2 cents on the low latency for MM thread...

From: Mike A. Harris (mharris@meteng.on.ca)
Date: Sun Jul 09 2000 - 07:14:08 EDT

  • Next message: Mike A. Harris: "Re: [linux-audio-dev] lowish-latency patch and toolchain"

    I believe that the whole low latency issue has pretty much been
    talked to death and that all parties have pretty much given their
    points, and nobody is going to budge on their thought processes.

    Linus has stated his piece allready, so it is pretty much written
    in stone.

    So, since it has allready been stated that Linus will not include
    Ingo's patches - WHY BOTHER ARGUING ABOUT IT? If ANYTHING, it
    will just piss off Linus and other developers, and not accomplish
    ANYTHING AT ALL for your cause.

    Linux is all about "show me the source". In other words, if
    nobody is willing to write what YOU want, then you need to either
    write it YOURSELF, or get someone else to do so for you, or HIRE
    someone to do it.

    In order for ANY of that code to get included in the FORMAL
    kernel, it will need to get past numerous 'code goons' (spoken
    affectionately of course) and finally Linus. This procedure has
    been occuring since 1991 and wont change soon. It has proven
    quite effective in keeping the QUALITY of code extremely high in
    the Linux kernel. It is also to note, that regardless of any one
    individual's motives or operating system needs, the QUALITY of
    code in the kernel ALWAYS comes before the QUANTITY.

    In other words, if code is not deemed the right way of doing
    things, or comes with bad side effects - such as poor
    maintainability, then it just will NOT EVER get included PERIOD
    regardless of the amount of complaining, whining, discussion, or
    threats of using another operating system. Nobody that uses
    Linux cares 2 shits, if someone threatens to use BeOS, Mac,
    Windows or any other operating system because Linux "can't do
    what they want." Other operating systems exist just for that
    purpose.

    So, now that we understand that unless a miracle occurs, the
    official Linux kernel will NEVER do 100% hard real time, we can
    move on to bigger and better things.

    How can audio people get their software for Linux then? By
    following other approaches. I myself would LOVE to use Cakewalk
    for Linux, and many other audio apps, including multitrack
    recording, mixing and digital effects, midi, etc.. NOT at the
    expense of turning Linux into a blob operating system. I don't
    want my system slow or kludgy just so my audio programs run
    well. That is rediculous.

    Here are thus some suggestions for the audio people and companies
    to persue for their cause:

    1) Patch the kernel on your own, or devise your own kernel
       addons, etc.. and distribute your OWN CUSTOM Linux
       distribution based upon RedHat or some other popular
       distribution. Call it "Audiophile Linux". Since you have the
       muscle with the audio software companies, it shouldn't be
       difficult to get it going, and also it shouldn't be difficult
       to get a niche market of customers interested in your low
       latency version of Linux. Recommendation is to come up with a
       fancy name illustrating the dist's high points of audio.

    2) Use RTLinux, and distribute it as above.

    3) Make a mutually beneficial business deal with RedHat or
       another popular distribution to distribute the 'audio-kernel'
       with the distribution as an install time choice for audio
       folk, or to create a custom dist for you and put their stamp
       on it. You could INCLUDE the audio apps such as Cakewalk, etc
       WITH the distribution. This would likely entice would-be
       users into purchasing the distribution.

    Brainstorm mixtures of my suggestions above.. I came up with
    them fairly quickly, so I'm sure that numerous other
    possibilities exist. So long as the kernel interfaces used are
    compatible between the audio kernel and the normal kernel, the
    software will not be stuck to a small crowd so long as the low
    latency patches are maintained by someone outside of the
    mainstream kernel. Perhaps you could supplement Ingo's income
    for doing so.

    Just some suggestions and food for thought to try and see this
    problem come to an end. I too would love to see professional
    audio apps running under Linux some day, however not until it is
    done CORRECTLY. That probably means a slight fork for this
    special purpose which is definitely NOT general purpose.

    Good luck with your solutions!

    Take care,
    TTYL

    -- 
    Mike A. Harris                                     Linux advocate     
    Computer Consultant                                  GNU advocate  
    Capslock Consulting                          Open Source advocate
    

    ... Our continuing mission: To seek out knowledge of C, to explore strange UNIX commands, and to boldly code where no one has man page 4.

    - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/



    This archive was generated by hypermail 2b29 : Sun Jul 09 2000 - 07:17:43 EDT