WebGL roadmap #11369
Replies: 2 comments 1 reply
-
@ianthomas23 I think this is a great list and order. The only thing I'd mention in addition is that (IMO) looking for any opportunities to converge the webgl uniform/vector notions with the Bokeh "dataspec" notion as much as possible over time would also be good as a background goal. |
Beta Was this translation helpful? Give feedback.
-
The original webgl backend used to work like this, though with a lot of limitations and unreliably. We decided to simplify this and let webgl focus only on rendering and leave any high-level processing to the data processing pipeline. If you can do better, that's great. However, keep in mind the maintenance burden and extensibility of existing capabilities when having effectively two implementations. |
Beta Was this translation helpful? Give feedback.
-
Now that we have some (the most commonly used) functionality of
line
,marker
andrect
glyphs working using WebGL and have demonstrated that the approach is worthwhile from a performance perspective, here are plans looking forward for the WebGL backend. Items are in priority order at the top of the list but not necessarily so at the bottom of the list.Separate javascript bundle (issue Separate javascript bundle for WebGL #11053, fixed by Restorebokeh-gl(.min).js
bundle #11371).scatter
with multiple different types of marker (done in PR Support multiple marker types in WebGL scatter #11840).Unify(done in PR Unify WebGL markers and rects #12040).marker
andrect
, asrect
is just a squaremarker
with a 2D size expressed in data not screen unitsline
tomulti_line
,rect
toquad
andvbar
, etc).multi_polygon
s. This needs tesselation of areas into convex polygons or triangles, which will requires an extra JavaScript library or perhaps a Wasm experiment.bezier
.If anyone has different ideas for priorities here, now would be a good time to mention it!
(Edited to strikethrough completed items).
Beta Was this translation helpful? Give feedback.
All reactions