In my previous two blog posts, I have talked about growing our community and communication. In this post I will be focusing on communication and how we can best improve it, and in so doing expand our community. A number of things can be done to improve our communication, some of which I laid out in previous posts. But one way or another we must improve – both within GNOME and the wider free software community.
In GNOME, as in most free software projects, the majority of communication takes place online. This in of itself, does not make it open. Many discussions take place primarily on IRC and are, in effect, private. When discussions eventually bleed over onto bugzilla and/or mailing lists, they are usually close to, or have already been resolved. The result is that most users and developers receive notice of changes only when they have already been finalized (or nearly so). Too often this is after they have been announced by a third party, or (even worse) a 4th, party such as Slashdot or Reddit.
Obviously, this is not an ideal way for any organization to project its message. The resulting fallout is generally negative, putting GNOME developers, designers and defenders alike on defense. Whats more, defensive responses are generally reactive, and are often perceived as dismissive. Of course, nobody likes to be dismissed and the result is another round of negative feedback, thus perpetuating the cycle. It is this cycle that we must break if we wish to remain relevant and respected.
Remaining relevant and respected should, I think, be a priority. In order to do so we must develop a better system to communicate what we’re doing, why it is important, and when it will occur. When we remove features, we need to explain our reason for doing so to users. For example, what was it about compact view mode that was truly harmful to Nautilus? We must also ensure that we avoid breaking API’s whenever possible. When we do so, we must explain to developers (especially those outside of our own insular community) why it was necessary, and how to ensure their work does not become obsolete. This means explaining ourselves, writing documentation and helping others whenever and however indicated.
If we continue to break API’s at random (like those in GTK+3 [1]), we will continue to alienate developers, and with them, users. This should be a concern, but does not seem to be for some GNOME developers. By ignoring the needs of others, we drive them away, which turns them from being supporters into detractors. When it appears that we are doing so on purpose, by breaking API’s and purposefully ensuring that extensions and themes cease to work, we anger and alienate users and developers alike, sometimes permanently.
But hope springs eternal, and change and growth for the better can (and hopefully will) occur. Primary to that is respect – for others’ work, ideas and opinions. Being open to others’ thoughts and a willingness to listen and consider outside ideas. Secondary to all of this is communication, on both sides. For those within GNOME, communicating what they are doing, and where they expect and hope to take GNOME in the future. For those on the outside, a willingness to listen, withhold judgment and avoid knee-jerk reactions. With just a bit more understanding and compassion, GNOME can continue to move forward, as a vibrant and relevant member of the wider free software community.
[1] https://igurublog.wordpress.com/2012/11/05/gnome-et-al-rotting-in-threes/
Todo en Tiempo 
November 13, 2012 at 8:28 am
You forgot to include the footnote of where we are breaking APIs at random. Because I’m not aware of that happening anywhere.
November 13, 2012 at 8:32 am
Updated – sorry!
November 13, 2012 at 10:17 am
Ah yes, the uninformed rant about GTK3 not working like GTK2. I elaborated on that in https://bugzilla.gnome.org/show_bug.cgi?id=687752#c9 and https://bugzilla.gnome.org/show_bug.cgi?id=687752#c13 and (most relevant to your blog post I think) https://mail.gnome.org/archives/gtk-list/2012-November/msg00044.html – so I’ll not say anything but the fact that your post felt uninformed and taking sides against what I’m doing.
November 13, 2012 at 11:05 am
It’s this flippant and dismissive attitude that is a large part of the problem.
The other part of the problem is illustrated by the 3 separate links to two different bugs on bugzilla and the mailing list and the final comment in one that it should be taken to IRC! There is no reason to spread discussions out in such a fashion, making them hard to find, let alone follow, unless your goal is to hide them from the general population… which brings us back to communication being a huge problem! 🙂
November 13, 2012 at 12:29 pm
Emily, that is not helpful *at all*.
IRC is what we use to discuss, as well as the mailing list. the bug was closed and the discussion moved to the mailing list, with the option of discussing in real time on IRC. all public, all available. nobody is hiding anything from anyone; if you want to involved in gtk+ development, as well as third party extensions like themes, then you have to go where the gtk+ developers are.
plus, there’s the whole “CSS ids and classes are not API” angle, but the discussion on the mailing list should already clarify that.
November 13, 2012 at 8:57 pm
Reading those bugzilla comments and mailing list posts, my first thought is that some people have their priorities a bit askew. Certainly, it might be nice if the Gtk theming APIs were a little more stable, but the grand assertions about it harming adoption are a bit comical. Of all the features of Gtk, I can’t imagine *anything* lower on the list of priorities than the ability to put a different skin on it.
November 13, 2012 at 10:34 am
I like how Mr. Otte acknowledges breaking APIs at random, but fails to acknowledge it being a problem. In the first bug post, he’s actually being an asshole about it too. Typical Gnome. Many people rely on an ecosystem created around Gnome, but the Gnome core does not care in the slightest about that community. And then they keep getting surprised that they’re being badmouthed and forked all the time. Because they’re impossible to work with, and they’re that way on purpose.
