RE: LTT's third Linux challange video, or "how I stopped raging and started loving the newcomer"

Just watched LTT’s third video where Linus and Luke try to run Linux as a daily driver on their home systems. The third video was the more “daily driver” part of the challange, containing what could be perceived as “daily tasks” using the Linux desktop.

I think that they in some way helped to point out a couple of small problems here and there, and I in some ways agree with Linus in regards to how Dolphin is a bit janked, though Linus sure took his sweet time and they both over-complicated a couple of tasks, but I wanted to adress how the video ended.

They had to underline the video by pointing out the backlash they had received and about how they perceived certain Linux desktop discussion threads online. The word used was: gatekeeping. One thing I kind of get a bit tired of is the sense of elitism found around linux user spaces, mostly because it hurts the original goal: to spread libre, free software.

When I reminisce over IRC logs of years past and certain “conversations” I’ve had on on and think about how perhaps the introduction of the code of conduct hoisted upon GitHub users was probably for the best. Reading certain support threads online makes me think that certain venues might be actively scaring away users from wanting to use libre software, and that scares me.

For me, I hope that we can create positive, welcoming communities, where we help people - or if we can’t help people, at least be able to give them some helpful hints, tips or proper directions on how they can reach their goal and solve their problems. It is after all part and parcel of making open source software better, as feedback is technically a form of quality assurance.

Elitism gets us nowhere, and even though I frequently use the CLI, I don’t really like how people are pushing it was a necessity as a justification for lacking functionality or “jank” user experiences in certain software that probably would benefit from actively getting feedback about certain issues.

Of course there’s a problem with fragmentation of communication channels, how to source proper documentation (the Arch wiki aside) or outdated community websites, etc. There’s a lot of hoops newcomers have to jump through to get acclimated - if they don’t just nope out and decide to try at a later date, or scrap the idea of using libre systems altogether.

For me the issue is clear: I want people to use Linux desktops because they are based in FLOSS principles. But that alone won’t usher in the mythical “year of the Linux desktop”. There are clear barriers to entry in the form of irritations, confusion and even toxic behaviour that prevents newcomers from settling in, on top of the classic hardware support issue that has given unfair advantages to proprietary operating systems like Windows.

I think we need to look at how we interface with communities and channels, with how we introduce and facilitate failtesting and troubleshooting, as well as creating places where UX problems can be discussed without getting “use the terminal” thrown in your face.

What do you guys think? Is learning the terminal really that vitally important, or should we be able to facilitate discussion, promotion and encouragement of usability improvements? Are communities really not that toxic, but it’s just a matter of having some fundamental understanding before you go into such a place? Let me know what you think.

1 Like

Wow, so as someone who started their use of PC’s back when they were Macintosh clones running 48K of memory or the new PC from IBM that touted 64K of memory, I just want to interject a bit of perspective here. First I would point out that in the beginning it was all proprietary or free, when speaking about what you ran, the third was your own code, and usually in basic. As interest grew, proprietary overtook, not because it was better, but because it was available and deep pockets financed it. The time of “value added reselling” had arrived.
Open source was born from the desire in the community of users at home and in the office, as well as in educational institutions and the industry plant floor, to be free to use the hardware as desired.
This ideology, goes directly against the model “sold” as the way to use a PC by the closed system folks. As this approach progressed, the desire to reduce issues by user complaints, opened the door to limiting user control over how the system configuration and setup was implemented. Again this is for the companies reasons, not the users of their provided product, and is not necessarily going to be the best or right solution, just the best or right for the shareholders of said company.
So if you desire to have everything about how you use your system predetermined for you, and by that I am referring to how you use it and configure it beyond the sane defaults provided, then I would suggest most Linux distributions are not really for you beyond wanting to be with the “cool kids”.
That is my opinion on use of any Linux distribution as opposed to the closed system alternatives available today.
Fedora Linux specific, and not just an opinion here. It is a leading edge Linux distribution that provides an OOTB usable OS with an entirely open sourced stack of packages with that first install. It ships with only open sourced software with clear licensing and freely accessible source code you can view and fix if you feel the need and share your solution with the community. This is a contrast to what many who come from the closed source proprietary uses are accustomed to, and from a convenience POV can be frustrating.
So yes learning how to use your system is a good thing, that includes the terminal. Bad habits should be unlearned, and just because it is easy and comfortable, doesn’t really mean it’s the right way to do something. Besides, in this world changes often come rapidly.
Discussion about usability is always open topic in the community I would think. Just because using the terminal once in awhile comes around, it shouldn’t be treated with apprehension or as some form of elite undertaking. Nor should it be misconstrued as rendering the system unusable as a consequence of needing to actually do something in a terminal.

