The Survey That GNOME Would Rather Ignore

Off the Beat: Bruce Byfield's Blog

Oct 27, 2011 GMT
Bruce Byfield

As you may have seen, the Phoronix site is hosting a private survey about GNOME. The survey still has several weeks to run, but, so far, neither the circumstances surrounding the survey or the replies show the GNOME project in a favorable perspective.

The survey was begun by Felipe Contreras, who first raised the idea back in July on the GNOME desktop-devel mailing list. "Lately I've [been] feeling that there's a lot of dissatisfaction with GNOME 3," he wrote. "Why not find for good what people are thinking with an user-survey?"

Conteras spent considerable time in the past few months refining the questions in the survey, sparking long and sometimes heated discussions. When the GNOME Foundation wouldn't host the survey, Phoronix agreed to do so instead.

Some of the mailing list participants, including Alan Cox, responded with helpful criticism and support. Naturally, too, the discussion included the immersion in details and bike-shedding that is typical of the free software community.

However, reading the various threads, you can't escape the suspicion that many of the responses from GNOME developers were deliberately obstructionist, providing reason after reason for rejecting the entire idea of the survey without offering any possible alternatives.

For instance, when one developer said that the survey still needed improvement but failed to give any concrete suggestions, Zeeshan Ali lectured Conteras that, "You are the one who proposed the idea of this survey and (more importantly) insist that a useful survey is possible for GNOME so *you* must address all criticism/concerns if you want us to take your proposal seriously."

With even more hostility, Josselin Mouette wrote, "Well, here’s a suggestion: since nobody knows how to address the correct target population or how to interpret the results, I suggest to spend our time fixing bugs instead." Similarly, Christian Fredrik Kalager Schaller announced a mock award was to be given to Conteras for "endurance." By September, Conteras was writing that he was "effectively banned from desktop-devel," since his emails were taking two weeks to be moderated.

In other words, the discussion was, as Alan Cox put it, "in blocking mode," In another message, Cox referred to the logo of Miguel's Izaca's old company Ximian, one of the major early developers of GNOME, commenting that, "Some days I think Miguel got the Ximian monkey dead right, except that there should have been three of them" -- presumably, see, speak, and hear no evil.

That Cox successfully summarized the mood of many of the responses is indicated by Christophe Fergeau's comment early in the discussion that the survey would be unrepresentative and would only result in people saying, 'upstream [GNOME] ignores their users!' -- to which Richard Hughes added, "To upstream maintainers, even a 'user survey' telling us to do X, Y or Z is probably going to be ignored. Users don't know what they want."

As you look through the discussion of the proposed survey, you can only conclude that GNOME developers' strategy -- officially or unofficially -- was to ignore criticism of GNOME 3 and to make sure that they never heard it, all in the hopes that it would go away.

Not a lot of love

If that was in any sense the strategy, GNOME was right to be wary of the responses. On Tuesday, Phoronix reprinted the first eight thousand comments in the survey. While some of the comments included stock phrases like, "Keep up the good work!", the lack of satisfaction with GNOME 3 was overwhelming.

If my sampling of a random hundred comments is in any way representative, then some 75% of survey takers -- even those who view the GNOME 3 series favorably -- have lists of technical improvements they would like to see in GNOME, including classic menus, icons on the desktop, and more and better customization and documentation. Some 25% view GNOME 3 with hostility, while, even by the most generous count, only about 13% are enthusiastic about it. 15% requested a return to the GNOME 2 series or the fallback mode, its equivalent in GNOME 3, and 3% talked about migrating to another desktop like KDE or Xfce.

Some of the hostile comments:

  • Gnome devs have previously been accused of not giving a shit about their userbase. Gnome3 is proof. Ya I'm bitching, but I was a gnome user for 4 years up till the 'switch'...Seriously, Vista was less horrible than Gnome3. Thanks a bunch for 'removing a choice' from the linux ecosystem.
  • Stop assuming that users are retarded.
  • GNOME 3.0 (gnome shell) was released way too incomplete. Even 3.2 is incomplete and /barely/ usable on a daily basis without significant tweaking and adding of various extensions. This is hurting your image.
  • Overall, GNOME seems to concentrate on forcing an increasing number of stupid ideas on people.
  • Please listen to your users.

