Wikipedia’s ‘complicated’ relationship with net neutrality | Hacker News Reader

Wikipedia’s ‘complicated’ relationship with net neutrality

washingtonpost.com

108 pointsdavmre85 comments

Wikipedians take it on faith that their website is better, more useful, more ethical, more informative, etc., than other websites. Thus they are able to believe that it is right that their website should receive special (non-neutral) treatment from ISPs.

The problem is, most website operators feel this way. I'm sure Google's executives believe that everyone would be better off if they had more access to Google; and Facebook folks think that about Facebook; Apple folks about iCloud; etc.

So, this is not actually a consistent framework for how ISPs should treat traffic. What if some team develops a better source of information than Wikipedia? They'll be hamstrung from the beginning because Wikipedia is "free" and they are not.

It takes the decision-making out of the hands of the consumer, and puts it into the hands of big ISPs and websites. Whether the website is run by a non-profit or not doesn't matter much...either way it is like the foxes guarding the hen house.

> Wikipedians take it on faith that their website is better, more useful, more ethical, more informative, etc., than other websites.

I've been a Wikipedia editor for about 10 years. I don't think that your perception of the community has any basis in reality. Lot's of Wikimedians are pretty vocal about their disagreements with the Wikimedia Foundation on a wide range of issues (including this one). There are lots of discussions within the community about the limitations of Wikipedia's model.

How do you feel about non-neutral treatment of Wikipedia traffic? Does this differ from your feelings on non-neutral treatment of Netflix traffic?

In the US or in Sub-Saharan Africa?

I tend to agree with the WMF that in parts of the world where millions of people are using a smartphone as their first computer, providing free information is key to wikimedia's mission. I think that the world would be better off if all websites were treated equally by all ISPs, but I also think the world would be better off if millions of people who have never had a computer before can have the access to information that we take for granted. Given that the WMF is unlikely to be able to change telecom policy or business practices in all these countries, I think it's better to work within the existing system and provide a resource that would otherwise be costly (perhaps prohibitively so).

Having said that, many Wikimedians disagree with me.

But you didn't answer the question, did you?

but I also think the world would be better off if millions of people who have never had a computer before can have the access to information that we take for granted

Like Netflix? Or only like Wikipedia? That's the question. Are you saying it's more important for Sub-Saharan Africans to get access to Wikipedia than Netflix? Netflix might disagree. scholarpedia.org might have something to say about it as well. So might stackexchange.com.

I think I did answer the question. I don't think that users should be billed for Internet access based on what they use, and I don't think that carriers should be able to play favorites when it comes to content providers. However, given that the wireless industry in many third-world countries works this way, and given that the WMF is unlikely to change either business practices or telecom policy in these countries, I would much rather see the WMF work within the system and try to provide free resources to people then refuse to participate on principle.

There's a huge practical benefit to Wikimedia Zero - people who have very limited access to information resources can get these great resources for free. Balanced against that, I just don't see how not participating benefits anybody, except, perhaps, by letting some people in San Francisco and Cambridge, MA feel good about standing up for an abstract principle.

As someone making a Wikipedia competitor I could not agree more.

This is an interesting topic, but kind of a weird article. Wikipedia wrote a long article addressing the topic of Net Neutrality and Wikipedia Zero a few months back. This is worth reading if you're interested in the topic:

https://blog.wikimedia.org/2014/08/01/wikipedia-zero-and-net...

Kind of odd that the Washington Post doesn't mention any of these points. Personally, I still feel the Wikipedia Zero program violates net neutrality, but mentioning Wikipedia's self-imposed constraints here seems important.

Good point. Here are the constraints:

> No exchange of payment. The Wikimedia Foundation does not pay carriers to zero-rate access to the Wikimedia sites and does not receive payments from carriers through Wikipedia Zero.

> 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.

> No exclusive rights. We try to partner with as many carriers as possible to maximize the number of users that can benefit from the initiative.

> Open to collaborating with other public interest sites. Our main goal is to promote free access to knowledge and we want to help other similar services interested in doing the same.

This is super interesting! Thanks for sharing.

I believe this is a good illustration of how most net neutrality advocacy is misguided. The (very real) problem net neutrality advocates are trying to solve isn't paid prioritization; rather, it's a lack of meaningful competition in the "last-mile" broadband market, and this market failure produces aberrant scenarios like NBC Comcast shaking down its competitors (eg Netflix). I'm hesitant to strongly support most mainstream net neutrality proposals because of situations like this, where network discrimination is net positive.

Anyway, I'd be curious to hear your thoughts.

You're absolutely right. However, the probability that meaningful competition will break through the enormous barriers to entry that Comcast, et al have erected before they manage to irreparably damage one of the few remaining growth industries in the US (tech) is near zero.

Sorry to pick on you, but the responses of net neutrality advocates vis-à-vis competition sound like the standard strategy of the Foreign Office when dealing with crisis from Yes, Prime Minister.

"maybe we should do something about it, but there's nothing we can do", it's a natural monopoly.

