2.8 project developer kickoff meeting notes

At the 2015 blender conference the attending developers sat down to discus things we as a group wanted the 2.8 project to be a software engineering perspective.  The things discussed below are intended to become effective with the 2.8 project and any changes in supported hardware will be kept as minimal as possible.

C++ 11 / C99

Blender is written in C, C++ and Python mainly. currently we use C++98, C89 and Python 3.4.

There is consensus to allow C++11 and C99 for the features that make sense and are supported by our current hosting compilers (Microsoft Visual Studio 2013 is the lowest common denominator here ).  This should let us write better code thanks to some stupid limitations being lifted but it will also need us to bump the platform requirements and in particular support for Mac OS X versions lower then 10.8 and Linux versions that ship with a glibc older then 2.14 would be dropped.

OpenGL

Currently blender uses OpenGL in a way that remains compatible with versions 1.4 of the standard. Over the last 20 years graphics hardware has evolved greatly and some of the concepts in accessing this hardware have also changes. In 2009 the OpenGL 3.2 standard was released that for the first time deprecated the old way of doing things. Today a lot of platforms even do not allow this old way of accessing the hardware and some disallow use of newer features when legacy calls are used (MacOS X is an example of this).

The developers universally agree that this will happen and is unavoidable. We also felt that this move away from immediate mode towards VBOs and GLSL will need to happen, regardless of the new viewport design. Antony Riakiotakis started this conversions, but there is a significant amount of work left, and it is unclear at this point how this is to be approached best.

This move will have some downsides, such as loss of hardware acceleration on early Intel i9xx cards. Any post-2008 Nvidia or AMD hardware should remain unaffected.

Scons

The Blender developers currently maintain 2 buildsystems (cmake and scons). Most of us use CMake, more than we use SCons, and collectively we feel that dropping one would free up a big enough amount of resources that the benefit would far outweigh the costs. There are buildsystem-specific bugs, it adds to difficulty of becoming a contributor, and the builds on both systems are currently inconsistent.

The remaining work lies mainly in supporting the linux release builds with cmake, and verifying the MacOS X release build against the scons version. Brecht and Martijn have volunteered to get this done.

Replacing or dropping code

There are various opinions on what parts of Blender are broken, hard to maintain, or lack a future. Mentioned were the sequencer, game engine, openimage-io, constraints, particle system, and OpenCollada. The only one we could reach some kind of consensus on is OpenCollada: the library and integration make up for 1/3 of the binary size of Blender, and we currently only have Gaia to maintain it (who was not present at the meeting). We decided to seriously consider dropping it for 2.8.

The particle system and constraints may need a complete overhaul.

The sequencer and game engine are in serious danger of removal, if we cannot come up with a good solution during the 2.8 project.

OpenNL was also discussed and it seems most of the usages could also be covered by the Eigen library.

Finally, it is good to remember that this discussion is about what could be good for Blender and the Blender developers from a software engineering perspective, and what could make it easier for us to deliver a better Blender. We make Blender for artists first, and in that sense this list cannot and should not be interpreted as a complete representation of the 2.8 project.

225 comments
  1. Sequencer with a Scene strip is the best video editor I know of.
    Tone mapping + Color correction + Lens correction nodes.

  2. We’re using the bge as well, PLEASE retain it!

  3. MORSE (https://www.openrobots.org/wiki/morse/) relies on the BGE which has a large potential in robotics! It should be kept in Blender.

  4. The love for certain features is strong on here. When you say “dropping” a feature, it sounds like it will be removed from code base and become lost and forgotten. Hopefully not. Being opensource, I would just split these features into their own code base (separate from Blender releases) for anyone to maintain on their own. Perhaps one day someone will make standalone tools that integrate with Blender or maybe even as Addons?

    I’m actually pretty glad you guys are finally considering the implementation of modern OpenGL as well as making use of newer C++ capabilities. Looking forward to this!

  5. The BGE is a wonderful part of Blender that we use extensively at my start-up. Keep it in there, please! When we get off of the ground, we will hire at least 1 full-time Blender developer to fix bugs and push the BGE forward. 1-2 years away…

  6. In my opinion I also feel the VSE sets Blender apart from the other software I have used and I find myself using it more and more lately. It is very useful and so convenient not to switch applications to get another part of a project done.

    If believe (after reading these comments) that the VSE will live on and just get overhauled; it is integral to so many people’s workflows.

    But, whichever way it goes; I will still support Blender. It is not only a piece of software but a type of personality that grows on you.

    So, thanks to all devs, Ton and the Blender community! Keep up the great work.

  7. If the entire blender program is upgrading to OpenGL 3.3+ then you don’t need to scrap BGE.

    Making VSE GPU based when rendering previews is all the VSE needs and should not at all be scrapped.

    Most importantly I feel the issue of texture painting as vital as it is between BGE and VSE, should be addressed first. It needs to be GPU based and have a sleeker GUI and layering system equivalent to krita, gimp, etc.. That is the only soft spot in blender currently and the more important than altering the BGE and VSE.

    On a minor note it would be nice to have Volumetric, and Smoke GPU rendering on OpenCL by 2.8 :). Or using HIP to convert CUDA code at the least.

  8. Personally i also think that, One software cant be made to support everything. If it does it will lose its quality. So the game engine part is really unnecessary. Blender is a great tool for Modeling, animation, sculpting. Next version should focus on them and specially on rendering performance and its quality.

  9. I read the above regarding BGE potentially being dropped and must also voice my thoughts as someone who is seriously impressed with all the many many hours of hard work that must go into producing such an exceelent piece of software. I am currently weighing up all options with regards to game development and believe it or not I was actually really hoping as a preference to be able to use BGE if I am to use a game engine.

    I just want to point that that there is a recent movement among programmers that Ton and the rest of the Blender developers should be aware of… HandMadDev – I honestly think that not only should Ton consider reaching out to Abner Coimbre regarding possibly being interviewed for the HMD Show, but also it would give the HMD community a chance to see the Blender project in more detail with potential interest to be stirred among devs and who knows, maybe some extra coding talent for the project.

    I notice that Tons main reason for considering dropping/detaching VSE and BGE are unwieldy codebase etc… The folks in the HMD community are all about ‘code done right’ and keen to put out quality, high performant, usable code.

    Just my 2 cents as someone who sees things from both sides of the fence.

    Thanks again to all the many, many people involved in Blender – ;o)

  10. Hey Ton and Blender crew,

    Just want to hop on the band wagon and say that the VSE is something I use on essentially every project I do, inclusive of commercial client work. It is my go to open source video editor and it’s extremely important. I would gladly help out where I can.

    I would also like to thank you all for the fantastic work you all do,

    Braeden

  11. How about raising VS2013 to V2015 as a new base to work on.
    I mean its free for projects like this.
    which would potentially lead to more developers (students etc).
    No one will start with buying VS2013 these days.
    As a bonus it has better debugging options too.

  12. Hi guys, I have a question. Will the new OpenGL branch of Blender improve texture painting?

    I ask because I’m starting to learn how to paint textures in Blender, instead of using Mari Indie. Hasn’t taken long to get used to; but, fiddly things like Masks only working in texture mode, and not displaying the texture underneath are a bit annoying.

  13. I do hope something can be done to save the game engine. I do like having it to play with to learn basic game engine design and flow. If it is removed I guess I will just have to keep an older copy of blender around to maintain the ability to use it.

  14. My two cents are that making a tool less useful is not a way to make it better.

    I think an extended metaphor describes it best:

    Dedicated tools will always be better at accomplishing their specific intended tasks. Given unlimited time, space, and resources in a perfectly organized workshop, one could use any number of tools, each ideal for its task, swapping between them and using each only for its intended task.

    But outside this imaginary workshop lies a place called the real world. In the real world juggling a dozen different tools is not efficient. In the real world, you use a hammer to open and stir a can of paint. In the real world, duck tape can fix anything. In the real world, the perfect tool is the one you can reach while holding something expensive together.

    Here enters the multitool. A single tool with many features which together make it an adequate tool for any task. It will never be the ideal tool for any task. But no single purpose tool will ever be as useful in the real world as that perfectly, universally, adequate multitool folded up in your pocket.

    That is Blender. A multitool. Not perfect. But indispensably useful because it can do almost anything adequately. And that is what it needs to continue to be. If features can be improved, improve them, and no one’s going to complain about taking away that obsolete corkscrew so long as there’s still a pick to replace it. But so long as something has a purpose, functions, and its function cannot be duplicated more efficiently within the program, it needs to stay, even if it isn’t ideal. Its having the option available that is vital.

  15. The VSE is the main reason I use Blender, it’s animatic capibilities is uncomparible with ANYTHING out there. There is no camera sequencer (and you could do that with the VSE), there is quick sequence to playblast (VSE could do that) and there is no way to quickly sketch in animatics (Grease pencil in the VSE) and all of those things are very necessary to a good animation production. There is no video editor that has a grease pencil that I know of. I think a huge mistake – to remove it unless there is better alternative – by the Blender community in my opinion…..

  16. would be nice to have a pro contra list where you have some estimated data like: module name, current state of the code, required ressource for update, whats good, whats bad, whats missing….

    in my point of view is optimize the code and analyse what part is mostly used. focus on what big studios(film and game) want as soon as they see that blender can be used as a core you get much more code from studios directly for blender. example gaffer from image engine could have been a wonderful addon in blender.

    BGE can be kicked for now and when the time is rigth you can make a project for the development of a new game engine. maybe you get a funding by the H2020.

    VSE remove it and plan a integration into the compositor. everybody can use a older blender editon if required.

  17. I never understood why there is two almost same things in Blender: VSE and Compositor with own not shared filters. I did compositing in VSE in blender 2.3x far before Compositor was introduced. I really love VSE, but IMHO VSE could be implemented as special time based representation of compositor. And instead of spending time to improve VSE developers could spend time to make Compositor near realtime.

    Also most outstanding feature is image proxies. When I work with overall video rhythm in VSE I do not really need full resolution image, 320×200 just enough to see timing. I believe Blender’s compositor can handle such resolution in realtime.

    Most of open source video editors tries to handle full res images all time, and it is very painful when video start lagging just after appending some color correction.

  18. Best to get rid of things that are no longer supported or obsolete like OpenCollada and perhaps the BGE as well since Blender is not a dedicated game engine like UDK. Please keep the important things like the VSE, particle system, sculpting, the animation system and the node editor. My advise is do a survey to see what features are most important to artist.

    Well its time for Blender to meet the current industry standards and since you guys are doing a complete rehall of the code, it would be a great time to impliment a better ui scheme like the one Andrew Price purposed awhile back. The goal is to make Blender workflow orientated. Blender needs to be more intuitive. A good example is the gamma node, by increasing the value it will actually decrease the brightness which is the opposite of what I expect it to do.

  19. Regards Open GL in Blender I use this a lot for actual final output due to high speed of turnarounds and instant changes required by our marketing department. The addition of depth of field and AO to viewport is great – wonder if potential for even more viewport rendering with a nod to the approaches made by VideoCopilots Element 3d Open GL

  20. Oh god….Please never get rid of the particle system. I rely on it heavily for my work load.

  21. Please focus on BGE, Games and Interactivity is the future!
    (VR filmmaking, ‘cinemaking’, collaborative environments..)

    Create an addon-system for BGE:
    Let user customize on how they want their game-engine to run (they can enable/disable any of the components just like the current addon system in Blender3D)..

    Choice of Renderer, Sound library, Scripting (c#,c++,javsscript,etc..),
    Hardware support (VR,Peripheral), Web-components (online multiplayer, Ad-framework), 2D framework / 2D/3D, GUI, etc.., database support, Archviz-components (most requested feature..)

    Don’t limit yourself to just linear-films.. Rendering images at some point will become real-time, the next phase is allow more accessible way of design things intuitively (Imagine Blender3D as a HUD Viewport for AR/VR.. the 3D space is no-longer necessarily static, it will act like James Cameron ‘Virtual-Camera’ project for Avatar.. with the perks that it’s open-source (open for extensive innovation and application)

    User can either model/design traditionally or just move their tablet (read: microsoft surface) or desktop-tablets with embedded cameras OR wireless cameras OR bluetooth cameras as an additional standard)
    and model/map while moving in real world space to virtual space.
    Remember the motion-tracking tool.. even more useful that at some point Blender could do its own Photogrammetry rather than relying on Agisoft or Autodesk pipeline..
    ([STEM]fields are looking closely for established open-source tools to bridge their prototypes together)

    It’s with thanks to your Node-system and logic bricks that you have inspired other companies like Epic, Crytek, and others to have followed route that Nodes/Logicbricks is proven to be very effective in accessible design for Artists and Artists at heart.

    BGE needs more support and development.. The future is in VR/AR..
    next we’ll need real-time collaboration (Think like Second Life or MMO but inside Blender3D; from Multiplayer to MultiDesigner/Multideveloper/Multi-user..).

    I also agree with the folks here that have mentioned in regards to VR moviemaking, we need a real-time engine. Do not abandon BGE.. Have BGE evolve (either by making it more modular: it’s own addon-system) or gradual design/UI/UX change and performance enhancements..

  22. For archviz/mechanical/tecnical, more tools to allow faster and precise work is more important than BGE, VSE etc, IMHO.
    On this side, Blender lack a lot of features. :-(

  23. Would be nice to have saved masking selections in sculpt mode. So if I select fingers I can cycle through the mask I’ve made to invert and sculpt them. Could play off the vertex groups somehow. For instance the mask you create allows you to assign the vertex group and selecting a vertex group creates a mask.

  24. IMHO replacing the GE with a public integration API to be used by other engines for seamless content pipeline integration would be the best. To achieve something kinda like what Autodesk do with their Stingray but Blender will only be concerned to provide other engines the services needed to integrate it as an editor. I think OGRE team started doing something like that for Blender to be a material editor for their 2.1 branch (http://www.ogre3d.org/forums/viewtopic.php?f=11&t=84227).

  25. Hi, quick question about OpenGL.

    Has any thought been given to looking at Vulkan?

    Its great all the hard work thats been put into OpenGL for Blender recently, but instead of continuing with OpenGL, maybe pause work on it and wait for Vulkan to come out so a comparison can be made between the two? Hopefully it would allow the developers to decide which might be better for Blender in long run.

  26. Don’t kill the sequencer and game engine !

    On the contrary, Blender has been ahead of the game with an integrated game engine, it could be that other 3d suites will integrate one by the time blender deletes it… because of the upcoming VR story development…

    Have a look at this:

    http://www.roadtovr.com/fabric-engine-is-working-to-enable-3d-storyboarding-and-editing-directly-in-vr/

    that is one of the potentials for the GE, you need a real time engine to do vr moviemaking… will blender ignore this whole upcoming market?