By contrast, few of those who said they liked GNOME 3 elaborated on the statement, or expressed any enthusiasm. In fact, those who said they liked GNOME 3 were about as likely to list technical improvements as those who were disliked it. In general, the level of enthusiasm didn't equal the level of dis-satisfaction, let alone of hostility.

Some Conclusions Become Inevitable

Results like these can be explained away with all sorts of time-honored rationalizations. You could say that only those with a point to make are likely to respond, and that survey takers are unlikely to be representative of GNOME users as a whole. You could say, too, that neither the methodology nor the questions in the survey are up to professional standards. Each of these rationalizations would have a kernel of truth, too.

However, so far, at least, the numbers aren't even close. As a result, it's hard to see how any rationalization could completely explain away either the dis-satisfaction or the hostility. Even allowing for generous margins of error, both seem undeniably real. In fact, their extent is surprising, even to someone like me, who has been voicing some of the specific complaints (and hearing them) for months.

Even worse is the response of what Cox calls "the GNOME oligarchy." From the evidence of the mailing list, GNOME leaders seemed to have suspected what the response would be, and done their best to obstruct and stonewall Conteras and his survey rather than face the truth. This kind of behavior is more what you would expect from the executives of a proprietary company rather the guardians of a large free software project. As Cox observes,

"Blocking a volunteer off [from doing] stuff and see what happens be it code or otherwise is usually the wrong thing to do." It's not in the best tradition of the community, and can only make the situation worse.

Maybe the final results of the survey will be different from the preliminary ones. However, I can't imagine that they will change very much. And already the conclusion is inescapable: GNOME 3 is a problem for most users. Moreover, evading the fact doesn't work.

The only questions now are whether the project will admit the unavoidable facts, and what plans it can develop to salvage a situation that should never have got so far out of control in the first place.


  • gnome will never learn

    It does not matter how bad gnome becomes. As long as Redhat (and others) push it wherever they can, no matter what, the gnome devs will continue to create a worse solution. There is no drawback. So why change anything? Lets ignore the users even more. They don't even know what they want.

    I do know exactly what I want. And because of that knowledge gnome 1.4.2 was the last usable release for me..
  • Not to offend anybody...

    Not to offend anybody...

    I would like to summarize what is going on with GNOME with the analogy from science. I hope this view will not offend anybody as it is the way things are in nature and you can't get angry at the truth. I'm not going to lecture anybody on evolution since I think Darwin did a great job in the first place. But it seems that this is exactly what is happening right now. GNOME 3 team changed the environment too quickly and perhaps for the better but as a result a few million dinosaurs died-off. What is left of the user base is either very agile users that perhaps just don't care which DE to use and will switch to something else in a heart beat or brand new users that are very very new to the whole Linux experience or simply put don't know any better. User base that depended on GNOME 2, the devoted followers that praised everything GNOME 2 has been killed off by the experiment. I'm one of the dinosaurs that just can't afford to wait for the environment of GNOME 3 to grow on me. To me GNOME 3 is simply brand new non-GNOME DE that I will never like, so I have to look some place else. And thing is I'm not even angry that GNOME 3 group did what they did. What I'm angry about is that they kept the name of GNOME. What they should have done is said that we are discontinuing GNOME and starting brand new desktop called “blah” that is nothing like the old GNOME. I think the world would understand and would be less hostile towards the new “blah”. Keeping the name GNOME seems like a cheap trick to try to keep the old user base and being endorsed into large and stable Linux distributions.
  • Re: are there any useful conclusions?

    "think that's what many people were afraid of when they were rejecting the survey. You just cannot make such conclusions on such surveys. "

    Yes and no. If the percentages were close, then you would be rash to reach any conclusions from them. But when the percentages are so far apart than even a large margin of error can't diminish them, then it becomes perverse to ignore them altogether.

    Or, to put it another way: You can't draw any fine-grained conclusions from such a survey, but you can see some broad tendencies.
  • are there any useful conclusions?

    "And already the conclusion is inescapable: GNOME 3 is a problem for most users."

