Hi Junio,
On Thu, Jun 11, 2020 at 12:41:56PM -0700, Junio C Hamano wrote:
> Denton Liu <liu.denton@xxxxxxxxx> writes:
>
> > Signed-off-by: Denton Liu <liu.denton@xxxxxxxxx>
> > ---
> > GIT-VERSION-GEN | 2 +-
> > 1 file changed, 1 insertion(+), 1 deletion(-)
> >
> > diff --git a/GIT-VERSION-GEN b/GIT-VERSION-GEN
> > index 06a5333ee6..7b0cfeb92e 100755
> > --- a/GIT-VERSION-GEN
> > +++ b/GIT-VERSION-GEN
> > @@ -1,7 +1,7 @@
> > #!/bin/sh
> >
> > GVF=GIT-VERSION-FILE
> > -DEF_VER=v2.27.0
> > +DEF_VER=v2.27.GIT
>
> I have been thinking about stopping this transition from ".0" to
> ".GIT" and left it at ".0" deliberately.
>
> - The target to create a tarball ("make dist") places the "version"
> file that overrides whatever value is set here, so those who
> build from a tarball will not care what the value described here
> is.
>
> - Those who build from a repository would use "git describe".
>
> - In addition, I do not create tarballs and upload to public places
> for a random version with .GIT suffix---I only do so for -rcX and
> the releases. If anybody else is doing so with "make dist",
> there would be the "version" file included in the tarball,
> recording what is obtained from "git describe HEAD".
>
> So there is no strong reason to care what this value is, and that
> was why I was experimenting with the idea of leaving it at ".0",
> in the hope of hearing from people who do want to see ".GIT" why
> they want it.
Thanks for the explanation, it makes sense.
> Unfortunately, your patch does not say why, either, so it hasn't
> quite helped yet ;-)
I tried looking on the list but I didn't find any mention about this
experiment so I assumed that you unintentionally forgot to update the
version. Good to know that it was done deliberately :)
-Denton
