Blender is really an amazing case study of open source software. Apart from the Linux kernel and web browsers/tools, it is perhaps the only open source software that managed to beat all the commercial software in its niche. It has rendered Maya nearly obsolete.
Meanwhile, in other niches, Microsoft Office still beats open source office suites like LibreOffice; Photoshop isn't about to give up its crown to GIMP; Lightroom isn't losing to Darktable; and FreeCAD isn't even in the rear view mirror of Solidworks.
I wonder what will be the next category of open source to pull ahead? Godot is rapidly gaining users/mindshare while Unity seems to be collapsing, but Unreal is still the king of game engines for now. Krita is a viable alternative for digital painting.
> I wonder what will be the next category of open source to pull ahead?
KiCad, for PCB design. They have been making massive improvements over the last few years, and with proprietary solutions shutting down (Eagle) or being unaffordable (Altium) Kicad is now by far the best option for both hobbyists and small companies.
With the release of KiCad 5 in 2018 it went from being "a pain to use to, but technically sufficient" to being a genuine option for less-demanding professionals. Since then they've been absolutely killing it, with major releases happening once a year and bringing enough quality-of-life improvements that it is actually hard to keep track of all of them.
From the type of new features it is very obvious that a lot of professional users are now showing interest in the application, and as we've seen with Blender a trickle of professional adoption can quickly turn into a flood which takes over the entire market.
KiCad still has a long way to go when it comes to complex high-speed boards (nobody in their right mind would use it to design an EPYC motherboard, for example), but it is absolutely going to steamroll the competition when it comes to the cookie-cutter 2/4/6 layer PCBs in all the everyday consumer products.
> Apart from the Linux kernel and web browsers/tools, it is perhaps the only open source software that managed to beat all the commercial software in its niche.
I can think of a few more: Git obsoleted an entire category of commercial software seemingly almost overnight, VSCode has become by far the most widely used IDE (not entirely open source, though), TeX still dominates mathematical typesetting AFAIK (as it has for as long as computers have been used for that), (lib)ffmpeg is used everywhere for video/audio transcoding and between them nginx and apache still account for the majority of webservers. Most popular programming language compilers/interpreters/runtimes are open source too, of course.
Let's not get ahead of ourselves, Blender has done far better than most open source software but Maya is still very much the industry standard. I don't think we can realistically say that Maya is beaten until Blender is battle-proven to the same degree, on the most demanding real-world production workloads (think Pixar/Weta), which for now it hasn't been.
What Blender achieved is that lots of university programmes have started teaching Blender or becoming 'tool agnostic'. Studios have also started diversifying their pipelines (this coincidences with studios adopting Unreal and increasing usage of Houdini).
So while Maya is currently the standard, I don't believe that it's growing.
It'll probably be around still in 20 years, with lots of studios having built their pipelines and tooling around it, with lots of people being trained in it, and because it's at the moment still better than Blender in some aspects like rigging and animation (afaik).
Blender is the go-to for struggling artists/developers, and industry outsiders, like me. I'm stuck at Blender 2.93.18 because I don't have the budget for better hardware, let alone a Maya license! However, even that version of Blender still gets it done for me.
And also, how can you say Blender is not battle-proven? I mean, the big studios use Maya like fortune 500 companies use Microsoft Windows - doesn't mean Linux isn't battle proven.
That is not a very big studio or very big production, Blender falls over in the pipeline department. It’s a constantly changing API that doesn’t allow for the extensibility needed to get a major project out the door, just the fact that only a Python API is provided is enough for most people who have worked on massive scenes with massive amounts of data to consider it a non starter.
I'm sure "major project" is a subjective label, but Flow made headlines earlier this year with an Academy Award (Best Animated Feature) and Golden Globe (Best Animated Feature Film)
Flow is good filmaking expressed through low-tech production, which is totally valid, but it doing a lot with a little isn't going to stop Disney from throwing 200 terabytes of assets at the next Zootopia movie so Blender needs to handle that angle too if it's going to become a catch-all solution for all kinds of production.
Not disagreeing that usage in large productions is something that Blender isn't really designed for, but I don't think that it's for a lack of Python API features (if a studio wants something specific it could just maintain an internal fork) or the ever changing Python API surface (the versions aren't upgraded during a production anyways)
VFX studios have been using Python APIs for twenty+ years, backed by C. They were one of the first industries to use it. That's where I learned it, around the turn of the century.
As a game developer, I'm really rooting for open source game engines.
Unity and Unreal are dinosaurs that target the shrinking console market. Godot is being built in their image. My hope is that something more versatile like Bevy becomes common so that we have something that could potentially compete with the next generation of Roblox.
Video editing? Adobe has set themselves up for failure there, everyone wants an alternative
Davinci Resolve is probably competitive with Premiere, but while free it's not actually open source. But either a viable competitor catching up or Davinci publishing the code could change that fast
> Apart from the Linux kernel and web browsers/tools, it is perhaps the only open source software that managed to beat all the commercial software in its niche
I use OBS regularly but vMix is definitely the superior option in the professional livestreaming world. OBS is missing many key features for professional operations and vdo ninja only covers some of those gaps.
> Meanwhile, in other niches, Microsoft Office still beats open source office suites like LibreOffice; Photoshop isn't about to give up its crown to GIMP; Lightroom isn't losing to Darktable; and FreeCAD isn't even in the rear view mirror of Solidworks.
Tbf, everything starts somewhere and all the proprietary apps you listed were not instant market leaders.
I can and do use all those FOSS tools just fine both as a hobbyist and professionally, my needs are meet. Others may not find the same, but I suspect there's just a lot of stickyness preventing even trying new workflows.
> I can and do use all those FOSS tools just fine both as a hobbyist and professionally, my needs are meet.
Mine aren't: GIMP is okay, FreeCAD is a complete joke. It is painfully obvious that their development is done primarily by F/LOSS enthusiasts rather than by industry professionals and UX designers. They are closer to being a random collection of features than a professional workhorse. You might eventually get the job done, but compared to the proprietary competition it is woefully incomplete, overly complicated, and significantly buggier.
The poor quality of FreeCAD is the main reason my 3D printer is collecting dust. As a Linux-only user the proprietary alternatives mostly aren't available to me, and FreeCAD is bad enough that I'd rather not do CAD at all. The Ondsel fork was looking promising for a while, but sadly that died off.
LibreOffice, GIMP, FreeCAD and Inkscape all have their quirks (and bugs), but they're probably seen as features by their core users so they won't change.
I’m rooting for gimp but they are truly a decade behind photoshop. I don’t know how they can compete. Luckily it does not seem like they are in the market for professionals so the need to match Photoshop isn’t quite as high
really? I haven't done 3D rendering in a long time, admittedly, but back then Maya and Lightwave were miles ahead of Blender. Rhino3D too. Even 3DSMax was better. Lightwave seems to have sadly fallen off (unfortunately, IMO it was the best at one point, had excellent ray tracing). I didn't really Blender had come such a long way -- that's great.
Since you mention niches: Adobe InDesign has no OSS competition at all, and Illustrator is still much better than Inkscape.
I’d really like to see something like blender come for the 3D CAD industry, at the moment it feels like the only people who would lose out are AutoDesk. The amount of money that flows in and out of 3D cad (as subscription and then value created) having a first class open source kernel and tooling, would be giving big industrial players freedom to modify and tailor to their needs as well as smaller / hobbyists get started for free!