    I think that's what many people were afraid of when they were rejecting the survey. You just cannot make such conclusions on such surveys. GNOME targets broad group of users. Users that care are relatively small group, I'd say 1-5 % of all users. My mum or dad will never take part in this survey neither will thousands of other users. But they are GNOME users as anyone else. So the actual problem is that if GNOME devs accept opinions and wishes of the small, but vocal user group it won't necessarily mean they will do something most users want. I think the only reasonable feedback would be usability tests on a neatly chosen group of people that will more likely represent GNOME user base. Something that Canonical has been doing for a while.
  • Phoniphication...

    I've been reading and hearing people complain for years that Linux had to stop trying to imitate Windows and should come up with something better.
    I suppose the Gnome leaders decided it was time to try it.

    Well... trying to stop being Windows, to switch over to trying to be an iPhone was NOT the solution.
    Especially a phone without a touch screen...
  • Re: Not such a good survey

    Corky Oct 31, 2011 12:05am GMT:

    "After reading the blog I thought I would participate in the 2011 GNOME User Survey. I was disappointed. Most of the questions are asked about Gnome without any differentiation between Gnome versions."

    With no differentiation, I would assume that the survey refers to the current release series.

    "Thank you Bruce for bringing the survey to our attention. Your observations about the Gnome developers' responses was interesting. Perhaps you could have offered a little critical judgement on the survey itself."

    Let me make up for that omission now: The survey could be much better worded, and ask more penetrating questions. To that extent, some of the comments about the survey are entirely correct.

    However, I think that the value of the survey doesn't lie in the specific questions, and probably wouldn't do so even if the questions were better chosen and worded. Instead, the survey matters because of what people are saying in their own words.
  • Not such a good survey

    After reading the blog I thought I would participate in the 2011 GNOME User Survey. I was disappointed. Most of the questions are asked about Gnome without any differentiation between Gnome versions. For example:

    04. Does GNOME do what you want?

    I am unable to answer this question properly because Gnome 2 does what I want but Gnome 3 does not.

    Thank you Bruce for bringing the survey to our attention. Your observations about the Gnome developers' responses was interesting. Perhaps you could have offered a little critical judgement on the survey itself.

    Sorry Felipe, you have not really provided an effective means for users to provide useful feedback about their experiences with Gnome 3.
  • Just a lowly, humble, common user...

    Ladies and Gentlemen,
    I am just an average-joe user who doesn't usually follow these blogs, but when I saw something about Gnome 3 I got curious. I'd like everyone to know I am grateful for the survey (which I stumbled upon by chance) and the opportunity to provide feedback to make things better, because Gnome 2.32 was just about perfect for me and my family, and now it sure sounds like the powers-that-be are going to take a great thing away from me and expect me to accept a very troublesome replacement. Maybe the feedback sounds hostile because the experience was "that bad." Don't take it as hostility...I'm just letting you know. You are trying new cutting-edge things out and I am giving you feedback. That should be a good thing; that's the way it is supposed to work. So here goes: I thought Gnome 3 was going to be really "cool," and I was lured by the promises on the web site. But the Gnome 3 user interface and functionality was painful enough that after a few weeks of it I was compelled to go on a week-long search for other desktops and other Linux versions. This made me even unhappier because of the time and effort involved when I should have been doing other things. I finally fell back to Gnome 2.32 on Fedora 14, where I will stay as long as I can or until Gnome 3.X or X.X--or any other worthy desktop-- makes my life (and my family's) better instead of worse. Whether the survey is precisely "scientific" or not, I think overall you are getting good honest and useful feedback.

    Sure sounds to me though that there is some defensiveness-- someone feels attacked, maybe by rival developers or overbearing critics, or people delighting in someone else's stumbling. I think "Hostile" is not the best label for most of the critical feedback partially summarized in this forum. I think the vast majority are not intending an attack to cause damage--rather users (like me) probably feel harmed themselves by having their excellent, very useful and productive version of Gnome taken away and replaced by something completely different, with functionality we've come to depend on missing, and an inefficient interface. So maybe we are just a little shocked and "agitated." Reminds me of when Windows Vista came out. For me it was a bad experience.

