Strip the Soul

We come to close of another year.  I thought I would report out on our out reach process.  By outreach, I mean the overall perception of the GNOME project by outside community.

I wanted to write this post earlier, but unfortunately, we’ve had a death in our family and  I was not able not able to really think about writing that post.  I feel that this post is important and so I’m taking the time to write this down.

Overall, we’ve made some great improvements in perception from the beginning of the year.  The 3.8 and 3.10 releases have been very successful and people have really appreciated the polish that have gone into these releases.  So, congratulations to all of you for all the great work contributors in GNOME have done this year!  As we look forward, I would like to summarize some of my perceptions from working with the FOSS community.

One of the things I want to highlight is that active participation in the community gets us great dividends as a project.  Thanks to creating community channels, core members participating in G+ communities, and twitter we have really stayed engaged!  Because of this engagement, perception of our project have become extremely positive from what it was last year.  As some of you know, I spend an inordinate amount of time arguing on GNOME’s behalf on the Internet.  I am happy to note that my job has become a lot easier with many people in the community who stand up and declare that they love GNOME 3,and willing to challenge others in the various forums.  Considering the huge backlash that we’ve had the past couple of years, I know that for a lot of people it has been a source of stop energy.  We spent the entire time at the GUADEC foundation meeting focused on this.   I am happy to note that we are seeing some great positive changes!

We have turned the corner and it’s thanks to all of you for being available, for answering questions.  Specifically, I would like to thank Allan Day, Ray Strode, Emmanuele Bassi, Matthias Clasen, and Jasper St. Pierre for always being available to respond to summons when I needed them.  Certainly for humoring me to be sure! :-)  My job has become a lot easier and participating in threads in http://reddit.com/r/gnome and http://reddit.com/r/linux are a lot easier as others have taken up the reins.  I know personally it has been gratifying to see the positive responses from the community after years of head butting and sometimes feeling the only person out there being positive.  Heck, even in Slashdot, we’ve seen some push back.  Thanks to all you in the community who do like GNOME 3, and are willing to say so.  It makes working on software so much more worthwhile!

It’s time to build on that success though.  We need to continue to open ourselves up and continue to build our community.  While I have been an advocate for GNOME to the community, I must also in turn become an advocate for the community in GNOME.  It’s up to the project to look at this feedback and respond appropriately.

Some issues worth addressing:

  1. Changes to Nautilus have been met with universal unhappiness – the changes that have been made have made a lot of people unhappy.  There just hasn’t been anybody I’ve met both who are fans of GNOME 3 and critics alike who like what the current nautilus have become.  People have either been silent or have encouraged the use of the 3.6 fork of Nautilus.  There are some reasons for it, things like  improvements have been promised but have not been forthcoming fast enough.   Secondly, the loss of being able to split the screen and be able to manage files have angered many.  The work around of having two nautilus windows and managing that way has not assuaged people.  They really miss this feature.   I’ve had a hard time arguing this as, I personally do not use nautilus to manage files.
  2. People have really been down on GNOME designers.  While they have made some great community outreach, specifically Allen Day and Jakub Steiner have always been available to talk about their designs.  Regardless, there is always this sense that whatever feedback is given will be ignored that everything is inevitable.  Even when asking module owners, they , to use an American idiom “pass the buck” and say, “ask the designers”.  Intriguing, is it not? :-)  In normal, free software, the maintainer of a project has the final say, but not necessarily in GNOME.  That puts significant pressure on designers.  What is the balance in working with community, but also have the time and motivation to work on designs in progress?  That is a hard challenge and I do not envy our GNOME designers their position as they must provide results and if you have to manage the community as well?.  As community, we should understand that some unknown, possibly significant work has now been shunted through three individuals.  I don’t know if that is particularly a good thing, certainly it is a stressful situation if you’re responsible for designing an entire eco-system.  I am not proposing a solution, I’m highlighting what I feel is a problem and a possible single point of failure.  Designers need to spend some small part of their time in public space.  Because like it or not, they are providing user experience of GNOME these days, and just about everyone else have faded in the background, proverbial noise.  Why?  Because it is strongly perceived that nothing can be changed without designers driving it.  I have some ideas.  I’ll discuss in another post.
  3. The community does not understand design.  This links to #2.  When you’re designing software, a lot of things depend on each other.  If you change something, it has ramifications to the entire design.  I don’t think users understand that.  When designers ask for feedback there is no agreed upon specific way to provide it.  Since users don’t understand that changing one aspect has ramifications elsewhere, it’s not that easy to revert something that is already planned out, so when people complain about regressions they don’t understand the context why those regressions occurred in the first place and reverting it back to the old behavior has consequences.   It would be good to know what manner of feedback will get a design a second look.  What would that feedback entail?  What is the contract with the community for design changes, if any?  We are in uncharted waters when it comes to this and if we are going to do design out in the public space and use our community as guinea pigs to a continually evolving set of designs then it is paramount that we create tools that will provide the right feedback.
  4. It’s very important that we are up front on regressions.  The gnome-terminal incident is a good example of this.  There is no doubt that transparency is a popular feature in terminals.  One could argue that they don’t contribute particularly to being useful, they do provide a kind of eye candy that  attracts the eye, like a bright paper to a magpie.  It’s not good when a feature that is used and is quite popular is taken away in a midst of a code clean up and we are ignorant of it.  I will ask module maintainers to be upfront to the release team when there is a significant regression like this.  In turn, release team needs to tell  the engagement team as well so that we are also ready to talk about it when it comes up.  Bad news travels, with social media it travels even faster and it takes on a life of its own.  Don’t keep the engagement team in the dark.  Be up front.  We have a reputation and branding that needs continual care and feeding.  We need to be always aware that any incident can last for days, weeks, years, or decades and can be very intense.  Let’s not put the engagement team into firefighting mode needlessly.  I have some ideas to mitigate some of this as well.

