GNOME and Ubuntu GNOME

[Stephen Michael Kellat]

On Sat, 27 Sep 2014 12:25:24 +1000
Tim <tim at feathertop.org> wrote:

> 
> On 27/09/14 07:42, Steve Langasek wrote:
> > Hi amjjawad,
> >
> > On Fri, Sep 26, 2014 at 10:06:23PM +1000, Ali/amjjawad wrote:
> >> Since I have joined in July 2013 to Ubuntu GNOME until this very
> >> moment, users never stop complaining about the fact that Ubuntu
> >> GNOME can not include the latest GNOME release and I have
> >> explained that many many times. The last time I did that was:
> >> https://lists.ubuntu.com/archives/ubuntu-gnome/2014-September/002278.html
> >> No matter how many times we keep repeating ourselves, this is just
> >> a bad/unhappy news to our users.
> >> I am writing to you after I have seen so many negative feedback
> >> here:
> >> http://www.omgubuntu.co.uk/2014/09/gnome-3-14-wont-included-ubuntu-14-10
> >> Is there anything whatsoever we could do? GNOME foundation and
> >> Ubuntu Release Team could work more closely, maybe? is there
> >> something I/we are not aware of? sorry to assume that - just
> >> trying to think the same way our users think - but is it because
> >> Ubuntu moved to Unity and left GNOME behind as not the default DE?
> >> can we find a way around that? maybe an official statement from
> >> Ubuntu Release Team? anything that could reduce the complaints?
> > In your mail to ubuntu-gnome, you write:
> >
> >   If you take a look at it, and know the fact that within one
> > month, we can't test Ubuntu GNOME 14.10 with the latest GNOME
> > (3.14) and then release it, etc etc ...
> >
> > This seems to be a decision of the Ubuntu GNOME Team, not something
> > that would involve a statement from the Ubuntu Release Team.  In
> > the past, when the Ubuntu desktop was more closely based on GNOME
> > and there was greater alignment between the two projects, there was
> > careful synchronization between the release schedules to ensure
> > that the latest GNOME would be included in each Ubuntu release.
> > This was certainly not a trivial amount of work; it required
> > several engineers to track the upstream betas and point releases in
> > the last months before release and update the packages regularly in
> > coordination with the Ubuntu Release Team.  And I certainly
> > understand if the Ubuntu GNOME Team aren't willing to commit to
> > this same work on a volunteer basis.  But ultimately this is a
> > decision for the Ubuntu GNOME team to make (with the support of the
> > Ubuntu Release Team), not for the release team itself.
> This is certainly not our decision, it was the Release Team (or
> desktop team) that decided to stick with GNOME-1. Without the current
> Gtk, we can't possibly ship the current GNOME. There are also a
> number of core components that are lagging behind, upower, bluez,
> gnome-desktop (pending ffe, but has been blocked for a year
> basically), network-manager, systemd. This makes it pretty impossible
> to ship the current GNOME
> 
> Then we have all the apps that are shared with Ubuntu, where the
> GNOME UI changes conflict with what Ubuntu want these can't be
> updated either. The end result being we are not even close to
> shipping a full 3.12 for Utopic, and this is in my view a much
> greater issue than being stuck on GNOME-1. And the gap seems to be
> growing each cycle, it is pretty typical that most of our updates (to
> even GNOME-1) are blocked until very late in the cycle, often having
> to land things after Freeze. Now when V-series opens, rather than
> start working on 3.14, we will be playing catchup for 3.12.
> 

With the ever-closer integration of GNOME and systemd, the decision by Debian to pick up systemd as default init that Mark then adopted pretty quickly for Ubuntu has some consequences here.  The timing of systemd-related things landing has implications not just for Ubuntu GNOME but for other flavors too that have to be addressed in transitions such as for upower.  LP Bug 1330037 is an example of a wide-ranging transition that has impact across most flavors except seemingly Mythbuntu and Edubuntu.  As the Ubuntu realm transitions to have systemd as its default init, actions taken with respect to Ubuntu GNOME are increasingly likely to have impact across all flavors due to GNOME's unique integration with systemd.  We have to tread carefully in this case.

As to the commentariat on OMGUbuntu, I refer to this: http://m.xkcd.com/1385/

> 
> >
> > I think it's too late to revisit such a decision for 14.10 without
> > significant downside for the quality of Ubuntu GNOME; but for
> > 15.04, I recommend that you explicitly consider at the beginning of
> > the release cycle whether you would like to release with the latest
> > version of GNOME, and discuss with the wider community to
> > understand how to make this work without negatively impacting other
> > flavors in the Ubuntu ecosystem.
> The real problem here is our large overlap with what is now
> essentially the legacy ubuntu-desktop, if we were shipping more or
> less an upstream GNOME stack (more inline with debian), this would be
> fine to track and ship current GNOME in 15.04, but the reality is
> that we have to contend with a massive ubuntu delta that is usually
> specific to Unity, and with the Canonical teams focussed on unity8
> and phone its been pretty hard to get the various blockers resolved.
> >
> 

Final beta is out now.  We are basically at the final polishing point to prepare for release and to finish up bug-squashing.  The best thing I can suggest moving forward is to make 14.10 as stable and presentable as it can be while also doing very critical assessment.  We're coming up soon enough on the Victorious Velociraptor or whatever Mark names it.  Take some time to carefully assess how this cycle played out.  Take time to set out S.M.A.R.T. goals that you feel you can reasonably achieve within a 26 week cycle where you can generally assume you have 16 weeks until Feature Freeze hits.  If there are still going to have to be transitions needed for GNOME DE landing in the Victorious Velociraptor cycle, start reaching out to the other flavors before Utopic Unicorn ends to build relationships to be able to work out such things amicably.  All these things need to be happening alongside QA actions up to and past release on October 23rd.

Details about what S.M.A.R.T. criteria happen to be and how they can be used in goal setting can be found here: http://en.wikipedia.org/wiki/SMART_criteria

Stephen Michael Kellat