In v8i I'm able to select these objects or snap to points along the yellow line THROUGH the glass box.
In CE it seems I'm not able to do so. What am I doing wrong?
PS. I have Disable Drag Operations checked in both v8i and CE, but I don't think this is related to the problem.
What setting have you got under;
File > Settings > User > Preferences > Input > Locate Interiors
In v8i the default was Never, and in connect it seems to be Visible
Thanks - I've now tried all options; Never/ Visible/ Always.
Never; will not select anything inside the glass box (no matter how far in I zoom)
Visible; First attempt at a tentative snap always shoots to a keypoint on the edge of the glass box, second attempt will select the object I want.
Always; same as Visible.
v8i quickly and accurately selects the geometry I want - something must be wrong with my CE settings, or is this standard behavior in CE. I really hope not.
By the way - it something wrong with the Communities site? Posts take several minutes to load.
Thanks for your sample file - the behavior is the same on yours as it is on mine (attached). It did cross my mind that my file (built in v8i) could be the problem, but it doesn't seem to be the case.
Are you able to Rotate View Dynamic > Define First Point > tentative on any of the cubes inside the glass box (either of our two files)?
Try also to select the yellow line in my file, by directly clicking on it. It works in v8i, but not in CE.
select through transparency.dgn
everything works fine:
Labyrinth Technology | dev.notes() | cad.point
Well then I'm stumped (and no closer to moving to CONNECT). Even if yours is working the way it's supposed to, then it's a step backwards from v8i (having to iterate through options, when you clearly intend to select the yellow line).
Thanks for your perseverance anyway, much appreciated.
Max Voigt said:then it's a step backwards from v8i
Because my tests are done in Locate Interiors: Visible mode, I think it works exactly as expected: When "something" is selected, an outer box will be always first and internal elements will be next in the list.
Max Voigt said:Well then I'm stumped
I think the problem is Locate Interiors: Never mode, because when used, it blocks the interior elements selection. It seems that some primitive solution was used by programmers: Any selection inside an element's face is canceled, instead of to be analyzed further whether there is (not) anything more in the selected area.
I think Service Ticket should be created and it has to be analyzed as bug.
Jan Šlegr said:Locate Interiors: Visible mode, I think it works exactly as expected:
Not really - the vertices ARE actually visible (through the transparent box). At least that's how v8i understood the terms.
Jan Šlegr said:I think the problem is Locate Interiors: Never mode, because when used, it blocks the interior elements selection.
So it's actually doing what it says it should - NEVER selecting elements inside another element.
With further testing, it seems to work better using the Transparent Display Style, although the terminology is no longer descriptive of what is going on. With Locate Interiors: Never selected, selecting geometry actually works accurately without scrolling through alternative snap points. With Visible and Always selected, it once again has to scroll through alternative snap points.
However, my point remains. v8i works much better than CE in Illustration mode. CE's Transparent model more closely resembles the behavior of v8i for my purposes (Locate Interiors: Never), but shows too much visible geometry in the view to be practical. CE simply does not work at all if I want to rotate a view around a selected point inside a transparent box (Illustration mode).
As a challenge, I compared the number of data points required to draw these red and blue lines from the top of their respective cones to meet various geometry inside the transparent box. The blue lines are to meet the vertices along the yellow line, and the red lines are to meet the hidden bottom corners of the five cubes.
Blue lines: v8i = 15 clicks, CE = 25 clicks
Red lines: v8i = 15 clicks, CE = 32 clicks (ie. CE is more than three times less efficient than v8i in this case)
Transparent mode makes it a bit easier, but is not the way I want to work with complex models.
2425.select through transparency.dgn