The big tech policy news yesterday was Comcast’s announcement that it will stop impeding BitTorrent traffic, but instead will respond to network congestion by slowing traffic from the highest-volume users, regardless of what those users are doing. Comcast also announced a deal with BitTorrent, aimed at developing more effective ways of channeling peer-to-peer traffic through networks.
It may seem natural to respond to a network issue involving BitTorrent by making a deal with BitTorrent – and much of the reporting and commentary has taken that line – but there is something odd about the BitTorrent deal, which only becomes clear when we unpack the difference between the BitTorrent protocol and the BitTorrent company. The BitTorrent protocol is a set of technical rules used by desktop software programs to coordinate the peer-to-peer distribution of files. The company BitTorrent Inc. is just one maker of software that uses the protocol – indeed, it’s a relatively minor player in that market. Most people who use the BitTorrent protocol don’t use software from BitTorrent Inc.
What this means is that changes in BitTorrent Inc’s products won’t have much effect on Comcast’s network. What Comcast needs, if it wants to change conditions in its network, is to change the BitTorrent protocol.
The problem is that you can’t negotiate with a protocol, for the same reason that you can’t negotiate with (say) the English language. You can use the language to negotiate with someone, but you can’t have a negotiation where the other party is the language. You can negotiate with the Queen of England, or English Department at Princeton, or the people who publish the most popular dictionary. But the language itself just isn’t the kind of entity that can make an agreement or have an intention.
This property of protocols – that you can’t get a meeting with them, convince them to change their behavior, or make a deal with them – seems especially challenging to some Washington policymakers. If, as they do, you live in a world driven by meetings and deal-making, a world where problem-solving means convincing someone to change something, then it’s natural to think that every protocol, and every piece of technology, must be owned and managed by some entity.
Engineers sometimes make a similar mistake in thinking about technology markets. We like to think that technologies are designed by engineers, but often it’s more accurate to say that some technology was designed by a market. And where the market is in charge, there is nobody to call when the technology needs to be changed.
Will Comcast and BitTorrent Inc. succeed in improving the BitTorrent protocol? Maybe. But it won’t be enough simply to have a better protocol. They’ll also have to convince the population of BitTorrent users to switch.
UPDATE (April 2): A reader points out that BitTorrent Inc bought uTorrent, one of the popular client programs implementing the BitTorrent protocol. This means that BitTorrent Inc has more leverage to force adoption of new protocol versions than I had thought. Still, I stand by the basic point of the post, that BitTorrent Inc doesn’t have unilateral power to change the protocol.
P.S.: I have never used p2p or any other mass mailing or other bandwidth-hungry softway. Like most private Emailers, any attached files are limited to “links”, *.docs, *.xls, and downsized & compressed *jpgs.
Blocking of my IPS by comcast and others was arbitrary. Occasionally, upon request for “unblocking”, it was granted, but immmediately re-applied upon trying to send another messages to their domains. In a word, it was permanent
Hello:
What I, as a humble personal Email user, cannot condone nor understand how Comcast, ATT and others can continue to block ISP’s for the “abuses” of bulk mailers, spammers and other bandwidth-hungry business users of the Internet. I have long been a victim of this practice and find it to be totally unfair and impossible to rationalize.
Eddie, you should check out my blog.
In place of the commercial privileges of copyright and patent I would propose nothing, indeed that no privileges should be granted anyone, that all men are equal and none deserve a law that privileges them above another, no matter the cultural benefit that is hypothesised as the result.
What we then revert to is a law that protects only mankind’s natural rights, human rights, to liberty, truth, privacy, and life.
This includes recognition of material and intellectual property, protects and compensates it against theft or plagiarism/passing off.
Moreover, there’s absolutely nothing except imagination in the way of artists or inventors selling their intellectual property, art or designs, to those who would willingly purchase it. However, the market for copies would no longer be subject to state instituted monopoly. Admittedly, it’s tricky for people who’ve grown up with monopoly to grok how on earth commerce could survive without it, but the paradigm shift can be made with effort, as it has to be, because technology has shifted things whether we can grok the ramifications or not.