"maybe there was something we could have done, but it's too late now", they already have a stronghold.

http://www.imdb.com/title/tt0751831/quotes?item=qt0190692

Don't get me wrong, I'd vastly prefer competition. There just isn't much reason to be optimistic it will arrive.

I do have a horse in this race, my employer just struck a very non-neutral deal with a wireless carrier: http://newsroom.t-mobile.com/news/more-music-freedom.htm

How is this critique useful?

I've never seen a net neutrality advocate be opposed to competition. I also don't know a single net neutrality advocate who was in favor of letting the wide variety of early ISPs decay into the sad monopolistic situation we have now.

All they're saying is that given that we no longer have an effective marketplace to reign in in the oligopoly in place, we need regulation to limit the harm. They are basically saying we should extend our approach to the phone network to the broadband network.

Why is it also their job to fix the conditions in the US that led to this? Net neutrality regulations are relatively simple: if the FCC (or possibly Congress) makes it happen, it will be pretty straightforward. But the lack of competition in the US broadband market can't be fixed with small regulation changes. It requires a wholesale reexamination of current economic dogma, deep changes in our approach to business regulation at multiple levels of government, and probably changing entirely the way we fund elections.

I just don't see much discussion around the topic; that may be because it's obvious for everyone inside the topic that it's unfixable, but as an outsider who lives in a very corrupt country (among the developed) but which nevertheless has plenty of broadband competition, I just find it very hard to believe that it almost requires a political revolution to increase competition in that market - or that at least that's obvious to everyone.

Yeah, there's a lot of history here, so I can see why it's mysterious to outsiders. But as somebody who has been following this since a teen, I believe a proposal of "let's increase competition" will go absolutely nowhere in the US. A very quick summary:

In the 80s a judge broke up our national communications monopoly [1]. That came as the result of action starting in the 70s, when business regulation had some teeth. To aid that, the Telecommunications Act of 1996 created the notion of the Competitive Local Exchange Carrier [2], allowing local competition for phone service by forcing the local-monopoly phone companies to share their wires. But that came just as telephone service was starting to become irrelevant, and didn't touch the local monopolies of the cable companies.

Eventually, AT&T basically put itself back together [3]. And that happened because the climate for business regulation had shifted; anti-trust enforcement has now withered away to the point that it's ok with oligopoly and near-monopoly. Why? Well, I think that's partly a US shift to managerialist thinking, aided by billionaires funding plenty of think tanks and professorships that just happen to promote then notion that billionaires and their companies should be able to do as they please.

This is all compounded by the insane US campaign finance system, where we end up with legislators that are very strongly pro-money. There is a lot of talk about supporting "free markets", but very little interest in healthy markets, ones where competition is a meaningful check on the action of dominant players. We had a very competitive ISP market, but it's basically gone now, and surprisingly few mourn it.

So basically, the whole political tide of the last 30 years in the US has been away from the kinds of governmental interventions that would restore real competition in the broadband market. It's so bad that 20 states have laws against municipalities providing broadband to their own citizens. [4]

So yeah, I'm just not expecting meaningful pro-competition action in the next 10 years. Net neutrality, though, is an achievable short term goal, because that's basically defending the status quo.

[1] http://en.wikipedia.org/wiki/Breakup_of_the_Bell_System

[2] http://en.wikipedia.org/wiki/Competitive_local_exchange_carr...

[3] Stephen Colbert sums it up here starting at 2:23: http://thecolbertreport.cc.com/videos/eamlaf/bears---balls--...

[4] http://arstechnica.com/tech-policy/2014/02/isp-lobby-has-alr...

I know all that history, though. But around here we went over a similar transition (single public carrier got broken up not that long ago), and it's hard to argue we're any less pro-money. I just don't find the argument that it's impossible that convincing.

Ok. It's not my job to help you understand US politics, so I can live with you not being convinced. If you would like to take some action based on your theory that the change is possible, I encourage that.

To me, the thing that I don't want to see is some web sites being throttled (or blocked!) on broadband.

Considering most (wired) broadband is bandwidth-limited at this point, varying that bandwidth limit based on what site I'm talking with just feels wrong, and I'd love to see the practice banned, because as you mention, a lot of users have no choice in what broadband they can subscribe to.

Wireless broadband is almost all capped in the quantity that you can use, however, and the T-Mobile "free streaming" options mentioned give you something extra for "free." This is in stark contrast to blocking/limiting something you're paying for, and I would be fine with this limited exception in the wireless space.

But what I would like to see, since this is effectively a utility, is for that "free streaming" to be available on an equal opportunity basis: And T-Mobile seems to be doing exactly that by allowing customers to vote on what services to include.

What feels like unfair competition is (as you mention) Comcast shaking down its competition by way of not providing customers equal access. And until that last-mile competition problem has been otherwise resolved (Google Fiber everywhere?), I think we do need those regulations to prevent monopolistic behaviors.

The T-Mobile deal seems great at first blush. You get more stuff for free - yay! But, what if you're a startup streaming music service, and T-Mobile does NOT include you in their list of cap-free streaming? Good luck competing with the other providers.