Probably requires something that is almost there then a sponsor(s) to throw in developers or funding to get the rest of the way. On the EDA side CERN did a lot to lift Kicad to the point of being a credible alternative that could breakthrough like Blender. Both those projects are over 30 years old and for a lot of that time were dismissed as too difficult to use or lacking in features. FreeCAD is only 23 years old. I don't know what the code base is like but if a large org put a couple of good devs into it for a few years who knows.
It must be difficult when so much management is short sighted and focused on delivering short term profits for shareholders. Even academia is run like a business now.
Unless a privately held rogue company like Valve got interested its probably going to have to wait for a government/ngo/scientific. Industry, particularly the tech industry, is notorious for leaching of free and open source software and in some cases building entire businesses on it and not giving back.
The problem with FreeCAD and every other free/open source MCAD project of note is the Open Cascade kernel they are built on. While Open Cascade is fairly mature, it has dealbreaker issues in a few key areas: fillets cannot consume connected faces and may fail for a number of other reasons, cylindrical and spherical faces require seams which often cause issues with boolean operations, and shapes like helixes are also often troublesome.
Originally open-source, but since taken back in-house. As I understand, which should not be construed as an accurate accounting, Sandia wants to flesh out the basics further before (potentially) open-sourcing it again.
I really didn't like how they both require a phone number and make all of your files public. I've stuck with Fusion which seems a lot more privacy respecting while also being basically free for home users.
take a look at https://Plasticity.xyz. It's not open-source, but it's got a small, highly dedicated team behind it. It's built on Solidworks' kernel, so it's quite robust.
Also take a look at solverspace, caligula, FreeCAD, ...
IDK how they compare to professional CAD tools, but I've heard good things about FreeCAD and OpenSCAD. I know that some people use Blender for CAD work, and there are even some extensions to make it easier, but I'm dubious that the representation of meshes that Blender uses are well-suited for CAD applications.
I just tried FreeCAD last week. I uninstalled it after about 10 minutes. The most basic actions to just get started were throwing errors. Maybe it was user error, but it was a very bad first impression.
Approached it with the same attitude at the same time, after 10 minutes decided to view some basic tutorial (for an earlier release) that made things clear and I could continue basic tinkering on my own.
But of course built-in intro of Solidworks was a way better UX.
I actually did the same thing so that I could figure out how to lay out my workshop!
What I'd do is:
- Spreadsheet workbench --> Create spreadsheet (name it "measurements"). (This is optional)
- Switch to Part design workbench --> Create body (name it "layout") --> select XY plane --> Create sketch --> Create Polyline
- Zoom out, start drawing the rooms in your house, approximately to scale.
- Before going into too much detail, add a dimension (select line --> "Constrain Distance") to the first line you draw, so that you can do the rest of your drawing approximately to scale. Then the general shape won't get messed up when you add dimensions to everything else.
- (If you have a photo or picture, you can import that to sketch over).
- Add constraints to match your room measurements, mostly vertical or horizontal distance constraints. Be careful not to overconstrain the sketch. (You can put the measurements directly into the sketch constraints, or you can put them into the top-level spreadsheet, create an alias for each cells, and then set the dimensions to reference those cells).
- Once the rooms are drawn, close the sketch and create a new sketch on the xy plane called "furniture".
- Draw some rectangles for your sofas / tables / etc, delete any horizontal and vertical constraints that get automatically added (they look like little | and _ icons), and instead apply perpendicularity constraints. Dimension your rectangles using only the "constrain distance" tool. Now you can drag them around the room and rotate them freely.
- If you want to make 3D models for these too, create new Part Design bodies for each room and each piece of furniture, create a shape binder referencing the master sketches in the Layout body, and then extrude the sketches using the "Pad" operation.
That's about as much tutorial as it makes sense to pack into a HN comment. If you give it a try, I hope it works out for you!
I like using paper and cardboard for this, dollhouse style, much easier to move things around and visualize that way and more fun than clicking a mouse to boot :^)
Inkscape is good for typing dimensions into rectangles tho
This was my plan B. I do have a scale I can use for it.
I’ll check out Inkscape as well. I’ve tried using some raster graphics in the past, but I couldn’t type dimensions and had to use the rules and guides with snapping. It mostly worked, but was a bit annoying.
During COVID I learnt Blender for 3D modelling. It is still my go to.
Many people complain about it being a mesh editor but it works for me. The sheer variety of tooling and flexibility in Blender is insane, and that's before you get to the world of add-ons.
I want to learn Geometry nodes and object generation as I think they will address a lot of the "parametric" crowd concerns. This v5 is meant to be a big step in ease of use of this.
Also, I'm not sure if the different tooling lets me see all the flaws of online "parametric" models, or whether I'm being pedantic. They get frustrating. I have Gordon-Ramsay-screamed "How can you fuck up a circle!".
In MCAD, “parametric” does not mean a high level part or feature is driven by editable parameters or procedurally generated features. Parametric refers to the underlying storage format representing part features in a parametric way rather than as a mesh. Mesh formats like stl cannot represent a circle by its position and radius, while a parametric format like step can. This distinction is more akin to raster (bmp) vs vector (svg) graphics. Both can be generated procedurally by “parameters”, but only with svg can sub-features be faithfully extracted or transformed.
>Mesh formats like stl cannot represent a circle by its position and radius, while a parametric format like step can.
This is where I think the Geometry nodes can help. A node (function) can be used to represent the circle with inputs and outputs set or changed as required.[0]
I have not fully explored this space though and so my "hopes and dreams" may well be as useful as thoughts and prayers...
I have some understanding of "parametric" vs "mesh". I looked it up when I saw so many people going on about it.
Maybe it is the export or something. I run the 3D toolbox and often models are not manifold.
I see things like two circles in slightly different positions but both are connected in different ways to the surrounding "single" instance model. Things like this mean you end up with "infinitely small volumes". There is no fully enclosed "volume" and so mathematically there is "nothing to 3D print".
As a model this makes no sense to do, and so it irks me.
But clearly the slicer software doesn't care or autocorrects and people make their 3D print happen just fine.
Depends if your goal is artistic or functional. Blender is good if you are trying to make character models, etc. It's not great when you are trying to make a part that has to fit something in the real world and after printing you discover one step half way through needs to be 1mm shorter.
It does take a different set of skills to regular CAD, but I haven't found it that bad for simple 3D printed models that need to be dimensionally accurate.
I have used it to make quite a few functional prints, with the help of making sure my scene units are correct and a CAD plugin.
I haven't tried any of the cad plugins for blender, but I'm not sure how you would retroactively change dimensions in blender. It's usually simple enough to create features to a certain size, but if you need to change them later it becomes significantly difficult.
If I put some holes in something that are 1mm from the edge, but then I print it and see it doesn't line up and needs to be 1.5mm, in Fusion I can just change one number and it all updates. Doing the same thing in blender would likely be very difficult.
Does Blender have NURBS? I don't even use NURBS in Plasticity, because curves are already essentially vectors. I don't have to worry about polygons at all, and then I choose the tolerances when I export.
Same, but I don't think it's possible without a large and sustained investment into a free geometric modelling kernel, which can probably be only done by a government.
Parasolid is powering practically every major CAD system. Its development started in 1986 and it's still actively developed. The amount of effort that goes into those things is immense (39 years of commercial development!) and I don't believe it can be done pro-bono in someone's spare time. What's worse, with this kind of software there is no "graceful degradation": while something like a MIP solver can be useful even if it's quite a bit slower than Gurobi, a kernel that can't model complex lofts and fillets is not particularly useful.
3D CAD is much harder than Blender and less amenable to open source development.
> Same, but I don't think it's possible without a large and sustained investment into a free geometric modelling kernel, which can probably be only done by a government.