The conclusion is, we are creating a product.  But we need to act like we are creating a product.   That will require closer teamwork between the various teams that we have before.  I’ll talk about this in another post.  But we don’t have everything set up for that.  We have gaps, and they should be addressed.

Like the Holidays, I provided a little sweet  and a little spice!

I would like to wish all of you a Happy and Prosperous New Year!  Thank you for all your hard work in GNOME this year.  To the folks using our software, I would like to wish you as well and that you will continue to love and trust where we are doing and where we are going even if you have minor disagreements in how we go about it.

8 Responses to “Strip the Soul”

  1. Sandeep says:

    @sri – why no love for Debian ?

    I mean I understand the Redhat love and all, but some of us have a lot invested into the debian ecosystem through our work/dayjobs/etc.

    It’s really hard to get excited about Gnome 3.1x when all we can do is gape from afar.

    Ubuntu hasnt been a big help either – 14.04 Trusty will probably not include a lot of the new gnome UX (including header bars).

    It would be great Gnome has some sort of a distro through Debian – right now, us debian guys have to hunt around through Pinguy, siduction, etc. and pray for a reasonably working distro.

    Hell, I’d be willing to support a Kickstarter on this!

  2. Sandeep says:

    AFAIK – the closest you will get to the Fedora setup on Debian is through Siduction

    http://news.siduction.org/2013/12/release-notes-for-siduction-2013-2-with-systemd/

  3. Friendly Penguin says:

    I’m glad to see that your hard work is paying off. However, I’m not sure that your assumption that you’ve turned detractors into supporters is correct. My own experience is that most of the critics have simply stopped engaging and gone elsewhere.

    [WORDPRESS HASHCASH] The poster sent us ‘0 which is not a hashcash value.

  4. Bastian says:

    Great read sriram, thanks for summing up the year! Really helps understanding the efforts the engagement team does and what needs to be adressed in the future!

  5. Stu says:

    Some great points. On the gnome-terminal front, I’m using ubuntu and dreading the day when it becomes unforked and transparency goes.

    It’s really useful to pop a full screen terminal on top of a full screen browser and type in something that written underneath.

    Sure, cut n paste achieves a similar thing, but it’s nice to not have to clobber the buffer.

  6. Zeeshan Ali says:

    Happy new year Sri and really good write-up. I had a comment on one of the points you made:

    ” I will ask module maintainers to be upfront to the release team when there is a significant regression like this.”

    I couldn’t agree more. I have been taking the time always to write proper release notes for every release of my modules. Sometimes this task of going through commit log and creating a summary for release team and distro packagers, gets really frustrating when releasing multiple modules but I think its worth it.

    I would suggest that we emphasis more on the need to provide proper release notes for every release and also the fact that its not the same as the commit log. Perhaps we could have some rule for this?

  7. Greg says:

    “When you’re designing software, a lot of things depend on each other.”

    Unfortunately, this also means that some bugs have a much greater effect than you might expect. Keyboard focus and window management has been handled poorly on X11 since day one, before GNOME began. This makes some designs untenable and so GNOME, KDE, et al. have moved toward less sophisticated designs than they could have achieved. Microsoft Windows, though not X11-based, has similar problems and a similar design trajectory.

  8. […] Strip the Soul « Sri Ramkrishna […]

    [WORDPRESS HASHCASH] The comment’s server IP (65.39.148.173) doesn’t match the comment’s URL host IP (65.39.148.219) and so is spam.