Change in the uname

Warner Losh imp at bsdimp.com
Wed Jan 27 19:20:43 UTC 2021


On Mon, Jan 25, 2021 at 3:03 PM Li-Wen Hsu <lwhsu at freebsd.org> wrote:

> On Tue, Jan 26, 2021 at 5:04 AM Ian Lepore <ian at freebsd.org> wrote:
> > That all sounds good, except the 'v' is a bit odd, normally it implies
> > a "version" number, which isn't exactly what this is.  How about 'n' or
> > '#' since it's just a count?
>
> Seconded, sorry for discussing the color of bikeshed, but 'v' is also
> not intuitive to me. Since it is still a counter, and we are removing
> the 'g' in the second part, cXXXXX-HASH is still distinguishable with
> the old cXXXXX-gHASH.
>
> Or, maybe call it gXXXXX-HASH which stands for git commit count XXXXX,
> with hash value HASH?
>

I can do 'n' instead of 'v'. I think that 'g' would be a bit confusing
since the connection to git is less obvious.

I happen to like the 'old school' versioning that would do X.Y(Z) where Z
is a count of changes (whatever that means) (either since X.0 started or
back  to the start of the thing). That would be branch(version)-hash. I
opted to not implement that, though. While I like it, I know I'm old school
and this style has fallen out of fashion since the 80s and 90s when it was
popular on the 'big iron' (TOPS-20, IBM, etc) as the 'big iron' culture has
given way to other schools of thought. It's also harder to script parse and
harder to implement.

All in all, I think 'v' or 'n' is best and most clear. # I'd like to avoid
because in the non-reproducible build case #x is the build number.

Any last comments before I roll this in later today to hit the MFC window?

Warner


More information about the freebsd-arch mailing list