Can you help me understand why this problem is so hard?
The other reply is really good. To add to that, the intersection of two bi-cubic Bezier patches (each being but one part of a cubic B-spline surface) is an implicit equation of degree 18^2=324. This simply cannot be implemented exactly in a geometry kernel and thus must be approximated. How do you want to approximate this? If you choose trimming (the industry standard) now you have to handle gaps in your geometry. If you choose remapping into an unstructured watertight spline, you need to solve a constraint system that is NP-hard. If you choose reparameterizing… well, see nVariate’s watertight Boolean technique (disclaimer: I once sponsored a project with nVariate).
Now, generally speaking, in a CAD model most surfaces will be “analytic” (plane, torus, conical, arc, line, etc). But whenever some complex surface that joins these surfaces is required, (NUR)B-splines are the principal technique for “covering” the gap.
The way modern CAD systems work is by having a tree of features/actions that is then used to construct an analytical representation of a 3D object. The features/actions can rely on "sketches" (2D drawings that are coupled with a real time geometric constraint solver) and can be "projected" into sketches, creating new reference lines, that can then be used by the sketch constrain solver, generating a sketch that can be used for more 3D features.
This is already complex and fiddly enough. Just having a stable 2D drawing environment that uses a constraint solver but also behaves predictably and doesn't run into numerical instability issues is already an achievement. You don't want a spline blowing up while the user is applying constraints one by one! And yet it's trivial compared to the rest of the problem.
Having 3D features analytically (not numerically!) interacting with each other means someone needs to write code that handles the interactions. When I click on a corner and apply a G2 fillet to it, it means that there's now a new 3D surface where every section is a spline with at least 4 control points. When I then intersect that corner with a sphere, the geometric kernel must be able to analytically represent the resulting surface (intersecting that spline-profiled surface with a sphere). If I project that surface into a sketch, the kernel needs to represent its outline from an arbitrary angle — again, analytically. Naturally, there is an explosion of special cases: that sphere might either intersect the fillet, just touch it (with a single contact point), or not touch it at all, maybe after I made some edits to the earlier features.
Blender at its core is comparatively trivial. Polygons are just clumps of points, they can be operated on numerically. CAD is hell.
> Same, but I don't think it's possible without a large and sustained investment into a free geometric modelling kernel, which can probably be only done by a government.
It's going to be years or decades before it's competitive though. Also, it looks like they switched to keeping progress updates private except to sponsors, which means I don't actually have any easily-accessible information about it anymore which is sad.
I'm very skeptical that one person can make a dent. Paging through the releases, they seem to focus on constructive solid geometry and code-driven shape generation, which I believe is a dead end.
The tricky bit is having a G2 (or even G3) fillet that intersects a complex shape built from surface patches and thickened, with both projected into a new sketch, and keeping the workflow sane if I go and adjust the original fillet. I hope one day we'll see a free (as in speech) kernel that can enable that, until then it's just Parasolid, sadly.
KiCAD was also a meh ECAD FOSS alternative 7-8 years ago, now it is by far the tool of choice for regular ECAD designs. I can see FreeCad getting there by 2030.
FreeCAD is probably the single most frustrating and unintuitive pieces of software I have ever used. I almost drafted hate mail to the devs after 15 minutes of crash coursing fusion360 got me further than 2 days of trying to use FreeCAD.
It seems like it has lots of capability but still "punch your monitor" levels of difficulty just trying to do the most basic stuff.
While it's a pain to learn and requires some plugins (addons) for basic ergonomics, FreeCAD absolutely works for parametric CAD modeling. YMMV depending on the project and complexity, it does the trick for laser cutting, bending and 3D printing.
Deltahedra is a great YouTube channel for getting the basics.
It suffers from too many "workbenches," some of which appear to be redundant or dated. You never know whether structures created by one are "compatible" with the M.O. of another (like "Part" vs. "Part Design").
And it presents nonsensical problems, like offering to create a sketch on the face of an object and then complaining that the sketch doesn't belong to any object. So you have to manually drag it under the object in the treeview. So gallingly DUMB.
Despite all that, I will wrestle with its ineptitude before giving Autodesk a penny. I get stuff done with it and respect those who give their time to develop it.
I tried it this year. Not in too much depth, but I tried Fusion and FreeCAD for the first time this year for 3D printing and found I was getting much further much faster on Fusion.
I'm sure I could grind harder and learn more and make FreeCAD work, but I'm not sure why I'd bother.
No arguement that fusion isn’t better and easier to learn. Their licensing and changes to their hobbyist offering were no longer tenable for me which prompted my change. I was pleasantly surprised at how well I was able to work in freecad after a few youtube videos.
Something that I've daydreamed of in the past is making a new constructive solid geometry "kernel" around which a CAD application could be wrapped based on projective geometric algebra (PGA).
The algorithms it enables are fundamentally more capable and robust than traditional kernels based on linear algebra (vectors and matrices). You can do really fancy things like interpolating in space and time robustly, find extrema in high-dimensional phase spaces, etc...
This could potentially allow straightforward and robust solvers for kinematics, optimal shape finding, etc...
Every few decades there's a "step change" where some new algorithm or programming paradigm sweeps away the old approach because suddenly a hobbyist can do the same thing solo that took dozens of developers a decade in the past. I suspect (but cannot prove) that PGA is one of those things.
You're on point that there's a tremendous amount of money captures by Autodesk for CAD software that could be better directed at the open source community instead.
Software like OpenSCAD and FreeCAD are obviously not suitable for much commercial work, and have very irritating limitations for hobbyist work, in my mind a big part of that is the UI and Blender has a good and established UI at this point so I'd love to see the open source CAD that provides an alternative to vendor lock in come from a Blender add-on instead of a separate program.
I am no expert but as I understand it the primary difficulty with developing good alternatives to commercial CAD software lie in the development of an effective geometric kernel.
It seems to me that if a developer of an opensource CAD program develops it as a Blender add-on they can effectively outsource the remainder of the development efforts to the Blender community while focus can be made on the CAD kernel itself.
I think at the very least they've found those rare individuals who can both code well and manage well. There are countless open source project leaders who might be the most knowledgeable person in the world on that area but have no idea how to communicate and collaborate in an open-source setting, especially as the project gets popular and attracts people who're detrimental (not necessarily malicious) to the project's growth.
Yeah this is a really good question. I'd love to know more about that too. There's some good history on the website here [1].
The creator apparently was selling it as freemium software in 1998, and then the bubble burst and the corp shutdown in 2002. But the creator created a non-profit called the Blender Foundation, launched a Free Blender campaign [2] (the forum post is still up!) to raise money from its users and bought out the rights to the software from the investors.
Every time Blender has a new update, I scroll and I'm amazed by how much is in it. Then I realize my scrollbar is only halfway through.
Radial tiling my beloved, and a seemingly far more straightforward array modifier <3 Faster volume scattering for non-homogenous volumes.
For those wondering "where the AI is", the new Convolve Node might be it :) Convolutions are a pretty generic signal processing operation (Hadamard product) which are also used in neural networks which work with images. Realistically though, this will be mostly useful for wonky hand-crafted blurs.
The new sequencer looks fantastic, too. I always went to DaVinci Resolve but I might be able to go full blender. Compositor modifiers in the sequencer is also very welcome.
I've been using blender since at least 2010; it's so exciting to see how much progress it's making.
I'm very excited to see the addition of structs and closures/higher-order functions to blender nodes! (I've also glanced at the shader compiler they're using to lower it to GLSL; neat stuff!) Not only is this practically going to be helpful, the PL researcher in me is tickled by seeing these features get added to a graphical programming language.
