Note: This post is part seven of a series of thoughts on the relationship between Canonical and GNOME.
The GNOME User Experience Hackfest of October 2008 spawned a science project to build a new user experience — GNOME Shell — and based on striking similarity in design, appears to have inspired Canonical’s private work on a “messaging menu” stack for persistent notifications.
By early December, too late for the GNOME 2.26 module proposal period, a plan existed to ship the messaging menu in the next Ubuntu release. Both Mark Shuttleworth and Ted Gould subsequently blogged about the design, but no code existed in the public sphere.
January 5, 2009: Ted Gould writes about application indicators — a step beyond just having persistent notifications in the messaging menu:
One thing that kept bothering me in the GNOME UI Hackfest was how little data applications export out to the desktop. […]
What I’d like to put forward is the idea of little flags that applications can hold up to say what they’re thinking or doing, which I’m going to call indicators. […]
First off I want to build something simple (start small, think big), the messaging indicator, which will mostly consist of a GNOME panel applet.
January 8, 2009: In a surprisingly highly-anticipated event at CES Las Vegas, Palm announces the Prē smartphone and its new mobile operating system, webOS. In later reviews, the notification system is widely praised as better than those found in Android and iOS.
February 18, 2009: That code exists for Canonical’s fancy new messaging menu and notification concepts is first revealed with their upload to the “Jaunty Jackalope” development branch. Design and architecture specs for Notify OSD are published on the Ubuntu wiki.
February 21, 2009: Mark Shuttleworth announces that the new work is available in Ubuntu:
Together with notify-osd, we’ve uploaded a new panel indicator which is used to provide a way to respond to messaging events, such as email and IRC pings.
A commenter asks, “Will canonical/ubuntu help on the new gnome-shell proyect [sic]?”
Yes, we helped by sending most of our user experience team to the user experience sprints, and have folks at the UI hackfests. We’re working on notifications in a way that will fit with both GNOME and KDE, and will contribute that at freedesktop.org. And we’re working on the panel and other pieces. We’ll package up the new gnome-shell in due course, but will wait to see what the best defaults for Ubuntu are.
Later in the comments, Mark points out that, “There will be a lot of new work coming from this team, though, and the best place to get it will be Ubuntu ;-)”
March 2, 2009: The GNOME 2.28 development series and module proposal period begin. Neither notify-osd or libindicate/indicator-applet are proposed for inclusion.
March 10, 2009: Mark Shuttleworth files a wishlist bug on Empathy to suggest it use Ubuntu’s new persistent notifications. Empathy developer Dafydd Harries replies:
Mark helpfully ropes in Ted Gould to answer any questions:
Sjoerd Simons tries to be encouraging under the circumstances:
Maybe you or Ted could explain how libindicate/indicators would improve things for Empathy? Screenshots/documentations/mockups/examples are appreciated, patches even more so
Also is libindicate something more gnome applications might start using in the near future? Or would it be an external dependency for gnome just for empathy’s benefit?
To which Ted replies with some links and documentation, but the bug chatter ends.
March 18, 2009: GNOME 2.26 is released.
March 31, 2009: A design specification for the messaging menu is published on the Ubuntu wiki.
April 2, 2009: Without a lot of forward movement since July (GUADEC) or October (Boston Summit and UX Hackfest), the release team decide to shake things up with a wide-ranging proposed plan for GNOME 3.0, written in release manager Vincent Untz’s signature franglish:
So let’s go to the core topic and discuss what the GNOME 3.0 effort should be. We propose the following list of areas to focus our efforts on:
As already mentioned, this is an ambitious plan and it will only be realized if everybody comes out and helps.
In terms of schedule, the release team are bullish but realistic:
Of course, we should be prepared to consider the fact that GNOME 2.30 might not be good enough for us to call it 3.0. […] That being said, we want the community to try as hard as possible to make GNOME 2.30 = GNOME 3.0 a success.
There is an immense amount of discussion, as you’d expect for a significant, project-wide release, revamp and rejuvenation plan.
April 5, 2009: Almost one month after Mark Shuttleworth suggested Empathy support Ubuntu’s new persistent notifications, James Westby provides a patch. Some less-than-diplomatic complaints are raised about the platform-specific nature of libindicate –it’s not even based on a freedesktop.org specification yet!
