Extrude manifold creates non manifold geometry #80440

Closed
opened 2020-09-03 20:11:19 +02:00 by Ludvik Koutny · 35 comments
Contributor

System Information
Operating system: Windows-10-10.0.19041-SP0 64 Bits
Graphics card: GeForce GTX 1080 Ti/PCIe/SSE2 NVIDIA Corporation 4.5.0 NVIDIA 451.77

Blender Version
Broken: version: 2.90.0, branch: master (modified), commit date: 2020-08-31 11:26, hash: 0330d1af29
Worked: Probably never

Short description of error
I have not seen this in a list of known limitations therefore I am reporting this as a bug:
2020-09-03_20-04-43.mp4
Extrude manifold tends to produce non manifold geometry if extrusion crosses an edge loop, which is in conflict of what the name of the tool suggest. One vertex connected to both closed and open edges in this way is not considered manifold geometry and also goes against common sense expectation of what kind of result this tool should provide.

Resulting geometry qualifies as non-manifold even according to Blender's own operator to select non-manifold geometry:
image.png

Exact steps for others to reproduce the error

  1. Create the geometry similar to the one one in the video (Cube with 2 subdivision)
  2. Select corner face on top of the cube and extrude it down past one or more edge loops, as seen in the video

Result: Resulting produced geometry is non-manifold:
image.png

Expected: The tool produces manifold geometry, as the name suggests.

**System Information** Operating system: Windows-10-10.0.19041-SP0 64 Bits Graphics card: GeForce GTX 1080 Ti/PCIe/SSE2 NVIDIA Corporation 4.5.0 NVIDIA 451.77 **Blender Version** Broken: version: 2.90.0, branch: master (modified), commit date: 2020-08-31 11:26, hash: `0330d1af29` Worked: Probably never **Short description of error** I have not seen this in a list of known limitations therefore I am reporting this as a bug: [2020-09-03_20-04-43.mp4](https://archive.blender.org/developer/F8836944/2020-09-03_20-04-43.mp4) Extrude manifold tends to produce non manifold geometry if extrusion crosses an edge loop, which is in conflict of what the name of the tool suggest. One vertex connected to both closed and open edges in this way is not considered manifold geometry and also goes against common sense expectation of what kind of result this tool should provide. Resulting geometry qualifies as non-manifold even according to Blender's own operator to select non-manifold geometry: ![image.png](https://archive.blender.org/developer/F8836981/image.png) **Exact steps for others to reproduce the error** 1. Create the geometry similar to the one one in the video (Cube with 2 subdivision) 2. Select corner face on top of the cube and extrude it down past one or more edge loops, as seen in the video Result: Resulting produced geometry is non-manifold: ![image.png](https://archive.blender.org/developer/F8836964/image.png) Expected: The tool produces manifold geometry, as the name suggests.
Author
Contributor

Added subscriber: @Rawalanche

Added subscriber: @Rawalanche

#81098 was marked as duplicate of this issue

#81098 was marked as duplicate of this issue

#80661 was marked as duplicate of this issue

#80661 was marked as duplicate of this issue
Member

Added subscribers: @mano-wii, @lichtwerk

Added subscribers: @mano-wii, @lichtwerk
Member

Changed status from 'Needs Triage' to: 'Confirmed'

Changed status from 'Needs Triage' to: 'Confirmed'
Member

Can confirm.

This was the thing where D5336 was really good at.
@mano-wii: do you think we can get the Extrude Manifold to match D5336 in that respect?

Can confirm. This was the thing where [D5336](https://archive.blender.org/developer/D5336) was really good at. @mano-wii: do you think we can get the Extrude Manifold to match [D5336](https://archive.blender.org/developer/D5336) in that respect?

Added subscriber: @ugosantana

Added subscriber: @ugosantana

Don’t think it’s a bug... I think it’s a limitation. The https://developer.blender.org/D5336 was much more feature rich. I was hoping the current version was going to evolve into that: remove edges, faces and end is a must for this tool.

Don’t think it’s a bug... I think it’s a limitation. The https://developer.blender.org/D5336 was much more feature rich. I was hoping the current version was going to evolve into that: remove edges, faces and end is a must for this tool.

