuitively and easily paint and textures with useful brush presets, masks, and not having to press sixteen buttons just to get a working brush alpha. Maybe even a brush node system would be possible (or perhaps bring back the now-neglected dedicated texture nodes)? Something to allow a more generative, non-destructive way of applying varied textures and detail without instead having to rely on third party tools like Ddo or Photoshop, or a huge and perhaps unoptimised library of self-gathered textures. Don’t make me struggle with my own disorderliness, help me out a little!

    I also agree that we need much more user-friendly presets, asset management and project structure. And I agree that the whole “fake user” thing is unintuitive and confusing, especially to new users (at the very least you could replace the F with an icon). And definitely better layer management. Let’s not stay shackled to the peculiarities of legacy Blender forever. Yes, it works, so maybe we’re reluctant to ‘break’ it, but does it work as well as it could? The principle of least astonishment should always apply!

    As for little niggles, why is there no blur node for cycles materials? Or a proper map value node? We need all kinds of kludges to get around such things. Better support for node groups too; add switches, dropdowns, or presets; let us make our own custom node groups that remain as simple and intuitive to use as the default nodes.

    I realise that everyone suggesting a million little improvements and changes isn’t necessarily the most productive way forward, because by trying to cater to it all maybe less gets accomplished. But I think it signifies what a passionate user base Blender really has, and how grateful we are for the opportunity to contribute in even just the smallest way. But I think with some proper reflection and direction, Blender’s future is assured.

  108. Hello blender developers,

    this is my first comment ever. Let me explain my experiences: I use blender since 18 months, starting from absolutely 0. I have no experiences with other 3 D programs but a lot of programming history. My desire is to visualize some abstract mathematical structures. So far I have used modelling, cycles, materials and python programming for mathematics and panel design; no animations.

    I completely agree to the comments of Mooncain.

    Blender is very powerfull, the amount of existing features is overwhelming, but it is really hard to learn, even if you use all kind of help: books, videos and internet. Beginners need a very high tolerance to frustration. For example I remember to have spent a whole rainy sunday just to produce a debug output file for my python program!

    Many times I thought to have found a bug, but it was always my fault!

    Although each release is announced to have many bugfixes I am pleased about the stability of software!

    I think that for the average, non professional user currently the focus of development must be to make the use SIMPLER. Therefor a creative pause to make this possible seems to me a very good idea.

    I would be glad to live 1 year without new realease, waiting for a big step forward to simplification.

  109. Yes please!
    If there’s one thing that would make Blender easier to use it will be an updated start screen with news, featured artwork, tutorials and more. Blender should also be able to auto-update, or at least inform and guide you when an update is available, keeping all your settings and such.
    And PLEASE, have dev builds for 2.8 when available! I would love to try it out…

  110. Some improvement in video editor?

  111. I just want to say thank you for all that!

  112. Blender lacks “advanced Crowd simulation”. but before, more speed is needed seriously:
    – official AVX2 build
    – more SIMD optimizations (how about using ISPC? )
    – GPGPU everywhere

    • Perhaps “crowd” or AI agent simulation could come from integrating GE into blender more. I really hope they don’t drop support for this with navigation meshes as blender is already totally unique for having a GE built in (with the ability to export runtime executables).

      After all, games are why GPU’s sell to the masses ever since the VooDoo days. The game engine as part of blender internal could be Blender’s secret weapon with all of it’s pre-set logic bricks, it could be really fun making animations with sensors, keyboard/gamepad/razer controllers and AI to animate and also render from that.

      BGE is isn’t unreal, but unreal/unity/cryengine won’t export your ‘crowds’ for offline rendering either at the moment, this could be genius.

    • DX12/Vlukan support for multi GPU rendering would be awesome, because it allows vram stacking. I already like the idea of bringing PBR rendering already, big hopes they make it MDL compliant so we could export materials for UE4 or Unity.

  113. I think that Blender just needs Alembic import and export to be piped into any film pipeline, either for taking work from Blender or taking assets to render in Cycles, Having FBX is great, but Alembic is what we’re all using, it would be amazing to have it.

    • +1 This is really important !
      I know a studio who is waiting for this, Right now, they have to use the octane plugin to export in alembic. O_o

      We could work with Guerilla, nuke and all other apps.
      I hope this will come as soon as possible !

    • Definitely +1 for this, this is major thing. Also I do agree with some notes Light BWK said, more layers, some kind of folders in Outliner (some redesign probably), project asset management.

