2021-03-16 Blender Modeling Meeting

Meetings
,
1

Attendees

  • Campbell Barton
  • Howard Trickey

Notes

Improve Onboarding

Ways to make the code-base more approachable.

  • High level documentation for BMesh & mesh editing how internals work & conventions.
  • Review TODO’s & assign “Good First Issue”, (we should be more proactive mentoring these projects).
  • Add-ons may be candidates for porting to C/C++.

Candidates for Sprints

As Blender is adopting “sprints”, smaller teams to focus on an area,
it is work considering modeling projects which could work as a sprint.

  • Performance: this is a big topic, see: T73360.

    A sprint need not focus on all areas at once, it could be limited to a few we know will be valuable.

  • Snapping:

    This area would likely benefit from more focused attention as well as user feedback.

  • Normal Editing:

    This could use a workflow/usability pass.

Other topics we considered, although they are lower priority and don’t necessarily require being part of a sprint.

  • UV Editing

    • There are currently many approved UV editing design tasks.

  • Text Curves:

    • Kerning, general improvements.

  • Curves:

    • Improved editing data structure.
    • Support for custom-data layers.

  • Python API: Memory views on internal mesh data

    This would allow for fast access to mesh data & may be useful for exporters,
    anyone moving data between buffers.

Practical Info

Howard has been added as an owner of the modeling module,
this meeting was the first step to cover areas we think
we could focus on.

Currently there are no plans for when the next modeling meeting should be.

32 Likes
2

Add #blender:meetings-archive tag so that more people see this

2 Likes
3

Hello! Sounds great that you had a modeling meeting, also, you discussed the T73360 topic (performance)

And having Howard and Campbell on this sounds even better! But I’ll not hide my worries on how it was tackled, specially that last quote

Currently there are no plans for when the next modeling meeting should be.

To me, and I suspect for a lot of people too, performace regresion is becoming increasingly annoying, even on top of line CPUs (like 5950x) you can feel the pain. So, Would I ask why this module meeting isn’t being periodically scheduled? also, Is there any other places that this topic is being discussed actively, specially T73360?

Thank you so much!

2 Likes
4

Until there is development time assigned to modeling, it probably doesn’t make sense to hold more meetings.
Keep in mind we have a module chat room where topics are raised and addressed, so it’s not as if the modeling team has no communication in the meantime.

Not actively, this would be discussed in #modeling-module chat root, or more formally as replies on the task.

1 Like