There’s more to the word “community” than “@gnome.org”. And “communicating” also involves “listening”, not just “telling”.
If you break other people’s things all the time, you’re doing it wrong. Period.
November 13, 2012 at 12:06 pm
I take offense at characterizing me as an asshole, in particular in situations where I’m honest about the situation we’re in. But then, calling me an asshole is certainly easier than trying to (a) understand and (b) solve the problems at hand. So feel free to continue doing that.
November 13, 2012 at 4:26 pm
The usability of mailing lists and IRC is just awful. It should be as easy as posting a comment on a blog, stackoverflow or G+. The FOSS community would benefit from a platform like G+ that could be privately hosted (e.g. on gnome.org) to replace mailing lists and IRC. If I recall correctly the Fedora Project (Mairin Duffy) is looking into something like this to improve the usability of mailing lists.
November 13, 2012 at 4:47 pm
I’m reading this as “stop using IRC”. That’s obviously not going to happen. You also didn’t address the particular point of what you thought was wrong with the way fallback’s removal was announced.
November 13, 2012 at 8:39 pm
“Stop using IRC” is maybe overkill, but it’s true that IRC has issues. It’s very good for rapid interaction, but it tends to exclude those not in convenient timezones, and it’s not good for record-keeping.
Logging the session is a partial solution to the latter, but only a partial one – the logs are often hard to follow for someone not involved in the original discussion. Ideally they’d be followed by someone writing up a summary of the discussion (including the arguments for and against) to send to the relevant mailing lists, but I can appreciate a lack of people volunteering to do that kind of work…
November 13, 2012 at 5:06 pm
We need to communicate better. I don’t think Emily is saying stop using IRC. She’s saying, start communicating better. We can start by communicating better with Emily in her comments! 🙂
She obviously sees a problem. Instead of telling her that there is no problem, or that she’s not helpful or that her suggestions won’t work, we should better understand the problem as she sees it. Then perhaps we can come up with a solution that will work for existing GNOME developers as well as those that are trying to help promote the project.
There is definitely a problem and we should definitely try to fix it.
My thoughts on the communication problem:
* People aren’t talking. We aren’t talking to each other, we aren’t talking on the mailing lists, we aren’t blogging. I think if we were to pull up traffic history for Planet GNOME or the Foundation list, we’d see that traffic has fallen drastically. If it’s moved some where else, we need to tell people.
* We aren’t doing any active press or user announcements. We’re letting others tell our story.
* We haven’t shared our vision or our roadmap for the future. Where’s the product going? What problem are we trying to solve? How are we going to do that?
November 13, 2012 at 8:25 pm
I was personally excited by the vision of GNOME OS that I saw proposed earlier at GUADEC. Now I don’t see anyone talking about that anymore. And the people I do see talking are often selfish, cynical jerks, who make up rules for their own convenience and punish people for breaking them.
I enjoyed participating in the Outreach Program, but I’m not impressed with much of GNOME’s community. My personal hope is that mailing lists, IRC, and trolling on other people’s bug reports become supplanted by other means of communication, simply because a new generation of contributors are using them to do the bulk of the work. I just see stuff like this and start to doubt whether I even want to be involved anymore.
November 14, 2012 at 4:57 am
It is a bit unfair to judge the GNOME community by the mailing lists, because as a matter of fact, a significant chunk of the GNOME community moved away from these mailing lists.
As you did, I always had excellent time with GNOME people face to face. What happens on mailing lists is something else.
Some are complaining from bad press about GNOME. The worst press I ever read on GNOME is right there, on GNOME mailing lists, written by a few GNOME people, always the same ones, sometimes long time GNOME developers. It is shockingly sad. But this should not be taken as the voice of the GNOME community.
November 14, 2012 at 2:47 am
Stormy – I’m working on fixing that problem. Developers aren’t great at doing these kind of things. Let developers be developers I say. Instead, get a team of GNOME enthusiasts or community managers deal with the community and then do the filtering. That way, instead of having to deal with this huge signal to noise ratio, they instead deal with a smaller group that understands what we are trying to do and can communicate within that context. Also point out pitfalls.
But it’s important that the project as a whole is sensitive about how we are perceived because eventually it is our brand that suffers. We used to get away with it before because most of the desktops had GNOME as the default, but since we lost Ubuntu we’ve lost a significant part of our user base since most people tend to just use the default. That means for them to switch they will need to know about us and want to switch. Bad press makes that harder.
November 13, 2012 at 7:26 pm
I think what Stormy is saying is true, but I also think that Emily is representing a marketing and user viewpoint (although she is also a clocks developer), and that these types of public discussion need to also take into account how frustrating it is to be criticized for being a GNOME developer and for making decisions. Working in the open is hard, so as we work to build better communication with our users it would be great if we could also try take into account the fact thatGNOME developers are also human 🙂
November 14, 2012 at 2:48 am
I agree. It is hard. Bunker mentality develops. See my write up above about community management.
