Announcement

Collapse
No announcement yet.

SphinUX OS Claims To Be ~150% Faster Than GNU/Linux

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

  • Originally posted by ChrisXY View Post
    Who is the author of that? This pretends to be from someone not involved with SphinUX but in many places it seems the author has no clue and just makes stuff up vaguely related to what you would find with a very superficial google search.
    Do you remember "Get the facts" campaign by Microsoft? At least MS bothered to pay "an independent company" to do the tests. And also at least MS compared parts of performance where it knew that the Windows Kernel was particularly strong.

    Comparing this with no facts given, it makes me think that this thread is eventually a publicity stunt and not deserved PR for the SphinuX team (not because Sphinux may be or not be faster, but because it didn't give to date a way to test their claims).

    Comment


    • oh man, that hurts

      esp, if you read through the reasons, most are simply untrue; eg. "windows has no way to detect and install new hardware automatically" ... except it has as about every noob can confirm who ever plugged in any device -> autoconfigure works pretty well and worst case is that windows update will have to search for a driver;

      same is obviously also true for nearly all other stuff on there, like the "backwards compatibility" stuff (didn't linus flame about someone breaking that not even a month ago?)

      and the "security" is just lol ^^

      Comment


      • A word from the Developers:
        Hello dear users, dear enemies, dear n00bs, and dear idiots
        well I'll be directing my speech to the users as they concern me more, I don't give a damn about others
        first off, I guess anyone can look for malicious and suspicious binaries, after our QA team explained earlier what is the oba script that sends data to the QA team, and they cleared that it's a bug reporting script, It's our turn "the developers" to troll on the some ignorant subjective bastards, well, where to begin with, first off, the claims were made by a guy who ports unity to archlinux "let's not talk about malicious Ubuntu products and introducing it to the arch users"
        well, that guy is benchmarking us, I'd love to see him complete his work, as it seems unfinished and subjective, he simply analyzed some scripts that does normal stuff (bug reporting, real mac address probing, discrete graphics card switching off/on, cpu architecture probing, etc)
        then he comes out with a brilliant idea, well If I can't disprove that SphinUX is better, let's just claim it has malicious features, classic, let's see his scripts analysis,

        note: our comment will start with a # sign

        /sbin/au
        Grabs MAC addresses of all network interfaces

        # and how is that bad ?

        /sbin/auther
        Segfaults

        # one of the main reasons why you should learn C programming language, is not embarrassing yourself when testing a binary file

        /sbin/besbes-otta
        Fake tool that benchmarks the time it takes to allocate a certain amount of memory

        # how did you find out that it's fake? see the sourcecode:
        http://sourceforge.net/p/sphinux/cod.../besbes-otta.c

        /sbin/getarch
        Prints 'x86_64' or 'i686'

        # probing the cpu architecture is a good thing

        /sbin/koko-wawa
        (Removed boost, so can't test right now)

        # too bad

        /sbin/lsx
        Symlink to /dist/sbin/
        /sbin/sau
        Sends /dist/usr/share/misc/tune to http://www.sphinux.org/stats.php, however that file doesn't exist (Screenshot: http://i.imgur.com/XcDOKk8.png)

        # deprecated script that does nothing since the file never exists

        /sbin/sendstat
        Downloads http://www.sphinux.org/56734 and does a string comparison with "SphinUS rocks other suck". The downloaded output is never used in an exec statement so command execution is not possible.

