Drazen Pantic on 24 Sep 2000 20:14:42 -0000


[Date Prev] [Date Next] [Thread Prev] [Thread Next] [Date Index] [Thread Index]

<nettime> Open Source Streaming Alliance


Open Source Streaming Alliance - Now!


Streaming on the Net has started to be almost the necessity of any
Internet savvy organization -- imperative is to send, and even better
stream, as much audio and video content as possible. The full agenda
is to broadcast rich multimedia content and reach the global audience,
delivering a message through a clear sound or crisp video, of course
with the minimal cost.

Everybody and everything is either a subject or medium or a creator
of the streaming content. Computer screens, clothes even skyscrapers
... any pixelized object or audio carrier is subjected - now or in the
near future - to the streaming content.

Not-for-profit and independent media organizations have been faced
with a very concrete problem: the development of the Internet has
evolved toward multimedia and streaming, while very few of these
organizations have the expertise or resources to implement high
quality audio/video streaming. They often suffer from inadequate
bandwidth, necessary to provide the number and quality of streams the
Internet audience requires. Establishing networks and alliances and
sharing resources is the the only way non corporate organizations can
address this issue.


Multicast: from Distribution Protocol to Public Domain Channel

The distributed set of audio/video streaming servers world wide,
exchanging streams through multicast and splitting technology could
provide a public domain channel for non-for-profit and non wealthy
organizations and individuals. This channel might be effective for
global streaming, without interruptions and "net congestion"
errors. Idea is very simple: when a user requests a content from one
of the servers in the network, he/she gets redirected to the closest
server relative to the Internet bandwidth topography. So, for example,
a user from New York that wants to listen to radio from Belgrade gets
seamlessly connected to a server in US, which on the other side
requests a stream from server in Amsterdam, being provided by a stream
from Belgrade. In that way, multiple users from US do not create
multiple connections to the low bandwidth server behind the
infra-structural terror in Serbia. Examples are numerous, but the basic
idea is clear, generate as less redundant traffic protocol for
streaming, and enable low bandwidth environments to rely on solidarity
in streaming from the better situated sites. 

Networking on the global level through multicast channels can create a
diversity of content, and enable global accessibility for all
voices. Advertising corporations have already realized the necessity to
bring their messages simultaneously to metropoles of the market economy
as well as to the distant villages in the underdeveloped
counties. Hence, local delivery techniques and sites are expanding
horizontally, through out the world, and vertically through the scale
of wealth.

One of the most active distributed delivery networks for multimedia
streaming is Real Network's Real Broadcast Network, bringing together
a set of distributed RealMedia servers and major global telco
providers. Implementation functions very well on the pure technical
level, but...it relies on the Real Media standard, specifications and
details are patented by Real Network, hence unknown to the rest of the
world. In their promo text RN states "Our sophisticated network
technology and management ensures high-quality and reliability,
allowing you to focus on what's important-your site and its content",
[1]. The meaning is that user(s) worldwide should be relieved from the
boring questions letting The One And Only Broadcaster to deliver their
content, regardless of the method and the cost. What brings us to the
second painful point of RBN -- the cost. Without further elaboration,
I will just cite the RBN's "Special Offer" page, [2], starting with:
"Enjoy worry-free Webcasting from RBN for as little as $9,995." RBN
example reveals the basic strategy:

- propriatery, closed standard;
- free distribution of end user clients;
- expansive content creation tools, servers and networking mechanism.


The alternative viable solution is establishing an Open Source
Alliance of Internet centers and sites that will exchange links and
streams using Open Source and/or freeware streaming solutions. The
basic premises should be:

* Open Source content creation tools and delivery mechanism;
* Totally voluntary and free membership in the alliance, without
contacts or centralized body;
* Sharing of technical and infra-structural resources and using the
collective intelligence in solving local problems of each individual
member of the Streaming Alliance.


Recent Examples

An example, at least in delivering audio content is available and not
very difficult to conceptualize. MP3 standard has reached incredible
popularity, combining the quality of sound and open source
approach. Collective intelligence of code creators worldwide, has
produced numerous tools: encoders, rippers, and streaming servers -
mostly like freeware software, available to everybody free of
charge. Moreover, open source standard has put MP3 in a perfect
position of total platform independence; any machine that has a
computer chip can play a MP3 file: from PalmPilots to mainframes
through stand alone devices. Servers are also available, look for
example [3] or [4].

Internet radio stations that use MP3 are just enumerable, and are
making their networks easy, exchanging content and streams. Temporary
global networks are being established every day, over various
occasions. Recent attempts of the broadcast industry to contain
distribution of MP3 files, resulted in the increasing popularity of
the distribution networks among individual users and non for profit
groups as well as with strategic planers in headquarters of the
transnational telco companies.


Recent developments show that streaming of video is not much more
complicated then audio. Traditional obstacles: expensive equipment and
content creation tools as well as insufficient bandwidth, as about to
be behind us. Very important initiative in that direction is Apple's
Open Source project, [6]. Darwin streaming servers for QuickTime have
proven be very stable, sophisticated and capable of a relatively
simple technology for exchange of streams. On top of that Darwin
servers are available for all operating systems, offering full
scalability. Content creation tools are also available as open source
and freeware Mbone software; see [7] and links therein.  Mbone
videoconferencing tools, developed in research laboratories in late
nineties, offer open source tools for live encoding, sending unicast
streams towards Darwin servers for further delivery. A full
description of using Mbone tools for live encoding and linking with
Darwin servers through split stream techniques is available on [8].


On the infra-structural side, things look also very good. The number of
non for profit groups and individuals world-wide have established
Internet servers with significant - or more precise with sufficient
bandwidth - to create a mighty network of streaming servers. (For
global distribution of the Internet bandwidth and traffic, see [5].)
Multicast technology is available globally through ISPs while all major
vendors offer routers and related equipment capable of supporting
multicast. Providers on the other side are not yet aware of the
multicast advantages, so some of them have additional charges for
customers that want to use multicast. But, commercial viability will
certainly force ISPs to realize that multicast in fact saves bandwidth
and reduces their costs. 

So, there is just one step needed: the will and awareness for creation
of the flexible and global network of distributed streaming
servers. The servers might just agree to exchange streams, and
establish a protocol for redirection of users. All that is needed is a
functioning example of sites and servers that mutually agree to
exchange streams and install software and pointers towards each
other. The first step is already made: an OpenStreamingNetwork mailing
list, [9], and it is certain that group of centers signed in to the
list will establish the network very soon. Now?




[1] http://www.realnetworks.com/rbn/about.html
[2] http://www.realnetworks.com/rbn/promos.html
[3] http://icecast.org
[4] http://www.shoutcast.com
[5] http://www.internettrafficreport.com
[6] http://www.publicsource.apple.com
[7] http://www.research.microsoft.com/research/BARC/mbone/MboneTools.htm
[8] http://pages.nyu.edu/~dp51/qt
[9] http://www.egroups.com/group/OpenStreamingNetwork

#  distributed via <nettime>: no commercial use without permission
#  <nettime> is a moderated mailing list for net criticism,
#  collaborative text filtering and cultural politics of the nets
#  more info: majordomo@bbs.thing.net and "info nettime-l" in the msg body
#  archive: http://www.nettime.org contact: nettime@bbs.thing.net