On Thu, Apr 29, 2021 at 12:24:17AM +0200, Aurélien Aptel wrote: > "J. Bruce Fields" <bfields@xxxxxxxxxxxx> writes: > > I'd rather see multiple patches that were actually functional at each > > stage: e.g., start with a server that responds to some sort of rpc-level > > ping but does nothing else, then add basic file IO, etc. > > > > I don't know if that's practical. > > Although it would certainly be nice I don't think it's realistic to > expect this kind of retro-logical-rewriting. AFAIK the other new > fs-related addition (ntfs patchset) is using the same trick of adding > the Makefile at the end after it was suggested on the mailing list. So > there's a precedent. OK, I wondered if that might be the case. I don't love it, but, fair enough, maybe that's the best compromise. --b.