        # two words, server authentication, please refer to the previous post by the QA Team for more information

        /bin/autodriver
        Detects graphics card and installs proprietary drivers

        # this is a deprecated script since we only use FOSS drivers, but it's there in case someone needed some help installing proprietary drivers which is something discouraged by us

        /bin/lsx
        Symlink to /dist/bin/

        # I'd love to see him analyzing our recursive symlinks, basically he couldn't understand it so he skipped it

        /bin/oba
        Also downloads http://www.sphinux.org/56734 and does the same string comparison
        Sends the following to http://www.sphinux.org/bug_report.php
        date
        lspci
        lsusb
        lscpu
        lshw
        lshal
        lsmod
        dmesg
        lsblk
        /var/log/boot.log
        whoami
        xwd -root ***WARNING: This takes a screenshot of the current Xorg session***
        This script presents a fake "# Authenticating ..." text when sending the data

        # already answered by the QA team

        /bin/readme
        Prints out /usr/share/horus/scripts/readme and pipes it to less

        # what's wrong with a readme?

        /bin/xhost
        ***WARNING: Potentially dangerous: Calls "xhost +"***
        Runs "/etc/init.d/autofs start"

        # not dangerous unless you understand it, xhost + disables the access control to X server, means that anyone can connect to your X server and use it, well unless the connection is refuesd, basically this script is used only by the user to disable access control LOCALLY, means it has no network effect since you get a connection refused when you try to connect remotely, and starting autofs which
        automounts the file system, this fixes an issue that KDE has which is not being able to mount the partitions even when entering the password, it keeps asking for password everytime you try to mount any filesystem, example for same issue in archlinux:
        http://linux.derkeiler.com/Mailing-L.../msg00143.html

        /opt/Synaptics/HKLM_Kernel
        Dump of Synaptics registry entries from Windows (includes some device IDs unrelated to touchpads)
        /opt/Synaptics/HKLM_User
        Some more Synaptics registry entries
        /opt/Synaptics/**/*.so
        Lots of libraries here. I have no idea what they are for

        # touchpad, touchscreens, wacom tablets need these shared objects and registry dumps to work

        /opt/firefox
        Custom(?) Firefox build. Tarball here: http://ubuntuone.com/2Xa1ggYqd7YvN5IDtBdFi6

        # it's the default firefox with nothing more , the customization can be found under /etc/skel/.mozilla
        as supporters of the Mozilla project, we'd like to send these configurations and tarball to the mozilla team to approve it

        /usr/bin/4L-cli
        Broken symlink
        /usr/bin/4L-gui
        Broken symlink

        # 4L for those who don't know it, is the Lightscribe discs solution for linux
        see: http://www.lacie.com/download/manual...erLinux_en.pdf

        /usr/bin/au
        Same as /sbin/au
        /usr/bin/auther
        Same as /sbin/auther
        /usr/bin/disoff

        # genius

        Calls some ACPI methods to supposedly turn of the discrete graphics card

        # he didn't even try it out to see if it works

        /usr/bin/dison
        Opposite of above
        /usr/bin/edu
        echos a short description of readme, rkhunter, nmap, ip, nbtscan, besbes-otta

        # we are educating our users about the basic unix tools to protect themselves from malwares and remote attacks

        /usr/bin/getarch
        Same as /sbin/getarch, except last line has "&>/dev/null 2>&1"
        /usr/bin/mangui
        Shows /usr/share/horus/scripts/readme with kdialog
        /usr/bin/powercontrol
        Performs some power management tweaks (CPU freq, VM write back timeout, SATA ALPM, etc)

        # and how is that bad?

        /usr/bin/sau
        Same as /sbin/sau

        Bottom line is, these guys are scared by the fact that we are better yes it's too early to judge that knowing that the project is still young, and has lots of work to be done, yet this doesn't justify the flame war that is about to start

        and now we challenge you to two benchmarks :
        - bring a core2duo machine with 2 GB of RAM, install the SphinUX DVD and the Linux distribution that you prefer "with KDE please", then try to run as much applications as you could, on both platforms, the first to die or have increased degradation of performance is the loser

        - we "the developers" are using the same version of SphinUX that you say is malware infested, since you seem a good l33t h4x0r we challenge you to find out the malware itself and use it against us, you'll find that we have secrets file that includes all passwords and login data to our servers and it's located inside all developers machines under /home/$user/Documents/secrets "pretty much careless isn't it "
        so the stakes are high, either you prove that you are a loooooser fanatic to the spyware Ubuntu who couldn't have good time with it's lousy performance, so you switched to the powerful archlinux and tried porting Canonical's malicious products "I quote RMS" to it so you can have more users' private data flowing into Amazon's servers, real classic, and look who talks about malicious products based on Linux?

