Eric Miller on Sun, 12 Mar 2000 19:39:52 +0100 (CET)


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

Re: <nettime> Open Source Streaming Alliance - Now!


hi all,
there were a couple of threads on this I wanted to throw in my thoughts on...
long post, sorry 'bout that.  flame away.

At 10:07 PM 3/11/00 +0100, Matthew Smith wrote:
>RBN charges so much because they can. If development time and performance
>were a factor in the pricing, how could they have offered the RN 5.0
>server at 500$ about 3 years ago? The comparable server now costs 2000$.
>And how can apple offer the QT server for free?

Well, RBN and a Real Server being two different things...RBN a service, 
Real Server
a product, dunno that there's really a comparison there.
I'm not sure how Apple's business model is allowing them to donate the 
server as
part of OSX, honestly.  But once again, I think this is apples and oranges 
(awful
pun intended) because the Real server is a more complex application and is
geared to support a much wider range of bandwidths.  Ever since the 6.x 
servers and
Surestream with all the new G2 data types, it's understandable to me that the
development overhead is greater.  Compare that to something like the Quicktime
or Windows Streaming Media server setups, which are much narrower in focus.

> >And MP3 decoding is processor-intensive.  even more so for MPEG-standard
> >video.  my 450MHz Pentium gags on MPEG.  So the day of Palm OS decodes of
> >MP3/MPEG is a ways off...even then, seems to me that the current wireless
> >trends for Palm won't support A/V-capable bandwidth for years.
>
>well, firstgen. PPC macs (61/71/8100) can easily decode mp3s, under linux
>u can even listen to them on a 486. we here at FirstFloor are
>experimenting with embedding various control signals in mp3 streams, and
>the main reason why we use mp3 is, as our resident coder says, because it
>is "the (VW) Beetle of audio codecs" - its free, its stable and its
>hardware requirements are reasonable. its bandwidth usage is scalable, and
>the low-end is also improvable.

I wasn't so much thinking of MP3 as MPEG decoding in this specific instance,
which often still requires hardware to do right.

One of the things that's unfortunate for me (but fortunate for Real) is 
simply user base.
When dealing with corporate target audiences, we're not only dealing with a 
small install
base of MP3 decoders, but a perception that says that it's an underground 
illegal
activity/format.  It's like talking to a blank wall when we even suggest it 
to clients.

>on last thing...when u use a fone, u listen AND broadcast at the same
>time, no? so your argument about everybody broadcasting and not listening
>seems a little naive (if not even lame), especially since that is a big
>part of the appeal and the goals involved in this effort to get people to
>overcome the oneway-media-paradigm, which was enforced in the dawn of
>broadcast media, such as the radio & TV, by choice of the governments of
>the 30ties....read up on your media history, take a close look at your
>broadcast RECIEVERs (TV,radio, in case ur wondering wtf im talking about)
>and give me one good reason why they rnt TRANSMITTERs aswell....

yes, a phone is full duplex...but it's going through a circuit switched 
network,
unlike the Net, where in order to do the broadcasting that Drazen was talking
about in his original post, there would have to be automated buffering flying
all over the place for any given piece of media...the traffic would be 
horrendous
as each individual with 20 megs of media has to cache that content on a regular
basis on a large distributed network of site hosts.  That's brutal for 
bandwidth,
eh?  That's the nature of the IP beast.

And the other fallacy is the pyramid scheme nature of "everybody broadcasts,
everybody listens."  There simply isn't enough time as it is to check out 
all the
personal sites, 'zines, newsgroup posts, and works of art in a given genre even
if you had nothing else to do in your life...What happens when everyone has the
ability to flood the network with their home videos?  Network traffic chaos,
decreased signal/noise ratio, lamer content, and a less informative and useful
Net.  If that makes me an elitist, so be it.  There's nothing stopping 
anyone from
posting their content right now, if they can spare $20 a month for hosting 
and some
development time.  I don't see that as a huge barrier to communication.

At 05:19 AM 3/12/00 -0800, Amy Alexander wrote:
>On Fri, 10 Mar 2000, Drazen Pantic wrote:

> > The alternative solution, 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
>
>Agreed - with the open source development efforts, MP3 is
>holding out a great deal of promise, at least on the audio
>end of things. My big concern, though, as with most open
>source efforts, is that enough attention be paid to
>user-friendliness, or at least avoiding user-antagonism. I've
>set up plenty of servers and server software, and I still
>find Icecast and Liveice tedious to deal with.... so I'm
>very concerned that an independent broadcaster trying to
>set up his/her first server, without necessarily having a
>systems administrator handy,
>is going to be frustrated trying to work with the current
>state of most open source software.

totally agree here as well...it's been my experience that most
open-source development results in an unacceptably steep learning
curve for the products.  usually the process seems to stop at "it works",
even if that means the end user has to be doing arcane command line
activity all the time.

>Some service providers explicity forbid the operation of a server
>in the service contract. (Though I imagine one might have to
>be fairly conspicuous in one's serving to be noticed.) Anyway,
>it's still a lot tougher to become an information producer
>than an information consumer, at least regarding issues like
>these (streaming media, potentially controversial
>material) in which it is necessary or at least highly advantageous
>to control one's own server.

I still haven't gotten an answer on why we WANT everyone to be a
content server, really.  Awfully unegalitarian of me, I know, but c'mon,
even though the principle is a noble one...I think we could agree that
for every user who pipes humanity-enriching content through a
distribution setup, we'd have 20 people sending stuff that no one wants
or needs to see.

I'm not saying this as an elitist, I'm saying it as a pragmatist.  The
"barrier to entry" right now is a level of technical savvy and desire that
pushes people to pay for/set up hosting and do the legwork themselves.
The ugly reality, though, is that if we make distribution free and easy
for broadband content, we'll see two things happen:
1. the net will become an even greater morass of irrelevant personal
content that decreases it's usefulness to everyone else.  The analogy
I'm thinking of is the "Tragedy of the Commons", an ecological piece
from the 60s.  check it out, http://dieoff.com/page109.htm
2. we will pay for it.  Our ISPs pay for the backbone traffic, but currently
spread the costs out over all the users (at least in the US).  If ISP
backbone bills suddenly quadrupled because of people pushing multicast
broadband all over the place, you can bet that there will be a cost
backlash.  So much for $20 access.

Like I said before, in an ideal world the 'web provides a broadcast medium
for anyone to speak their piece.  and if you're resourceful enough, you can
do it very cheaply through budget hosting and access.  But broadband
SERVING access combined with redundant broadcasting/caching rings
closer to anarchy than a meritocracy of content.  And then no one wins.

Eric


#  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