this post was submitted on 13 Mar 2025
156 points (100.0% liked)

Linux

6615 readers
446 users here now

A community for everything relating to the GNU/Linux operating system

Also check out:

Original icon base courtesy of [email protected] and The GIMP

founded 2 years ago
MODERATORS
 
you are viewing a single comment's thread
view the rest of the comments
[–] [email protected] 76 points 1 week ago (15 children)

Most Linux users disable Telemetry, so I think it's probably missing around 10%

[–] [email protected] 5 points 1 week ago (14 children)

For the fuck of it, I asked AI to guesstimate the market share

Based on the Mozilla Graphics Telemetry data provided, Linux accounts for 3.4% of the sampled Firefox sessions. However, this is likely an underestimation of the true Linux market share for the following reasons:

  • Telemetry Opt-Out: A significant portion of users may disable telemetry for privacy reasons. It's reasonable to assume that Linux users, who often prioritize privacy and are more technically inclined, might be more likely to opt out of telemetry compared to Windows or macOS users.
  • Pi-hole and Ad-blockers: Tools like Pi-hole and ad-blockers, which are also more commonly used by technically proficient users (a demographic that overlaps with Linux users), can block telemetry data from being sent.

To guesstimate the true market share, we need to consider a correction factor to account for these missing Linux users in the telemetry data. Unfortunately, precise data on telemetry disablement rates per OS is not readily available in the telemetry data itself or through a quick search.

However, we can make a reasoned guesstimate. Let's assume:

  • Baseline Telemetry Rate: Assume that for Windows and macOS, the telemetry data is relatively representative, and a smaller percentage of users disable it.

  • Higher Opt-Out on Linux: Assume a significantly higher percentage of Linux users disable or block telemetry. Let's arbitrarily assume that twice as many Linux users disable telemetry compared to the average across Windows and macOS. This is a rough guess, but it acknowledges the higher likelihood of Linux users opting out. Applying a Correction Factor (Illustrative Example): This is a simplified example for illustration. To be more accurate, we'd need real data on telemetry disablement rates.

  • Assume a hypothetical "average" telemetry disablement rate across Windows/macOS: Let's say, hypothetically, 10% of Windows/macOS users disable telemetry.

  • Double that for Linux: Based on our assumption, 20% of Linux users disable telemetry.

  • Calculate the "visible" Linux user proportion in telemetry: If 80% of Linux users enable telemetry (100% - 20% disablement), then the 3.4% in telemetry represents 80% of the actual Linux user base within the Firefox user population.

  • Extrapolate to the "true" Linux market share: To find the 100% (true) Linux market share, we can do a simple proportion: If 3.4% represents 80%, then X% represents 100%. X = (3.4% / 80%) * 100% = 4.25%

Therefore, based on this highly simplified and assumption-based guesstimate, the true market share of Linux might be closer to 4.25%. Important Considerations:

  • This is a very rough estimate. The actual percentage of users disabling telemetry on each OS is unknown without specific data from Mozilla.
  • Firefox User Base vs. Overall Market Share: This telemetry data is specific to Firefox users. The overall operating system market share might be different, as Firefox usage itself might be skewed towards certain operating systems.
  • "True" Market Share Definition: "Market share" itself can be defined in various ways (desktop OS, all devices, etc.). This guesstimate is within the context of desktop/laptop users who might use Firefox.

In conclusion, while the telemetry data shows 3.4%, a more realistic guesstimate, considering telemetry disablement, could be in the range of 4% to 5%. To get a more accurate figure, one would need access to data on telemetry opt-out rates per operating system or look at alternative market share data sources that employ different methodologies (though these also have their own limitations).

[–] [email protected] 1 points 6 days ago* (last edited 6 days ago)

Don't listen to the haters but it would have been nice if you collapsed this because it is very long and generalized to the point that it is pretty much an eyesore. Plus most people can ask their AI of choosing semi-random topics. I don't see what was interesting in the AI response at all. It states some blatantly obvious facts and is rather too wordy. I intentionally include into the system prompt or "personalization" about how I like things to be kept short and to not reiterate what I had posted especially if it just sounds like the "AI" is thinking out loud.

load more comments (13 replies)
load more comments (13 replies)