What is a troll bug? It’s a bug that taunts you and makes you repeat the same thing over and over until you want to rage quit. The best way to deal with them is to remove them from the source instead of feeding them with your precious keystrokes and mouse clicks. Lucky for you, we got rid off a bunch of them this time around, so you can channel your energy into more productive and harmonious tasks, like designing sweet products in Fusion 360 that do good for the world. Here are the notable fixes and improvements we made for this update.
Joint and joint origin now stay intact after edits
TrippyLighting reported a joint-related bug where editing a joint and joint origin back and forth resulted in some really poltergeist-ish behavior with the joint origin jumping around for no reason. We called the software ghostbusters and were able to catch em all so that joints work as they should again.
Component issues fixed
We fixed a weird component naming issue in the browser where the first time you double click a component, it didn’t highlight the text. Pressing CTRL+A to select all didn’t work either, and pressing delete with the intention to delete the name actually deleted the component itself. What a hot mess. We’re glad that this is sorted out.
Bold.stelvis reported an issue where he was seeing x-ref’d designs nested in other components becoming out-of-date, even though no changes have been made to them. Turns out the out-of-date messaging was falsely triggered by some wonky code. We came up with the fix in November, but spent some time testing it to make sure that it works. Now it’s fixed 🙂
Mesh to Brep conversion improved
Since the introduction of the mesh workspace, we’ve been receiving a number of crash reports related converting mesh bodies to brep bodies. We’ve been incrementally improving the conversion performance, and this update should make it a lot better. Let us know in the forum if this is working better for you.
Also related – previously it would take about 10+ minutes to file > export a design with very large mesh bodies as SPD file format. Now it takes about 4 minutes.
The trackpad scrolling troll bug is gone
Cekuhnen reported that if you were scrolling through your browser with your trackpad and let go of it to select the item you were looking for, the browser would jump up just enough for you to select the wrong item. At first you’d think that it was your own fault, but no. It’s the trackpad. Trolling you. The more you did it, the more you were feeding it. Don’t worry, we took care of it.
Proxy server improvements
We’ve made a bunch of proxy server improvements, such as automatic proxy detection for Mac users so you can move between an office proxy to a home proxy without any issues, as well as
better install performance for corporate environments with PAC scripts enabled.
Python upgraded to 3.5.1
Previously we were running Python version 3.3.5, but now we’ve upgraded to 3.5.1. Our code jockeys believe that it will reduce the number of SSL related issues during the installation process. We are actively tracking this and seeing if it does, so if you run into install issues, please let us know.
Rendering & Graphics
STLs are no longer black when rendered locally
So remember that bug where your STL bodies turned into pitch black as soon as you kicked off the local rendering engine to do its thing? Yeah, that’s fixed. Now all STLs will adopt your default appearance material (Steel Satin, or whatever you’ve set it as in your preferences). Hooray!
Model disappearing from workspace issue fixed
We got a design where when you tried to orbit around it, it would literally disappear from the workspace. Poof. Gone. No errors, no way of getting it back. Even the viewcube went away with it. We later learned that this was due to a graphics corruption issue and got it fixed. Spooky.
View scale drop-down menu now stays open
We fixed a wonky UI bug where if you tried to expand the the View Properties dialog and wanted to select a different base view scale, the drop-down list wouldn’t stay expanded. It would drop down for a second and then collapse back up, taunting you, testing your ninja reflexes. This isn’t a dojo. Now when you expand the list, it’ll stay expanded like it should.
Open slots now works better
Good news! We fixed an issue where recalculating an open slot toolpath resulted in an an error where the toolpath did not fit the open slot anymore. The fix in HSM means that cutting open slots in Fusion should now work a lot smoother with much better performance.
HSM Kernel updated
We’ve also updated our HSM kernel, which fixes a number of outstanding issues we couldn’t get to in previous product updates. It also fixed a series of nasty turning issues related to gouging parts due to invalid toolpaths, checking for gouges, and machine tool crash scenarios.
The updated kernel also fixed a tool path trimming issue for 2D Contour, where the incorrect trimming occurs when stock contours were defined.
Apparently there was a ramp taper angle setting threshold that causes the cutting Z level to not go to the correct final depth, which resulted in a tool crash. This is now also fixed.
Data no longer disappears
jonathanZRMZM reported a glaring issue where moving ANY operation directly after a folder or pattern causes the folder to empty, leaving its children attached to the operation. If the operation is regenerated, the toolpath for the operation is removed, a bunch of important data gets lost. What’s worse is that the issue won’t throw any errors until it’s too late. Ouch. We quickly jumped on this and got it fixed.
Swarf bugs fixed
We caught a Swarf bug where the direction of Face chain couldn’t be changed. Oh and it gets worse. When a swarf operation chooses the wrong side of a part, you have to check “machine other side” and the cut will go from conventional milling to climb milling. There was no option to specify which you want to use, which was annoyingly frustrating. Now you can change direction and machine the other side.
Fusion 360 freeze issues fixed
This issue was found on both El Capitan 10.11, 10.12 Sierra, and Windows 7. Regenerating an adaptive operation, then clicking “Yes” to confirm the regeneration, but then right-clicking the operation and then aborting it will make Fusion 360 get all frazzled and crash.