        now the first challenge will prove our claim, the second will disprove yours

        finally I will talk some more about the kernel, what's wrong if our OS is based on the Linux kernel? isn't Google doing so with Android? the benchmarks you are doing indicates that you have
        no understanding of how Operating Systems should be benchmarked, we are already in our way to drop the apt/dpkg into a more solid solution, we are already working on several customizations
        to the Linux kernel itself to run better in virtualized environments and have better performance and smarter drivers support, we are already doing stuff that no one wants to do anymore,
        why would you call us scammers unless you have a proof that we are doing so? why don't you just push our development forward, our work will be good for the whole GNU/Linux based technologies
        instead of doing so, a non-subjective Free software supporter who benchmarks a community product should help and co-operate with the developers, one day our work might be a good base for most Linux distributions, one day, someone
        will create forks and distributions of our work, the same way we forked and redistributed Linux, this is why the Freesoftware movement started in the first place, to allow freedom for us
        the guys who believe that knowledge should be transferred and it's not ethical to stop technologies from being evolved by the shared ideas of different people, this is what I heard from
        a guy named RMS 17 years ago, I believed in him, I supported him, and I will always support his idea as long as I live, what's wrong with allowing people from different cultures cooperate
        to achieve the better good for all of us, is it me or I am feeling that corporations have a hand in this diversion of the Linux world, I am getting the feeling that most of Linux users
        stopped caring about freedom anymore, Corporation based opensource products are not the good alternative to the community based development model, we have a large community that reviews our
        code and stops any not intended action from controlling the user, we don't have fancy corporations supporting us to milk our minds in exchange for technologies, we are already fighting Canonical
        , Rehat, Microsoft and Apple inside Egypt, we are constantly being bothered by the regime to stop us from spreading our word and reach more users, and now you call us scammers? you should be
        ashamed of yourself, you sir support Canonical that violates the user's privacy and use him to generate more profits, you sir are fighting the GPL license, Linux should be redistributed under
        GPLv3.0, and those who fights this, are the bad guys, not us, too shame that you are supporting the bad guys.
        I am sure most of users don't care about the license nor the name, they care about good quality software that doesn't violate their privacy, you care more about brands and names, and technical
        superiority of corporations, well in that case, I would like to make a comparison, when the world found out that Canonical violates the users privacy and RMS talked about it
        most of you defended Canonical being the poor angel of FOSS, when some unknown guy in phoronix calims something without a proof, and we reply with proofs, most of you ignored our reply, and became satisfied
        with his claims that we have backdoors or whatever, without finding out what is his proof, but now we are the bad guys, who tricks users into using malicious software and collect their data?
        finally I would like to add that either that guy accepts our challenge, or we'll be releasing a series of videos benchmarking Ubuntu vs SphinUX in a cam recorded HD video with all the proofs needed to prove that SphinUX is the evolution of the Linux Kernel, we'll be contacting the FSF to help us remove any closed source binaries and give us legal consultations on how to relicense our forks on the Linux kernel to be compatible with the GPLv3 to make sure that freedom will prevail, either we go aggressive and fight back, or he proves that we are scammers that violates their users

        sorry for the long word and excuse me for any grammar or typos, I wrote this while I was talking on the phone and eating lunch

        Sincerely Yours,

        Ahmed G. Elnil
        SphinUX Community Founder
        Alexandria, Egypt
        ......................