  114. Great initiative! And I think “floating” perfectly describes our current situation.

    This trooper is energized for 2.8 and can focus on:
    – viewport, OpenGL, GLSL performance
    – GLSL authoring (shaders as content)
    – platform specific/GHOST work
    – anything involving Blender –> Unreal Engine

    “C++11 everywhere” in 3.0? :D

  115. This for Blender 3.0
    – Relink Assets when link is broken, Blender Aid is not the solution for this.
    – Spline IK with only one rig system.


  116. current blender version, the UI is enough great, at least I thinks this, interaction is good, very human nature, have blender have self style, much easier to operate than Autodesk. all nice interaction, the blender not only blender, and also an art, you know, the modeling, the operation, the function, all like magic, it’s amazing in this software which less than 100M. it’s very cool,and hope more nice on 2.8.

  117. I’m hope new Blender can add something focus on FX, e.g. physics simulation, all works in nodes,stronger particle, fracture modifier, much stronger simulation on smoke and fluid, and wish cycles get more optimization, more quickly and less noise, and the pbr viewport is great, sorry my bad English :)

  118. Please focus on a few basics that keep new users from adopting blender. Here are obstacles I keep seeing, version after version. I see that LIGHT BWK and PGT touched on some of the same pain points.

    1. When a user touches a number key, stuff disappears. You and I know that he just toggled a layer’s visibility, but he doesn’t know it, and blender doesn’t make it obvious. When he tries Undo, it doesn’t undo, and his stuff is still gone. Look at the Layer Management add-on, please. Layers deserve a panel. Hiding a layer should probably be undo-able.

    That’s the single biggest obstacle. A lot of users download it, try it, can’t find the stuff they just made, and quite reasonably look elsewhere for an app that respects their time.

    2. Blender sometimes secretly deletes content the user created. He creates a material for his cube, but later, makes another. He saves. Tomorrow, both materials should still be in that file, but they aren’t. The whole “fake user” idea looks like a kludge that only insiders could possibly know about. The button doesn’t even say what it does, just a letter F. And that’s not the only place where blender assumes you’d rather have a small file, instead of the art you just spent hours creating.

    Don’t delete by default. SAVE by default, whether it’s a material, a brush, or a color. Here’s an idea: the Outliner pane should have an “unused” section that lets me find my old material out and reapply it to my cube. The material it replaces doesn’t get deleted, because *I* didn’t delete it. It goes to the “unused” folder.

    3. Speaking of color: it’s an art app. Why no color palettes? This is a must. The currently included add-on (as of Monday, July 20, 2015) is a start, but needs improvement. Speak the same color languages as the users you want to please. If you only want math experts to use blender, don’t change a thing, but if you want artists to like it, give them some familiar color input choices they know how to use. Usually, I want to see the same RGB numbers I see in GIMP or Illustrator, for example.

    4. “Reset to Default Value” seldom does. It usually resets to nonsensical 0 or a value that would never be used. Either implement this properly (take responsibility for the default value, pick one, and ensure it gets reset properly) or remove it.

    5. Images: it’s a visual art app, but dealing with images is very un-visual. The interface for image editing needs basics that users (rightly) take for granted. It’s hard to preview, load, rename, move, or replace images, or even see 2 side by side… You should have these things by now. Users should see something like the tree view in the Outliner pane, maybe even folders.

    If a feature requires an image (like Texture Painting), make one by default if one’s needed! Don’t leave the user guessing why he sees no paint strokes. And don’t delete the default image when he saves.

    6. You should have an option to use the computer OS’s default file browser. Your file browser has some cool features, but it lacks much that users have come to expect, like clicking a header to sort files. It doesn’t remember *anything*. Every time I open a file, I have to click the Date icon again to see files sorted that way… every time. Not helpful. It doesn’t remember my layout for the awesome shortcuts on the left, but instead always list System shortcuts first, top of the list, the least useful shortcuts of all. It doesn’t do anything for me that my OS doesn’t already do, except for the nice “P” shortcut and the “+” shortcut. I’d rather show new users an Open or Save dialog that they already understand.

    These same issues are wearing down long-time users, too. Too much time spent working around these probs, or just clicking the same thing again & again because blender won’t remember that I was looking for the latest file last time, not the one whose name begins with A.

    The Outliner pane is one of blender’s best additions, finally making it easier to visually see what’s in my project. Develop it, please! Make it more awesome without making it weird. It’s OK if you make it look like Windows Explorer; at least we know how to use that.