Perhaps a solution would be to classify types of services, and then make ANY company in that service eligible for the cap-free streaming. So, if you're a startup in the music streaming space, you are automatically included along with all the others. Then the telcos could provide packages of free services without harming innovation and competition.

I get that. But streaming for free over mobile is far from the only way to compete.

T-Mobile is the third-largest carrier in the US. Probably a small minority of T-Mobile users even know about the free streaming feature. If your business plan requires that tiny market segment to bootstrap you, then you need a new business plan.

I just don't see it as a big enough differentiator to kill a new service, even if other carriers started doing something similar to T-Mobile.

A new streaming service doesn't need to rely on free mobile streaming at all to bootstrap; if its focus were mobile streaming, a better answer could be to pre-cache hours worth of streams when on wifi. I end up in dead areas frequently, and I end up in roaming areas where T-Mobile won't give me data at all. I'd love a service that would auto-cache exactly the music I like to listen to so I could play it when roaming.

I've heard Spotify CAN cache, but it's too expensive for my tastes. Something like a cacheable Pandora would be nice.

But my point is that while yes, it grants an edge to certain established businesses, it's not an insurmountable barrier. Not like Comcast throttling Netflix; if you can't use the service in its primary form, it's DOA. And T-Mobile seems to be adding new services pretty quickly: Even indie stations like Radio Paradise are on the list.

> But what I would like to see, since this is effectively a utility, is for that "free streaming" to be available on an equal opportunity basis: And T-Mobile seems to be doing exactly that by allowing customers to vote on what services to include.

That's not an equal-opportunity basis, as consumers can't (or won't) vote for a company that doesn't exist yet.

This raises the barrier to entry for startups even further, because now they have to get enough brand recognition to beat out established competitors in this vote just to reach some segment of the market[0].

As if the barriers to entry for startups disrupting music distribution aren't already high enough! :)

[0] T-mobile has a small market share, but what if Verizon did it? Remember that there are only four non-virtual carriers in the US, so if just one of the two big guys (Verizon and AT&T) does it, it'll dramatically tilt the odds away from startups even more.

See my comment here https://news.ycombinator.com/item?id=8683425

> The (very real) problem net neutrality advocates are trying to solve isn't paid prioritization

Well, duh, net neutrality has been an issue for nearly a decade, and paid prioritization is a much more recent focus of discussion. Paid prioritization can be a manifestation of the problem net neutrality is designed to address, but its never been the main issue.

> rather, it's a lack of meaningful competition in the "last-mile" broadband market

No, that's not the problem that net neutrality is trying to solve. That's a constraint on the solution space, given that there isn't a magic "make meaningful competition appear and remain" wand that can be waved. Certainly, there are reasons to suspect that if wishing for such competition would make it so, that would also go some way to solving the problems net neutrality seeks to address, but since that rests on a counterfactual assumption, its not really all that significant to dealing with real-world issues.

There's no magic, but there is an argument to be made that the time it takes to iterate on one of those decisions contributes to the difficulty. Instead of being able to just pass a regulation for a set period of time and then revisiting it, the target company raises a huge stink, sues, and gets their buddies to sue.

If we'd have had a few chances, this would be a solved problem by now. Now we have to fight to the death for every single attempt.

It's also complicated by the fact that competition in mobile broadband is much stronger than in cable broadband. Personally I would like to see net neutrality limited to cable broadband.

"Misguided" would be a mischaracterization. I doubt anyone advocating for imposing net neutrality on providers would pass up a chance for more competition. However, without that competition, a net neutrality rule is one part of what it takes to keep the US internet a good environment for innovation.

Even in a competitive market, Comcast wants to, and can, vertically integrate with NBC then punish its competitors for the same reason Apple wants to vertically integrate content/software/hardware and push out anyone that competes with any part of that stack. There's a huge, technical, advantage to having the content company also own the network: pushing caching nodes to the edges of the network, consolidating redundant data flows on the backbone, taking advantage of multicast, etc. It's not all that different to the advantages of vertically integrating iTunes/iOS/Apple Ax-based hardware.

> It's not all that different to the advantages of vertically integrating iTunes/iOS/Apple Ax-based hardware.

It's different due to the fact that Internet is a "utility" now in the same way gas, electricity and water are (isn't it considered a human right in the USA now?). An Apple iPhone is not in any way a utility, and its functionality can be made up, 1:1 (fanboyism aside, thank you) by another device. iTunes (the store) is not a utility, and its functionality can be made up, 1:1, by purchasing CDs or using Google Play / Spotify / Rdio. iTunes (the app) is not a utility for obvious reasons.

Yes Apple locks you into an ecosystem, however that ecosystem is not a fundamental "requirement" to function as a business. There's simply no replacement for the Internet and, in a lot of places, there's no replacement for $ISP.

It's different due to the fact that Internet is a "utility" now in the same way gas, electricity and water are (isn't it considered a human right in the USA now?)

