Wow, I did not expect this. I started looking at Fusion 360 a couple years back because it seemed very hobbyist friendly, and while I primarily use it for 3d printing, I've dabbled in laser cutting and CNC too.
I always had the impression that Autodesk had positioned Fusion 360 as their gateway drug into AutoCAD or some other Autodesk software, but that it was remarkably full featured for that purpose.
Last year (or was it the year before?), they added some restrictions to their free version. No biggie, I thought. Most of the cool hobbyist functionalities remained intact.
Now, they have taken more away. The file export limitations in particular are crippling. No .step? .dxf? No extensions??? I can't even buy stress simulation credits?
US$495 a year to subscribe to the "full" version (they have "generously" granted a 40% discount until Oct, hurry before it's too late!). No cheap versions for hobbyists like me who just like to dabble and for whom a yearly sub is out of the budget.
Any suggestions for free/cheap and functional local-install alternatives to Fusion 360? Parametric modelling has really saved my butt, and motion joint functionality is very useful too.
I totally resonate with your reaction. I wrote a response to their email out of hopelessness. Pretty sure it won't be read, but as it summarizes my thoughts I can post it here as well:
> Hi Shannon,
> I thought for some years that Autodesk was a smart company.
> Most people I know in the makers community were using Fusion 360 because it was free, and when you are a hobbyist this makes a big difference, you are not gonna to spend hundreds of dollars for simple side projects. Your previous approach of allowing people to use Fusion 360 for free unless they made money out of their project was really smart. These people were anyway not going to be your clients so you (almost) did not "lose" money, and on the other side they got used to your product so when moving to a professional environment they would push for Fusion 360. It was kind of a win-win approach, with the additional effect that Autodesk looked like a hobbyist-friendly company.
> Today you did not just break the main use-case for hobbyists by disabling the export in DXF, but you also showed that you prefer to spend money to block existing features of your software instead of improving it with new paid features. That's an indicator that you don't care about the product itself anymore, but that you are just trying to extract the maximum money out of your current users instead.
> I don't see the point to support a company like that anymore, so I deleted my account. It is now useless anyway.
> You did not only lose a user, you lost my trust. And probably not only mine.
FreeCAD is the big name, but I'd check out SolveSpace too. Then there is of course OpenSCAD which works well for the things it happens to work well for, but I wouldn't call it full replacement for a CAD. To wrap up there is the oldie and honestly bit intimidating BRL-CAD, which I'm kinda trying to learn whenever I feel like I need to do mechanical design.
I feel LaTeX relates to MS Word like OpenScad to FreeCad.
The scripting makes making technical parts often much easier, and more maintainable. I for instance wanted to replace a metal rod in an antenna with a smaller diameter, and changing several mounting parts for the rods became changing a single variable. Plus, being eble to comment scad files makes it much easier to meaningful share parts (the STL files on Thingiverse are hard to modify).
The usual caveats apply: people tend to either shy away from the few days it takes to learn OpenScad, or, once they invest the time, there's a tendency to over-engineer scripts and not focus on tasks at hand ;)_
has anyone used form•Z recently? i remember liking it much better than autocad in college (waaay back in the day), simply because it was much easier to get started modeling (but not nearly as integrated into the engineerng workflow).
The natural result of this change will be piracy, especially for non-commercial users. People tend to forget that paying $0 for a crack is always an option and entirely within the scope of the free market. Either way i hope Autodesk sees an apparently plummeting user base and reverses course.
I'm using FreeCad for simple 3d-printing, and it mostly works, has all the features I need, but you have to struggle with everything.
The dependencies between sketches break when you change anything earlier in the chain (like drawing a circle for a hole in a wall that was 6 polygons and now that wall is 8 polygons and the circle jumps to another corner or somewhere) because automatic naming of stuff isn't consistent.
You can work around this by always using datum planes as the basis for each sketch but it's very inconvenient and makes editing sketches hard because datum planes occlude other stuff on the sketch. There's a display mode that renders only lines, but the planes remain semi-transparent and mess with selecting points behind them.
Parametric modeling is possible but it also breaks very easily if you edit anything up the chain because of the above problems, to the point it's almost useless.
You also can't export to anything used by 3d printers other than .stl, and the tesselation configuration is weird (you have to change configuration of minimal angle between faces generated from round surfaces in totally unrelated configuration tab - it took me a long time to find it).
I haven't tested 90% of the features, including the motion joints.
Also it crashes sometimes (like once a day) - you have to save often.
But it's free and it mostly works and I didn't had to deal with the licencing and registration, and it runs locally and I can use git for backups, so that's good.