  27. I thought we were planning on upgrading the Blender Game Engine rather than killing it: http://www.blender.org/press/18-anticipated-blender-development-projects-of-2015/

    The game engine needs to be more closely integrated with the rest of Blender. For example, since you can change material colors in Blender, you should be able to do the same thing in realtime. Have you ever tried to make a car game where you could change the RGB color of the car in-game? It’s not possible without keyframes. I understand that this is a HUGE project, but PLEASE don’t simply remove the game engine entirely!

    Thank you for reading this.

  28. Allow me to add my voice to the crowd:

    The VSE is powerful, extensible, understandable (even to newcomers), and vital to my and countless other’s use of Blender, especially on Linux.

    For the sake of all that has ever been good, DON’T REMOVE IT!!!

  29. Huge thanks to all the people who programming blender.

    My wishes and hope for the future:

    Cycles Noise Level in
    Global illumination:
    If you need a GI not exactly perfect but some GI without the noise.
    There is no simple decent-effective solution.
    It is not possible to obtain noise level down unless you render 500-2000 samples.
    Which is substantially 30-60 min or hours for one frame.
    I use a partial solution to separate Diffuse Glossy Transmission and use BILATERAL BLUR.
    But this is not perfect it depends on the scene that is quite smooth but remain here and there one pixel black dots.
    If it could be a global illumination blur depending on 3D normal?
    Like this on
    Youtube: Hardware-Accelerated Global Illumination by Image Space Photon Mapping.

    Open CL Physics:
    Physics, Particle, Cloth, Fluids, add-Granular, use OpenCL or Full CPU USE
    Present Particle System is not good – especially computation run like 20-30% CPU
    Breaking object like in Blender Molecular Script would also cool with OpenCL.
    Blender Molecular run like 10-20% CPU.
    Ideally in one particle system to simulate complex physics interaction.
    Like: Nvidia Flex (Which is only closed source futures for games Geforce GPU)
    Youtube: Nvidia Flex
    Youtube: Cloth – Fluid simulation with elastic solids

    The current situation is that these calculations take hours.
    If the processor would go to 100% this process could run 3x 4x faster.
    If it run on GPU OpenCL could run 20x faster.

    Fluid Open Soruce:
    DualSPHysics v3.1 web dual.sphysics
    fluids3 web: rchoetzlein

  30. Just so I’m sure I understand: even if you decide to drop OpenCollada, you will retain support for the format through some other library, right?

    There are of course many good reasons for using Blender, and we all have different needs, but Collada support pretty much a make-or-break for me personally. Without it, I would have to look elsewhere — which I *really* don’t want to do.

  31. For years I’ve been watching developer notes looking only for Collada improvements. AND NOW THIS??!! I’m very disappointed!
    You must realize BLENDER IS NOT THE MOST USED 3D PACKAGE out there!!.
    Soooo, one of the MAIN priorities should be, IMO, interoperability with other tools.
    I know a good effort has been done in FBX, which is very interesting, but I’ve always thought that Collada was the great hope for open standards.
    At least it’s the only way I can import rigged and animated meshes.

    Note: maybe Assimp project could provide a good starting point for an alternative.

  32. I’m migrating from 3ds max to the Blender, but Animation layers like the Autodesk’s CAT system would be great! Or something like base layer and adjustment layers… =)

  33. How about improving animation tools? Like adding animation layers and complete nla overhaul, which will behave like for example motion editor in “source film maker”. The code rework is need to be done, true. But at first blender is the software for making animations. After a year of developing can we expect to see some improvements in this area?

  34. I repeat myself but why not try to appeal to kickstart ?!
    It is a crowdfunding platform for music, games and many more

  35. What i understood so far is:
    -the BGE could be part of an action node system in the future.
    -the VSE needs to be updated
    -these two points are from a list of options to get a faster, consistent development.

    What i don’t understand is:

    -where comes the time pressure from? Who cares if development speed is slow? Blender is a free tool, not a stock corporation. If it takes ten years to fully integrate the BGE into the rest, then i guess we will enjoy what you already managed to code meanwhile. The VSE is amazing as it is.

    -some people: they use a free software and complain that its usage could enable others to get a free access to their assets of their non-free games. Do i understand that right so far? So why don’t those people just leave us idealists alone and choose a commercial substitute? Noone forces them to use blender for their business. So they shouldn’t push the spirit of open source in the money corner. Please take care of that spirit. I’m already alarmed by the rise of the blender market. It will divide the community into first and second class. I do understand that many formerly idealistic developers get older, maybe get children, and start seeing the need to get some wealth. But i’d rather like a future with slow but free progress than a two-class-society. And that i mean in general, not only blender-specific.

  36. removing the video sequence editor without some sort of replacement would be a huge loss of functionality (and therefore a big mistake) in my humble opinion

  37. don’t drop the VSE, i worked in a web agency making 3D clip with blender for clients and the VSE was my main Editor, moreover i used it to edit video montage for my company.
    it is really the only good multiplateform video editor i found

  38. This is really exciting news. I learned blender with 2.4x, and the 2.5 overhaul was amazing, so I am really looking forward to the next large overhaul.

    I am an architect, and I see a really amazing potential for Blender in the mainstream for architectural use. For this occur, there are a few features that would make it play well with most other software that architects use:

    1. Better layer system. Many times, we import large files with many layers. It would be great if the layers from the imported file are preserved

    2. Wireframe colors. Along with layers, in program like Rhino and 3ds max, it would be great to assign a wireframe color, as it helps differentiate within a very busy architectural scene, different types of objects

    3. Materials. Similar to layers, when importing files, it would be great to preserve materials as much as possible based on the original. I know there has been a lot of progress with this with cycles, but it isn’t as good as it is in other programs.

    4. Selection filter. Rhino and 3ds Max have great selection filters, where, only a certain type of object can be selected in the scene – lights, meshes, bones, etc). Although I know that we can select all by type, it would be great to have a filter that can be turned on and off, so within a very busy scene, only camera can be selected, for example.

    With the features mentioned above, this would simplify the workflow for many architectural designers and remove a large barrier that the software has at the moment for newcomers.

    Second priority for architectural design:

    5. Nurbs. If there is some way to continue what was started in Gsoc 2013 with importing and working with nurbs more natively, it would be amazing.

    6. Similar to what many others have mentioned, it would be great to have a centralized plugin repository. Dynamo, a relatively new grasshopper-like software that works with Revit, has done a great job with this. They have a repository that can be called from within the software, and immediately add a package, and just as easily remove it.

  39. Man, so many people out there beg for mercy to BGE it’s tempting to turn troll and demand the death sentence for it! :-) But I’ll be merciful and suggest turning it into a physics/simulation tool that can be used along with the things like fluids, flames/smoke, cloth, etc. Of course the poor thing will never, ever come close to the glory of UE4, not even it’s little brother Unity (sorry, couldn’t resist this little bit of trolling!) and I, for one, would sooner revert to chessboard than actually play any game made with it, but with some out-of-the-box thinking it just might still be useful. :-)

  40. any plans about the buffered page particles project from lukas?

  41. The Sequencer is one of the main Reasons for me to use Blender. It is a very important Part for making movies, trailers and so on. Blender would become kind of uninteresting for me if it would be removed.

  42. Add modelling nodes the way Sverchok does for Blender. http://nikitron.cc.ua/sverchok_en.html https://github.com/nortikin/sverchok/
    If you want a workflow then the modelling needs further automation and Sverchok nodes process the vertices, edges etc.
    Nodes are used very well in many places in Blender. They need to be the modelling area.

  43. I’ve been using Blender since I was 13, just after the 2.5 update. I’ve done a lot of 3D work, have enjoyed making games, and am now working as a professional video editor.
    I would be in favour of dropping the BGE and the VSE for the following reason: I think Blender now has to make a descision whether it wants to be considered as a serious, professional tool in the industry. The BGE will (as it stands, at least) never be used to make a wide reaching game, and the VSE is not viable as a professional video editor. If we hang on to these features that are only used in non-professional context, we will not be taken seriously in the areas where Blender does excel, i.e. 3D modeling and rendering.

    I have nothing against Blender being used by hobbyists and non-professionals (heck, I’m both of these), but I would hate to see some peoples reliance on features (that are not best for the job) stop Blender from becoming a more widely in the industy.

    As I have said I dabbled in game development, but I used Unity, as it’s designed for game development. The BGE dosn’t have a good workflow for making games.
    My job is as a video editor, and I use Adobe Premiere. The VSE simply dosn’t have the workflow to allow it to be used for professional video production.
    In case people think I just don’t like Blender, I would say that I tried 3DS max for about 15 minutes, found it’s workflow wasn’t good enough to let me model well, so I went back to Blender.

    In short, I think workflow is crucial in all software, and if we don’t have the resources keep the workflow clean in a particular feature, it should be dropped.

    • i fail to see why you think professionals are dead set on using a single piece of software. there are things a software does well and things it doesn’t. so in general they use the parts that work for what they’re doing. just because blender is an entire suite of tools doesn’t mean you have to use them all.

      another thing, blender being considered a serious professional tool in the industry is going to have more to do with the amount of visible quality work produced in blender, and if it will work for their needs, rather than if blender has the VSE or BGE.

      besides, the VSE isn’t really blender’s draw and i doubt by far that most people interested in blender are interested for the VSE. though it certainly does work better than what you’re giving it credit for and it is a valuable asset.

      as for the game engine, you’re right that it may never make a wide reaching game (as it is now) but it has been used in scientific research
      and areas not strictly “game” related and has been a valuable asset for the ones who used it as such. i even remember reading that the BGE was chosen because it’s integration with blender allowed it to be uniquely suited to what they wanted to do.
      they’ve presented the idea of merging the BGE into the viewport to better integrate that into blender. personally i think this would be a great step forward and would enable not only the BGE to improve with better integration but it would also give more utility to blender as a whole that currently only exists within the BGE. that’s pretty much a win/win if you ask me.

  44. I use the VSE and compositor heavily. I have never used the BGE so I can’t speak to that. In my opinion, removing the VSE would be a huge mistake for a number of reasons.

    1) It is the best openscourse video editor for linux. I have tried others but they just don’t work as well.

    2) It allows for a fluid method of dealing with the entire project as a single output from the 3D space or the compositor.

    I would LOVE to see the VSC and the compositor integrated better.

    I think we all agree that the VSE’s workflow needs a lot of work. Looking at what is common in the other editors out there can help. To me the main issue I have is the limitations of the main editing window.

    1) No separation of audio and video.

    2) The inability to easily add strips/channels between existing channels without moving all the above strips which is cumbersome.

    3) The worst is having to have multiple strips stacked when a strip is modified. The edits become overwhelming with many useless strips.

    4) It allows for a freelancer or small studio to have the tools they need.

    Overall however I still use the VSE to the exclusion of others like Cinelerra (What a mess) and Shotcut because it IS better.

    Thanks Ton and all the developers and add-on creators, Blender is just what so many of us needed.

  45. Hi,

    To me the ability to produce an edited animation video from scratch within a single piece of software is one of Blender’s big strengths. Blender’s VSE is actually quite good (probably the best open source NLE) I for one would sorely miss it, should it be removed from Blender :(

  46. I am sad to hear that blender will change so much. You will need much money to maintain and improve blender such as we know today?
    You can not make a call on a network such as: https://www.kickstarter.com/ ??? If I was rich. .

  47. dropping constraints? what’s next? dropping modifiers?

    • when they say that they mean temporarily until they can get it fitting into 2.8+ properly.
      based on what they said, i’m pretty sure they don’t mean to permanently drop anything but some of it may change forms.

  48. As a relative newcomer to Blender i offer the following radical suggestion……

    Split this hulking monolith that is Blender up into separate modules , each module would be light weight , with only those features and buttons for that module.

    Use the blend file format to transfer the data from one module to the other as needed.

    That would then also free up some of the buttons and keyboard short cuts for each module.I’m sure it would make each module easier to navigate for beginners as well.

    Each developer would also only have to code for a module.

    End users could then decide which modules they need to download for the task at hand !

    anyways, it’s just an idea.

  49. Could you leave OpenGL1.4 support as a legacy option buried in the preferences similar to 3dsmax? Don’t need to support it with the planned viewport rework, just put it aside with the current options.

    • Nope! :D The legacy GPU option is to keep using 2.7x.

      Moving to a higher version of GL as a *baseline* lets us code without checking to see if every feature or function is supported. Like “can I use VBOs here? Can I use shaders?” For Blender 2.8 we can answer YES to these and so much more. For the few GL features above the baseline we still have to check at runtime and provide alternatives/fallbacks. But not the pervasive GL paranoia — or potential crashes if we forget to be paranoid — that exists today.

      Dropping support for old GPUs (> 6 or 7 years) allows us to be bold when implementing new features. It’s a good thing! I hope you agree and enjoy the results.

  50. Hi, sounds like a great blender time for all…coming soon again…
    Thx ….
    Hope we can work with a better color Management system in 2.8 thats near by C4d, MAX …!

    A color Management thats helps when it comes to baking or that is full implemented in the node systems and VSE.

    The VSE and a REBUILT …yeah would be great..

    And ai hope internal render with here Workflow(OpenGL) and Cycles can float more together…thx the new Viewport !?

    P.s. cycles with render Noise detecting and frühe fastest Cycles for 2.8 please..
    Dream…
    ö.ö’
    —-

    Best regard Jacky

  51. hey..!
    someone (a German developer topspr)made progress in realtime gi and more for panda game engine….open source!
    https://github.com/tobspr/RenderPipeline
    youtube videos:
    https://www.youtube.com/channel/UCK32ZFlXF7_mBSVjLgxWFkg
    may usefull or possible for Blenders Game engine development!

  52. I like the VSE, but would love to see a revamp. It’s the only cross-platform NLE free editing suite I know of, which makes it a great addition of tools for my shortfilm group.

    The game engine I personally don’t really need/use. I’ve been following Godot (FOSS game engine) and think it is going to advance a lot within the next years. Of course there aren’t that many examples out yet as it only released last year, but it’s pretty powerful. It helps that the engine is MIT licensed as that means if people make games with Godot, they can make/write a solution to bring their games to closed-source platforms, whereas with BGE that is not possible (GPL is incompatible with NDAs for the big platforms)

    The Godot developers also have a Blender-plugin called “BetterCollada” which seems okay to me for now.

    Tl;dr, I would love to keep the VSE and chip in if there’s a crowdfund for a revamp. Game Engine I personally don’t need as much.

