It's a wonderful day in the Neighborhood. Would you be my Comrade?

May 27, 2009 11:24 AM

"When masses of people who own the means of production work toward a common goal and share their products in common, when they contribute labor without wages and enjoy the fruits free of charge, it's not unreasonable to call that socialism."The New Socialism: Global Collectivist Society Is Coming Online, a provocative article in the new Wired magazine, examines the effects of the growing influence of online collectivism. I thought this might make for an interesting read and discussion by members of an online community.
posted by Benny Andajetz (63 comments total)

I'm only a page in, but the pull quote you posted seems to have a problem in that masses of people still don't 'own the means of production' - the means being a physical server somewhere. Without hardware, no one has ANY means of production in his examples.
posted by spicynuts at 11:36 AM on May 27

A few things undermine the articles point:

1. Wired's poor record of breathless prognostication (c.f., the long boom).


2. Wikipedia's declining reputation; or rather, the decline from overhyped saviour of knowledge to the more modest reputation of "useful for many bits of common knowledge, but not particularly authoritative."


3. Linux's decade long failure to realize "the year of the linux desktop"


4. The incredibly high ratio of failed open source/community projects to successful ones.

What's really happened is that digital socialism has been around and successful long enough for us to see its limitations: the fact that, as Mitch Kapor puts it, there's an old-boy network at the heart of every working anarchy, means that such efforts are often and easily derailed by a powerful minority's groupthink (c.f., Wikipedia's 'deletionists'); the way Linux and the BSDs are very powerful in certain areas like networked servers and still poor in areas that prevent widespread adoption among common users (e.g., printing, sound); and the way that only a chosen few examples of digital socialism ever really succeed in building a large community.
posted by fatbird at 11:39 AM on May 27 [10 favorites]

spicynuts: further along the article seems to imply the product is intellecutal property and the means of production are (naturally) the computers and things we all own. But yeah, it depends on what level you want to look at it (i.e. as the internet as an abstract "place" or as actually being made of servers and wires)
posted by selenized at 11:44 AM on May 27

This is a cute quote from the article:

The organization behind MySQL, an open source database, is not romantically nonhierarchical, but it is far more collectivist than Oracle.

Since Oracle now in fact owns MySQL (after buying Sun), and the project itself has forked about six different ways, and no one's sure anymore which fork should be considered the proper inheritor of the MySQL legacy.
posted by fatbird at 11:44 AM on May 27

metafilter:
posted by infini at 11:46 AM on May 27 [7 favorites]

oopsie, finger slips

metafilter: masses of people who own the means of production work toward a common goal and share their products in common, when they contribute labor without wages and enjoy the fruits free of charge
posted by infini at 11:49 AM on May 27

You'd be surprised how much our physical (I believe the nerds call this "meatspace") society depends on collective action. You know, things like Little League, Meals on Wheels, community gardens, church groups, community associations, parent-teacher associations, unions, professional associations, etc.

Wired, as usual, is breathlessly waxing poetic with its head up its ass.
posted by KokuRyu at 11:54 AM on May 27 [5 favorites]

spicynuts: further along the article seems to imply the product is intellecutal property

Yeah I'll have to give it a more thorough read but I don't think it works. With the internet, you cannot separate the product from the means of distribution. The intellectual property accessible on the internet CANNOT exist without the physical hardware to store and display it and there is zero ownership of that tools of production of that hardware by the masses. Given that, then a digital arena for IP is infinitely less socialist than a park bench where two people hang out talking about ideas.
posted by spicynuts at 11:55 AM on May 27 [1 favorite]

The incredibly high ratio of failed open source/community projects to successful ones.

Does the incredibly high ratio of failed restaurants to successful ones undermine the idea of eating out?
posted by l33tpolicywonk at 12:04 PM on May 27 [2 favorites]

From each according to his e-bility, to each according to his feed.
posted by joe lisboa at 12:05 PM on May 27

Wired, as usual, is breathlessly waxing poetic with its head up its ass

Well in all fairness if your head is up your arse it is going to be difficult to breathe.
posted by spicynuts at 12:09 PM on May 27 [4 favorites]

metafilter: masses of people who own the means of production work toward a common goal and share their products in common, when they contribute labor without wages and enjoy the fruits for $5.
posted by furiousxgeorge at 12:10 PM on May 27 [3 favorites]

Does the incredibly high ratio of failed restaurants to successful ones undermine the idea of eating out?

No, why should it? The high ratio of failed open source projects doesn't undermine the use of Apache or Linux in roles that are appropriate.

