Announcement

Collapse
No announcement yet.

Systemd 199 Has Its Own D-Bus Client Library

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

  • #76
    Except that what LP fans call "Features" Is actually little more than bloatware. Literally, in just about every case where a "feature" is touted it is usually completely unnecessary for almost everyone. It seems like systemd wants to do 100% that 0.1% of the userbase needs.

    Its an extreme case of trying to cover edge cases at the expense of the majority cases.

    Comment


    • #77
      Originally posted by frign View Post
      I couldn't have put it better!
      The day will come where will realise how important sucklessness and smartness is in software. When Lennart Poettering is so good at puking code, he should start working on the Kernel and stop trying to reinvent the wheel with bloated user-space-"solutions" attempting to fix problems which could be solved more efficiently or aren't even existent, but I guess this would blow his scope, because you actually have to take care of making your code work once you enter Kernel-space.
      Same with systemd: It generally is a collection of tools already existent and very efficient with OpenRC and others. When systemd-zombies attempt to push you to the wall stating the feature-richness of systemd, then they do it in attempt of F.U.D. without knowing these features have been existent in the current
      solutions _for years_!

      I am glad to read I am not the only one opposing this guy and his agenda so harshly.
      Its not that we "Systemd zombies" dont know these features didnt exist already in other programs, systemd just made them easier to use and more consistent / streamline.

      Socket activation of services? Xinetd. But its still service activation...why shouldnt the init manager do that? Why add yet another piece of the puzzle when the fundamental goal of that piece is already covered by init?

      Resource management? Technically that was handled by nice levels...but nice levels are fundamentally broken in the modern age of mutli-process programs. See my apache example above.

      Poweroff, suspend, hibernate, restart, boot? All of that was handled by a combination of init, kernel and acpid. Seriously, we had a tiny little daemon that was barely maintained and who's only goal was broadcast "POWER_BUTTON_PUSH" "SUSPEND_PUSH" "LID_CLOSE" "LID_OPEN" why not just let the kernel or init handle it and be done? You could also argue that it was PM-Utils job to handle suspend, but that was only the case because of broken hardware--especially video. Now drivers have been fixed, suspend is less of an issue and it can go back to being the kernel and inits job. (Why init? because you gotta make sure services come back up cleanly)

      ^Addendum to that, in a non-systemd world if you run Gnome, KDE, XFCE and have acpid running on the system, you'll get double events. One of the nice things about DBUS is PM-inhibitors. 1 process gets the job of handling ACPI events, thus preventing double-events.

      general logging in / logging out? Traditionally handled by ConsoleKit and PolicyKit and before that, i guess...bash? Not sure, I wasnt around in the FOSS community back then lol. This is where the break between the systemd component, and the systemd umbrella name is made. Logging in / logging out isnt actually handled by systemd, its handled by logind. Gracefully user-switching, graceful TTY spawning, mutli-seat if you need it.

      Hostname and timezone? That was mostly about just consolidating all the distros random BS changes. Theres no reason Debian, Ubuntu, Fedora, Arch and Gentoo all needed different places to set their hostname

      Comment


      • #78
        Originally posted by duby229 View Post
        Except that what LP fans call "Features" Is actually little more than bloatware. Literally, in just about every case where a "feature" is touted it is usually completely unnecessary for almost everyone. It seems like systemd wants to do 100% that 0.1% of the userbase needs.

        Its an extreme case of trying to cover edge cases at the expense of the majority cases.
        Systemd has the same goal as the kernel: To be scaleable in all directions. To be able to be used on a router to a smartphone to a desktop to a supercomputer. If you want the features they are there, if you don't then you just pass a compile time flag to ./configure and it doesnt compile what you dont want. Watch Lennarts talk from linux.conf.au I think it was called "Systemd: two years later"

        Comment


        • #79
          Originally posted by frign View Post
          When Lennart Poettering is so good at puking code, he should start working on the Kernel and stop trying to reinvent the wheel with bloated user-space-"solutions" attempting to fix problems which could be solved more efficiently or aren't even existent, but I guess this would blow his scope, because you actually have to take care of making your code work once you enter Kernel-space.
          You do understand that those kernel features don't really mean anything if they aren't actually used by the userspace? systemd exposes and takes use of many of the modern kernel features like namespaces, seccomp, cgroups and soon kdbus. To my understanding one the problems with PulseAudio was that it didn't try to workaround problems in the kernel (ALSA drivers reporting false timing information) and therefore the drivers were fixed. Now it's pretty much the same with systemd; if they need something from the kernel they request it to be added or write it themselves (kdbus, firmware loading...).