I also noticed that sketches can be messed up if you make changes earlier in the chain. BUT: if you get in that state, you can use the 'Map to Sketch' feature to repair it. It's a bit clumsy- even after mapping, you have to click on the sketch, and the object made from it, to get it to recalculate. But this has saved me a fair amount of rework when fixing something earlier in the chain.
Also, when you're working on a sketch you can go back to the object tree and use SPACE to disappear objects temporarily if you are having trouble selecting.
I don't know what you mean by tesselation configuration- ah, well that explains why I had undersized holes (I print a lot of parts with shaft holes, or holes for bearings).
v0.19 is a lot more stable than any previous version I've used- I still have a crash a day too.
To fix that I had to increase details, and to do that I had to select completely different task and go to configuration of some other (non-stl) format and change maximal angle error or sth like that. I don't remember but I remember it was far from obvious.
I know you specified local-install but I've found onshape better than any of the alternatives I've tried. It's web-based but I've never had performance issues and having easy access to my designs from any computer is an added bonus.
No surprises. For all the arguments around FAANG and user hostility, Autodesk are the absolute pits. They have been price gouging for decades. They release buggy software and hold various industries hostage with their pricing. For an Architect with 100 users of Revit as an example, the fees are eye watering, and the support 'contract' is always with 3rd parties, which is invariably extremely limited. Horrible organisation.
Some of the changes make sense - IE they won't store your documents for free any more, and you can't use their server rendering farm for free. But others are just punitive, like not letting you export certain file types.
Does Blender even have any CAM/CAE functionality? Parts tree, version management, integration with CNC? It seemed like Blender is to CAD as MS Word is to Visual Studio. Different worlds that are superficially alike.
I used OpenSCAD for several years of hobby-Level 3D printing modeling. I switched to Fusion360 during lockdown and it was night and day more productive, especially for modeled threads, fillets, and constraint-based models.
I've always found it a little surprising that companies haven't built/funded a open-source organization for parametric mechanical CAD, similar to Blender for games or KiCad for electronics/PCB design.
The upside for a ROI seems like it would be enormous, and pretty easy to make happen. Industrial seat licensing for products like Solidworks, CATIA, Inventor, and (now) Fusion are enormously expensive. It's not just private organizations either--CAD proficiency is such a basic skill that every serious engineering school has an organizational license for their students too, which I'd imagine also costs a bundle. It's not as if the tools themselves are expanding functionality at some sort of rapid rate, either; I haven't done much CAD in the last two years (so maybe I somehow missed some sort of feature explosion), but while a regular user between 2013-2018, I saw basically no change in the vast majority of my most-used tools for several different CAD programs, with the exception of some improvement in out-of-the-box simulation capability.
The existing FOSS alternatives just aren't at par. I've tried FreeCAD, OpenSCAD, etc., and most mechanical engineers/CAD specialists wouldn't touch a code-based editor. They're certainly better than before, but the rough edges exist and some of them appear in areas that need to Just Work (like assembly and drawings). From my understanding of their contribution graphs, private organizations putting even two or three full-time developers working on those projects could push them much, much closer to being a drop-in replacement for a lot work that gets done, and potentially even save those orgs some money in the short-term by reducing the total number of seat licenses they need in to function.
Edit: Added mention of KiCad, bc that's kind of important too
I think the tipping point for KiCad was when CERN realized how much they were spending on professional EDA licenses, and decided to pour those resources into improving KiCad instead. That's how I understood it, anyway.
I suppose some large-ish entity that relies on CAD will need to take the same view. I have less faith in academia there, however -- I suspect many of those sitewide licenses are heavily discounted in order to cultivate the addiction, to profit from it later. That's just speculation, though.
This is where we really need good software bounties. I'm sure a whole lot of scorned F360 users would throw a few bucks a month into a Patreon or something to advance the state of FreeCAD. I'm not even directly impacted by this; I have a F360 license through my employer, but I'm offended enough on principle to contribute if there's a place to do so.
> when CERN realized how much they were spending on professional EDA licenses
I've worked in national labs, and the bigger concern is how to guarantee the ability to open or update the current files in one or two decades.
(It was common to see foot-high printouts that faded over the years and became illegibile, to give an example of the time scale we're talking about.)
Hobbyists have a similar problem in that they want to be able to "hibernate" a project and update drawings in 5 years, but without thinking about monthly cloud fees. So free or perpetual licenses also make more sense for that use case.
Storing drawing files long-term in a vendor's cloud is undesirable if you really want local files for archival or distribution purposes.
>> I've always found it a little surprising that companies haven't built/funded a open-source organization for parametric mechanical CAD, similar to Blender for games or KiCad for electronics/PCB design.
