The Problems with Linux No One Talks About (Featuring @RaidOwl)
#Problems #Linux #Talks #Featuring
“Learn Linux TV”
What’s wrong with Linux? In this video, Jay and @RaidOwl discuss some of the issues that surround Linux and prevent wider adoption in the desktop market. This video is part of Learn Linux TV’s 45 Drives Creators Summit coverage.
To see the full content, share this page by clicking one of the buttons below |
*Check out 45 Drives*
• Check out their site here ➜…
source
Concluzion: The Problems with Linux No One Talks About (Featuring @RaidOwl) – Linux,gnu/linux,LearnLinuxTV,Learn Linux TV,LearnLinux.TV,Learn Linux,Linux Training,Linux Tutorials,45Drives,45 drives,Storage,Enterprise Storage,Enterprise Linux,Linux Desktop,Linux Desktop Adoption,Linux Workstation,Ubuntu,Debian,Fedora,SUSE,Windows migration
Typical Windows/Mac user complaints. Cliché
I also hate repos. Thank you for this insight.
Lewis Cynthia Young Patricia Lopez Brian
The Issues for me is the following:
1. Drivers. Broadcom chipsets on a new laptop are not picked up right away. Compared to Intel Wifi or Ethernet. Intel works out of the box. I had to replace the laptops internal broadcom WiFi to Intel and works.
2. Active Directory. On Mac it joins Active Directory easily and works. On Pop, Kubuntu, KDE Neon does not. Ubuntu has the feature to connect to AD domain in the installation but when you try to add the AD domain it will not work. You have to do back end stuff to make it work. MacOS just works and very easily add Macs to AD. I wish this was easier to finally switch all my computers to Linux and still test some AD features as just basic AD login. I understand there is Opendirectory on MacOS but not as pushed on Linux.
3. I tried PopOS, Ubuntu, KDE Neon, Garuda and settled on Kubuntu. The distro hopping is cool but at the same time annoying.
4. With SteamDeck and Raspberry Pi RetroPi builds and homelabbing are making people look toward opensource more. Which is good but then you get stuck being the support person for the family.
5. DRM for games. With Valve's push to opensource with time I think it will go away. I am sticking to PC gaming over console gaming.
Make connecting to network resources, Linux or Win., easier via the GUI in desktop. And vice versa. Terminal free sharing and connecting to resources on the network. Active Directory, too. I have fewer problems connecting to and from Ubuntu Server* than with Desktop Linux.
The future of Linux is Installation (up & running), integration (in current enviros.), and migration to a full-on Linux enviro. (IIM). Linux devs. MUST start thinking of users in the context of IIM.
Linux must move beyond adolescence to maturity.
*Post No Distro. Bills.
The kinda annoying thing to me was all this talk about how there are too many different distributions of Linux, only to then disagree with design choices of the Linux they have used. If there was only 1 Linux, they'd be be stuck with those designs. At least now they can find a different distro that has people agreeing with their ideas.
But I guess what they mean is "There should just be 1 good Linux, and good means all of my opinions on how to do things."
Nice discussion anyway though, just cool to see what different people think.
Hmm. I'm 7 years into 'trying' Ubuntu. Now I'm on Fedora. Yes, some hardware won't work. Some hardware doesn't work on Windows or Mac either. It's just most hardware vendors have ignored Linux. So use Linux, and they won't be able to ignore it. smh
I think one thing Microsoft did early on was too make Windows pretty easy to use for beginners.
Sure there is stuff you can dig into the weeds on, but most folks don't need to.
Linux took the path of powerful and robust, but not easy to get into, however extremely customizable
I saw a post on reddit about a guy ranting about how 'new' users coming over from windows are lazy and won't rtfm and how tired he is of answering noob questions. I told him, people like him are the reason why linux doesn't get big. He sees new users as a problem, I see new users as an absolute win. If more people keep coming over, things will only get better for the linux community. device manufacturers will start building drivers, software companies will start shipping their software with linux support. Please people, help new users. Have some patience. We all know how much of a terminal genius you are and how you were banging out shell script inside your mom's womb. No need to flex that. No need to scare away new users. by doing that, you are harming the community and blocking the entry. Do not freaking gatekeep.