  53. It’s really cool to see so much support for the video/sequence editor and the game engine. Here’s how you can help:

    – More developers! We always need more devs – especially the experienced ones. Send those to the ‘get involved’ section of the blender site.
    – Subscribe to the development fund. Also your company should! http://www.blender.org/foundation/development-fund/
    – Documentation is important. Join our doc project, or help developers with (video) tutorials about features.
    – Spread the word! Write articles, blog, use social media. Make sure people don’t take it for granted that there’s a free/open source 3d suite.

    It took a huge effort to get where we are now, and if we want to improve Blender it will only be more work. Even when we can attract more volunteer coders, we especially will need more full-time paid people to work on important features.
    Partially that now gets solved by the studios/companies in our network who contribute code, part by the Blender Institute (and Blender Cloud), part by the Development Fund, but the biggest leap forward is probably only possible if we can hire an additional 5-6 developers with a 2 year contract to tackle the the main issues. And that probably means a massive campaign to get the funds in for it.

  54. Please don’t drop the sequencer. It’s my second-most used feature and I’d hate to see it gone. True, I could just use an older version of Blender for it, but I already run two distinct versions due to compatibility issues and would rather not make it three.

  55. I think dropping the Game Engine is a bad idea.
    It may need a major rewrite but I find it useful and interesting.
    Many people become attracted at first by this feature of Blender.
    Many young people at least in my experience beco
    thanksme attracted to Computer graphics by this particular feature of Blender

    Thanks!!!!

  56. Still need to read the whole comments thread…but my use of Blender involves OpenCollada Exports and imports andI hope you can also find a workaround for not removing that feature set for Blender.

  57. I’ve been using Blender since 1999, I personally very rarely use the VSE and I never used the BGE.

    I’m more concerned about losing constraints and the particle system. How can you do any rigging without bone constraints? You can’t be serious about dropping the particle system. :(

    • Of course we keep particles and constraints, the code needs to be fixed or replaced. Same goes for Collada, VSE or GE actually.

      • Bless you!

        And, what about Richard Colburn’s offer to share his improved cloth simulator?

  58. i would trade 50 game engines for a better VSE. However I have noticed over time the performance has depreciated along with its speed. I think the VSE was at its prime in the 2.59 era and then after that it became a tad bit slower with it becoming further depreciated. It’s an essential feature that makes blender a powerhouse of production. The game engine however is perfect for human sacrifice. Either push it forward or cast it into the fire.

  59. Blender is one of the most powerful tools out there, bar none. I for one have been simply blown away by Blender’s capabilities. Even someone new to Blender, who has a good grasp of modeling and animation principles, can become productive in Blender. Mastering Blender, however, is a daunting task and I fully believe more would use Blender if it offered an easier path to something beyond amateur efforts.

    If I could make a change in the course of development, it would be to greatly simplify some tasks in Blender and improve documentation and tutorials. I know from incidental contacts that some people who were impressed by Blender’s capabilities were frustrated enough by its complexities to give it up for something less powerful but easier to deal with. I think this is an issue that needs to be addressed early on. It may make the some feel good that one must be a rocket scientist to squeeze the really good stuff out of Blender but I think such attitudes defeat the whole purpose of Open Source.

    1. Make Blender easier to use, provide better and more illustrative documentation.

    I think it was a mistake to continue to bundle the Game Engine with the core modules. I think BGE is a fantastic tool but it should’ve been spun off into its own project quite a while ago. Keeping it a part of Blender itself has gone toward diluting what both could become. This has also resulted in the Game Engine not being able to advance, as well. Coupled as they are together, they mutually inhibit each other. Separate Blender and the Game Engine, keeping their “roots” as intact as possible, then develop both to to their full potential. Open BGE to other tools, as well. With the emphasis on the development of the asset “library”, I can think of not time better to accomplish this separation. Asset management is just one support role Blender offers to BGE.

    2. Separate Blender and the Game Engine into separate projects.

    At its core, Blender is a modeling, animation, and simulation package. I do not know the mix between sharing objects across applications and just producing renders and movies, but I believe both are important to Blender’s soul. Thus, Blender must keep all of those features to remain as powerful as it is. Ditching the particle system or the VSE (which could be spun off as well but since it already works well it could be kept and improved as resources are available) would greatly undermine Blender. The simulations should be kept, improved and expanded where possible.

    3. Keep the modeling, animation and simulation cores of Blender and expand and refine them where possible. Keep the VSE and prioritize it when possible but not to the exclusion of other features. There are a number of alternatives to the VSE (free or low-cost) that work well with Blender output.

    The 3D world owes a tremendous round of applause to the Blender developers. Maybe you should do a credit video in the future showing some of the gangstas working on Blender. )

    Arnold Townsend

  60. A lot thanks for the GREAT work.
    We need a tool-chain tools, like Houdini, more expandable. Please split all the models, keep the GUI, keep the core, or develop a LLVM like core, then JIT the others model or data.

  61. OpenCollada: Please drop it. It’s a pig and when distros like Debian don’t even have a working version that builds against Clang/LLVM due to OpenCollada hard coded references to old ext_hash functions that were dropped by GCC 4.8 and newer you have to wonder who the hell is driving the car on that project.

    OpenShading would be great. OpenImageIO is great. Same with OpenColor and seeing as they are all interdependent on one another it’s rather asinine to discuss any one of them being removed.

    What needs to be discussed is with OpenShading not being current with LLVM/Clang 3.6 or newer, due to the MCJIT redesign.

    Larry still hasn’t updated OpenShadingLanguage stack due to legacy clients for SONY. Then again, they should build a private branch for those clients who pay them to maintain it. He knows how it works.

    Dumping the Particle system is moronic, especially when you really should be working on Bullet OpenCL 3.x support and expanding the Particle System.

    OpenGL baseline should be 4.1.

    C++11/C99? Try C++11/C11. It will make the OpenCL 2.x transition far less painful.

    Dump Scons and focus solely on CMake. That guarantees your compiler suites from LLVM/Clang, SPIR, and all the AMD driven initiatives like Vulkan and more to not become a Clusterfrack down the line.

    The Cmake still needs clean up for OpenMP when passing the -fopenmp=libomp for LLVM/Clang 3.7+ but does work when you pass those additional flags via the CMakeCache.txt file.

    OpenEXR/OpenImageIO/OpenColor/OpenShading are a big family that you guys should coordinate with Gritz on.

    OpenSubDiv still doesn’t build from git source inside Blender. Then again GLEW and other OpenCL additions never were fixed until I pointed out they were broken from the hackneyed Google linked sources.

    If you’re going to drive using a lot of these third party libs you better be prepared to make sure the upstream sources are current with your modifications. Makes it so much easier when building from trunk.

    Best improvement in Blender has been Cycles. Bravo. Fast as hell.

    • I largely agree with you. We are however held back by Visual Studio in the C11 department.
      All the Open* libs are great but i think we could do more there. I agree on the OpenCL comments and we seem to be going there.

  62. I wish Blender have way link file but still can orient, scale, position, duplicate linked objects, just can’t edit their internal structure. If some person change a linked object, automatic update in all blend file use that object, no need delete and reload objects.

    Also, often animators want test animation with temporary objects before have real/final objects. 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.

    This work method allow interation and divide work more easy than current link or add object Blender use now.

  63. Hello all,
    Last time my comment was deleted for no reason so I hope not this time.

    I would really be grateful if we could keep the particle system. I do mostly architectural rendering and I’ve used it for many projects, for both interior and exterior rendering. I’m not sure how I’ll work without it.

    I have only touched a bit the BGE. But I’m amazed what its devoted users can achieve with it. I do agree it needs to evolve. I think that creating one with the advantages of the actual BGE and resolving its main issues. But in one comment above someone talked about paradox. I do not know how it would be done but if this could be integrated to blender that would solve quite some issues.

    Lastly let me add as I saw comments say above, I love how blender is like a one stop shop. Some say its a weakness, but for me its an enormous strength. Modeling, sculpting, texturing, animating, programming, rendering, tools for postpro, and so so much more its candy land :). That’s why for me and for many others, Blender has such a special place in our hearts.

    • This blog holds all posts from 1st time contributors. With one message approved you then can freely post everywhere.
      Thanks for sharing the feedback!

  64. Dropping the blender game engine?? Why?? it’s super important!! that’s why we have all the graphic development done in the past versions for the 3d view!! The sequencer, I think you mean the video sequencer, right??? if you do that’s something that should be overhauled, not dropped away. By the way, Keep the Blender internal, that’s another super useful tool blender has. I mean, not everybody can get neat super graphic cards to use cycles and the blender internal with some compositing can do the job just as good as cycles. Thanks.

  65. especially on linux losing the sequencer would be a huge let down. it may have it’s issues and not be the easiest thing to learn or work with but it is honestly quite powerful and a great tool so i prefer it over the easier but often only good for simple things options.
    if it gets removed i’m sure i’ll keep around an old version of blender until we get it back.

    i’ve never really used the game engine but i did like the thought presented of integrating a number of it’s uses (and logic) into the viewport and having the viewport based on unreal’s shader system. that combined with some of blender’s other features (freestyle, particles, smoke, cloth, physics, sims, compositor, render layers) would make an excellent render engine for those of us without the strongest hardware. even if the other stuff slows it down a bit from realtime having the base visuals that much faster would be great. plus being able to bake PBR textures for game engines and online viewers from blender would be fantastic as well.

    i’m not really sure what’s up with particles and constraints on the development side but they’re working great for art.
    i’d certainly hate to lose either of them as i can’t really imagine blender without them. i especially use particles a lot but i also enjoy constraining plains with previously rendered (fire and special effects) animated, alpha mapped textures to be tracked towards the camera for some simple low cost visuals.

    if you’re going to integrate logic into the viewport would mixing constraints in with that somehow be a viable option? i’m not coder so this could be way off base but constraints seem like a logic based thing at least as far as the constraints that i’m familiar with using.

  66. The best of Blender is able to perform any type of CGI work without the need to change software. Delete features would be a disaster.
    If it were not because it is yet a bit complicated to manipulate images directly in Blender, or even would use Photoshop.
    Remove the VSE would be horrible for me, I use it in all my work, and even some use only the VSE.
    The BGE not use it much, but I see it is a very interesting feature. And some animations are very easy to generate with.
    I love particles, break my heart if they disappear.My proposal:
    -back to to do or better integrate VSE, especially to be reciprocal with the composer.
    – Particles improve please.
    -The BGE letting it run its course in development, but not eliminate, the chances are very great.

    1. In all cases, if the problem is the cost of developing, generate donations for specific lines of each case.
    2. If the problem is continue its integration as addons let them with its own team releases update would prefer them to stay as they are, to disappear.

  67. | The sequencer and game engine are in serious danger of removal…

    …if, by “The sequencer”, you mean the VSE (Video Sequence Editor) aka “the only Open Source, cross-platform video editor in existence”…then, echoing other comments here: I’m horribly disappointed for Blender 2.8’s future. The VSE needs more features, not to be removed! Do all the Devs only care about the modeling aspect of Blender? Don’t any of em wanna enhance the video editing parts of Blender? Blender itself has released several movies…weren’t they edited in Blender’s VSE?…or just modeled/animated in Blender?

    Yes, I agree that Blender is, 1st & foremost, a modeling & animation program (& it’s great at that!)…but the VSE & Game Engine are useful too…& as I just said, there’s no alternative to Blender’s VSE — there’s no alternative video editor.

    The Open Source community is really hurting for a nice video editor, for example…

    * VLMC (VideoLAN Movie Creator) is VaporWare (only a few alpha versions & *POOF*, nothing else)

    * Kdenlive (a video editor for Linux) is totally useless to me cuz it’s only on Linux (sorry, it might be great, I’ve never used it {cuz it’s not cross-platform}).

    …& now Blender’s VSE is on the chopping block! How is anyone supposed to edit videos for free?

    Yes, in the case of Kdenlive, I could download a Live CD & use it that way…but really, if Windows is my Primary OS, do I really wanna: shut everything down, boot up Linux (so I can run Kdenlive), edit a video, then reboot into Windows? No, not really. I’d rather use Blender (or some other F/OS video editor on Windows), even if the VSE leaves a lot to be desired as a real “Video Editor” (it ONLY JUST got the “Text Effect Strip” in 2.76 — & you can’t even change the font!…or position the text “properly”).

    I’d really hate to see the VSE be removed from Blender, cuz that would leave the Freeware/Open Source community with ZERO cross-platform video editors. I would, however, like someone to take Blender’s VSE seriously & make it as good as it could be (to rival proprietary video editors — Blender is already better than proprietary modeling software, great job guys!)…so, technically, I don’t care if the VSE stays in “Blender”, per-se, but the code for it needs to go somewhere, like to a new Open Source project…or maybe the Kdenlive Devs should do a cross-platform rewrite based on Blender’s VSE code? (& the Linux-only code they already have)…& maybe throw in the VLMC code, if any of it is useful. No, I’m not saying that mixing the code for 2 or 3 different open source video editors would be easy, but it kinda needs to be done, cuz as it stands now, there isn’t a “good” Open Source, cross-platform video editor.

    This may be off-topic, but: OBS has already solved the “screen recording” problem, now we just need a program to edit those videos! That’s what I want/need a video editor for: I recorded gameplay with OBS & I was TRYING to use Blender to edit it & add text on top of it. Which is when I noticed Blender (2.75) didn’t have ANY way to simply “add text on top of the video” (how did you do it for the movies you created?), which is also when I found out that 2.76, JUST ADDED that feature. Which is when I noticed that that new feature was unfinished: not being able to change the font & the alignment options don’t work as I’d expect.

    | note that currently comments are moderated, so please, don’t send your comment multiple times.

    …I haven’t clicked Submit yet, but, to mitigate this, you should display the comment to the person who posted it & simply say it’s being moderated (but you may already do that)…as a bonus, you should also let the person edit it while it’s in the moderation queue (even if not logged in). Case-in-point: I’m submitting this blind: I have no preview or any way to know what it’ll look like. I would’ve preferred to use bold in some places (instead of uppercase). (test: bold)

  68. Don’t remove (drop featurings) anything from Blender (sequencer, game engine etc), please, — the current components set of Blender make this software unique in all CG-world. Just improve this components constantly.

  69. I made me first game in bledner.
    If you remove BGE, you remove about 25% of the software.
    I think I now 1 way all of this can be good, make blender as a multi_windows program, 1 of the game alone, and 1 of the Blender render, and 1 for the Cyclus render.
    the Particle System must be able in the Game engine.

    VSE is a powerfull video editor tool, and I don’t i now how we can you blender to make videos and films without.

    VSE and BGE is about 50% of Blender, and I don’t like it when a tool is removed.