Unigraphics was owned by McDonnell Douglas and then sold to EDS, which was a subsidiary of General Motors. As far as I know, GM still uses Unigraphics as their primary CAD system.
> I've always found it a little surprising that companies haven't built/funded a open-source organization for parametric mechanical CAD, similar to Blender for games or KiCad for electronics/PCB design.
Salome is supported by EDF (French national energy) and BRL-CAD by the US Army.
It's a shame that Autodesk is cutting back on features for free personal use, especially the removal of STEP export (an open interchange format) and multi-axis CAM (used to convert 3D models into instructions for milling machines & lathes).
But - $495 for a CAD program as fully featured as Fusion 360 is a steal compared to $$$$ Solidworks & competitors. Plus it's cross platform which you don't typically find in the world of CAD.
Is FreeCAD a viable alternative? Not even close.
Should there be a usable FOSS CAD program? hell yes!
Unfortunately the FOSS foundations required to build a real CAD program haven't materialized in cohesive way. There are so many pieces, with many requiring teams of PhDs to produce: CAD kernels, constraint solvers, file format interop, sketching tools, assembly support (mating), simulation, etc... There are some advanced projects out there like the OPENCASCADE kernel, but Parasolid is more reliable and has real support through Siemens.
Ugh. AutoDesk (sort-of) did this with Eagle (PCB design) when they acquired it -- in that case it drove me to learn KiCAD which in the end was what I needed to do! (and I think the AutoDesk acquisition of Eagle also invigorated interest and spurred development of KiCAD!).
I recently got a 3D Printer and have been learning Fusion360 and designing my first parts. Guess I'll take another look at FreeCAD.
In general, I'm happy to pay for things that "have value", but the pricing here ($495/year) is just too much to justify for a hobby.
The expectations that the blog sets up are in the positive mood. It should really start off with, "This is going to really suck, everyone, and in fact it might be the worst thing you heard about this year. Just kidding, it's slightly better, but...". But instead it builds up positive expectations. Anyone sufficiently jaded just has to fight the urge to believe this fluff while desperately scanning to find the place where they tear it all down and tell you what is happening with the product, after needlessly building up for paragraphs at a time.
I think this announcement is fine. No language on earth would make free users happy, it gets right down to business, and it includes a clear and specific table showing what’s being changed. There isn’t really anything else to be done - an apology would just be decried as throwaway corporate speak or as mocking the helpless.
You can't make them happy, so at least be honest. The lede is buried, and user-blaming language like "We need to eliminate both the confusion and the misuse that exists within our offerings," implying that the users did something wrong, when it was in fact the company that got the pricing wrong.
Not to mention the misleading infographic with the checkmarks (which displays the important information in a more dim, harder-to-notice color and displays old outdated information in bold black which goes top to bottom) or the interstitial with the dark pattern button that says "I don't want to save 40%" in order to dismiss.
I imagine there's a lot of 'free' users that are using the product to make a significant profit, as they seemed to quite precisely be cutting out the steps where they are part of a 'toolchain' and not just an 'all-in-one basics' solution. They're clearly also worried about the cloud costs that were incurred by the free users in various respects, which suggests that too many people were not paying for it vs. costs. That confusion, that misuse, is clearly their own fault for allowing it to happen at all. The quote you provide doesn't read as user-blaming to me, so I'm not sure how to respond to that point.
I just uninstalled and wrote "fusion 360 hobby announcement" as the reason. I'm just a hobby user, so I don't have time to figure if/when I'll run into their new limitations. Just removed it and I'll go learn FreeCAD or something similar. I also use OpenSCAD where applicable.
There really isn't anything. For at least a couple of years, Fusion 360 has been available for "free" (users had only community support, had to write their own POSTs for cam, etc) so as far as I know no one has been developing or considering writing anything else.
One thing that's been needed for years in the open source space is a decent mechanical CAD program and integrated CAM system.
The issue is that such things are still dark arts to write.
There are open source projects that are out there, including BRLCAD, Blender, and other special purpose programs.
Fusion 360 is head and shoulders above them in terms of usability and integration with the CAM system.
FreeCAD definitely does not have the full feature set and is still quite rough around the edges but it is improving at a remarkable pace. You still have to deal with some silly things like multiple workbenches for the same thing, lack of decent (any?) part mating support, lack of huge databases of fasteners and standards and so on, but the fundamentals are there. It's pitched as a free replacement for professionals rather than for hobbyists, but the kind of features that make hobbyist CAD software easy to use are the type of UI design that's incredibly expensive so... you get what you pay for.
Fusion 360 is a huge beast that is really ten+ applications in one.
