qsl Fair enough, but I will say the whole process of chroot and whatnot is enough to scare many folks away from linux forever.

I'm not quite sure how my comment (which was intended to explain to @Kevinsotovalle that the process of creating new ISO's is complicated and time-consuming for the team) morphed into a discussion of the user-unfriendliness of Linux in general (if that is what you meant by "the whole process of chroot and what is enough to scare away many folks from Linux forever"), but if that is what you meant, I agree with you.

I have used Windows and Linux in parallel since I started using Linux in 2006, and I have coached a number of friends trying to adapt to Linux. All found it difficult, and many (most?) decided after a while that the rewards did not justify the effort, abandoned the effort, and returned to Windows. I am suggesting to most of the people who ask me about moving to Linux when Windows 10 reaches EOL in a couple of years (non-technical senior citizens for the most part) that they would be better off buying a new Windows computer rather than trying to adopt Linux on their existing computers.

I know that my view is heresy on a Linux forum, and I have been pounded around the head and ears for saying this on other forums, but Windows 11 is an excellent operating system, user-friendly, requiring almost no technical knowledge on the part of the user for the most part, and "just works" for most people. Linux remains less so despite years of development. I see no compelling reason for most computer users to move to Linux.

I've argued for years that Linux will not gain significant market share unless and until (1) the Linux community becomes more focused and disciplined (that is, heeds Torvald's 2014 advice to reduce the number of distros and apps dramatically and focus on quality rather than quantity/choice), (2) Linux can be installed, used and maintained without resort to the command line, (3) Linux architecture evolves in a more modular, "plug and play" direction, instead of tossing users into "dependency hell" requiring user intervention to get sorted, and (4) Linux gains enough market share that OEM's and manufacturers supply reliable Linux drivers as a matter of course. To gain market share, Linux has got to get to the point where Linux is as easy to install, use and maintain as Windows, Android, macOS, iOS and ChromeOS. I don't see that happening any time soon, if at all.

I think that Solus (focused as it is on ordinary desktop users, extremely well-designed and maintained by the team, and carefully curated) comes as close to meeting those criteria as any distro I know, but it is still a long way from entirely meeting those criteria. Solus, being Linux, gets stuck in "dependency hell" more often than it should, still depends on the command line for troubleshooting/repair, and too often requires user intervention to get proper drivers in place. Windows, Android, macOS, iOS and ChromeOS don't present those issues to users, for the most part.

qsl If there is a learning curve, it should be showingcasing how the linux approach is better than any other OS. Otherwise you are just being esoteric just because, which may be fine for a private club, but makes for a shitty OS.

Well said.

In the last week, I updated a friend's computer from Windows 10 to a clean install of Windows 11, did a clean reinstall of Solus Budgie (to remove any cruft as I moved from an private Edge eopkg to the Edge Flatpak) and spent several hours trying to figure out if I should move my railroad computer (Dell Latitude 7390, i5 with Intel onboard graphics and Intel wifi) from Windows 11 to Solus.

The experiences were instructive.

The Windows 11 installation took about an hour and 10 minutes, most of it consumed by automatic updates, the Solus reinstall took about 40 minutes, again most of it consumed by updates. Post-installation setup (user preferences and so on) were all handled through GUI, as was app installation except for installing the Edge Flatpak, which required CLI. Because my friend syncs her Windows instances through her MSA, her Windows preferences were set up automatically for the most part, and because I use Solus Budgie OOB for the most part, I had to change a half dozen settings and I was done. Easy as pie in both cases.

But then, to the laptop. As I thought about it, I was reminded of the limitations of Linux in several respects.

I didn't have to worry about basic drivers, of course, because the laptop is vanilla Intel, but I remembered the testing I did on that computer in April-May, testing power usage and battery life (about 70% of Windows, with outside case temperatures several degrees hotter than Windows) and thought, why do I want to do this?

The next thing I thought about was setting up the display (the laptop is 13.6" running 1920x1080 at 150% scaling in Windows), and 100% scaling on that laptop doesn't work for my aging eyes. I read forum discussions about how to do deal with this, decided against xander, and set my fonts to display at 150% on Solus Live, which was an acceptable, if not entirely satisfactory, solution. But again, I thought, why do I want to do this?

And finally, I thought about MyRadar, which is a simple weather radar app available on iOS, macOS, Android and Windows, but not Linux. There is no Linux alternative, and I have looked high and low. I decided that since I had the app on my iPhone, I didn't need in on the laptop, but again I thought, why do I want to do this?

In the end, I decided not to switch the laptop from Windows 11 to Solus Budgie. I wanted to, but it simply wasn't the right choice for me.

