andre klapper’s blog.

This is an update about cleaning up the GNOME stack for GNOME 3. This has also been posted to the desktop-devel mailing list.

This status report refers to the aims listed in the 2.27/2.29 schedule and the automatic statistics available (which now also covers the Mobile section, hence results can be worse than last time).

Maintainers: I have listed available PATCHES AWAITING REVIEW.
Please take a look if your module is listed and review/commit NOW so the changes can receive enough testing for 2.28.

THE PROBLEMS: What migration paths are missing?

• libgnomeui provides retrieving thumbnails of files. There is no substitute yet. This blocks deskbar-applet, gnome-mag, evolution and probably many others.
• Currently EggSMClient gets copied into all apps. That’s not cool but the way to go until Session Management support in gtk+ gets resolved.
  See Thomas’ blog for a current list.

This list is of course not complete. Also see LibgnomeMustDie.
Feel encouraged to add your issues.

ZERO modules with Glib-Deprecated-Symbols

NOT COMPLETED (”Reopened”) now that we also check external deps and the Mobile set:

• Still to do: gconf-dbus, evolution-data-server-dbus.
• External deps to do: dbus-glib, hal, libnotify, mono. PATCHES available: dbus-glib, libnotify. FIXED: farsight2, libnice, poppler.

Officially ANNOUNCE libglade as deprecated in favor of GtkBuilder

DONE.

Less than 35 modules depending on libglade.

COMPLETED.

• low: 25
• average: 5 (dasher, gnome-media, gnome-panel, gok, zenity)
• complex: 2 (gnome-control-center, evolution)
• PATCHES awaiting review/commit: gnome-control-center, gdm, gnome-nettool, gnome-mag, gnome-media, gnome-menus, gnome-panel, gnome-session, gnome-system-tools, gtkhtml, sound-juicer, zenity, tracker. Maintainers please review/commit.

Clear a11y plan and schedule for 3.0

NOT COMPLETED.

Less than 12 modules depending on libgnome

NOT COMPLETED (Progress compared to 2.27.1: 22->15).

• low: 10
• average: 4 (Evolution, gnome-media, yelp, anjuta)
• complex: 1 (gok)

Please share experiences and knowledge.

Less than 12 modules depending on libgnomeui

NOT COMPLETED (Progress compared to 2.27.1: 15->12).

• low: 9
• average: 2 (Evolution-Exchange, gnome-panel)
• complex: 1 (Evolution)

Please share experiences and knowledge.

ZERO modules dependening on gnome-vfs

NOT COMPLETED (Reopened):

• average: 1 (gst-plugins-base)

Gtk-Deprecated-Symbols

• low: 8
• average: 7 (gnome-control-center, evolution, gedit, metacity, glade3, gconf-dbus)
• complex: 2 (gnome-games, gnome-media)
• PATCHES awaiting review/commit: gnome-control-center, gedit, metacity, yelp, glade3, policykit-gnome

Evolution-Data-Server must be migrated to D-Bus by default

NOT COMPLETED. Evolution schedule currently under discussion.
A Git branch is available.

WebKit status report for 2.27.5

IN PROGRESS. WebKitGTK+ has been proposed as an external dependency.
See d-d-l for the status.

Evolution to get rid of Bonobo by 2.27.3

NOT COMPLETED and postponed for 2.29.1.
See KillBonobo for the status. Testing and reporting bugs is HIGHLY welcome. See Matthew’s blog for more information.

Complete migration from HAL to DeviceKit-* by 2.27.3

NOT COMPLETED.
According to “jhbuild rdepends hal –direct” the following modules still depend on HAL:

• brasero
• cheese
• gnome-power-manager is FIXED

More important stuff to take a look at:

Not yet covered in the stats but required to fix are also:

• GTK+/GLib Single includes (Metabug):
  • To Do: gdm, gail
  • PATCHES awaiting review/commit: gnome-mag, gtksourceview
• GSEAL:
  • To Do: A lot. Developers please start taking a look at this.
• LibSexy deprecation:
  • To Do: Vino
  • PATCHES awaiting review/commit: anjuta, tracker, PolicyKit-gnome

Nice to fix:

• Adding Introspection support.
• AM_SILENT_RULES / shave bugs
• Porting to PolicyKit 1.0: PATCHES awaiting review/commit: gdm, gconf, gconf-editor, gnome-applets, gnome-panel, gnome-session

GNOME Showstoppers

For GNOME 2.26/2.28, I have posted a Showstopper Review earlier this week. Feel free to take a look, test & help out, get things done.

Other activity

Kudos to the progress that has been made so far!
Getting rid of Popt is basically DONE.
ZERO modules dependening on Esound is DONE.
ZERO modules dependening on Gnomeprint is DONE.
The Website revamp front is rocking, and the Documentation team also has some great momentum currently.

GNOME 3

Today I have released a GNOME release schedule proposal for 2.27 and 2.29 (please keep discussion streamlined on desktop-devel mailing list instead of e.g. the comments section of this blog).

GNOME 2.30 will be GNOME 3.0 (if it’s considered to be ready in an early 2.29 stage – else 2.32 will become 3.0), as proposed by the GNOME release team at GUADEC 2008 (Remember some blog entries and press coverage?).

All the additional stuff (”additional” compared to former schedules) listed in the proposal is technical and under the hood, as this schedule proposal also covers streamlining of the platform by getting rid of deprecated modules.

