Slashdot Mirror
Net Neutrality vs. Technical Reality
penciling_in writes "CircleID has a post by Richard Bennett, one of the panelists in the recent Innovation forum on open access and net neutrality — where Google announced their upcoming throttling detector. From the article: 'My name is Richard Bennett and I'm a network engineer. I've built networking products for 30 years and contributed to a dozen networking standards, including Ethernet and Wi-Fi. I was one of the witnesses at the FCC hearing at Harvard, and I wrote one of the dueling Op-Ed's on net neutrality that ran in the Mercury News the day of the Stanford hearing. I'm opposed to net neutrality regulations because they foreclose some engineering options that we're going to need for the Internet to become the one true general-purpose network that links all of us to each other, connects all our devices to all our information, and makes the world a better place. Let me explain ...' This article is great insight for anyone for or against net neutrality."
Article summary: "Hello, my name is Richard Bennett, and I'm an industry insider who's been bought off by big money to say net neutrality is bad in the same way climate scientists got bought off by big money to say environmental protection is bad."
I hate printers.
One word: Multicast .
YouTube on steroids is geographic caching. But even if two people on the same network are watching the same video, it should be an option to receive the networkdata that is for the position the other person is currently watching.
But the problem with multicasting is not that there are no tools, but it is not 'neutrally' implemented across different carriers that deploy access networks.
Support Eachother, Copy Dutch Property!
Parent post has to be one of the most clear, cogent, and effective rebuttals of the arguments made in the original article. One must always be mindful to consider the social, economic, and regulatory environment in which engineers--and by extension, the technologies they create--operate. And the author of the article simply fails to do this by viewing the problem as (in the words of parent post) "purely an engineering question."
I had mod points a few days ago but they expired. So this is my way of making up for not being able to mod the parent up.
No, It is not just simpler to add udp management capabilities. The Management capabilities in TCP are built into the NETWORK STACKS! And these network stacks may not play by the rules either. Read up on TCP RENO etc... Then you will understand what he is talking about.
Network Neutrality Squad Network Measurement Agent. It's LGPL.
Never mistake "can" for "should".
No, it seems to me you understand it perfectly. However TFA seems to be blurring the lines between net neutrality and treating traffic differently. Here is the main technical problem that TFA ignores entirely, and it is the central problem that network neutrality seeks to resolve: QoS and filtering aren't just applied to protocols and ports -- they are applied to individual IP Addresses, and to suppress new services!
I'm perfectly happy to give VoIP ports a higher priority QoS than file transfers, which tend to be more "bursty" anyway. I just don't think the ISP has the right to determine that VoIP connections to Vonage or Skype have higher priority than connections to my freebie/personal SIP/VoIP service. If no one on my route is currently using VoIP, there's no reason my BitTorrents should be going any slower. If the ISP had provisioned their networks properly, there's no reason any service should ever be going slower, when I'm not maxing out my personal bandwidth allocation!
The TFA is willfully ignorant about all the central problems with the way ISPs configure their filters and their network asynchronous transfers, which favor particular host-protocol combinations, not just individual protocols and ports. And when they do configure QoS, they do so to limit or segregate the customers, NOT to provide better service, unless a specific QoS toll is being paid by their "preferred" customers.
There has to be some money link between Richard Bennett and the incumbents -- this is just too obvious an oversight to not be intentional.
NACK/ACK (old S&F/RUID terms) is not an IP responsibility. ACK/NACK for TCP packet delivery failure is only noticeable at the destination client/server computer .... The IP part is the only part used by the IAP (CableCo/TelCo) infrastructure there is no consideration of the content TCP packets, failure to deliver, and/or the order/time of delivery. The TCP origin of an email/file does not need any ACK-confirmation that a packet was received at the intended destination, but the TCP origin does require a NACK-notice (to initiate a resend specific packet) when a specific packet was not received or corrupt (no need to resend the whole email/file); So, in some ways it perhaps prevents a great deal of unnecessary Internet traffic.
