Fusion360 still crashes too much OS X

[cekuhnen]

The past days I gave fusion a lot of tests and encountered many crashes. The software is still extremely buggy and prone to say good by even with very simple files.

[roambotics_scott]

Yeah there is definitely some amazing potential there and the constraints are huge though there are lots of little things getting in the way of fullfilling that potential.

I imagine / suspect that it'll be insanely great if they keep pushing on it for a few years (which is part of what makes it so frustrating now; there are so many great and promising things but there are also so many gaps and show-stoppers).

[cekuhnen]

My fear is that surface modelers and engineer solid modelers simply remain what they are because also complete industries grew around them.

take a look at Word it is one of the worst to use - but everybody uses it because of the peer pressure.

I find it very interesting to explore how in Inventor or SW one would work but always shake my head at all the prepare steps they have to do before they get into designing compared to Alias.

[roambotics_scott]

Not only does everybody use Word but almost everybody (even Apple to some extent) blindly copies it.

Instead it'd be better to spend a long time studying the options, steal the best parts from each, throw away or fix the worst, and make something new and great.

Also (this discussion and Ideastation, aside) most users should be generally ignored becasue when you actually do build something new and dramatically better (like Ribbons in Word), a lot of people will complain about it simply because it's different from what they're used to without regard to how easy it is to learn for a new user or how much more efficient it makes your workflow once you «get it».

Actually (done well) I think that's part of the promise of this product (and it's part of why I kinda think Autodesk should push it out of the nest and spin it off into its own start-up with only enough cash reserves for a ~9-12 month runway).

[cekuhnen]

I agree with most while I think Pages is actually fantastic at what it should do - process text.

I was hoping Fusion would turn into a hybrid but it seems now more like a Inventor light with a design timeline and great a TS module.

Marketing wise it somewhat makes sense - while sad.

[rpagewood]

Has anyone had experience with Siemens NX 9 on Mac? That runs natively on mac and is a really high end cad app with powerful tools. It's really and end To end solution for design. It a also has really nice solutions for sub-d modelling and direct editing.

It would be interesting to know how stable this was.

[roambotics_scott]

Just to be clear - I think Pages (now) is very much there and its own thing (I was thinking more of AppleWorks there).

I do feel like marketing has really been steering the product.. the saddest part about that being that if it actually did what (I think) it set out to do (and still has the potential to become) then it'd almost not even need marketing because people like me would be happily sharing it with everyone who asks.

Funny you should mention NX 9 - our intern was supposed to get the trial version to do some physical simulation. I think he ended up just using Blender's rigid body dynamics instead but I still want to try it (definitely interested in hearing anyone's experiences with it).

[Oceanconcepts]

I have Cobalt as well, and there are some things I like a lot- the ability to draw lines in 2D or 3D space without entering a different “mode” pretty much replaces what sketching does in Fusion, and I think is a more intuitive and natural way to create- but that may be just my background. You can be drawing lines and toggle easily back an forth between constraining lines to a work plane, or drawing them in 3D space. They are just another sort of object, and you can edit them at any point without entering a special “edit sketch” mode. The ability to organize them in user nameable, function related layers and toggle them on and off goes a long way towards reducing the browser clutter that is a problem in Fusion.  It has snaps and alignment tools (with hints!) that are for more extensive than what is currently available in Fusion, and has great information and editing options for objects that you select in the drawing, always present. I still use it for creating 2D dimensioned drawings without starting Windows. It is, unfortunately, a lot more expensive than Fusion- now about $2500. 

However, I do find myself doing boolean operations with closely aligned bodies, or just deleting faces to heal bodies in Fusion, and I still have an involuntary cringe when I do, as I know those kind of operations would have crashed Cobalt. The problem is that Cobalt has not been really updated in many years.  The underlying kernel in Fusion feels a lot more robust to me- this is as opposed to the top layer stuff that seems to be causing the crashes. If the stability can be fixed, Fusion has the potential to be an amazingly capable program. 

i do hope the Fusion designers have a look at the alternatives to the sketching modality, long term. It does feel like a throwback to working from blueprints. 3D sketching in Fusion is so hidden in the interface (having to go to preferences to enable it?) and limited that I have very rarely used it.

I think the Word comparison is spot on- these are interface methods that are  there because users of dominant packages expect them, not because they make the most sense for a free form, creative  program.  And you can fail by asking people to take too big a leap, so perhaps it makes sense. Then there are people like me who open Word only a few times a year, and have found simpler alternatives.

[cekuhnen]

I feel for 2D drawing the sketch engine is ok - because it also functions are mainly all other solid modelers. But in all other solid modelers I feel the 3D sketch engine is literally a joke. It feels forced. Cobalt since the beginning had fantastic curve drawing and snapping tools. It took Rhino many years to get close to that.

And the lack of snapping in many modes in Fusion is just a major show stop to me - but I know they know that and are working in a solution.

But I agree also with you. I would throw out the idea to make the 3D sketch like in all other solid modelers and for designers sake give us a tool we work with that is truly a 3D sketch engine. It cannot be so hard when it is so common.

Regarding Pages it was an interesting observation. Sure Apple radically but back features when they made the web iOS and Mac version the same - bot interestingly many said afterwards that this is finally an app hat is just great at writing text.

In one of my recent furniture projects with a new client we got the CNC DWG data send back because they said their CNC cannot deal with splines and needs poly lines and I found it odd after google what they need that there are methods that are so old they dont change because also the hardware was expensive to buy in the first place.

Thus I can understand that when engineering 2d sketching is just done this way because everybody does it that Fusion follows it - just wish like with Pages maybe it could be further improved or extended into doing both.

[cekuhnen]

I am also not sure what Fusion should really be. Truly it should not be an Inventor killer or competitor but should it remain just a light CAD app?

