On Mon, May 21, 2018 at 10:40 AM, Robert P. J. Day <rpjday@xxxxxxxxxxxxxx> wrote: > On Mon, 21 May 2018, Elijah Newren wrote: > >> I had always assumed prior to your email that 'known to Git' meant >> 'tracked' or 'recorded in the index'... > > i *know* i've been in this discussion before, but i don't remember > where, i *assume* it was on this list, and i recall someone (again, > don't remember who) who opined that there are two categories of files > that are "known to git": > > 1) files known in a *positive* sense, those being explicitly tracked > files, and > > 2) files known in a *negative* sense, as in explicitly ignored files > > can anyone refresh my memory if that happened here, and whether that > was the consensus after the discussion was over? Perhaps this: https://public-inbox.org/git/EEC5BA1D5F274F02AE20FC269868FDEF@PhilipOakley/ ? > If that's the > definition that's being used, then this passage makes sense: > > "Normally, only files unknown to Git are removed, but if the -x > option is specified, ignored files are also removed." > > that pretty clearly implies that ignored files are considered "known" > to git. Yes, _if_ that's the definition used, then that passage makes sense. But if that's the definition used, then the other two passages I pointed out in Documentation/git-commit.txt and Documentation/git-rm.txt do NOT make sense and need to be rewritten. Junio has already chimed in elsewhere on this thread and stated pretty clearly that the intended meaning for 'known to Git' was just (1), not (2), and even provided a suggested wording fix for Documentation/git-clean.txt. Putting that into a patch format and submitting along with an update to Documentation/glossary-content.txt as Duy suggested look like the two todos to me, though perhaps others want to discuss ways to just avoid the phrase 'known to Git' (as suggested by Jonathan).
