Net neutrality is a very confused idea. Once one tries to define a testable law, it becomes a lot less desirable.
Since I started paying attention, the definition has changed from “QoS on the last mile” to “slowing certain protocols” to “blocking competitive services” to “blocking any services” to “not building enough capacity” to “fast lanes that aren’t CDNs” to “throttling any user” to “paying for interconnect”.
In the Wikipedia case above, it’s clearly a violation of NN and it’s clearly good for consumers. It’s a very progressive policy, I support it – and it’s not neutral.
The justification Wikimedia offers here is that they are serving the community and are not exchanging $$ with the ISP. Is this part of the definition now? And if so, do we want to start defining which sites we feel serve the community?
Since I started paying attention, the definition has changed from “QoS on the last mile” to “slowing certain protocols” to “blocking competitive services” to “blocking any services” to “not building enough capacity” to “fast lanes that aren’t CDNs” to “throttling any user” to “paying for interconnect”.
Effective net neutrality regulation has to address all of those things. If it seems like the goalposts have shifted, it's because ISPs have kept pushing the boundaries of what they can get away with in their quest to gouge their customers and extort Internet companies.
A neutral ISP only has to provide two things at minimum:
1. Guaranteed minimum throughput at peak times, with customer-controlled QoS determining which packets get dropped, and no discrimination on source, destination, protocol, or content that cannot be disabled by the user for free.
2. Sufficient upstream bandwidth, to those third-party services and networks customers are paying to use, to satisfy all minimum guaranteed demands under #1.
> Net neutrality is a very confused idea. Once one tries to define a testable law, it becomes a lot less desirable.
It certainly is. It's not clear to me that a neutral internet is possible, enforceable or desirable either. It seems like what people really want is for ISPs to not show some of the poor behaviors that they've shown (blocking bittorrent, poor connectivity to netflix, etc).
A much simpler solution is local loop unbundling, leading to a competitive marketplace and better behavior from ISPs. Local loop unbundling is easy to measure, easy to regulate, and easy to legislate: we already have the law in the Telecommunications Act of 1996, it's just that the FCC decided that the provisions didn't apply to cable internet, and when DSL providers complained that it wasn't fair, the FCC decided that they didn't apply to any internet.
(Local loop unbundling wasn't perfect, but it's a good start)
The argument that QoS would violate net neutrality is FUD from net neutrality opponents. The only time I've seen proponents call it that was because ISPs would block bittorrent for "quality of service" reasons, which of course isn't QoS at all.
I am pro-net neutrality, and I think that particular packets getting QoS guarantees is exactly what I'm against - and pretty much completely describes what I'm against.
I don't know what net neutrality with QoS guarantees even means. Unless QoS is a promise that the ISPs will be making to me, guaranteeing that I will always have a particular amount of bandwidth - but I don't want any of the packets that I am sending or receiving to get favorable treatment of any other packets that I'm sending or receiving.
I'd even like to see law about the variables that upstream caching algorithms can use, and I think that situations like the Netflix appliance could absolutely legitimately be declared a no-no. Packet throughput should be fungible, and it is before people start slipping each other cash.
QoS is a highly desirable feature of a network, iff you, the end user, control the QoS selection. If I'm playing a fast-paced online game, talking to a group of friends with VoIP, and downloading a movie from one of the online stores to watch later, I'd like to place the priority of the movie's packets below the game and voice packets. But if the ISP does that for me, gives me no choice in the matter, and/or prioritizes my minimum guaranteed allotment movie packets above or below the voice packets of other customers, that's not okay.
I've mentioned this strategy a few times before on HN, starting ~4 years ago:
Yes; QOS has no place crossing the boundaries of a network segment - It should allow me to control the ingress and egress of my network, but be stripped before it gets to a network "trunk" like an ISP.
Further on that; If bandwidth on a segment is constrained, the ISP should divide it as fairly as possible, while also trying to maximize throughput. Each user should get an equal share, and if they're using less than that, the remainder should be apportioned to the other users evenly.
It seems to me that some level of either user-controlled QoS or congestion notification is desirable at the ISP level, since your software (e.g. Netflix, VoIP) might not know that the ISP network is congested, but you'd still prefer to tell your ISP's routers to prioritize the VoIP over the Netflix (within your allotted share of bandwidth, of course).
>QoS is a highly desirable feature of a network, iff you, the end user, control the QoS selection.
Agreed, although I think of that as something that I do at my own router, but that only covers my outbound. Being able to control what comes down to me would only be an improvement.