Michael,
I clearly misunderstood what you were originally saying; it seems we pretty much agree, but approach the issue from different directions. While I support the concept of copyright (if not maybe the exact implementation we have today), the current RIAA/MPAA fight to preserve copyright-as-they-know-it is doomed to failure. It’s a never-ending game of whack-a-mole. The only long term solution that will avoid this arms race requires a change in business models.
When you referred to the astroturf, are you referring to the fact that I’m responding to Richard Bennett as if he’ll listen? I know he posts here with a narrow point of view and that he reduces the whole argument to “pirates vs rights owners” whether or not it fits.
Crosbie,
It would be an interesting discussion (although off topic here?) to see what you would propose in the place of current copyright/patent law.
The BitTorrent CTO told me today they have no plans for any Comcast-specific changes to uTorrent, so don’t worry Brian, be happy.
This casts a very curious spotlight upon BitTorrent Inc’s recent purchase of μTorrent.
As you mention, the reference BitTorrent client accounts for a minor fraction of the population; and changes to it, to accommodate a given business model, would not largely affect the torrent community. μTorrent, on the other hand, is one of the more popular clients. Now i’m wondering if things are going to happen to it that will make me glad that i save old versions.
Well, Jason, it’s like this. The Pirate Bay is a pirated content indexer. You can get that from the name and from their legal troubles. When people who frequent such a site form a group to kick around ideas for a new P2P program, and they say they’re hoping to provide protection from anti-piracy outfits, I figure they’re interested in doing piracy. This may be a leap, I know, because maybe, um, they’re really out to collect money for the Girl Scouts and they’re just using the piracy talk on the piracy site about piracy tools to cover-up their real intentions. So you might be right.
But at the end of the day, I figure the main benefits of protocol obfuscation, source privacy, content obfuscation and destination anonymity are more and better piracy. And yes, there may be some good anti-totalitarian crusader in China who will benefit from these measures to hide his identity from Google and the government of China, but that’s what you call a fringe benefit.
So I see you’re neither naive nor ill-informed, you’re simply choosing to live in a state of denial. Which is your right, of course, but don’t expect me to follow suit.
And what do you think of the proposed new scheme for BitTorrent to encrypt tracker announces? Let me guess – it’s a step forward for free speech and democracy, of course, and nothing to do with making more illegal content available to more people faster.
Right.
The irony.
You said:
I refuted that statement.
You said:
strongly implying that there is such software being widely
distributed. I challenged you on this. There isn’t such a client. What
you refer to was a comically inept attempt at a P2P algorithm by
technically incompetent montebanks that never got passed the wild
pie-in-the-sky planning stage and is now presumably dead (the page
with their ‘proposal’ has gone dark).
And that project is not called Hypercube by the way. Hypercube was the
plain-old Bittorent tracker the PirateBay was using before December
2007. To my knowledge it now uses another plain-old Bittorrent tracker
and definitely no custom protocol.
In addition, the “protection against anti-piracy outfits” was
primarily to stop spammers and fake files. You would imagine that this
would be a desirable characteristic for *any* tracker, piracy or
no. Their proposal provided no anonymity.
And that’s the point I am making and you have seemed to have
missed. What you call “optimized for piracy” are higher level goals that
apply more generally to data communication. Anonymity, protocol
agnosticism, privacy: These aren’t piracy specific. Piracy may benefit
from them but so does it benefit from high-speed Internet —
that doesn’t make high-speed Internet ‘bad’ or
illegitimate.
You have this bizarre hard-on for piracy that seems to largely
influence your opinions on network management which are often
thinly-veiled ‘anti-piracy’ measures.
Goodness. Who knew that the only reason not to use a product licensed by two profit-making corporations, and possibly bearing any number of interesting restrictions (say a non-exclusive license to retransmit any date shared by means of the protocol, as with Adobe’s first attempt in this arena) was to commit unlawful acts? I guess wanting to use open-source software is mens rea in general…
Eddie,
There is no coloring of wrong or right here. I don’t support copyright infringement. I don’t download content.
But I don’t live in a dream world either. People want to do it. People are going to do it. You have to accept that and move forward with that in mind as you try to structure networks, software, and business models. Fighting it directly has been proven many times to be pointless.
Don’t turn your ankle on the astroturf, friend. It grips tightly.
Saul Hansell has an article in the New York Times, Comcast’s Blurry High Definition Picture.