    • Excellent suggestions! Hopefully a year of polish will make things easier on artists streamlining and speeding up mundane needlessly complicated tasks.

      I love blender, sometimes it just leaves me confused, lol.

  119. Hello Blender members ! ;-)

    After GooseBerry’s project, it will be a game project ?
    If yes, can you make, game’s nodes (with physique’s nodes) like unreal Engine 4… ?

  120. PBR would be really, really useful for creating assets for games, really miss it now that i’m doing a unity related project.

  121. I really like the idea of taking the time and refocusing. There are plenty of features in the current code base to keep artists busy and there are definitely some strange moments of cruft left as well that need to be re-thought.

    The current GE is a prime example of code that was added all at once to Blender, but isn’t really integrated or consistent with the 2.5 code. The same with the VSE vs. the Clip Editor. Those two can and should be brought together. Not to mention the workflow for compositing with the VSE. While each of the main components of Blender have been getting love, there’s definitely a need to rethinking how they all work together, and making the codebase consistent as well.

    I personally think it’s also worth making 2.8 a breaking change, ie. updating an importer for earlier versions, but backwards compatibility not required.

    There is one other aspect to this effort I personally think is important and it has to do with the with the mindset of the development in general. I get the feeling that a lot of the core development revolves around integrating wonderful work that becomes available from open source projects like OpenVDB or Alembic.

    What I would like to see is a focus on not just matching programs like Maya, Houdini, and ZBrush but innovating and _surpassing_ commercial counterparts. Making Blender a must have, go to tool that EVERY studio or aspiring film maker should be using. Because it’s simply the best. This 2.8 target seems like a real opportunity to lay the ground work for such development, and I’m very excited to see what comes out of it!

  122. So we won’t have help text near a command that says that it will execute the said command and in exchange we’ll get real help text that says what this button/option does? Strange times…

  123. Blender proposes a very good Python API that allows users to tweak most of problems by themselves. It is better if devs keep working on more complex things that require a deep understanding of the blender’s code and that would have a larger and generic utility for the community whatever we do with Blender. Then we can all tweak blender to make it more like C4D (animation nodes addon) or others softwares as soon as the we can access everything with python. Expecting a good out-of-the-box 3D software for everyone is nonsense, but expecting a good functional set of simple and efficient tools that you can combine with python to suit your own workflow is doable.

    Right now we have seen how much the community can support Blender. So how Blender can bring new members that would financially support Blender? Those members are potentially CG studios, small or medium, that are already looking at Blender but they usually miss some stuff like a modern viewport and pipeline integration. I am glad to see that it will be done.
    Thank you very much Blender!

  124. yes, this is exciting!
    Take all your time Blender Foundation ;)

  125. 1º congratulations to all staff. Leveraging that will make changes only for next year,are some tips to think: indie game producers with annual income above $ 20,000 or $ 40,000, go to contribute 5% of the products sold. A means to raise money, just do not know if the license allows it.The money could even at first,to hire some external programmer, to develop new features GE. After, resources for the entire program. The indie productions, increasingly grows are sold increasingly in steam. While some indie productions, sell their work, BLENDER foundation can ensure a means to raise funds. GE has great potential, just needs to be improved (RTT MAP with colors to detect properties, Particle systems, selection line “RTS” for multiple objects, incorporation of multiplayer addon, ocean incorporation , position counter to racing games “based on the objects in the scene or the property contained in objects”) these essences things you have on other GE. All win, the indie producers to publish their work and the foundation to secure funds.

  126. Fantastic plan! The difficulty will be to keep people in the feeling that progress is done, especially for those missing the posts about it. They might interprete it as a slowing or stopping of development, i think a better report structure than the develeopment weekly reports might be a good idea. (more visuals, more insights)

  127. Particle nodes is what I’m really excited about, is there a branch to check out the progress so far?

  128. YAY!! waiting for PBR

  129. As a professional blender artist I feel there are two main points which need work: General speed of everything and a streamlined workflow. We have some fantastic tools at our disposal, but at times I feel like things are slower and the interface clunkier than they should be. 75% of my time is spent compositing but with a few streamlined options my compositing time could be reduced considerable amounts. Some pretty important options like grabbing objects in viewport to move them or resize them just aren’t there and should be there. Time spent tweaking % or pixels is just huge.. HAVING to go into movie edit mode to mask things, each separately, bring them in and review the scale of everything if what I’m compositing is a smaller part of a scene is also a huge time waste….. I know blender is more of a 3D program than a compositing program, but with a bit of a workflow review on this sort of thing it would become a super streamlined powerhouse. I’ve got used to working around this stuff, but it’s my main frustration at the moment.