          I hope you do also understand that Lennart wrote PulseAudio when he was still a stundent on his freetime. It however was widely adopted by free software community... because you know maybe because it was actually needed? It's nowadays used by pretty much every single desktop Linux distribution and also mobile and IVI platforms like Maemo, MeeGo, Tizen, webOS, Ubuntu Touch...

          Originally posted by frign View Post
          Same with systemd: It generally is a collection of tools already existent and very efficient with OpenRC and others. When systemd-zombies attempt to push you to the wall stating the feature-richness of systemd, then they do it in attempt of F.U.D. without knowing these features have been existent in the current solutions _for years_!
          You couldn't do worse job at advertising OpenRC. Do you honestly think that projects like MeeGo, Tizen, Fedora, openSUSE, Red Hat, Arch Linux, GENEVI Alliance and embedded industry at large had adopted systemd if they didn't have very good reason to do so? Sure systemd takes best from launchd, Solaris SMF and Upstart but the point is that the end result is better than anything of the past projects could do.

          Originally posted by frign View Post
          I am glad to read I am not the only one opposing this guy and his agenda so harshly.
          You seriously got to be fucking kidding me. The anti-Lennart hate is essentially a meme, the guy has gotten death threaths, there has been petions wishing him to stop developing software, his wikipedia pages have been vandalized, some people seem to think that his projects are worth critizing just because they were written by him...

          Comment


          • #80
            Originally posted by cjcox View Post
            I'm sort of waiting for Linus to speak up... and he will when his Linux distro of choice stops working well....
            Q: What's your view on systemd? [...]

            Linus Torvalds: I actually like a lot of what systemd does. My personal biggest issue with systemd is: the people involved seem to think that change is good for it's own sake. I've seen Lennart Poettering, for example, talking about how something is bad because it's something that has been done for thirty years, and old is by definition bad. Which makes no sense at all to me because I'm saying if it's been working for thirty years, it's clearly doing something right. This is my standpoint while some of the systemd people have the exact opposite, which is saying ``If it's been working that way for thirty years, it's about time we changed it.'' That mentality makes me very nervous. They seem to sometimes make changes for the sake of changes and worry less about what people are used to. That's probably why systemd has generated so much negative feedback, because it takes people out of their comfort zones and doesn't feel bad about that at all. At the same time I think, a lot of what it does is interesting. So I'm a bit nervous about the development model and willingness to break things, which I think is a huge mistake, but I do think that it's showing a lot of promise.

            http://bambuser.com/v/3084584

            Comment


            • #81
              Originally posted by frign View Post
              It is part of the Kernel, because there is active support for its user-space-implementation. -- This stands in sharp contrast to PulseAudio, which is not endorsed by the Kernel; that makes me feel warm inside
              I seriously fail to see your point. ALSA has been shaped for the needs of PulseAudio for years because it's its primary user in the userspace. Most of the work has been done on fixing the timers of ALSA drivers. If there's some Jack specific code in the kernel; please point me to it.

              Originally posted by cjcox
              Just because I person can spew forth a LOT of code and create massive scope creeping projects doesn't mean he's "most notable"... unless you mean notable going both ways (positive and negative
              He's notable because many of his projects (PulseAudio, Avahi, systemd... among other things (or at Ohloh.net)) have been adopted very widely in the open source community.

              Comment


              • #82
                Originally posted by ceage View Post
                Q: What's your view on systemd? [...]

                Linus Torvalds: I actually like a lot of what systemd does. My personal biggest issue with systemd is: the people involved seem to think that change is good for it's own sake. I've seen Lennart Poettering, for example, talking about how something is bad because it's something that has been done for thirty years, and old is by definition bad. Which makes no sense at all to me because I'm saying if it's been working for thirty years, it's clearly doing something right. This is my standpoint while some of the systemd people have the exact opposite, which is saying ``If it's been working that way for thirty years, it's about time we changed it.'' That mentality makes me very nervous. They seem to sometimes make changes for the sake of changes and worry less about what people are used to. That's probably why systemd has generated so much negative feedback, because it takes people out of their comfort zones and doesn't feel bad about that at all. At the same time I think, a lot of what it does is interesting. So I'm a bit nervous about the development model and willingness to break things, which I think is a huge mistake, but I do think that it's showing a lot of promise.

                http://bambuser.com/v/3084584
                Thanks Ceage. I will agree with Linus that breaking things for the sake of breaking things is bad, but changing things isn't necessarily bad. Its a chance to experiment with a new architecture or new design idea, its a chance to try something new. We'd still be stuck in the dark ages of computing if people were so afraid of breaking things that they didnt question the status quo or try anything new. I fear the day that the motto of programmers becomes "Well its worked for 30years, so its good enough." Fuck that. What about striving to do better? What about being willing to take risks and experiment?

