obliterate functionality?

cowwoc cowwoc at bbs.darktech.org
Wed Mar 19 20:47:47 CDT 2008



Matt Mackall wrote:
> 
> Sure. But first, who's allowed to do it? There is no central authority.
> SUN does not have the authority to broadcast obliterations of the
> changes in my OpenSolaris branch, and I don't have the authority to
> obliterate theirs. Though it might be fun to try to obliterate some
> security fixes.
> 

They don't have the right to obliterate *your* branch, but they have a right
to obliterate theirs. If you pull code off their main OpenJDK branch and
they obliterate files there then future pulls should obliterate on your end
as well. If you want to create a private branch of OpenJDK as a diff on top
of the main branch that's fine and obviously they wouldn't be able to
obliterate copies there. The point is they own the main branch and you own
your own branch.


Lastly, there's no way to enforce such a flag.

We're not talking about enforcing it. We're talking about Sun modifying
their official repository to remove some confidential information. This
pleases the lawyers just fine. To be clear: we are *not* trying to solve the
problem of guaranteeing the copy is obliterated on remote machines who
already pulled a copy of the code. If their client follows the protocol then
it will obliterate their local copy in the main branch, but obviously they
could have made a copy beforehand or hacked their client. That is okay.

Gili
-- 
View this message in context: http://www.nabble.com/obliterate-functionality--tp16114445p16169812.html
Sent from the Mercurial mailing list archive at Nabble.com.



More information about the Mercurial mailing list