Announcement

Collapse
No announcement yet.

What Are The Biggest Problems With Linux?

Collapse
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • #91
    Originally posted by birdie View Post
    This: Why Linux is not (yet) Ready for the Desktop (a.k.a. Linux problems), 2012 edition

    There's one problem no amount of money can solve: for most Open Source developers Linux is a playground, a thing they don't care beyond their aspirations, thus we have constantly broken features and API breakage every odd moon cycle. With such an attitude there's no way Linux will ever attract a big number of serious ISVs. Of course, people will be quick to point that already available Open Source software is enough for everyone - but that's a serious myopia. No, it's not enough, very very far from that.
    The kernel has an extremely consistent Userspace-kernel API. Regression tests are commonly done by dropping a dozen different kernel versions into the same userspace. With a few kernel config options you could drop the 3.4 kernel into a 2.15 userpace and use it.

    We've recently seen many breakage in the X server and the mesa/DRM subsystem, however these have been associated with significant performance improvements.

    Things are rarely broken just for the hell of it. I don't the the changes themselves are so problematic, so much as the fact that they are out of sync.

    And we know it's not the main issue, because FreeBSD which is careful to avoid any sort of breakage doesn't have a flock or ISV's purueing it. The main issue is simply users.
    Last edited by WorBlux; 06-11-2012, 12:30 AM.

    Comment


    • #92
      Too many package formats
      Too many differences in vital config files
      Too much jargon in UI;s
      No standard distro., every distro has a different combination of kernel, service, and library versions.

      But really I think the biggest problem from a user perspective is that nobody really uses it. If dear old grannies has a question about the computer she bought with Ubuntu, her go-to nephew for computer issues probably has no idea how to start, much less actually solve the problem. A

      Comment


      • #93
        The stable Linux distro argument. Talking about Debian not Ubuntu.

        A lot of work goes into keeping Debian stable, stable but it gets overlooked because it lacks the new features out there.

        "A couple of years wait for a new release is just too long".

        Then these same people go back to Windows wondering why they even bothered with Linux and resume again the few years wait between releases.

        Some of the more popular applications can be kept more recent by the use of backports and deb-multimedia and the like if you don't want to get into mixing with testing and/or unstable.

        Comment


        • #94
          Originally posted by chmmr View Post
          If I had infinite money to throw at helping Linux become everything it could be, here's where I would start:

          - Pay Adobe to port Creative Suite. There are vast armies of professionals who rely on these tools to do what they do. Some of them are AAPL converts for life, but many others would jump ship.
          - Pay Unity Technologies to port Unity - not just the web player, but the development application. This engine is already a major force in game development and will only become moreso over the next decade. If we have Steam and Unity we will have Gamers.
          - Send Valve a big pile of candy and flowers for doing what they are already doing

          If Adobe or Unity have technical concerns about Linux as a platform for their software, work with them to fix those. (this will undoubtedly get into the issues with open source video drivers, etc, which in turn need to have resources and help poured onto them).

          In a nutshell, work backwards from why we don't have the killer apps that keep people tied to the other 2 major operating systems. Swallow our pride and admit that Linux is already being used by people unlike us (programmers, nerds), and accept that the entire landscape will be better for this new blood, because choice (eg to use weird, power user configurations) will never go away as a core value of Linux.
          The wasteland 2 project is going to use the Unity engine. From what I understand there was a bit of work done in this direction, and they plan to collaborate to get the port to be usable.

          Comment


          • #95
            Fragmentation

            I think the main problem with Linux nowadays is fragmentation. Fedora (e.g.) obviously does its own business and does not care about what the rest of the world likes, they consider themselves as innovation leaders.

            Some changes appear like innovations first, mostly not of big use anyway, but they make rather big problems when it comes to installing software.

            A package has no chance finding out in which environment it is running. What does it need to do to install a save daemon process? Jailed or paravirtually encapsuled, or whatever? There are dozens of answers, depending on the constellation of installed packages and version numbers, and, depending on the machine, possibly all wrong since the user has decided differently.

            Comment


            • #96
              Heroes of the Computer Industry tomorrow are going to be Linux figures. Make a name for yourself by making quality productions and projects now, if you're interested in being influential in the field. Be vocal and represent your works.

              Comment


              • #97
                Smartphones are here to stay while OSes are bound to blur into the background

                I would say the biggest issue is from a user's viewpoint. Among all the computers where I have been pushing a linux distro (Ubuntu/Mint/Slitaz), the biggest turn-off was syncing a smartphone.
                And it's only the beginning. N00bs with iPhones can't stand anyway to have something more complicated to set up than iTunes. But Android smartphones... Since more and more of our digital lives are transferring from desk/laptops to smartphones, this will be a major battle.

                I really feel than most people on Linux will find it difficult to imagine what the average user mostly needs. I have been able to push for Linux on some people's computers since it has made so many advances in the past 6 years to work with lots of hardware out of the box.

