If anyone is thinking about replying to my email last week
(particularly other "Synch or Stream"'ers), please let me know of your
intention via email, even if it doesn't contain your "final" write up (I
know Susan's planning on writing something up, anyone else?).

   I'm attaching my current draft of "business cases" for your perusal,
and, hopefully, your comments.  The Content Hosting case needs work for
sure, the others could probably use considerable tightening up.

Lynn
http://www.free-expression.org/
------------------------------------------
Case for Content Developers
---------------------------

   Free expression is a fundamental human right that deserves and
demands active protection.  One of the dangers is to rely on any
particular company (or person for that matter) to provide the means to 
freely express yourself.  Actually, dependence on corporations to
provide the means of freedom of expression is almost inherently going
to be counterproductive - see Noam Chomsky's work for examples and
reasons of why this is a troubling relationship.  Even if a
corporation begins with leadership who support progressive values, the 
purpose (as demanded by law) of a for-profit corporation is to make
shareholders money.  Leadership can (and will) come and go, only the
profit-making purpose remains a constant in the life of a
corporation.  Hence, reliance on the goodwill and beneficience of a
corporation to provide you with the means of expression and
distributing that expression is undesirable - particularly if you are
one of the disenfranchised groups or a group with an unpopular
message, who have a particularly strong need to get your message out.
Then there are non-activist artists who need to distribute their work, 
and have this point needed a corporation to provide recognition and
monetary success (at least in the US).  Empowering them by making the
means of distribution easily accessible encourages the widening of the 
art available to society, and the benefits of the diversity that will
develop.  

   Currently, many independent groups who cannot afford the means of
distribution offered by the dominant corporations (such as
RealNetworks servers) have tried to work around the limit of
concurrent stream license limits by pooling available time and
resources.  For example, the World Service project provides a
scheduler that allows one to schedule time for one's programs on a
server that has some time available for broadcast.  Another response
has been developing "public spaces", such as freespeech.org, that
centralize resources to provide public access to the tools for
distributing media on the net.  These are both well and good, but to
some extent do not address the problem directly - the problem being
that the servers are expensive, and even if affordable, can come with
restriction on use that may be intolerable.  If you want an example,
try downloading RealNetworks' "free" Basic G2 Server, and carefully
read the license that is sent to you via email.  Unfortunately, it
explicitly forbids redistribution (of the license itself) and is not
available directly on the web (as far as I can tell).  

  Another issue is the level of hardware required to run the
software.  Generally speaking, software with source available and
modifiable _is_ modified to make it run decently on older machines
(to the extent that is possible).  The goal of the Free Expression
Project is to write well-engineered software that is extremely
flexible - usually this means the software is somewhat heftier than a
more specialized piece of software would be (flexibility and
efficiency have to be traded against one another).  However, we plan
on addressing these issues in two ways: (1) encouraging adaptations of 
the software to be more efficient on low-end hardware (at least older
Pentium systems, hopefully 486 class systems as well), and (2) by
developing dynamic code optimization components that eliminate the
inefficiencies (that flexibility requires) at run-time.  Hopefully,
by having the software run on older systems, we can ameliorate the
difficulties presented by the requirement of having a computer
(e.g. in third world countries, where free circulation of ideas is of 
particularly critical importance).

