I did not have the "auto project geometry on active sketch plane" option enabled. However enabling it does not change any behavior. In the example I shared it still kicks the selected face off screen. So I retain the believe that the logic being used is broken. I leave that project option off because I usually do not want all edges included, I want specific edges at most, and I appreciate that the option to turn it off exists.
What's even more irritating is that fusion is clearly aware of the way the user had the model oriented, because if you don't re-orient a sketch after it rotates to make X horizontal, clicking end sketch rotates you BACK to the view you had. It feels like the pieces should all be there to allow user preferences on how this option works.
My comment about wondering if developers use the software stems from years of seeing just annoying UI issues in the software that I cannot image how they get through testing. Eventually it starts to feel like either developers aren't using the tools, or they are using them in a very different way, even if that's an unfair assumption and statement. Some examples
1. this issue described in this thread.
2. For reasons I have not yet been able to categorize, when you have a timeline that is longer than your screen can display, if you drag the history marker back in the timeline, then edit a sketch/feature near the marker, often times when you finish working on that sketch/feature the timeline auto scrolls back to the end or beginning instead of leaving the marker on screen, requiring that you scroll back each time. To be clear I mean scrolling the timeline view, not scrolling the actual marker. The view is changing itself so that you no longer see where the marker is. It doesn't happen every time, and I've been unable to find the combination of things that recreates it reliably, but it's frequent enough to be productivity breaking.
3. Rotation centers for the model seem to not consider what is actually visible. If I have a large model, and I have a component isolated so it's the only visible component on screen, but that component is located significantly far enough away from global origin that the origin is off screen when zoomed to the component, the rotation center for the model still defaults to global origin which results in some wild rotations. This may be a spacemouse specific issue, it's been ages since I used middle click rotation. But it requires a right click, "reset orbit center" to get it to snap to a point on the visible component. And that must be done every time you exit a sketch because it resets to origin each time. This is annoying enough that I have a macro programmed to end a sketch, then right click and select that option so rotation always acts predictably.
4. The save popup (and other pop up menus) seem decoupled from any activity in the Data panel. If I go through the effort of creating a new folder inside a project to save a new design into, the save popup defaults into the location it had last used, requiring that the user navigate to that new folder again. This has caused me to "lose" designs in the data panel countless times because it just seems natural that there should be context associated with creating a new project/folder and then saving a new design. I am much better at catching it now, but it still seems like odd behavior.
5. I've had a bug for years now where the fusion sketch environment will randomly register extra clicks resulting in tiny extra line segments that often require extensive hunting to find. Meaning I will click once to start a line, click again to end it somewhere, and wind up with 2 lines and 3 points, one of which is extremely close to the first point. If I ever have a sketch that isn't registering as fully defined when it should be, I always have to go hunting for that extra unintended line segment. I have swapped out keyboards, mice, and even computers, and I've tried unplugging every single other USB device connected to the computer. I see the bug on multiple computers and my co-workers see it too. I even went so far as to make fusion the first piece of software I installed on a new laptop, using only the integrated track pad, and saw it happen on that machine. But the only response I've gotten when asking about it is blaming my hardware.
6. The construction toggle randomly locks in sketches and won't respond to the keyboard "x" shortcut, requiring that you click the toolbar button to "wake" it up again. This can be after a successful toggle from the keyboard. So I'll hit X to go to construction, then change my mind, hit X again and it won't toggle off.
7. Fusion electronics has been years of updates full of bugs that truly feel like the user base being treated as the actual beta testers for each release. From a bug that was causing actual files to be deleted that a co-worker and I found, diagnosed, and submitted several years ago, to toolbars that auto size themselves to take up so much of the screen that the available workspace is unusable, constant crashes, overall buggy behavior, if it wouldn't be an enormous time hit to migrate all our files to a different e-cad software package we'd have done it already, and probably will once we get a minute to breathe in the schedule.
There are more, but the cumulation of little annoyance bugs like this adds up to the end user feeling like updates just simply aren't being tested as much as they could be. Is it unfair to imply that no one on the team tests the software? Yes. I was frustrated, even if that isn't a good excuse. Is it unfair and counter-productive to complain in a wall of text on the forums? also probably. I do my best to only post when I think there is a real issue with something, I'm sure there is a lot of noise here from other users. So I'm sorry for adding to it, but I do feel the behavior described in the start of this thread is an actual issue that should be addressed.