If you haven't heard of Blender before, or if you think AI will replace all the work done in it, fair enough. But I'd still strongly suggest looking into what it is and how it works.
So...... was there ever a resolution with that persistent incompatibility in ROCm/certain AMD drivers and Cycles that made it impossible to render in almost every version of blender? As in, it literally doesn't even detect the GPU outside of eevee
The SDF geometry nodes open up a wild amount of new options. That and node closures solve so many difficult problems, especially in complex scattering. If you haven't checked them out, here is a bit on them and volume grids: https://code.blender.org/2025/10/volume-grids-in-geometry-no...
It might be possible to reproduce the same effect in EEVEE using geometry nodes. I know people have done that for automatic level of detail work. That being said, IDK if subsurf as a geometry node will take a non-constant number of iterations.
Is this the first blender release where you can change the working color space? I thought that you could in previous versions but it caused issues with some nodes.
Now I want to look into it more, but I'd imagine that "Blackbody" and sky generation nodes might still assume a linear sRGB working space.
> Now I want to look into it more, but I'd imagine that "Blackbody" and sky generation nodes might still assume a linear sRGB working space.
Since people are always asking for “real world examples”, I have to point out this is a great place to use an agent like Claude Code or Codex. Clone the source, have your coding assistant run its /init routine to survey the codebase and get a lay of the land, then turn “thinking” to max and ask it “Do the Blackbody attribute for volumes and the sky generation nodes still expect to be working in linear sRGB? Or do they take advantage of the new ACES 2.0 support? Analyze the codebase, give examples and cite lines of code to support your conclusions.”
The best part: I’m probably wrong to assert that linear sRGB and ACES 2.0 are some sort of binary, but that’s exactly the kind of knowledge a good coding agent will have, and it will likely fold an explanation of the proper mental model into its response.
Display P3 (distinct from cinema display P3, because names are hard ig) is used as a render target color space. ACES (and its internal color spaces) are designed as working spaces.
If you make a color space for a display, the intent is that you can (eventually) get a display which can display all those colors. However, given the shape of the human color gamut, you can't choose three color primaries which form a triangle which precisely contain the human color gamut. With a display color space, you want to pick primaries which live inside the gamut; else you'd be wasting your display on colors that people can't see. For a working space, you want to pick primaries which contain the entire human color gamut, including some colors people can't see (since it can be helpful when rendering to avoid clipping).
Beyond that, ACES isn't just one color space; it's several. ACEScg, for example, uses a linear transfer function, and is useful for rendering applications. A colorist would likely transform ACEScg colors into ACEScc (or something of that ilk) so that the response curves of their coloring tools are closer to what they're used it (i.e. they have a logarithmic response similar to old-fashioned analogue telecine machines).
The key point is that your ray tracing color space and your display color space don't need to be the same thing. Even if your monitor only displays SRGB colors, it still can be useful to have more pure primaries in your rendering system.
This seems like an unnecessarily unkind response. The post you're replying to is sharing their opinion and asking what people who are in the industry think about it.
I design projections for independent theatre in Baltimore. I use AI in my workflows where it can help me and won’t compromise on the quality of what I’m making. I frequently use AI to upscale crappy footage, to interpolate frames in existing video (for artistic purposes, never with documentary archival stuff) and very occasionally to create wholesale clips in situations where video models can do what I need.
I recently used WAN to generate a looping clip of clouds moving quickly, something that’s difficult to do in CGI and impossible to capture live action. It worked out because I didn’t have specific demands other than what I just said, and I wasn’t asking for anything too obscure.
At this point, I expect the quality of local video models (the only kind I’m willing to work with professionally) to go up, but prompt adherence seems like a tough nut to crack, which makes me think it may be a while before we have prosumer models that can replace what I do in Blender.
If AI is at the point where it is exactly as capable of your average junior 3D professional in 10 years, it will probably have automated a ton (double digit percentage?) of current jobs such that nothing is safe. There's a lot of complexity, it's fairly long time horizon, it's very visually detailed, it's creative and subjective, and there's not a lot of easily accessible high quality training data.
It's like 2D art with more complexity and less training data. Non-AI 2D art and animation tools haven't been made irrelevant yet, and don't look like they will be soon.
As someone who's actually used Blender for small video projects, I'm fairly confident you'll still need this type of tool even with AI assistance doing some of the work in it, especially for at least the next 10 years.
AI coding agents didn't make IDEs obsolete. They just added plugins to some existing IDEs and spawned a few new ones.
I don't think the "what will it train on" argument is bullet-proof, but animation and 3D art can encompass so much more than just things that exist in the real world.
Blender is really an amazing case study of open source software. Apart from the Linux kernel and web browsers/tools, it is perhaps the only open source software that managed to beat all the commercial software in its niche. It has rendered Maya nearly obsolete.
Meanwhile, in other niches, Microsoft Office still beats open source office suites like LibreOffice; Photoshop isn't about to give up its crown to GIMP; Lightroom isn't losing to Darktable; and FreeCAD isn't even in the rear view mirror of Solidworks.
I wonder what will be the next category of open source to pull ahead? Godot is rapidly gaining users/mindshare while Unity seems to be collapsing, but Unreal is still the king of game engines for now. Krita is a viable alternative for digital painting.
> I wonder what will be the next category of open source to pull ahead?
KiCad, for PCB design. They have been making massive improvements over the last few years, and with proprietary solutions shutting down (Eagle) or being unaffordable (Altium) Kicad is now by far the best option for both hobbyists and small companies.
With the release of KiCad 5 in 2018 it went from being "a pain to use to, but technically sufficient" to being a genuine option for less-demanding professionals. Since then they've been absolutely killing it, with major releases happening once a year and bringing enough quality-of-life improvements that it is actually hard to keep track of all of them.
From the type of new features it is very obvious that a lot of professional users are now showing interest in the application, and as we've seen with Blender a trickle of professional adoption can quickly turn into a flood which takes over the entire market.
KiCad still has a long way to go when it comes to complex high-speed boards (nobody in their right mind would use it to design an EPYC motherboard, for example), but it is absolutely going to steamroll the competition when it comes to the cookie-cutter 2/4/6 layer PCBs in all the everyday consumer products.
> Apart from the Linux kernel and web browsers/tools, it is perhaps the only open source software that managed to beat all the commercial software in its niche.
I can think of a few more: Git obsoleted an entire category of commercial software seemingly almost overnight, VSCode has become by far the most widely used IDE (not entirely open source, though), TeX still dominates mathematical typesetting AFAIK (as it has for as long as computers have been used for that), (lib)ffmpeg is used everywhere for video/audio transcoding and between them nginx and apache still account for the majority of webservers. Most popular programming language compilers/interpreters/runtimes are open source too, of course.
Let's not get ahead of ourselves, Blender has done far better than most open source software but Maya is still very much the industry standard. I don't think we can realistically say that Maya is beaten until Blender is battle-proven to the same degree, on the most demanding real-world production workloads (think Pixar/Weta), which for now it hasn't been.
What Blender achieved is that lots of university programmes have started teaching Blender or becoming 'tool agnostic'. Studios have also started diversifying their pipelines (this coincidences with studios adopting Unreal and increasing usage of Houdini).
So while Maya is currently the standard, I don't believe that it's growing. It'll probably be around still in 20 years, with lots of studios having built their pipelines and tooling around it, with lots of people being trained in it, and because it's at the moment still better than Blender in some aspects like rigging and animation (afaik).