What the comparably high rates of failure undermine is participation. Just as many shy away from opening a restaurant because the odds are against them, so many examples of digital anarchy in action go nowhere or taper off and wither away simply for failing to be the next Linux. To waterboard this analogy a little more, Kelly is saying that restaurant ownership is the wave of the future, when it's obviously not because it has definite limitations.
posted by fatbird at 12:14 PM on May 27

Ok, i've dutifully read the article. Its all about openness and sharing, stuff for free and all the online goodness you can inhale.

then how do we reconcile it's message (which one assumes is global and not just the bay area) with snippets like these? Build walls to keep the 'unprofitable' 4/5ths of the world out ?

Last year, Veoh, a video-sharing site operated from San Diego, decided to block its service from users in Africa, Asia, Latin America and Eastern Europe, citing the dim prospects of making money and the high cost of delivering video there.

“I believe in free, open communications,” Dmitry Shapiro, the company’s chief executive, said. “But these people are so hungry for this content. They sit and they watch and watch and watch. The problem is they are eating up bandwidth, and it’s very difficult to derive revenue from it.”
posted by infini at 12:16 PM on May 27 [1 favorite]

But there is one way in which socialism is the wrong word for what is happening: It is not an ideology.

One thing in common for all ideologies is that those who share it would never call it an ideology.
posted by daniel_charms at 12:22 PM on May 27 [1 favorite]

Wow, the author missed one simple fact, and this completely blows up his argument.

This isn't socialism. This is capitalism. But it's capitalism that takes advantage of zero marginal production cost.

In physical capitalism, you have an upfront cost to start manufacture of something, the capital goods required to create, say, cars or sofas or Tinker Toys. This could be the simple physical workspace and basic tools for furniture, the extremely complex tool-and-die infrastructure with cars, or the unbelievably complex infrastructure required to create computer chips. Then you have a marginal cost of production, which is how much it costs to create each individual car or sofa or Tinker Toy, after the production facilities are built.

Typically, over time, the cost of goods tends towards the marginal cost of production plus a profit that's not quite adequate to pay for startup costs. It may dip below that for a period of time, but if it does, then producing the good becomes unprofitable, and manufacturing slows or stops until the price returns to a profitable level. If it gets too high, then the available profits encourage investors to create new capital goods to produce those items.

So, with the advent of digital media, suddenly your marginal production cost is zero. Almost the entire cost of the good is front-loaded into the first copy. From there, you can make millions or billions of copies, with the very small cost being absorbed by existing paid overhead.

So, you are seeing market forces organize themselves in new ways around the fact that CREATING digital goods is expensive, but REPRODUCING digital goods is free. It makes perfect sense to cooperate in the creation process and then share the reproduction with everyone. This is the price tending toward the marginal cost of production.

Some companies are able to make a small profit by selling services to go with free digital goods. If the profits get too high, more service companies will spring into being; if the profits get too low, some will go away. Mr. Market is, once again, exerting his invisible hand to keep things in balance.

That's all this is. It's just sharing the high cost of creation. And, unlike socialism or communism, it's entirely voluntary -- you don't HAVE to work on open source anything if you don't want to. If you're smart enough to stay ahead of the people doing it mostly as a hobby, then you can make a very comfortable living. See: Microsoft or Electronic Arts or Activision. Socialism and communism are horrid because they're enforced at gunpoint; copyleft is enforced simply by depriving you of the ability to use copylefted software if you don't abide by the terms.

The RIAA and MPAA will, no doubt, argue that they're being forced into a communistic relationship with file-sharers, but I disagree. This is, again, market forces responding to a zero marginal cost of production; millions of individual people are taking over the "manufacture" of movies and music in their own homes, because it doesn't cost anything. The recording industry is desperately trying to use the guns of the government to preserve an obsolete business model by force, so they're far more communist than the digital sharing community could ever be.
posted by Malor at 12:23 PM on May 27 [19 favorites]

First, neat packages like socialism are hard to use in reference to online matters. To make the analogy, politics would be much different if everyone could migrate whereever they wanted and the means of production were given away.

Two, as fatbird says, online communities run in a "socialistic" way have been pretty much part and parcel of the Internet since the beginning. Usenet (there is no cabal), RFCs, MUDs, The Well.

To be as breathless as Wired is to ignore the massive amount of drama, the blow ups, the slow withering away of communities, and the assorted failures of the years. Few of them rarely reach the critical mass of "too big to fail" like Linux.