  70. These days I’m a professional freelancer and I need to deliver weekly completed animations to my clients.

    The Sequencer is my most important tool and I’ll explain why:

    a) One thing about clients is that they have always some new idea once they see something I make. Instead of re-rendering an entire animation, I only render the frames that are bad and fix things in the video sequence editor.

    b) The client (and me personally) we want to edit sound as we animate, most often I have to match voice and sound effects and the sequencer is a great,fast way to do it.

    c) I can edit scenes in it even without rendering the shots and see immediately how it flows, how awesome and useful.

    d) It’s provided to me dozens to hundreds of reference images to work, what i do is load them in the sequencer and by scrubbing the animation I see them in a small window in the corner, instead of having one windows in top with an image viewer that is always on the way.
    It’s also awesome for when you have a reference video for animation. The movie player on blender isn’t as flexible as the sequencer, I often need to edit the reference as I animate. It helps the creative process.

    e) One work we had consisted in hand animate mouth and eyes to put in the 3D model. On all the lines we had to fix things in sequencer to correctly match the sound, instead of having to re-draw the bad frames I just copy pasted existing frames and re-render it again.

    I make everything in Blender, I don’t really want more software, for what!?
    I don’t have time to open windows, clients are waiting for feedback. I’m a worker, I don’t want to carry an hammer, a chainsaw, a concrete mixer, a screwdriver if I can use just one Sci-fi, all in one, Laser magic do everything tool.
    A few years back this was mentioned to be the goal and was the best idea I ever heard/read.
    For me blender is a complete production tool… This is my dream job :) I’m doing very well as a freelancer thanks to current blender way of doing things…

    Blender Game Engine it’s important too, but I’ll have to leave the explanation to some other time. It’s the fact blender can do anything that impresses the people I’ve promoted it to.

    Best regards
    -Rogério-

  71. Since this seems to have the attention of Ton, I have a question: Why is Blender’s Timeline only for display of keyframes? Why is it not also an editor of keyframes? I come from a background of 30 or so years using Maya, Softimage, 3dsMax and Motionbuilder to name a few of the software I have used professionally and all of them use the Timeline as a keyframe editing tool. I have brought this question to other Blender sites and I was recommended to use dope sheet. The dope sheet is a good editing tool when you want to get in deep into the animation, but it is a bit cumbersome for use as a tool to help blocking animation. This is where you could up to double speed of blocking animation.

    On the topics being discussed above, having made no use of VSE or BGE, I have not much comments about them VSE is unique, in my experience, to Blender. I hope that particles remains in some form. My primary experience with particles has been hair and fur and I love the Blender implementation of this. It is very intuitive and beats the pants off of my experience with Maya and 3dsMax hair for ease of use.
    I am fairly new to using blender, but I have been watching blender evolve for years. I am amazed at what you guys have done that you offer for free, although I have contributed $ to the development of blender because I believe that this is an important project.

    • The dopesheet and the action editor are also “time lines” with scrubbing and editing possibilites.
      Several animators who give us tool/ui feedback come from Maya backgrounds… didn’t hear this comment before. We have an animation system mailing list for people who want to be involved with design and testing – check this url http://wiki.blender.org/index.php/Dev:Doc/Contact

      BTW: Posting design ideas (videos, mockups, texts) on blogs or social media is always useful. It might not have a direct effect, but the good ideas tend to trickle up and get attention.

  72. I want to see the VSE and the BGE receive improvement.
    If we can rewrite BGE and give it an open license that is compatible with commercial production then I’m positive we will see an explosion in its use. The BGE is the only tool of its kind; not only for game development, but for prototyping, simulating and more.
    I will donate money to support a rewrite of the BGE.

    • I second this. If the developers choose to crowdfund 2.8 development, I would surely donate for a new game engine.

  73. In regards of the GE, had you considered making Blender more “friendly” with current GE likes Unity or UE?

    I know many indie devs use Blender to create 3d models for those GE`s, but the process is not as streamlined as it could be. Having a closer integration or at least making it easier for the art pipeline to move between Blender and Unity/UE could be a great deal.

    In any case, if the Blender GE is to be maintained, you still have to consider that the GE environment nowadays is very competitive and most GE´s are multiplatform, flexible and provide a lot of features that Blender GE currently does not (and may divert devs from working on other Blender features that are the most important ones right now)

    • Check the blog posts here about the Viewport project and the 2.8 project.
      For example: we consider to build the viewport based on Unreal’s shader system.

  74. For the sake of clarity, i’d recommend making a prominent EDIT section on the post, clarifying what Ton and Sergey said here in the comments. Otherwise people will just react the same way again and again.

    • I moved my remark to the top of the stack (hacked the date ;)

  75. I am going to take a chain and fasten me to the Game Engine if someone is planning to cut it down and burn it… It is what I am gone do(I wouldn’t do something like that with trees, though:D)…

  76. I don’t understand people complaining about features they don’t use.

    For those who think dropping BGE would represent a saving of money, and that coding BGE is a waste of time, just consider the BGE is mainly maintained by volonteers (not paid) who spend their own time to fix and improve it, because they love it.

    I don’t understand why the fact that the BGE is less powerfull than UE4 or Unity 5 could be an argument to drop it. IMO, it’s a great tool to learn python, make mini games, write and test nice GLSL shaders. I know that Blender is used in some schools in France to learn design, mathematics, programmation to the students. It’s also a great tool for many basic realtime stuffs. And of course a cool hobby for thousands people.

    If Blender internal needs a rewrite, and makes the BGE incompatible, I’m convinced:
    1) That BGE devs will do as best they can to adapt it to the new code.
    2) That the BGE community could mobilize to a fund raiser to hire new devs for a periodic employment contract.

    Anyway, fortunately, the BGE won’t disappear tomorrow. You could still download 2.7 version and enjoy :)

  77. Please do not drop neither the BGE nor the VSE!
    I’m not an artist and I am just learning to use Blender and checking out the scope of it.
    For the BGE I suggest an interface to other programs to transfer values to the BGE sensors and back. This programs could be e.g. chatbots, databases and logic languages like PROLOG.
    The particle system needs also some improvement, e.g. the colliding of the hair particles does not suffice. So I understand your wish to overhaul the particle system.
    I use AMD Graphics with OpenCL 1.2 and OpenGL 4.4.
    It is possible to query via OpenCL the graphic card and use features beyond OpenGL 3.2 for newer graphic cards. The same is possible for OpenCL. So it might be possible to introduce new shaders to the cycles render engine without dropping the support of older ones.

  78. The discussion the developers had was focused on the code and guided by the code, not by usages metrics.

    It is good to also remark that removing a piece of code does not imply there will be no replacement or that there are more or less features.

    There is no consensus for dropping anything, but from a coder point of view there are issues with the VSE. For the GE there could be a lot of overlap with the new viewport. but that in itself is not enough to remove it.

  79. I think dropping some parts of Blender to help bring it forward are an excellent idea. Although, it would be sad to see the VSE going. Even if there are alternatives.

    Because you are doing such an overhaul of the system during the 2.8 project, have you thought of rebranding it Blender 3.0? This would define a clearer break between versions. Then you could make files forwards compatible from there.

  80. Please don’t drop video sequencing in Blender. Rewrite or refactor it if you must, but dropping it as a feature means breaking a promise to your userbase of having an end-to-end 3D content creation suite. It’s no good being able to render something if you can’t finally create a full movie at the other end, complete with sound.

    Video sequencing doesn’t need to be fancy but I feel it needs to be there in some form to cover the _entire_ pipeline.

  81. I find it better to ditch older outdated features and maybe replace them with more modern tools.

  82. So, I guess seeing comments posted up to 3 days later than mine showing up, and mine not showing, I guess this tells me i’m not welcomed here. Respectful Blender user for the past 6 years, supporter for blender’s projects, hundred of dollars spent, OK, i’m no part of your club. It is Clear. Thank you, i’ll go spend my money elsewhere, i guess. Alienating people for no particular reason is not a good decision. Any simple explanation would of been really appreciated, any really. Even trough email. I think I need to talk to Ton then.

    • Sorry, the moderators here all coincidentally were not checking the queue for 2 days. Now it’s all back.

      • Thank you very much Ton for your help and for your time.
        It is really appreciated. I was wondering what I did do be put in the corner wall punishment :D

        Thanks again.

        Best regards, Aclariel.

  83. About Blenders Game engine:
    maybe a solution!?
    if technically possible integrating and using “Paradox Game engine” in Blender and using blueprints -logick bricks node editor system like Unreals Blueprints !
    its going open source
    http://paradox3d.net/

  84. Noise, noise, noise. What about Cycles noise? You’ll talk about this trouble?

  85. For my part, the VSE is very important. I’d also like to see some integration between the compositor and the VSE (as others have already said). I’d be willing to suffer through a partial or even complete rewrite if that was needed to bring it more closely into the Blender workflow.

    I’ve stayed away from learning the BGE because its future was previously stated to be uncertain. If a rewrite ends up happening, count me as another potential BGE user (FWIW.) For my purposes at least, the licensing issues don’t exist. But I suspect that figuring out how to handle distribution and licensing issues for commercial applications is probably going to make or break whether the BGE should continue to live (please correct me if I’m wrong about the licensing being an issue. I’m somewhat ignorant of the issues surrounding distributing games.)

  86. BGE & VIDEO EDITOR SHOULD NOT BE REMOVED!! INLCUDING PARTICLE SYSTEM!!!!

    – First of all lets make this clear, BGE is actually a great engine for beginners, it doesn’t need programming, and it can achieve some cool graphics close to unity, but many people mostly care on gameplay not graphics, minecraft is an example of shitty graphics, but good gameplay, and made with shitty software, and hey look! it made milltions of dollars. BGE doesn’t have to be perfect, just continue to add some cool things, but not challenging for developers, because they will get frustrated with their work, they don’t have to rush. They drop features, In my opinion because they probably didnt write efficient code that can be understood & be edited easily in the future, which leads them to tremendous amount of time to add features n stuff.

    – Video Editor. Video editor isn’t great as others like premier or vegas pro, but personally I use it a LOT for making video tutorials, and I don’t want to spend huge amount of money on my videos, it could affect some people who use it to edit videos, and they have to pay huge amount of money to adobe n stuff for videos, there is nothing special about video editing its just about cutting, fading, volume control, and some basic stuff to make professional videos.

    Concluding:

    JUST BECAUSE NOT MANY PEOPLE USE THAT PART, DOESN’T MEAN ITS BAD, BLENDER FOCUSES ON MODELING, BUT HAVING SOME OTHER STUFF WOULD BE NICE AS WELL, WHICH MAKES IT MORE SUCESSFUL

  87. Two posts waiting moderation approval written yesterday Nov 2nd.
    Can someone explain why my post always needs to be waiting for approval?
    Just curious. I’m a respectful user, been using Blender for years, paid for supporting its development, bough material from the store. Is there something I am missing?

  88. Was going to say something about the ‘complete pipeline’ being important, particularly at 3Dami, but I think that is understood, especially given Sergey’s comment, so not too worried:-)

    But on the subject of the game engine (which I don’t use!), my solution would be a plugin system. Right now you can have an external renderer added via a plugin – it can appear in Blender, fully integrated with a customized node editor for the materials and even do view port renders. Why not the same with game engines? Plugin per game engine (open source or not), that gives you customised tools, e.g. a logic editor with nodes, a text editor for game engine specific scripts etc. The Blend4Web guys already seem to have done something like this. I would then ‘pull a Cycles’ – separate development of a second game engine with far more modern features, from scratch, with a more convenient license (Apache again?), make that a built in option for Blender, or even a one-click install if its too big for the default download, and watch as everybody stops using the current game engine, and doesn’t care when you remove it entirely (Ok, Blender internal is still a valuable part of Blender, but I think the game engine would be more amenable for this).

    Just an idea, but thinking in terms of ‘default plugins’ improves flexibility, keeps everybody happy and allows the development of Blender’s crazier features to operate with some independence. Plus it would then be a safe bet that people like Valve would get (them or community) a plugin for their Source game engine that works like that, and that super tight integration with immediate feedback would make lots of people very happy, myself included as I want to run a computer game variant of 3Dami:-)

    • ^ I concur and agree with Tom!
      plugin-system much like how blend4Web does it..
      or Cycles approach (separate development);

      also excellent idea on that 3rd party companies could make plugin of their engine to “blend” in blender..

      Imagine drop-list of following engines:
      Blender Render
      Blender Game (external); Apache? [with its own custom add-on system]
      Cycles Render
      Renderman
      Blend4Web
      Blend4Unreal4
      Crytek-B
      Blender Unity Unity4Blend
      Processing
      Blender VVVV
      Source Game Engine
      Godot
      Blender VR (for StarVR, VIVE, Oculus, FOVE, Morpheus, ..)
      Hadoop4Blend
      Scilab4Blend

      3rd party engines and more.. (all hooked-up like plug-ins and/or addon)
      Houdini4Blend
      Reaper4Blend
      Blend4Live Live4Blend
      Cinder
      Enigma Development Environment
      … and more..

  89. Remove VSE?
    But it was exactly what I thought it would improve.
    it makes no sense

    I thought the proposal was “break with release cycle” to improve the blender for film animation and games

    Remove VSE and BGE does not seem to improve it.
    Will have another video editor instead of VSE? Will be another way to edit videos in the blender?
    If not, this sounds crazy, it makes no sense, remove and leave no substitute will not improve blender in absolutely nothing.
    If VSE will not receiving improvements, so keep it as it is.

  90. PLZ BLENDER DEVs, DONT LET GO OF THE GAME ENGINE. There are THOUSANDS of bge users who are making games and will update blender. They only want development in bge, thats what u’ll do, but plz DONT drop the bge, particles system and plz bring the ‘mirror’ option of the blender render to bge (if u can, if cant— NO PROBLEM)

  91. not destroy this perfect software. Remove things just because they give work is not the best way to get rid of that work. BLENDER GAME ENGINE. I want to help prevent it from being removed. How can I get to the Game Engine development team?

  92. I don’t mind dropping VSE, if it is coming back and being better. If it’s not coming back later, then it should be forked to be a standalone program. I really think it is that good. Worth of saving and improving.

  93. My thoughts on this:
    -Collada can probably go. This might be necessary to import from Google Sketchup, though, last time I checked. (That was years ago.)

    -The VSE is a niche feature but an important one. Keep it and have a developer or two work on the needed upgrades.

    -Game engine, particles, constraints and physics all seem to suffer from problems with integrated systems.These present a complex problem, but I think the community is already kind of looking at the solution: nodes. We’re seeing new node systems pop up, such as the Blend4Web logic editor and animation nodes project. So here’s where I could see that going.

    -Make particles, logic, constraints and physics into a node-based interaction system(maybe modifiers too), and improve those systems where they are broken to provide what users need in the process. The goal is to create a streamlined pipeline for internal workflow. Some of these systems might need a complete rehash in the process, but churning up the waters will be worth the result

    -Provide more support for the Blend4Web project, as an semi-universal external player for interactive media created using this system.