I guess I disagree then, because when I think of QoS, I think of what has to be done to keep a network and services on it working smoothly. I consider it a violation of net neutrality when a packet is blocked or throttled for political or business reasons, but not when a packet is blocked or throttled to keep the integrity and functionality of the network intact. Where this still intersects with politics is when an ISP makes claims like that what its customers use the network for is itself causing problems, which really is a business problem--if all your customers are using Netflix and it's choking your network, that's not a technical problem per se--it's a mismatch between the features your network offers and the features your network users desire to use. Which is a business problem. So I am against the usage of QoS as a FUD tactic, that is to say, "we have to throttle your traffic because you're abusing the network by using it."
I don't consider technical QoS decisions to be a violation of net neutrality, because everything that is permitted is not good. I can saturate my upload with bittorrent traffic so that it makes looking at websites unusable. Dropping my bittorrent upload just enough that it doesn't cause this problem doesn't violate my rights as far as I'm concerned.
The main argument against QoS-style throttling is not the bittorrent example, but rather the example of Netflix and friends - cable companies have a vested interest in throttling by content type in order to limit quality of online video services and sell their own cable TV/video on demand.
The QoS issue wasn't that bittorrent was blocked, but that ISPs were heavily deprioritizing bittorrent and potentially other unwanted traffic through packet inspection. It is and was very much something opposed by net neutrality advocates.
I didn't feel like going into that level of detail, but since you brought it up you are right. ISPs were using tools like Sandvine to detect bittorrent traffic and throttle it down to unusably low speeds. The ostensible reason was that the Bittorrent protocol was using a ton of data transfer (that the customer thought they had bought, boo hoo) and that the packet storm was causing the performance of everything else to drop to shit quality (legitimate.)
I guess QoS could mean many things to many people, but in this example, I think it was techinically incorrect for them to say they were doing it for the performance of the network when they were dropping the speeds to far below what would have simply stopped the bufferbloat problem. They were mad because people were bittorrenting rather than just sending an email now and then and checking out a webpage.
So when I say FUD, I mean you can justify a lot of rotten shit to prop up your business and call it QoS. I may be wrong because this is not my area of expertise, but you can prioritize packets based on protocol in ways that keep your network working without engaging in draconian throttling and caps. I mean, if the argument is that NN means everyone can bittorrent at full speed even if it makes the entire network run like shit, well then I would be against Net Neutrality too.
There are various interests in play, but what prompted the movement was the idea that ISPs were going to start billing websites to provide unrestricted access to their customers and invert the web. That idea was nearly universally condemned online and the first few people who came up to name a movement contrary to that called it net neutrality and started trying to come up with principles to fit what had horrified so many people.
So the basic idea behind it is that ISPs shouldn't be able to indirectly charge content providers by threatening to block or restrict their content when it's requested by the ISPs customers. In other words, ISPs have no legitimate business meddling with other peoples' businesses, but different people have different ideas about what it is or isn't now.
Net neutrality is a catch-all phrase at this point but it always was just a business dispute between telcos/cable and the tech/internet industries who are both jockeying to accrue the cost savings to themselves. It's just cloaked in language that pretends that this is an issue in the public's interest.
There are billions of websites out there and each one of them is some kind of information. Giving wikipedia free access but charging others may seem to be innocent. But I believe on the contrary, it is opening one door that should not be used.
There are thousands (millions?) of wikipedia articles and all have some kind of linked citation. How will you draw the line? Wikipedia is open information but these citations are not?
In my country we have operator giving twitter and facebook free access. And I always thought that was another nail on net neutralities coffin. Wikipedia is nailing one more too.
Public benefit organizations like Wikipedia already operate by a somewhat different set of tax laws than traditional for profit corporations. This is precisely because governments around the world believe there is high value in organizations other than governments that serve the people with a public good or service.
Why can you not imagine similar carve-outs for non-profits when it comes to communications regulations on the Internet.
Also I'm not against free wikipedia access. I'm against using connection provider power to limit access to any where.
If an ISP starts with free wikipedia where would it land then? I believe "You can freely connect wikipedia but you can't use whatsapp or you must reimburse us for the lost sms revenue" is only one step ahead.
+1 If it becomes the norm in these countries for wikipedia articles to not have citations, it becomes that much easier for those in power to control/manipulate information.
Just because it's Wikimedia doing it, doesn't mean that all of the sudden I'm all for it. I believe it's still discriminatory against small organization/websites and against net neutrality principles.
>Wikipedia Zero cannot be sold as part of a bundle. Access
> to the Wikimedia sites through Wikipedia Zero cannot be
>sold through limited service bundles.
Sounds to me like it is ONLY available in some sort of bundles.
