It’s so user hostile and constantly getting in the way of getting work done. Years ago (2001 ish) when I first started to learn CAD I used PTC’s Pro/Desktop. It’s a simple package and not without its faults but when you consider the pace of technological development since then it’s appalling how much CAD has stepped backwards. Everything is clunky, cloud strangled, ribbon interface befuddled and runs like molasses.
There's a learning curve, of course, for any CAD system. It takes WEEKS of regular usage to get reasonably fluent and even then you'll get snagged with a "how do I?" problem regularly. The hardest part is unlearning bad habits. There's easy ways and hard ways to do things. Sometimes the hard way is initially more intuitive, but ends up costing effort and time.
It's helpful, I think, to watch pros use it and see what they do. That's another good feature. It has a decent user community and there's lots of help available.
Autodesk's present scheme seems to be to make as much of their software cloud-connected, with the purpose of entrapping all your company's vital IP within.
I definitely couldn't replace Solidworks or Fusion360 with FreeCAD in terms of speed of use, especially if I was sitting down with all 3 for the first time.
I think what you are really comparing at this point is FreeCAD, the application, to Autodesk, the brand.
With that in mind, yes, FreeCAD doesn't have those features, but they exist in other open source applications.
I really wish that was true. It would make my life much better.
I think what you are really comparing at this point is FreeCAD, the application, to Autodesk, the brand.
I'm intentionally not. Because if I where I would also be talking about Navis, Infraworks, Revit, 3DS and so on.
And just to be clear I greatly dislike AutoCAD and if I never had to use it again it would no doubt improve my quality of life.
The problem is that everybody uses a different 80%.
"a good number of features that AutoCAD is missing"
Everyone needs a different bunch of features, and for some of them that would exclude AutoCAD.
AutoCAD was basically a UI for a pen plotter.
One notable exception was Duff Kurland. He liked hanging out with us users and would chat us up. He'd get our bugs fixed. We loved him.
Generic CADD was pretty good, much loved. So Autodesk had to kill it, natch.
Only Bentley's MicroStation was more user hostile.
Later, an intern at Bentley created an amazing auto snapping & aligning input tool. Especially for isometric drawings. I don't know came of it.
User interface researchers like Suresh Bhanavi did some really good work. He was hired by Army Corp of Engineers to determine why CAD work had poor quality and ROI (often negative), compared to just using drafting boards. He tried to advocate his findings. The CAD vendors weren't interested.
I doubt Autodesk or Bentley ever addressed their distain for users.
Gods, writing this now, I'm angrily remembering helping coworkers fix their SolidWorks and 3DS Max problems.
A few startups were more user and domain focused. ArchiCADD and Rhino come to mind.
Source: Was a CAD jockey in the 80s/90s. My "chaotic mutant user interface" critique (allegedly) got wide circulation within Autodesk. Briefly served on Bentley's steering committee; waste of time.
PS: Just looked at Autodesk's current product list. Wow. They're the Computer Associates / SoftBank of design software.
It's ridiculously powerful, and the cost is about 1 cup of coffee per day. I'm running it on an $800 laptop that's a year old, and it's snappy. My biggest complaint is lack of Linux support, which is abominable since Linux is the most widely used OS on the planet, but what can you do?
The CAM works, the simulations work, the assemblies work. It’s Cheap for what it is. I don’t even touch the sculpt, electronics, or sheet metal workspaces and those are included!
My personal bugbear is that the rotary 4th axis cam is so expensive.
Linux is not the most widely used desktop OS - that is windows by a long shot and that is the segment that matters here. Why would a software maker invest the time and effort in catering to a negligible slice of the desktop OS market?
> Why would a software maker invest the time and effort in catering to a negligible slice of the desktop OS market?
Why indeed? I disagree with "negligible" here, but for any sort of advanced and automated workflow, Linux is the best choice.
I've tried Fusion 360, and I agree, it is just such a pain compared to SolidWorks which I use for work. I already have a paradigm for design in my head due to how long I've used CAD and Fusion 360 doesn't match that. I'm somewhat surprised that Solid Edge isn't more popular in the maker space, it is similarly free for hobbyists and is legitimately CAD software that professionals use.
CAD is bad because there are too few users.
Work backward from the Altium statements. Altium has about $200 million in revenue. Altium website claims licenses are about $2K-ish. That's a maximum of about 100K paying users (reality is likely closer to 25K).
If you have 1,000 features, you have 100 users per feature.
In reality, those features follow a power law. Everybody uses the same basic features, but if you use a slightly off-path feature, you may be the only user.
Open source is the only path forward in this space because code can accumulate over time and users can consolidate over time.
The problem is that CAD coders are even rarer than CAD users, and they can get get paid quite a lot in their day job--why should they do this for free?
CAD is bad because it is made by SW developers who implement requirements and have no idea how the SW is used. I do not need all the toolbars at once. If the buttons are too small they are difficult to press, if they are too big they eat valuable space.
> If you have 1,000 features, you have 100 users per feature.
:) Most of the features are used sporadically.
> In reality, those features follow a power law. Everybody uses the same basic features, but if you use a slightly off-path feature, you may be the only user.
No. There is a workflow. During this workflow you may, or may not need special features depending of the object being designed.
> Open source is the only path forward in this space because code can accumulate over time and users can consolidate over time.
There are some very good open source tools (like Magic layout system) but they are good because the developers also use those tools.
> The problem is that CAD coders are even rarer than CAD users, and they can get get paid quite a lot in their day job--why should they do this for free?
AFAIK there are no "CAD coders". There are SW developers who got the job to write the code. Finding people qualified in multiple domains is hard.
Sorry, the Magic layout system is NOT a good tool.
It has a lousy DRC/extraction system based on corner stitching which requires that everything be lambda-based rectangles. Ever seen a professional VLSI layout person use rectangles?--they almost never do. There's a good reason why they use paths.
Good layout systems also link the schematic and layout in real-time. A lot of PCB systems even do this now.
I can go on and on and on about Magic. None of it is good.
That said, there is a makers version of SolidWorks coming to the market this year.
But it will eventually be good enough. For some things it already is.
KiCAD gave me hope. It is possible to have professional grade open source CAD software.
But it's definitely behind the curve in terms of how friendly and discoverable the user interface is. It's a powerful modeler, but I'm not sure that it's in a good state to be used by a large engineering team on multiple large projects. In particular, a lot of the PDM options for FreeCAD seem sort of questionable; for better or worse, a lot of organizations are going to need it to interface with their current data management software. I haven't seen the effort to do that yet.
It's an incredibly intuitive and well designed 3D CAD design tool for beginner to intermediate complexity 3D printer projects.
For more complex design I'll use blender or onshape. I've tried Fusion and just find it too convoluted!