1 Like

I can see your points, largely because I started around RedHat 6.2 (note: not RHEL, but back in the GNOME 1.x days). I’ve seen the lack of support and availability that had to be augmented by community input, but that’s sort of my point.

Using the terminal is all well and good, but making it mandatory in some sense is kind of ineffective. I’ve spent a fair amount of time in chat teaching people how to use journalctl and where to find logs, how to interpret them, how to operate certain daemons/services, how to figure out where the source of a problem is, so that they can then either find a solution or at the very least provide a support ticket to the right project.

That time spent is due to the fact that there is really no easy way of doing it at this time and date, let alone easy ways to report or find the right projects to report to. Not to mention I have problems on my own system, but instead of going down the rabbit hole of finding where to actually file bug reports, I let them be as I have other projects in the pipeline. Sometimes I really can’t be bothered, and that’s the problem.

My issue is that we have a discrepancy, a rather big red line between the average user and technically savvy users, and that said line needs to get thinner. I thoroughly believe that if we could create a more welcoming environment and easier mechanism to actually contribute back that we would see much more yield in regards to hardware support, UX improvements accessibility improvements (because having to navigate with a straw in your mouth means the terminal itself is kind of hard to use).

I think that if we want a “Year of the linux desktop”, it will only come when we have made OSS projects ready for the intake of various problems and that we see a renewed commitment to UX and design in regards to functionality, i.e not having to rely on terminal all the time. If someone calls Microsoft or Apple and gets the help they need they will be happy. If they go into a chat on Matrix and meet the wrong people it’s going to be severely demotivational and probably contribute to them fleeing back to proprietary land, something I think we’d like to avoid - at least if we think that FLOSS should be the standard to protect peoples privacy, autonomy and generally speaking their liberty from being subjugated by these large corporate entities.

1 Like

Here’s a bug ticket that shows a fundamental problem with reporting processees. It relates to a problem that I currently have, but since tickets are on a per-system version basis it will be EOL’d along with the system in question. This isn’t very intuitive, as now the entire search result for my problem is littered with outdated and unhelpful bug tickets. It’s indicative of usability issues rather than anything else if you ask me.

But it is effective at cleaning up the tickets found on the redhat bug tracker, in a sense that it is designed to help developers and maintainers exclusively. Bug tickets could be sources of further exploration and discovery to solve certain problems, rather than being a dead end. It becomes a fundamental problem of design and purpose if you ask me.

Btw, I decided to try and use GNOME Logger for the purpose of trying to find a viable example, something which was frustrating since it was the Flatpak version from Fedora and clipboard functionality is broken there. For a keyboard warrior such as myself that was frustrating. But for someone who’s used to right-clicking and copying text it wouldn’t be a problem.

Bugzilla for Redhat is monitored and maintained by people paid to do this for Redhat, whereas the issues reported against Fedora Linux are dependent upon the community at large to help resolve, which is a voluntary role for anyone stepping up. There is no product guarantee when you download and install Fedora Linux that would be comparable to what you get from purchasing a Redhat subscription for support for instance, or even an install for personal use RHEL which comes with a finite life of support.
Having said all that, your point(s) around accessibility are something that are in the forefront of some discussions around here and at ask.fp.o frequently. In most cases I find people in general will be okay with resolving whatever issue they are confronted with provided there is ample documentation and an obvious path to it. One thing that is apparent with Fedora Linux Documentation is that it suffers from the release cycle frequency and also the pathways to the relevant info are not obvious always (shameless plug for :fedora: documentation volunteers here). Combined with certain (I hate to use the term) corner cases that don’t get caught during QA cycle of a release (more shameless plugging for :fedora: testers) contributes to the general feeling of dissatisfaction from the newcomer who experiences something like the issue you highlight.
That issue could be solely related to the actual chip used in the particular production run of your hardware for instance, so maybe it is just waiting for someone with the right hardware to do some test for, I don’t know, isn’t hardware usually the realm of kernel dev’s?