——————————————
“Not only has Comcast been slowing down Internet users exchanging files with the BitTorrent protocol, it has been quietly reducing the quality of some high definition television networks it carries as well.”
Eddie, I submit myself as a candidate member of the set of thinking persons who hold that copyright per se is unethical (and immoral if you prefer).
NB Whilst I also find patents similarly unethical, I wholly support the concept of intellectual property and laws against its theft.
Michael Donnelly Says:
> Paul: sounds like a win-win-loss situation. Comcast wins,
> BT-the-company wins, and the only loser is: the consumer.
> (snip)
> People want to do things with BT-the-protocol that many companies do not
> like. People have the technical ability to do these things, regardless of the
> actions of the ISP and other legal groups. People will find a way to do these
> things.
Gee, you’re rather supporting the viewpoint of Richard Bennett, except from the opposite side. Michael, in your scenario, the only loser is the consumer who wants to do certain illegal things. The consumer who is only interested in doing legal things would appear to win. You appear to be one of those who doesn’t support copyright law?
In areas where gun ownership is legal, people have the technical ability to kill others with a gun. This does not make it right, or moral, to do this. It is true that there are those who feel that if they are technically allowed to do something, then they have the moral right to do so. (Or they just don’t care about morality or legality.) Your argument, taking in whole, appears to support the actions of those people.
Am I equating copyright violation to murder? No, of course not. And I recognize that illegality is not always equal to immorality. I’m just pointing out that something being possible does not equate to it being moral or it being something that society should support. IMHO, in addition to being illegal, copyright theft is immoral, and no thinking person should support it.
Richard Bennett Says:
> Piracy actually does exist, and anti-piracy measures aren’t
> “anti-consumer†unless you have a warped idea of “consumption.â€
Richard, the first half of your sentence is certainly true. The second half may or may not be true,, depending on the facts, and does not necessarily follow. It depends on the specific anti-piracy measures. I felt that the anti-piracy measures by Sony on their music CDs, for example, were notably anti-consumer.. These measures, although currently stopped, have decreased my willingness to purchase any Sony music. (Which, of course, means I have to do without.)
A great many anti-piracy measures, in practice, end up being somewhere from moderately anti-consumer to strongly anti-consumer. Many also lend themselves more directly to being used as vendor lock-in devices than to use as an anti-piracy device.
If you’re legally trading copyrighted material, given the prevalence of nasties on the internet it would probably be a good idea to encrypt headers and payloads.
Jason, you’re either really naive or well uninformed. Go see Torrent Freak’s discussion of the piracy-enhanced tracker: “The Pirate Bay is without a doubt the most popular BitTorrent tracker. Unfortunately their current tracker system is not performing as it used to and has reached its effective limit. The TPB team is currently working on a more efficient Open Source tracker system that, among other things, will guarantee better protection against anti-piracy outfits.”
The new piracy tool is called Hypercube.
No. You encrypt headers and payloads to avoid protocol specific traffic shaping, independent of what data you are transmitting. What on earth is the pirate version of BT your spouting about? Do some rudimentary research. You are ignorant of the facts. And stop conflating piracy with network usage.
You optimize a protocol for piracy by encrypting payloads and headers, and by hiding peers behind off-shore proxies. Winny and eDonkey do this real well, and the pirate version of BT they’re kicking around on the piracy sites will follow those models.
Piracy actually does exist, and anti-piracy measures aren’t “anti-consumer” unless you have a warped idea of “consumption.” Consumers of legal content will find it in their interest to upgrade to clients optimized for legal uses, and others won’t.
I’m waiting for a BitTorrent RFC, like the RFC we have for FTP. Anybody got an idea when that’s coming?
Uhuh.
How can a customer produce a congestion? The customers pay for an advertised bandwidth, and that is what they should get. The truth is that customers rarely get what they pay for and the ISPs are simply lying! The problem is not bit torrent or any other protocol or user activity, the problem is that the ISPs overbooked their networks.
Paul: sounds like a win-win-loss situation. Comcast wins, BT-the-company wins, and the only loser is: the consumer.
If Comcast allows good-BT-the-protocol and optimizes it within their network, but then goes back to choking evil-BT-the-protocol, it’s not going to make much difference.
People want to do things with BT-the-protocol that many companies do not like. People have the technical ability to do these things, regardless of the actions of the ISP and other legal groups. People will find a way to do these things.