Added subscriber: @Kent-Davis

Added subscriber: @Kent-Davis
True https://developer.blender.org/T80661
Member

Added subscriber: @grosgood

Added subscriber: @grosgood

Added subscriber: @ideasman42

Added subscriber: @ideasman42

Changed status from 'Confirmed' to: 'Archived'

Changed status from 'Confirmed' to: 'Archived'
Campbell Barton self-assigned this 2020-09-22 10:13:26 +02:00

Even though this would be nice to support, it's not a bug/mistake in the code.

Closing.

Even though this would be nice to support, it's not a bug/mistake in the code. Closing.

@ideasman42 Why you make invalid. I already test and there BUG! Extrude Manifold have many bug!

@ideasman42 Why you make invalid. I already test and there BUG! Extrude Manifold have many bug!

Look what I show you there have many bug. https://youtu.be/KslV6N7uPlw

Look what I show you there have many bug. https://youtu.be/KslV6N7uPlw
Author
Contributor

In #80440#1020789, @ideasman42 wrote:
Even though this would be nice to support, it's not a bug/mistake in the code.

Closing.

An issue of a feature which is explicitly called "Extrude Manifold" creating Non-Manifold geometry is considered "nice to support" instead of bug?

The level of mental gymnastics used on this bug tracker to justify dismissing bugs as feature request just keeps surprising me on daily basis.

> In #80440#1020789, @ideasman42 wrote: > Even though this would be nice to support, it's not a bug/mistake in the code. > > Closing. An issue of a feature which is explicitly called "Extrude **Manifold**" creating **Non-Manifold** geometry is considered "nice to support" instead of bug? The level of mental gymnastics used on this bug tracker to justify dismissing bugs as feature request just keeps surprising me on daily basis.
Member

@ideasman42 : would also think this is wrong to close.
Do we have another task that would cover shortcomings/issues with this tool that already covers this? If not, we should have this open (it would be reported again anyways).

@ideasman42 : would also think this is wrong to close. Do we have another task that would cover shortcomings/issues with this tool that already covers this? If not, we should have this open (it would be reported again anyways).

@lichtwerk Yes correct your answer please open again.

@lichtwerk Yes correct your answer please open again.

Added subscribers: @Germano, @juang3d

Added subscribers: @Germano, @juang3d

@ideasman42 @Germano what we are seeing here is that users expect this tool to work in a way that is as it should work, and it's not working as expected, so while you may consider there is no bug in the code per-se, this should be considered a bug, it may be a design flaw on the tool or the code, but the tool is not working as expected, so probably the whole tool should be considered in a buggy state, and maybe even remove or relocate it under the experimental options in the next point update of 2.90.x because it's not working as expected at all.

Keep in mind that no matter if the code works as expected, that does not make the tool to work as expected I'm afraid :S

@ideasman42 @Germano what we are seeing here is that users expect this tool to work in a way that is as it should work, and it's not working as expected, so while you may consider there is no bug in the code per-se, this should be considered a bug, it may be a design flaw on the tool or the code, but the tool is not working as expected, so probably the whole tool should be considered in a buggy state, and maybe even remove or relocate it under the experimental options in the next point update of 2.90.x because it's not working as expected at all. Keep in mind that no matter if the code works as expected, that does not make the tool to work as expected I'm afraid :S

We do use Blender 2.91 Alpha daily build I always download daily. I can see still bug

We do use Blender 2.91 Alpha daily build I always download daily. I can see still bug

Added subscriber: @Ace_Dragon

Added subscriber: @Ace_Dragon

The idea of user-expected behavior merely being a nice to have, is a very FOSS thing that I thought would start to go away now that the BF has over a million dollars coming in annually for development.

I know Germano has a patch that provides a significant improvement to this tool, which will be great when it lands. I also acknowledge the the tool is already useful in many cases. However, the BF in its marketing and announcements has made it clear that they want Blender to be seen as a software for professionals, and to succeed in that requires making polish a priority.

The idea of user-expected behavior merely being a *nice to have*, is a very FOSS thing that I thought would start to go away now that the BF has over a million dollars coming in annually for development. I know Germano has a patch that provides a significant improvement to this tool, which will be great when it lands. I also acknowledge the the tool is already useful in many cases. However, the BF in its marketing and announcements has made it clear that they want Blender to be seen as a software for professionals, and to succeed in that requires making polish a priority.

EDIT: Agree, better not go off-topic here, sorry.

