Currently, stdlib has a bespoke plot API which is useful for fast static rendering. However, our implementation is quite limited in the types of plots it can produce. The goal of this idea is to refactor our plot API to build atop of vega (or its specifications). For this, we'd need to migrate to an async plot generation API, which is probably necessary regardless if we want to support WebGL or some other async rendering engine.
Ideally, we would retain the same plot API and internally generate a vega specification.
Expected Outcomes
We can generate simple plots using the new plot implementation.
Involved Software
This will involve using vega (or something similar depending on whether vega is sufficiently maintained). We will want to transpile to ES5 and vendor in order to ensure that we can support our supported Node.js versions.
Prerequisite Knowledge
JavaScript, Node.js.
Difficulty
Intermediate/Hard.
Project Length
350 hours. This project has the potential to spiral out of control, as there are many unknowns we'd need to answer. Mentor would likely need to be actively involved in order to perform R&D and properly scope.
Idea
Currently, stdlib has a bespoke plot API which is useful for fast static rendering. However, our implementation is quite limited in the types of plots it can produce. The goal of this idea is to refactor our plot API to build atop of
vega(or its specifications). For this, we'd need to migrate to an async plot generation API, which is probably necessary regardless if we want to support WebGL or some other async rendering engine.Ideally, we would retain the same plot API and internally generate a vega specification.
Expected Outcomes
We can generate simple plots using the new plot implementation.
Involved Software
This will involve using
vega(or something similar depending on whethervegais sufficiently maintained). We will want to transpile to ES5 and vendor in order to ensure that we can support our supported Node.js versions.Prerequisite Knowledge
JavaScript, Node.js.
Difficulty
Intermediate/Hard.
Project Length
350 hours. This project has the potential to spiral out of control, as there are many unknowns we'd need to answer. Mentor would likely need to be actively involved in order to perform R&D and properly scope.
Potential Mentors
@kgryte @Planeshifter @rreusser