Section viewport: inconsistent behaviour

[line-weight]

See the attached file. (VW2018)

 

There is a solid block, and three zero-height rectangular extrusions.

 

Each of those three extrusions is identical. They have a red dotted linetype.

 

The section viewport is set up to use objects' own attributes for the cut plane.

 

The three extrudes are rendered differently in the section viewport. One with a black elevation line in the background, the others without. Why? Is this another bug?

 

So far I've lost about 5 hours today trying to sort out this and an associated problem.

 

VWsect.vwx

[Matt Panzer]

The problem is that the geometry beyond is rendering and the line of the one extrude beyond is showing along with the cut line. While sections try to be smart about which lines beyond to hide, sometimes they show. This is usually not a problem because the cut is typically a thicker solid line than the line of objects beyond. What you have is a sort of 3D annotation object being sectioned as part of the model. While I completely understand why you want to do this, there are conflicts between how the model should be sectioned and how an annotation is drawn. For example: If the 3d sectioned "annotation" happened to be drawn on top of parts of the actual model (eg: top of a slab), you would want that line beyond drawn. Sections do not know that you want particular objects to be treated differently. What we really need is a built in way to have 3D grids so viewports will know what to do with them.

[line-weight]

Yes, that's exactly what I assumed it's doing: drawing the geometry beyond.

 

And you're right that there are situations where you do want to see the geometry beyond.

 

But what is frustrating is the inconsistent behaviour. Why does it only happen with one instance of the object?

 

I got this to work previously by using a 2d polygon with some of the edges turned 'off' so that there were no edges to see, and that was fine until I realised that the section viewports were drawing them in the wrong place, which is covered in another thread I posted earlier today.

[Matt Panzer]

31 minutes ago, line-weight said:

But what is frustrating is the inconsistent behaviour. Why does it only happen with one instance of the object?

 

Right. I certainly understand the frustration. I wish I had an answer of why it's not consistent. As we continue to improve our section viewports, we do want to eliminate these inconsistencies whenever possible.

 

40 minutes ago, line-weight said:

I got this to work previously by using a 2d polygon with some of the edges turned 'off' so that there were no edges to see, and that was fine until I realised that the section viewports were drawing them in the wrong place, which is covered in another thread I posted earlier today.

 

Yes. I saw that post and did investigate a bit. I'll add a comment to the thread.

[line-weight]

42 minutes ago, Matt Panzer said:

 

As we continue to improve our section viewports, we do want to eliminate these inconsistencies whenever possible.

 

Since moving into 3d properly, section viewports have become central and crucial to my workflow*. The various issues with them cause me quite a few headaches. They are one of those things which it's really important work well and are rock solid. Because if you build your model perfectly and then a section VP messes up when it renders it, it's impossible to correct those errors manually. It means time consuming and frustrating workarounds in the way the model is built.

 

*At present I use them for all 2d views of the model. That means not just sections but all elevations and floorplans too.

Edited February 20, 2019 by line-weight

[Matt Panzer]

Given the improvements added to vertical and horizontal sections in VW 2019, I hope it's evident that we feel the same. 🙂

[line-weight]

17 hours ago, Matt Panzer said:

Given the improvements added to vertical and horizontal sections in VW 2019, I hope it's evident that we feel the same. 🙂

 

That's certainly good to hear for the future, but it would be even more evident if multiple SVP bugs were not allowed to remain in the final release of VW2018.

[Matt Panzer]

34 minutes ago, line-weight said:

 

That's certainly good to hear for the future, but it would be even more evident if multiple SVP bugs were not allowed to remain in the final release of VW2018.

 

I hear you. My point is that we have been, and will continue, improving section viewports.

The bug you state in this thread would be considered minor (IMO) for the reasons I stated earlier. The other bug you mentioned, however, is more serious and I stated as such in the report I submitted.

[line-weight]

3 hours ago, Matt Panzer said:

 

I hear you. My point is that we have been, and will continue, improving section viewports.

The bug you state in this thread would be considered minor (IMO) for the reasons I stated earlier. The other bug you mentioned, however, is more serious and I stated as such in the report I submitted.

 

Sure. Thanks, and I appreciate that.

 

For the record there's also this one:

 

 

And this one (labelled as resolved, but it's not resolved in 2018):

 

 

I agree that one or two of these, on their own, can be considered 'minor'. The cumulative effect of them all, though, isn't. And that's before I even get started on what happens in design layer section viewports, which are borderline unusable.

 

Edited February 21, 2019 by line-weight