Non-neutral network that does proper QOS by throttling bandwidth-heavy protocols that don't behave themselves on the network is acceptable.
Stop getting D/DOS attacks and/or badly configured networks confused with TCP/IP. Yes, TCP/IP is an overhead heavy protocol, but there are legitimate reasons, and a lack of QoS bandwidth is always the problem on the Internet for ISP content/services providers and customers.
Quit listening to IT product marketeers (AKA: vendors with and agenda)
Unaccountable leaders are masters, and unrepresented people are slaves. How do US and EU fare?
LOL Richard Bennett is the guy that repeatedly spammed the wikipedia Network Neutrality article with his garbage, over a long, long period. He tried to rewrite the *definition* of network neutrality about a dozen times maybe in bizarre ways, he repeatedly deleted referenced material (usually describing them as 'lies' in his subject line), rewrote stuff, and in every possible way you can imagine tried to spin reality in ways that were self-evidently harmful to the balance of the wikipedia article.
He even deliberately misquoted another engineer to say the exact opposite of what they said; to the point that they logged onto the wikipedia talk page to complain. This was even after it was pointed out they never said what he wrote them saying and that the references disagreed.
He also thought that it was a good idea to get interviewed in articles in 'The Register' and then quoted himself in the wikipedia to 'prove' his points.
Oh yeah, and he used 'sockpuppets' to continue to also push his point of view while temporarily banned.
I could go on about this sleazebag for quite a while. When you even try to list the stunts he pulled it runs to several pages.
I would also challenge some of his depth of understanding, for example, at least at one point in time he didn't seem to have the slightest clue what a contended service is, which is kinda... basic. Really, really basic.
Really, he's just a bizarre guy, with bizarre views, and personality wise he's a total asshole.
(See wikipedia RFC, which contains references to a small fraction of his 'work' in the wikipedia if you want to get a measure of the man).
-WolfWithoutAClause
"Gravity is only a theory, not a fact!"Oops: RFC:
http://en.wikipedia.org/wiki/Wikipedia:Requests_for_comment/RichardBennett
-WolfWithoutAClause
"Gravity is only a theory, not a fact!"I live in Central PA, and we've basically got a duopoly between Comcast and Verizon (and no FIOS, just DSL). There are a few smaller companies like D&E or Commonwealth Telephone that run DSL a few smaller suburban areas, but they only operate where Verizon doesn't. One company that I know of, PA Online, leases bandwidth from Verizon, so they're stuck hoping that their far better customer service is worth the extra $7/month that they have to tack on to Verizon's price. The big question is whether or not skipping the totally aggravating call that you have to make to Verizon every 2-3 years for service is worth the extra cost.
Well, where I live - along the border between 2 cities, I have access to 2 cable companies, Wow and Camcast, and AT&T. (There is also stelite broadband, of course, but my neighbors tell me the QoS is bad). There are several companies that resell AT&T DSL, and a few that run there own DSL over lines leased from AT&T, but all of these are dependent on AT&T's infrastructure, so are not really competitors. (I often wonder if AT&T saves money from not having to bill or provide customer support for "indirect subscribers" - just provision circuits and let the "little guys" worry about the rest.) In my experience, these 3 big ISPs, and their vassels, behave more like a cartel than competitors.
Don't try to out wierd me, three-eyes. I get stranger things than you, free with my breakfast cereal. --Zaphod Beeblebr
In fact, they tend to operate in reverse economies of scale - bandwidth gets cheaper the more you buy (think of it as complexity O(x+n) due to fixed costs and the simple 1 to 1 nature of links), but management gets more expensive the more you do it because the 1-to-1 nature of links gets subsumed by having to manage the effects of all connections on each other n-to-n style for O(x+n^2). I think that's a little simplistic - Building faster links gets *very* expensive because you end up having to employ some bleeding edge technologies to effectively switch higher volumes of traffic.