        Comment


        • Originally posted by SphinUXOS View Post
          (...) why would you call us scammers unless you have a proof that we are doing so? why don't you just push our development forward, our work will be good for the whole GNU/Linux based technologies
          instead of doing so, a non-subjective Free software supporter who benchmarks a community product should help and co-operate with the developers, one day our work might be a good base for most Linux distributions, one day, someone
          will create forks and distributions of our work, the same way we forked and redistributed Linux, (...)
          People said about SUOS are scammers and other bad words were because:
          - in the front page there were claims that could not either be tested or being tested (using common sense) did not give the numbers were claimed to be achieved
          - the description pages did not make clear how the amazing numbers can be achieved even at request
          - the statement of "<Any OS> rocks others suck" shows a disrespect of other projects you forked from, but in place you request respect from community, even these projects gave supposedly the freedom to make your work possible
          - the "comparison chart" made a lot of claims with no methodology of how things are tested. Many (!) benchmarks and papers (if this is a paper this, is a joke, because it has no quotations, references) state clearly that the runtime/compiler/OS/hardware may work different in different workloads or environments. There are no such disclaimers anywhere to be found in SUOS page! (another miss of respect)

          This brings the second part - "why cooperate with SphinUXOS"? Where is the Google Talk or a conference in clear English (if the Community Leader doesn't speak clean English, ask a student in Foreign Languages University to read a script and put slides) which shows the advantages or tradeoffs? Or a paper link with quotations where people can see what your OS is about? Or a blog, etc.

          It looks to me that this OS is bullying by bringing at least unsubstantiated claims. This maybe is not scamming, but FUD. Is to be expected that you don't justify all your claims, but most claims that are sensible, should be fairly easy to say how to test. I'm working in some(what) performance critical components, and I do make performance claims, but I test them and many of my performance claims are like this: "For locking a document it works 20% faster on average compared with version X", or on my tests with a compiler project I'm working in my free time I write a claim like: "In NBody this program is faster by 25% using Mingw 4.7 64 bit vs .Net 4.5". These statement do imply that this would not happen for all operations.

          SpOS should say: "it is 50% faster in high lock contention" or "is 30% faster in reading in parallel 30.000 files of 8K in size".

          So, it is about honesty and the claims to be provable as true. SpOs seems to fail on both, so why expect love and support even you may be faster, or whatever you claim you have?
          Last edited by ciplogic; 06-25-2013, 02:28 AM.

          Comment


          • Originally posted by SphinUXOS View Post
            - bring a core2duo machine with 2 GB of RAM, install the SphinUX DVD and the Linux distribution that you prefer "with KDE please", then try to run as much applications as you could, on both platforms, the first to die or have increased degradation of performance is the loser
            LOL - Never heard of a "Benchmark" like "We drink milk till we puke and who pukes last is the winner"
            --
            Originally posted by SphinUXOS View Post
            ....a guy named RMS 17 years ago, I believed in him, I supported him...
            Originally posted by SphinUXOS View Post
            that SphinUX is the evolution of the Linux Kernel
            When you are such a FOSS Guy why dont you contribute your "Evolution" to the Kernel? So please share your amazing, groundbreaking, revolutionary space-time beding Hardcore Kernel (you know, things like what you did and why you dont contribute to the Linux Kernel or GNU Hurd since RMS is your Idol)
            --
            Originally posted by SphinUXOS View Post
            xwd -root ***WARNING: This takes a screenshot of the current Xorg session***
            This script presents a fake "# Authenticating ..." text when sending the data

            # already answered by the QA team
            Please give a TL;DR explaination.


            Im open for new things, but everything i read and have seen looks like a 14 year old Kid forking Debian, making dubious claims with lack of Proof and says that the Fork "Rocks" and the Base "Sucks" (Therefore SphinuxOS must Suck too when all others, therefore Debian, your Base, also Sucks). Very childish. But please, give us some reason to take you Seriously.
            Last edited by pythoner; 06-25-2013, 04:28 AM.

            Comment


            • Originally posted by SphinUXOS View Post
              well, that guy is benchmarking us, I'd love to see him complete his work, as it seems unfinished and subjective, he simply analyzed some scripts that does normal stuff (bug reporting, real mac address probing, discrete graphics card switching off/on, cpu architecture probing, etc)
              How is that "benchmarking"?

