The Industry Standard - BitTorrent: Register report on protocol change is "utter nonsense" - Comments

You might enjoy my

Richard Bennett — Fri, 05 Dec 2008 17:17:13 -0500

"Didnt i read somewhere that

Richard Bennett — Tue, 02 Dec 2008 21:21:40 -0500

"Didnt i read somewhere that this new feature is off-by-default,"

No, you didn't. The new feature is ON by default, which is why DSL Reports has shown that it evades some widely-used management systems.

http://en.wikipedia.org/wiki/

Henrique — Tue, 02 Dec 2008 09:07:20 -0500

http://en.wikipedia.org/wiki/Datagram_Congestion_Control_Protocol

Even if it is not the exactly right-sized answer for BT, the development process for DCCP certainly has many "lessons learned" and "best practices" the uBT people ought to know, and should implement on BT-over-UDP, particularly in the congestion control area and traffic fairness control.

Half the traffic? Didnt i

Rock — Tue, 02 Dec 2008 06:56:37 -0500

Half the traffic?

Didnt i read somewhere that this new feature is off-by-default, unlikely to kill the internet when the portion of users that actually need it turn it on. Like those who's ISPs throttle not for 'fair use' but minimising their costs by not delivering a broadband service but still charging for one.

I don't doubt that Simon

Richard Bennett — Mon, 01 Dec 2008 21:32:57 -0500

I don't doubt that Simon Morris, like all marketing managers, has nothing but good intentions. But I think there will be unintended consequences of this move that the folks at BT and in the IETF's LEDBAT task group need to consider.

Shifting the transport for half the Internet's traffic from one protocol to another is no small change.

Richard Bennett, Alarmist

UDP + Forward Error

pinhead — Mon, 01 Dec 2008 21:14:11 -0500

UDP + Forward Error Correction = bliss.

http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.23.9399

If you bothered to follow

Phil — Mon, 01 Dec 2008 20:14:35 -0500

If you bothered to follow the link, you would see the explanation you are looking for. Here it is.

uTP is our UDP-based implementation of the BitTorrent protocol.
Normally BitTorrent is implemented on top of TCP which is the standard congestion control mechanism for the internet.
It so happens that the congestion control mechanism inside TCP is quite crude and problematic. It only detects congestion on the internet once “packet loss” has occurred – i.e. once the user has lost data and (probably) noticed there is a problem. The problems of TCP are fairly well known in technical circles, but it doesn’t get fixed as TCP is one of those protocols that is implemented in every OS, client and server, on the internet. Co-ordinating a giant upgrade is a very long process.

Because BitTorrent publishes the world’s most popular BitTorrent clients AND because these clients are talking mostly to each other (not to web servers), then we have an opportunity to detect end-to-end congestion and implement a protocol that can detect problems very quickly and throttle back accordingly so that BitTorrent doesn’t slow down the internet connection and Gamers and VOIP users don’t notice any problems. This is our objective.

A bald denial from

Golodh — Mon, 01 Dec 2008 19:41:15 -0500

A bald denial from Bittorrent, but nothing more. A sensible explanation of why The Register allegedly mischaracterizes the the protocol as "greedy" would do much to increase the credibility of this denial.

Unfortunately The Register's claim that the UDP protocol is slanted towards high-priority communication is a solid one, and for that reason alone it's a grave abuse if software like Bittorrent were to use it.

If ISPs needed any further justification to use deep-packet inspection to regulate internet usage this would be it. Just like The Register claims.

BitTorrent: Register report on protocol change is "utter nonsense"

Lincoln Spector — Mon, 01 Dec 2008 15:56:57 -0500

BitTorrent has slammed a report in The Register that suggests a planned protocol change could threaten the stability of the Internet.

According to blogger Richard Bennett, developers of the BitTorrent client uTorrent are reportedly taking steps that could slow gamers and voice-over-IP throughput, and even bring the Internet to a crawl. Writing for the The Register, Bennett described uTorrent's preference for the User Datagram Protocol (UDP) as a warning of slow times ahead.

As an Internet protocol, UDP is not as well-known as TCP, which has important traffic control features. Bennett says uTorrent's switch to UDP as its default peer-to-peer protocol could make the amount of ungovernable traffic could go through the roof. "By most estimates," according to Bennett, "P2P accounts for close to half of internet traffic today. When this traffic is immune to congestion control, the remaining half will stumble along at roughly a quarter of the bandwidth it has available today: half the raw bandwidth, used with half efficiency, by 95% of internet users. Oops."

ISPs could control the problem by slowing down UDP traffic, but Bennett argues that "such throttling will utterly destroy VoIP."

Is Bennett right? His article makes some pretty broad assumptions ("most [P2P downloaders] have a sense of entitlement where their etiquette gene should be") and offers few references. And as one Slashdot.org forum participant said of Bennett's article," the bottleneck for end users is typically the uplink on their last mile connection, so this probably won't bring the internet down or crash any ISPs, but it will make life worse for people sharing the connection."

When contacted by the Industry Standard, BitTorrent marketing manager Simon Morris described the Bennett's report as "utter nonsense," and said that the switch to uTP -- a UDP-based implementation of the BitTorrent protocol -- was intended to reduce network congestion. "It completely mischaracterizes what we're trying to do with uTP," Morris said in an email. "We're trying to roll out a protocol that is latency sensitive/performance neutral, NOT a greedy one that kills the internet."

More news, commentary, and predictions from The Industry Standard:

Analysis: ICANN: Yes, we're toothless
Analysis: DPI discussion highlights a Net Neutrality weak spot
Analysis: Bogus anti-Scientology DMCA notices sent to YouTube linked to Wikipedia user
Analysis: Deletionpedia: Where Wikipedia entries go to die
Analysis: Bad signs for Blu-ray: Free discs, cheap players, and declining market share