Xavier Claessens (not generally known for subtlety or a calm demeanour) asks:
Ted Gould corrects Xav’s mistake, knowing full well that optional dependencies are fine:
After some reasonable chatter about the design motivation for Ubuntu’s messaging menu, persistent notifications, and applicable freedesktop.org specs, the discussion trails off.
April 15, 2009: Jon McCann’s first post to the GNOME Shell mailing list. Bryan Clark joins the list and offers some suggestions, including:
You probably don’t want me trying to arrange this stuff for you, but I’d recommend gathering the existing docs and blog posts you have into something more consumable.
Jon McCann replies, “I’m a bit out of the loop too but here’s some background.” It is his first post to the mailing list, and based on the responses from active contributors, Jon’s ideas about GNOME Shell are pretty firmly stuck in October 2008.
April 17, 2009: Mark Shuttleworth blogs about “meta-cycles” across FLOSS projects, clearly aware of the significant — and potentially risky — plans for GNOME 3.0 and the continuing work on GNOME Shell:
On the other side of the fence, GNOME is now more aware of the limitations of indefinite regular releases. I’m very excited by the zest and spirit with which the “user experience MATTERS” campaign is being taken up in Gnome, there’s a real desire to deliver breakthrough changes. This kicked off at the excellent Gnome usability summit last year, which I enjoyed and which quite a few of the Canonical usability and design folks participated in, and the fruits of that are shaping up in things like the new Activities shell.
But it’s become clear that a change like this represents a definitive break with the past, and might take more than a single six month release to achieve. And most important of all, that this is an opportunity to make other, significant, distinctive changes. A break with the past. A big bold move. And so there’s been a series of conversations about how to […] break with the tradition of incremental change, in order to make this vision possible.
April 22, 2009: Canonical’s Ayatana project is announced as an umbrella for the DX team’s various projects, beginning with notify-osd and indicator-applet. The desktop experience and design teams have grown since they were announced in mid-to-late 2008.
April 23, 2009: Ubuntu 9.04 ships with Canonical’s fancy new notifications (notify-osd) and the messaging menu (libindicate/indicator-applet) as features worthy of mention in the official press release.
May 18, 2009: The GNOME 2.28 module proposal period closes. Despite being shipped in Ubuntu 9.04 a month earlier, neither notify-osd or libindicate/indicator-applet were proposed for inclusion.
Late May, 2009: Jon McCann moves from internal projects to Owen Taylor’s section of the Red Hat desktop team — the group assigned in 2008 to work on the results of the UX Hackfest, which became GNOME Shell.
May 25-29, 2009: Ubuntu Developer Summit in Barcelona. The desktop team discuss the plan for GNOME 3.0 in Karmic Koala:
Ubuntu is a GNOME based distribution, having the new versions and components available early will benefit the distribution, users and GNOME.
Users will want to try the new GNOME components and be able to test GNOME3, it will be easy to do so from Ubuntu.
May 25, 2009: Ted Gould submits a talk about the messaging menu for presentation at the Gran Canaria Desktop Summit.
June 15, 2009: Jon McCann blogs about pulling together the design work already done on GNOME Shell, and encourages participation in daily IRC discussions about the design.
July 3-11, 2009: Gran Canaria Desktop Summit.
Ted Gould presents messaging menus.
Jon McCann publishes the GNOME Shell design bible, expanding on design concepts first introduced at the UX Hackfest in 2008.
Owen Taylor presents the latest work on GNOME Shell, including an entirely new design for notifications. While the central location for notifications, and separation between system indicators and notifications from 2008 still exists…
This is not an area that is variously called the Notification Area or System Tray and should not be used by applications (foreground or background) to indicate their status.
… the drop-down panel menu “stack” is replaced with an idea clearly inspired by Palm’s webOS: the message tray.
GNOME Shell appears to be coming together well, so the Release Team recommit to the goal of shipping it with GNOME 3.0… if it’s accepted for inclusion, that is! The team plans to conduct a review, prior to the next release cycle, to decide whether the project is ready to ship 3.0 in March 2010.
August 8, 2009: KDE hackers begin work on a KNotificationItem-based spec for status icons, hoping to encourage a more flexible cross-desktop replacement for the old-school XEmbed-based “system tray” icons used by GNOME and KDE:
This is a draft standard for the management of visual items, usually icons used for reporting the status of an application to the user or provide a quick access to common actions performed by that application. It is intended to be complementary but not directly related with the Freedesktop’s Desktop Notifications specification and is aimed as a replacement to the Freedestop Systemtray specification.