    I think developers should keep a longer-term view and listen to feedback, or perhaps they won't have much feedback in the future.... users may all have fled. Run MORE and different surveys rather than fretting over making a single perfect survey at one point in time. As for Gnome 3 in particular, the developers can do what ever they want as they see which cutting-edge capabilities work, and which don't..and so can users (i.e, we can go elsewhere). It would be a shame though if the Gnome developers don't fix the way things are now, because to me, they had a really good thing going in Gnome 2.32.

  • Misinterpretations?

    Richard Hughes wrote:

    "The "quote" misinterprets my email completely. "

    Yes, I could have more summarized the rest of your comment -- the gist of which is that users don't always know what is required when they ask for a feature. But that wouldn't change the main point that many of the postings, including yours, tended to discourage the whole idea of the survey.

    Granted, it wasn't clear whether you were expressing a personal opinion or making an observation. But that doesn't change the general effect of the post.

    Zeeshan Ali wrote:

    "Thats quite a misrepresentation of what happened. My email came after a long discussion (there is even v4 in the title of the thread to give you an idea) between GNOME developer (myself included) and Felipe. We all tried to explain to him how a *useful* user-survey is not possible and unless all concerns presented were properly addressed, there is no point in conducting the survey. Felipe's insisted that all those concerns can be addressed but failed to inform us how. He seemed to have been under the impression that we share this baseless optimism with him and therefore should help him address those concerns. All I did in the quoted email here was to inform him about this fact that he continued to miss throughout the discussion."

    Few people actually gave concrete suggestions about how to improve his drafts. When people act that way, it certainly gives the appearance of wanting to obstruct instead of help, and that is the impression I get from reading the threads -- and, yes, I did read all related ones in the mailing list.

    Incidentally, I should point out that an interpretation with which you disagree is neither automatically a misinterpretation or a misrepresentation. Here, it's a difference of opinion and perspective, or, perhaps of emphasis. True, the impression that I came away with may be different from what participants consciously intended -- but that's a long way from justifying the implication that I deliberately distorted things. Not being in anyway psychic, I can only interpret from what is said.
  • Bruce, I expected better..

    The "quote" misinterprets my email completely. Please read in full before just using headline grabbing quotes. I really expected a better quality report than this from Linux Magazine...
  • "Zeeshan Ali *lectured* Contreras"

    Thats quite a misrepresentation of what happened. My email came after a long discussion (there is even v4 in the title of the thread to give you an idea) between GNOME developer (myself included) and Felipe. We all tried to explain to him how a *useful* user-survey is not possible and unless all concerns presented were properly addressed, there is no point in conducting the survey. Felipe's insisted that all those concerns can be addressed but failed to inform us how. He seemed to have been under the impression that we share this baseless optimism with him and therefore should help him address those concerns. All I did in the quoted email here was to inform him about this fact that he continued to miss throughout the discussion.

    Oh and btw, *I* asked Felipe to bring this idea of survey to the mailing-list, in hopes that he'll be reasonable enough to listen to GNOME developers on why exactly they are not interested in the idea of a survey but obviously I was wrong.
  • The real reason behind the curtains...

    I think that the real reason behind gnome developers' attitude is a deep fear of a massive reaction from the users against the new workflow concept of the Shell. There are users hating GNOME Shell and users that like GNOME Shell - like me -. But even between the users that like the Shell, we are many that we don't like some decissions or we think that GNOME Shell needs maaaaaany improvements or changes. But there isn't a "medium point" between "lovers" and "haters".

    Like most humans do, we usually dislike situations where many people tell you that your child is ugly big-smile. If they admit that their ideas are "wrong" from the view of a big percentage of users, they think that their "battle" wold be lost, and the dev egos would suffer.

    "Normal" users are very conservative people in tasks involved with computers... As an example, two years ago i performed a migration from Windows XP to GNOME and KDE desktops in a programmers shop; admin people liked GNOME simplicity over KDE, but they felt more comforrtable using an interface that resembled Windows XP - bottom task bar with start menu -, and most programmers preferred KDE for the same reasons and the better integration with subversion - SVN -. All said that GNOME ran "lighter" than KDE, but preferred the known model... except two Mac users that remained with GNOME big-smile.