    -Improve integration with outside software and standards to make multi-software workflows easier, almost seamless.

  94. First things first, it’s nice to see Blender moving forward! Keep up the great work!

    That being said, I can’t help but agree that there’s some controversy in the original post which had better been clarified in the post itself as opposed to the comments… I mean, the way the post reads to me is, for the most part, “for 2.8, let’s bump the system requirements and drop a bunch of useful features, including particles, constraints, VSE and BGE, among other things”. The original post says very little about what the dropped features are to be replaced with, if anything at all, and whether the said features will simply go into oblivion or will be rethought in some brand new (and no doubt much better) way.

    Obviously, simply seeing a lot of good stuff disappear from Blender with no replacement is a horrible perspective. While some of the named features may be controversial and/or not very useful to an average Blender user, at least in their current state, many of them are useful on a regular basis (particle systems, for instance, which in my mind includes hair/fur creation as well; VSE which may not be very useful as a video editor but which is awesome for creating videos from frame-by-frame PNG renders and syncing them to audio and whatnot; constraints, which include the IK constraint that is very useful for character animation, etc. etc.). It would have been much better, in my opinion, to clarify that features like that are not planned to be simply dropped but rather “reinvented” in a better way (as Sergey clarified in one of the comments above).

    Dropping major features and not replacing them with better/alternative ones may have long-going implications and may turn away potential users, even long-time users. There is actually very little in the original post about what will be improved barring some rather technical details (OpenGL, C++11 – things like that may not communicate well to your ordinary *artist*, at least not in the sense that they translate directly into “usefulness for the creative process”). At the same time, the focus of attention closer to the end of the post seems to be on removing stuff, which seems to hint that Blender is about to become a less useful tool for an artist than it was before.

    Sorry for the little criticism here, guys, I know you mean well and I’m sure you’ll find a way to make Blender better for everyone! I wish you the best of luck in the 2.8 project!

  95. A major issue I, and I expect a lot of other artists, come up against is the incongruity of the various physics systems.

    A very welcome refit to start with in 2.8 would be to drop the discrete soft body and cloth systems, replacing them with the Bullet library’s soft body simulation. This way, rigid and soft bodies would reside in the same “world,” and could seamlessly interact with each other, which would be much more intuitive and logical to the artist than the state of things today. This would also allow a number of redundant parameters to be unified. Furthermore, better leveraging Bullet would put physics simulation in a better position to have its calculation handled by the GPU in the near future.

    The particle system is invaluable, and needs attention rather than removal. Particles are how artists will interact with and control point clouds. Between Dynamic Paint, SPH, Pyroevil’s add ons, Freestyle and Bullet, the particle system has more potential than ever for use in, or as a lightweight alternative to, physics simulation, and maybe more importantly in motion graphics and NPR. However, it very much needs more of itself exposed to the user, and in a more intuitive way than the imposing expanse of shifting and arcane-looking parameters we use today. During the shift from 2.4 to 2.5 the particle system was actually hobbled in some ways, such as the excising of reactor particles, and just hasn’t had enough attention paid to it since. Dusting off and completing Phonybone’s particle nodes project would likely simultaneously solve most if not all of the system’s current drawbacks.

    When it comes to physics and particles, “integration” and “consolidation” should be key words for 2.8 and the future.

  96. Sounds nice. Can’t wait to see the new blender. I would still like to see the video sequencer in the future. If not, I still have older versions of blender that I can revert to. I’ve love blender for its bulk of capabilities so I don’t have to switch programs.

  97. I can definitely live without the BGE, never saw a need for it, there’s so many great game engine alternatives out there.

    Particles, definitely need. revamp as needed.

  98. I don’t think drop the sequencer and the BGE will be such a good thing. Manly because the BGE is getting known within the game developers, we use it for prototyping just because it does not support as many platforms as we need, but it is indeed an awesome prototyping platform. Blender is an awesome software manly because it is almost all-in-one development kit, I liket the idea that I can be good in everything instead of awesome in one thing. Please, invest a time on the BGE, rework, rethink the idea. I think, with love, that you have to rework the BGE, not drop it. Think wisely, Game is art too, and it is growing <3

  99. Please don’t drop the Game Engine, it’s very important for realtime interactive 3d! please….

  100. Please don’t remove the sequencer: or just fork it into a separate “video editor program.” I use your sequence editor as my editor of choice, and it is the best free editor for windows out there! Drop the game engine, if anything.

  101. Thank you Sergey for your clarification.
    Don’t want to loose most of those features we love so much.
    Like others, my entire workflow is base around most of blender’s current features.

    Regards.

  102. Ok I’m not a technical Blender user, or a professional, I’m not even a good amateur, but I am learning and I never add my voice to stuff, but I am now. I read the comments and to hear that they are talking of dropping the BGE and VSE that was to much, drop it for what, it may not be perfect, so what, neither am I.
    I hear people saying Blender is no good, well I cant compare other software, because I have not really used other software like 3dmax or V-ray.
    But I still see images and films made by other software and I admit some of the work is fantastic, and yes I heard someone this week bleating his gums that Blender artists were not up to scratch, but I also see Blender users produce some work that is fantastic to and in lots of cases even better.

    Only this week we had the video of the sparrow flying. I went to and had a look at artstation, and like every other forum there is good and bad work, we no different there no better, but I had to question some of the work I saw, £9000 of software to produce something that looked like it was made in windows paint.

    So it goes to prove no matter how much software you got, it not the software that going to make you better, its you that going to make you better by learning the software.

    And you hear people say about Blender, it got a learning curve from hell, well you want to be good at anything, you got a learning curve from hell, two click software is not going to get you what you want, the most expensive chisels will not make you a good carpenter, skill is what does that. And that comes from learning and practice and more practice.

    People say its pipe line isn’t that good, what they mean to say is, we used to what we already using.

    People say that it interface is to complicated, my god didn’t that Aussie bump his gums on like a whining pommie, well I suggest you go and look at some of the other software used in industry. my mum is 73 and she didn’t start using a computer till see was 65, and she uses home edition digitizing software for embroidery, you think Blender got a learning curve from hell, you should see the interface on that. And that is another world that is full of people who want two click software to to produce magic.

    It don’t happen, you want magic you got work at it,

    Now we hear that we may be dropping stuff like the BGE and VSE, oh my god, has Ton been smoking. Why? Have the developers allowed those that want two click software to cloud their better judgement, influence how they work.

    It is true that most people who use Blender started with it because it was free.
    It is true that a lot of people then drop it because they think it difficult, when i first downloaded it, I looked at this square and went yep what now.
    Does it crash, if I don’t blue screen my machine at least once a night I’m not pushing hard enough.
    But it also true that those that stick with it, found out that Blender can do magic, as long as your willing to work at it.
    Many people start with Blender, listen to the hype, Maya is better, V-ray is better and swap over, maybe they are, but they forget that what make you better is you, not because one program uses predefined assets.

    The thing about Blender is you can take a square primitive and go from nothing to a full film and never have to leave Blender at all,

    I see people doing archiviz, in 3dmax, trees from whoever, using railclone, then export into v-ray,

    I can do the same and never leave Blender, and some people are producing as good if not better results.

    We have cycles, people complain it slow, so what it powerful, yes there other render engines, but are they as flexible, can you plug a node into the wrong socket and go, that a different result but I like it, and people write new nodes and code to make it faster, better stronger. I believe it one of the few that is a brute force raytracer so it can produce true photographic images. Wonderful.

    The game engine, ok we cant make block buster games because of the copyright law or what ever it is. Ok then call it a visual studio, and let people interact with objects, which is what we sort of thought that was the route that Ton was going to go down,

    You will have the archviz people dancing in the street, even if it is 3d.

    The VSE, I’m totally crap at animation, I make square balls and hollow solids, but I’m saved by the VSE, I do an animation, I see that my thumb is wrong or my walking cycle, I can go back in 3 clicks to the animation edit, fix it, re-render those frames that need to be done and then check it. All in one program, wow. No exporting, no changing file systems, no losing textures. Who what’s to lose that flexibility.

    I’m working on a still scene right now, it 13 gigabyte in size, I use a laptop, has an I7, but I only use 6 cores, Blender allows that, and it stops it from over heating, I have a GPU 650 with 1 gb of memory, and the laptop has 6 gb of memory. so I use scenes. Scene one is the main scene, and it always using the CPU to render, it holds the basics lights, and a box that the camera can work on, but it invisible. the next scene I can set up to use the GPU, so if I want to do a cup on a table, scene 2 will have the cup and scene 3 the table, and both scenes can use the GPU and I can set the border boundary just to render that part.
    This is all in the same file, Blender can use the CPU and GPU, and have different settings, different global illumination, linked lights, different render cycles.

    So I got about 8 scenes so far, and the compositor uses the CPU, to stitch it all back together, to me that is amazing that I can do that, but more amazing is that the developers allowed Blender to be that flexible that is can actually change what it will be using to render and composite on the fly.

    I understand hardware is becoming more technical, wont use old code and we got keep up, and there only so many developers working on it.
    Yes we have have thousands of people making add ons and stuff and that’s great, and it was Blender than made me think, I want to learn python. I want to code. I can say hello world.

    We not Pixar because we cant cram all the riggers, artists, developers, who ever into Ton’s office,
    but people who rave about I realised I wasn’t such a good artist, when I walked though the doors at Pixar forget that if it wasn’t for Disney saying make a full animation, they were still making shorts, and that if it wasn’t for a large cash injection from Steve Jobs, they might have been another studio that went under.

    Blender is alive and kicking, has a great community, PEOPLE who share and the software, may have a learning curve from hell, but so what, it needs to be updated I agree, but do we need to lose some of it functionality, I don’t want V-ray, or 3dmax, I want Blender, I’m a Blender user and proud to say so, and I love it that everything I want to do I can do inside Blender, I still got loads to learn.

    Improve it yes, make it faster yes, but take stuff away, no way, I’m sure you can find a way to make it work.

    I Watched the videos from lasts year Conference and what got me was not just the artist or animators using Blender, but the vast group of people from archaeologist to scientists who prefer to use Blender because of its flexibility and it functionality and that everything you need is there.

  103. Do you kill videosequencer after Cosmos Laundromat? No please, I think that its a powefull tool that make the difference. Thanks!

  104. Thank you Sergey for your rectification.

  105. Dropping a sequencer is a big NO-NO. It’s really usefull, just leave it as it is.

  106. so just to clarify collada support is not being dropped?

  107. I am really scared of the idea of dropping the BGE cuz man. there is a comunity of thousands of people that use it (including mysefl). There are FB groups specially made for people that just use the BGE and i see in this comment section that alot of people preffer taking out the BGE without
    knowing that, that part of blender is pretty important for us cuz we like it and we don´t upgrade to UE4 or other engine cuz man. WE LOVE IT!
    I just hope that the BGE will remain been improved :)

  108. another VSE thought: Blender and open source offer digital artists a level of security that corporate packages do not – it’s hard to be motivated to build a small studio and commit to long-term career building around a specific toolset if a package can be whisked away at any moment –

    Adobe Creative Cloud is guilty of this – focusing on a specifc toolset is important if one wishes to become proficient – take Photoshop: after years of use one becomes proficient but, at least in my case, the bells and whistles are incidental to the core functionality – v7, v12, CS5, CS5.5, CS6, my basic use of Photoshop hasn’t really changed for decades – and then along comes Creative Cloud and removes accessibility for a whole class of users –

    this is not to say one can’t take a left turn into open source but, fingers crossed, open source will not also remove accessibility – for me, the sense that all the tools comprising Blender will be there tomorrow makes it possible to lay out long-term projects with confidence and commitment –

    the VSE may be simple and no-frills but it is reliable and, I believe, contributes (along with the rest of the package) to independent artists having the confidence to invest time and heart as they build their projects and studios and workflows –

    the idea is that, through repeated use and familiarity, digital art tools get out of the way and become transparent – if those tools also come loaded with the potential for inaccessibility, mounting unease builds as one invests years and heart in their use (and the developer becomes demonized!) – this can harm confidence and motivation and perhaps affect goals and aspirations and goodwill –

    maintaining the VSE (and perhaps adding formats as things develop) would be an ongoing gift to independent studios and artists such as myself – preserving it as a dependable rock solid option would be a wonderful outcome – cheers, SN

  109. Suddenly I’m no longer expecting for blender 2.8… but those dropping feature are killing me, specially VSE :'(

  110. I kind of agree Blender is literally a superweapon in my toolset because of all the features that ot has just in it. Dropping any f them would be a massive backslide for all of blender. The software has become my almost mainstay for anything need because I know that blender will get the job done. I think dropping features in blender should not even be considered the software is the powerhouse that it is because of everything in it. No i doesn’t whoop on after effects in video editing but the truth is its easy to dominate an area when your software is mega specialized for what you are doing but lets face it Blender can get the damn job done forgive the french but I love that knowing that when I turn to Blender its going to close the deal because it just can. Changing anything and removing thins would change this and weaken Blender alot. Sorry just my two cents.

  111. Removing some of the key features mentioned in the article is a very bad idea. Example, we use VSE to key frame all animations. We love the ability to import audio, create key frames for lip sync, and use reference video for animations. We can do an entire sequence in less than a day because of VSE. What other FREE and OPEN SOURCE software can do all of this on Linux? NONE to our knowledge.

    It would be better to fix/improve what’s broken or improve what needs attention than to arbitrarily remove features so many depend on. If not, then any users will end of leaving Blender in favor of commercial solutions.

  112. I think particles and constraints definitely need a rewrite to allow for nodes and chaining multiple particle setups together (like MODOs particles). I would also argue that Blender Render needs a rewrite, as there needs to be a flexible and fast raytracer people can use when they need faster or more stylized renders than those Cycles can provide. Blender Render at the moment feels so messy and inefficient when compared to other raytracers I find I can’t really use it for anything except the Freestyle integration.

    I think it’s also really important to get rid of any features Blender does not require to focus the development more on the features people need, personally I see the game engine as an unnecessary waste of valuable time and resources as there are much better solutions out there for people who seriously want to make games.

  113. Whoa! VSE is the best free video editor for Windows. New features had been added recently, why drop it?

  114. I’m really looking forward to what comes out of the 2.8 project. I am hoping it will bring a great many improvements to parts of Blender that are lacking so that Blender can stay competitive with current commercial packages. I say competitive, but as Ton mentioned in his talk, also complementary. Blender is unlikely to replace dedicated tools like ZBrush or Substance Painter as an industry standard, but it can certainly complement them and fit into artists’ tool sets if it gets developed in the right way.