When testing on a Surface Pro 4, we ran into a freeze issue when trying to move and edit a tool in the local folder of the Tool Library dialog using the touch screen. We dug into the code and saw some mutated code that apparently came in contact with the green ooze. Luckily we had an antigen to reverse the effect. Now it’s fixed.
UI wonkiness fixed
We found a weird Mac OS UI issue where the Tool Renumbering dialog appeared to be unnecessarily enormous for only 4 entry fields and a couple of checkboxes. Now it’s back to a more logical size.
Metals with nonlinear behavior added to Materials Library
When you conduct a nonlinear study, you would typically use metal materials that have inherent nonlinear material behaviors. Problem was that we didn’t have any of these materials in the Materials Library in Fusion 360. Now the update includes new metals with nonlinear behavior. You can access them via the physical material command. Under the library drop-down menu, select Fusion 360 Nonlinear Material Library and you’ll get an assortment of metals to choose from.
This should be a solid foundation for doing nonlinear analysis. We have plans to expand this library to include other materials types such as rubber in the near future.
UI weirdness fixed
We found some conflicting UI behavior where if the name of your study was very long and you went into the Compare environment to compare results, the study name would be cut off when it reaches the Navigation tools. As a compromise, we shortened the allowed space so that it doesn’t clash with the Nav tools, and if you name is still cut off, hovering over it will show a tooltip with the full name displayed.
Compare option no longer disappears
Speaking of the Compare environment, we fixed a bug where the Compare option would disappear from the workspace switcher after you closed and reopened a design with results are already solved. The Compare option is now available as soon as you open a design with solved results, ready to be compared.
There were cases where if you went into your preferences and changed your Load Glyph Graphics from Screen Size to Model Size, continued to exit out of the design and then opened it back up, a giant load arrow would flash for a second, followed by an empty canvas and an invisible viewcube. These troll bugs really know how to mess things up. Luckily we got this sorted out as well.
Another weird UI bug had to do with the materials dialog growing bigger and bigger every time you open it. If you do it repeatedly, the dialog would take over your whole window. Another case of some voodoo magic going on. Luckily we broke the spell and now our UI works as it should.
Apparently the Event Simulation Contact Manager forgot to include the instance number for your components. If you haven’t renamed your components (which you really should be doing), it seemed like you only had one component to work with. Now the contacts manager will correctly show the number of instances.
Probe bugs squashed
There have been a number of Surface Probe issues reported in the previous version of Fusion 360. Issues like when you go into Surface Probe, change Displacement result type to something else in the legend bar, and then move your mouse, the result type would switch back to Displacement for no reason.
And then there’s this annoying bug. Say that you ran a study and the results show some deformation of your model. The undeformed state is shown as wireframe. When you tried to put a surface probe on an area of the model where the undeformed and deformed state were very close, Fusion 360 would think that you want to put it on the deformed one (which is wrong), and will plop it on the wrong state. We’ve taught the software to play nicely with the two states so now you can place your surface probes correctly.
Web viewer naming issue fixed
Apparently troll bugs weren’t just messing with us in-product. They managed to get into the web browser as well. In the Design details of the browser model view, one troll bug managed to take one of your study names and replaced all of your folder names under that study with the same exact name. We decided to take the high road and quietly put an end to this nonsense. Hopefully this didn’t get too many of you.
X-ref modeling performance back to 100%
There were some cases where running a Simulation on a model that included x-ref designs made the modeling performance worse. Things got slower over time, which shouldn’t happen. We untangled some knots in the backend and now modeling performance is back up to speed.
Suppress/Unsuppress issues fixed
We fixed a couple of suppress/unsuppress body issues related to Simulation. One was an instance where suppressed bodies were somehow also included in Event Simulation’s Contacts Manager. We called the Event Simulation bouncer and tossed them outta there.
The other one was the suppress/unsuppress command not actually working for x-ref designs that were sub-assemblies in a component. Now it does.
New FindBrepUsingRay and FindBRepUsingPoint methods
We’ve added FindBrepUsingRay and FindBRepUsingPoint methods on the Component class, allowing you to find B-Rep entities in a model. These can come in real handy for certain workflows.
New viewToModelSpace and modelToViewSpace methods
We’ve also added a couple of new methods to the Viewport class to let you go between the 2D space of the view and the 3D space of the model. I know most of you haven’t been losing sleep over this but if you have needed it, this will be a welcome addition.
- Previously when the image command input was activated in the the Table command input, InputChanged event didn’t fire. Now it does.
- Nnikbin reported an issue where Matrix3D.translation doesn’t actually do what it said it would do in the documentation, which was super confusing. Another bug bites the dust.
- Niklaskp found that when he tried to set the distance value in his Sweep API, it always turned to 0, which made the command to fail. Our developers found that creating sweeps with API have generally been just downright buggy. We dug into it and beefed up the code so now it should work a lot better.
- Burzins noticed an issue where API calls for non-uniformly scaled Cylinders and Cones were written incorrectly as regular Cylinder and Cone, instead of EllipticalCylinder and EllipticalCone. This is now fixed!
- We found an issue while working with one of you where API used to automate timeline functionality failed when a feature compute failed. The timeline API should have succeed regardless whether there features failed or not. Now it does.
For all you API junkies, look forward to a lot more goodies coming in the January update.
That wraps up this update.
Thank you for your continued support! As always, let us know your thoughts in the comments section below or the forum. Onward and upward!
Keqing and the Fusion 360 Team