No, it isn't, and no, it isn't.

I disagree. Rather than Internet.org giving free access to wikipedia and facebook, they should just give people 100mb free data and let them decide what is best for them.

This is different to giving free access to Wikipedia. You can download Gigabytes of Wikipedia articles every month without paying for it. A better alternative would be to enable users to choose, say, five domains that don't count. But then clever users choose one web proxy site and get all their internet for free...

I don't think you'll be guzzling gigabytes of data on the low end phones people use in the poor parts of Africa that we are talking about. If they did, I doubt internet.org could afford to give it away.

We are not talking about broadband providers, but GSM/Edge internet on what is essentially feature phones.

I think since it has gained exposure through a public statement by the president, the business news has started to take interest in it. Now a lot of sense is starting to get expressed on the matter. The FT also had a good opinion piece recently: http://www.ft.com/cms/s/0/de089e8c-6999-11e4-9f65-00144feabd...

Net Neutrality is really a fight between two industries about shifting their costs onto one another, with really good PR from the tech industry that has enlisted normal people onto their side.

T-Mobile's zero-rating of data from certain music services supposedly violates "net neutrality", but customers like it and it allows them to compete with ATT and Verizon. It may be unfair to some other music service start up, but who cares? It's unfair in the same way that an incumbent has more resources to advertise their product than a start up.

There are generally 4 economic players at minimum:

  Network service like Netflix.
  Their ISP, like L3.
  Consumer's ISP, like Comcast (also a Netflix competitor).
  Consumer.

The end points are paying money to the ISPs in the middle.

There's also content providers, content owners, or other ugly term. They aren't participating in network traffic, but deeply shape structure and demands of at least network services and consumers above, consumer ISP in some cases.

This is a bullshit stance by Wikipedia. If when Wikipedia started there had been some popular ad supported encyclopedia that received preferential ISP treatment Wikipedia would not exist today.

Maybe it makes sense for the established internet companies to oppose net neutrality because they're the ones that would get packaged up by the ISPs. But it's bad for everyone else.

Does the end result of more people having free access to the information wikipedia provides, justify the zero rating/preferential treatment by ISPs of the site?

No, Because it closes out competition to newer, better information.

Like Demand Media? Would it be a bad thing for people to have free access to Wikipedia and Open Street Maps? They could always pay for Demand and GMaps if they wanted to.

>No, Because it closes out competition to newer, better information.

But that would only make sense if Wikipedia created some sort of insurmountable barrier for newer, better information. It doesn't. Newer, better information can reach people through Wikipedia with just as much ease (provided it fits Wiki's publishing standards) as it would through any other platform.

For all intents and purposes, Wikipedia and other similar open and free general knowledge sources can (and perhaps should) be regarded as a public good. We don't complain that the government's monopoly on building public roads closes out competition to 'newer, better roads'. We recognize that these roads are vital public infrastructure. Platforms like Wikipedia should be no different.

Wikipedia creates an enormous barrier to better information due to its size and scope, and its co-dependent relationship with Google.

In many cases Wikipedia's information is substandard when compared to other sources, both in presentation and content. The wiki layout is useless for certain content types, such as video.

For example, many people prefer to use imdb for movie information (which is also free to the end user), but it comes lower than Wikipedia on many search results. There have also been several recent studies that say Wikipedia's medical information is unreliable and dangerous.

So giving free access to Wikipedia, over other sites is not actually in the public good at all. The answer is more competition.

IMDB does have better quality information about their niche topic, but their layout sucks. Outside their niche topic, their information sucks. Trying to get even basic information about animals, major religions, or non-entertainer public figures, and IMDB is just plain awful.

As for video, both IMDB and wikipedia 'pop up' a box in which to play video, independent of the parent page layout; they don't seem to differ in that respect, and IMDB is a specialist site that's all about video.

There have also been several recent studies that say Wikipedia's medical information is unreliable and dangerous.

As opposed to...? Online medical literature in general is noted to suck, even the specialist websites. Even paid professionals - my housemate returned from the doctor two days ago after a norovirus scare... and the doctor claimed the incubation period before symptoms was 2 weeks... when it's actually 1-2 days.

Anyway, Wikipedia is a generalist site - basically you're arguing that it's not as good as specialist sites, so to free up access to information, all that better info should be free, hence all (useful) sites (effectively) should be free. I'm not sure that's going to go down well with ISPs.

> As opposed to...? Online medical literature in general is noted to suck, even the specialist websites. Even paid professionals - my housemate returned from the doctor two days ago after a norovirus scare... and the doctor claimed the incubation period before symptoms was 2 weeks... when it's actually 1-2 days.

Now get some good quality evidence - something like a Cochrane review or NICE guidance - and try to edit those medical pages.

> Wikipedia creates an enormous barrier to better information due to its size and scope, and its co-dependent relationship with Google.

Too silly to address.

> Wikipedia's medical information is unreliable and dangerous.

