Ask HN: How do you maintain multiple SDKs?
Our product is open source and has several SDKs available in different languages. The commonly used ones are maintained by us, but we also have several that are maintained by the community.
The problem of feature parity is becoming increasingly obvious. We tend to release new features for the few important SDKs that we are more familiar with, while the remaining trail behind.
Also, there are concerns like bundle size and bug prevention that we consider important for our users yet we have no control over when it comes to community SDKs.
Has anyone faced this issue before? Any existing solutions out there? A lot of SDKs are just boilerplate HTTP calls with associated Model files. You write an openapi spec and then use tools to generate the SDKs, cleaning them up as necessary In our case, we do a lot of client-side computing (for data privacy reasons). We also support hashing, and real-time events. These all have different implementations per language that aren't trivially scaffolded. well generate what you can and focus man-hours on the actual work in each SDK Or implement it in C/C++ and then create bindings for each language