Use this thread to talk about this problem:
https://devtalk.blender.org/t/manifold-extrude-not-working-as-expected-maybe/15112

EDIT: Agree, better not go off-topic here, sorry. Use this thread to talk about this problem: https://devtalk.blender.org/t/manifold-extrude-not-working-as-expected-maybe/15112

Added subscriber: @rjg

Added subscriber: @rjg

Changed status from 'Archived' to: 'Confirmed'

Changed status from 'Archived' to: 'Confirmed'

@lichtwerk I agree, the operator should not produce non-manifold geometry or it should be given a different name.

@ideasman42 I'm setting the ticket status back to Confirmed. The implementation of the operator does not match the description. IMHO even if this issue is a known limitation of the current implementation the ticket should be left open as a ToDo.

@Ace_Dragon While I understand the frustration, please avoid going off-topic.

@lichtwerk I agree, the operator should not produce non-manifold geometry or it should be given a different name. @ideasman42 I'm setting the ticket status back to *Confirmed*. The implementation of the operator does not match the description. IMHO even if this issue is a known limitation of the current implementation the ticket should be left open as a ToDo. @Ace_Dragon While I understand the frustration, please avoid going off-topic.

Changed status from 'Confirmed' to: 'Archived'

Changed status from 'Confirmed' to: 'Archived'

There has been plenty discussion and feedback in the meantime, both on the linked DevTalk thread and on the Blender-Coders channel. I hope this exchange helps to improve the development process to incorporate user feedback early on and document/communicate limitations better.

While I still agree with the criticism of the current implementation, I'm in no position to overrule @ideasman42 decision to close this ticket. Hence, I will restore the previous ticket status. This doesn't mean the issue is forgotten. All developers involved in the implementation of this feature are subscribed to this ticket or have been tagged in the discussion on Blender Chat. If I'm reading the proposal in #75913 correctly, there are already further improvements planned.

There has been plenty discussion and feedback in the meantime, both on the linked DevTalk thread and on the Blender-Coders channel. I hope this exchange helps to improve the development process to incorporate user feedback early on and document/communicate limitations better. While I still agree with the criticism of the current implementation, I'm in no position to overrule @ideasman42 decision to close this ticket. Hence, I will restore the previous ticket status. This doesn't mean the issue is forgotten. All developers involved in the implementation of this feature are subscribed to this ticket or have been tagged in the discussion on Blender Chat. If I'm reading the proposal in [#75913](https://developer.blender.org/T75913) correctly, there are already further improvements planned.

@rjg That again you put invalid why?
Did you read I show you youtube?

@rjg That again you put invalid why? Did you read I show you youtube?

Regarding closing this issue. We already have #75913 (New tool to add and subtract prismatic volumes), don't let closing this be misinterpreted as being ignored.


Note that extrude-manifold does not perform the kind of boolean operation users might want/expect when extruding into existing geometry.

Even if this exact case is handled (overlapping with a subdivided cube), we will still run into cases where one half of the face overlaps, the other half not. Making the fix for this issue more of a workaround.
Where it would be trivial to make simple examples that still fail.

Time spent on some special case detection is probably better spent on #75913, although I'll double check with @mano-wii.

Regarding closing this issue. We already have #75913 (New tool to add and subtract prismatic volumes), don't let closing this be misinterpreted as being ignored. ---- Note that extrude-manifold does not perform the kind of boolean operation users might want/expect when extruding into existing geometry. Even if this exact case is handled *(overlapping with a subdivided cube),* we will still run into cases where one half of the face overlaps, the other half not. Making the fix for this issue more of a workaround. Where it would be trivial to make simple examples that still fail. Time spent on some special case detection is probably better spent on #75913, although I'll double check with @mano-wii.
Author
Contributor

In #80440#1021311, @ideasman42 wrote:
Regarding closing this issue. We already have #75913 (New tool to add and subtract prismatic volumes), don't let closing this be misinterpreted as being ignored.

The issue is that it's been originally closed as invalid, not closed as resolved, and labeled as "nice to have" instead of bug. If the user expected behavior of the tool which is literally named after what it should do not being met can be categorized as nice to have instead of bug, then it sets a precedent where it's borderline impossible to report bugs for Blender as it's just impossible to tell what is one.

Here's an analogy: Imagine a "Extrude Face" operator, which upon execution, would only create new 4 vertices at the specified extrusion depth, but would not generate new triangles to form surface between and among the original 4 and new 4 vertices. Upon reported as a bug, this bug report would be closed as invalid because Extrude Face operator actually creating faces is a "nice to have" feature request, not really a bug, and there's a "feasible" workaround of user selecting sets of vertices themselves and using the Fill operator.