    These are a few things I’d like to see come out of 2.8/Viewport (and later perhaps):

    1a. An massively improved real-time viewport that can use PBR materials created by tools like Quixel and Substance along with some default shaders that make using such materials easy.

    1b. Please give us proper viewport transparency sorting too :D

    1c. Please do NOT remove wireframes! Rendering with GL_LINE drawing is unimportant, but the concept of being able to see a wireframe is incredibly useful to a massive section of the Blender user base – anything where topology is important. I guess Ton was just ranting about Open GL wireframe drawing, but he sounded serious in his intent to get rid of wireframes…

    1d. It would be AWESOME if we could make shaders to create our own rendering modes and add them to the menu or replace the existing modes. Currently we have Rendered, Material, Texture, Solid, Wire, Bounding Box and matcaps. If you expose just a few parameters to the shader (a couple of object properties, selection status, mode, layer info etc.) then we could make shaders to help visualise things unique to our particular workflows etc. and share them with other Blender users. Think of it as matcap on steroids.

    2. Setting up materials for BI is not very intuitive (e.g. normal map, alpha channel, matcap style normal based lighting with texture etc.), so it would be great if this saw improvements too. Perhaps materials could be created with nodes, and we could specify inputs in the nodes that are accessible from the Blender UI materials panel to make material setup not require knowledge of the node system?

    3. More of Blender opened up via the Python API and nodes. For example, the ability to use Python or nodes to make modifiers. If users can create modifiers that people want, then there is a good chance the code could be ported to C++ later if the performance isn’t good enough. Setting up the dev environment, learning c++ development and making a separate Blender build is a huge barrier to entry for people who might otherwise be able to make a good modifier, and even if they do make something cool, the audience for users willing to try the modifier is much smaller than that of a script. Please try to open up more of Blender via the Python API and nodes if there is a good opportunity to do so during 2.8.

    3b. I REALLY hope that the manipulator/widget project will be completely accessible to scripts so that we can create our own manipulators. Give us some nice API functions to simplify some of the trickier math stuff we might need, and write some good docs for it.

    4. Nodification of more systems. This has been talked about, but I’m looking forward to the animation system being fully nodified. I saw Ton reach out to the creator of animation nodes during his 2.8 talk, and judging by the response to AN in the Blender Artists thread, it is deserving of full integration. If this was approached with care, and we also got some nodes to deal with input sources (keyboard, mouse, on-screen buttons, websockets or similar to allow arbitrary data input) then users would be able to do a lot of the things they do with the game engine. (I’m not a BGE user, so sorry if I am totally wrong here).

    5. I really love Blender’s OpenGL UI. However, I’ve heard that the UI code is one part of Blender that uses a lot of legacy OpenGL code, so I guess it will receive a lot of attention during the 2.8 project. However, please do not drop it for some generic UI Toolkit that adds dependencies and loses some of the unique abilities of the Blender UI. It feels great, but it could use a bit of work to make management of panels a little easier, such as swapping panels, some kind of way to pin functions to parts of the UI (like maya’s toolshelf) and perhaps collapsing of multiple panels at once (you can’t drag one panel onto another if the boundary edge doesn’t perfectly match).

    6. A better layers system. An unlimited number of named and coloured layers, with freeze/disable, hide, isolate layer functionality.

    7. A hugely improved Outliner with real object grouping. Maya’s Outliner is not perfect, but it’s a lot more intuitive and powerful than Blender’s, and something like it would be a great starting point. Batch object renaming is fairly easy to code, and would be a great feature to have too.

    8. Hopefully the custom normals system will finally see final implementation with a real editor. This would go well with the viewport, as they are both things very useful for making real-time content.

    9. Layers in the texture paint tools would be amazing. Maybe a lot of work? :)

    I realise some of these things are probably not even on the table for 2.8, and even just the things that have been proposed are a huge amount of work for the small development team and community. You guys are all doing an awesome job and I look forward to watching 2.8 develop.

    • Pretty much all of 1 (a-d) and 2 are planned for 2.8! 1b is the only one I’m not sure about but it is a goal.

      Regarding 5, don’t worry we’re not moving to another UI toolkit. Still going to draw with OpenGL. Just need to update the hundreds of places that draw UI code.

      6 & 7 would be awesome! I hope someone tackles grouping/layers/organization of content.

  115. When Ton said at BCon15:

    “And of course, no more wireframes! Shocked? And why not? Well, wireframes are the stupidest things from the ’90s, the 80s… wireframes are stupid!” …was that a joke?

    I use wireframes every day for my technical animations, so if they were to be removed Blender would need a beautiful new viewport object x-ray / transparency system with edges/vertices overlaid on top.

  116. No not again the fear of dropping the BGE, one of the old unique features of Blender… I agreee that there are better engines but for speed of prototyping and techning people realtime graphics nothing comes close.

    And the VSE is also very usefull, even in times where we have HW accelerated NLE systems which produce a great video with a click of a button “Make Magic Film”. hahah.

  117. I love the VSE , would be a shame to lose it.

  118. Oh no! Please not the VSE! It’s really the best video editing software out there for Linux. All others I tried just couldn’t compete. I totally understand that it’s hard to maintain so many features, but as long as there is no good alternative it would be such a pity to drop it! If you know another really good open source video editing please let me know. For compositing I use Natron, but what to use for editing then?

  119. Blender game engine is much more handy than Unity or Unreal especially for indie developers. Don’t remove it please.

  120. Hey-hey, relax, guys!

    There’s some quite big difference between dropping feature and dropping the code. Sometimes the legacy nature of the code makes it so difficult to improve functionality that you need to spend some time re-thinking some area design and re-implement it with keeping new features in mind. And in order to do so it’s easier to fully remove code in the 2.8x branch and then bring stuff back together.

    Such approach was already used for 2.5x project and for some smaller projects like depsgraph where we literally took blender fully apart and then brought it back together.

    It’s also to be stressed that we’re still strong belivers in “don’t remove anything without bringing something totally new ad fresh to artists” rule.

    Hope it makes it more clear what’s the actual intention was in there.

    • I agree, but the article last point is “Dropping Features”… Blender IS awesome in space just because is SO versatile that thinking of “dropping features” is just a bad idea imho, if the main point for soing that is “there is something better around”. That could (arguably) be said also for Blender modeling features, or node system…

      I think most users could even stand only 1/2 releases/year, instead of up to 4, but keeping all the features “in” (that would also help documentationa s bughunt, always imho).

      • It’s quite fuzzy title of the topic, but hard to think of 100% clear name for me either. While for VSE it’s really just dropping old code to replace it with better one, OpenCollada library is to be dropped. It’s 300meg library (prior to strip) and is mainly used as just an XML parser. We could do it from python (in fact, there is already WIP collada addon in python) and save maintenance time spend on this library.

        OpenNL library is to be just removed and replaced with the solvers from Eigen library.

        BGE is a bit controversial topic. It wasn’t really pushed forward for a quite some time now and the idea of extending Blender’s animation system with actuators and sensors idea might just make BGE obsolete in terms same exact functionality will be available from “regular” Blender.

        Particles will be fully reconsidered. Some it’s features will be removed in favor of instancing nodes (which is just om just an ideas level). That would make it easier to have flexible instancing and will make particles themselves much easier to maintain and use area.

        Such clarification should have been in the original post IMO, but i didn’t have time to edit it before it went online..

        • Could you comment on “no wireframes” speech? Was that a middlefinger to blender users about useful and already coded feature coloured wireframes that will never get in master or was Ton serious in removing wireframe mode alltogether?

        • “While for VSE it’s really just dropping old code to replace it with better one…”

          Sergey, don’t you think that this sort of thing deserves more visibility? Because until your comment here the public knowledge was that 1) VSE has no maintainer at all, 2) GSoC 2013 project to rewrite bits of it was incomplete and not going to be merged.

    • Thanks for the clarification Sergey.

      I was worried about the VSE, I’ve cut, shaped and re-encoded a lot of video with it over the years that I’ve been VJing. It’s the backbone of what I use Blender for even when I’m not working in 3D and what I use to put together records of performances.

      As long as it stays even if it doesn’t get worked on for a bit I’ll be a very happy bunny.

    • Thank goodness! I love the BGE and the sequencer, but sometimes things have to get worse before they can get better. Blender is THE most amazing software on the planet. I trust the dev team will come up with a good solution to the GE and sequencer problem.

  121. Personally, I won’t miss BGE (a better interface with Unity and/or Unreal would do a better job for game developers), but dropping VSE? Might as well drop all animation and/or FX features: they’re useless, really, without it.

  122. When I heard Ton at BCon15 talking about removing wireframe, I was horrified. I use wireframe *every day* to align objects in my technical animations. Wireframe must remain as an option or I will not be able to use Blender any more.

    The Blender team must remember that there is a very large number of ‘technical’ users. We are not ‘artists’ making squidgy/sculpted cartoon characters and fluffy organic models – we are technical/scientific/CAD people who use Blender to visualise and animate industrial hard-body models. We couldn’t care less about BGE or sculpting, but must have wireframe and the VSE!

    • Hey, I was just making fun of the topic :) Wireframes have a use, but we also have to look forward and invent 3d tools using full high quality near-photorealistic real time rendering. There is the future.

  123. The BGE is the only feature of BLENDER that allows an end user to make dynamic inputs that can alter the way a BLENDER “work” behaves. Animations are all well and good and fun to watch but they are inherently passive. They cannot be altered or changed in any way by the viewer. Eliminating the BGE eliminates from BLENDER any possibility of user interaction, such as simulations, scientific modeling, educational tutorials, and, of course, games. Rather than dumping the BGE, more effort should be spent updating it — just my opinion. If I had the skills I’d offer to do just that.

  124. from http://www.blender.org/about/:
    “Blender is the free and open source 3D creation suite. It supports the **entirety** of the 3D pipeline—modeling, rigging, animation, simulation, rendering, compositing and motion tracking, even video editing and game creation. ”

    Now, I thought 2.8 was to modernize **the code**, not to remove features because is difficult… and so make blender a lesser thing. All this years spent promoting Blender as the only way to keep (if wished) all production in the same file and same program… to the trash?

    Even in the BC15 videos stated that the words “free”, “opensource” and “complete” was the definition of Blender.

    No, I vote to keep features. Maybe start with the others, but the long plan should be to keep them all in.

  125. (note: for odd reasons my comment got truncated from using arrows for emphasis)

    Blender Game Engine is essential!
    here some solutions:

    1. turn blenderplayer GPL to BSD or follow lead of Unreal4 with a custom EULA..

    2. make game engine modular (have its own addon-system): hardware, controllers, physics, libraries, render(PBS/PBR, sprite, etc..)

    3. meta-logic bricks, game logic nodes into a custom logic-brick (just like python is a custom brick for programmers; nodes of game-mechanics, etc.. turn into meta-logic brick)

    4. again make it modular: forking the game-engine may work, but in this case have all of blender as seperate blocks and one can make custom pipeline of blender: modelling, animation, simulaiton, painting, texturing, game-engine, sequencing, compositing, etc..

    5. blend-script (why does it necessarily have to be constrained to python for game-engine; or if so, then python C#,javascript or C++

    I agree that as long there’s a tactical reason to detach some components and then further reattach when it gets refined/mature then that would be a nice plus.

  126. Blender Game Engine is essential!
    here some solutions:

    1. turn blenderplayer GPL to BSD or follow lead of Unreal4 with a custom EULA..

    2. make game engine modular (have its own addon-system): hardware, controllers, physics, libraries, render(PBS/PBR, sprite, etc..)

    3. meta-logic bricks C#,javascript or C++

    I agree that as long there’s a tactical reason to detach some components and then further reattach when it gets refined/mature then that would be a nice plus.

    • The game engine being GPL really hinders distribution of popular platforms — Apple comes to mind.

      We can’t just relicense it under a permissive license to get around this. What about all the people who contributed to it as a GPL project? What about the rights of people using it under GPL? It’s messy. Unlike Cycles which is Apache licensed and can be brought into other projects.

      Like other folks have said, making Blender work well within another game engine’s pipeline seems to be the way to go.

  127. Dropping the VSE would be disastrous for my pipeline, since I’m using it a lot for quickly setting up animatics and tweaking them. Having it in the same environment with the 3D, Grease Pencil, Compositing and Animation environment enables me to work in an extremely fast and powerful way…

    IMO Blender’s strength lies in its variety of tools as a Generalist software. Even if the individual parts of it might not necessarily fully compete with some of the specializing software out there (some actually do compete easily),it all adds up to more than just the sum of its tools.
    Great for freelancers and small firms!

  128. For me Sequencer is part of Blender that keeps me using it. I know of people who use Blender solely because of Sequencer.

  129. While I’m excited for 2.8 overall, I have a few things I’d like to say after reading this blog post. Whenever any project comes close to a major rework, it is natural to see some extreme radical ideas.

    I think blender has come a long way since the 2.5 Overhaul. For me, blender is a software unique in the way it can be adapted to various workflows, and play several different roles in a pipeline. For this reason, I believe decisions regarding dropping any existing features or capabilities need to be tackled with great caution.

    Of course, I’m talking about the VSE. I myself do not use the VSE all the time, nor is my primary reason for using blender; But the built-in VSE is still my first preference for editing animations or image sequences. With the recent developments and improvements to the Grease Pencil, having a VSE opens up a whole new 2D/Traditional frame-by-frame pipeline!

    I think any animation software is incomplete without some editing functionality, and the blender VSE is one of the better implemented examples of this idea. It does need improvements in efficiency and usability, but considering to drop it altogether seems to be a bit of a premature overreaction. Suggestions of the same will just lead to general panic.

    On the other hand, the BGE is a completely different story. In it’s current state it faces several key issues compared to other GEs, to list a few:

    1. Issues with the GPL
    2. Limited Platform support (No iOS, Android etc)
    3. The logic brick system needs an overhaul to bring it in line with other Node-based workflows
    4. The (temporary) lack of a proper Asset Manager in blender
    5. The lack of alternatives to Python as the scripting language

    These are just some of the issues that immediately come to mind. These issues may not seem serious to those just beginning to explore the BGE and starting out in Game Development. But any serious developer or studio will consider these to be major deal-breakers.

    It is because of this, that dropping the BGE and replacing it with an “Interactive Mode” as Ton proposed makes much more sense. Forking blender to create a Game Engine Only branch, is also an option which will give those developing and using it some freedom to take some much needed drastic decisions and give it a major facelift. That however will take additional dedicated workforce, and it will be a challenge to keep both the main Branch and the GE branch in the same release cycle.

    TL;DR:
    Let’s keep and improve the Sequencer; Lets replace the BGE with a better alternative or fork it.