Blender is the go-to for struggling artists/developers, and industry outsiders, like me. I'm stuck at Blender 2.93.18 because I don't have the budget for better hardware, let alone a Maya license! However, even that version of Blender still gets it done for me.
And also, how can you say Blender is not battle-proven? I mean, the big studios use Maya like fortune 500 companies use Microsoft Windows - doesn't mean Linux isn't battle proven.
The studio that makes Evangelion moved from 3DS Max to Blender as their primary 3D software according to this article:
https://www.blender.org/user-stories/japanese-anime-studio-k...
That is not a very big studio or very big production, Blender falls over in the pipeline department. It’s a constantly changing API that doesn’t allow for the extensibility needed to get a major project out the door, just the fact that only a Python API is provided is enough for most people who have worked on massive scenes with massive amounts of data to consider it a non starter.
I'm sure "major project" is a subjective label, but Flow made headlines earlier this year with an Academy Award (Best Animated Feature) and Golden Globe (Best Animated Feature Film)
https://flow.movie/
https://www.youtube.com/watch?v=ZgZccxuj2RY
https://www.blender.org/user-stories/making-flow-an-intervie...
Flow is good filmaking expressed through low-tech production, which is totally valid, but it doing a lot with a little isn't going to stop Disney from throwing 200 terabytes of assets at the next Zootopia movie so Blender needs to handle that angle too if it's going to become a catch-all solution for all kinds of production.
Not disagreeing that usage in large productions is something that Blender isn't really designed for, but I don't think that it's for a lack of Python API features (if a studio wants something specific it could just maintain an internal fork) or the ever changing Python API surface (the versions aren't upgraded during a production anyways)
VFX studios have been using Python APIs for twenty+ years, backed by C. They were one of the first industries to use it. That's where I learned it, around the turn of the century.
The fact you can point out specific examples of when Blender is used says a lot. It tells me it is the exception.
As a game developer, I'm really rooting for open source game engines.
Unity and Unreal are dinosaurs that target the shrinking console market. Godot is being built in their image. My hope is that something more versatile like Bevy becomes common so that we have something that could potentially compete with the next generation of Roblox.
Video editing? Adobe has set themselves up for failure there, everyone wants an alternative
Davinci Resolve is probably competitive with Premiere, but while free it's not actually open source. But either a viable competitor catching up or Davinci publishing the code could change that fast
Kdenlive is awesome. I am only a hobbyist but I cancelled my subscription to Premiere when I found it.
Resolve Studio vastly outshines Premiere IMO. For $300 flat it’s a no-brainer.
> Apart from the Linux kernel and web browsers/tools, it is perhaps the only open source software that managed to beat all the commercial software in its niche
OBS is on line 2 ....
I use OBS regularly but vMix is definitely the superior option in the professional livestreaming world. OBS is missing many key features for professional operations and vdo ninja only covers some of those gaps.
> Meanwhile, in other niches, Microsoft Office still beats open source office suites like LibreOffice; Photoshop isn't about to give up its crown to GIMP; Lightroom isn't losing to Darktable; and FreeCAD isn't even in the rear view mirror of Solidworks.
Tbf, everything starts somewhere and all the proprietary apps you listed were not instant market leaders.
I can and do use all those FOSS tools just fine both as a hobbyist and professionally, my needs are meet. Others may not find the same, but I suspect there's just a lot of stickyness preventing even trying new workflows.
> I can and do use all those FOSS tools just fine both as a hobbyist and professionally, my needs are meet.
Mine aren't: GIMP is okay, FreeCAD is a complete joke. It is painfully obvious that their development is done primarily by F/LOSS enthusiasts rather than by industry professionals and UX designers. They are closer to being a random collection of features than a professional workhorse. You might eventually get the job done, but compared to the proprietary competition it is woefully incomplete, overly complicated, and significantly buggier.
The poor quality of FreeCAD is the main reason my 3D printer is collecting dust. As a Linux-only user the proprietary alternatives mostly aren't available to me, and FreeCAD is bad enough that I'd rather not do CAD at all. The Ondsel fork was looking promising for a while, but sadly that died off.
Maybe try OnShape, which is browser based and quite professional?
LibreOffice, GIMP, FreeCAD and Inkscape all have their quirks (and bugs), but they're probably seen as features by their core users so they won't change.
Possibly Audacity, given the direction v4 is taking. Great video essay / update here: https://youtu.be/QYM3TWf_G38
>FreeCAD isn't even in the rear view mirror of Solidworks.
I think FreeCAD might be on a distant hilltop in their rearview these days, check it out again.
>>It has rendered Maya nearly obsolete
Hadn’t heard that. How many AAA vfx studios have left Maya for Blender?
I’m rooting for gimp but they are truly a decade behind photoshop. I don’t know how they can compete. Luckily it does not seem like they are in the market for professionals so the need to match Photoshop isn’t quite as high
> It has rendered Maya nearly obsolete.
really? I haven't done 3D rendering in a long time, admittedly, but back then Maya and Lightwave were miles ahead of Blender. Rhino3D too. Even 3DSMax was better. Lightwave seems to have sadly fallen off (unfortunately, IMO it was the best at one point, had excellent ray tracing). I didn't really Blender had come such a long way -- that's great.
Since you mention niches: Adobe InDesign has no OSS competition at all, and Illustrator is still much better than Inkscape.
I’d really like to see something like blender come for the 3D CAD industry, at the moment it feels like the only people who would lose out are AutoDesk. The amount of money that flows in and out of 3D cad (as subscription and then value created) having a first class open source kernel and tooling, would be giving big industrial players freedom to modify and tailor to their needs as well as smaller / hobbyists get started for free!
Probably requires something that is almost there then a sponsor(s) to throw in developers or funding to get the rest of the way. On the EDA side CERN did a lot to lift Kicad to the point of being a credible alternative that could breakthrough like Blender. Both those projects are over 30 years old and for a lot of that time were dismissed as too difficult to use or lacking in features. FreeCAD is only 23 years old. I don't know what the code base is like but if a large org put a couple of good devs into it for a few years who knows.
It must be difficult when so much management is short sighted and focused on delivering short term profits for shareholders. Even academia is run like a business now.
Unless a privately held rogue company like Valve got interested its probably going to have to wait for a government/ngo/scientific. Industry, particularly the tech industry, is notorious for leaching of free and open source software and in some cases building entire businesses on it and not giving back.
The problem with FreeCAD and every other free/open source MCAD project of note is the Open Cascade kernel they are built on. While Open Cascade is fairly mature, it has dealbreaker issues in a few key areas: fillets cannot consume connected faces and may fail for a number of other reasons, cylindrical and spherical faces require seams which often cause issues with boolean operations, and shapes like helixes are also often troublesome.
On a scale from "big chunk of work" to "complete rewrite", how much work would it take to fix those issues in Open Cascade?
Sandia seems to have some form of kernel, but only Federal-associated entities can get access to it.
It would be interesting to see if they would license that out further for some amount of money.
If you're referring to Cubit, they license the ACIS kernel under the hood.
They’re (possibly) referring to “Scalable Geometric Modeler” (SGM)
https://github.com/sandialabs/sgm
Originally open-source, but since taken back in-house. As I understand, which should not be construed as an accurate accounting, Sandia wants to flesh out the basics further before (potentially) open-sourcing it again.
Onshape has blown me away with its browser interface and how quick it all loads. And as long as your projects are public, it's free.
Wish they had something between $0 and $1500/yr.
I really didn't like how they both require a phone number and make all of your files public. I've stuck with Fusion which seems a lot more privacy respecting while also being basically free for home users.
agreed!
take a look at https://Plasticity.xyz. It's not open-source, but it's got a small, highly dedicated team behind it. It's built on Solidworks' kernel, so it's quite robust.
