Re: [decentralization] One to many broadcast should go thru wires?
- clay said:
> Not just that. Chat is the human version of SYN/ACK. Chat's advantageYou bet. Chat is like a duck in water on the internet because of that. And audio
> is not just low bandwidth, therefore, but tolerance of widely varying
> latency, so long as its beneath a conversational threshold.
streaming on the internet is like a duck in a snowsuit: it's not going to fly no
matter how hard it flaps.
It's not that Real Audio doesn't work fine over broadband. It's that out-of-band
transmission is cheaper, more scalable, uses well established open standards like MP3
and HTTP, doesn't force me to re-download material, and allows me to listen according
to my own schedule.
- Justin said
> > What I would like to see is some economic signals back to theLucas said
> > end user that tell him streaming content costs more, so that
> > the system can be brought to equilibrium, and provide a more
> > accurate allocation of what people really want.
> I don't think thats necessary. Users will soon start to realize the
> drastically better speed and quality of cached content over real-time
> content and tend to gravitate towards the cached. Content providers will
> further push this by realizing massive bandwidth savings for cached
> content over real-time content.
> It's not that Real Audio doesn't work fine over broadband. It'sAny scheme which involves users in choosing efficient methods of getting
> that out-of-band transmission is cheaper, more scalable, uses well
> established open standards like MP3 and HTTP, doesn't force me to
> re-download material, and allows me to listen according to my own schedule
audio and video content, has an implicit incentive system somewhere.
If people can get what they want with RealAudio or its open source
counterparts, why would they look one inch further? The same goes
for content providers (i.e. those 100,000 small broadcasters.)
Beneath your plan lurks the soviet planned economy in which, for
lack of any other economic means of allocatoin, consumer goods
were allocated based on who was willing to stand in lines the longest
number of hours. That is how our highway system works too. Why
not hold your noses and let people pay pennies for bandwidth?? Is
there something morally wrong with that?
Remember, we are in an era of abundant bandwidth and it need not be
costly on a micropayments basis. Avoiding a micropayments infrastruc.
plays into the hands of centrallized power i.e. the telcos and ISPs
will pay for the whole thing by extracting money in other ways.
Wishing all those routers and fiber were really free, or trying to
steal or freeload bandwidth is a losing game. I want to win.
I will pay my $5 or whatever it costs, per month (this will tend
to keep the spammers and commercial broadcasters out, by the way)
I do honor your ideas and your accomplishments. I will not
pester you further with these ideas.
> Beneath your plan lurks the soviet planned economy in which, [...]This makes me wonder if its time to update Godwin's Law.
> for lack of any other economic means of allocatoin, consumer goodsThis would make sense if it were merely a matter of nose holding, but
> were allocated based on who was willing to stand in lines the
> longest number of hours. That is how our highway system works too.
> Why not hold your noses and let people pay pennies for bandwidth??
people don't _want_ to pay pennies for bandwidth, so any competitor
offering all-you-can-eat pricing will tend to win.
The Case Against Micropayments
> Is there something morally wrong with that?Nope, its just a practical issue.
Note that this does not mean that price cannot be used to segregate
classes of service, as in Odlyzko's "Paris Metro Pricing" model.
It just means that micropayments are doomed as the method for doing so
for any consumer app.
Clay Shirky | shirky.com - Essays on the Internet:
http://www.shirky.com/ | Culture, Economics, Globalization
- At it's core the Humancasting system I'm trying to
develop, has caching at its core. Video isn't blasted
all over the internet, but instead intelligently tries
to work its way effeciently to the user. Its one to
many without a "scattershot" approach...
--- Todd Boyle <tboyle@...> wrote:
> Dave Winer said=====
> > Little-known fact: Most of the videos you see on
> TV news are many hours if
> > not days old.
> > It's a pure tradeoff. The aha is that for some if
> not most video, a few
> > hours makes no difference at all.
> I fully agree with the goal of caching whenever it
> can *possibly*
> be achieved. e.g. music clips or nonurgent news
> Who could disagree with THAT?
> Since there are also times when realtime audio or
> is needed and desired, it will of course be achieved
> by some
> software, regardless of anything we do or say. And
> users *will* clog the pipes if nothing is done about
> What I would like to see is some economic signals
> back to the
> end user that tell him streaming content costs more,
> so that
> the system can be brought to equilibrium, and
> provide a more
> accurate allocation of what people really want.
> If they want to hear ANY music clip immediately
> regardless of
> whether it's in the cache, then, they are going to
> get it! The
> challenge for decentrallization workgroup is to find
> a way to
> provide the same content cheaper faster via cache or
> routing, etc. And to politically and economically
> isolate the
> streaming pigs so they don't consume the whole
> commons without
> paying for it, one way or another.
> IN traffic management this has never been resolved
> and in Seattle
> for instance we've decided for some decades now, to
> let the
> highways fill up with cars and let the delays
> themselves provide
> the message. That is not an optimal way of
> everybody getting
> what they need.
> In conclusion I urge developers of decentrallized
> routing and
> communication systems to build in a micro general
> ledger into
> the router so that it can conduct itself like a
> business. Its
> GL rows would charge for resources and also maintain
> just as Mojonation does.
> Why use a GL structure? Because it interfaces with
> real business
> systems transparently, and, makes every node a
> general purpose
> business transaction platform. This enables the
> exchange and
> aggregation of balance sheet balances such as
> payables and
> receivables, as rows in a table, in a database like
> This uncouples, finally, the decision to settle/pay
> them from
> the choice of a settlement provider. All of today's
> and settlment schemes are hardwired to a particular
> bank (duh).
> Thus, the byproduct of these little GL structures is
> to enable
> individuals and small businesses to buy, sell, and
> send and receive
> payments over the network outside of banks. I
> posted a detailed
> example of how ledger settlement can work on the
> tipster list
> last week, after meeting with Jeff Kandt,
Do You Yahoo!?
Yahoo! Photos - Share your holiday photos online!