[trurl]

The key question for me is what about the project made Facebook uninterested in using the language or continuing development. On paper, Skip sounds quite compelling, so I assume there is some reason, if just politics.

[skip-throwaway]

(I work at FB, not on Skip.) The project was originally started with the hopes of converting Facebook's entire (Hack) web codebase to use Skip and support reactivity. It seemed like the Skip team hoped that it would be possible to convert incrementally while reaping performance wins incrementally too.

Converting a multi-million-line codebase takes many years, and in practice, it turned out that huge swaths of the code would need to be converted before any wins were had. (To fully leverage reactivity, all your dependencies also need to be reactive.) Our codebase is intertwined and doesn't have many self-contained parts that could be fully converted in isolation so this wasn't possible.

Essentially, Skip is still very promising as a language for new projects but it feels like Facebook management couldn't justify staffing a permanent team given that we don't have a path to move huge chunks of development to it except in projects being developed more from scratch, which we tend to have few of.

I wouldn't see any of this as an indictment of Skip's potential generally.

[X6S1x6Okd1st]

Thanks for the insight. That seems quite reasonable, especially the fact that you'd need your dependencies to implement it before seeing any speed up.

[Serow225]

Thanks for the insight.