[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Bug#1020413: nmu: bind-dyndb-ldap_11.6-3



Timo Aaltonen kirjoitti 23.9.2022 klo 11.21:
Paul Gevers kirjoitti 22.9.2022 klo 22.26:
Hi,

On 22-09-2022 20:38, Salvatore Bonaccorso wrote:
I honestly don't know because I don't use this package, but I think
it might prevent the users using the bind-dyndb-ldap users from
upgrading the bind9 package.

Why is this binNMU actually needed? bind9-dyndb-ldap has the
following:

Depends: bind9-libs (>= 1:9.16.15), libc6 (>= 2.14), libkrb5-3 (>= 1.6.dfsg.2), libldap-2.4-2 (>= 2.4.7), libuuid1 (>= 2.16), bind9 (>= 9.11)

which is satisifed as well after the bind9 update via
bullseye-security, and updates are possible. Do your request imply
that the relationship would be too lax?

I think there was a change after the bullseye release. The package in unstable has a strict relation instead of a larger-or-equal relation:

Depends: bind9-libs (= 1:9.18.6-2), libc6 (>= 2.34), libkrb5-3 (>= 1.6.dfsg.2), libldap-2.5-0 (>= 2.5.4), libuuid1 (>= 2.16), bind9 (>= 9.11)

bind-dyndb-ldap (11.9-5) unstable; urgency=medium

   * support-9.18.diff: Fix build with bind9 9.18. (Closes: #1006014)
     - drop patches that aren't needed anymore with this
   * control, rules: Use a strict dependency on bind9-libs that the
     package was built against, in order to avoid bind9 updates breaking
     the package. (Closes: #1004729)

  -- Timo Aaltonen <tjaalton@debian.org>  Wed, 23 Feb 2022 13:17:07 +0200

So, Timo, is the package in bullseye broken with the security update and does it need a fix, or is it fine?

It needs a rebuild, because the bind9 library sonames get bumped every time the upstream version changes. That's why the silly strict dependencies were introduced in sid.

Ondrej, let's talk about merging bind-dyndb-ldap to src:bind9 for bookworm, I'm all for it ;)

I'm sure you've seen my MR #21 to add it there, but it got zero comments. I noticed afterwards that you had some concerns about shipping the GPL2 plugin alongside withe the MPL2 bind9, but is it really a problem?



--
t


Reply to: