01:30:10 RRSAgent has joined #html-3d
01:30:10 logging to https://www.w3.org/2019/09/18-html-3d-irc
01:30:25 rrsagent, draft minutes
01:30:25 I have made the request to generate https://www.w3.org/2019/09/18-html-3d-minutes.html wseltzer
01:30:41 rrsagent, make logs public
01:48:48 koalie has joined #html-3d
01:49:04 RRSAgent, make logs public
01:49:06 koalie has changed the topic to: https://w3c.github.io/tpac-breakouts/sessions.html
01:49:09 koalie has left #html-3d
02:13:50 Roy has joined #html-3d
02:23:59 yaochangli has joined #html-3d
03:19:50 Guest19 has joined #html-3d
04:23:37 Guest19 has joined #html-3d
04:34:42 Roy has joined #html-3d
05:39:45 Roy has joined #html-3d
06:20:53 Roy has joined #html-3d
06:54:51 Roy has joined #html-3d
06:55:37 Zakim has joined #html-3d
06:55:47 RRSAgent, make minutes
06:55:47 I have made the request to generate https://www.w3.org/2019/09/18-html-3d-minutes.html Roy
06:56:33 Meeting: HTML 3D Element & Native glTF
07:02:56 RRSAgent, make minutes
07:02:56 I have made the request to generate https://www.w3.org/2019/09/18-html-3d-minutes.html Roy
07:10:11 zhiqiang has joined #html-3d
07:23:58 dom has joined #html-3d
07:28:12 xfq has joined #html-3d
07:30:27 xueyuan has joined #html-3d
07:31:20 Present+
07:31:29 zhiqiang: good afternoon, everyone
07:31:39 Meeting: HTML 3D element & Native glTF
07:31:42 ... we will have two topic here
07:31:43 RRSAgent, make log public
07:31:50 suzuki has joined #html-3d
07:31:58 xfq_ has joined #html-3d
07:32:01 mdrimages has joined #html-3d
07:32:11 ... zhiqiang from Huawei, mobile platform
07:32:33 ... 3D is becoming a popular type of resource, much richer
07:32:48 shaoboyan has joined #html-3d
07:32:55 jiajia has joined #html-3d
07:32:59 ... to achieve more good experience
07:33:19 ... few example
07:33:46 ... System level 3D/AR Integration, all platform could play 3D
07:33:46 kzms2 has joined #html-3d
07:34:10 ... you could even use phone to do 3d model, there is a example for this
07:34:23 yigu has joined #html-3d
07:35:17 ... once more, for e-Commerce, we could use 3d model to do the goods show
07:35:46 ... social platform also suit for 3d
07:36:08 myles has joined #html-3d
07:36:10 present+ myles
07:36:11 ak has joined #html-3d
07:36:12 ... you also could use 3d model to show the search results
07:36:21 dino has joined #html-3d
07:36:28 Jared has joined #html-3d
07:36:34 jehdh has joined #html-3d
07:36:40 ... This is our Usage Scenario
07:36:50 klausw has joined #html-3d
07:37:04 ... cloud and client side could do interactive
07:37:05 cabanier has joined #html-3d
07:37:16 ada has joined #html-3d
07:37:26 ak_ has joined #html-3d
07:37:36 AmeliaBR has joined #html-3d
07:38:00 ... challenge here: low-level Web3D interface,Inconvenient, Web 3D Performance is not Good, No Easy way for content + processing sharing to do AR
07:38:06 xfq has joined #html-3d
07:38:19 xiaoqian has joined #html-3d
07:38:20 xfq has joined #html-3d
07:38:36 ... our goal is to single line of code to achieve AR and 3d
07:38:53 kip has joined #html-3d
07:39:02 ... this is gITF idea for 3d
07:39:12 s/gITF/glTF
07:39:20 ... there is a simple example of gITF
07:39:40 s/gITF/glTF
07:39:53 ... we bring something in mobile, here is the demo
07:40:31 ... quick app demo, you could use phone to put a car in surround area
07:40:41 ... with a specific tag
07:40:55 ... only one tag included
07:41:41 ... another demo, we could transit 3 model from one device to another
07:41:54 smfr has joined #html-3d
07:42:43 ... I will introduce the api later, we could use this API for the different device
07:43:11 ... 3d will have alive feather
07:43:44 ak has joined #html-3d
07:43:55 ... from developer side, we create something called xmodel
07:44:17 ... attributes list in slides
07:44:36 present+
07:44:44 ... ar, autoplay = true...
07:44:59 ... simple view for the architecture
07:45:27 ... we have sceneView in native
07:46:06 ... there are some thoughts on standardlization
07:46:22 ... khnoros have some word on 3d
07:46:51 ... from W3C side, I want to proposal Xmodel as highlevel
07:47:07 ... also WebXR for lowlevel
07:47:41 ... there are some sample code from Apple, Google, MagicLeap
07:47:56 ... the last one are our code
07:48:01 q+
07:48:03 q+
07:48:17 present+
07:48:19 ...