I can't speak for all of its use-cases, but even then: I haven't found any 3D modeling software that is as easy to use & feature rich as Fusion. However, for 3D print slicing and compiling after the model has been made, Ultimaker Cura is really a gold standard in the space, and open source (AGPL3 I believe).
An often-overlooked option is Salome . Although it's primarily intended as part of a pipeline for computer simulations, it's honestly about as functional as FreeCAD (with which it actually shares a lot of code, lately including the 2d constraints engine) and a good deal more stable, in my experience. I find it a good deal easier to use, and it has a few nice features such as the facility to dump your model as a Python script; the Python API is simple and very well documented.
Yeah, I haven't given solvespace a try yet, but it looks very promising. Its workflow is just the same as the one I was using most of the time in autodesk Inventor/Fusion/Solidworks: 2D plan, place constraints like parallelism, angles, lengths, collinear, etc. extrude. Repeat, with maybe 3D constraints in some places (coplanar etc).
The only thing I'm unsure of is if it supports some functions such as milling, threads, etc. I think interfacing it with OpenSCAD just for this use-case would be huge (define the function in OpenSCAD, apply it in solvespace).
I tried hard to use FreeCAD, but maybe not hard enough, as I really had some difficulties using that workflow (IIRC, finding the 2D planner interface was difficult enough, and constraints didn't work as expected). It did have an OpenSCAD interface, though.
The last point is something worth mentioning: an opportunity FLOSS has that proprietary software often lacks is interoperability, either with the save formats, or directly with APIs, or calling each other for specific parts. I wish this was leveraged a bit more in this space.
I spent enough time in FreeCAD to become really proficient. It's not as refined as F360, it lacks a number of features, but the latest version (0.19) is fairly mature. I've designed and built full assemblies with gears, and then printed or CNC'd it (FreeCAD has some support for CNC toolpathing).
I've tried F360 a few times and each time dropped it thinking "this is designed to lock me in and make me pay for it".
Saw this come a mile away when they started jacking up the subscription cost for hobbyist licenses (~late 2019 according to my emails.) Instead of being an on-ramp for their serious offerings, it was just too good and needed to be hooked up to a milker.
These restrictions seem fairly reasonable to me, most cause for concern would be further future restrictions I suppose. No generative design at all is a bit of a shame though, people do play with that ('needlessly') for hobby 3D prints.
Odd that it doesn't mention small businesses though - wasn't it previously personal or <100k USD pa?
But also.. why? If it's already personal use only, and someone has an automatic tool changer for their personal use CNC mill, why not let them use it? Get hooked on it? Continue to want to use it when they have a saleable product idea?
> And I'm sure ATC can be added back in as a post-process, since it must at least stop for a manual tool change, or its CAM would be useless for even hobbyists
I believe ATC moves are implemented as macros by the machine. So the code emitted looks like:
and the ATC behavior is filled in with a macro for the specific machine.
To nerf ATC users, Fusion is preventing toolpaths with multiple tools from being exported in the first place. So you have to generate individual toolpaths for each tool.
The sad part is that hobby machines also make use of the "M6 T<tool number>" macros to move the machine to a place where it's easy to change tools and then do automatic touch-off before continuing. It's not only professional-class machines that will be affected.
It should be theoretically possible to run the multiple generated files through a script to put them back together, but this is likely too much complexity for the vast majority of users.
> Fusion is preventing toolpaths with multiple tools from being exported in the first place. So you have to generate individual toolpaths for each tool.
Oh that's a shame, I'd assumed it would use the 'compulsory stop' code, so no ATC, but allowing the opetator to. I hadn't considered that non-ATC machines might do something helpful with ATC codes. (I don't have, but have recently become interested in having, a CNC mill or anything.)
I actually test software in a Virtual Machine with no internet access before I use it on my main machine. I'm sure a dedicated developer could trip up my test by only doing an online check after two weeks or something, but it seems to work.
The difference is the deer and turkey don't get to read and accept the ToS. Autodesk seems clear about what you sign up for (1 year, non-commercial licence). You're choice to accept the terms and risk not being able to use this professional software for free.
Agree, You can't expect a company like Autodesk to offer such a expensive tool for free forever. They seem to be preventing abuse with these changes, while still allowing hobbyists to continue to use it for small projects. And that's all you should use it for. Building a busines on a hobbyist license or trusting all your complex big designs to their cloud is asking for trouble.
Oh, I mean I save the thing as an STL and then run a Python program on the commandline to do the conversion. I didn't even know that there was a provision for making that even easier.
