I think making a git branch which has auto-merging (but only if not conflicting with the upstream) is more time consuming than just making a branch at the current state.

More likely this involves some kind of release engineering; anyone who has worked on large projects will tell you that it's not an effort to take lightly.

Personally I feel like this fork is not really serious, but the issue might be that it either becomes serious or Rust's leadership will not change course.

So, I'll give them the benefit of the doubt and quietly hope that I will never need to use crablang.