    The new model really rocks if you run tablets or netbooks, but in tasks concerning many programs running and multiple windows for each program the new workflow is worse than the old GNOME way... With the new workflow you need to have tab support in your apps if you want to be productive, a solution that many users don't like and many programs don't support well - or don't support it at all -. This is a BIG problem for some people, like programmers or system admins that must check many services and servers - i was one of them blunk -. Try to check quickly the logs of 4 or more servers in GNOME Shell and tell me how good it was without using tabs - and searching the "correct" terminal tab if you use them - big-smileDDDDDD.

    With this new workflow we have another interesting "feature" that would be productive or a "nightmare" for users: you must be thinking in ONLY one app at the time; GNOME "forces" you to adopt the mental process of "first I must do this, and do that other task after this one"...

    Imagine that I'm developing a web page, and todey I must implement some Javascript events and the design of their related "action buttons" in the page; because changing between apps is a bit onerous in time - click on "Activities" or use "Win" key, search the other app window , open the window and wait to have it on screen -, I tend to do all the design before and all the coding after big-smile... This situation is an example of "forced" behaviour in the user.

    GNOME devs should learn from KDE folks; when they implement the new "start menu" with "sliding" categories and the new "icon containers" - happy - desktop , they left the option of a classic menu and a classic desktop for the "old-style" users, and after two or three vesions most people liked and used the new menu adn desktop model. In the GNOME side, there is no "reasonable" migration path from the "old style" workflow to the GNOME Shell workflow, leaving the users with a "take it all or take nothing" decission.

    And most users decide to leave the GNOME ship, like many did inthe KDE 4.0 fiasco... But the worrying thing is that many of the lastest went back to KDE when the 4.x versions were more complete and stable. I think that today's GNOME Shell doesn't allow the same for "lost" users...
  • changes??

    not all changes are made equal. enough with this nonsense. both a person getting a haircut and another undergoing a leg amputation are experiencing "changes". one is good (hopefully) and another is bad. that's how it is with changes.

    so there's no point comparing "the changes" in KDE and those in Gnome. i use neither on my main, work computer btw but KDE 4 was a good concept that was broken in implementation, and has since stabilized. I have it on a spare notebook and am impressed with its features and also a lack of bloat - when put on top of pure debian.

    as for gnome shell and unity, they are flawed in their very essence, not in its implementation. unless both concepts and designs are actually abandoned for something else, i don't see how the mere passing of time and minor fixes will make any difference - unlike with KDE.

  • I moved to KDE

    For me the story went something like this... When I was reading blog Planet GNOME posts about GNOME 3 before it was released I was quite excited. But then the release came out and I tried using it for about half a year. I just couldn0t get used to it. It was always getting in the way and it was missing way to many features. So after Installing the new version and seeing that it hasn't improved I just decided to try my luck with Ubuntu Unity. Unfortunately I found Unity even worse than GNOME 3. At this time I also did what was quite unthinkable for me: I decided to give KDE a try again (I tried it when it was in version 3 and I hated it back then). To my big surprise, KDE 4.6 worked like a charm for me and I even liked the look of it a lot. After a couple of months I really got used to KDE and discovered a lot of new things which it offers and now for me it is the best desktop I have ever used. The desktop is very flexible and it is easy to make it work just as you want it. Even for KDE apps it seams they each offer something more than GNOME counterparts. And the KDE community is also very nice and responsive, it feels quite a bit more friendly and social compared to GNOME community. What can I say, I found a new desktop I love and I don't see myself coming back to GNOME anytime soon. Unless they start to listed to their users and change GNOME in a big way.
  • Community vs Corporate Upgrades

    @Luka I suspect the reasons for these changes on the Windows side (Metro desktop, Ribbon panels) are somewhat related to the 'new and shiny' used to sell more software when the old software likely worked fine. Behind these new interfaces there are changes in the code and to stay interoperable with new MS software, everyone else must get the new MS software also.

