Insights for future development from the Guix Survey

[Steve George]

Hi,

Part 1 [0] and Part 2 [1] of the Guix User and Contributor survey's results are 
out. These cover the user-focused parts of the survey, with the contributor 
part still to come. I think the results have some great insights for where Guix 
development should/could focus, hopefully it will spart some discussion at Guix 
Days! ;-)

Here's 5 insights from each one - adoption (post 1)

1. There are lots of new Guix users: almost 50% of respondants had been using 
Guix for less than 2 years! This could have implications for how we provide 
end-user documentation and the focus of tooling to assist new users.

2. Almost 50% of adopters start by using Guix as a GNU/Linux distribution, 
predominantly as a desktop. While a third start by using it as a hosted package 
manager on top of another distro. This could have implications for how the 
project prioritises packaging and testing.

3. Adopters struggle with (a) lack of how-to's and examples (rather than 
reference documentation), (b) Guile/Lisp syntax, and (c) differences in Guix's 
approach from other Linux distros.

4. Users who stopped using Guix focused on how the complexity of maintenance 
was too high, missing packages/services they needed and issues around 
drivers/proprietary software. There's really good comments in this section 
which bring up lots of ideas for future development!

5. Overall, 67% were satisfied with their initial adoption experience. This is 
great, and the survey shows lots of ways which could improve the how user's 
adopt Guix.


Some insights from general usage (post 2):

1. Using Guix as a graphical desktop is the most popular: 73% use it on laptop, 
workstation hardware. It's used in a server configuration by roughly a third of 
users. Many users start with the desktop configuration, and when they become 
familiar they expand to server usage.

2. Proprietary drivers are used by a majority of the participants. The lack of 
all drivers being included by default in Guix was a problem for about a third 
of users. This is a much-commented area within the survey and I know it will 
draw lots of attention - worth reading through the set of views and comments, 
whereever you stand on this area.

3. Overall satisfaction with Guix was good about 70%. There's also some clear 
limiters and the same themes come up. About 50% of participants would like to 
see better runtime performance (e.g. guix pull), there's challenges with needed 
packages/services and problems with error messages/debugging.

4. In general, both in the user survey and in the contributor section there was 
a lot of focus on making the latest version of packages available (package 
freshness) and more packages (more is better) with an outright rejection of a 
smaller/more-focused package set. There's lots of potential implications for 
where the project could automate and put energy across these areas.

5. I know that all sounds bad but it isn't! There's lots of positive love for 
Guix - the comments are wonderfully uplifting - great love for the friendliness 
and hard work of everyone involved in the project - worth reading!

Being a bit self-serving (it's been a lot of work) - I really encourage 
everyone to read through the posts and scan some of the comments. I hope it 
will give you ideas for topics to discuss at Guix Days, and areas you'd like to 
focus on. 

Here's the posts:

[0] 
https://guix.gnu.org/en/blog/2025/guix-user-and-contributor-survey-2024-the-results-part-1/
[1] 
https://guix.gnu.org/en/blog/2025/guix-user-and-contributor-survey-2024-the-results-part-2/


Steve / Futurile