New module decisions for 2.30


The release team met last week to eat some ice cream, chat about who
will be the villain(s) in the next Batman movie and play some football
(or soccer if you're living in a country where there's another football
game ;-)). We still found some time to discuss the new module proposals.

Many thanks to the people who contributed to the discussion on the list,
and to the authors and maintainers of the proposed modules!

Short summary

Please make sure to read the details for modules that are of interest to
you, as the release team generally comments on why a module is approved
or rejected, with recommendations that we'd like to see followed.

  gmime (external dependency)
  libdb (external dependency)
  vala (external dependency)
  gnome-packagekit (desktop)
  nautilus-sendto (desktop)

In, but not as expected:
  tracker (external dependency instead of desktop)
  dconf (not for 2.30, but pre-approved for 3.0)

Blocking on external issues:
  clutter-core (see details below)

  couch-db, evolution-couchdb (desktop)
  globalmenu (desktop)
  nautilus-actions (desktop)

Withdrawn by maintainer:
  emerillon (desktop)
  libvtemm (desktop)


 + gmime (external dependency)
   - needed by totem-pl-parser (in 2.28 already)
   - would be needed by tracker
   => approved
   => it would make sense to make evolution-data-server/evolution use
      gmime to remove code duplication.

 + libdb (external dependency)
   => already approved by mail

 + vala (external dependency)
   - would be needed by parts of dconf and tracker
   - gathered quite some interest already
   => approved
   => the release team encourages maintainers of modules using vala to
      put generated files in tarballs to enable compilation without vala
      being installed.

 + clutter-core (desktop)
   - already adopted by the GNOME community
   - already an external dependency
   - not hosted on GNOME infrastructure, but tarballs and API docs are
     now there (missing: git and bugzilla)
   - copyright waiver possibly limits contributions:
   - copyright assignment is also an issue
   => the release team is working with the Foundation to investigate the
      copyright waiver and copyright assignment, and with Intel to find
      an appropriate solution.
   => feedback from the community at large on what solution would be
      appropriate is welcome.
   => at least bugzilla should be moved to the GNOME infrastructure.
   => rejected, until those (non-technical) issues are solved. We still
      support the project as we believe it's really essential for GNOME,
      especially in the GNOME 3 context.

 + couch-db, evolution-couchdb (desktop)
   - not really used widely at the moment
   - it seems the mozilla javascript engine is a highly recommended
     dependency. Until we come to a decision about the mozilla vs webkit
     engines, this is an issue.
   => rejected: too early for wider adoption.

 + dconf (desktop)
   - agreement it's the way forward
   - concerns about migration of settings
   - concerns about the lack of planning for admin tools (pessulus and
   - concerns about the fact that we need stuff in glib but that's not
     there yet (although we know there's a plan for this)
   - a massive migration from gconf to dconf would be preferrable
     (instead of having some modules using gconf and some other modules
     using dconf). We know it might not be realistic, though.
   => rejected for this cycle, but pre-approved for the next cycle
      (assuming glib gets the required API for the next cycle). The
      additional time should be used for careful planning of the above
   => we encourage developers to look at it and to create gsettings
      branches for their modules (like devhelp and gedit).

 + emerillon (desktop)
   => withdrawn by maintainer

 + globalmenu (desktop)
   - still considered as a hack by platform maintainers
   - does not align with the GNOME 3 plans at the moment
   => rejected

 + gnome-packagekit (desktop)
   - the PackageKit integration in GNOME offers nice features
   - the PackageKit tools are not used widely by all distributions right
   - some parts of gnome-packagekit have a place in the desktop suite,
     but not all of them, so it's not clear what to do right now.
   => approved
   => we believe packagers will only take the parts of gnome-packagekit
      they want in their packages anyway. Packagers can contribute
      patches to make this easier for them as Richard is open to this

 + libvtemm (desktop)
   => withdrawn by maintainer

 + nautilus-actions (desktop)
   - some think it's mostly targetted at advanced users, and so it's
     fine to keep it out of the desktop
   - should potentially be integrated in nautilus itself in a
     non-intrusive way if possible
   => rejected, but we can give it more visibility as part of our goal
      to advertize more GNOME applications

 + nautilus-sendto (desktop)
   - widely adopted
   - integrates in a non-intrusive way to nautilus
   => approved
   => it might make sense to integrate the feature directly in nautilus
      if possible (and have the plugins directly in the applications

 + tracker (desktop)
   - this is tracker 0.7 (not 0.6)
   - mixed feelings in the community
   - the tracker team has done some good work recently, and we want to
     give a chance to integrate their work
   - we need more integration first; right now, it seems to be mostly
     integration of the search and not the metadata
   - we think applications should not rely on the indexer being enabled,
     since people might disable it. This should hopefully already be
   => approved as external dependency
   => we encourage tracker developers to clearly separate the indexer
      from the store to avoid confusion in the mind of non-tracker



Les gens heureux ne sont pas pressés.

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