To me, those experiences encapsulate what I've been saying about Linux and market share. Solus Budgie is as good as it gets in Linux, but Linux needs to improve dramatically to become an OS for "Joe User".

    tomscharbach

    I'm not quite sure how my comment...

    More an orthogonal comment/observation on my part concerning the frequency of ISOs, the less than ideal results, and a means of possibly avoiding the issue altogether.

    And I agree with you with usability issues in linux for the most part, heavily tempered by the fact it is a comparison of the output of billion dollar corporations vs (how many people on the team?) essentially Geppetto's workshop. For the resources available, Solus is freakishly good/great, and the fact that it can even be considered as in the realm of a drop-in replacement for numerous people is a testament to just how considered the development has been (I could point to numerous problems with Win 10, how OOB it hides too much from users, has way too much cruft, or as the last post by kyrios makes clear; you are giving up more than the treadmill of regular computer upgrades by using it).

    At this point, I'm probably less interested in the "everyman" aspect per se than seeing where this mutant branch leads to.

    It would be cool if other OSs start taking cues from ideas Solus pioneers.

      qsl Solus is freakishly good/great, and the fact that it can even be considered as in the realm of a drop-in replacement for numerous people is a testament to just how considered the development has been ...

      I have often described Solus Budgie as "near perfect" for home desktop use, and I continue to believe that, in part because I just spent four months using Ubuntu Budgie daily. Sous is exceptional, but remains subject to the limitations of Linux in general.

      qsl I'm probably less interested in the "everyman" aspect per se than seeing where this mutant branch leads to.

      We've got something of a chicken and egg problem, in the sense that Linux and Linux OEM/manufacturer support will not get up to speed unless and until Linux market grows, but Linux market share is unlikely to grow unless and until Linux becomes an OS that works for "everyman".

      qsl It would be cool if other OSs start taking cues from ideas Solus pioneers.

      Solus points to the future of Linux desktop, and I try to support it as best I can.

      Every week is overkill.

      I think the base mission of the ISO is to be able to get any PC to be able to install Solus as smoothly as possible. The last thing the user should experience is having nothing pop up on screen or having to fall back to the command line to get things working.

      It doesn't necessarily have to be perfectly tested in terms of the packages on the ISO. The idea would be that the user installs it, runs an update, restarts then gets on with whatever it is they want to do.

      To that end: I would suggest that a new ISO be prepared before the Intel ARC and Radeon RX 7000 series GPUs launch. I don't think the current kernel/Mesa/llvm in the ISO has the bits needed to get those GPUs to a basic working state (though correct me if i'm wrong). So if someone buys a new PC with those bits and uses the current ISO....what actually would happen?

        Neumie It doesn't necessarily have to be perfectly tested in terms of the packages on the ISO.

        I have a different oppinion. I don't think it helps Solus if new users go through the download and installation process, but then crash their system because of an error that wasn't caught and fixed. That's not what "curated" means. We need to get as close to perfect as we can.

          WetGeek I don't think it helps Solus if new users go through the download and installation process, but then crash their system because of an error that wasn't caught and fixed.

          I agree. The installation process is a new user's introduction to Solus. It is the one part of Solus that we should take special care to ensure is absolutely flawless. If my fresh installation the other day is an indication, the team is doing this. If flawless installation means that the ISO is refreshed less often than it otherwise might be, so be it.

          WetGeek Sorry I probably should have phrased it a bit better.
          Yes of course the installer should be as stable as possible. The ISO needs to get the user to a point where they can boot into their newly installed system and run updates. That's the critical functionality of the ISO imho. Make that as smooth and as simple as possible. So CPU support, GPU support and networking are critical for the ISO.

          But the entire contents of the ISO outside of that doesn't need to be production ready. e.g. if sound doesn't work on the newest mobo because the kernel on the ISO doesn't have support for the sound chip on your mobo...fine. It's not the end of the world. Updates will fix it. If the power management for your CPU/GPU doesn't work straight from the ISO and your PC is stuck at low clocks....not a problem. It's a minor inconvenience. Run updates and a newer kernel will have that fixed.

          Consider: if the RX 7000 series and Intel ARC GPUs released today and someone had them in their PC. Would the current Solus ISO be enough to get a new install completed in a smooth and trouble free way? Or would they need to pass in extra kernel parameters? Or would they be stuck looking at a blank screen?

            Neumie I see your point and it makes sense.

            The problem, of course, is that drivers are a kernel issue and the kernel is an ever moving target. If Solus were to release a new ISO today, the kernel would be 5.15 rather than 5.18.

            That's not the end of the world (Ubuntu Budgie 22.04 LTS uses the 5.15 kernel, as do most of 22.04 Linux releases), but it highlights the problem with OEM/manufacturer driver support. Intel is good about getting current drivers into the kernel quickly, but other OEM/manufacturers aren't, and community drivers are catch-as-catch-can.

