Idea: Yearly NixOS Community survey

Development
1

Some time ago i had the idea to create a survey about Nix(OS), so we better understand which problems should get prioritized and where is potential for innovative features. It would also help to understand how the Community uses the software to better promote it.

Some quick ideas what we could learn:

  • Which desktop environment does most users use on their main NixOS machine? (so we could focus to improve it and maybe make it default in graphical ISO)
  • What OS did you use before NixOS?
  • Where do you use NixOS? Desktop/Server/At Work/Embedded/???
  • How would you describe yourself? Developer/Administrator/Linux enthusiast/Student/Scientist/…
  • What do you like most about Nix/NixOS?
  • What is the biggest problem Nix/NixOS has?
  • Do you have any idea to improve Nix/NixOS?

But also a path like: Do you use NixOS? → No, Have you tried NixOS? → Yes, What was your experience trying NixOS?

The survey should be anonymous and the data released as open data and the results released under an open license. It should also be short, like 5 minutes.

What do you think about this idea?

Would you participate in such a survey?

Would you like to work on the survey and evaluation?

I would propose to create a working group for this tasks.

Next steps could be:

  1. find people to work on this
  2. evaluate survey software (TellForm looks very modern and user friendly, LimeSurvey is popular, formr might be interesting)
  3. collect questions that can give useful insights, also over time
  4. create the survey, find good time to get many participants (maybe shortly after a release to also get people who just tried it out?)
  5. do the survey, promote on all channels (IRC, Discourse, Twitter, Reddit, …)
  6. evaluate the results und publish them
  7. goto 3. for next year

We should also talk to other linux distros or projects how they do surveys. Maybe someone from academia has experience with survey design and evaluation.

(right now i struggle from burn-out and have decided to only care about my health, so the earliest i might be able to actively work on this is next year. but the idea keeps coming back to me, so i want to share it and maybe others would like to work on it)

14 Likes
2

Hi,
IMO since, even today, even people involved in FOSS and aware of
tech trends still use or at least have a GMail account for their
Android portable device… Well a Google docs survey is the
quickest “web point&click style” option. People who take the
survey do not need any account.

Another option is simply post here (via ML mode) and analyze the
result, with enough disciplined users it can be automated in few
minutes, perhaps rediscovering and sharing a bit of “text tools”
experience too many “youth user” do not know even that’s exists…

However I often notes that a distro is NOT a commercial product
with a seller and few buyers but a project made by some people
for themselves that may turn out good for others that use it
sometimes without contribution, sometimes with valuable or
minor help. So “focus” IMO is what contributors needs and for
that there is no survey but simple personal knowledge and
discussion… People who do not want to take time at least lurking
casually post are simply not useful for the community…

Anyway if you or someone else make the survey I’ll participate,
if I can help otherwise I’m here. For now my answers are:

  • DE :: EXWM for me, Gnome SHell for few desktop users I support

  • OS :: my timeline:
    Irix - my first “PC”, an old dismissed O₂
    FreeBSD - first modern OS on x86 desktop
    Mandrake - first laptop
    RH - tired of Mandrake
    Debian - tired of RH bugs and absence of pkg mgmt
    Gentoo - tired of APT bugs
    FreeBSD - tired of Linux 2.6.x and general evolution
    OpenSolaris - because it was the best OS out there
    Ubuntu - because after osol FreeBSD was not much usable
    on modern crappy desktop hw
    NixOS - because after Unity I have no reason to remain on
    a buggy mess that’s is Ubuntu now (and in general more and
    more upon any new release)

  • Where :: personal desktop, home server, few supported desktop

  • Myself :: admin

  • Like :: IaC / a simple, reproducible, text config

  • Problem :: documentation

  • Idea :: in my priority order

    1. documentation
    2. few derivations improvements
    3. other OS support (OpenBSD, FreeBSD, DragonFlyBSD)
    4. more Emacs integration
    5. development platform changes

Point five is essentially avoid depending on single proprietary
platforms, no matter how friendly and known they are. For instance
I have nothing against GitHub as a simple “external public storage”,
an official mirror like also GitLab, … but not being tied to their
proprietary functions, starting from PRs. I know developer have
chosen to ditch mailing-list for Discourse and most are happy with
it, but at least discourse is “self hosted”, not too much a third
party services.

IMVHO these days distro should teach newcomers free tools even at the
price of less contributions. FOSS is more in danger than ever for
tons of reasons, many involving the most common platforms people know
and use everyday.

For instance, just in topic of desktop usage: in the past ten years I
succeed, even with frictions and hard work, to push “enterprise
GNU/Linux desktop for generic users” to a certain extent. Starting from
the end of 2018 a bug after another on proprietary software or due to
few companies decisions I have for the first time in more than TEN years
hearing people asking me “what if we came back to Windows?”. For more
than ten years I do not see much hw-related problems, since few years
crappy graphics on laptops and even desktops start to be again an issue,
even more than 15+ years ago. Support for crappy IoT/mobile spamware also
became a problem (like “hey how I can sync my spyphone contents on my
shiny FOSS desktop?!”). As a result while I see less and less “generic
users” I see also that most techies start to go for “back to basic” or
tiling WM etc rediscovering some “old” tools and discovering that modern
environments in the end are unuseful, unproductive monsters. I consider
generic users a lost cause but at least push good tools for newcomers
willing to learn is IMO a “social good thing” to do.

Sorry for long post and bad English.

– Ingmar

2 Likes
3

“youth user” here, 1st computer was an eee pc 701 running Xandros lol. replaced wm with iceWM and installed backtrack 3 as soon as i saw Xandros’ gui.

Desktop, home server, 2 routers, 2 laptops, + various local VMs and ec2 machines 100% nixos now, if there is an Xserver it is running xmonad.

problem: I can’t run nixos on my phone yet… I forgot how to manually manage systemd units in Arch… mutating etc files on other systems feels like a criminal act…

5 Likes