You should fork it if you're doing development work like Apple, Intel, AMD, Nvidia, etc because that's how open source works, but it's not necessary if you're using the tooling as-is.
It might be helpful to consider how other people will read your comments before getting defensive next time. It wasn't until I looked at your links that I understood what you were trying to say because it's not the meaning most people are going to assume.
He tried, but people weren't interested in reading his contributions. It's hard to expect people to force contribute something to a collective which is simply not interested.
Downvoting without asking for rationale was the reason he stopped worrying about comment votes, not the other way around.
The fact that lots of people are doing some thing, doesn't make this thing right. After all, "I did it because the other guy did it" is an excuse for children.
This depends a lot on what you're doing with LLVM. If you are just using LLVM as a code generation backend for your language frontend, you generally do not need an LLVM fork.
For example, while Rust does have an LLVM fork, it just exists for tighter control over backports, not because there are any modifications to LLVM.
You should have ordered them differently to not distract from your point: That most LLVM-based compilers do eventually need to write patches for it and usually ship them before upstreaming, they can't just use an existing release. I at least have always found that to be true to build such compilers from source.
Oh interesting I didn't realize the role of adjectives has changed recently. So war and my nuclear war mean the opposite things now. I'll keep that in mind.