                Comment


                • #83
                  So I take that to mean that systemd is a risky experiment... OK, I can dig it.

                  The question then is why the hell are so many distro's so quick to adopt such a risky experiment? It seems like a wave of retardedness has overrun our world.

                  Comment


                  • #84
                    Originally posted by Teho View Post
                    You do understand that those kernel features don't really mean anything if they aren't actually used by the userspace? systemd exposes and takes use of many of the modern kernel features like namespaces, seccomp, cgroups and soon kdbus. To my understanding one the problems with PulseAudio was that it didn't try to workaround problems in the kernel (ALSA drivers reporting false timing information) and therefore the drivers were fixed. Now it's pretty much the same with systemd; if they need something from the kernel they request it to be added or write it themselves (kdbus, firmware loading...).

                    I hope you do also understand that Lennart wrote PulseAudio when he was still a stundent on his freetime. It however was widely adopted by free software community... because you know maybe because it was actually needed? It's nowadays used by pretty much every single desktop Linux distribution and also mobile and IVI platforms like Maemo, MeeGo, Tizen, webOS, Ubuntu Touch...

                    You couldn't do worse job at advertising OpenRC. Do you honestly think that projects like MeeGo, Tizen, Fedora, openSUSE, Red Hat, Arch Linux, GENEVI Alliance and embedded industry at large had adopted systemd if they didn't have very good reason to do so? Sure systemd takes best from launchd, Solaris SMF and Upstart but the point is that the end result is better than anything of the past projects could do.

                    You seriously got to be fucking kidding me. The anti-Lennart hate is essentially a meme, the guy has gotten death threaths, there has been petions wishing him to stop developing software, his wikipedia pages have been vandalized, some people seem to think that his projects are worth critizing just because they were written by him...
                    It renders your argumentation quite invalid considering there are good reasons to hate Lennart Poettering.

                    Comment


                    • #85
                      Originally posted by duby229 View Post
                      So I take that to mean that systemd is a risky experiment... OK, I can dig it.

                      The question then is why the hell are so many distro's so quick to adopt such a risky experiment? It seems like a wave of retardedness has overrun our world.
                      Because apparently the distros like the benefits and the design that come with systemd. Its an experiment, yes, but its an experiment that people are liking apparently. One of two things is gonna happen.

                      Either systemd is gonna work and the linux world will be better for it. Or it won't, and the developers and users will learn something from it to take into the future. Either one is good, a mistake is only a mistake if you don't learn from it.

                      Comment


                      • #86
                        Originally posted by frign View Post
                        It renders your argumentation quite invalid considering there are good reasons to hate Lennart Poettering.
                        The fact that Lennart writes free software is a good reason to hate him as a person? That's just sad.

                        Comment


                        • #87
                          Originally posted by duby229 View Post
                          So I take that to mean that systemd is a risky experiment... OK, I can dig it.

                          The question then is why the hell are so many distro's so quick to adopt such a risky experiment? It seems like a wave of retardedness has overrun our world.
                          Indeed, everyone is insane and you're the only one who is not. That makes total sense.

                          Comment


                          • #88
                            Originally posted by Ericg View Post
                            Because apparently the distros like the benefits and the design that come with systemd. Its an experiment, yes, but its an experiment that people are liking apparently. One of two things is gonna happen.

                            Either systemd is gonna work and the linux world will be better for it. Or it won't, and the developers and users will learn something from it to take into the future. Either one is good, a mistake is only a mistake if you don't learn from it.
                            Yeah its a mistake that people have to deal with until then unfortunately. And thats the whole problem with LP fans. They seem to think that "This is the retarded way that I am doing things, so you have to do the things my retarded way at least until there is an uprising and we can't make you anymore."

                            Comment


                            • #89
                              Originally posted by bkor View Post
                              Indeed, everyone is insane and you're the only one who is not. That makes total sense.
                              Oh its faaaar from everyone and just me.

                              Comment


                              • #90
                                Originally posted by frign View Post
                                It renders your argumentation quite invalid considering there are good reasons to hate Lennart Poettering.
                                Reasons such as that when you ask Lennart a technical question, he will give a technical answer. When you have a technical discussion, he won't just resort to attacking you as a person immediately, he will first explain technically why you're stupid, then say out loud that you're an idiot.

                                Loads of reasons to hate him indeed!

                                Comment

                                Working...
                                X