Add to this the owners of the means of production (webservers, etc.) are either companies (which can decide to revoke these means) and individuals (who can do the same), you're left with something that can't be relied upon to be around forever, or even through the end of the year.

I wish that more Internet prognosticators would be a bit more like Clay Shirky. He at least includes the negatives and problems.
posted by zabuni at 12:31 PM on May 27

Socialism. Well, no. It's cool when people get together and build useful stuff gratis, but they've been doing that for a long time under capitalism (barn raising, softball teams, meet-ups). But the stuff that actually costs something to produce, well, you're gonna have to pay for that. Or you can just steal it, call it "file sharing" and proclaim a bright new day in the history of humanity.
posted by MarshallPoe at 12:33 PM on May 27

It's interesting to note that there have been efforts to create internetworked systems where the users really would own everything, and there would not be a dependency on some giant corporation running backbone fiber and charging everyone rent via their ISP bill.

As far back as the 1980s, ham radio operators were working on an AX.25-based packet radio networks that would allow various types digital communication between stations in the network, without any infrastructure besides the user-owned endpoint equipment. Today we'd call it "peer to peer" or maybe "P2P with supernodes," but the key was that all you'd need to do was set up a radio and computer, make sure it could talk to another station or stations in the network, and you'd be able to contact anyone else. They weren't generally TCP/IP (that was added later), but the intent was similar.

There have been similar efforts at creating mesh networks over the years; the more recent ones use 802.11 gear on 2.4GHz and don't require licenses. To my knowledge they've all failed.

There are technical problems with such networks — they can run into nasty scaling issues if the traffic isn't symmetric between nodes — but the main problem is that "owning the means of production" requires an upfront cost that most people don't want to pay. It's a lot easier, and in the short term cheaper, to rent your connectivity from someone else. Plus, most decentralized networks, especially the radio-based ones, are very slow.

I remember back in the early 90s there was a lot of excitement over this kind of stuff; some people really did think, I believe, that mesh networks would displace the phone company. (If you want to get a sense of that, just read any of Doctorow's early books.) Instead, the "phone company" just moved from carrying analog voice signals to IP packets, but really didn't need to change that much of its business model. (At least at the higher, long-lines levels; the local telcos have obviously been affected by competition with cablecos and other last-mile options.) They still own the really important infrastructure, the backbones.

The modern Internet has virtually everyone paying rent to the owners of the core infrastructure, but that rent is so low relative to the perceived benefit of the service that there's not a huge incentive to create alternatives. So what you really have is people generating a ton of intellectual/social equity — all those free software projects and other stuff that's given away, including things as trivial as MeFi comments — using rented facilities.

I don't really think that's a bad thing; users own the products, or choose to give them away such that nobody really "owns" them or they're owned collectively in some sense, but they certainly don't own "the means of production" except insofar as the key means of production is their brain (which hopefully haven't mortgaged that out). As long as the rent you pay to access the Internet stays relatively low, it's a pretty good deal.
posted by Kadin2048 at 12:36 PM on May 27 [5 favorites]

I've never understood why the more appropriate metaphor of 'barn raising' isn't used instead. It avoids the civilization-level baggage of terms like socialism, and far better captures the feel of (immediate) community participation in a shared enterprise that benefits all.
posted by fatbird at 12:39 PM on May 27 [3 favorites]

I pretty much agree with all the points made, but I took away a different impression of the article.

I don't think the author was referring to doctrinaire or "political" socialism; he was referring to more of attitudinal socialism. People are way more comfortable with the process of creating things through cooperation and ad-hoc synergy (for lack of a better term). Class, color, background, location, politics, etc. are becoming less and less relevant to a large group of earthlings.

The most interesting point to me was at the end:We underestimate the power of our tools to reshape our minds. While I think his analogy might be a little shaky, he's more right than wrong.

As for using "barn-raising"- I think that fails to capture the global notion. This whole thing is about community writ large.
posted by Benny Andajetz at 1:13 PM on May 27 [2 favorites]

The article links to the wrong John Barlow.

John Perry Barlow.

I guess I'm a socialist now.
posted by mecran01 at 1:13 PM on May 27

Class, color, background, location, politics, etc. are becoming less and less relevant to a large group of earthlings.

A large group of middle-class earthlings.
posted by stammer at 1:19 PM on May 27 [4 favorites]

The incredibly high ratio of failed open source/community projects to successful ones.

What about the incredibly high rate of failure in conventional startups? Does that undermine internet capitalism?

And, unlike socialism or communism, it's entirely voluntary -- you don't HAVE to work on open source anything if you don't want to.