1 Like

Yeah, I see that, but it’s irrelevant which tracker it is. Part and parcel of my problem regarding usability is not fragmentation per say, but interoperability between platforms, communities and the ability to get credible information. It becomes a needle in a haystack problem.

And no, it isn’t always about hardware regarding opcodes. Sometimes it’s because the service or application (in this case bluetoothd or bluez) that can’t handle certain exceptions. So the problem should be deferred to them.

The fact that the problem persists after many years means the developers have not been given the right data or telemetry to see the problem up close, either that or it’s because it simply isn’t a priority.

For bug tickets we are dependent on the user to help out. If a user files a bug and it isn’t fixed and it is going to be closed because the version is EOL, the user needs to update the version. Every bug I’ve had I’ve kept updating the version so it doesn’t get closed. I do test to make sure the problem hasn’t gone away.

For the most part my kids work with linux without needing the command line. I think we have created a pretty good GUI even if there is room for improvement.

Here is a new user’s perspective.

In its present form, I don’t think there will ever be a “year of the Linux desktop.” Its never going to catch on with the general public. I am here because I am an enthusiast who wants to learn the Terminal and how an operating system works, but when I tell friends and family about my how excited I am by Terminal commands, their eyes glaze over. The vast majority of people just want the point and click OOBE that Microsoft or Apple provide. They just use a computer because they have to, not because it excites them. Those of us who enjoy Linux are already far more proficient with a computer than the average Joe. Most people couldn’t even figure out how to install Silverblue, much less use it. We also see the advantages of FOSS that most people don’t think about.

As to your point about a welcoming environment, I think we already have it. When I was using Clear Linux, not only was there no support from the community, responses to questions were often dismissive and not welcoming. The Clear Linux community takes the view that if you have to ask a basic question, you shouldn’t be using their OS. It’s reserved for the gods of Linux only. In contrast, the Fedora community is very welcoming and inclusive. I am most appreciative of the support I’ve received. It makes me want to continue using Silverblue, and contribute to it in whatever small way I can. In my view, the newcomer is already welcome here.

1 Like

I think this is actually a really good example of the issues I outlined in this topic about our Bugzilla documentation. The issue you’ve got there is a real one, and a serious problem with Bluetooth that we’d like fixed. However, we only have a couple of kernel maintainers in Fedora for all of the kernel. They’re not really equipped to dig into each of these issues. The main way these things really get fixed is by the bluetooth kernel stack maintainers upstream.

Filing the bug isn’t useless — if we get a lot of those, it helps identify particular problem areas that maybe the kernel team should focus on. But it’s just not really possible for it to get individual attention, which sets up users to be disappointed and the maintainers to feel guilty because they want to help.

I think it’s better to track these kind of things as issues — like “bluetooth is a problem area” for the Fedora Workstation team, and as help topics on Ask. Then, there might be individual bugs tied to those, but not necessarily ones in Fedora in specific.

Does that make sense?

I’m super glad to hear this, Joe! We can’t do much about the YouTube commenters or the attitude of Reddit users, but we can work to make Fedora discussion spaces positive and awesome — and that’s definitely been an area of focus in the past few years, so I’m glad to see it’s feeling like that to new folks coming in.

1 Like

Thanks for posting :grinning: I enjoyed watching this video, and I found the guys to be genuinely interested in solving the challenge in whichever way they were able to. I have not watched previous videos, but judging by this one they are not out to smear the Linux community at all (they mentioned in the video that some had accused them of that). Rather, the contrary - to help improve it.

Overall, I found their points to be valid and I think it’s important for our inclusivity to see how newcomers approach an unfamiliar platform. Yes, sometimes over-complicating, but that’s not unexpected in some cases if they are not THAT familiar with the platform, even if they are experienced computer users. And now consider people who are not even experienced computer users :slightly_smiling_face:

I am a happy Silverblue user, and I have set up my father’s desktop also with Silverblue. He is NOT very computer-savvy, but overall he manages to do what he needs (email, scanning, printing), although there have been a few complications with the sandbox permissions of the flatpak apps, I must say. I ended up installing Thunderbird for him as an RPM overlay, and things went easier.

1 Like