Also take a look at solverspace, caligula, FreeCAD, ...
* Node-locked (up to 2 machines)
Hard nope.
IDK how they compare to professional CAD tools, but I've heard good things about FreeCAD and OpenSCAD. I know that some people use Blender for CAD work, and there are even some extensions to make it easier, but I'm dubious that the representation of meshes that Blender uses are well-suited for CAD applications.
I just tried FreeCAD last week. I uninstalled it after about 10 minutes. The most basic actions to just get started were throwing errors. Maybe it was user error, but it was a very bad first impression.
Approached it with the same attitude at the same time, after 10 minutes decided to view some basic tutorial (for an earlier release) that made things clear and I could continue basic tinkering on my own.
But of course built-in intro of Solidworks was a way better UX.
Do you mind sharing what you were trying to do? I love FreeCAD so I'd be happy to help you do it if you'd be willing to give it a second try.
My main goal is to reproduce the floor plan of my house, so I can figure out how to best layout the furniture.
I actually did the same thing so that I could figure out how to lay out my workshop!
What I'd do is:
- Spreadsheet workbench --> Create spreadsheet (name it "measurements"). (This is optional)
- Switch to Part design workbench --> Create body (name it "layout") --> select XY plane --> Create sketch --> Create Polyline
- Zoom out, start drawing the rooms in your house, approximately to scale.
- Before going into too much detail, add a dimension (select line --> "Constrain Distance") to the first line you draw, so that you can do the rest of your drawing approximately to scale. Then the general shape won't get messed up when you add dimensions to everything else.
- (If you have a photo or picture, you can import that to sketch over).
- Add constraints to match your room measurements, mostly vertical or horizontal distance constraints. Be careful not to overconstrain the sketch. (You can put the measurements directly into the sketch constraints, or you can put them into the top-level spreadsheet, create an alias for each cells, and then set the dimensions to reference those cells).
- Once the rooms are drawn, close the sketch and create a new sketch on the xy plane called "furniture".
- Draw some rectangles for your sofas / tables / etc, delete any horizontal and vertical constraints that get automatically added (they look like little | and _ icons), and instead apply perpendicularity constraints. Dimension your rectangles using only the "constrain distance" tool. Now you can drag them around the room and rotate them freely.
- If you want to make 3D models for these too, create new Part Design bodies for each room and each piece of furniture, create a shape binder referencing the master sketches in the Layout body, and then extrude the sketches using the "Pad" operation.
That's about as much tutorial as it makes sense to pack into a HN comment. If you give it a try, I hope it works out for you!
Thanks! I’ll save this and give it a shot soon.
I like using paper and cardboard for this, dollhouse style, much easier to move things around and visualize that way and more fun than clicking a mouse to boot :^)
Inkscape is good for typing dimensions into rectangles tho
This was my plan B. I do have a scale I can use for it.
I’ll check out Inkscape as well. I’ve tried using some raster graphics in the past, but I couldn’t type dimensions and had to use the rules and guides with snapping. It mostly worked, but was a bit annoying.
Blender is a decent option for low effort 3d modeling for 3d printing in my experience
During COVID I learnt Blender for 3D modelling. It is still my go to.
Many people complain about it being a mesh editor but it works for me. The sheer variety of tooling and flexibility in Blender is insane, and that's before you get to the world of add-ons.
I want to learn Geometry nodes and object generation as I think they will address a lot of the "parametric" crowd concerns. This v5 is meant to be a big step in ease of use of this.
Also, I'm not sure if the different tooling lets me see all the flaws of online "parametric" models, or whether I'm being pedantic. They get frustrating. I have Gordon-Ramsay-screamed "How can you fuck up a circle!".
In MCAD, “parametric” does not mean a high level part or feature is driven by editable parameters or procedurally generated features. Parametric refers to the underlying storage format representing part features in a parametric way rather than as a mesh. Mesh formats like stl cannot represent a circle by its position and radius, while a parametric format like step can. This distinction is more akin to raster (bmp) vs vector (svg) graphics. Both can be generated procedurally by “parameters”, but only with svg can sub-features be faithfully extracted or transformed.
Sorry, separate point:
>Mesh formats like stl cannot represent a circle by its position and radius, while a parametric format like step can.
This is where I think the Geometry nodes can help. A node (function) can be used to represent the circle with inputs and outputs set or changed as required.[0]
I have not fully explored this space though and so my "hopes and dreams" may well be as useful as thoughts and prayers...
[0] https://docs.blender.org/manual/en/latest/modeling/geometry_...
I have some understanding of "parametric" vs "mesh". I looked it up when I saw so many people going on about it.
Maybe it is the export or something. I run the 3D toolbox and often models are not manifold.
I see things like two circles in slightly different positions but both are connected in different ways to the surrounding "single" instance model. Things like this mean you end up with "infinitely small volumes". There is no fully enclosed "volume" and so mathematically there is "nothing to 3D print".
As a model this makes no sense to do, and so it irks me.
But clearly the slicer software doesn't care or autocorrects and people make their 3D print happen just fine.
Alibre has a free option, which does not include sheetmetal bending but otherwise solid software
Depends if your goal is artistic or functional. Blender is good if you are trying to make character models, etc. It's not great when you are trying to make a part that has to fit something in the real world and after printing you discover one step half way through needs to be 1mm shorter.
It does take a different set of skills to regular CAD, but I haven't found it that bad for simple 3D printed models that need to be dimensionally accurate.
I have used it to make quite a few functional prints, with the help of making sure my scene units are correct and a CAD plugin.
I haven't tried any of the cad plugins for blender, but I'm not sure how you would retroactively change dimensions in blender. It's usually simple enough to create features to a certain size, but if you need to change them later it becomes significantly difficult.
If I put some holes in something that are 1mm from the edge, but then I print it and see it doesn't line up and needs to be 1.5mm, in Fusion I can just change one number and it all updates. Doing the same thing in blender would likely be very difficult.
I use Plasticity to model for 3D printing. Having to worry about polygons in Blender is really annoying.
So model using the NURBS tools?
Does Blender have NURBS? I don't even use NURBS in Plasticity, because curves are already essentially vectors. I don't have to worry about polygons at all, and then I choose the tolerances when I export.
Same, but I don't think it's possible without a large and sustained investment into a free geometric modelling kernel, which can probably be only done by a government.
Parasolid is powering practically every major CAD system. Its development started in 1986 and it's still actively developed. The amount of effort that goes into those things is immense (39 years of commercial development!) and I don't believe it can be done pro-bono in someone's spare time. What's worse, with this kind of software there is no "graceful degradation": while something like a MIP solver can be useful even if it's quite a bit slower than Gurobi, a kernel that can't model complex lofts and fillets is not particularly useful.
3D CAD is much harder than Blender and less amenable to open source development.
> Same, but I don't think it's possible without a large and sustained investment into a free geometric modelling kernel, which can probably be only done by a government.
Can you help me understand why this problem is so hard?
The other reply is really good. To add to that, the intersection of two bi-cubic Bezier patches (each being but one part of a cubic B-spline surface) is an implicit equation of degree 18^2=324. This simply cannot be implemented exactly in a geometry kernel and thus must be approximated. How do you want to approximate this? If you choose trimming (the industry standard) now you have to handle gaps in your geometry. If you choose remapping into an unstructured watertight spline, you need to solve a constraint system that is NP-hard. If you choose reparameterizing… well, see nVariate’s watertight Boolean technique (disclaimer: I once sponsored a project with nVariate).
