[LINK] More Geist: Myths and Fallacies About Usage Based Billing

Tom Koltai tomk at unwired.com.au
Sat Apr 2 13:38:23 AEDT 2011



> -----Original Message-----
> From: link-bounces at mailman.anu.edu.au 
> [mailto:link-bounces at mailman.anu.edu.au] On Behalf Of Kim Holburn
> Sent: Saturday, 2 April 2011 11:16 AM
> To: Link list
> Subject: Re: [LINK] More Geist: Myths and Fallacies About 
> Usage Based Billing
> 
> 
> 
> On 2011/Apr/02, at 1:19 AM, Paul Brooks wrote:
> 
> > On 1/04/2011 12:25 PM, Kim Holburn wrote:
> >> He reiterates 3 important points here that we could do 
> well to note:
> >> 
> >> http://www.michaelgeist.ca/content/view/5718/125/
> >> 
> >>> Myths and Fallacies About Usage Based Billing
> >>> 
> >>> Thursday March 31, 2011
> >>> Bill St. Arnaud has posted a paper on usage based billing that 
> >>> challenges some of the frequently made claims on UBB. St. Arnaud 
> >>> notes that the paper demonstrates three important facts: 
> 1. Internet 
> >>> video streaming services actually reduce costs for 
> Internet backbone 
> >>> networks operated by telephone and cable companies, even 
> as traffic volume grows; 2. There is no correlation between 
> volume of Internet consumption and costs for telephone and 
> cable company last mile providers and that congestion, if 
> any, is more of an artifact of design assumptions made by the 
> operators; and 3. Cable and telephone companies operate 
> competing video streaming services over the same last mile 
> infrastructure used for Internet access services, which 
> generally are not priced based on usage, and yet somehow seem 
> to able to avoid congestion as well as provide the service 
> for fraction a price of what they charge for delivery of the 
> same video content delivered over the Internet.
> > Unfortunately, the paper and the conclusions seem to be based on 
> > underlying assumptions that do not hold in Australia for 
> the majority 
> > of ISPs, and describes the architecture of an incumbent, 
> local access 
> > network owner's retail services, where the provider owns 
> outright all 
> > long-haul transmission infrastructure, and does not need to lease 
> > bandwidth in any region of the network. The underlying 
> assumptions are 
> > not true for an ISP that uses any of the Australian wholesale DSL 
> > network services to reach its customers, and probably won't 
> hold for 
> > NBN-connected RSPs.
> > 
> > If the underlying assumptions are incorrect, the three 
> conclusions are 
> > suspect.
> 
> I'm not sure what you mean here.  I live in Canberra and for 
> a while had a Transact connection.  The third point is 
> exactly like Transact, they offer TV and VOD as well as 
> internet.  Can you get internet through cable in Australia anywhere?

Both Optus and Telstra offer cable based internet access.

> I can't see how point 2 is not a completely general point 
> regardless of country.  I'm not sure I understand what he 
> means in point 1.  
> 
> > 
> > Its too late to create a full analysis tonight however. I 
> suggest you 
> > read the full paper before accepting the conclusions as gospel in 
> > Australia.
> 
> OK.
> 

The Koltai interpretation of Arnaud's comments. [Great speaker BTW]

1. Internet video streaming services actually reduce costs for Internet
backbone  networks operated by telephone and cable companies, even as
traffic volume grows; 

Multicast services that service several simultaneous users obviously
decrease capital works costs.
Youtube/TVU/Boxee/Miro/Vimeo push users towards the most popular content
to assist in attaining multicast bandwidth savings.
(Youtube users viewing the top ten would notice the speed increase which
degrades rapidly as one travels down the top 1000 YouTube videos
watched.)

P2P shared content (that's the technology not illegal file downloading)
between users obviously reduce capital works costs and is the basis of
the new Telstra delivery model.

Akamai CDN's are almost ubiquitous with the first thirty seconds of all
sponsored video content cached.

And of course, whilst users are all watching videos they cant play those
stupid bandwidth hogging Ajax Facebook/Tencen/QQ/Zynga games.

2. There is no correlation between volume of Internet consumption and
costs for telephone and cable company last mile providers and that
congestion, if any, is more of an artefact of design assumptions made by
the operators; and 

The copper was paid for on installation.
The Cable Headends were installed with state of the art technology (C&W
in 1996 and ). That of course has now been supplanted and needs to be
upgraded.

The increases in capacity on cable, copper and fibre over the last few
years has meant that only the switching equipment, the headends and the
CPE need replacing, the transit technology remains as it was when first
installed.

3. Cable and telephone companies operate competing video streaming
services over the same last mile infrastructure used for Internet access
services, which generally are not priced based on usage, and yet somehow
seem to able to avoid congestion as well as provide the service for
fraction a price of what they charge for delivery of the same video
content delivered over the Internet.

This to me would seem to be the multicasting and P2P portion of 1.
above.

References:
Background Reading
The CAN http://www.moore.org.au/comms/04/04_comms.htm

/body




More information about the Link mailing list