Wikipedia is an encyclopaedia not a DIY medical textbook. If you see dangerous information please feel free to edit it out — Wikipedia emphatically does not give medical advice (https://en.wikipedia.org/wiki/Wikipedia:Medical_disclaimer).

> The wiki layout is useless for certain content types, such as video.

New-ish media viewer allows video playback fullscreen or inline, it's no youtube but unclear in which way you consider this useless.

> So giving free access to Wikipedia, over other sites is not actually in the public good at all. The answer is more competition.

Firstly Wikipedia Zero does not preclude, and indeed hasn't precluded, Facebook zero etc. Regarding competition, you know Wikipedia is CC-BY-SA right? You and everyone else has a right to fork the entire project, along with the mediawiki software...

>Wikipedia is an encyclopaedia not a DIY medical textbook. Wikipedia emphatically does not give medical advice.

That disclaimer (hidden in the terms and conditions section) does not change the fact that many people do use Wikipedia for medical information. Each page that contains medical content should have its own disclaimer stating that the information contained in it has not been checked by doctors for accuracy and could have been changed by anyone at any time and could be incorrect. As this article points out [1]:

They discovered that 90% of the entries made statements that contradicted latest medical research. Lead author Dr Robert Hasty, of the Wallace School of Osteopathic Medicine in North Carolina, said: "While Wikipedia is a convenient tool for conducting research, from a public health standpoint patients should not use it as a primary resource because those articles do not go through the same peer-review process as medical journals."

>If you see dangerous information please feel free to edit it out

But how would I know if it's dangerous? I'm not a doctor. Your comment encapsulates the problem. You are asking me, a non-trained person, to make medical judgments on behalf of others.

When Google chooses Wikipedia pages to be the top result for medical queries, instead of peer-reviewed information then that is an enormous barrier to entry.

[1]http://www.bbc.com/news/health-27586356

I think you are completely misunderstanding the purpose of Wikipedia Zero if those are the use cases you're showing me. It is not meant for people to get information about movies and watch videos. It is meant to give them some basic level of information about the world around them. No other online resource has the breadth and variety of Wikipedia.

> There have also been several recent studies that say Wikipedia's medical information is unreliable and dangerous.

Nobody is going to use Wikipedia to diagnose diseases. But suppose I live in a country with a cholera problem and search for 'chronic diarrhea' on the web. Wikipedia will be able to point me to a few probable causes, and will give me some primary hints about what I'm supposed to do.

If there isn't a doctor nearby to tell me this, and there isn't a public library with a medical encyclopedia nearby, and I can't afford a PC with internet access, how else am I supposed to find out?

Trying to fight against the deletionists on Wikipedia with their spurious claims is a barrier to entry.

We do (and did) complain about Microsoft giving its browser away for free; in the end, we got lucky that there was a well-funded nonprofit (Mozilla) and another monopolistic company (Google) willing to put in the massive investment to compete in that market, but it took years.

Similarly, I may happen to trust Wikipedia itself, but the tactics it is using are very abusable to keep a for-profit's inferior loss-leader product on top.

But my contention is that Wikipedia should be regarded as fundamentally different from Microsoft or Google. They are not trying to compete with other encyclopedias. Why would they? They have nothing to gain by beating the competition. They are not making monetary profits. They are not paying to get themselves this advantage.

Suppose an NGO decided to provide expensive vaccines to children in Africa for free and enlisted the help of a local delivery company to ensure the vaccines reach as many people as possible, would you accuse them of trying to out-compete drug companies?

Why is Wikipedia different from this? They too are trying to provide an essential service (information) to people who could not afford it under normal circumstances. They are doing so without expectations of profit. Their tactics are in no way similar to those of Microsoft, purely because they are not competing in a market economy.

Net non-neutrality used to overcome severe data caps so some poor buggers can browse the Wikipedia?

This really sounds like trying to add one wrong to another to try to make a right.

This really confuses issues too; metered access to a network isn't quite the same thing as non-neutral access where some content gets much better latency and bandwidth.

Cutting off access to all sources of content at a given data cap does in fact meet the definition of neutral: No Netflix for you, and no Usenet, no Skype, no IRC, no git, no ssh, ...

Data caps in fact hurt the high-bandwidth premium services that easily bust data caps, whose purveyors support non-neutrality.

This seems very similar to the postal system when it give free-of-charge service for humanitarian aid. It seems a very strict interpretation of the common carrier principle to disallow it on the basis that its discriminating against those who has to pay.

What is the threat model to net neutrality for cases when there is no exchange of payments?

The authors didn't address the point that all the current use cases for zero rating are on wireless, not on traditional broadband.

The Google-Verizon proposal from 2010 had this "wireless exclusion". I was opposed to it in principle, but now as a T-mobile customer I really love the fact that Spotify doesn't count to my usage limits. I can completely see how zero rating could benefit the developing world and even the elderly in developed countries - these demographics are very likely to own a cellphone even though they might not use home broadband.

I don't see anything wrong in providing a throttled experience for free as long as all paid traffic (including non-zero rated versions of Wikipedia that include pictures/video) is treated equally.