Now, generally speaking, in a CAD model most surfaces will be “analytic” (plane, torus, conical, arc, line, etc). But whenever some complex surface that joins these surfaces is required, (NUR)B-splines are the principal technique for “covering” the gap.
The way modern CAD systems work is by having a tree of features/actions that is then used to construct an analytical representation of a 3D object. The features/actions can rely on "sketches" (2D drawings that are coupled with a real time geometric constraint solver) and can be "projected" into sketches, creating new reference lines, that can then be used by the sketch constrain solver, generating a sketch that can be used for more 3D features.
This is already complex and fiddly enough. Just having a stable 2D drawing environment that uses a constraint solver but also behaves predictably and doesn't run into numerical instability issues is already an achievement. You don't want a spline blowing up while the user is applying constraints one by one! And yet it's trivial compared to the rest of the problem.
Having 3D features analytically (not numerically!) interacting with each other means someone needs to write code that handles the interactions. When I click on a corner and apply a G2 fillet to it, it means that there's now a new 3D surface where every section is a spline with at least 4 control points. When I then intersect that corner with a sphere, the geometric kernel must be able to analytically represent the resulting surface (intersecting that spline-profiled surface with a sphere). If I project that surface into a sketch, the kernel needs to represent its outline from an arbitrary angle — again, analytically. Naturally, there is an explosion of special cases: that sphere might either intersect the fillet, just touch it (with a single contact point), or not touch it at all, maybe after I made some edits to the earlier features.
Blender at its core is comparatively trivial. Polygons are just clumps of points, they can be operated on numerically. CAD is hell.
> Same, but I don't think it's possible without a large and sustained investment into a free geometric modelling kernel, which can probably be only done by a government.
Fornjot has been attempting this: https://www.fornjot.app
It's going to be years or decades before it's competitive though. Also, it looks like they switched to keeping progress updates private except to sponsors, which means I don't actually have any easily-accessible information about it anymore which is sad.
I'm very skeptical that one person can make a dent. Paging through the releases, they seem to focus on constructive solid geometry and code-driven shape generation, which I believe is a dead end.
The tricky bit is having a G2 (or even G3) fillet that intersects a complex shape built from surface patches and thickened, with both projected into a new sketch, and keeping the workflow sane if I go and adjust the original fillet. I hope one day we'll see a free (as in speech) kernel that can enable that, until then it's just Parasolid, sadly.
FreeCAD is the front-runner for me.
KiCAD was also a meh ECAD FOSS alternative 7-8 years ago, now it is by far the tool of choice for regular ECAD designs. I can see FreeCad getting there by 2030.
FreeCAD is probably the single most frustrating and unintuitive pieces of software I have ever used. I almost drafted hate mail to the devs after 15 minutes of crash coursing fusion360 got me further than 2 days of trying to use FreeCAD.
It seems like it has lots of capability but still "punch your monitor" levels of difficulty just trying to do the most basic stuff.
I use FreeCAD and it's pretty good. But I think it's impossible to learn by trial and error.
MangoJelly has done an amazing job in churning out high quality tutorials for FreeCAD: https://www.youtube.com/watch?v=t_yh_S31R9g&list=PLWuyJLVUNt...
(this is just one playlist, there's a lot more on his channel).
While it's a pain to learn and requires some plugins (addons) for basic ergonomics, FreeCAD absolutely works for parametric CAD modeling. YMMV depending on the project and complexity, it does the trick for laser cutting, bending and 3D printing.
Deltahedra is a great YouTube channel for getting the basics.
It suffers from too many "workbenches," some of which appear to be redundant or dated. You never know whether structures created by one are "compatible" with the M.O. of another (like "Part" vs. "Part Design").
And it presents nonsensical problems, like offering to create a sketch on the face of an object and then complaining that the sketch doesn't belong to any object. So you have to manually drag it under the object in the treeview. So gallingly DUMB.
Despite all that, I will wrestle with its ineptitude before giving Autodesk a penny. I get stuff done with it and respect those who give their time to develop it.
how long ago did you try? the recent advances have turned me into a believer as a hobbyist compared to the first time i checked it out.
I tried it this year. Not in too much depth, but I tried Fusion and FreeCAD for the first time this year for 3D printing and found I was getting much further much faster on Fusion.
I'm sure I could grind harder and learn more and make FreeCAD work, but I'm not sure why I'd bother.
No arguement that fusion isn’t better and easier to learn. Their licensing and changes to their hobbyist offering were no longer tenable for me which prompted my change. I was pleasantly surprised at how well I was able to work in freecad after a few youtube videos.
Something that I've daydreamed of in the past is making a new constructive solid geometry "kernel" around which a CAD application could be wrapped based on projective geometric algebra (PGA).
The algorithms it enables are fundamentally more capable and robust than traditional kernels based on linear algebra (vectors and matrices). You can do really fancy things like interpolating in space and time robustly, find extrema in high-dimensional phase spaces, etc...
This could potentially allow straightforward and robust solvers for kinematics, optimal shape finding, etc...
Every few decades there's a "step change" where some new algorithm or programming paradigm sweeps away the old approach because suddenly a hobbyist can do the same thing solo that took dozens of developers a decade in the past. I suspect (but cannot prove) that PGA is one of those things.
Plasticity is the closest thing to this, I think. It uses Parasolid, which Blender does not, and supports xNURBS, which Blender does not.
Plasticity is closed source though?
You may be interested in CAD Sketcher: https://www.cadsketcher.com/
You're on point that there's a tremendous amount of money captures by Autodesk for CAD software that could be better directed at the open source community instead.
Software like OpenSCAD and FreeCAD are obviously not suitable for much commercial work, and have very irritating limitations for hobbyist work, in my mind a big part of that is the UI and Blender has a good and established UI at this point so I'd love to see the open source CAD that provides an alternative to vendor lock in come from a Blender add-on instead of a separate program.
I am no expert but as I understand it the primary difficulty with developing good alternatives to commercial CAD software lie in the development of an effective geometric kernel.
It seems to me that if a developer of an opensource CAD program develops it as a Blender add-on they can effectively outsource the remainder of the development efforts to the Blender community while focus can be made on the CAD kernel itself.
What does Blender do differently that makes it such a successful open source product?
It’s powerful and pleasant to use. Even the release marketing page is beautiful and well-made.
I like open source as much as the next guy, but outside of developer tools there is little that comes close to Blender in terms of utility and UX.
Is it funding? Specific individuals? Are there PMs and designers? Whatever it is, it’s working!
I think at the very least they've found those rare individuals who can both code well and manage well. There are countless open source project leaders who might be the most knowledgeable person in the world on that area but have no idea how to communicate and collaborate in an open-source setting, especially as the project gets popular and attracts people who're detrimental (not necessarily malicious) to the project's growth.
Yeah this is a really good question. I'd love to know more about that too. There's some good history on the website here [1].
The creator apparently was selling it as freemium software in 1998, and then the bubble burst and the corp shutdown in 2002. But the creator created a non-profit called the Blender Foundation, launched a Free Blender campaign [2] (the forum post is still up!) to raise money from its users and bought out the rights to the software from the investors.
[1] https://www.blender.org/about/history/
[2] https://blenderartists.org/t/free-blender-campaign-launched/...
Every time Blender has a new update, I scroll and I'm amazed by how much is in it. Then I realize my scrollbar is only halfway through.
Radial tiling my beloved, and a seemingly far more straightforward array modifier <3 Faster volume scattering for non-homogenous volumes.