RTFM? Write TFM, publish TFM (in print or complete downloadable format like pdf), link to TFM. Even then, there are people who don't do well w/ manuals, but there is a paucity of assembled, accessible, indexed documentation.
If you're into Adobe, Linux isn't going to work for you. At this point I feel this is a known failing use-case we see repeatedly from youtubers.
It would be nice to see someone do an "enterprise office" thing. It doesn't have to be all on-prem & self-hosted, but can we do linux as an enterprise desktop?
in Linux we trust, nothing more
Fingerprint sign in is still not really implemented unless you intall a hack named "grosshack" in pam modules. I wish that was fixed.
If you see someone else on a forum giving an unhelpful "RTFM" to someone asking a question, here's a similarly short response to give them: ATFQ – "Answer The Fucking Question!"
My experience with Linux communities has been positive.
I stopped using LR, I now use DarkTable on Linux Mint.
I am full time Linux for many years.
Such an interesting conversation: I can definitely relate with some of the problems with switching, especially moving away from known procedures and software. But with the growing compatibility of games on linux and AMD attitude towards open source, I think switching now it's relatively easier for gamers and general users. Plus, most work now is really browser based, until you need a specific closed source software that won't run on linux.
Linux is often like Nissan Skyline. It’s great if you know what you’re doing, but for most people, VW Passat is much better choice.
The old …..keep it simple…..mantra is how to help people use and understand linux. IMHO linux messiahs start simple …and as a newbie you think .."Okay ! Yeah…I understand that. Then for some inexplicable reason ….they veer off into the weeds and you go …"WTF?!"
You get the normal Windows user confused or intimidated and they will immediately retreat to the friendly confines of Windows. Show people an easy desktop version of a linux distro like Zorin or Mint and how to use a few of the types of apps they use on Windows. email, video etc.
It's weird that Brett encountered "oh, you need a GUI for that??" if he was hanging out in Linux Mint spaces … they kinda frown upon that sort of behavior. Unfortunately AFAIK there is NOT a GUI for that even on Mint, yet, but y'know … that could exist. It should exist. But definitely that "RTFM" attitude has been a problem since 1997.
I like the direction of smallish immutable distributions with containerized apps is the direction for the end-user desktop. It's probably not the ultimate solution for everyone, but it's a good default—sane defaults is what Linux really needs to strive for.
I think an update to this video in a year or two would benefit from collection actionable points from this one, such as:
Hardware support. Is there either a better chance Stuff Will Work or at least you're able to determine in advance what will and what won't? (Recent changes with Nvidia might give a couple free points on that one already!)
Network config and sharing configuration: Can this be done from a GUI or at least without learning how to write config files yet? I'd look at basic setup for wired and wireless, personal firewall because you need one, and file sharing. (Printer sharing? IDK.)
Software availability: Can you get the stuff you want? Like do you know where to get it from and can you get a reasonably recent version? How do you know where to get it from and what's involved. Pick several test apps.
Stuff like that. And put those points out there somewhere soon so that people can look at this list of pain points and maybe address them. And maybe pick a few distributions to test out? Let's give them real report cards. Observable, measurable, actionable. I think this sounds like an interesting project, honestly! Wouldn't mind contributing to it. I wouldn't mind testing stuff to see how it fares either.
I'll throw in one of my own right now: Input devices. Keyboards, mice, … and game controllers. Wired probably works, but what about wireless? Do any special features work? What if the wireless is BLUETOOTH, how stable is it? And I'm including game controllers because yes, Linux gaming is becoming more of a serious thing. Oh, and for wireless devices … can you do things like read battery levels? If there are settings for the device, what works? (Perhaps that gets into a hardware database a little?)
Stupid stodgy design.
Lack of uniformity in apps.
