2023-05-15 Grease Pencil Module Meeting

Meetings
1

Practical info

Date/Time: 2023-05-15T14:00:00Z
Link: https://meet.google.com/xxo-tyin-rem
Attendees:

  • Amélie Fondevilla (Les Fées Spéciales, developer)
  • Antonio Vazquez (Grease Pencil developer)
  • Dalai Felinto (Blender)
  • Daniel Martínez Lara (grease pencil team)
  • Falk David (Grease Pencil developer)
  • Matias Mendiola (Story Artist - Gease Pencil core team)
  • Nick Alberelli (community Member)
  • Samuel Bernou (ADV studio, Addon dev)
  • Sebastian Parborg (Blender developer)

Topics

  • Grease Pencil 3.0 development updates.
  • Discussion on having the experimental option or not.

Meeting Notes

  • Grease Pencil 3.0 development updates:
    • Falk: Bastien finished his review and raised concerns with how the layer tree is stored. This will have to be reworked and some implementations of the API have to be rewritten. Work on this has already started.
  • Antonio raised concerns with the timeline and the general plan for the project of GP 3.0.
    • Wants to make sure that we don’t rush the release and “force” people to use GP 3.0.
      • Falk: Believes people working on projects won’t switch to a new major Blender version immediately. Blender 3.6 LTS will be available. Agrees that we need to make sure GP is functional and that conversion works without issues.
    • If we re-implement all the modifiers, why not move to a Geometry Nodes based solution directly?
      • Dalai: Thinks that porting the modifiers will be much less work than trying to replace it with a GN based solution directly.
      • Falk: Agrees. Thinks that moving to GN is a much bigger step with more design decisions to be made beforehand.
  • There was a discussion on the idea of having an experimental option vs. a separate build of blender:
    • This will need some follow-up discussions.
    • Pro (experimental option):
      • We’re working in main, collaboration is easy.
      • Users don’t have to download a separate build (just the normal daily build).
    • Pro (separate build):
      • We can already replace the current grease pencil, making development easier.
  • Follow-ups:
    • Check if it is possible to have the old API existing as deprecated during the 4.x series.

Next Meeting

  • Time/Date: 2023-05-30T14:00:00Z
  • Link: TODO
  • Agenda:
    TODO
2 Likes
2

For the records, after the meeting I talked with other developers (Brecht and Sergey mostly) who were still making a case for having Grease Pencil as experimental as oppose to a separate branch.

Falk will be at the studio next week and we will discuss this further. Antonio made a few points in the grease-pencil-module channel as well to be considered.

3

why not having both but prioritizing one and the second wait until when the patches need a review then try to make the second catch the Favorite one ? I’m not aware of the time that is needed duplicate the code in the second but for me it’s appears it’s faster when we want to replace the legacy after … if the duplications would take more time then i have no opinion here. sorry :laughing:

4

From an end-user perspective, as someone who has spent considerable time in learning grease-pencil from bits & tricks scattered everywhere:

  • I think there is a need for blog-post clarifying the difference between GP-current version, and GP 3.0, and the ground reason/explanation for an overhaul, like brief details about current limitations. I don’t think a lot of people EVEN know at all that current GP is 2.0 or something.

This should help with receiving more community inputs that will help with development in various ways.

  • I agree with Sergey & Brecht… in a conditional way about implementing GP 3.0 as experimental feature. imho, most people will download that latest ‘release version’ of blender to not miss out the other major feature that come out with that version. And due to release schedule change, 4.0 is further away. And people are people, psychologically some may perceive 3.x to 4.x as a bigger jump then it actually is.

For now I’m in favor of making GP 3.0 as experimental feature that you can turn-on. And close to 4.0 release date,

  • if it is polished enough to replace GP 2.0, then release it.
  • if it is NOT polished enough to replace GP 2.0, then keep it experimental, and release it in 4.1

The question is how much difference in functionality would be there close to release. I think a decision should be made then. Also, the number of users that download ‘branches’ to give feedback is way lower then people who just turn on experimental features out of curiosity.

I have EEVEE-NEXT feature turned on from day 1, even though I don’t use it, just because of curiosity. I would not have downloaded a separate branch.

  • As for between GN and Modifiers, I again agree with Dalai on this one. From end-user perspective, most of the tutorials on Grease Pencil follow a modifiers based workflow. Lot of people are still putting off Geometry Nodes, because it is whole new subject, and mentally people often don’t have time to mentally commit to learning it (even though there are many wonderful tutorials of each level)

    I have learned GN already, so I don’t mind personally, but I seriously think that GN related approach should be implemented afterwards. Besides, its not a pressing issue.

Take Outline for example. I know there is a GN based method someone created for outlines fairly recently. But many people are still using the old Solidify Modifier approach, specially when they are going to render in Blender.

6 Likes
5

After a short discussion with @brecht, @sergey and @dfelinto, it was decided that we stick to the original plan for now. Experimental option in main. If we run into major issues at any point, @sergey mentioned that we can still branch then if needed. But for now the plan is to avoid that.

2 Likes