Fwd: Re: Tag subset patch (gitorious merge request 99)
- 2014-12-09 @ 17:08
On 08/12/2014 23:32, Eric S. Raymond wrote:
> Loz <firstname.lastname@example.org>:
>> No problem, I'll do that. However I think the change to
>> cvs_commit_latest should either go in, or we should revert to the
>> previous tagging mechanism (I'm not sure it does any better,
>> though). Dead commits don't get backlinks. I found a concrete
>> example of this in netbsd-pkgsrc that caused me to write the change.
>> I'll get some more details and pass them on.
> Reasonable. OK, presase resubmit that change with a comment near it
> reminding us that it may have to be revisited if we can't detect spuriously
> dead commits and turn them live again.
In netbsd-pkgsrc: "pkgtools/pkg_install/files/lib/pkg_signature.c" 1.1
is pointed to by tag "pkg_install-renovation-base"
In this case, 1.1 is a cvs synthetic commit indicating the file was
added on a branch. Also it's from 2008, so you would expect a reasonably
recent version of cvs.