This is all great feedback, everyone. I'm glad to see that this many folks are interested enough in Fusion to comment.
I would like to discuss one of the early sub-threads: Sketch performance. The points raised here about sketch solve performance when the sketch is "large" (not everyone will agree about what defines a "large" sketch) are completely valid. We in the Fusion team agree that improving sketch performance is a high priority, and we will work hard to improve it. So, don't give up on us just yet!
If you're interested in some history, read on...
When Fusion first started focused on direct modeling, we first thought we could get away without a sketch solver at all! Because we were exploring the Direct Modeling paradigm (no timeline, no history), we thought that we could extend this paradigm even into sketching. AutoCAD went for 25+ years without a sketch solver. So, we built a sketch system that did not have constraints and dimensions in the way that other CAD products had them. We soon became convinced (when we tried to use it ourselves) that this route was not going anywhere good. So, we started over and started building a more traditional sketcher.
Most of us in the team have come from a parametric Mechanical CAD background. Most products in that market have separate 2D and 3D sketch environments. Some of that is historical, some is because of the solvers used. But, the idea of not being able to constrain 2D and 3D sketch geometry together, or to convert sketch geometry from one to the other, we saw as limiting. We also wanted to have a more wireframe modeling like experience that you might find in conceptual design and surfacing tools So, we made a somewhat risky decision: Because we wanted to someday deliver a unified 2D/3D sketch environment, we would leverage an internal 3D solver as our sketch solver (it's also the same solver that is used as our assembly/joint solver).
While this solver is proven as an assembly solver, it had never really been used in a sketch environment. The solver team put a lot of work into it over the past couple of years to make it handle geometry and constraints it had never had to deal with before. Here are some videos showing Fusion 360 solving as compared to other CAD tools:
Wedge Videos
The video below show Fusion drag behavior on a sketch with some lines and tangent circles
The video below shows Solidworks dragging the same sketch. Notice that the circles show some undesirable behavior
Arc Videos:
The video below show Fusion drag behavior on a sketch with tangent arcs
<
The video below shows Solidworks dragging the same sketch. Again, notice the behavior
We've been pleasantly surprised so far. The technology has held up well. However, as this thread points out, it is not yet perfect. The team is working hard every release to improve it. Hopefully you can see this improvement over the past several releases.
Now, what about that unified 2D/3D sketch environment? Unfortunately, we have not yet been able to completely realize it just yet. We have been focusing on getting the basics to be right, stable, and performant (that last one is the source of this thread). You can see hints of what we hope to accomplish in the fact that you can take a 2D sketch, use the Move command, and drag a sketch point or a line out of the sketch plane. It's all still the same sketch. This is not possible in some other MCAD parametric systems. Anyway, we are excited about the future in this area. We think we can come up with a pretty unique approach.
That's probably more than you all really cared to know, but hopefully some folks found it interesting.
Jeff Strater (Fusion development)
Jeff Strater
Engineering Director