I hate when people say “oh we had 100% growth!” Ignoring the fact they went from 4 to 8 clients. Meanwhile the one that only shows 0.5% growth considers 8 people a rounding error.
There’s liars, damn liars and statisticians.
I hate when people respond to a post with a little anecdote that is completely irrelevant to the original post.
I’ll feed the troll… it’s very relevant with the bottom part of the graphic. It’s touting that it’s got the biggest increase of traffic change at ~32%, however overall it still only has a 3.2% share. That ties in exactly what what I’m stating and the xkcd that was posted too. There was nothing anecdotal or that was irrelevant in my post.
Ignoring the fact they went from 4 to 8 clients.
You think the 3.2% is on the order of single digits of machines? You think 3.2% market share is 8 people?
Obviously you don’t. It’s 10s of thousands of machines and you exaggerated the actual situation so far it no longer made any sense.
Okay, so here’s some more easily interpreted numbers. In addition to having the largest change in traffic share relative to its previous share, Linux actually had the largest change in actual traffic share. It was just BARELY above Mac OS. If more significant digits were allowed here, Linux rose 0.0085 while Mac OS rose 0.0082.
| OS | 2023 Share | 2022 Share | Share Change |
|---|---|---|---|
| Windows | 0.632 | 0.647 | -0.015 |
| Mac OS | 0.292 | 0.284 | 0.008 |
| Linux | 0.036 | 0.027 | 0.009 |
| Chrome Book | 0.029 | 0.026 | 0.003 |
| Other | 0.011 | Unknown* | 0.005* |
2022 percentages computed as:
share_2022 = share_2023 / ( 1 + relative_percent_change )
and percent change computed as:
absolute_percent_change = share_2023 - share_2022
* The relative percent change for “other” is not reported in the graphic. There is a remainder of 0.005 in the Share Change which is, presumably, attributable to change in the “other” category.
Last year’s results
I just found last year’s results. It looks like my math bears out, though it appears that they rounded differently.
Notably, last year’s changes were very different.
| OS | 2022 Share | 2021 Share | Share Change |
|---|---|---|---|
| Windows | 0.647 | 0.646 | 0.001 |
| Mac OS | 0.284 | 0.273 | 0.011 |
| Linux | 0.028 | 0.028 | 0.000** |
| Chrome Book | 0.025 | 0.025 | 0.000** |
| Other | 0.011 | Unknown*** | 0.015*** |
** Linux and Chrome Book’s absolute change was less than 0.0005.
*** As with 2023, the percentage change for “other” is not reported in the graphic. There is a remainder of 0.012 in the Share Change which is, presumably, attributable to change in the “other” category. This seems large, though, given that the total traffic share for “other” is only 0.016. That would imply a change from 0.001 to 0.012. Looking at the 2021 numbers this doesn’t appear to be correct.
Relevant xkcd: https://xkcd.com/1102/
Where did they come from? India?
Do we really want to be bigger anyway? I kind of like where Linux as a desktop isn’t really big enough for all the scammers and malware makers to care.
(And I know it’s huge for servers and malware also targets that, but they are usually maintained by professionals, not your parents that would probably run every shell script they are offered as help)
If Linux would become the biggest desktop os you are going to find so much more bad advice whenever searching for help online. I wonder if the nice people we have now are really ready for when the terrible people invade the community.
Yeah kind of partially agree but not entirely sure what to say man. I’ve had my windows machine and my Linux box hacked in the past, didn’t do much besides ruin my ability to do my homework and general productivity, so I can’t say much.
I would prefer there to be more actual meaningful stuff out there for there to be hacked and or made different but a lot of the time almost all complex systems compound into or towards static failure, just look at the USA with being a military hegemony like Sparta or China slowly running out of people to sell junk to, the big oil companies slowly trying to micromanage the shift to renewables, people that believe in conformity, confucianism and “the myth of stability” ironically usually slow down all of societies progress rather than supporting an actual stable diffusion of change
Do we really want to be bigger anyway?
YES. It needs more market share to influence companies financially to make products for it.
It’s truly starting to make inroads recently, but it still has a ways to go.
I kind of like where Linux as a desktop isn’t really big enough for all the scammers and malware makers to care.
It’s also not big enough for gaming companies to truly care, unfortunately.
You know, I do wonder how many of these statistics are influenced by Linux users tendancy to use adblockers and block tracking. Linux could be more popular than it looks.
Also, they should tell us how much of that increase is due to the Steam Deck. :P
The issue is whether the data is collected by a known third party tracker domain which would be blocked by an adblocker
Large sites usually have their own first-party analytics. Often they don’t want third parties to be in possession of detailed analytics data.
Unfortunately, due to the constant willful or untested shenanigans of various website I have set up all my system’s Firefox profiles to spoof by default its user agent (and other JavaScript properties) as Windows 11, x86_64, Firefox LTS (even if I use latest, Aurora or beta). Some blantant recent example: YouTube uses lower quality options on browsers running on Arm-based systems — misreporting as an x86 CPU appears to be a widespread browser fix
Doing so has helped me and many friends/family I switched to a flavor of Linux (mostly Mint, but sometimes LMDE or Ubuntu or specific requirements/demands) avoid numerous dumb problems.
Even on mobile sometimes UX breaking issues creep up.