                I'm waiting for the ubuntu phone, but that will be 2014. As of now I haven't found a simple interface that would help unify the linux/smartphone experience with a dashboard of checked syncs, sets of data available in the cloud/physically and working with Ubuntu One and various cloud services.

                Comment


                • #98
                  1) https://bugtrack.alsa-project.org/al...ew.php?id=1798

                  2) X.org

                  3) API changes too often

                  4) complex configuration files

                  5) too much bloat

                  Comment


                  • #99
                    So much fud and nonesense in this thread, let me atleast try to debunk some of it

                    Originally posted by birdie View Post
                    It's not that Linux is hard to learn, the real problem is that once it breaks (and it does break often) you are helpless because you cannot even say what and why has broken.

                    Nooooo, documentation has nothing to do with that. The lack of stable APIs/ABIs is what makes ISVs shun Linux like a plague. Can you imagine Microsoft intentionally breaking core libraries?
                    <snip>
                    That's like a red flag for any ISV. "Transitions? WTF??? Are you saying we cannot even rely on any libraries in Linux?"
                    Please, unstable API/ABI of the kernel is a _feature_. It more or less forces vendors to either support their stuff, or GTFO. Seriously, this doesn't need to be sugarcoated either. What vendors WANT to do, write a driver once, and hope it works for the next 10 years or so after they claim it's to old anyway. Same happens on windows too. "Oh your old printer driver doesn't work anymore on Vista? To bad, buy a new printer." Seriously? Yes that is their response/stance on hardware that still works perfectly fine. Why doesn't it work anymore? ... Guess what. API/ABI breakage.

                    How does windows/applications fix that? Delivering older versions of libraries to go with the product (which obviously only works to some extent). Windows Vista/7/8 has a 'XP compatibility mode'. What do you think that does? And why do you think every game tries to install DirectX? Or why do many programs have DLL's in next to their exe's that are installed many times, including in C:\windos? Right, API/ABI breakage and compatibility.

                    As pointed out elsewhere, you can dump a current linux kernel into quite an old userspace. And really really old binaries still work to this day on modern systems. It is not as bad as you make it out to be, so stop trolling.

                    Comment


                    • Originally posted by Setlec View Post
                      biggest linux problem? there is no standard btw the distos. on the programing point of vue, no multimedia API equivalente to DX blob. (this makes quite hard to make games for linux or port stuff from winblob) X.org is old just ditch it for crying out loud wayland could get more love and get to a faster development. PULSEAUDIO SHIT AS SOUND SERVER?!? wasn't ALSA working better? oh right b4 using alsa we used OSS that we ditched it too.
                      Multimedia API equivalent would be a combination of packages. OpenGL for graphics (which pre-dates DX), OpenAL for sound (it is actually quite heavily used in games) and SDL for 'a lot else'. With these 3 components, I'm pretty certain you can get extremely far.

                      As for pulseaudio, yes it's not perfect, but it's not a replacement to alsa. I'm not a big fan of pulseaudio, but do think it is needed. If every sound card would have properly supported hardware mixing, then the need would be far less important. It would offer features, such as instantly move input streams to output systems.

                      Alsa replaced OSS because of many reasons. For one, OSS wasn't as opensource as you wanted it to be, rather nasty stuff even.
                      Again, pulse is not replacing alsa, alsa is the audio 'framework' so to speak, and pulse still does it's output through alsa.

                      Originally posted by Setlec View Post
                      sometimes when i see a new nightly build of haiku, i always think that if GNU/linux had a single dev group the OS as a whole would be way better than what it currently is. (distros could modify it at their taste but come on the GNU/linux dev community is so fragmented that i'm currently spending more time on windows than on linux for daily routine due to stupid breakage with sound card driver or power consumption and etc..
                      A huge single dev group would be nice in some respects, yes there's too much fragmentation now. On the other hand, it's healthy evolution at work. Weaker distro's/packages fade into obsolescence and the strong survive. In the great scheme of things, it's not bad.

                      Comment


                      • For me it has to be interactivity:


                        Run a process that needs more RAM than available, and as soon as the process starts using swap everything becomes unresponsive (not just the process). If you have patience you can kill the process by ssh'ing into the box, but it usually takes like half an hour to do so.

                        Have a CIFS mount, and as soon as the mounted device becomes slow or stops sending packets, your system will become unresponsive too (GUI, any terminal window), even if you aren't accessing any files from it. You can't even unmount the device.

                        Or just click or right-click anywhere. Open windows. Click on menus, etc. Everything reacts slowly. You can verify this filming the screen with a videocamera, then later measure delays by analyzing the video.
                        You can even see the repaint of every window, and how the painting isn't uniform (i.e. frames, part of the content, etc. everything painting at different times).