From where I'm sitting, capitalism isn't voluntary. I have to have a job to make money to feed, house, clothe, and provide health care for me and my family. I may have more options as to how to do these things than in a socialist or communist society (or I may have the perception of more options when in fact my workable options are quite limited), but participation in the system is mandatory.
posted by vibrotronica at 1:21 PM on May 27 [2 favorites]

A few things undermine the articles point:
Wired's poor record of breathless prognostication (c.f., the long boom).

A reasonable point

Wikipedia's declining reputation; or rather, the decline from overhyped saviour of knowledge to the more modest reputation of "useful for many bits of common knowledge, but not particularly authoritative."

I'm not really sure of that, if anything the only thing Wikipedia has really lost is its novelty. I mean, You want to look anything up and there it is Wikipedia really is, for almost any topic is what we all were promised with the with the world wide web back in the early to mid 90s. It isn't that Wikipedia failed to deliver on its promise, but rather that it succeeded and has since become boring and commonplace. Just like how sending a man to the moon seems impressive but not unthinkable, just like flying across the country in a jet is for most people a tedious annoyance rather than an incredible marvel of human ingenuity.

Wikipedia sacrifices the promise of accuracy for convenience, for sure. But other "authoritative" sources also have errors. And for most people searching for their own curiosity Wikipedia solves the problem so well that other sources aren't even needed (or in many cases even available quickly).

Most Wikipedia articles also include references for people to look up the authoritative data if they want to.

Jimbo Wales is a huge duchebag, obviously, but wikipedia worked I think more despite him then because of him and his friends.

Linux's decade long failure to realize "the year of the linux desktop"

Well, Linux dominates on the server, and T-moble has sold tons of Android phones, which run Linux. The palm Pre will also be based on Linux so in a few years millions of people will be doing lots of "personal computing" on Linux based platforms. There's also the firefox web browser.

The incredibly high ratio of failed open source/community projects to successful ones.

This is a bit silly; most Open Source projects are either personal projects that the sole author blows off, or the last gasp of corporate products no one cares about. Those projects wouldn't have survived if they hadn't been open source. Open sourcing software won't make people interested in software they wouldn't otherwise care about.
Complaining about the "high ratio" of failed open source software is like complaining about the "high ratio" of unfinished to finished novels in the world.
---
I do think we're likely to see far more of an 'open culture' type viewpoint in the future as more people who grew up digital and are used to the internet become more prominent.
posted by delmoi at 1:25 PM on May 27 [4 favorites]

You know, things like Little League, Meals on Wheels, community gardens, church groups, community associations, parent-teacher associations, unions, professional associations, etc.

Are volunteer institutions as strong in (relatively more Socialist) Europe as they are in the US?
posted by ZenMasterThis at 1:34 PM on May 27

Kadin - it's roads vs railroads.

Roads and highways are built and maintained by government for the benefit of all, and the money comes out of taxes on property, gas, licencing.

Railroads are built and run by private companies and they won't build or provide service if there's no direct short term benefit to them for doing so.

Guess what? In North America, roads go everywhere, most freight is trucked, and railroads currently suck.

Oversimplified, sure, but this is the situation we're coming up to with the Internet. The private companies that built the Internet infrastructure (and not initially of their own initiative, but under contract to government) are now seeking every opportunity to wring more profit from their investment, and this could has taken the form of traffic throttling and filtering, or creating "priority' packet service which for extra cost will receive preferential carriage, by shifting bandwidth from others. Other limitations and tolls will doubtless be enacted, unless the network providers are regulated into providing unfiltered connection for all.

So, yes, good deal now, but not guaranteed forever.
posted by Artful Codger at 1:35 PM on May 27 [2 favorites]

You mean you guys don't get paid to comment here?
posted by Faint of Butt at 1:36 PM on May 27 [1 favorite]

What about the incredibly high rate of failure in conventional startups? Does that undermine internet capitalism?

Not at all. It wasn't my point that the high failure rate is somehow fatal to digital anarchy itself, only that it's limited in its ability to conquer all, contrary to what Kelly is implying.

I'm not really sure of that, if anything the only thing Wikipedia has really lost is its novelty.

That's part of it, but Wikipedia has definitely hit some walls in terms of usefulness that it seems unable to overcome. The reign of the deletionists is part of it, but more generally its reputation seems to have capped at "useful for most things in a general way, but don't depend on it alone." Which is to say that Wikipedia is a valuable communal effort, but falls short of its own mission of encapsulating all human knowledge.