            My own view is that Solus should release a new ISO every six months or so, tracking the schedule of point-release distros. I think that is the best that we can expect.

            I would agree the ISO needs to be updated more frequently then it currently is.

            While I agree with @Neumie that features such as sound are not important, it should still probably be kept in mind when releasing an ISO.

            Normally when checking hardware compatibility (or debugging hardware issues), I generally suggest downloading the current ISO and testing the Live Environment. But that's not really true for Solus, we've just cross over the year mark since Solus 4.3 was released and there's no sign of 4.4 that I'm aware of.

            My thoughts on the ISO issue are two fold:

            Maybe have only one ISO that is constantly updated (bi-yearly/quarterly/whatever make sense), and have the others update on the point releases 4.4, 4.5...
            Personally I wouldn't care which ISO was chosen. As long as I know my hardware will work, I'll spend the time needed to get my DE of choice installed

            How difficult would it be to build/update "minimal" ISO that has an Installer and little to nothing else?
            I'm thinking should be possible to make the Installer able to install any Desktop Environment, as well as connecting to the repositories to download and install a fully up to date OS.
            While it would be nice to be able to do minimal hardware testing testing of hardware (sound, display...), I question the feasibility of this.

              taking what QTC said, and running with it:

              I think if I do a fresh install again, I will use a 3-year old live iso usb stick. Who needs a new iso?
              (except for those who don't have a 3-year old solus live usb).

              On paper at least an old usb or new .iso is identical: after updates they are both equal, no?

                brent Who needs a new iso?

                I dunno. Maybe anyone with a new enough computer? As long as you're installing it on a 3-year-old computer, a 3-year-old .ISO file oughta work just fine.

                  WetGeek yeah but how do you know? if current isos are giving old and new computers fits, isn't using an older live usb playing it safe?

                    brent yeah but how do you know? if current isos are giving old and new computers fits, isn't using an older live usb playing it safe?

                    Yeah, as long as it boots and your video card likes it. I've installed plenty of Solus systems on VMs using 4.2 .ISOs.

                    I never heard of the old 4.2 isos giving anyone a headache. As far as adaptability, nothings perfect as you know. Some will see a blank screen with no install no matter what the distro I think.

                      QTC Maybe have only one ISO that is constantly updated (bi-yearly/quarterly/whatever make sense), and have the others update on the point releases 4.4, 4.5 ...

                      How difficult would it be to build/update "minimal" ISO that has an Installer and little to nothing else?

                      Both are interesting ideas, and there is a long discussion (from 2019) about these concepts at Minimal option for Solus install?. It is a very interesting thread, although JoshStrobl added a "No" tag and I've heard nothing to suggest that the "No" has changed.

                      brent I never heard of the old 4.2 isos giving anyone a headache.

                      You might want to read back into the threads from 2020 or thereabouts. Discrete graphics card drivers and wifi drivers have been a perennial problem, and that hasn't changed.

                      Neumie Consider: if the RX 7000 series and Intel ARC GPUs released today and someone had them in their PC. Would the current Solus ISO be enough to get a new install completed in a smooth and trouble free way? Or would they need to pass in extra kernel parameters? Or would they be stuck looking at a blank screen?

                      Neither is supported by the ISO kernel (5.13) or the current kernel (5.15), so both would have issues running under Solus at this point. It isn't clear to me whether the 5.18 kernel supports either/both or how well. Kernel support for the newest hardware is always a problem in Linux, because OEM's aren't diligent about delivering drivers to the kernel and the kernel always lags.

                      That is not to disparage/deny your basic point, which is that the Solus ISO needs to be updated relatively frequently (twice a year, maybe) in order to keep up with the pace of hardware development.

                      Edit/Update: I am scratching my head again this morning, trying to sort out whether or not to install Solus Budgie on my railroad laptop (Latitude 7390), rather than continue to use Windows 11 on that laptop. My brain says "No" (scaling and battery life issues) but my heart says "Yes". However that sorts itself, it reminds me that Linux presents issues beyond the state of the ISO for many users, even those using vanilla hardware.

                        I thought the install experience was good. The GUI installer in the Live ISO is nicer and easier to follow than other distros (I'm looking at you Fedora). Afterwards, went to the software center and updated. It took a long while to download and install the updates, but it didn't require any intervention on my part.

                        tomscharbach You might want to read back into the threads from 2020 or thereabouts. Discrete graphics card drivers and wifi drivers have been a perennial problem, and that hasn't changed.

                        I was here, I read 'em. Had that caveat in my reply, i.e., it could happen to anyone

                        murbert that would all have to be graphics card/kernel incompatibility I suppose---I mean for the ones that can't boot the .iso...
                        ...for some reason I thought going backwards may improve that situation.