Case for Researchers
--------------------
   One of the goals of the Free Expression Project is to provide an
infrastructure that makes it easy to investigate and develop new
protocols, codecs, policies, and file formats.  To this end, we will
be developing development tools (basically special purpose languages)
that generate new code that can be loaded into the existant software.
For example, to develop a text-based protocol, you should be able to
simply specify an Augmented BNF grammar (as used by IETF RFCs) and
their semantics with regard to various services.  For a fixed format
protocol (such as the Real Time Protocol), you should be able to
simply specify how the information is stored, and the semantics
(i.e. how different services use the information).  
   Most of the system will be accessible through a GUILE interface
(i.e. Scheme) to facilitate writing and testing new policies.  By
policy, we refer to the separation between policy and mechanisms.  For 
example, you should be able to write a buffering policy (in scheme)
that determines how much information is buffered, in what order it is
released, etc.
   Other areas of research will find this project of interest as
well.  We plan on writing a garbage collection system tailored to this 
system.  Another subproject will be writing a run-time code optimizer
that eliminates objects once they have been fully specified.  For
example, in our server, a particular streaming thread will be composed 
of a protocol module (e.g. RTSP or RTP), a service
(e.g. video-on-demand or teleconferencing), and buffer objects (one for
the network, one for the file).  Once all these are known, the code
should be composable in place to eliminate the objects - the resulting 
code could then be stored for later use (rather than re-optimizing the 
same set of components over and over).
   For codec developers (and mathematicians) we will develop a general 
harmonic analysis toolkit that uses compiler technologies to provide
optimal code.  As part of the toolkit, we will provide some basic
bases to work with, such as wavelets.
    There will also be room for working on other areas, such as
intelligent formation of server coalitions (to save bandwidth), secure 
transmission paths, and e-commerce technology.

Case for Content Hosting Services
---------------------------------
   Does your business depend on the license of another?  Are you a
small business facing competition from large corporations?  If so,
then you would undoubtedly benefit from having the option of streaming 
media software under a free license.  You are probably already
familiar with the benefits of having the Apache web server available
under a free license.  Most ISP's probably would not have been able to 
get started without the combination of a free OS (such as Linux or
FreeBSD) combined with Apache and other OSS software.  The
availability of such software enhances the services you can provide to 
your customers for a reasonable price, while giving you the highly
touted reliability that OSS is known for.  How much benefit would
your customers get if you were able to offer them streaming media
tools - servers, encoders, and players - unencumbered by EULA's and
per-license pricing structures?  How about being able to customize the 
software to fit your business?  Providing support/consulting services
to your customers?
   
Case for End Users
------------------
   If you're like most users, you probably haven't read the End User
License Agreements that accompany most of the software you purchase.
But you should read them, and see what kind of restrictions the
vendors you purchase from think they can apply to you.  They are
usually quite intrusive, or would be if followed.
   One particularly important area of your life to keep free of
corporate and government intrusion is communications software.  The
Free Expression Project is dedicated to building software that
facilitates your using your computer to communicate what you want, how 
you want, to whom you want, and _only_ to whom you want (e.g. through
encryption of your data).  We have no interest in invading your privacy 
to further our corporate interests.  Since the source code of our
programs is open and free, it would indeed be difficult for us to
include privacy invading "features" without some programmer
discovering and loudly complaining (and offering a version without
said "feature").  This does not preclude, however, content developers
from "watermarking" their product in non-obvious ways.  
    We also try to assure that our software will
run on your platform of choice, be it Windows, MacOS, UNIX, or some
system hardly anyone else knows exists.  If it doesn't, you can always 
have it modified to run on your system - we won't complain.  You can
even give it to other people you know using that system - we still
won't complain.  That's what our software is for - to enable you to
communicate with others, through whatever media type you choose.  
   Furthermore, we will attempt to make sure it runs on even somewhat
obsolete hardware, by encouraging developers to test it out, and
perhaps having them develop special versions.  Hopefully this will
extend back to 486 class hardware (386 class hardware will probably
just not be able to support the computational demands of the software
in any form, though we'd be happy if someone proves us wrong).  The
key to success for us (and you) is that none of our information is
proprietary - that is, no one has to pay or ask for permission to see
how our software works.


 |   |   |   |   |   |   |   |   |   |   |   |   |   |
(a) (c) (o) (u) (s) (t) (i) (c) ( ) (s) (p) (a) (c) (e)
 |   |   |   |   |   |   |   |   |   |   |   |   |   |
information&comunication channel | for net.broadcasters
http://xchange.re-lab.net  (Xchange)  net.audio network
xchange search/webarchive: http://xchange.re-lab.net/a/