              Originally posted by SphinUXOS View Post
              /sbin/auther
              Segfaults
              # one of the main reasons why you should learn C programming language, is not embarrassing yourself when testing a binary file
              How is that a good answer to a segfaulting program? Defensive programming, do you do it?

              Originally posted by SphinUXOS View Post
              Bottom line is, these guys are scared by the fact that we are better yes it's too early to judge that knowing that the project is still young, and has lots of work to be done, yet this doesn't justify the flame war that is about to start
              Bottom line is that you haven't provided any documentation that shows how any of your novel ideas are actually implemented. Or have we all missed it?

              Originally posted by SphinUXOS View Post
              and now we challenge you to two benchmarks :
              - bring a core2duo machine with 2 GB of RAM, install the SphinUX DVD and the Linux distribution that you prefer "with KDE please", then try to run as much applications as you could, on both platforms, the first to die or have increased degradation of performance is the loser
              No. I challenge you to make a reproducable benchmark.

              Originally posted by SphinUXOS View Post
              the benchmarks you are doing indicates that you have
              no understanding of how Operating Systems should be benchmarked
              Like "try to run as much applications as you could"?

              Originally posted by SphinUXOS View Post
              we are already working on several customizations
              to the Linux kernel itself to run better in virtualized environments and have better performance and smarter drivers support, we are already doing stuff that no one wants to do anymore,
              Any documentation at all about that?

              Originally posted by SphinUXOS View Post
              Corporation based opensource products are not the good alternative to the community based development model, we have a large community that reviews our
              code and stops any not intended action from controlling the user, we don't have fancy corporations supporting us to milk our minds in exchange for technologies, we are already fighting Canonical
              , Rehat,
              Actually it's completely irrelevant whether the code was developed by a corporation. If you don't like the online search from canonical at all, then you can simply fork it yourself and remove it. Freedom.

              Originally posted by SphinUXOS View Post
              ashamed of yourself, you sir support Canonical that violates the user's privacy and use him to generate more profits, you sir are fighting the GPL license, Linux should be redistributed under
              Actually he doesn't.

              Originally posted by SphinUXOS View Post
              finally I would like to add that either that guy accepts our challenge, or we'll be releasing a series of videos benchmarking Ubuntu vs SphinUX in a cam recorded HD video with all the proofs needed to prove that SphinUX is the evolution of the Linux Kernel,
              Why not simply do that anyway? That's all that was asked of you after your bold claims anyway.

              ("You" meaning anyone from the Sphinux developers)

              Comment


              • @SphinUXOS

                I did not dl your iso, but do you have got a repository or something like that with your Linux kernel patches? As the Linux kernel has GPL licence you need to publish it as soon as you provide a dl link to the binary. It would be better to test just the kernel changes and not a different system.

                Comment


                • It seems it was too good for us, we were not worthy XD
                  http://sourceforge.net/projects/sphinux/
                  "Whoops, we can't find that page."

                  Comment


                  • Originally posted by ashkbajw View Post
                    It seems it was too good for us, we were not worthy XD
                    http://sourceforge.net/projects/sphinux/
                    "Whoops, we can't find that page."
                    We've decided to remove our gem from SourceForge as they were moaning about the licence; something that we couldn't care less about.. We'll be ditching the linux kernel in the next release and maybe then we would consider uploading the project back to SourceForge

                    Comment


                    • Maybe you should also remove the "rebranded" Xen Sourcecode from your "system". It's also GPL code and based on your explanation about the module "X" is based on Xen

                      X is just a (badly) renamed "collection" of some Xen headers and sourcefiles. Look into "xmod/public/version.h" and compare it to "xen/include/public/version.h" from the Xen source. They just changed "XEN/Xen/xen" to "X/x" and did not notice that even they emailaddress of Keir Fraser is no @xsource.com

                      I would call this "fail" xD

                      And since the project was removed from sourceforge no sourcecode is available anymore and you are just violating the GPL

                      Comment

                      Working...
                      X