please stop "pre-promoting" packages

[Matthias Klose]

"pre-promoting" packages is a practice moving a package from universe to main 
without review. please stop it! It may save the promoter a few hours, but it 
adds to the workload of others, and undermines security, QA and the MIR process 
(I have hardly seen a completed MIR for a pre-promoted package).  Things seen 
during the maverick and natty release cycles:

  - pre-promote a package and don't subscribe the MIR team. These go
    unnoticed unless noted for other reasons, e.g. cleaning up NBS, or
    uploading for library transitions.

  - pre-promote a package which did fail to build on some architectures.
    If it FTBFS it's not fit for main, and pre-promotion leads to
    massive backlogs.

  - pre-promote a package, submit a MIR "this looks easy" and leave the
    report alone to rot.

  - pre-promote a package, submit a MIR, get feedback from the MIR team
    "does the daemon run as root?" setting the report to incomplete,
    and let the report expire.

  - claim that feature X is super-urgent for desktop feature Y, not
    wanting to wait for security reviews.  Maybe a configure option
    to en/disable feature X would be better for the time?

Most main inclusions can be planned for.  The urgency of a MIR often comes from 
starting a major update and not looking for MIR's at all or not looking at 
further dependencies of a package wanted for promotion.  One possible exception 
are our auto-syncs from unstable, which shouldn't be too much of a problem with 
the debian freeze.

The current practice doesn't work, so I would like to ask archive admins to stop 
pre-promotions until we decide to change or keep the current MIR process.

   Matthias