svn commit: r195626 - in head: contrib/amd/amd contrib/ntp/scripts contrib/sendmail/cf/sh secure/lib/libcrypto sys/contrib/dev/acpica/compiler sys/dev/hptrr

John Baldwin jhb at freebsd.org
Mon Jul 13 15:32:53 UTC 2009


On Sunday 12 July 2009 7:42:30 am Colin Percival wrote:
> Brian Somers wrote:
> > On Sat, 11 Jul 2009 22:30:37 +0000 (UTC) Colin Percival <cperciva at FreeBSD.org> wrote:
> >>   Remove build timestamps from the following files:
>  >>   [...]
> >>   After this commit, the only remaining build timestamps are in the kernel,
> >>   the boot loaders, /usr/include/osreldate.h (the year in the copyright
> >>   notice), and lib*.a (the timestamps on all of the included .o files).
>  >
> > This is a fantastic change.  Our new-branch-per-project methodology
> > @work made us implement this change locally.
> > 
> > Another change that was necessary along the same lines was to remove
> > any hostnames or path names that leak into object code, the most noticeable
> > of which is of course one of the first kernel boot message lines.
> > 
> > Are there plans to change kernel code and/or hostname/path name stuff?
> 
> I'm not particularly concerned about the hostname or paths, since there are
> workarounds (i.e., set the same hostname and use the same path, and files
> will come out the same).
> 
> The timestamps in the kernel and boot loaders can probably be removed, but I
> wanted to consult with the community first, since it is conceivable that those
> timestamps are useful in some situations.  I'm even less certain about the
> timestamps included in library archive files -- if anyone with binutils-fu
> can confirm for me if these are used at all, and if not, how to get rid of
> them, I'd be very very happy.

I find the timestamps in kernel builds quite useful when doing work in feature
branches.  OTOH, back when I used sendmail as my MTA on my mail server I found
the date in sendmail.cf useful as well.  I don't use sendmail anymore so I am
no longer worried about that one, but the kernel date one I would hate to
see go away.

-- 
John Baldwin


More information about the svn-src-head mailing list