[PATCH] D69764: [clang-format] Add East/West Const fixer capability

Mon Aug 9 07:46:36 PDT 2021

erichkeane added a comment.

In D69764#2934535 <https://reviews.llvm.org/D69764#2934535>, @MyDeveloperDay wrote:

>> My 'requires changes' is that this needs an LLVM-project-level RFC to change the charter of one of its projects, doing so in a 15 month long patch, against the wishes of TWO maintainers is a violation of the LLVM community practice.  I'm completely willing to disagree-and-commit here once that happens, but allowing this patch in without that decision being made intentionally by the project seems like a violation of trust.
>
> Ok, thats fair and thank you for verbalising what the changes are. I'm not closed to the idea of the RFC just didn't want to go down that road if it just ended up with 2 opposing views and not getting to a conclusion.
>
> I will challenge a couple of things:
>
> 1. I'm not sure there is currently a "charter" that says we won't modify the contents of a TU, and actually in my view that has already changed when we added. (include sorting, namespace comments, javascript requoter, trailing comment inserter).
>
> 2. I agree if we want to use this as formalising that "change" in charter then I'm ok to try via the RFC but I think we'll get 2 very opposing views, and likely no concencus. So I don't want to just cause a rift in the community any more than this is already.

For better or worse, RFCs are our way of changing these things. RFCs definitely succeed after discussion sometimes. _I_ would be completely against this patch unless some level of community consensus was formed via RFC, and I believe a few others above have made the same point.

> 3. As for the "TWO maintainers", I don't deny their extremely excellent contributions, far greater than mine could ever be. But in fairness they are not frequent maintainers here! On the other hand I am and have been for a number of years. When @djasper and @klimek stepped back a bit, I've really tried to help by filling even a little, the void of their enormous shoes.
>
> I can't even think of emulating all those peoples amazing efforts, but I do this in my free time as I assume other do, and I like to think there is value in me continuing to improve and debug clang-format.

I was referring to @rsmith and @aaron.ballman (to clarify), both are maintainers in 'clang', the former of which is the 'superset' maintainer of this format project based on its directory. While Aaron is a peer-maintainer to this project, his contributions are immense, and his points are too-well-reasoned and motivated to be dismissed.

> So I'd like to think my view isn't disregarded just because others with more muscle disagree, I mean I assumed this was at least a democracy where we could find a fair concencus.

It _IS_ a democracy where we can find a fair consensus!  And the mechanism with which to obtain said `fair consensus` is an RFC.

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D69764/new/

https://reviews.llvm.org/D69764