- cross-posted to:
- linux@programming.dev
- cross-posted to:
- linux@programming.dev
Linux Myths
A compilation of linux myths and misconceptions, busted and explained
Purpose
To catalog and provide useful responses to common linux misconceptions and myths. To serve as a useful reference for new and old users alike.
I’m not affiliated with the website or its creator(s).
You must log in or register to comment.
Myth: That this year, the year 2024, is indeed and finally the year of the Linux desktop.
It was for me. Been using Windows for 20 years, installed Aurora after all the MS craziness this year and haven’t looked back.
In my case it’s turned out to be a whole lot better - my laptop runs cooler, battery last about twice as long, and I no longer have any issues with going to sleep when I close the lid.
Well, it was for me. ¯\(ツ)/¯
Current Linux market share worldwide for desktops is at ~4%. There’s also ~2% ChromeOS which is Linux based so I don’t know why it’s listed separately. As well as ~6% other which is probably Linux with privacy settings turned on.
If we go back 5 years in Linux desktop usage, the high end is including the “Other” category.
2019: ~2% to ~9% 2020: ~2.5% to ~5% 2021: ~3.5% to ~11.5% 2022: ~4.5% to ~10.5% 2023: ~6.5% to ~10% 2024: ~6% to ~12%
There is definitely a growing trend, the user base has grown somewhere between 33% and 300% depending on whether you include the “Other” category, which I personally think is a pretty safe assumption since for most PC users if it’s not Windows or Os X, it’s Linux.
Here’s where I got the data from: https://gs.statcounter.com/os-market-share/desktop/worldwide
If it helps, desktops outside of enterprise are mostly dead. Sure there are still some among PC gamers, but the average household no longer is likely to have a desktop PC. Laptops, tablets, and smartphones have fully supplanted most of the demand for PCs.
The main Linux myth is: “Linux is OS”
Interjection imminent
Stallman, it’s time to get over it.
The section about stability vs “bleeding edge” gives me the strong impression that the author doesn’t really know what they’re talking about and only parroting something they heard someone else say.
Thanks for reading through it and giving your thoughts!
Could you elaborate on the mistakes/oversights found in the “Stable vs bleeding edge” section?
IDK. Gentoo is considered stable, but fedora “leaning unstable”?
Anyway what is that whole un/stable supposed to mean anyway? All non-rolling distros try to be stable. What can break are third party repos and stuff you compiled yourself. With fedora that can “break” twice a year. With a rolling distro that can “break” on every updates
Gentoo is considered stable
Honestly, I’m too unfamiliar with Gentoo to make a proper assessment on this. Though, even my (simple) understanding allows me to understand it as follows:
- Gentoo is not a point-release distro. Hence, by definition, it satisfies the definition of a rolling-release distro.
- Furthermore, regarding Portage’s branches, this page suggests only two branches; Stable and Testing. With Stable being set as the default one. Furthermore, AFAIK, tests occur for over a month over at the Testing branch before updates enter the Stable branch. Hence, more time is taken compared to other rolling-release distros.
Which, I believe is what’s alluded to here: “The update philosophy of a distro is generally not related to its release cadence, as you can have rolling release distros that are relatively stable (for example, Gentoo) and point release distros that are relatively bleeding edge (for example, Fedora).”
Is there any reason why you would deem Gentoo as not stable? If so, what?
but fedora “leaning unstable”?
For the sake of completeness, proper quotation would have been “leans bleeding”
I’ll give you that the article is definitely not exhaustive and/or properly clarified. Perhaps for the sake of brevity, idk. Hence, I believe that this confusion is justified. However, again, I think the raised point is justifiable based on the following:
- Fedora is known to push new tech first. Heck, it even adopts it first; e.g. PulseAudio, systemd, GTK4, Wayland, PipeWire etc. Hence, this causes Fedora to feel bleeding edge; i.e. because its users are literally the first to test it en masse.
May I ask why you think Fedora does not lean towards unstable?
Anyway what is that whole un/stable supposed to mean anyway?