At this point, I’m slightly optimistic on the possibility of Comcast doing the right thing: which is work with BT-the-company to make the protocol work better on their network. Period. But a big part of me knows that part of that process is might be adding a bunch of extra rules to make the protocol worthless for what Comcast’s customers want to do.
If the Verizon p4p experiments are a scalable guide, Comcast and BitTorrent have a huge amount of headroom to play with in in convincing users to step over to new clients. The report was 4x the performance at lower cost to the ISP; even if you cut that in half, a lot of people would be thrilled to have their torrents complete in half the time.
And there are also ways in which negotiating with BitTorrent the company rather than (impossibly) with BitTorrent the protocol can be a winning strategy for Comcast: once the two companies have come to an agreement, all of the software that relies on the original protocol can be depicted as “unlicensed” or “obsolete” or “shoddy knock-offs”, and Comcast can go right back to forging reset packets for it. Which might make BitTorrent the company perfectly happy as well.
I think the ‘illegal protocol’ and ‘optimized for piracy’ straw arguments are most interesting for calibration of the guy making them.
That said, he doesn’t seem to understand that, just like you can’t negotiate with a protocol, you also can’t arrest it.
Baseless drivel. What constitutes an illegal protocol? How do you optimize a protocol for piracy exactly? Obviously not complying with whatever the ISPs dictate is not an indication of illegality nor an optimization for piracy in particular.
As a long time linux user, grabbing ISOs via torrent has become second nature. If people are concerned that BitTorrent, Inc. has purchased uTorrent, then simply switch to an alternative: MonoTorrent/Monsoon.
http://www.monotorrent.com/s3_fastrampup.png
Look familiar? Almost just like uTorrent. Only in C#. And FOSS. The uTorrent WebUI’s even being ported to it.
uTorrent may end up dying fairly quickly if Monsoon is noticed on the larger torrent trackers, as said trackers currently link to uTorrent, but will change their link to whatever the current popular client is.
The only thing I miss is Azureus’s awesome pie-chart-visualization of chunks coming in and going out to other nodes. Absolutely genius!
I just don’t feel like having a large java vm installed to run it. Go GTK#!
Lessig is a nice man, and he has got a few things right. His “code is law” formulation flows from the more general principle that “law is whatever we want it to be until we change our minds.”
As far as BT is concerned, I rather suspect that the protocol will balkanize into a one or more versions optimized for legal file transfers and one or more versions optimized for piracy. As BT, Inc, knows which side of their bread is buttered, they’ve decided to lead the charge for legal a legal BT that’s optimized for use within highly-tuned ISP networks. This leaves their competitor Vuze in an interesting place, because they can try to be champion of piracy at the expense of their business or follow BT, Inc, down the road that they’re taking. Vuze is presently trying to have it both ways, but I doubt that’s going to work for long.
Richard, note, this thread is a good example of what Lawrence Lessig was talking about in the book _Code and other laws of cyberspace_. There’s a viewpoint that since BitTorrent-the-company cannot legal compel adoption of any changes to BitTorrent-the-protocol, that’s the end of the story and it’s significant to keep pointing it out. The other perspective is that broad technological changes under the control of large institutions can funnel people to doing what BitTorrent-the-company implements to BitTorrent-the-protocol. That is, if you adopt the changes to BitTorrent-the-protocol recommended by BitTorrent-the-company, your P2P exchanges will work a lot better than otherwise. And this will be technically (not legally) enforced by Comcast’s traffic-shaping. Code == law (metaphorically).
Bet you never thought of yourself as a Lessig-defender 🙂
As far as I know, the current Bittorrent specification is free to implement by anyone. There are dozens of clients that are not controlled by BT Inc (even uTorrent was one of them until recently). Some are backed by competing commercial entities (Vuze/Azureus). Some of them are open source, some are not. But that’s not the point: BT Inc cannot decree that all current implementation shall now behave in the way it specifies. So any ‘control’ BT, Inc. has over the Bittorrent protocol (or whatever you want to call the protocol + extensions that is currently implemented by the various clients) is superficial and consists of simply slapping their trademark on the modifications to the protocol.
Neither BT Inc nor anyone else will dictate how other clients operate.
