Bryce I think the core of this topic is how to cope with situations where potential inference points are so numerous and close together that snapping to the right one becomes difficult. Rather than zooming and orbiting to reduce the ambiguity, these users want to be able to turn off (or selectively enable) inference snapping.
Personally I think this is a “be careful what you ask for” situation. 3D modeling without inference snaps would lead to many more broken models than we see already created by people who ignore the inferences.
@Bryceosaurus
Hey, I know this is an old thread but I was unable to find any other place where to give feedback about interferencing in SU 2017. The down arrow is very cool but there are some cases which give me a hard time: eg. drawing a line that is perpendicular to a face, or drawing a line that is perpendicular to another line but is not the perpendicular SketchUp thinks I need.
It would be great (at least for 2018, if not in an update for 2017) to have these features:
-
Line perpendicular to a face: Constructing the end point of a line, when the down arrow is pressed while the cursor is On Face (blue mid-angle square), the interactive point would stick to a line that is perpendicular to the face and comes through the start point.
-
Plane perpendicular to an edge: Constructing the end point of a line, when the down arrow is pressed while the cursor is On Edge (red square), the interactive point would stick to a parallel line (this is already implemented). When the down arrow is pressed for the second time, SU chooses a perpendicular it thinks is the most important (this is also implemented). When the down arrow is pressed for the third time, the interactive point is locked on a plane that is perpendicular to the edge and comes through the start point.
.
I hope you understand what I mean by all this as I’m not a native speaker and I struggled a bit with the description 
The second feature is more complicated but indeed very useful. If you think three times the down arrow is too much, then replace the second press, because SU is not an oracle and often guesses a wrong perpendicular.
I think these features are very important and have a variety of use cases, without them it it very hard eg. to create a perpendicular transversal of two lines in general position.
One last thing: would it be possible that SU prefered sticking to vertices instead of sticking to axes? When I need to connect two vertices that are not parallel with an axis just by a tiny little bit, I need to zoom in really, really close to be able to make a line between them. When too zomed out, the end point sticks to the axis. This is even worse when the model is too complex and plane clipping occurs – this means I have to change models axes to make the line visibly non-parallel.
Later than last: I think what @brveldkamp and @patfan1222 mean by disabling interferencing is disabling sticking, so that you’d be able to lock onto something manually, but wouldn’t be forced to eg. stick to an axis or connect to a nearby vertex. This would however make modelling much more difficult… Maybe you could add options like “Stick to axes” into preferences, so that they could be selectively (or even through hotkeys) turned on and off.
or how about a snap override, no more time wasted hunting for midpoint or intersection would be awesome.
It’d a long time since I used AutoCad, but I seem to remember a pop up with check boxes, where you could choose what types of inference to use (check the box) or ignore (uncheck the box).
The SU equivalent could include end point, midpoint, intersection, on edge, on face, parallel to, perpendicular to, on axis, centre point, origin point, cpoint, guideline, and maybe a few more that exist but I rarely use or forget at the moment.
3 Likes