November 13, 2012 at 8:35 pm
Hi, I speak from the user point of view. One of the things that was very frustrating about Gnome 3 was seeing the project moving towards the wrong direction and feeling powerless to do anything about it. While I’m not a programmer or a designer, I do have some ideas about user interfaces. It’s not that complicated. I have been using GNU/Linux for more than 10 years and have used many different window managers and desktop environments. I could see from the start where Gnome 3 was headed, because many of the assumptions that the UI designers were making about the users and the UI were just plain wrong. However, it was clear that they had already made the decisions and there was no way for people like me to influence them because the developers had their “vision” and were determined to go with it. I wouldn’t be that bad if Gnome was more configurable and also I think it would help getting non-technical people involved with the project. There’s people who can’t write code, but they sure can arrange panels and applets and menus the way they like, and they can share their config files or themes with others. This is also a form of involvement. But, alas, the designers of Gnome 3 also think that configurability is bad, for it goes against their vision! So, it’s really a dead end. Anyway, I’m glad that some people from inside the project are beginning to realise that something has gone wrong, maybe there’s still hope.
November 14, 2012 at 2:58 am
Ernest – I understand your frustration but designers of GNOME are not the bad guys here. Perhaps we haven’t been clear exactly what we are trying to do or the philosophy of what we are attempting.
I’ve done a lot of community outreach since 3.0 and I’ve been part of this project for 15 long years. I can point out the people bitching about GNOME were the same people bitching about GNOME 2 when it first started. We didn’t do community outreach then either and that’s why we have greater vitriol now.
GNOME 3 is in fact very configurable. So configurable that Mint was able to make a completely different desktop out of it. Wouldn’t you agree that isn’t a bad design if you can completely change just about every behaviour and even override our defaults? There are some bumps in the road of course. Extensions tend to break after reach release and users get sick of having to wait for whatever extensions they were using to get ported.
If you’re talking about exposing options, then yes there is a difference in philosophy, but we’ve held the same philosophy even during 2.0 when we declared that we were going to make usability “Just works” as the main design goal around GNOME. You ended up liking the end result although it took some time. Trust us again, because for the most part it is the same people.
We are not done yet. So, I hope you’ll continue to be part of our community and see where this takes us.
November 14, 2012 at 12:34 pm
Sri – Yes, I agree that modularity is a good design. I have to say that my criticism is directed at the user interface, not the underlying system. The problem I see is that I don’t share at all the philosophy with which the Gnome shell was created. For instance, I don’t agree that “mental models with categories and hierarchies” should be avoided. I don’t want to avoid them. For me categories and hierarchies are very useful. I prefer an applications menu that presents me a list of applications grouped by category (e.g. Graphics, Internet, Multimedia…), over a list of recently used applications or a search-based approach. I also prefer browsing a directory tree in order to find a file rather than searching for files or selecting files from a meaningless list. It’s not that I don’t trust developers. I tried to do it their way (I’m pretty open minded), it simply didn’t work out for me.
November 14, 2012 at 7:45 am
“It was clear there was no way to influence” sounds like you didn’t even try. Next time you could consider joining and bringing in your voice instead of standing at the sidelines and criticizing afterwards. It’s not that mailing lists or IRC are invite-only, and if “it’s not that complicated” (which I personally consider an underestimation), here you go!
November 14, 2012 at 10:56 am
From reading through your personal blog, you admit yourself, that it’s harder for outside people to join discussions.
I agree though that it would be nice to have some logs to see how decisions are made in GNOME.
“With regard to criticism which sometimes comes up on the transparency of decisions: It is a fact that many discussions happen in real time on IRC (or via other channels, like Google hangouts), in the timezone of the developers, and not on mailing lists only.
IRC makes it harder for interested people to follow those discussions if you live on the other side of the world or are not online all of the time. My very personal opinion is that IRC logging might help to be able to get a better understanding of the reasons why and how some decisions were taken.” (http://blogs.gnome.org/aklapper/2012/08/03/guadec-2012-discussions/)
November 15, 2012 at 9:02 pm
Mailing lists? Seriously? Mailing lists should have died over a decade ago. I don’t want to suscribe to a mailing list and fill my email account with garbage to discuss with someone who doesn’t want to listen. Forums are a much better way to make decisions, as it doesn’t have the inconvenience of mailing lists, and you don’t have to be in a specific timework to chat with the developers. More importantly, almost everyone is familiar with forums, as opposed to mailing lists. Why are we still using mailing lists? It seems as if the GNOME devs don’t want to listen to users.
November 16, 2012 at 6:44 am
“GNOME 3 is in fact very configurable. So configurable that Mint was able to make a completely different desktop out of it.”
When will you understand that not users are developers? “Configurable” means different things for a user and for a developer, is that so hard to understand? Or perhaps you understand it all too well and are just grabbing the usual fig leaf …