  130. Don’t remove Game Engine !!!!!!!!!!!!!!!!!!

  131. BGE is essential, it’s the only tool of its kind to make rapid interactive prototypes for ARTISTS!.. Why not translate the Logic bricks into Nodes or open up a framework for artists to design custom-logic bricks (for example: a meta-logic brick where it behaves like a “compositor” for game-mechanics, math, etc..);
    OR
    Have Blender the ability to run/compile multiple languages: python -> c#/boo/javascript (Unity3D), python -> c++ (unreal4)..
    OR
    An add-on system for Game Engine: (just like addons for Blender3D where the plugins/addons are made by python) suppose an addon system for Game Engine (it can be in this case customized: not everybody wants to make AAA ultrarealistic graphics, some want 2D based [sprites], NPR, some want for other purpose [Medical, Science, real-time visualization], or some want some hardware based [think like Processing or VVVV or MAX/MSP* with 3rd party support], how about VR, or AR, Motion Capture, Mobile-device as virtual Camera, Custom hardware profile or controller mappings[omni,cyberlith,leap motion,kinect,hololens,etc..]; maybe it could be used as a visual-tool or viewport for big-data [R,Hadoop,etc..]).. * – even MAX/MSP had made custom nodes based on its own scripting language, and Unreal with unrealscript, suppose we have our own script: blendscript [and blendscript can be customizable with its own addon system: network based, game-based, analytics, HTML5-based, console, etc]
    .. just like python nodes on Blender for graphics.. and we already have Animation Nodes, then why not Logic Nodes or take notes from Unreal4’s approach on Kismet; On the other side of the fence, other companies are closely watching on Blender3D to see it’s next move, nobody wants the hassle and burden with licenses and invasive installations..

    Blender is by far the most accessible (portable, platform agnostic), flexible tenacious tool. It’s up to the users(Artists and Artists) and communities to further customize their blender to suite their needs.

    Quick Cap with extras cents:
    1. meta-logic brick (like python script/module) but nodes into a meta-logic brick for ARTISTS.. and workflow customization and tweaks..
    (Artist can have it their way: kismet style, vvvv;max/msp, gamemaker, scratch, etc..); and why not nodes that we could arrange in 3D space (especially for VR/AR generation; why fixed to screens, let’s get immersed and already have stacks or custom layouts arrange anywhere in 3D space, 3D paint, 3D sculpt, true approaches in 3D space.. or move characters in sequences like chess-style or table-top game and BAM! animation baked and/or logic baked)

    2. Blender’s python -> umbrella of different scripts to become a bridge to 3rd party engines.. [heck, even blender scene with logic brick = exported assets to translated scripts: logicbrick -> python -> script [e.g: c#, lua, javascript, etc..]

    3. Addon system for Game Engine:
    – custom support of graphics, features, peripherals, physics, sound, function/purpose, 3rd party, render..

  132. PLEASE do not drop Collada!

    It is the only reliable bridge that I know of,
    between DAZ Studio and Blender,
    that ports fully rigged and morphed and animated figures.

    Materials can be translated from 3Delight to Cycles via http://mcasual.deviantart.com/art/UPDATE-mcjTeleBlender-3-11-A5TheDrifter-557740819
    and a number of people really use it. E.g. http://parrotdolphin.deviantart.com/art/Blender-Cycles-Test-2-517490444

  133. My thoughts are:
    Removing the Sequencer is a bad idea, the Blender guys should team up with Kdenlive guys to merge the projects inside Blender, since kdenlive is the only decent -or at least almost decent- editor in linux. If that is even possible (no idea :D)
    Removing the Game Engine is probably the right thing to do in order to focus the development in other more vital parts of the software. I’ve never seen anyone actually using the BGE except some posts on Blendernation from people experimenting with it. Having a game engine INSIDE a 3dsoftware doesn’t make sense imho, there are better solutions for free and the development should focus on the 99,9% of the people that use Blender for its main purpose: 3D/Video production, which it does very well (yeeey).
    About the other features to-(may)be-removed, I’m not sooooooo sure about them, they seem quite vital for any 3d software :/.
    (Just my thoughts 8))

  134. adding my two cents worth: the VSE is an excellent no-frills tool and should be retained, at the very least maintained, if for no other reason than to provide an open option instead of Premiere Pro or Resolve or similar – of course Premiere Pro and Resolve offer all the bells and whistles but, as a professional filmmaker, I can say that Blender’s simple VSE is a breath of fresh air, fast, pain-free and unencumbered by bloat and corporate sociopathy – I keep a custom version of Blender set up as a standalone VSE and use it for my day-to-day previz, rough cut and transcoding work – if the Compositor gets a few upgrades I can see the VSE and Compositor working as a powerful no-nonsense option for finaling too – thanks for listening, SN

    • I agree, I think we should figure out what the VSE and compositor should become. For that we need users who really care and some coders to make improve or rewrite parts of the underlying systems. Stick those people together in some place for some amount of time.

      Making blender great for the workflow you seem to describe would even be fun. The VSE is on the list as the code really should be modernized or rewritten.

  135. game engine is important for crowd simulation.

  136. Well, from my perspective…

    1. Abandoning sequencer and (even more) particle system – VERY BAD idea. Without a sequencer Blender will no longer be a “3D animation suit”, ending up as a “normal” 3D modeling/texturing/sculpting package. A particle system is something that is absolutely essential to make an art in Blender. How can You do realistic hair (for example) without it? Asteroid field? Literally anything that has many pieces distributed (semi) randomply, without giant amount of unnecessary work?

    2. Dropping Collada – well, never used it, so for me it’s “whatever”

    3. Killing BGE – VERY GOOD idea. There are numerous better and more convenient way to create games and/or interactive presentation. There is no chance BGE will reach the level of (partially free) commercial competitors, so allocating resources in it has no sense, it’s a waste of time and money

    4. Updating to OpenGL 3.x – good idea. Yes, I know that some years old integrated graphic cards will be left alone with no BLender support, but let’s be honest: who even a bit serious about 3D graphics work on Intel i9xx graphics or worse? I really understand You care about the users, but imho this is absurd.

  137. Hi, I understand getting rid of the sequencer and BGE, as there are other open source projects that are dedicated to delivering these tools.

    But I do have a question about OpenGL. Might it not be better to postpone work on OpenGL, until Vulkan comes out? And then perhaps compare the two to see which is better for Blender in the long run?

    • We’re definitely keeping an eye on Vulkan! But we can’t hold our breath for Apple to support Vulkan in Mac OS, since they have their own Metal API. OpenGL 3.2 works today on Mac, Linux (with Mesa or vendor drivers), and Windows. OpenGL supports graphics cards several years old up to the latest GPUs. So it continues to be the best option for now.

      • Makes sense to me. I look forward to seeing more OpenGL improvements.

        Also, I’m sorry. I submitted a similar post to this a week or two later after this. I wasn’t able to find this post again to see if anyone responded.

  138. The proposed changes sound great for this next-gen Blender! With all the breaking changes in the code and entirely new workflows, surely it would be a good thing to call it v3.0?

    It will be a lot easier for casual users to distinguish the difference between opening older scenes in the 2.5x series compared to the 3.0x series. Compared to the small difference in naming from say 2.78 to 2.81 which would not be compatible, it may lead to a lot of confusion.

    Aside from that, it is all very exciting news and bodes well :)

  139. But they had already announced for the PBR with the replacement BGE? Now they want to remove it. This seems illogical, however there are advantage and disadvantage as to fork for BGE or other game engine support (Godot, gameplay, etc …) but it ask us to do technically, the least down is that some of between from us don’t have no way to buy Unity …

    • Unity has a free edition and Unreal is free. Its better that they make Blender better in its core toolset. It already works great with these game engines.

      I believe that Blender game engine is not up to date and is too far behind its it capabilities. They are doing the right thing by possibly dropping the ball on something that requires too much work to bring it up to par.

      • No amigo lee bien las condiciones para el uso de esos engine (la letra pequeña) de “libre y gratis” nada

  140. Dropping BGE could be the most terrible thing you guys could do, I personally would be very disappointed, because BGE has a very large and solid community and dropping it could literally mean a suicide for you guys.
    If 2.8 is going to be this bad, I think I’ll start learning how to use Autodesk 3DS Max :P

  141. Need the sequencer..Period

  142. what in the world is going on here? whats all the hating on bge about? it seems like most people here are drinking from the hate kool aid and have clearly never even tried to use bge lol

    bge can be used for soo much more than just games….why not just change its name to “blender coding simulator” or something to stop it from being compared to other “game engines”

    i have not used sequencer that much so i cant comment on that

  143. Please dont “kill” the Game-Engine :( It would the terrible thing that ever happens with blender. A much of Potential will be wasted if you remove the Game Engine :( Please dont do it…

  144. the VSE is the best free sequencer editor, I can’t believe no VSE in blender, if no VSE and BGE, the blender will not full complete, This is the worst news I’ve heard today, oh GOD, PLEASE HELP:(

  145. Please keep the VSE

  146. Very interesting set of propositions. Here’s my 2c:

    1) BGE could probably go. Godot has constant progress and is much more feature rich. BGE feels like a bolt on to me to the degree that I’ve never felt compelled to use it.

    2) As far as the sequencer, I don’t think this would impact people much or for too long. There are some decent video editors out there for Linux that can do just as good and better than Blender.

    For both of these, however, if they really are on the chopping block, they are pretty big features and I think it would be better to nuke them in Blender 3.0 where it will be less likely to catch anybody by surprise. Again, my 2c.

    • Believe me, I work with video in Blender in Linux by 10 years: There’s no better video editor than Blender.

  147. im a happy user of blender,and i’ve tried other 3D software’s you name it, maya,XSI,3d’s Max,C4D,Modo,Lightwave, non of them like blender, for what i think is that dropping the BGE sound like a good idea, for many artist or blender users, they never open the BGE and some of them maybe dont even know that blender has one game engine inside of it, blender is a 3D software, okay, game development in our time is a big area of 3D,blender has awesome features, is a complete suit but i think that the BGE is one thing that can be droped out, for better, i dont use the Sequencer to much but it is a great tool, worth keep it, and maybe improved it,cause it has some future if you gave him some love

  148. Extremely interesting notes and I agree with almost everything said, however. I believe dropping featured in blender would be unwise. I understand that it is inevitable but we should try and minimize it. Having blender so just about everything eve if it doesn’t do those things the best is one of the things that makes blender unique. I love that when I want to do a project I don’t have to think what softwares do I need? I think, blender, blender for everything. This mentality has been pushed further since I started using blender for my post work on images I a fashion very similar to Krita of Photoshop. I’, not saying using multiple softwares is bad, but being able to have blender handle everything is extremely freeing.

  149. Don’t know why but my previous comment never showed up.

    Anyways, just my thoughts. Blender + BGE must the the only tool (commercial or not) that:
    A) allows you to build an entire game without ever exiting the program. Modeling, UVs/texturing, Game logic… the whole pack. All the big game engines show off their editors, editors that in this time and age can only do some primitives or import objects to build an object, character, scene etc. Blender does it all!
    B) Allows an ‘artist’ that doesn’t know a thing about code and all to build he’s own game, test his ideas and then move on with refining and even hiring someone to work on another engine to build the final version. You can’t beat BGE in game prototyping.

    The VSE is very important too. You can put together your reels, clips, tests; the BF movies were also edited using the VSE. It’s pretty big imo having a tool that takes you from your 3D objects to the final image that your client or friend will see. Blender, finally, doesn’t have to follow the ‘industry standards’ as a 3D application. I think it’s here to fill people needs, make them creative and happy.

  150. Is there a difference between the Sequencer and the Compositor?

  151. Guys, please don’t take away the Video Sequence Editor. It is a great tool. Without this Blender will not be a complete tool for animation anymore.

  152. I use Blender mainly because of the VSE (maybe 75%) and the other 25% is modeling, texturing, etc.
    I also see it as a “swiss army knife of graphics” like someone above mentioned already.
    Please only consider dropping the VSE if you have a better substitute for it.

  153. The sequencer is great, and is my one of the more powerful video editors I have actually gotten to work on linux. it would be a shame to loose it

  154. Reducing round tripping is a huge plus that Blender has always offered that helps set it clearly apart from other software combinations. Losing the VSE would be a huge shame because of this and also because it compiles and works hassle free on various LTS GNU/Linux distros, unlike current versions of Kdenlive, which favours more bleeding edge distros. Seems like a small thing, but in a studio setting, this really adds up as yet another hurdle. Please, consider all the angles before jumping to a decision on this one. Perhaps, a reduction and refinement of the code would be warranted, but this is also about stress on resources on the Foundation’s end. No easy answers here, but consider the artist and studios in your decision making process, as promised.

    I can’t believe that particles would ever be in threat of being on the chopping block. How is that even possible? A massive overhaul, sure; but, dropping it altogether? It’s ESSENTIAL, pure and simple. There was all this talk of nodifying it completely over time after the depsgraph project, which excited many beyond belief, and now this? I’m having a hard time believing this. PLEASE, CLARIFY.

    Particles can never go, in my opinion. What would happen to all those features added in Cycles and smoke systems and the list goes on and on… These are tremendous advantages in Blender over other software out there. I’m sorry, but I’m a bit flabbergasted, at the moment.

  155. What is the concern with the Blender VSE being dropped? Is the concern that it is presently un-maintained? Please clarify. The VSE is a profoundly useful part of Blender, and has a growing user-base. Blender is still the best video-editor on Linux. Being able to animate anything is very very nice in creating video productions that rival the output of After Effects. If a new maintainer is needed, say so…

  156. In my opinion Blender 3d is the most flexible Fast and powerfull 3d programm in the world! Thank to all Blender Developers!!!
    Why?
    because you can work very fast with the non blocking user interface you have all the tools you need around you in the right place and panels-buttons etc.fast modelling-texturing- rigging- animating rendering – a “swiss army knife” and very stable… a very Intelligent User Interface… THE BEST!! and it looks very proffesional!!!

    i agree that the Sequencer is very very Important inside Blender!(needs Development in some areas
    Blenders particle system is also very powerfull!!why remooving must be in development we need particle system for the game engine integrated also
    if you believe me or not Blender with the Game Engine integated is also the best and right choice !!Blender´s Game Engine is not Unreal and not Unity…etc in quality speed etc …but who knows in the near future
    …. one day i see a youtube Video:in this Video some Big Bosses from various and famous Game development Companies talking about how
    the best game engine in the world must looks like work like and you know what if you hear them, they talking about Blender (not about Quality, speed and tools etc. but about the way you work
    Why!they speaking about Instant acess of all tools without Exporting and Importing all the time stuff in other applications Rigs Models Textures they talking about the possibility to Rig and Animate Paint Sculpt change modifie a model a terrain a level inside the game engine itself (in the same Programm) making some rapid changes and then pressing the button and start the game engine….to see how it looks the exit and made other changes…etc

    so developers dont remove anithing from blender!)dont destroy the power of Blender!
    Please

  157. “BGE” could be a very important feature of “Blender 3D”.
    It just isn’t as powerful as others egines,(UE4,UNITY,CRYENGINE etc…)i know that blender is an open source software but you really should concentrate on the “BGE”, because in my opinion “Blender render” and “Blender cycles” are already the best.
    So give us new features, new rendering system, more power like the modern engines..
    Then you’ll have more users/developers working on it.
    So please, think more about the blender game, thank you.

