Skip Navigation

Linux user survey!

pad.tchncs.de

Encrypted Form

To get an idea of this community, and to try the cool CryptPad Survey feature, I created a pretty big Linux usage survey!

The data is anonymized and the content encrypted on the server. I plan on publishing the results.

Have fun!

It works on hardened Firefox on a phone, but the experience is better on a PC.


live results

Notes

  1. I am very sorry but the question "it is okay that my above message gets published" cannot reasonably be respected, as the text is just dumped into a single block
  2. Lag caused some empty questions to appear, removed
  3. A question about disk encryption and "why do you use other OS" got mixed up
  4. i changed the wording of some questions or added more options, so there may be duplicate old answers or too little new ones. You can edit your submission and update your answers.
56 comments
  • Duo to the way Crytopad works please remember to add a "Don't want to say" answer to questions with radio buttons, since you cannot deselect them meaning you can't back out of questions you don't want to answer. Hit me up if you wanna do that survey again. I'll gladly help you with structuring and formatting

  • 62 questions. I guess I am lazy.

    • I kinda escalated but that form feature is supercool, those are not really 62, they expand depending on what you select.

      For example if you use GNOME and used KDE before, it asks for reasons.

      Thats probably a bug

  • Very biased survey in general. Should have more open-ended questions and less tailoring of questions based on previous answers.

    • I experimented with the platform.

      The "anonymous" feature means I cannot do a lot of things one would normally do. For example I have no idea "users that use ubuntu and stayed with Ubuntu will likely dont care about xyz". Because there is no correlation.

      Having questions depend on the previous answers makes it possible to filter out groups of people to get more interesting data.

      And for sure the questions are biased, try to do an unbiased survey that is still interesting or funny. But I tried my best.

  • What is the difference between "Versioned" and "Stable"? And which one is NixOS?

    • I am not sure haha, I think versioned means there is a certain amount of stability but closer to upstream, while stable is like "I run Kernel 5.14 and do my patches myself".

      It is the same but with a slight difference in how extreme it is.

  • So I am going to try and give some feedback. I think I was a little harsh at first but that was because I found this to be upsetting.

    I think it is inappropriate to ask about income. I know that might be a useful metric but I think it is touchy subject. Next, do not ask if someone feels represented. That is a vague question and needs to be rethought.

    For the survey itself I think it is way to long. Try breaking it into smaller pieces that are spread apart. Surveys should be around 4 simple well worded questions. With the smaller surveys there is less data combined so it would be much harder to single out individuals. It also is a time saver.

    For the "special brain" question you should instead ask about disabilities. This is a dangerous question as if a person hits do not answer that is a data metric. This question should also probably be in its own survey.

    For gender I think that isn't all that big of an issue. However, the fact that only non males matter is problematic. You are singling out people who may not want to be singled out. This also applies to disabilities or any trait that could put them be used against them.

    I really think you should break these surveys down. Also I would have a look at the surveys by the Linux Experiment. He did a good job of doing them in a way that isn't discriminatory. He also didn't ask for information such as disabilities.

    Lastly, you should offer a other button with a custom response. Sometimes we do not fit in boxes.

    • Thanks, I will read your reply.

      it would be much harder to single out individuals.

      It is all anonymous, I dont get any individually sent data. Which is pretty problematic as it would be interesting to combine certain types of people, traits, behaviors etc.

      For example "people using Ubuntu back then and staying with it tend to use X11" or whatever.

      This question should also probably be in its own survey.

      Yes probably, multiple smaller ones make way more sense.

      This also applies to disabilities or any trait that could put them be used against them.

      As the ratio is so extreme, I disagree that focussing on just non-male people is wrong.

      It is a scope though, and you cannot make a fair and representative survey.

      But I agree that other disabilities could have been added.

      You are welcome to join, I and maybe a few more people plan on doing more, smaller surveys in the future.

      have a look at the surveys by the Linux Experiment.

      Yes but I didnt find it that interesting. We can argue about discrimination, but in an anonymous survey it is not problematic to ask for "sensitive" info and tons of people gave their answer there.

      These just interested me, and they are more personal than Nicks for sure.

      Lastly, you should offer a other button with a custom response.

      Yeah for sure. I want to make a followup especially on the "other" boxes that got tons of checks.

      But CryptPad seems to not have that feature.

      As a test of this feature I think it is interesting. I will try Limesurvey and Nextcloud forms the next time.

56 comments