                        Or compare the time to launch an app like Kcalc with the time to launch the calculator from Windows XP.
                        Or compare the time to launch a native app in Linux vs the same app ported to Windows but launched from Linux through Wine.


                        Then go to benchmarking sites like this one and cry because interactivity isn't benchmarked and nobody seems to care.
                        Also benchmarking clusters of ARM computers, but no benchmarks of 6 year old PCs. WTF.

                        Comment


                        • Originally posted by birdie View Post
                          Linux constantly evolves? OMG. You even boast about that.

                          Windows evolves, but APIs and ABIs are rock solid. On Windows 8 I can run software written for ... Windows 3.1 which was released over 25 years ago. Try this feat with Linux software.

                          Worthless points? Great, go and convince a single company to port its applications from Windows to Linux. Whereas you are theorizing, I deal with big ISVs and I know what they want from Linux. But you are free not to agree with me, just forget about Linux having more than 3% on the desktop.
                          As said above, yes it's an intended 'feature' if you will. Windows 'evolves' (it doesn't) very very slowly and breaks API and ABI's just as often. There's just more support in place to have old cruft around (Which as pointed out before, can be done in linux just as well, if you really would want that (hint, you really don't)).

                          And also again, i'm sure ISV's want to write once and forget. But no. Bad bad bad bad. Try to explain why it's bad to your ISV.

                          Comment


                          • Originally posted by e8hffff View Post
                            Biggest problems

                            * Developers need to have the mantra to make Linux/GNU the best. Latest versions Quality, quality and more quality. Speed, size, and functionality.
                            * Developers and groups wasting time and being a detriment, porting for other operating systems that are already commanding the scene. Like "Microsoft/Windows can go to Hell", needs to be the mantra. "'F' Azure", another.
                            * Lack of updates. Example slow repository updates or scope of projects.
                            * Lack of shunning system for hardware makers that don't support Linux/GNU. Example, Canon should be shunned. Buyer should be aware not to buy Canon products.
                            * Developers should also consider GUI along with CLI.
                            * Naming of concepts, projects, and programs. Some real weird names or acronyms out there and no clear explanations.
                            Just wanna brush on the naming of things. Some/most of the time, they can't use some name due to copyright bs etc.

                            Take as example Mesa. What a silly name that is? Why not OpenOpenGL? or really, just opengl? No, they where not allowed to, so they had to get creative. Sometimes they have to avoid even sounding similar to whatever it is. Like OpenGraphicsLib or OpenGPU may have gotten them in trouble. I'm guessing 90% it's politics.

                            Comment


                            • Originally posted by ua=42 View Post
                              1. Bad power consumption - most people have laptops. This should be a priority.
                              2. Audio In - why doesn't it work reliably? People use microphones for chatting over the internet. The fact it doesnt work is bad.
                              3. Stable interfaces/api/etc so that if a core component is updated to a newer version, things that depend on it won't break or become glitchy.
                              4. Stable method for video drivers (binary or otherwise) to hook up with the kernel/video system. So that we can use binary drivers with newer kernels, etc. even if the driver hasn't been updated (or support for it has been dropped)
                              5. 64bit OS should be able to run 32bit binaries.
                              6. The ability to update individual programs to their latest versions. Why do they always have to be compiled for your specific os?
                              7. HP - why do I have to go to their website, download and run their program to get all the binary bits to make the printer work properly. Why isn't this automated?
                              Audio In can be troublesome on any OS. Many inputs, many options which device is the recording device? Volume set correctly etc etc. (Yeah I have had to support audio issues on all 3 major OSes). Does it need to be improved? Yes, but generally it's hardware that's at fault if anything.

                              About stable api/abi interface for video and the rest, read my post above

                              64bit Linux can run 32bit binaries without a problem. You need to enable the option in your kernel (when compiling yourself) and need to install libraries (if not using some major distro who has it default installed).

                              And you can install many programs to their latest versions usually via PPA's (on ubuntu) or sometimes they even support packages (deb/rpm). If your complaint is that it's not in the default repo, well, they do try to keep things stable and tested. E.g. a 'known configuration' to prevent things to randomly break in places. If you need to be up to date, constantly, use a 'rolling distro' (Ubuntu IS actually talking/thinking about this).

                              Er I have had a few HP printers, all have been automatically detected by my OS's cups. I have installed hpijs at atimes and ran hp-setup (which creates config files for cups) and also that works fine. Not sure what the difference in print quality is between the two though. So this works automatically for ubuntu and gentoo to my knowledge. No need to download anything from HP.

                              Comment


                              • Originally posted by e8hffff View Post
                                Another GRIPE:
                                Lack of GUI to visualise sound channels, be that mainly for recording recording
                                Not sure what distro/tools you use, but Ubuntu does this reasonable. And depending on your recording app, that should do it for you? I know audacity does it?

                                Comment

                                Working...
                                X