    On the Community side, however, there isn't nearly as much money in the KDE or GNOME desktops. They work and either you like them or you don't. Having these developers force these changes on us when there isn't money being made, just frustration with the FLOSS 'release early release often' model, gets people up in arms.

    Note the Gimp changes to have a Single Window Mode was a community-requested change for a long, long time that was ignored by those developers; now its implemented but its an optional setting — the best of both worlds.

    I personally love KDE4 and didn't much like GNOME2, but I've not tried GNOME3. I can't imagine that more simplification is something I want. I took the GNOME survey (they wanted anyone), and here were my suggestions

    @Bruce thanks for the report, I'll have to check out the results!
  • The Survey That GNOME Would Rather Ignore

    I believe that Gnome 3 may take more time than Unity. It will gain a lot of ground when it's finally ready.
    I mean seriously, Gnome has been doing this for far more years than the Unity scene, and they will adapt or wither away, which I don't see them doing. Just my opinion.
  • Phoronix

    I've been reading the first 100 or so user comments on the Phoronix article, and my impression is that the general tone of the comments is much more positive than the author of above article made me believe.
  • On changes

    There is a movement in software industry away from known ways of doing things like we did for the last decade or two. KDE moved to 4, Gnome moved to 3, Microsoft to Ribbon, browsers are ditching everything but the web page, GIMP will be available in single window mode and look what will Windows 8 bring. Changes used to be a good thing and we, the users constantly expect them. So why this hostility when a change is introduced? There was a huge move of users from KDE when they released version 4. It should be expected the same for Gnome with this major change. But most of the users adapt. The computers from a couple of years ago are perfectly capable of running Gnome3 (but not for Windows Vista or 7). For me the Gnome 3 is a move in the right direction. I am using it for work and for fun. I can say I am a bit more productive (let say 10%, but that is hard to measure). I use it for programming, web design, photography and office work. There are some features that I miss, but not one that would block me from using Gnome3.
    Give developers some more time. Surveys are helpful and on some level the developers will eventually use the results that are constructive. When they have the time and resources. Come on, KDE became useful again in what, 4.4? As for those that do not agree with Gnome, there are always alternatives. This is Linux, is it not?
  • Stay The Course

    GNOME Devs, please stay the course. gnome-shell is an excellent interface. I can find and launch any program with a few keystrokes, the overview mode does a better job than the task list for finding windows, and I do not miss the icons on the desktop (they are always obscured by windows, and apps and the file manager are not hard to reach anyway).

    The workspace management is leaps and bounds ahead of what I knew with Compiz.

    I reiterate. Stay the course, and consider folding some of the excellent extensions out there into the default package (I've grown accustomed to your policy of sane defaults he he). One I would strongly recommend is the "native window placement" extension.

    This is the most seamless experience I've ever had with a Linux desktop, and I've been here for 6 years.

    Thank you.
  • On GNOME 3

    I don't think the survey said anything that wasn't already known, there is a vocal group of people who dislike GNOME 3. There was similarly a vocal group of people who disliked GNOME 2. Usually because it didn't permit them to tweak some particular setting they'd come to love from something that came previously. People either move to something that provides what they're looking for, or the community creates tweaks, alternative docks, and whatever else to let people customise their desktop the way they want.

    GNOME is never going to be everything to everybody. For 10 years now, we have been focused on making there be one way to do something, and have that way work well; rather than six preferences, none of which work at all. GNOME 3 continues this, but is now also very designer driven. Personally I think this is a strength. It should be our goal in free software to evolve and explore new ideas, not just to clone the ideas coming from the non-free world.

    Some of the complaints and missing features will be addressed. Unfortunately the project only has finite resources, so you have to focus on the big things first.
  • Hostile comments

    Hey, Bruce -- First, excellent article overall. But -- and you knew that was coming -- under the hostile comments you have, "Please listen to your users." How can a comment be hostile if they're saying "please," to say nothing of the basic sentiment of listening to users?

    Larry Cafiero
comments powered by Disqus
Subscribe to our Linux Newsletters
Find Linux and Open Source Jobs
Subscribe to our ADMIN Newsletters

Support Our Work

Linux Magazine content is made possible with support from readers like you. Please consider contributing when you’ve found an article to be beneficial.

Learn More