August 10, 2009: The GNOME 2.30 development series and module proposal period begin. libindicate/indicator-applet are again not proposed for inclusion in this series.
August 22, 2009: indicator-applet is split into three: libindicate, libindicator and indicator-applet. The libindicate client library still uses a custom DBus protocol to talk to the indicator host user interface, despite using the org.freedesktop.indicator namespace.
(The org.freedesktop.indicator namespace has never been mentioned on a freedesktop.org mailing list, let alone the XDG specification collaboration list. libindicate has been raised once, in May 2010, in a completely different context.)
September 5, 2009: The KNotificationItem specification is first raised on the XDG list:
In the past few months in KDE we worked on a new way to represent the systemtray icons to overcome the following limitations…
September 23, 2009: Aurélien Gâteau raises his work on Ayatana notifications for Plasma on KDE’s Plasma development mailing list:
I would like to present to you some work I have been doing for Canonical regarding notifications in Plasma.
Before going further, let me state that my goal with this mail is not to get this work incorporated upstream, as I understand the Plasma project have a different position than Canonical on the subject of notifications. I decided to write this mail because I believe it is more correct to let you know about it in a more “personal” way rather than you discovering it in a blog post or a Kubuntu review.
A lengthy discussion follows, during which some familiar concerns are raised. From the pen of Aaron Seigo:
it has been made very clear that Canonical is intent on following through on these ideas as they are. it is an insular project without outside participation, which is something they are perfectly free to engage in and i fully support their right to do so.
now, whether anyone else outside of Canonical thinks it is better or worse seems to really not matter. any discussion on it is therefore a waste of time. and i’d prefer not to waste our time on this list.
… to which Celeste Lyn Paul replies:
Canonical did a really bad job involving upstreams in the design process and are now trying to make up for it by keeping upstreams aware of what is going on. It may be too little too late, but if we (KDE) can at least humor them, then they will be more willing to listen how they can make their software work better with KDE (even if it doesn’t follow what KDE wants to do) instead of just forcing Kubuntu to ship whatever they produce.
September 24, 2009: GNOME 2.28 ships.
October 13-15, 2009: Canonical developers quietly begin early development on a replacement for libindicate called… libappindicator. A “quicklauncher” prototype is blessed as the starting point for a new project called… Unity.
October 29, 2009: Ubuntu 9.10 ships with the libindicate-based messaging menu and “me” menu, which replaces the old FUSA applet.
November 3, 2009: The release team gathers feedback from teams throughout the community to determine if GNOME 3.0 should ship in March 2010, as planned, or be delayed until September 2010.
At this point, we’ve reached the beginning of the libappindicator story, though as you can tell, there was quite a bit of history behind it! Aside from the occasional crucial event in the timeline, I will avoid repeating the details of that saga in this and future posts.
December 17, 2009: Mark Shuttleworth announces that from March 2010, he will focus on “product design, partnerships and customers” and hand over the CEO’s reins to long-time COO, Jane Silber.
I’ve become very passionate about design and quality, and want to spend more time figuring out how we harness the collaborative process to build better, more insightful products. I can’t think of a more interesting challenge, and luckily I couldn’t think of a better person to take over my formal management and leadership responsibilities at Canonical than Jane.
Concerned that the transition might be misunderstood, I leave an encouraging comment on Mark’s blog:
For those who don’t know Jane — as she doesn’t have the most public profile among the stars of Ubuntu and Canonical — be assured that she is all kinds of awesome, far beyond the capacity of English to describe. This is a fantastic transition for Canonical (it’s great that Mark has someone so awesome to hand the reins to), which will in turn be fantastic for Ubuntu.
… and totally cool that Mark will have more time to focus on the stuff he finds fun and satisfying. Further arse-kicking will ensue.
Wonderful news all ’round, congratulations both!
Disclosure: I enjoyed working for Canonical from 2004-2006, and although I have occasionally been accused of shilling for Ubuntu since then, I suspect few at Canonical would regard me as their #1 fan at the moment. I haven’t been involved in GNOME for quite some time, and generally try to avoid thinking about it very often.