Hacker News new | past | comments | ask | show | jobs | submit login

> it's roughly feasible

What do folks even mean by "Google's index"?? Google results combine tons of signals, including personal histories for each user. Sharing metadata for the top billion urls wouldn't cover half the functionality, or make a competitive engine. And on the other hand, there may not be a single other organization in the world prepared to manage a replica of the entire data plane that impacts seatch. The proposal is somewhere between underspecified and nonsense.

Thanks, this is mainly what I came here to say. And I just don't see even the vaguely defined "index" as the crown jewel. If anything, it's "relevant results", which is something quite different.

Applications are open for YC Winter 2020

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact