I’m a great fan of Andrew Odlyzko who I think you can picture as the network economics equivalent of the early Jakob Nielsen, long before he got drunk with power and started challenging beautiful waif-like designers to chicken-wire cage matchs. By that I mean Odlyzko is rather good at explaining what you at some level instinctively understand, but maddeningly no-one else higher than your pay grade does. Handily he explains these characteristics using actual facts that you can clutch onto in the next futile argument you are forced to re-justify over the self-same self-evident truths. You will still lose, because the people who believe the opposite of you are your employers, but you will go down in full righteousness instead of just bleating “but.. but…”
In this recent short note on streaming, Odlyzko explains, as slowly as he can, why content and telecom business executives (erroneously) think that streaming movies need them to demand special network-neutrality-busting queue-barging privileges on the Internet.
It turns out it’s because they think that real-time streaming means that the bytes have to move in exactly real time — that’s to say, if you’re watching a movie encoded at 6Mb/s, you need a constant, unbreakable, 6Mb/s stream over the Internet. In other words, no-one told them about buffering:
I have been asking listeners [at his networking presentations] to raise their hands if they saw any point at all in faster-than-real- time transmission of multimedia … The highest positive response rate I have observed was at a networking seminar at the Royal Institute of Technology in Stockholm, in September 2007. It was about 30%. Twice, at networking seminars at CMU and Stanford, the rate was about 20%. Usually it is far lower, almost always under 10%. And sometimes it is close to zero. I had two similar audi- ences, on two separate continents, of about 100 people in each case, consisting of (mostly non-technical) mid-level telecom managers as well as government research agency staff and others connected with communications, where among the approximately 200 attendees in all, just one hand went up, and that one very tentatively.
Apparently, advocates of streaming have never wondered what that little grey bar that preloads your YouTube clip before you watch it, so you can cope with drops in the download rate, means. Or, more likely, have never actually watched a YouTube clip in their life, and just sit behind their desks wondering what everyone in the open-plan bit of their office is giggling at.
“The purpose of data networks is to satisfy human impatience.”
August 21st, 2008 at 8:21 am
[…] O’Brien points out that this isn’t true, because you can use buffering: It turns out it’s because [opponents of net neutrality] think that real-time streaming means […]
August 21st, 2008 at 8:27 am
This is essentially a rehash of the circuit switching v. packet switching debate. it seems the IT industry is continually rehashing the same debates as technology moves on, with the younger practitioners not being aware of the debate that’s gone before.
For example, in client-server systems, do we try to do most of the work on the client or on the server? Traditional websites asre essentially thin clients, but when you have Javascript, Ajax, Flash, etc, it becomes more of a fat client.
Or, when compiling, do we compile to machine code, or to some intermediate code (such as the JVM, or Pascal p-code)