Handling of this bug report falls exactly into the same category.

> In #80440#1021311, @ideasman42 wrote: > Regarding closing this issue. We already have #75913 (New tool to add and subtract prismatic volumes), don't let closing this be misinterpreted as being ignored. The issue is that it's been originally closed as invalid, not closed as resolved, and labeled as "nice to have" instead of bug. If the user expected behavior of the tool which is literally named after what it should do not being met can be categorized as nice to have instead of bug, then it sets a precedent where it's borderline impossible to report bugs for Blender as it's just impossible to tell what is one. Here's an analogy: Imagine a "Extrude Face" operator, which upon execution, would only create new 4 vertices at the specified extrusion depth, but would not generate new triangles to form surface between and among the original 4 and new 4 vertices. Upon reported as a bug, this bug report would be closed as invalid because Extrude Face operator actually creating faces is a "nice to have" feature request, not really a bug, and there's a "feasible" workaround of user selecting sets of vertices themselves and using the Fill operator. Handling of this bug report falls exactly into the same category.
Member

Added subscriber: @Massivetree

Added subscriber: @Massivetree

Also, it may be worth considering that extrude the geometry through the end of the mesh(at which point it seems to flip the normals), it should just remove that geometry all together. So you could essentially punch holes through the mesh with the tool.

Also, it may be worth considering that extrude the geometry through the end of the mesh(at which point it seems to flip the normals), it should just remove that geometry all together. So you could essentially punch holes through the mesh with the tool.
Sign in to join this conversation.
No Label
Interest
Alembic
Interest
Animation & Rigging
Interest
Asset Browser
Interest
Asset Browser Project Overview
Interest
Audio
Interest
Automated Testing
Interest
Blender Asset Bundle
Interest
BlendFile
Interest
Collada
Interest
Compatibility
Interest
Compositing
Interest
Core
Interest
Cycles
Interest
Dependency Graph
Interest
Development Management
Interest
EEVEE
Interest
EEVEE & Viewport
Interest
Freestyle
Interest
Geometry Nodes
Interest
Grease Pencil
Interest
ID Management
Interest
Images & Movies
Interest
Import Export
Interest
Line Art
Interest
Masking
Interest
Metal
Interest
Modeling
Interest
Modifiers
Interest
Motion Tracking
Interest
Nodes & Physics
Interest
OpenGL
Interest
Overlay
Interest
Overrides
Interest
Performance
Interest
Physics
Interest
Pipeline, Assets & IO
Interest
Platforms, Builds & Tests
Interest
Python API
Interest
Render & Cycles
Interest
Render Pipeline
Interest
Sculpt, Paint & Texture
Interest
Text Editor
Interest
Translations
Interest
Triaging
Interest
Undo
Interest
USD
Interest
User Interface
Interest
UV Editing
Interest
VFX & Video
Interest
Video Sequencer
Interest
Virtual Reality
Interest
Vulkan
Interest
Wayland
Interest
Workbench
Interest: X11
Legacy
Blender 2.8 Project
Legacy
Milestone 1: Basic, Local Asset Browser
Legacy
OpenGL Error
Meta
Good First Issue
Meta
Papercut
Meta
Retrospective
Meta
Security
Module
Animation & Rigging
Module
Core
Module
Development Management
Module
EEVEE & Viewport
Module
Grease Pencil
Module
Modeling
Module
Nodes & Physics
Module
Pipeline, Assets & IO
Module
Platforms, Builds & Tests
Module
Python API
Module
Render & Cycles
Module
Sculpt, Paint & Texture
Module
Triaging
Module
User Interface
Module
VFX & Video
Platform
FreeBSD
Platform
Linux
Platform
macOS
Platform
Windows
Priority
High
Priority
Low
Priority
Normal
Priority
Unbreak Now!
Status
Archived
Status
Confirmed
Status
Duplicate
Status
Needs Info from Developers
Status
Needs Information from User
Status
Needs Triage
Status
Resolved
Type
Bug
Type
Design
Type
Known Issue
Type
Patch
Type
Report
Type
To Do
No Milestone
No project
No Assignees
10 Participants
Notifications
Due Date
The due date is invalid or out of range. Please use the format 'yyyy-mm-dd'.

No due date set.

Dependencies

No dependencies set.

Reference: blender/blender#80440
No description provided.