I just googled NX 9 OS X which is not native, it runs in X Windows but I was surprised to see that NX 9 as well as also now ProE all start to have a freeform tool similar to T-Splines build in.

What type of physical simulation did you try to do in NX 9 your intern was maybe able to do in Blender? Just curious.

[roambotics_scott]

I know what I want it to be (at least roughly) - a unified place where you can do everything from concept drawings to surfacing, physical layout, mechtronic simulation, all the way through production (ideally with usable versions, branching, and documentation).

I don't want it to be just an Inventor killer but also an Alias killer and an Eagle CAD killer and bunch of other things (probably not as feature-rich in any one area but something that startups could reasonably use to go from concept all the way through scaling).

He was just working on some simple rigid body mechanics and some numerical modeling (both of which it turns out Blender is passable with for making decent movies with some work but it takes a ton of work to get any sort of usable simulations out of it).

[cekuhnen]

I am not sure if that wish will come to reality. I think Autodesk keeps Fusion below Inventor for a reason. That's also ok. I as a designer do not want to have to use a software that offers a toolset I might not use at all - meaning a tool set that is too much geared towards what an engineer would do. But I like the idea to work on a design and passing it through in a way so that the CAD data has not to be rebuild by the engineer. From what I have seen that is one of the reasons SolidWorks got so popular.

It would be great if the work done in Fusion can be passed on and utilized by other systems. Inventor and Alias had/have such a connection where Alias designs can be linked into Inventor.

Fusion is very often advertized as design differently and I cannot see it being differently besides the design time line and the cloud sharing. I think inter connections and more better surfacing and sketching tools could really help to design better as I feel each method has their strengths but also weaknesses.
As great as the 2d sketch engine is for drawing basic lines and acrs it falls short compared to how I can draw in Rhino with all the line tools.

So I would like to see not just the Inventor sketch tools but just a mix of that with more common surface modeler curve tools.

Regarding Blender simulations in general I think take a lot of fine tuning work. I use the software myself a lot and also teach it along Fusion and Alias. They all complement each other well - specifically the poly modeling and render abilities. Fusion is really fantastic to either turn poly models in TS for organic shapes or flat Breps when the design is more linear. Because of a lack of usable snapping tools furniture design would be a pain in Fusion right now for me.

[roambotics_scott]

I'm sure they don't want it to becasue as awesome as it'd be, it'd cannibalize Alias and Inventor (at least). That said though - I think it's really what Fusion 360 /wants/ to be. It's one place where you can do surface modeling and detailed design and jump back and forth without having to constantly push things back and forth between two environments (and two distinct UIs). When you add in the documentation, CAM, and rendering, it's increasingly feeling to me like a one-stop shop for hardware startups.

That's another reason why I think F360 should be spun-off of Autodesk. As long as it's part of the company, they (paradoxically) have too many resources and aren't hungry enough to really fix things (because they can essentially just go on forever without having to worry about running out of runway), and on the other side, they have pressure not to get too good because if they do then they'll eat into two big revenue streams.

I don't think it should be geared toward engineers or designers but rather both (in particular, people who can and would like to at least dabble in both), and even in bigger organizations where the roles are divorced, there's a lot to be said for having a single common environment (assuming that it can be done well without too much compromise).

Interconnections and an API to make it a platform would be great (and I think that'd also be rapidly more evolved as a spin-off fwiw because they would need to form partnerships and wouldn't have the resources to do everything in-house).

On the Blender simulations - I definitely wouldn't want to do anything serious there. I can see how great those are for proceedural animation but it's annoying to know how simple it'd be (in principle) to build real rigid body dynamics with force fields and actuators into F360, and for the kind of stuff we're working on that'd be huge.

[cekuhnen]

I agree that an app that combines designer and engineer tool sets would be
great. It feels that SW is still an engineering tool. ProE the same. Some
of my students who work for Rubbermaid now have to leave rhino and only
work in ProE and I see that being problematic.

If the design timeline remains but the add base feature could also include
surfacing workflows like alias rhino it would be a blessing for the product.

Currently the direct modeling has no design history making it as useful as
rhino for concept refinements. Sketches have no relationship to surfaces.
With a real 3d sketch tool I think the DM area would be perfect if they
also make the features then editable. Fillet you can adjust loft you cannot.

[rpagewood]

Regardless of what F360 is or isn't, the reason I like using it is because it's fast and intuitive. That being said, I just wish it was more stable on my mac.

P.s. I do think NX is available as a native mac program, it's just extremely hard to find the system reqs on the Siemens website.

[cekuhnen]

I saw screenshots of it and those were in the X Window system on OS X.
http://architosh.com/wp-content/uploads/2013/10/nx9-full-screen.png

[cekuhnen]

Because we talked few times in this thread about sketch engine I just saw a screen recording from NX 9 and they seem to use all three methods solid modeling with 2d sketch planes sub-d and they have normal 3d splines and surface tools. That would be so ideal for Fusion.

[roambotics_scott]

Yeah the stability is really the killer. It's an amazingly powerful program but the only way I've managed to get any serious work done is through monumental force of will and I'd expect that no sane or reasonable person would do the same.

[cekuhnen]

or allowing you to work offline 

[roambotics_scott]

Oh yeah - I completely forgot they had scheduled downtime today.. it does kinda seem like it should be within their power to do zero-downtime deploys and to have both development and production back-ends.

Dependence on the cloud really is its Achilles’ heel - and the sad thing about that is that as nice as the cloud sync and backup are 99.9% of the time I’ve been using it as a single user.

[cekuhnen]

one of my clients engineers is interested in Fusion because of what it can do and collaborate on the same file online but the company owner questioned out of fear the internet idea. And it seems that he has a point. If this would be during a work day this would be a no go.