For those wondering "where the AI is", the new Convolve Node might be it :) Convolutions are a pretty generic signal processing operation (Hadamard product) which are also used in neural networks which work with images. Realistically though, this will be mostly useful for wonky hand-crafted blurs.
The new sequencer looks fantastic, too. I always went to DaVinci Resolve but I might be able to go full blender. Compositor modifiers in the sequencer is also very welcome.
This is incredible for me.
I've been using blender since at least 2010; it's so exciting to see how much progress it's making.
I'm very excited to see the addition of structs and closures/higher-order functions to blender nodes! (I've also glanced at the shader compiler they're using to lower it to GLSL; neat stuff!) Not only is this practically going to be helpful, the PL researcher in me is tickled by seeing these features get added to a graphical programming language.
If you haven't heard of Blender before, or if you think AI will replace all the work done in it, fair enough. But I'd still strongly suggest looking into what it is and how it works.
I've used blender since 1999. It's my favorite open source software. Simply amazing
So...... was there ever a resolution with that persistent incompatibility in ROCm/certain AMD drivers and Cycles that made it impossible to render in almost every version of blender? As in, it literally doesn't even detect the GPU outside of eevee
Can you link to a Blender issue?
The SDF geometry nodes open up a wild amount of new options. That and node closures solve so many difficult problems, especially in complex scattering. If you haven't checked them out, here is a bit on them and volume grids: https://code.blender.org/2025/10/volume-grids-in-geometry-no...
Very very sad that the adaptive subdivision is touted as a Blender feature but unfortunately it's a Cycles feature.
Always nice to see these updates though, Blender has really come a long long way.
It might be possible to reproduce the same effect in EEVEE using geometry nodes. I know people have done that for automatic level of detail work. That being said, IDK if subsurf as a geometry node will take a non-constant number of iterations.
Blender is amazing for 3D Modeling. But metaverse is still dead. RIP
Awesome
The first thing on the website is a Cloudflare Captcha box :/
But after that, all the other things on the page are AWESOME! I’m super stoked about the proper HDR support and all the new node improvements.
Yeah. the HDR support is very nice. ACES got their system right the 2nd time around thankfully.
Is this the first blender release where you can change the working color space? I thought that you could in previous versions but it caused issues with some nodes.
Now I want to look into it more, but I'd imagine that "Blackbody" and sky generation nodes might still assume a linear sRGB working space.
> Now I want to look into it more, but I'd imagine that "Blackbody" and sky generation nodes might still assume a linear sRGB working space.
Since people are always asking for “real world examples”, I have to point out this is a great place to use an agent like Claude Code or Codex. Clone the source, have your coding assistant run its /init routine to survey the codebase and get a lay of the land, then turn “thinking” to max and ask it “Do the Blackbody attribute for volumes and the sky generation nodes still expect to be working in linear sRGB? Or do they take advantage of the new ACES 2.0 support? Analyze the codebase, give examples and cite lines of code to support your conclusions.”
The best part: I’m probably wrong to assert that linear sRGB and ACES 2.0 are some sort of binary, but that’s exactly the kind of knowledge a good coding agent will have, and it will likely fold an explanation of the proper mental model into its response.
why ACES and not something like P3?
Display P3 (distinct from cinema display P3, because names are hard ig) is used as a render target color space. ACES (and its internal color spaces) are designed as working spaces.
If you make a color space for a display, the intent is that you can (eventually) get a display which can display all those colors. However, given the shape of the human color gamut, you can't choose three color primaries which form a triangle which precisely contain the human color gamut. With a display color space, you want to pick primaries which live inside the gamut; else you'd be wasting your display on colors that people can't see. For a working space, you want to pick primaries which contain the entire human color gamut, including some colors people can't see (since it can be helpful when rendering to avoid clipping).
Beyond that, ACES isn't just one color space; it's several. ACEScg, for example, uses a linear transfer function, and is useful for rendering applications. A colorist would likely transform ACEScg colors into ACEScc (or something of that ilk) so that the response curves of their coloring tools are closer to what they're used it (i.e. they have a logarithmic response similar to old-fashioned analogue telecine machines).
no monitor uses ACES so it always needs to be converted to P3 to even see what you're doing right?
or you are saying if there is some intermediate transform that makes color go beyond P3 it will get clipped? then I understand...
The key point is that your ray tracing color space and your display color space don't need to be the same thing. Even if your monitor only displays SRGB colors, it still can be useful to have more pure primaries in your rendering system.
Looks like they picked a bad day to do a major release.
That's the whole internet now. That or Anubis.
Or do the rational thing and rate limit GET requests to human speeds.
Based on what fingerprint?
Fingerprint: *
What’s the consensus on the future of this type of 3D tool? Especially for video animation/CGI in movies/tv/ads?
Seems like in 10 years AI will basically make it pointless to use a tool like this at least for people working on average projects.
What do folks in the industry think? What’s the long term outlook?
If you don’t work in the industry, you have zero chance of accurately evaluating whether or not, or how, it will be impacted by any new technology.
The fact that it “seems easy” is a great flag that it probably isn’t.
Industry has no idea how they’re going to be impacted either.
Really no one can predict the future.
This seems like an unnecessarily unkind response. The post you're replying to is sharing their opinion and asking what people who are in the industry think about it.
Zero?
"AI will make this pointless" is so exhausting.
Directors spend a LOT of effort trying to keep continuity and that's the weakest part of AI.
What blender and other CGI software gets for free is continuity. The 3D model does not change without explicitly making it change.
Until we get AI which can regenerate the same model from one scene to the next, the use of AI in CGI will be severely limited.
I design projections for independent theatre in Baltimore. I use AI in my workflows where it can help me and won’t compromise on the quality of what I’m making. I frequently use AI to upscale crappy footage, to interpolate frames in existing video (for artistic purposes, never with documentary archival stuff) and very occasionally to create wholesale clips in situations where video models can do what I need.
I recently used WAN to generate a looping clip of clouds moving quickly, something that’s difficult to do in CGI and impossible to capture live action. It worked out because I didn’t have specific demands other than what I just said, and I wasn’t asking for anything too obscure.
At this point, I expect the quality of local video models (the only kind I’m willing to work with professionally) to go up, but prompt adherence seems like a tough nut to crack, which makes me think it may be a while before we have prosumer models that can replace what I do in Blender.
If AI is at the point where it is exactly as capable of your average junior 3D professional in 10 years, it will probably have automated a ton (double digit percentage?) of current jobs such that nothing is safe. There's a lot of complexity, it's fairly long time horizon, it's very visually detailed, it's creative and subjective, and there's not a lot of easily accessible high quality training data.
It's like 2D art with more complexity and less training data. Non-AI 2D art and animation tools haven't been made irrelevant yet, and don't look like they will be soon.
As someone who's actually used Blender for small video projects, I'm fairly confident you'll still need this type of tool even with AI assistance doing some of the work in it, especially for at least the next 10 years.
AI coding agents didn't make IDEs obsolete. They just added plugins to some existing IDEs and spawned a few new ones.
Can there please be one post on this godforsaken website where there is no attempt to shoehorn it into the AI craze?
Oh you must think you are reading Hacker News, sorry about that, this is actually AI Optimism News.
You build apps for Shopify.
You are asking for industry predictions from industry professionals in an industry you know nothing about while assuming a lot about that industry.
Why do you think they should do all the heavy lifting for you?
You might as well ask ChatGPT what it thinks because it seems you already have an idea of what you want the answer to be.
What will AI train on?
3D scans of the real world?
I don't think the "what will it train on" argument is bullet-proof, but animation and 3D art can encompass so much more than just things that exist in the real world.
Famously, all 3D art is of things only found in the real world.