Peter Eisentraut <peter.eisentraut@2ndquadrant.com> writes:
> In my experience, changes to function declarations in header files
> happen a lot in forks. So applying the pgindent change to backbranches
> would cause some trouble.
> On the other hand, it seems to me that patches that we backpatch between
> PostgreSQL branches should normally not touch function declarations in
> header files, since that would be an ABI break. So by not applying the
> pgindent change in backbranches we don't lose anything. And so it would
> be better to just leave things as they are.
Maybe we could wait awhile and see how much pain we find in back-patching
across this change. I have to admit that the v10 pgindent changes have
not been as painful as I expected them to be, so maybe this round will
also prove to be just an annoyance not a major PITA for that.
Another thought is that, at least in principle, we could re-indent only
.c files not .h files in the back branches. But I'm not sure I believe
your argument that forks are more likely to touch exposed extern
declarations than local static declarations.
regards, tom lane