BitTorrent, Inc, owns uTorrent and it’s not open source. Bram Cohen also controls an open source project called BitTorrent, but it’s not as full-featured and high-performing as the closed uTorrent implementation. The open source thing is written in Python, a very nice interpreted language, but hardly a speed demon. And then there’s all the BitTorrent DNA stuff that’s not part of the open source project.
It appears that BT, Inc. is happy to develop a proprietary system that works better for commercial applications, while the open source community is more concerned with better piracy. That’s what the protocol encryption and DHT is about, after all.
From my talks with the BT people it’s clear that they want to leave the pirate past behind and go straight.
Call BT, Inc. authoritarian if you must, but to me it’s simply good business.
Wrong. Nearly every design enhancement to Bittorrent was created by non-BT Inc implementations (protocol encryption, UDP trackers, DHT trackers, etc.). There is no open source version of the protocol, it is the same protocol with possibly some interoperable extensions. I do not know what feature you believe to be lacking in open source implementations, please elaborate.
I don’t know what you mean by ‘property’ of BT, Inc. They cannot force other implementations to adopt their modifications. So changes made by BT Inc. are largely irrelevant unless they can convince other implementors that they are worth adopting. And if they go ahead without support of other implementations they balkanize their torrent client.
Authoritarianism doesn’t work here. Sorry.
Seth, Andrew, and Tony are right, Ed, it makes perfect sense for Comcast to work with BitTorrent, Inc., to improve the BitTorrent protocol, and such work has been ongoing for some time in the P4P Forum. Contrary to popular belief, BT is not strictly “Open Source,” it’s the property of BT, Inc, and all the design enhancements are controlled by the company. There is an open source version of the protocol, as I understand it, but it’s not full-featured.
In any event, BT is not an RFC-defined Internet Standard, and is still very much evolving.
They’re only about 15 years behind corporations in that sense.
Back in the day when I was still in college, the CEO and/or Chairman of Reuters was given a demo of this shiny new-ish thing called The Internet at the CS department from which I was about to graduate.
He was very impressed. Allegedly one of the first questions he asked was “I want to buy it. Who do I talk to?”
At least he didn’t get confused with plumbing infrastructure. like a certain Senator did recently.
If ISPs were really interested in reducing non-existent network congestion, then perhaps they should require their customers to either install anti-virus software or perform an online anti-virus check every month.
I look at my router log and the number of incoming packets being blocked is mind boggling. It seems that a couple of Shaw and Comcast customers have zombie machines. Perhaps Shaw and Comcast could terminate these accounts.
But that all makes too much sense.
You’re missing the fact that extensions to the protocol that take network topography into account are beneficial to the users too. Why? Because what Comcast cares about is primarily minimizing BitTorrent traffic *outside* of its own network (where it has to pay by volume), BT traffic within its own network is much less of an issue.
On the other hand, traffic within the same network is generally speaking faster — usually significantly so — than traffic between distinct or even distant networks. So if BT client prefers peers from the same or nearby networks, it results in faster transfers.
Currently, BT clients pick peers more or less randomly, completely ignoring their real-world location. If BT contained extensions (it’s easily extensible with new stuff, fortunately) for determining topology of [relevant parts of] the network, both BT users *and* ISPs would benefit: users would get faster downloads, ISPs would get less expensive traffic.
So this is actually a smart move — BitTorrent Inc. was saying that it want to work with ISPs on exactly this kind of thing for a long time now, it’s nice to see that things are finally happening. By the way, BitTorrent Inc. is “just one maker of software”, but they’re the principal author of the protocol and they’re work is likely to be widely adopted if it’s useful. And you’re not quite right about them being minor player: their uTorrent (which they acquired to replace the original Python “mainline” implementation) is arguably one of the most widely used clients.
http://www.news.com/8301-10784_3-9893915-7.html?tag=nefd.top
This story came out a couple of weeks ago. It’s about trying to keep the number of P2P hops down by peering according to geography, reducing bandwidth use and increasing download speed simultaneously.
Everybody seems to win in this scenario, as long as there is an open protocol, no content-induced traffic-shaping, etc. You could probably get everyone to switch to a next-generation P2P. You might even be able to do it seamlessly on top of current systems, just sorting the list of peers by distance.