This is a brilliant display of the good that non-NN can bring to the world while showing the potential sliding slope. Interesting stuff.
This is a difficult choice. On one hand I believe that Wikipedia should be FREE in all senses of the word. On the other hand making one exception will create more exceptions. Why not make facebook and Gmail free to access so that people can communicate? I want this exception for Wikipedia. But I am worried.
"We believe in Net Neutrality, but it doesn't apply to us." This is bullshit. How many of the big incumbent websites would like to be in the blessed exception category? Facebook is one (http://internet.org/press/introducing-the-internet-dot-org-a...), and they are most certainly a for-profit corporation with many competitors (which are not included in the Zambia deal, of course). If the intention were truly to benefit these parts of the world, and not just to drive more users for the existing big players, why not give free access to local sites, to help drive growth in the local economy? Of course, then there is a risk of them competing with American websites, and they might have a cultural/linguistic advantage. Surely there is a way to provide a basic, inexpensive free tier of Internet service that is content-neutral.
I am from India. Bandwidth is costly here. Internet enabled devices have surprisingly have become cheaper. Making some access 'Free' could change the world of many people.
I agree where there is food and shelter in short, giving Internet free may not help to fight these. But it will help to fight corruption and bureaucracy. Many important government services are online nowadays. In West-Bengal, government stopped online college admission just to keep bureaucratic system going. They fear free Internet, it is a killer for corruption here.
Allowing certain organizations to bypass data caps or data rates and not others is an incredible opportunity for corruption.
It's like the "licenses Raj" days all over again. Nehru might have had good intentions, but it just created a massively corruption system of extortion.
I think it breaks pretty cleanly on whether the content provider is for-profit or not-for-profit. Not for profits organized as public charities (e.g. Wikipedia) could arguably be granted some treatment that is not available to for-profit entities such as Facebook and Google. This is consistent with their treatment in the tax code, etc.
Gmail and Facebook very clearly allow people to communicate but their reason for existence is not that but to deliver advertising, a commercial objective.
Likewise, this is a messy grey area. It looks like Wikipedia is trying pretty hard to avoid the slippery slope of cable packaged internet here, but they're still playing by the ISPs' non-neutral internet rules. That alone really isn't a good sign.
Facebook Zero is an initiative undertaken by social networking service company Facebook in collaboration with mobile phone-based Internet providers, whereby the providers waive data (bandwidth) charges (also known as zero-rate) for accessing Facebook on phones via a stripped-down text-only version of its mobile website, located at 0.facebook.com or zero.facebook.com
It is hard to say if Wikimedia will stay credible regarding their net neutrality effort without disabling WP0. The no-money-paid and no-exclusive-deals approach sounds reasonable at least.
>Imagine a world in which every single human being can freely share in the sum of all knowledge. That’s our commitment.
Wikipedia can never be the sum of all human knowledge and anyone who thinks so is seriously deluded. The web in its entirety still doesn't include even a fraction of all knowledge. To think that one website -- a poorly designed one, with terrible software, design, media handling, policies and leadership -- is claiming this is simply laughable.
That goes for Wikipedia "protecting the internet". Protecting it from who exactly?
Wikipedia's leadership and its acolytes constantly use false utopian goals to justify their expansion into areas that they are ill-equipped to deal with (see: dangerous and incorrect medical data). It seems they are trying to use the same lofty ideals to carve out market protection in Africa.
Ok. Instead of simply spouting vitriol, suggest an alternative service, solving the same problems, that you think genuinely cares about a fair internet.
Or do you dismiss anyone who strives to make things better as "false utopian" and "acolytes"?
That depends what you call a "fair" internet. There is a large body of people who don't believe Wikipedia embodies any kind of fairness. There are also many people who are trying to provide better information than Wikipedia. I have my own claim in that area and am happy to support others who see Wikipedia's failures for what they are -- right here, right now -- rather than some imaginary future ideal. Whatever the case, I don't think it is right for Wikipedia to claim special privileges and considerations based on utopian goals that will never happen -- that's politics not progress.
Since I started paying attention, the definition has changed from “QoS on the last mile” to “slowing certain protocols” to “blocking competitive services” to “blocking any services” to “not building enough capacity” to “fast lanes that aren’t CDNs” to “throttling any user” to “paying for interconnect”.
In the Wikipedia case above, it’s clearly a violation of NN and it’s clearly good for consumers. It’s a very progressive policy, I support it – and it’s not neutral.
The justification Wikimedia offers here is that they are serving the community and are not exchanging $$ with the ISP. Is this part of the definition now? And if so, do we want to start defining which sites we feel serve the community?