Theme changes occur often enough and for enough instances that for many Open edX providers, this can become cost-prohibitive in terms of CPU time. (In one specific case: “1000 hours of compute time in an m5.large machine for building all of them”).
Product/Platform Value
This furthers adoption, as it reduces the friction for providers that require custom theming, by reducing the need to rebuild images. And in addition to reducing the number of builds, this has the potential to reduce build time. It would also remove the blocker for multi-tenancy sites to upgrade beyond Koa.
Acceptance Criteria
A theming system where artifacts were built, and possibly hosted, separately would solve the problem.
Related or in-progress work
OpenCraft has begun preliminary work via these two co-dependent ADRs:
Problem
Theme changes occur often enough and for enough instances that for many Open edX providers, this can become cost-prohibitive in terms of CPU time. (In one specific case: “1000 hours of compute time in an m5.large machine for building all of them”).
Product/Platform Value
This furthers adoption, as it reduces the friction for providers that require custom theming, by reducing the need to rebuild images. And in addition to reducing the number of builds, this has the potential to reduce build time. It would also remove the blocker for multi-tenancy sites to upgrade beyond Koa.
Acceptance Criteria
A theming system where artifacts were built, and possibly hosted, separately would solve the problem.
Related or in-progress work
OpenCraft has begun preliminary work via these two co-dependent ADRs:
Contingencies
https://github.com/openedx/tcril-engineering/issues/418