Wondering why you don't see an issue with zero-rating - it's essentially making an unlevel playing field, with the big players like Facebook/Spotify buying an anticompetitive moat against any new entrants.

I don't see how zero-rating is going to help network congestion either - won't zero-rated Spotify for my neighbors causes my calls to drop or my maps to not update?

Is zero-rating some bandwidth somehow magically evading Shannon's Law?

You brought up three issues: (i) Anti-competitiveness in music streaming (ii) Anti-competitiveness in social networks (in developing regions) (iii) Effect of zero-rated services on congestion

(i) http://www.t-mobile.com/offer/music-freedom-list.html This is a large enough list that I wouldn't worry about anti-competitiveness. It is still them vs. everyone else, but looking at the list right now, the barriers to entry cannot be that high, which means any anti-competitive effect cannot be that high either. I and most others in my income category (college undergrads) are willing to accept this in order to have free music streaming.

(ii) Facebook (in some parts of the developing world) is a slightly different issue, and not necessarily one that I agree with. Still, zero rating popular websites in the developing world seems to be win-win-win for consumers, carriers, and content providers, and ignoring market forces is simply not a solution. Especially if the target demographic simply cannot afford to pay for data - this is an entirely different market and I don't have the experience to form an opinion on this. Edited my original post to remove Facebook.

(iii) Let's assume all paid traffic should be treated equally. And to keep things fair to paying customers, paid traffic should be prioritized over free traffic. I'm willing to bet this already happens indirectly - by lowering the music quality when the network is congested.

But what if I have Spotify Premium - I (indirectly) pay for my traffic, so why should your maps app be any more important?

---

Obviously there are tough questions here, and a "one size fits all" net neutrality approach for wireless and broadband might not be the best solution.

If wireless was as cheap and fast as broadband, I would be all for net neutrality on wireless. But it's not. They are different 'products' in economic terms (even if they have a similar end result and "feel the same"), and I think it's therefore okay for policies that govern them to be different.

While I appreciate your thoughtful response, all of your points evade my questions.

Especially your (iii) which makes a pretty invalid assumption about paid traffic all paying the same rates. Do you know this is true for current paid traffic? I would bet it's entirely not true. Even if it is, paid traffic being prioritized over free traffic is essentially reinforcing the internet slow-lane.

Nothing of what you said replies to the difficulties in competing against Facebook or other internet giants when they're paying to prevent competition from newcomers.

Sorry, but you sound like a telecom/cable apologist.

I didn't say anything that could be construed as an apologist argument for the cable industry, or even the wireless industry. You can only accuse me of being a T-mobile apologist - and what they are doing hugely benefits me, considering my limited budget. And I'm not alone.

  Nothing of what you said replies to the difficulties in competing against Facebook or other internet giants when they're paying to prevent competition from newcomers.

1) I argued that zero-rating music services isn't as anti-competitive as it's made it out to be. Since 27 services (many of which I hadn't heard of) made it onto the list, it can't be that hard for a newcomer to get onto the list.

2) Facebook Zero is a developing world issue. I said earlier that I don't have the experience to comment on developing-world issues, but let me take a stab at it now (this is sort of long): if you lived in a part of Africa that had no electricity but had cellphone reception (this is quite common), there's a good chance you would rather have Facebook and Wikipedia for free on your phone, as opposed to having a level playing field among all services yet being able to afford none of them. (<-- TL;DR)

Of course Facebook isn't being purely benevolent in offering this for free - they want to increase their MAUs and ad revenue, and get a hold on the market. Which is why I said I don't necessarily agree with that they're doing. BUT, you can see how the argument has its merits.

Yes, this makes it hard for another social network to gain usage share in that market. Duh. It would be hard if you were a startup from SoMa, and it would be equally hard (but much more soul-crushing) if you were a startup from the same region and can't get what Facebook does from the local wireless carriers.

What do customers in such markets want? The bottom 90% of customers in such a market will almost certainly want Facebook and Wikipedia for free.

3) As for congestion: I just wanted to pose the Spotify Premium vs Maps question to you. No wireless company ever mentioned "reducing congestion" as the reason for their zero-rating policies (it doesn't make sense anyway), so the subject of congestion is out of place here, and your questions about congestion are completely irrelevant here. That part of my response was very poorly worded now that I read it.

  paid traffic being prioritized over free traffic is essentially reinforcing the internet slow-lane.

Not quite. The important difference is that it's paid for by the consumer, not by the content provider. Even today, those who pay more for a better internet package get their traffic prioritized over those who pay less and everyone's fine with this. See, for example, Google Fiber's slower free tier in Kansas City.

Are these services paying T-Mobile to be zero rated?

This is kind of at the peripheries of net neutrality. I don't recall any proposed "net neutrality" related regulations -- whether the 2010 Open Internet Order, the NPRM from earlier this year, or any of the concrete alternatives proposed -- that would have prohibited zero-rating on mobile networks that have data volume charges.

