Move mouse in a specific direction (negative blue axis in my case).
Type the distance “0mm” and hit Enter.
As expected a guid point appeared at the location that was clicked in step 1.
I subsequently removed the pre-existing point from step 1, and can still see the new zero-distance guide point.
But this guid point can’t be interacted with in any way. I can’t snap to it. I can’t select it. I can’t delete it.
Is this a bug? In that case, where should I report it?
How can I get rid of these guide points? (Regardless if it’s a bug or not, I want to get rid of them).
What’s even odder is that I can find that guide point and add it to the selection using Ruby. That suggests it is indeed a glitch somewhere in the GUI.
Here’s a way to select the elusive zombie GP. @TIG’s Selection Hide/Show plugin has an Invert Selection command.
Select All > Invert Selection … and the unselectable Guide Point is selected.
I suspect it is to do with the lack of a tail. I think all native guide points are at the end of a tail aren’t they, only extensions create actual naked guide points.
At least I can’t think, off the top of my head, a native way to make a guide point without a tail.
Edit: No, find center will, hmm so I’m probably wrong.
I have the same problem I cut my model from around the points and copied to a new template, when everything was deleted from the model I had only 3 guide points and still totaled over 18000kb They are annoying little devils.
That’s an interesting speculation. There are details of guide points that are not exposed in the Ruby API, so there’s no useful way to investigate. For example, in the Ruby API if you add a ConstructionPoint (the API name for a guide point), you get just a bare point, no tail. In the Ruby API there is no such object as a guide point with a tail! If you want a tail you have to add a finite-length ConstructionLine with appropriate ends. Conversely, in the GUI aside from the trick demonstrated in this topic there is no way to create a guide point without a tail (ok, find center will also do it). Obviously the back-end has a more complex structure to handle both the guide point and the tail as a unit, but this structure is not fully exposed via the Ruby API. My suspicion is that when you cause the GUI to create a zero-length tail the back end gets confused about how to handle the (in its view) defective tail, and that blocks selection of the guide point.
It’s clearly a bug. The question, then, is whether the development team will see it as important enough to assign anyone to fix it.