On 15.08.24 09:20, Michael Banck wrote:
On Wed, Aug 14, 2024 at 06:05:19PM -0700, Jeff Davis wrote:
That looks good to me. Does anyone have a different opinion? If not,
I'll go ahead and commit (but not backport) this change.
What is the rationale not to backpatch this? The error message changes,
but we do not translate configure output, so is this a problem/project
policy to never change configure output in the back-branches?
If the change is too invasive, would something like the initial patch I
suggested (i.e., in the absense of pkg-config, add a more targetted
error message) be acceptable for the back-branches?
But it's not just changing an error message, it's changing the logic
that leads to the error message. Have we really thought through all the
combinations here? I don't know. So committing for master and then
seeing if there is any further feedback seems most prudent.
(I'm not endorsing either patch version here, just commenting on the
process.)