Well, Linux dominates on the server, and T-moble has sold tons of Android phones, which run Linux. The palm Pre will also be based on Linux so in a few years millions of people will be doing lots of "personal computing" on Linux based platforms. There's also the firefox web browser.

I'm not sure your example of phones and PDAs makes the point about digital anarchy that Kelly is trying to make, since in those cases capitalist entities are taking what the anarchic effort produces and packaging it up very nicely, specifically to address the shortcomings of the anarchic effort. Apple also falls into this category, with Darwin being the open source kernel but with Apple's own developers adding what's needed to make it a successful consumer-level desktop system.

I'm not sure that Firefox is a good example, either, since the Mozilla Foundation is heavily dependent upon funding from first AOL, then Google, and most of the key developers are paid by the Mozilla Foundation.

Complaining about the "high ratio" of failed open source software is like complaining about the "high ratio" of unfinished to finished novels in the world.

I'm not complaining about it, just observing that digital anarchy's model has its problematic side--just like internet startups and restaurants. Many anarchic efforts simply fail to gain or keep the interest of a sufficient community. To borrow your analogy, what Kelly's doing is proclaiming the age of the novel for everyone because Tom Clancy and Dan Brown are demonstrably successful.
posted by fatbird at 1:52 PM on May 27

This guy really sounds like Lenin - except that where Lenin claimed that socialism is power of the soviets plus electrification of the whole country, this guy thinks that it's "Volunteer group work a la Wikipedia plus real-time Twitter and RSS feeds".

On a more serious note, I think that if he wants to call Wikipedia 'socialist', he's free to do so. 'Socialism' doesn't denote some concrete, fixed object; it's a general term that covers a wide range of ideas, which often have little in common. You can easily fit all the things he talks about under some really general notion of socialism. Of course, this raises the question whether there's any point in calling Wikipedia, Linux, Twitter, etc 'socialist' phenomena. What does it give us? Well, it could help put these recent developments into a historical perspective, relate them to the collectivist ideas of 19th century, but the guy who wrote this piece somewhat oddly seems to be trying to avoid any parallels between historical socialism and "new" socialism*. Why, then, use the term 'socialism' in this context? What's the value of this word? In my opinion, this is the most important question he should have tried to find an answer to - but it seems like this is one question he didn't even stop to ask.

* I'd write this down to his little knowledge of socialist theory and the history of socialism - it's as if all he knows about it he's learned from History of the Communist Party of Soviet Union: A Short Course...
posted by daniel_charms at 1:52 PM on May 27 [1 favorite]

It's like since Obama got elected that the entire Western world has re-discovered socialism and is intent on completely misinterpreting Marx.

yes, this is really about zero-cost marginal production and the completely radical economics of digital goods. As much as I dislike people labeling the old as new, labeling the new as old is equally annoying.
posted by GuyZero at 3:08 PM on May 27

Dude doesn't know shit about the history of socialism, the theory of socialism, or the history of socialist theory. This article's basically a mass of know-nothing bloviation and should not be paid attention by anyone.
posted by Pope Guilty at 3:10 PM on May 27 [1 favorite]

"some people really did think, I believe, that mesh networks would displace the phone company."

So did the phone companies, which is why they fought tooth-and-nail to stop stuff like municipally-owned WiFi.
Until such time as people can understand and be motivated by their own power, the 'divine right of kings' will continue to prevail in their worlds.
posted by Twang at 3:34 PM on May 27

I wonder if there isn't a core fallacy here that digital reproduction on a massive scale is really zero cost. Wikipedia is constantly struggling to raise funds to keep up with storage, bandwidth and power. YouTube may be losing something like 450million /year from what I've read. Facebook has yet to turn a profit. Most online newspapers are losing money. Having run some big sites in my time and seen these distribution and maintence costs on a P&L, I think we have made a big mistake assuming "free" distribution. Even MeFi costs $5.
posted by humanfont at 3:36 PM on May 27 [2 favorites]

"I don't really think that's a bad thing; users own the products, or choose to give them away such that nobody really 'owns' them or they're owned collectively in some sense, but they certainly don't own 'the means of production' except insofar as the key means of production is their brain (which hopefully haven't mortgaged that out). As long as the rent you pay to access the Internet stays relatively low, it's a pretty good deal."