This smells of a face-saving move on Comcast’s part, given their past history. However, if we assume that bit torrent is driving the evolution of the protocol, what is wrong with cooperation between those working with network technology and those working with end-to-end applications? If Comcast had had had a more cooperative attitude from the start, perhaps the past situation could have been avoided.
Congestion management mechanisms in networks are needed to ensure responsive, efficient and fair sharing of resources and these need to work well with end to end protocols and applications. The needed engineering will be more likely to appear in a cooperative environment.
Mike,
I couldn’t agree more. Comcast is now in the same position as many of the other big, ignorant media players in this game. They can “solve” the problem by working with BT-the-company to add DRM or otherwise strip functionality, which will leave them facing people that fork the old protocol into new and fun methods of encryption or just plain obfuscation to avoid technical countermeasures.
Or they can accept that people want to use it to get what they want. You can’t negotiate with the protocol and, more importantly, you can’t negotiate with the people. Start with that as the Bottom of the Pyramid that cannot be changed and build around it.
Of course, Comcast is an ISP, not a media company. Their concern should be in the healthy operation of their network.
This is why, if Comcast is smart, they will try to make their proposed changes to BitTorrent as beneficial to the users as possible. For example, they might design it so that it adapts better to network usage and therefore doesn’t need the sort of throttling they want to impose on heavy users. Having a new client available which won’t get throttled the way other clients do would provide a big incentive for people to switch. Simply slowing things down to impact the network less will result in nobody switching.
Of course, knowing how Comcast operates, they will completely ignore the political realities of the situation and simply push for changes which suit them best, even though such changes will never be adopted on a wide scale.
To extent Michael’s argument a bit, it seems to me that network effects working against any new BitTorrent Inc protocol. Most users will still be using the old protocol, so if the new protocol is incompatible users will shun it.
Seth,
I think it’s more than just journalistic shorthand. The ability for a company to really alter a protocol like is different from protocol to protocol. How much could someone change how FTP works vs, say, SMB? Those examples are extreme, but they show the framework of the argument.
The torrent protocol is not solely owned by BitTorrent-the-company. In fact, the strength of open-source software will be shown in this case if the company negotiates changes to the protocol that the general public does not like.
BitTorrent-the-protocol users do not care about the protocol. They care about getting what they want. If BitTorrent-the-company changes the protocol so that those people can no longer get what they want, then they will:
a) Use another product to get what they want.
b) Keep the old BitTorrent-the-protocol to get what they want because they can.
Ed’s basic point about negotiating with BitTorrent-the-company is spot-on.
It may not be fair to describe BitTorrent Inc. as only a minor player in the BitTorrent software market. uTorrent, which they recently acquired, is easily the most popular BitTorrent client for Windows. In fact, it would seem that most BitTorrent users (at least on Windows) do use software provided by BitTorrent Inc.
Here are some relatively recent data on this:
http://torrentfreak.com/utorrent-gains-popularity-azureus-loses-ground-071216/
Yes, but to some extent this is also just journalistic shorthand. Like ‘The White House said today …”. A building can’t say anything! Building are mute, immobile structures. They have no life. There’s a difference between the building and the people in it. Only people can say anything …
So it all really means “Comcast and BitTorrent-the-company cut a deal so that BitTorrent-the-company will alter BitTorrent-the-protocol to work with Comcast’s network management needs. As part of the deal, both sides will be accomodating to each others interests rather than hostile”.
Though it’s rather hard to write it so bluntly in terms of what you’re allowed to say in polite company.
And right, BitTorrent-the-company has to convince the userbase of BitTorrent-the-protocol to switch, but at this level, that’s a different topic.
When you say : “What this means is that changes in BitTorrent Inc’s products won’t have much effect on Comcast’s network.” – that’s not necessarily correct. It’s assumed that BitTorrent-the-company has the major effect on the usage of BitTorrent-the-protocol. That assumption may not be true, but examining if it is in fact true may also be a distraction in reporting a piece of news.
I wonder if BitTorrent hopes to use a DRM approach where it will introduce digital signatures to identify ‘licensed’ BitTorrent sharing?
We then only have to wait for fraudulent impersonation of such signatures, but it could be the sort of thing Comcast is hoping for. It can then throttle all ‘unlicensed’ high-bandwidth protocols.
Tell me this approach won’t work – please.