[sdiy] OT: sequencing over the internet

[Torbjörn Hörnfeldt]

Thanks!
This was what I was worrying(?)/thinking about but did not make explicit in my question before.

Torbjörn

----- Original Message ----- 
From: "Karl Helmer Torvmark" <karlto at go.telia.no>
To: "Torbjörn Hörnfeldt" <torbjorn.hornfeldt at telia.com>; <synth-diy at node12b53.a2000.nl>
Sent: Friday, March 16, 2001 11:52 PM
Subject: SV: [sdiy] OT: sequencing over the internet


> ----- Original Message ----- 
> From: Torbjörn Hörnfeldt <torbjorn.hornfeldt at telia.com>
> To: <synth-diy at node12b53.a2000.nl>
> Sent: Wednesday, March 14, 2001 9:51 PM
> Subject: Re: [sdiy] OT: sequencing over the internet
> 
> 
> > Not updated on problems with "collision detection" (concept right?) of USB and Firewire. Like to know more...
> > 
> > Torbjörn
> 
> USB supports a special mode, isosynchronous, which is especially made for real-time data -
> you get a guaranteed bandwidth - but no error detection/correction (this is used to transfer 
> digital audio to USB speakers, for example). Firewire supports something similar.
> In other words, they are FAR more suited to real-time applications than Ethernet-TCP/IP.
> 
> The big technical difference between USB 2.0 and Firewire is that USB is centered around the concept
> of one master (the computer) - the only message a device can send without being polled by the
> master is a wake-up token (enabling a pheripheral to wake up the computer). Firewire supports
> several masters, and is therefore somewhat more complex.
> 
> Thinking of conventional networks, with a local LAN running 100Mbps or more, I think pretty good real-time
> performance may be achieved in practice. Over the Internet? Not for a long time, especially if
> you think MIDI doesn't have acceptible timing....
> 
> 
> -------------
> Karl H.
>