07:48:56 mounir has joined #html-3d
07:49:07 q?
07:49:16 q+
07:49:36 q-
07:49:38 -> https://www.w3.org/2018/12/games-workshop/papers/new-html-3D-element.txt A new HTML element to display a 3D scene, Yasushi Ando (Kabuku Inc.)
07:49:38 q+
07:50:22 -> https://www.w3.org/2018/12/games-workshop/ W3C Workshop on Web Games
07:50:24 Sushanth: I am come from Microsoft, from immersive WG
07:50:51 -> https://github.com/immersive-web/proposals/issues/52 Native GLTF #52 in Immersive Web CG
07:50:52 ... there is a proposal in immersive WG
07:51:21 -> https://github.com/immersive-web/spatial-favicons Spatial Favicons in Immersive Web CG
07:51:22 ... 3D usage in the web
07:51:58 RRSAgent, draft minutes
07:51:58 I have made the request to generate https://www.w3.org/2019/09/18-html-3d-minutes.html dom
07:52:06 ... you could find 3d people could use for e-commerce purpose
07:52:32 q+
07:52:33 ... you could change the car color and give a house overview
07:52:40 q+
07:52:46 ... or decorate a room
07:53:23 ... you could think about api in highlevel model
07:54:01 ... here is a model creat example
07:54:30 ... you could add model by code, its convinent
07:54:56 ... introduction for gLB and gLTF
07:55:33 ... gLTG JSON represents a scence graph
07:56:02 ... we have already done for implantation for gitf 2.0
07:56:09 ... @@
07:56:34 ... there is a Demo, you will see a 3d demo
07:57:24 ... you could see the details for the model
07:57:40 q+ to ask about accessibility support in glTF in general, and its exposure in a Web browser
07:57:52 ... also we could change the color by gitf
07:58:56 ... there are something not including by gITF
07:59:13 scribenick: dom
07:59:50 Sushanth: we wanted to compare native with JS-based implementation from performance
08:00:02 ... I've used a synthetic benchmark to gauge performance
08:00:26 ... we put tons of gltf models together - you get much better fps in native vs threejs
08:00:42 ... in another test, we compare when we take animations into account - again big difference in performance
08:00:55 ... but that's only true when it uses CPU
08:01:17 ... native only addresses main-thread congestion - it only helps with CPU bottleneck
08:01:25 ... FPS is gated both by GPU and CPU
08:02:20 ... looking in more details, with the three.js based implementation, the bottleneck is the main thread
08:02:41 ... with the native implementation, the bottleneck is the GPU
08:02:50 ... there are opportunities to reduce congestion on the main thread
08:03:05 ... thinking about this further in comparing native vs js
08:03:25 ... you don't want to miss frames in a VR environment
08:03:48 ... another advantage might be to enable model protection - not sure if it's actually needed
08:04:08 ... please bring comments to Immersive Web Proposal github issue
08:04:13 q?
08:04:20 ack cabanier
08:04:28 cabanier: work for Magic Leap on the Helio browser
08:04:33 zhiqiang has joined #html-3d
08:04:35 q+
08:04:40 ... we did introduce the model tag - it's a custom element, not a native HTML element
08:04:51 mdrimages has joined #html-3d
08:05:03 ... we could easily subtitute with another component e.g. the model-viewer component from Google
08:05:14 ... I don't think we need a tag as much as some low-level JS binding
08:05:39 ... people are very peculiar about how to integrate their model, it's hard to do it declaratively
08:05:48 ... I think an API to load a model and render sounds like a better approach
08:05:56 ... I'm happy to also give a demo with my device
08:06:24 ... the rendering is done directly by the OS in our case - the browser is a scene graph
08:06:25 mattwoodrow has joined #html-3d
08:06:34 ack mounir
08:06:46 mounir: some of my questions have been answered
08:07:06 ... what would be the main differences with the custom elements from Magic Leap and Google?
08:07:23 ... you mentioned accessibility - it would be good to hear about specific use cases
08:07:43 ... similarly to go down into more details in the performance issue - may be useful feedback for WebGL or WebGPU
08:08:15 ... one issue for model viewer is stability for an HTML element - feels hard to reach
08:08:23 q+ to ask about a x-org web component?
08:08:33 Sushanth: main difference between JS and native, it seems to be about garbage collection
08:09:14 ... Wrt JS, can it achieve the same performance? Native can do multithreading, which even with workers would be hard to do with JS
08:09:34 ... native has the upperhand on this; they can also use platform primitives (e.g. on Apple)
08:09:45 Mounir: can WASM help wrt GC?
08:10:02 Sushanth: there are still issues when talking back to the DOM since that remains single-threaded
08:10:32 Zhiqiang: the tag would enable native performance, particularly important on mobile device
08:11:20 ... if we have a pre-defined tag, it enables any runtime to use gltf interactive format in a performant manner
08:11:40 ... we can integrate all the gltf spec features and bring a good 3D UX
08:12:05 cabanier: the low level bindings talk directly to the OS - we don't go through openGL in Magic Leap
08:12:10 ... you don't actually parse glTF in JS
08:12:19 q?
