Subject: Re: [linux-audio-dev] Gigasampler Clone
From: Benno Senoner (sbenno_AT_gardena.net)
Date: Fri Jun 30 2000 - 01:15:17 EEST
Just ignore the SW-patents : here in Europe we have no problems, and if you
are unhappy with US patent laws, move to other countries of fight SW-patents to
death
:-)
I think that it will eventually happen (more later than sooner) that the US will
be the only country left "believing in SW-patents".
I know that the US is trying to persuade (ENFORCE is a better term) Europe
that SW-patents are a good thing.
But they will lose this battle (see german government which declared SW-patents
dangerous for the future of the IT sector).
And developing a free sw project, why should you pay $$$ to the patent holder
when you make no money out of it, plus you wrote the code all your own
without looking at the patent.
And what makes me furious when the patent office allows IDIOTIC simple
things to be patented.
see the amazon 1click patent for example
or this from-disk-streaming stuff is just so simple that it make look the
patent office really ridiculous ..
and what does mean "realtime-low-latency triggering of samples stored on disk"
and even if the sample is COMPLETELY stored on disk without an initial part
in RAM, what happens when disk seek times get down to 1ms ?
Is then XMMS breaking the patent , because when I pressed PLAY the sound
came out of the speakers 3ms later ?
:-)
Plus what does mean realtime ?
guaranteed 100ms latencies is "hard-realtime" just as 5ms.
bah ... from now, let's ignore the patent issue and focus only on the technical
ones.
Benno.
On Thu, 29 Jun 2000, Richard A. Smith wrote:
> On Thu, 29 Jun 2000 12:36:28 -0600, Archer Sully wrote:
>
> >Does anyone have the number on Gigasampler's patent? I didn't find
> >anything matching "gigasampler" at IBM, unless they are using Invision
> >patents. Which, quite frankly, should have been rejected out of hand,
> >as they only seem to patent standard practice in software synthesis
> >circa 1990, albeit with a twist that the algorithm is selected by software
> >according to the amount of CPU cycles available. I think that if this
> >feature is left off then a system would not infringe, but IANAPL.
>
> That patent dosen't really sound enforceable to me. IANAPL as well but I have a really
> good resource if someone can get me the patent number.
>
> Plus it's probally still possible to get around the patent the way LAME gets around the MP3
> where source distribution dosen't break the patent only the binary.
>
> That of course limits distribution heavly but you could implement the actual patent
> infringing stuff as some sort of plugin that was only distributed via source.
>
> So it looks like I just started myself a project. Ick.. I was REALLY hopeing that someone
> was going to point me to work already going on but oh well. Sleep is overrated anyway.
>
> So lets start the beginnings of a framework. Ideas?
>
> WARNING: I am not a mature linux sound device developer. Just a linux hacker with a lot
> of embedded system design and programming experience.
> I am not even that much of a musician.. Piano is what I TRY to play but years of non-
> practice and not being very good to begin with leave me with nothing more than fumbleing
> through stuff I used to have memorized.
>
> I got into thinking about a linux gigasampler because I friend of mine is starting a
> recording studio and I am trying to work out a way for him to remotely set up 2 or 3
> computers running gigasampler based on some type of pre-saved config file. And it dawned
> on me that linux would be perfect for this kind of stuff.
>
> I am kind of his technical consultant when it comes to specing out computer type hardware.
> A bullshit filter if you will..
>
>
> --
> Richard A. Smith Bitworks, Inc.
> rsmith_AT_bitworks.com 501.846.5777
> Sr. Design Engineer http://www.bitworks.com
This archive was generated by hypermail 2b28 : Fri Jun 30 2000 - 02:23:34 EEST