glTF Working Group Meeting 2022-01-27 #52
Replies: 1 comment
-
|
Meeting notes: OMI GLTFAttendance
NotesOMI_audio_emitter explainerReviewing PR for explainer: #46 Need to reach out to people that contributed, even if people haven't contributed they should be listed, even if they haven't signed the CLA. Action Items:
In order to contribute you have to sign the CLA tim: Should have people and not deny listing them as contributors. Legally we aren't providing specifications. We're disallowed, so the SLA stuff doesn't matter yet. We'll need to un-opt-out of that for the SLA to apply and for us publish tim: We should block work until we ammend the charter. It's only fair if we would block listing people if they didn't sign the CLA ifire: It's a legal requirement for contribs to sign the CLA tim: I think everyone that contributed should have been recognized So we'd need to remove their contributions before releasing? tim: WHat if we gather the list first and go from there? ifire: Proposal to create an issue and have a checklist and block until we merge time: Should have list and talk about it, before the list we shouldn't finalize it ifire: So we should ahve a list and talk about contributions, then ammend the charter? tim: Yes, we can't publish specs ifire: Do the list right now? tim: Remove the contributors section? lyuma: Should have steps written up, what are the steps we should take, where does the contributor list take place? Tim: CLA should not equal "did contribute" ifire: e.g. degrees to rad change is a substantial contribution ant: in wordpress the criteria is: did comment inside tickets to track issues, anyone that contributed code, anyone that contributed an issue. Implies you're a user of the system and these are contributions. It's how we measure and avoid the grey area. If you contribute hosting a meeting, that's not rewarded props Rick: I don't think lurking in the voice chat is worthy of credits antpb: I think for this doc, we're contributing code. In my case I'm making a pull request, and why my PR was signed by the CLA lyuma: I agree I think you should be listed there, I don't think I should be a contributor or listed there, haven't signed the agrreement ifire: I recognize you also added the change lyuma: It was a single word, I don't think it counts and don't want it to be. ifire: autoplay tim: The reason for the CLA is to account for the potential of people doing stuff while on the clock. ifire: if the contribution was ok with the w3c and was legally met I think that would work too tim: All I'm saying is opting out doesn't help here. We'd need a release or something, otherwise it's a grey area tht persists antp: I think we need to define it and none of this will get merged until that happens, another implementation is another thing we need. ifire: table this? antpb: not time sensitive lyuma: Not even stage 1 yet Mauve: Mauve reviewed omi-collider and a omi-physics body spec based on previous conversations from github and discord. Roughly based around Hubs and other game engines. You can add an extension to a node in a gltf scene called omi_collider. Schema is that it is an object with a type that defines what type of collider it is. In the case of hull, it is convex. Mesh is a full mesh. Each type has optional properties, like box can have a radius which sets w/h/d to defined value. Or you can specify an extents property which is an arraty that contains w/h/d. Mauve: Todo do implementers need to set radius standards? ifire: If no default should you set a value? Mauve: use default in the spec if not set. For something Antpb: for things like radius, its hard to anticipate radius so maybe no default? Mauve: todo, grammar and spelling ifire: limit poly for hull to 255 similar to Unity. Avoids conflics. Lyuma: we should find the lowest poly limit to set as the max for hull Mauve: Callout a node can have a mesh assigned ot it and a mesh with a collider assigned to it Mauve: Discourage folks from using Mesh without good reason. Maybe document better options. ifire: regarding mesh limit, should be the normal mesh limit. follow gltf limits Mauve: we should discuss the usage How can you have multiple meshes and colliders. Compound colliders are an option. Node could have one collider under a physics body. Or a node that has physics bodies that are colliders. ifire: is an array of ids bad? Mauve: you can have more than meshes as compound colliders. you can have multiple types merged together. Box, mesh, sphere example. One alternative is an array of colliders, but how do you position? Mulitple nodes means you can transform the nodes and the collider will inherit. Array of colliders should specify those trnasofmraitons. ifire: so you can group based on parent? yes! :D Mauve: capsule, how do you dtermine and change orientation. One option is to change the orientation in the collider object. Use trnsforms from scene graph. Lyuma: unity is limit to what can vbe in the compound collider You need to use cardinal axis. Limit how things can be rotated. If you want compound, no rotate. can but aligned to the axis. Rick W: Local axis. You can rotate via transform, bound to xyz. Mauve: wording will need to account for this. Mauve: will research Mauve: we have a basic physics body spec as part of this spec. Marks nodes as a physics body. Single property type, static, dynamic, kinematic, trigger ifire: static for me means it can never move. Kinematic is movable by script. Lyuma: in godot static can be moving at constant velocity. Rick: When I wanted to move a 3d object by collider, use the physics system that you have. Sometimes its okay to lean on the physics engine. You need to use scripts ot interface ifire: dynamic means it can be moved, static cannot be impacted by physics engine but are in the world and can do things. ifire: if we dont define kinematic it leaves it open for extension. ifire: we need use cases first. Mauve: wonder if there can be extensions for your extensions Mauve: Table animation retargetting, prioritize next meeting |
Beta Was this translation helpful? Give feedback.
-
OMI glTF Working Group Meeting 1/27/2022
This meeting will be on 1/27/2022 at 10:30 PM (UTC) / 2:30 PM (PST) in the OMI Discord within the #omi-weekly-meeting channel. During the meeting, we will be using the #omi-gltf-extensions channel to manage a speaker queue, post links, and for any sidebar discussions.
To be notified of this meeting and others, subscribe to the OMI Meetings and Events Calendar or add yourself to the @omi-gltf-subgroup role in the #roles channel of the OMI Discord.
Agenda:
To propose another item for the agenda, comment below (preferably before the meeting).
Beta Was this translation helpful? Give feedback.
All reactions