It is unfortunate when I read the list of things they are doing to make the life of a hobbyist harder, because Fusion 360 has been pretty useful to me. They should just make it something you buy for $50. Would instantly pay.
FreeCAD is pretty cool and there are plenty of other great FOSS CAD/CAM tools worthy of mention. Is there something similar to the Blender Foundation geared towards CAD/CAM?
There are some great non-SaaS tools out there if you're willing to spend a little money. Rhino 3D is one stand-out in this space. And if you're looking for a decent CAM solution, check out Estlcam. It's a stand alone tool with a ton of great features at a reasonable price.
It's still significantly less expensive than OnShape, which seems like its closest competitor. Unfortunately it's hard to actually differentiate between "hobbyist" and "small company" without cutting some really critical features. Most people, even professionals, just don't use that much of the overall feature set.
Update from the FAQ: "You can still export DXFs, but only from a sketch. To do this, you’ll need to right click on the sketch in the browser and choose ‘Save as DXF. For more information, review How to Save Sketch as a DXF."
Fortunately for me, this is exactly what I'm writing up.
So the only hobby competitor (barely) is SketchUp which was acquired and is just being squeezed for profit while adding no new features, lacking MANY needed features and being quite buggy with terrible UX. So Auto desk pretty much has their hobby users in a corner.
There's only negative comments on the thread, but what most people missing here is software development takes resources. Heck, just keeping an online product alive costs money (check GCP/AWS pricing). A software engineer in the Bay Area costs $200k/year on an average. If you want to hire quality engineers, it takes a lot more.
They're not blocking it behind a paywall. It's still free. I understand that for hobbyists, it can be expensive. But if you are using things beyond the free tier, that's more than just a hobby. And I know this is an unpopular opinion, but being a Software Engineer, I really hate it when anytime a software company tries to charge for things, it's unfair and it's price gouging. But people fail to acknowledge that it is the only industry where you can even pick things up as a hobby for free. Literally anything else you want to pick up as a hobby costs money upfront. If all of the 645,000 users were willing to put up with those upfront costs, they wouldn't have had the need to price their plans so high.
I guess most hobbyist would be pretty content with 2006-era solidworks if it were readily available: look at how Eagle was used without updates for years, as a shareware.
I expect that most of the new features cater to professionnals, except for basic maintenance and a few simple features. The workload necessary to match hobbyist expectations is nowhere near a single full-time dev.
But these products are monstrously complicated for wat they are, in part due to the company looking for ways to extract value at every point it can (forcing cloud mechanisms: they cost money for sure, but are mostly used for locking customers in and as an excuse to require a permanent connection). And for every possible revenue stream, it seems...
I'd say this is a good thing if it makes FLOSS alternatives a bit more attractive. Hobbyists are usually hard to switch to alternatives once they find something "good enough". In turn, they tend to produce lots of documentation, tutorials, and training material. Just look at the number of people you can ask for help with, say, plotting a graph in Microsoft Excel vs doing it in LaTeX.
When reading through the comments, it's mentioned that one of the reasons for the license change is because of people abusing the free hobbyist license:
> [...] as talked about in the announcement, we did not make these decisions lightly, but it is necessary in order for us to continue support Fusion 360 for Personal Use as a free offering, cut down on abuse of the license type, and continue to develop advanced capabilities.
The issue of charging from the software is distinct from the issue of enforcing the requirement that the license is for non-commercial purposes. Even if they had a low cost hobbyist license with all the features unlocked, the people currently using the hobbyist license for commercial purposes would just use the inexpensive one as opposed to paying for the full commercial one.
Autodesk could've taken the approach of sending sternly worded letters to commercial users using the hobbyist license, or even deactivating commercial users with the wrong license type, since their software is partly cloud-based and requires an Autodesk account and internet connection. Instead, they've taken the approach of changing the feature set, which is why some people are annoyed.
Fine. It's just a hobby. You paint as a hobby, you buy paint, canvass and brushes. You read as a hobby you buy books. You bike as a hobby you buy a bike. You use CAD as a hobby you pay for CAD. Not everything is cheap, some hobbies are expensive, others are not. Buying books is cheap, buying a bike is not. I'm tired of people talking about free software as their right. People who make the software also have families to feed.
Then they should have charged for it from the very beginning. It's manipulative to use hobbyists as beta testers (while giving the impression of being benevolent and caring for the hobbyist and maker community) and then when the software is mature making it impossible for most hobbyists to financially buy into the software. This is not a one-time $500. This is a yearly or monthly subscription that is very expensive for most hobbyists who do not make money using the software. So while I agree that it doesn't need to be completely free, it does not make sense to charge a hobbyist the same amount as for the business user who's making money with the software.