Note that such a “boring” platform cleanup is not meant to completely define GNOME 3. This schedule proposal is just a proposal for being a part of what will be called GNOME 3.

By purpose, the schedule proposal does not cover any potential UI changes, any potential redefinition of GNOME, any blingbling UI, any complete rewrites from scratch or any implementations of a semi-working kitchen sink. If you are interested in such fields and topics, I am not the person to talk to. Please see other more generic threads about GNOME 3, especially Vincent’s d-d-l posting and blog post and Lucas’ blog post.

Also, this plan does not cover by purpose stuff like gnome-shell vs. gnome-panel, gnome-zeitgeist, or Vala, PolicyKit, PackageKit, etc. Discussing them is recommended, but IMO these modules are not crucial for the GNOME3 schedule.

On the other hand, gconf vs dconf is also not yet covered by this plan, but definitely crucial to discuss.


So… if you are also interested in cleaning up your favourite desktop and getting it to the next level: Join the discussion (e.g. about the future of GNOME’s a11y), or take a look at the status overview and provide patches in GNOME Bugzilla. Help is highly welcome, especially when it comes to getting rid of libgnome and libgnomeui. Code needs to be written here. Soon.

GNOME 2.26

I didn’t blog about our beautiful 2.26.0 release as I’ve been busy with relocating and job, but I must mention my heros for 2.25 – those brave folks that provided a lot of patches to clean up and helped on the way to GNOME 3. This is definitely not a complete list, but names like Thomas Andersen, Cosimo Cecchi, Luis Menina, Maxim Ermilov and Saleem Abdulrasool come to my mind. Thanks guys, you rock!

• Oh yeah, there will be a 2.26 Showstopper review on Monday. For those interested in sneak previews:
  • Current list of 2.26 blockers (if there’s something important missing, please ping).
  • Current list of 2.28 blockers.
• Congrats to muelli: No (public) NEEDINFO bugs older than six months left in GNOME Bugzilla. There simply IS no untidiness!
• You want to be a code monkey? You want to write boring patches? You love reading non-existing API documentation? (Uh, wait, I didn’t actually say this!)
  Spent some time filing bug reports (some of them quite trivial) against modules that use deprecated Gtk symbols, deprecated Glib symbols, or other deprecated stuff, so now there’s even more gnome-love bugs that you can choose from to provide an awesome patch for. Or simply take your beloved module and remove its dependency on libgnome(ui) (yes, that might be a bit more complicated, but we all love challenges, don’t we?).
  It’s all highly welcome as it helps on the way to the next major version of {Gtk+/Glib/GNOME}.

More news to come in the next weeks, e.g. a schedule for 2.27.

This is a GNOME Release team service announcement.

Yes! We want you to think about potential new modules for 2.26.
If you’re a maintainer, go wild and propose your favorite new modules for inclusion in GNOME!

How to propose a module? Click here.

The new modules proposal period will end on Monday November 24th at 23:59 UTC.
We expect discussion to heat up about those proposals at the beginning of January and to reach a decision around January 12th.
Also note that early feedback&testing help maintainers/developers to work on issues. Don’t start testing in January so developers don’t have time anymore to work on concerns.

More information about the GNOME 2.25 release cycle.

Congrats everybody to 2.24.0! Been hard work sometimes, but always fun.

Now while you, dear user, are probably going to send zmrzlina, chocolate, cheese and kisses to your beloved favorite hackers*, you might also consider becoming involved. Here’s some offers I have in mind (unfortunately all unpaid positions). Not all of them require coding skills (I don’t code either, so “Why don’t you do it yourself?” comments are unrequired).

• nautilus-actions is looking for a new maintainer, see the announcement. Interested in helping out? Or even fixing the most favourite bug? Feel free to contact the former maintainer if interested!
• Update ancient project websites. We have lots of old projects sites in http://www.gnome.org/projects/. You don’t want to code but you know about HTML and a bit of CSS and your Wnglish is good. A list was provided by element3260 as part of his GHOP work, available here – this should be transfered to a wikipage at live.gnome.org first, see the Gnome Goals pages for an example how this could look like.
• Push the Gnome Goals to clean up our stack for 3.0 by providing patches. There’s still enough to do. Also see GioPort and GtkPrintPort.
• Take a look at bugs in Bugzilla with gnome-love keyword set. They are a good place to start working on the code of your favourite project.
• Unrelated to this, worst translation issues will become blocker bugs for GNOME 2.26 too. Providing good applications definitely also includes a well-localized UI, and this is impossible when not fixing them.
  Some examples (not to blame developers, but to get a feeling what this means): Missing context, split sentences, english structures, broken plurals.
  Most (not all) translation issues are easy to fix and gnome-love bugs too. Just take a look at bugs with L10N keyword set.
• Identify unmaintained modules. I’d love to have a small script that could run on my svn checkout and printing the date of the last svn commit (svn log –limit 1) that was not an updated .po file. But I don’t do code, so if anybody has too much time… *cough* Would make it easier to identify rotten modules in order to remove them from Bugzilla and the translation statistics, so nobody wastes time anymore on them.Another shameless request – I’d also love to have information on the language team pages about the date of their last commit for that language. Would make identifying non-active teams or maintainers easier.
• Nothing in that list for you? Don’t want to code? For more ideas, visit JoinGnome.

(* “hackers” refers to developers, translators, artists, marketing folks, bugsquadders, doc writers, …)

Yay! Catch me if you can.