Issue, Hide Selection in 3D

[thesleepofreason]

The new Hide Selection in 3D function is view specific - which is great! So you use the function in a view to hide an element, redefine the view with current window setting and the element/s stay hidden in that view.

However, there seem to be a rather counter-intuitive and critical issue with the current implementation.

If you create a new element it will not show in any view that has hidden elements in it. 

So rather than hiding the selected elements in the view the new function seem to limit the view to only show the inverse of the selection/s. The difference is crucial.

If I'm not mistaken then this could compromise workflows using 3D views with regards to QA. Because in order to rely on a view one have to constantly redefine (show all, redo any eventual hiding, redefine) due to the possibility that the view has, perhaps unintentionally, been defined with a hidden element. If this is not done, then there's a risk with elements created after the view was defined don't show in the view although the haven't been actively hidden. The fact that there is no way to tell if a view contain hide selection operations makes it worse.

The current implementation also means that a lot of the function's potential is lost. Say that you use the function to quickly set up a view that hides any element that is irrelevant to what you want to model/show. The current implementation means that you are limited to modelling in the specific view and that you, unprompted, need to remember to redefine the view before changing the 3D window if you want any new elements to be shown in the view. If you have to use a different view, like 2d plan to place a new element, or forget to redefine before changing the 3D window then you have to redo the hiding. 

I think that this shows that there is a real and urgent need to rework how Element Visibility Control work in AC. 

[Minh Nguyen]

Hi Barry,

Thank you very much for the detail explanation! That's totally understandable!

We consulted with the development, and indeed, the Hide Selection function is more or less the same as Show Selection command, but it works straight in 3D. Therefore, we would expect the same functionality between both:

- In case of Show Selection + Save View, the view will store the selected element ids. When view is opened only the stored elements will be shown.
- In case of Hide Selection + Save View, the visible elements of the 3D windows are stored but the selected elements are not stored. So when view is opened only the elements visible when storing will be shown. 

There is a text about this on View Settings/3D only page: 3D view limited by stored selection

I hope this is understandable, albeit it's a bit counterintuitive. Please let me know if you have any questions or comments. Thank you very much!

Best regards,

Minh

[thesleepofreason]

I appreciate the response although I must say that it's quite disheartening that the implementation works as intended.

There is a real and crucial difference between actively showing and actively hiding. Given the implementation it's nothing more than a shortcut for Ctrl+A > deselect element > F5. To call this 'hide' is counter intuitive to the degree of confusion. To call it a new feature is deceitful.

For a user ignorant of the inner technicalities it's hard to see why we don't have an expression based visibility control where hidden elements are handled as an exception to the list of shown elements. 

[gdford]

Dang  - I just painfully figured this out - MyLord - what a mess