On Fri, Sep 10, 2021 at 10:19:04AM -0400, Jeff King wrote: > But there is one thing that does get weird if we don't do this patch. If > we later take the approach of checking: > > #ifdef CURLOPT_PINNEDPUBLICKEY > > then that will subtly shift the cutoff point from 7.44 to 7.39 anyway. > _If_ we are going to do that conversion in a later patch (as this series > does), I think it makes sense to shift the version number explicitly in > a commit with an explanation, as this commit does. Ah, nevermind. I just saw Ævar's re-roll where that strategy turns out to be a bad idea anyway. So that is no longer a compelling argument. :) -Peff