There tend to be some quite serious step points of cost where as the bandwidth goes up the costs don't change much (so economies of scale do apply) until you get to particular points where you have to depoly a different technology base. This point is expensive and is why a lot of the smaller ISPs fold; they simply cannot afford to go to the next level.
Ancient history. Very few Ethernet links today are CSMA/CD. Full duplex Ethernet is simply a point-to-point serial link which has no utilization degradation, and since switches replaced hubs, virtually all Ethernet links are full duplex.
If the clients want to start watching at arbitrary times rather than wait for the "next slot to come along" you'll probably need to cache "near" the clients anyway.
The benefit from multicast isn't that great in that scenario.
Multicast is good if you have many clients downloading the same thing at the same time.
That's not quite what people are doing with Youtube.
I will take my offtopic moderations just to make sure other people who post legitimate replies to his astroturfing know what they're getting into.
Once he starts replying to himself with multiple accounts, anything he had to say becomes irrelevant.
The twitter monologues. Click on my homepage and be amazed.
Do they have to throttle P2P to control bandwidth usage, or are they just rolling out the same kind of propaganda as American network operators? After all, Sony owns a lot of content.
The engineers have to implement technology that supports the activity that people want to do. That the market demands. Their design decisions are constrained by their bosses, who are prioritizing competing with other sources of content over any kind of service quality to their customers. You're talking about these engineers as if they're actually being charged with keeping the networks clear, rather than build out a strategy that constrains bandwidth to enhance the perception of scarcity, which is the basis of the value that the network operators are selling.
Yes, people's bandwidth demands will expand to fill whatever capacity is delivered. But networks sell that bandwidth. The problem is that WAN operators have long counted on an optimization, bandwidth oversubscription, that sells people what they could never all actually get. That era has passed. Network operators have to expand the capacity of the shared links to accommodate the capacity of the edges. When they do that, get the proportions correct to the modern demand models, they will have a hugely valuable product to sell, that will earn them a lot more money as it helps others make a lot more money, either off the delivery of content or just elsewhere in order to consume that content.
But they're so used to getting everything, and giving the least possible, that they're just doing more of that, with their new monopoly powers. They want to sell content in competition with their content delivering customers, and of course (because they always have) they want to use their competitive advantages to crush those customers who are competitors. Even if that means also stepping all over the demands of their customers who are consumers.
That's why the telcos and cablecos create the false dichotomy. They're the ones who are saying they must deploy tiered pricing and bandwidth caps, even though their own research shows that increased bandwidth is cheaper and more effective at solving the problem, while bringing extra benefits (more bandwidth overall to sell).
A content-neutral network is a primary benefit of the Internet. Keeping that value as a design goal is necessarily the job of the government, because the telcos/cablecos are ignoring the economics of the basic problem in favor of a more complex strategic goal. A goal that puts the telcos/cablecos' interests in conflict with their customers at both ends of a transmission.
I think that the current means of Network Neutrality enforcement is in fact wrong, because it focuses on the wrong layer of the overall problem. If Internet bandwidth providers were prohibited from the kind of vertical bundling that defines other monopolistic industries, like say Microsoft's, then they'd all be falling over themselves to build more bandwidth, more product to sell. If they were also required to continue the "common carriage" policies that everyone knows is essential to essential infrastructure, that combination would give us a level playing field that would include content neutrality, and a lot of the rest of what makes the Internet healthy.
But instead, the cablecos/telcos are running the legislation with their lobbyists, finding a conflict only with a few content providers rich enough to stand up, newcomers like Google. Consumers are totally absent, except that the content providers prioritize them (because they're more sensitive to market demands than are telcos/cablecos, because they're not monopolies who can ignore their customers). That's a tragedy, because of course Congress is supposed to represent those consumers first, as their population is vastly greater than the number of executives at the cablecos/telcos and content providers combined.
So in the meantime, at least Network Neutrality protects part of consumers' interests. Since economics (at least the basic, sus
--
make install -not war