Nish Aravamudan writes ("Bug#861364: dgit: empty directories are not
representable"):
> On Wed, May 3, 2017 at 8:47 AM, Ian Jackson
> > Call it a bug in git ? :-) I know that doesn't really help.
>
> Yeah -- Robie is working on fixing git.
Coo!
> > My initial plan was to simply ignore empty directories and hope that
> > it didn't matter. If they do matter then we probably don't have a
> > good answer.
>
> Agreed -- I think it is generically not possible to know if a given
> empty directory matters or not. But it does mean if a goal is to
> represent the exact contents of a given source package in the
> corresponding git commit -- that's not achievable in the presence of
> empty directories, at least by default.
Yes.
> I wanted to file this mostly so our importer and dgit did the same
> thing for cases like this :)
Stuff like this is one reason I haven't done a big historical import.
After reifying such an import, it will be difficult to retrospectively
fix this kind of problem.
If you have an actual plan for fixing this then I would certainly like
to hear it. If you have any non-awful workaround ideas, too.
Thanks,
Ian.
--
Ian Jackson <ijack...@chiark.greenend.org.uk> These opinions are my own.
If I emailed you from an address @fyvzl.net or @evade.org.uk, that is
a private address which bypasses my fierce spamfilter.