FreeBSD ports and the like solved the repo issue long ago; the AUR is a close second. Otherwise, with appimage and flatpak, etc., people are just starting to handle sandboxing versus interoperability issues (I'm thinking how GIMP works with Darktable). But having a kernel developed independently of an OS is the root cause of the problems. For example, Debian-based distros have a sudo that deals with paths differently than, e.g., Arch-based distros. Whose security model is better? Even among Debian-based distros, those using bash may handle .bashrc and .profile differently. So developing for "Linux" means setting up complex queries to handle whatever individual decisions may be out there. Until Linux-land adopts the wisdom of BSD and develops a kernel integrated with a core OS, with actually decent docs, the fractured, fractal nature of the Linux community will prevent any economically-viable dev support. Mac has a different problem due to proprietary costs of development and VSCode. But if there were a core, official LinuxOS that guaranteed certain functionality, and offered well-defined dev support, that would create a foundation for computing that just works (TM). One could address other issues after the basics were dealt with. Remember that the growth and adoption of UNIX occurred precisely because there was a central standard that could be adapted to meet local needs. Linux has at least three "standard" distros (RHEL, Arch, Debian).
good one
That KDE 4.0 thing… I think their mindset is that their version doesn't say anything that it is ready for production. Instead what is production ready is that is passed QA of OS provider. Like, is it found way to Debian stable. It wasn't 4.0. It was 4.4.5. There are several stages how QA is done. It is same in Windows. Origin of code is developer machine, they push them to internal testing, after internal testing it ends up to Windows insider. And it takes time to get it from there to production.
So, there is freedom to choose OS but it is important realize that example Fedora is close to Windows insider while Red Hat Enterprise is the production version. Same thing openSuse vs Suse Enterprise. Rolling releases are even more bleeding edge than those, they are close to minefield. Don't expect to have stuff that is not made to production in laptop found from store shelf.
Those "standard distros" do exist. They are those who stand their own and made for production. There are several of them focusing on different areas. Then there are derivates of them when some people are not satisfied and want to change new theme or something. Linux indeed work, but when talking OS you should specify what OS you are talking.
Also it is not that uncommon to wait next OS release to get software updated. I do agree that it can make sense if applications are taken out from repository that offers platform. However it is harder to define what belongs to "platform". I think it is something that can be kept stable in way that others can build something depending on that. Most upper layers in software stack are only dependent on "platform" so these can in separate repository, and lower level is often very much standard. Hard to draw line here because some applications do have intefaces for plugins, so application can be also part of platform. Example GIMP. Do you want to GIMP to be in application repository and when it is updated it can break some business critical plugin?
LibreOffice can of course used to write native ODF documents. This kind of productivity suite of course need to be standardized in organization because they all handle files differently, especially if it is not native to software. That is why Google suite and MS Office works from browser.
Thanks for the podcast!
Mounting network shares in Debian:
-add package: nfs-common
-command as root user: mount -t nfs remotehost:/path /thishostpath
-making it permanent needs to add it to /etc/fstab like: remotehost:/path /thishostpath nfs defaults 0 0
Details can be found from NFS documentation.
My personal 2 cents: as far as I was able to see during my 26 years of using Unixes and Linuxes both professionally and at home, Linux (and Open Source as a whole) got widespread as software geek ideology intermingled with a "free beer", "stability" and "UNIX-like". While "free beer" and "stability" are OK with your average desktop users, they care a ton less, if at all, about ideology or "UNIX philosophy" (because you never do such things w.r.t. an appliance).
What you said about LibreOffice reminded me about an argument with my kid: after he showed me the latest and greatest LibreOffice Writer (on Windows) going temporarily black while saving a short simple document, I just gave up and shelled for MS Office 365 family license.
There are so many abandon projects in linux how can I be sure that if I move my business processes to a native linux app that it will be supported? Devs are so fickle now at the first chance they are on to a SAAS solution which is always in your pockets with fees. For that experience we have MS.
"installing an OS on a refrigerator" is a great paradigm shift I hadn't really considered before. I "knew" that most people don't consider their OS, but it's hard to really duplicate that feeling or sympathize. I felt if you use any two devices with different OS: Windows, Android, iOS, MacOS, ChromeOS, heck even Blackberry/flip-phone era or game consoles, you'd consider how you like to interact with devices. Not everybody has access to multiple devices or upgrades, but considering how people felt about Windows 7 to 8 to 10, it seemed like the effect should have been wider when including other interfaces as comparison.