Review: Syncing from testing a success?

Kees Cook kees at
Thu Apr 8 21:38:48 BST 2010


On Thu, Apr 08, 2010 at 09:41:48PM +0200, Martin Pitt wrote:
> So, in your opinion, did syncing from testing
>  (1) help to avoid introducing larger breakage into Ubuntu (for the
>      domain you are usually watching)

I have a few hats, but for security, this doesn't apply as much since
security uploads in Debian unstable tend to quickly move to testing.

>  (2) meant a smaller or larger amount of review and sync requests

For me, this was larger, as I had to request syncs with some what more
frequency to updates that were sitting in unstable (and usually couldn't
wait 10 days for it to move to testing, for example).

>  (3) made it easier or harder to merge with Debian and get changes
>      integrated back upstream

Didn't notice a change here, other than what was noted for 2 above.

>  (4) made library transitions easier or harder

I wasn't directly involved in any, but my gut said this was actually
harder, since it created some issues when suddenly everything for
a transition appeared in testing, instead of following the same
order/progression as they went into unstable.

>  (5) anything else that caused or eased problems that you can think
>      of

My general opinion is that for _development_ we should return to unstable,
and during _stabilization_ we could sync from testing.  I've long suspected
that the auto-sync stops much too early.  If we auto-sync from unstable
until the current cut-off, and then auto-sync from testing for a while
after that, I think we'll get the best of both worlds.

Barring that, I'm strongly in favor of returning to syncing from unstable.


Kees Cook
Ubuntu Security Team

More information about the ubuntu-devel mailing list