  130. That sounds like a good strategy ! I can’t wait.

  131. What about Qt library for GUI and C++ API and material library system?

    • +1 QT library is used by Maya and Houdini. Blender really has big problem with gui.
      For example, Blender GUI has no custom panels, custom windows, custom tabs, custom TreeView elements. QT Gui is future of every 3d package.
      As far as I know even Modo 901 has QT support now.

      I hope Blender developers will hear our cry about GUI.

      • Highly against this, we can add support for custom panels/windows/etc at any time, but we’d loose a lot of existing GUI features only we have, when switching to QT or similar. See

      • Though I like and use Qt privately a lot, I do not think that it would be a good solution for blender. I’ve seen Maya and I have to say that we can do better, but I can’t see us doing that with the Qt library, as it is useable for standard window apllications, but in my opinion just not fit for such a complex project as blender.

      • Just because QT is used by other DCC apps does not make it a good choice. These other applications most likely chose QT because it’s one of, if not THE only viable existing ui framework with good python bindings. In other words, QT wasn’t chosen because it was their best choice, QT was chosen because it was their only choice. I think Blender can do better than that.

        As Julian points out in his link, there would have to be a clear advantage to switching to QT.

        I personally find coding UI’s in QT to be a giant pain in the butt. It’s great for small little UI’s that just need a few buttons, but once you need something sophisticated (with things like lazily loaded thumbnail model/view/delegates) the development time increases in orders of magnitude. In short, I think time would be better spent improving Blenders UI API than switching to QT.

  132. Kudos! This is exactly what Blender needs now. There’s a time for smaller improvements and a time for larger projects.

    Since you mention keeping the same data and code design, is there any chance to put some work on developer docs too? I mean not only BPY, but core architecture details. For instance a large, in-depth article about the RNA/DNA system. Or some brief overviews on how each module is organized and designed.

    I know documenting sucks, but it’d help ease new developers into Blender.

  133. greetings to all the team, I am a sculptor 3D, I began in Blender and I would stay .but there is a thing in thine heart for all the sculptor modeling, the Multiprocesing to Multiprocessor when a Blender? thank you again for everything.

  134. Wonderful plan. Don’t forget link asset and renderlayer granular overrides!!

  135. Fantastic idea Ton. Blender was developing very fast recently, almost to the level when changing versions was a little bit frustrating, but there wasn’t many game changing features.
    PBR viewport would be an outstanding addition. More and more people are using blender in game development and current trend is PBR. Having ability to see actual materials inside blender, while preparing textures would be so much easier. Programs like marmoset would not be needed anymore.

  136. I’m writing this representing BNPR and our over 3000 members on the most active non-photorealistic forum online ( With such a big members, we admins can see where artists are struggling with Blender from NPR’s pov.

    1. Everyone is struggling with normal editing. There are 3 solutions, yet no real synergy. In NPR vertex normal editing is as important as lighting for photorealistic.

    2. Better ways to rig. Instead of just 1 Rigify add-on for whole body, we should break it down to many more pieces. For example, we can have many arm rigs. One for basic arm rig (game/mocap), another for more control and another for arm with animated muscle. Also we need to explore how rigs are done in other DCC for more inclusive perspectives on the meta rigs.

    3. Better primitive. It is often done yet seldom mentioned that in Blender we do the same thing many times. One of them is human character. A base humanoid primitive should be available with few basic parameters to adjust size and height.

    4. Better rendering API. This will make the viewport improvement fresh for new features all the time. BNPR will be participating with the development with direct discussion with artists and studios involved.

    5. Texture painting is OK, but has many workflow traps. They are:
    a. saving painted texture (everyone who have lose their paintings will agree).
    b. hidden masking ability (just too many steps to turn it ON and OFF).
    c. cannot paint in layers (can only see & paint 1 layers at a time).
    d. cannot handle big textures (extreme lag at 4k textures painting).
    e. smear and paint cannot be turned ON at the same time for more natural blending (this decrease texture painting time as much as 75% as experienced in other software such as Photoshop and Krita).
    Solving these will make Blender texture painting more on par with other DCC software.