08:12:37 Zhiqiang: there are different levels - low vs high
08:13:19 ... if we want to just show a model with some common interface, I think if we can have a related tag for that, and hide all the processing we can get very good performance and power consumtion
08:13:25 q-
08:13:38 ... it also makes it easy for users to integrate content
08:14:01 ack klausw
08:14:15 klausw: wrt your native gltf proposal - excellent discussion there
08:14:20 ... unclear if a tag is ready
08:14:26 ... what if you want to compose models
08:14:28 ... it's uncelar
08:14:38 ... we shouldn't prematurely standardize things
08:14:51 ... I think it would be good to explore with JS solutions for a while
08:14:59 ... unless it's really unusable in JS land
08:15:11 ... it would be great to have a universally supported tag
08:15:18 ... on apple, you would need uszd
08:15:23 ... support
08:15:35 ... which you would have to take that into account in a model viewer
08:15:44 Zhiqiang: we can have different formats like in video
08:15:55 ... e.g. h264 vs vp8
08:16:08 ... I agree that we need to explore more before proceeding, with a polyfill
08:16:12 ack myles
08:16:36 Myles: at a high level, the combination of simplicity and perf is compelling, but several issues need to be addressed
08:16:48 ... different renderers across platforms with different capabilities
08:16:53 ... finding a common subset would be important
08:17:12 ... it would also be a shame if the animation and event didn't align with what is available in CSS
08:17:19 ... that requires lots of analysis
08:17:28 ack AmeliaBR
08:17:43 AmeliaBR: thank you for the presentation of the two proposals - very encouraging
08:17:51 xfq has joined #html-3d
08:17:51 ... different priorities but similar approaches
08:18:05 ... having compared your proposals, any concerns on the one or the other?
08:18:17 ... will you work towards merging?
08:18:37 Sushanth: no much different between the proposals
08:18:52 Amelia: it sounds like xmodel is a high level blackbox like the img tag
08:19:03 ... whereas the native gltf proposal exposes more of the internal
08:19:18 ... and thus exposes some of the possible instability of the underlying format
08:19:38 ... the use cases may be a bit different
08:19:55 ... would be good to understand the differences of use cases and iterate on the features necessary to compare
08:20:01 ... iterate with custom elements
08:20:22 Zhiqiang: we've started sync-ing - we agree on the tag approach, overall the ideas are similar
08:20:29 ... a combination of a tag and native rendering
08:20:38 ... for better performance and quality
08:20:48 ... The JS API would also be useful for controlling the tag
08:20:55 ... it would also be useful in some cases
08:20:57 q?
08:21:09 ... we need more time to explore the use cases
08:21:21 ... the general idea seems to provide support in a simple way
08:21:50 Sushanth: are there any Web developers that would have similar needs to what e proposed?
08:21:59 Jared: I would use it
08:22:21 Dino: a few conflicting desires here
08:22:49 ... if you take the extensible Web manifest as gospel, this is a great example of starting with low level technologies and we're now getting feedback that people want a simpler approach
08:23:00 ... the downside is that you need interop
08:23:17 ... it's relatively easy for 2D (e.g PNG), but for 3D, you get very different rendering
08:23:27 ... we've even changed rendering across iOS versions
08:23:39 ... that's even truer across platforms and OSes
08:24:03 ... even if it's clear that's what the users want, there will still be significant challenges in interop in rendering
08:24:41 ... we've got feedback that taking a model from sketchfab and rendering in iOS led to unexpected results
08:24:43 ack ada
08:24:43 ada, you wanted to ask about a x-org web component?
08:24:50 ada: there are 4 different implementations of model viewers
08:25:08 ... would be worth collaborating on a single web compoennt to flesh out what web developers need
08:25:17 ... once that's done, standardizing something around it might be easier
08:25:28 ... it would already be awesome to work with CSS 3D Transforms
08:25:50 Sushanth: when you need to integrate a library to show models, this create resistance
08:26:00 ... even for a very popular libraries
08:26:15 Ada: it could live in the Immersive Web WG
08:26:19 RRSAgent, draft minutes
08:26:19 I have made the request to generate https://www.w3.org/2019/09/18-html-3d-minutes.html dom
08:26:23 Roy has joined #html-3d
08:28:29 smfr has joined #html-3d
08:28:57 dontcallmeDOM has joined #html-3d
08:30:38 xueyuan_ has joined #html-3d
08:30:53 myles has joined #html-3d
08:31:26 smfr has left #html-3d
08:33:02 dino has joined #html-3d
08:33:33 xiaoqian has joined #html-3d
08:37:02 xfq has joined #html-3d
08:37:10 xfq has left #html-3d
08:38:37 xueyuan_ has joined #html-3d
09:15:45 mattwoodrow has joined #html-3d
09:17:47 xueyuan has joined #html-3d
10:10:53