I agree it causes more confusion/conflation that it has any right to.
All non-rolling distros try to be stable.
It depends on the used definition of “stable” 😅.
- If “stable” is used here in the context of name for used branch (of the repository). Then yes, but this also satisfies rolling-release distros; as they, by default, ship software with their own designation of “Stable” (even Arch). (For the sake of completeness, I’m aware that some distros default to testing/unstable branches.) Hence, using this definition of “stable” is not very productive.
- If, instead, “stable” is used in the context of stability. Then, also yes. And, yet again, this also satisfies rolling-release distros. It’s not like any reasonable distro is out there to deliver software that’s known to cause issues and whatnot. The distros only differ in how exhaustive their testing is. Which gets us to…
- If, finally, “stable” is used in the context of how well-tested the distro is. This also ties in to the earlier presented definition for name of the used branch (of the repository). Because we all know that Arch’s Stable repository is wildly different from Debian’s Stable repository. And here, unsurprisingly, we find wild differences that are also actually helpful in a productive conversation.
- (Surprise,) tied to the previous point, “stable” could also refer to how often the distro requires you to update. With “stable” being used to indicate that updates are only required between (infrequent) point-releases. However, non-intrusive security updates should be able to get through regardless.
What can break are third party repos and stuff you compiled yourself.
Sorry, I can’t agree with you on this. Even if this is said in the context of non-rolling distros, my experiences with Fedora suggest otherwise. Granted, Fedora is sometimes referred to as semi-rolling release distro. So, perhaps it (and direct derivatives) are the exception.
With fedora that can “break” twice a year.
Agreed (with earlier mentioned caveat*).
With a rolling distro that can “break” on every updates
Agreed.
phew long answer. I wouldn’t call Gentoo unstable. I was rather interested in why it’s supposedly more stable then Fedora.
I just wrote from my limited experience. I never had something break on Fedora. I just updated a system from 35 to 41. The stuff that broke was something I compiled against old dependencies. (That’s why I didn’t update so long)
My Gentoo experience is >15y old. I had numerous incompatibilities, because I used the tools the system gave me. But sure that’s on me if I cutomize my system with USE flags. And it’s probably better now.
phew long answer.
Yeah, lol. My apologies 😅. Thank for reading through all of that 😊!
I just wrote from my limited experience. I never had something break on Fedora. I just updated a system from 35 to 41. The stuff that broke was something I compiled against old dependencies. (That’s why I didn’t update so long)
Interesting. Within the last two years, we had issues with mesa and codecs; some devices couldn’t even boot up. I’m glad to hear you haven’t had any issues though. BTW, for the sake of completeness, I’ve been daily driving Fedora Silverblue for over two years now*.
I just updated a system from 35 to 41.
LOL, my first version was version 35; which is a curious coincidence. With 41, do you mean Rawhide?
My Gentoo experience is >15y old.
15 years is a long time 😅. Do you recall if Gentoo had dependency resolution back then?
Was on the phone and only quickly looked up the latest version. So I only updated to 40, not rawhide.
Sure Gentoo had dependency resolution. Does Gentoo still have use flags? Because that makes dependency resolution much hardere It’s not enough to know the dependeicies, you also have to know all the use flags you dedend on. And if a maintainer adds a use flag for a feature you depend on, you have to add that dependency as well or people who disable that flag break with your package.
I’d be surprised if gentoo was considered stable, if you make heavy use of use-flags - if they still exist.
edit Maybe your “dependency resolution” is a new automatic thing that identifies dependencies including use flags automaticallt? It was automatidally done, only if the maintainers put the right stuff in their ebuilds.
Does Gentoo still have use flags?
AFAIK, it still does. Though I wonder if it still comes with a lot of added complexity that you note. I’ve yet to give Gentoo a good try. Therefore, you could be totally right with the following:
I’d be surprised if gentoo was considered stable, if you make heavy use of use-flag
Regardless, it was an informative talk. Thank you for that! Have a nice day 😊!