    6. Project and asset management. In other 3D DCC software, you have the option to create a project (with file structure templates) or create a Blendfile. This will directly solve problem present in 5a. This also introduce good habits for managing projects & assets, something Blender users should have a long time ago.

    7. Enhance the outliner to have 1 extra “mode” for users to arrange and rearrange objects in folders.

    8. Lifts the 20 layers limit. With bigger projects, 20 layers are just ridiculously too little. When you have more than 10 characters in a scene, everything is very hard to handle. Solution such as LightWave is very nice, since it will not change current UI that much.

    These are just few big things that we are struggling with on the daily basis. Thank you for your attention.

    • 5) Base human model, good idea, would be handy as props to fill scenes.
      maybe stick to model like second life ?, just because that model has so many textures etc, other model could do too.

      7) would be nice indeed, complex scenes require a lot of naming stuff and filtering stuf, manual movement (in a placeholder tree, would be nice).

      9) i’d like to add one, BI is on its end, but some good game like render engine not real like cycles, with perhaps the support of a layered based scenes (layers on top of each), as in many games is used.

      and as with (9), 10, better support for text, text in a layer on a place fixed position.

      9-10-11, and yes no realist rendering is important too, besides ofcourse a super fast cycles engine. Maybe we could have some more and better denoise filters in compositor. Or perhaps something like a morph frame, so uneven frames get morphed.. just thinking aloud howto speedup animation rendering.

  137. But what about canceling BGE thing,I know people who stopped working on blender game engine because it is going to be canceled,Can someone explain this point?

    • Rather than see BGE cancelled, I’d rather see it converted (or simply extended) into being a material editor and renderer, like Marmoset Toolbag. It seems perfect, if you ask me.

      • Correction: “…into being a REAL-TIME material editor and renderer…”

    • Hey! Why don’t you like BGE? Do you think that it is unused? Come on – it is only engine that I like + my computer can run it pretty well. I can’t even imagine Blender without BGE.

      • from the blender conference talk about this ( ) i got the impression that the default viewport stuff could very well end up encompassing what the game engine does now as well as being a PBR render engine to replace blender internal. so if i understood that correctly in the end it’d be all about how you used it rather than being a specific for that purpose kind of thing.

  138. i am not being funny but i am surprised that blender foundation isn’t doing this every couple of years already lol, but anyways yay it is about time that they refocus blender 3d.

  139. Hm i dont think the UI is bad, i’m mostly a keyboard user :)
    But maybe have some UI access python/html like, so people can create their own ui & skins, so resources wont get lost into a super fancy viewport which does do the same stuff

    But maybe time to ditch BI, some will hate it but a part of code that cannt be touched shouldnt be in it i think, its a conflict area now.
    (people will still be able to use blender 2.7x for that), for the rest i hope better nodes and python support preferable in most view pannels, like compositor python access to image layers.And nodes based particles.

    However eventually one ends up bugfixing no matter what version.
    Thats just code cylces. If thats the part costing a lot of time, then maybe think of re-organize the coding proces / who aproves / devide groups up in small groups of 3 for example. each group can update and release their code parts (multiple small grouped dll’s, its how projects are done these days). Might also checkout some automatic code refractor solutions.

    Oh and if possible, put some development on youtube or so, use a screen recorder, other people can learn from it.

    • BI is much faster than Cycles, and still produce good renders.
      I don’t think it’s time to drop it.

      • BI is in non-maintainable state since years now, idea is not to merely drop it, but to replace it with some modern, nearly-real-time OpenGL-based renderer that would do pretty much same things as BI, but with all modern shaders etc., and a design allowing it to keep evolving as new possibilities will be unveiled by newer GPUs.

        That goes for all ‘drop things’ listed in original post – ‘replace things’ would have been better, we do not want to remove those features, but rather to remake them.