This isn't even the same thing as paid prioritization (which the recent NPRM may still have allowed, within some bounds), where edge providers are given a choice to pay the ISP or have packets get lower delivery priority -- the ability of the customer to connect and the quality of service provided on the connection isn't affected by the payment, the customer gets exactly the service they would get connecting to an edge provider that hadn't made a deal with the access provider. The only difference is that the customer doesn't have to pay for the data consumption (which essentially is equivalent to the edge provider reimbursing the consumer for the data cost, except presumably the costs are lower commensurate with the simplicity of a direct edge provider to access provider payment route.)

So, while certainly this kind of thing might raise concerns related to net neutrality, its not what the focus of the net neutrality debate and regulatory effort has been.

> So, while certainly this kind of thing might raise concerns related to net neutrality, its not what the focus of the net neutrality debate and regulatory effort has been.

Well, that's because this wasn't a threat (in the US) until relatively recently, unlike paid prioritization and/or throttling. That doesn't mean that zero-rating doesn't violate the same principles as paid prioritization does. If you take a look at the doomsday scenario that Fred Wilson outlines in "VC Pitches In A Year Or Two"[0], the situation could just as easily apply to a zero-rated world.

You can think of zero-rating as "slow lanes", the speed on the slow lane is set to zero. (In a way, zero-rating is actually worse than paid prioritization, because it achieves the same end result, but placing a more consumer-friendly facade on the same rotten interior.)

> This is kind of at the peripheries of net neutrality. I don't recall any proposed "net neutrality" related regulations -- whether the 2010 Open Internet Order, the NPRM from earlier this year, or any of the concrete alternatives proposed -- that would have prohibited zero-rating on mobile networks that have data volume charges.

I also don't recall any large-scale[1] efforts to impose zero-rating on mobile consumers until the last year or so, which is why net neutrality legislation hasn't focused on it. In general, legislation around consumer protections tends to be reactive rather than proactive, and net neutrality has definitely followed this pattern.

[0] http://avc.com/2014/01/vc-pitches-in-a-year-or-two/

[1] (Yes, they existed in some cases, but they were very limited and relatively obscure.)

There was no concept of 'zero-rating' before telecoms/ISPs established the concept of a 'rate-per-byte'.

well, early residential connections were paid per minute, so I don't see how that was any better. There's nothing wrong with rate-per-byte, because some of the costs actually are per-byte.

From memory:

* Dial-up in the 28.8k era was per-minute. Occasional $700 AOL bill. Mom was pissed.

* Dial-up in the 33.6k through 56k eras was flat-rate. Start a download before bed, enjoy it in the morning.

* Broadband, pre 3mbps era. Flat rate.

* Netzero: Back to dialup, but now free. (Especially if you replaced their dialer software with a pencil-and-paper key to the Caesar cipher they used for auth.)

* Cellular. Sorta based on SMS rates.

* Broadband, 10mbps. Some companies have caps. Only a subset of users notice them, nobody had their back.

* Mobile. Smart phones! But the device is the limiting factor, not the network.

* Broadband, every major provider has caps written into the contract, but few enable technical enforcement of the caps. I smell deception: "We've always had these caps, you must have increased your usage!"

* Mobile, post 3g: pay per byte, period.

Precisely. I am trying to see the "complex" moral dilemma part of this issue that Wikipedia claims but I can't. For a fixed price, people should get a fixed speed - on wireless or broadband. That's it. Every byte of data should be treated the same. Disadvantaged consumers should not get free FB or Wiki. Instead they should get free 64kbps or 1mbps (or whatever the market can bear).

"Zero rating" is harmful because content providers and telecoms get to decide what is worth providing for "free". It's never really free because in the end someone is paying for bandwidth.

Voice calls, text, and IP are all just data. If I pay for 20/5mbps, give me that. If I have a fixed speed of 20mbps, I do not have "unlimited data" so there is no need to monitor and cap. It is 20mbps / 8 * 86400 seconds * 30 days = 6TB per month max. 64kbps is 20GB/month max. Nobody's browsing 24/7 at max. speed on mobile so there is very little chance a mobile user in free data plan of 64kbps is really going to use 20GB of data per month.

Wikipedia should consistently support Net Neutrality throughout the world regardless of local deals. If they really want to help consumers in poor countries, fight for them to get a slow data plan that can be used not only for Wikipedia but also Khan Academy, MIT OpenCourseWare, or anything else the user wants to see.

While this is technically correct, it suggests a causation that just isn't there. Metered per-byte rates have existed since at least the 1990s. Datacenter bandwidth has historically always been metered, though typically on a 95/5 rate basis.

This is also how mobile data has historically been billed: data packages really only took off with the relase of the iPhone, and before that most people just paid per-kb (which is why nobody actually used their mobile data capability).

> data packages really only took off with the relase of the iPhone

Please, no more iPhone revisionism. T Mobile Web & Walk packages launched in 2005 with 1GB and 2GB monthly caps; the latter was the 'Pro' package IIRC. Other carriers quickly followed.