  158. Hmm, reading the above I feel there is some concensus. But ther is also a need. We do not want to drop functionality which makes life easier or is a part of a pipeline. This is exactly what Ton was proposing, how to be a part of a pipeline. It is that could be the solution to a lot of problems. How to integrate better with other software (open or not). If the integration is a succes its easier to drop something if it can be replaced and has an equivalent way of working.

  159. I’ve been a user of Blender for almost 15 years now, but I think that, blunt things to be said, but is because for Blender community’s sake:

    a) Every animation system should be working like the Dinamic Paint system in terms of integration among systems (e.g. modifiers, particles,hard-soft physics, force fields, working in harmony, meaning like meshes being affected by other objects and being deformed, ripped apart by a tornado made in particles, for example).

    b) Integrated physics in armatures at least a spring constraint-like, a Spline-IK that works as intended. Sorry to be blunt here, but many, many constraints are pretty damn good ideas but many are half baked, like the “limit distance” which have years of reported “for transform” not working.

    c) Blender strives to be useful for artists, we all know that, but people likes to be paid and need to be in professional enviroments to grab a salary, to use Blender as a tool in the chain, dropping Open collada, and other tools for real support in real production chains is not wise.

    d) Let’s be honest Blender can’t do commercial games by it’s own, since you can’t keep painstakingly made assets safe from others eyes; for making assets it’s a real neat monster, I agree on that. as for dropping the game engine, ask users, was tried and failed a while ago.

    e) There are no other real alternatives for the sequencer out there. Just works as intended, no one said it was going to be “Abode Lumiere”

    f) Don’t tell these are hard, and not doable, that no one is interested in, that I should learn C++-% , and learn to walk over razor sharp blades, I’m a animator and rigger and HAD to learn Python because I had to, but it’s lazy to say “do it yourself man”, if you are asking for things that could be made better in Blender, those are, I have a few more. A quite more challenging changes like the 3d cameras has been made.

  160. Video sequencer is useful for previz. We can make previz editing with OpenGL views of different scenes and change camera in one scene with camera override option or multicam selector strip in a second.
    I don’t know if it would be possible to an external video editor to be as flexible for previz.
    Although animatics are made with grease pencil; editing of order, timing, focus of scenes is still needed.
    Simple previz is something that autodesk softwares copied from blender.
    VSE is also a way to create animation according to sound or to edit sound according to animation and for a mix of both.
    VSE and Game Engine are things that are facilitating prototyping, research work. It is faster to check actions blending into animation tool than to re-export them at each minor changes.
    IMO, keeping them as previz tool, animation synchronizing tools is making sense.

  161. Dropping so many needed and belowed functions, like BGE or VSE is dangerous. We already suffer this when Reactor Particles were dropped for no clear reason and they never came back. IF such parts of Blender are about to be dropped, there must be a good, integrated replacement. And who is going to write better VSE in place of the current one?

  162. I have to be frank- the Blender Game Engine can do quite a lot despite its shortcomings. I’d rather have the option to distribute games made with the game engine without having to use Unity or Unreal. The workflow is amazing compared to these other tools. Many say it’s great for prototyping, but why not for complete development?

    Perhaps I’m in denial, but I’ve become rather attached. With node materials recently providing decent realtime PBR, along with a sufficiently powerful interface for quickly programming, the BGE has become an invaluable tool for me. If I had the right kind of experience (and time) to fix it rather than scrap it, I would try.

    Just my two cents. I’ll be sad to see it go, and I may continue to use it even if it’s removed in 2.8 and beyond. I like the ideas of integrating it as a ‘realtime interactive mode’ a la Marmoset Toolbag and bringing the Blender and BGE codebases closer together, but I really want to keep the distributable runtime and powerful scripting capabilities. Of course, with all this talk of physics visualizations, I assume we would still be able to run these projects in some kind of player, if not Blender itself. I don’t imagine it would be difficult, then, to create a new Blender Player based on the integrated realtime mode if someone were sufficiently motivated.

    So really, whatever we need to do on the technical side of things is fine by me so long as the tangible benefits for game developers are still present. I’m aware of the BGE backend and its problems, but it’s ultimately irrelevant to me so long as I can use PBR node materials and program in Python. The next best open source / free software engines barely rival the usability and rendering already present in our ‘broken’ solution. I’m not trying to dictate what must be done- I only want to express my concerns and hopes for the future.

    I’m afraid that people like me don’t take enough time to express how much we value the game engine, and I’m sorry for that. I figure I ought to at least speak up before possibly having to say goodbye to something I’ve found so useful in my personal and professional work.

  163. I’d love a light-er version of Blender but it would be sad to see the BGE and VSE go. BGE is awesome for game prototyping and the node-based logic is one of a kind (I’ve only seen UE4 also doing it right). It allows people that don’t know how to code to build a full game, that’s a big deal imho!

    BTW are there any open source node-based game engines around?
    VSE is also used by many for putting their shorts and renders together, heck even the BF films were edited in the VSE! The one feature I rarely use is the video tracking, but guess it’s important to many for FX and all. So… I don’t know…
    Collada takes 1/3 of the binary size of Blender? Wow!

  164. Here http://opensource.com/life/15/1/current-state-linux-video-editing Blender is winner in compare to any other video editing software on Linux including kdenlive. And I choose blender sequencer! Good work, maybe you make a separated project?

  165. The bge is plenty powerful,
    it needs a better render,
    also its maintained by volunteers mostly,

    Besides drawing/openGL, design issues I can do just about
    anything I wish in the bge.

    To make the bge much much better.

    S C A node graph,

    Properties can be used to set sensor values,
    and sensors can set properties when true.

    so

    Radar(set target)——-and——-steer to target

    Ray(set hitPoint)——–and——-add bullethole
    mouse——————–/

  166. I’d say focus on exchanging resources with game engines in a fluid manner to allow fast iteration instead of having a BGE.
    A game engine is very resource demanding.
    Working on integrating blender made content with other game engines would be better. Integration would focus on export/import of assets, content but could extend to game engine preview.
    To create animations, having physics would be tremendously helpful.
    But a video editor is not necessary. It’s perfectly possible to use another video editor together with blender. I guess a video editor isn’t an absolute necessity for blender. Just an interface to render video and export the video.

    • A game engine is very resource demanding qua developer time.
      Real-time game engine preview with live updating assets would be awesome though. Integration could be done in a bunch of export scripts. One script for each different game engine. And have some standard blender game engine export format and conventions that can be accommodated by game engines.

  167. I’m also confused about the idea of dropping the Sequencer. I feel like someone is talking about something that has passed me by called the Sequencer. We rely on keyframes for our animations and game development, without it, we’d be on Maya. What gives? Please clarify?

    • The sequencer isn’t the same as keyframes. We’ll never drop the ability to animate :)

    • The sequencer is referring to the Video Sequence Editor (VSE), not the timeline or any of the other animation tools.

    • They’re talking about the video sequence editor where strips of video and audio can be cut, edited, matched together with some limited effects, cross fades, color matching etc.

      It’s a good facility and some of us use it a lot, but it does need rework.

      If love to see a compositor effects strip fit instance.

  168. In my opinion, one of the great things about blender is it flexibility as a tool. I like to think of it as the “swiss army knife” of the 3d packages. So maybe the Game engine could be more a “realtime-oriented rendering” feature for game development, like marmoset toolbag software and combining the sequencer with some GPU-node based filters, as this feature is in my opinion extremely useful when you need to blend output images for fast results or effects.

  169. My honest opinion, and do not take it personally :D
    Blender can not be good in all aspects, and it is a huge mistake and waste of money trying to support all these features.

    1) Godot Engine is Open source, much more powerful than the BGE, support for 2D/ 3D/Win/Linux/Mac/Android/iOS….. not to mention Unity/UE4…

    2) Kdenlive is amazing video editor with better performance and various features that save your life.

    3) Natron is an extremely powerful and complete composer, compatible even with Nuke plugins… it would be absurd an integration with Blender, would be something like the “”Maya/Nuke open-source”” xD

    Anyway, Blender integrating with other open-source projects would help to evolve faster and these resources could be spent on solving current problems as Viewport, Sculpting, Retopo, Alembic, OpenVDB, Physics System, Particle System, PBR… etc …

  170. Second NICO on VSE. Despite its drawbacks, it’s still IMO the best video editor on Linux. If the sequencer can be GPU-accelerated and some very basic editing tools implemented for a more hotkey-based workflow, the VSE can be really great!

  171. Why drop the video sequencer ?
    How would we create animation movies without it?
    Dropping game, its better i think to look in the direction of blend4web, and use their logic nodes.
    It would be handy to have a render option as fast as a game render.
    So something not Cycles and not BI, remind that blender is used for making game things and doing simulations. Blender never can compete against parties who create only a game engine. But Blender doesnt have to be a full blown game engine either, just the basics, to be a great tool for physic simulations (thats not only game simulations.. i’d like to see more of bullit engine).
    It might be better to join teams who do make opensource game engines / or to only focus on a few popular export formats later.

  172. No sequencer and particle system would make Blender pretty much useless for me. =(

    • I don’t think the particle system will ever be completely removed, but I think we can expect an overhaul at some point as meantioned.

      As for the sequencer, it would probably be a good idea to get rid of it (along with the GE). I do use it every once in a while but I feel like it probably shouldn’t be there since the main purpose of blender is to be a 3D modeling, animation, and rendering software.

      • i agree with Ton that the sequencer is a great addition to blender.

        you do realize that most animations aren’t made in one long shot. they tend to be shot in scenes. the sequencer allows you to put those scenes together and thanks to being integrated into blender it does allow you to do some interesting things as well.

        it does need some work but even so it’s still my go to video editor and it’s honestly more powerful than most people give it credit for.

        • If the sequencer and the animation system were unified such that it could be used through the entire pipeline, I think that would be interesting. You could plan shots and sequences before rendering, and then just press “render” and then get back your sequence. It could all be set up like a multicam editor, with dropdown menus that select viewports in a traditional a/b setup!

          If you could define how much excess footage is to be rendered before and after the edit, you could even edit after the fact the render is complete. You could very easily re-render certain clips on-the-fly by selecting the edit and hitting “re-render”. In a networked setting, an editor could continue working even while the footage is being rendered using real-time previews.

          Imagine if keyframe/curves and video could be all edited together in one space? It would totally open all sorts of editorial possibilities!

      • Allthough, you dont use the bge, there are many who use it on a regular basis, take the bge fb page for example, it has over 5000 members and is the only one I know of where you can easily edit and add meshes in the view port without import/export. It is probably the easiest and fastest engine to use, to many, (myself included) the abandonment of such a vital feature is terrifying! I really do hope there is a solution to the bugs features and optimization, the rasterizer needs alot of work, somebody was close to cracking it a while back but stopped development due to all the change due in 2.8 see;
        http://blenderartists.org/forum/showthread.php?227199-BGE-Development-Rasterizer-and-Scenegraph-optimization-(proposals-discussion)
        http://blenderartists.org/forum/showthread.php?336095-Moerdn-s-Rasterizer-and-Scenegraph-optimization-s-Updated
        thanks and goodluck to the devs

  173. Very interesting notes. I definetly look forward to the next big leap.
    Of course there are some fears on my side, for features being dropped that I consider very useful, even if badly maintained or out of date.
    I’d love the sequencer to stay or be replaced by a new one, and i don’t care if the BGE was dropped, however i definetly see how someoneelse be the exact opposite.
    I am optimistic I will be sold when the first versions of 2.8 come around.

  174. BTW: I was the one who coded the original VSE and GE in Blender. Both I consider great additions for a 3d suite like Blender. Both have serious maintenance issues and will need a recode or a serious overhaul.

    For the GE I believe the best future is to get it using more and more features from Blender itself, like the new viewport and the animation system.
    For the VSE: we just need a devoted developer (or small team) to get the system refactored or refreshed. Could well be added as a 2.8 target.

    • Thank you very much Ton for this reassuring note.

      Would it be perhaps possible to create some sort of fund raiser for any specific feature, like for the beloved VSE?

      Blender is already an amazing tool as it is right now, I can’t wait to see what 2.8 will bring to this software.

      Thank you Ton for this priceless peace of software, and to all the developers. Like others have said, maybe updates only after six months, or perhaps yearly update would be accepted and more than enough. It would give more time to developers to focus on the features.

      Regards, Aclariel.

      • Yep I consider a massive fund raiser, but I want to wait a bit with decisions until we have enough feedback.

        Needless to say – the Development Fund could use a lot more subscribers too! If 100 people would spend as much on Dev Fund as they do on 1 Maya license, we could hire 5 developers extra and get 2.8 kickstarted right away.

        • Hi Ton,
          Could you contact me or Jeff Bell at Tangent Animation to discuss how we could help you out please. We have some resources we could allocate.

          Brian.foster at tangent-animation com
          Jeff.Bell at tangent-animation com

    • You’re a genius, Ton; PhD well deserved
      (and that was before the great GUI overhaul of Blender 2.50!).

      Now, as for Richard Colburn’s offer to share his improved cloth simulator:
      have you had a chance to see it?

    • First off thank you for all your hard work and dedication for making a remarkable software.
      We are currently using Blender on a feature film as the principle software package. We chose it because of it’s unique abilities in several areas. One of these areas is the vse for previs and layout. In this area it is unique among software packages for the ability to mix media, including live scenes for evaluation.It would be a major shame to give up on this feature as it is incredibly useful for maintaining shot continuity. Please leave this on the list and develop it further, it helps set Blender apart from others.

    • Yes! Thank you. This is exactly what needs to be done. I understand that this is a very large project, but removing the GE is a huge leap in the wrong direction.

    • I’m agree with you

    • The main problem with the sequencer is that it duplicates functionality from the compositor, which is much more powerful, but harder to see what is happening.
      Perhaps the compositor could just get another interface, in the style of the sequencer?
      Some features might be, for example, to have some way to graphically position things (like the centre of sun beams), to see and move compositor keyframes, and other things that would make it easier to use.

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