        • Oook.
          But also a plain-simply-cpu-based-always-everywhere-working render engine it’s not a bad thing… ;-)

  140. i agree with Marcus…
    Blender is absolutly great but it deserves an awsome tool like the “mograph” for C4D.
    Today Blender is getting better and better but not for the motion designers. Of course we can fake some stuffs with the use of various scripts. However it’s not always fun to work with them… and Hard btw…

  141. what I read is a realistic model of managing a project like Blender. I agree with Too and the devs regarding shedding ancillary code to replace with more modern concepts based on contemporary mathematical and theoretical CG that are established at this time. SIGGRAPH is the source of great developments and since the opening of Pixar to the community by making renderman free and in addition engaging in participating to the development of Blender, these resources represent a unique opportunity to engage toward new frontiers in CG, Animation, rendering, modeling etc.. Artists are lucky to be served so well by Blender Foundation and the community.. It is very courageous to assume such a position.. Bravo Ton

  142. I agree with most things , but you must put a greater focus in Blender from the artist’s point of view . Blender is becoming technical , and lot of strange buttons everywhere that are hard to understand the meaning of . Blender should be a tool where you as an artist will not get caught in a jungle settings , when you in fact just want to start create something. Blender has succeeded is how fast you can model compared to other programs , but understanding material is hard för a beginner.
    Materials could be simplified a lot , like ” Keyshot “, drag n drop principal, then if you want to ”tinker”, you can enter the ” node-tree” as an advanced option.

    I also think that the blender should be more suited for motion graphic artists when it comes to animation , UI and so on. Maybe introduce presets or a similar engine that mograph in Cinema 4D?

    • People make material libraries on websites for free. I’m not sure what your problem with that approach is. It’s not hard to google this stuff.

      Besides, one should learn the “why” of something. It’s part of what makes Blender so great. If you learn the why, then you can contribute to the community so others in the community can also know the “why”. That’s part of being an artist. You don’t just have a button that sets up automatic compositions, do you? Or paints that pain the picture for you? Setting up Materials are about as artistic as anything else. There is planning, a style, a look. I completely disagree with you especially since you can go on a website and download pretty much whatever you want as far as basic materials are concerned and “drag and drop” them into your scene, as well as delve into the nodes to “tinker.”

    • Too simplistic methods of creating materials and you end up with much less options. I find the Cycles nodes actually pretty easy to understand so far. There is always some basics everyone needs to know about how to write shaders for your materials.

      For the actual subject there is, I am glad that they decided to actually take a break from the current release schedule to concentrate on working on the next best update, that will finally eliminate legacy stuff that surely prevents Blender from being up to par with the current trends in modeling, with PBR material support (hopes are high here, will we have metal/roughness, possible compatibility with Substance materials as well?) they can then go further and add support for some of the best PBR material formats import-export like Substance, which in return will help a great deal in the workflow between Blender and game engines like UE4 and others that already have full support for native Substance PBR materials.

  143. I completely agree. At last some wise words from Ton! Good luck!

  144. I agree. The developers are core to Blender and doing something that is not a pleasure anymore can cause the development team to lose valuable contributors. We need them and heck, we appreciate them!
    Let it become fun for them again as it is fun for us to use Blender.

  145. I think this is a fantastic idea. Having a development pause (so to speak) and allowing the team to concentrate on what they have will bring focus back to the project. Plus, you’ll all be chewing at the bit for the 2.9 / 3.0 cycle.

  146. This sounds like a wonderful plan to me. By focusing on the bigger picture and a release further in the future the development can find and give direction to Blender and its UI and workflows instead of being thrown from one bug to another and being mostly reactive rather than proactive and creative.

    A focused and mindful effort to improve UI (with everything it entails, from workflows to tool consistency, interoperability and everything related to user interaction) is exactly what Blender needs, and I feel like it’s crucial that the devs take the appropriate time to get this right and build a solid foundation, both technologically and in terms of design guidelines and ‘Blender principles’.

    Both thumbs up for this proposal!

    • I’m still waiting for an updated help file lol. I shouldn’t have to watch videos in order to figure out how to work this program.

      • Or you could upgrade the documentation as you learn? That’s more in the spirit of blender development.

    • I’m not sure what’s actually better though. Is it better to release a chunk of stuff that is really buggy and needs to be fixed anyway, in which it’s too large of a release and certain features just stay bugged because it gets left behind, or to retroactively fix bugs in smaller releases? Maybe one is better for the developers but they should want to do what’s best for a complete program.

      You have to remember that this is all driven by user-base. Constantly adding small releases piques interest and keeps users active. It also drives up user-base. The above model is actually horrible for other programs…3DS Max, Maya, etc. Everyone always complains at release how buggy things are and that really paints a negative attitude. I’d rather see smaller releases, like we have now, because smaller means less expectations and happier users when things are fixed right away.

    • I desperately wish for them to put in alembic….Blender nerd plz plz. alembic.

In order to prevent spam, comments are closed 15 days after the post is published.
Feel free to continue the conversation on the forums.