Certainly my Nokia E smartphones of the pre-iPhone era all used such packages for browsing / e-mail / IRC.

95th percentile billing is not like the per-byte billing that ISPs and other parties try to enforce. 95th percentile billing uses your instantaneous bandwidth utilization, which directly correlates to how you have to size network equipment, and therefore to how much it costs to operate the service.

Per-byte billing does not directly map to how much it costs your provider to transfer your data. I can utilize a link to send 1kbit/s for a month, which is 2.47Tbit of data transferred. That sounds like a lot, but what burden does 1kbit/s place on your provider? Virtually nil. And yet carriers with data caps will charge massive overage fees or heavily throttle usage because of usage like this.

AWS bills per-byte for outbound data, so I'll use that for a concrete example. Lets say that I have a small service that uses 1gbit/s for 30 days.

  Commodity transit:
  billed at 95th percentile
  $2.00 / mbit
  $1000 circuit commit
  = $2000 / mo

  AWS:
  billed per-byte
  $0.05 / GByte (using their cheapest tier for simplicity)
  = $16,200 / mo

There is no major difference here between what it costs the provider to transfer my packets. If anything, Amazon can get better deals on transit than a small startup can. Yet I would pay Amazon nearly ten times as much, simply because their billing metric does not bear any relation to their cost to serve. This is rent-seeking behavior and it should not be allowed to become the default way that ISPs and other service providers bill customers.

You comment is informative but it contains a magnitude error. It should be 2.47gbit, not 2.47tbit.

https://www.google.com/search?q=1+kibibit+per+second+*+30+da...

You're right. 1kbit/s over a month would result in 2.47gbit of data transferred, which is a bit less damning. To get to 2.47tbit, you would of course need 1mbit/s transferred (which is still not that much!).

However flawed my example, I think my point remains valid: data usage measured over arbitrary time periods makes no sense when the underlying infrastructure is built with blocks that only care about instantaneous transfer rates.

I agree.

If your meritocracy is just a snapshot of whoever's best at a specific time, then you don't have a meritocracy. In other words, what happens when a new, better source of information comes online?

Wikipedia Zero creates a financial penalty for users who want to preserve their privacy by using Wikipedia over Tor. I wonder what folks here think about the ethics of that.

Serious question, as I don't really have a strong opinion on zero rating either way:

In a scenario where zero-rating is available to all actors at equivalent prices, is there still a problem? It would raise the barrier to entry for internet companies perhaps, but probably not by too much. I guess you could argue it encourages bandwidth capping?

I'd like to hear arguments from both sides.

Not mentioned in the article is that Wikipedia refuses to pay to participate in zero rating. In the world you envision only profitable commercial sites would be available.

Unless that price is zero, that's exactly what net neutrality is trying to avoid- an uneven playing field.

>In a scenario where zero-rating is available to all actors at equivalent prices, is there still a problem? It would raise the barrier to entry for internet companies perhaps, but probably not by too much. I guess you could argue it encourages bandwidth capping?

An alternative, would be to allow zero-rating through free prioritization. Give either web services or consumers the choice of "this is an important connection: per-byte-rate and give higher priority," or "this isn't so important: zero-rate and give lower priority."

That way VoIP and other latency/rate-critical services could operate more reliably, while BitTorrent, home cloud storage and others could operate without congesting the network and without data cap problems. If this choice was allowed in an automated way, it would not seem to cause any unfairness, or keep smaller/newer players off the market.

For those who think Wikipedia is right in defending Wikipedia Zero, could you please elaborate on how this will not impact the adoption of any alternative encyclopaedic website?

Can you fork Wikipedia? I remember back in the day it was all open licensed and downloadable, but it's grown quite a bit since then.

This isn't a dilemma. You can have net neutrality and also have Wikipedia Zero -- it isn't mutually exclusive. It might make it a little harder and Wikipedia may need to ask for more donations, but this is their goal, not anyone else's. Wikipedia isn't the internet.

A better solution: no data caps, even if it has to be enforced by regulations.

But bandwidth is a limited medium. It is fair IMO that those who use it a lot should pay more,

Which doesn't mean having a data cap. It could be 1$ per GB on wireless devices (where bandwidth is really limited because of collisions with everybody's else traffic) or whatever amount it makes sense. Wikipedia pages are quite skinny.

Servers usually pay per GB with many VPS providers right now, with a monthly quota included in the base fee. If you have your own datacenter you pay per GB in other ways.

That said, subsidising the access to one own website could be seen as a fair commercial practice or anticompetitive behavior. It depends on so many factors that I don't even want to get into it.

There's a better solution than zero rating on mobile plans: a very low bandwidth (4-8 KiB/s) unlimited connection once the full speed limit is reached. TIM does this in Italy and it doesn't care what sites you access in this low bandwidth regime.

The thing is that zero rating can be promoted by platforms such as Wikipedia and Facebook, who want to promote net neutrality; unlimited low speed connection is an ISP-only thing, and I haven't seen many big ISPs (really) promote net neutrality.

Reply on news.ycombinator.com