Actually, it's not a good deal. We (the US taxpayers) subsidized the building of the networks which we now use, and some which aren't yet available (lots of dark fiber). The idea behind deregulation is that we would create incentives through tax breaks to build out a network which would be 100Mb/s+ in exchange for some common carrier rights, like anyone could compete over the network, and nobody would hold the keys to the kingdom (e.g., the way it works now is, if you want to be a DSL provider other than the phone company, in most areas you are subject to whatever the phone company wants to do with the partnership, which usually results in a severe disadvantage for the partner/competitor). We already paid for it. It's not really built, because the companies which made these promises did not live up to them, but they took our money anyway, and now we aren't even allowed to compete on a level playing field.

This IS our network already. We paid for it.
posted by krinklyfig at 4:17 PM on May 27 [1 favorite]

From where I'm sitting, capitalism isn't voluntary. I have to have a job to make money to feed, house, clothe, and provide health care for me and my family.

Well, fundamentally, humans have to eat and drink, and need shelter and clothing to survive. This is not optional. Never has been. You are not free to do whatever you want, assuming you want to keep eating. You must provide yourself with these basic necessities, or you will die.

Communist and socialist societies may hide this fact from you, by compelling people who grow food to give you enough to eat, but SOMEONE has to do the work to keep you breathing. In that sense, we are not free, and never will be.
posted by Malor at 4:35 PM on May 27 [3 favorites]

Malor: You're not one of them Randians/Libertarians, are you? We LOL them hard around these parts.
posted by ZenMasterThis at 4:49 PM on May 27

I can't tell if repeatedly reminding people that there are people all over the US who still have minimal to no home internet access and minimal to no access to people who can teach them how to use it when (or if) they DO get such access makes me a protector/defender of the underdog or a crank. I know that it's important to have people who dream big dreams and who can act like Flickr's 3 billion photos somehow implies that every other person on the planet put a picture on the Internet. Realistically, however, 3 billion people on the planet have never made a telephone call.

This new collectivism is just a different way to slice through social strata so that you can interact with people who share certain interests that don't happen to be geographic. However one of those interests still need to be technological in nature, pretty much. This is good news for dorks like me who love technology and also living someplace rural [but not so rural that I can't pay for broadband] but this isn't great news for people who are being asked to file taxes and apply for jobs online and have a hard time figuring out how to double click. Meanwhile the public librarian, often the overseer of the only public broadband in town, struggles to even get health insurance on her 16 hour a week job.

And Verizon who was the telco here until last year, ran away from Maine, New Hampshire and Vermont because they'd made promises that (for example) they'd get 80% of VT on broadband (or with access to it) by 2010 and 2008 came around and they weren't close. They sold their lines to another company who is struggling while they kept their non-unionized wireless company which is still profitable because it's significantly less regulated. I swear sometimes the only reason parts of Vermont have electricity at all is because Rural Electrification finally made someone pay for it. So, yeah, capitalism isn't doing the best job here but collectivism is notoriously bad at building infrastructure. People can put content on servers all day long but at the end of the day if you don't have root access, you're not doing anything but lining someone else's pockets.
posted by jessamyn at 4:49 PM on May 27 [4 favorites]

3 billion people -- and dropping -- on the planet have never made a telephone call.
posted by ZenMasterThis at 4:54 PM on May 27 [1 favorite]

ZenMasterThis: You're not one of them Randians/Libertarians, are you?

Lowercase-l. When some companies would use pregnant mothers as building material if they were legal and ten cents cheaper than lumber, uppercase-L Libertarianism would be hellish.

The lowercase flavor, though, looks a lot like the dictionary definition of liberalism. Don't confuse this with the modern Democratic party, which appears to have mostly gone insane.
posted by Malor at 5:09 PM on May 27 [1 favorite]

I've never understood why the more appropriate metaphor of 'barn raising' isn't used instead.

Because once a barn is raised, you have one barn. More barns require more barn-raising.

Open Source software is more like the creation of a meta-barn, from which infinite barns can be instantly cloned, as long as you have some land on which to place a new barn. Also, your barn will be incompatible with several kinds of cows, but you're free to patch the meta-barn to improve compatibility with your cows. Later, when Meta-barn 2.9.1 is released, other cow owners will benefit from your cow patches.

Wait, I think we just had a serious metaphor collision there.
posted by verb at 5:10 PM on May 27 [4 favorites]

Is a cow patch something you put ON a cow, or something you put a cow IN?
posted by ZenMasterThis at 5:18 PM on May 27

That depends. What version of Meta Barn are you running?
posted by verb at 5:23 PM on May 27

I wonder if there isn't a core fallacy here that digital reproduction on a massive scale is really zero cost.

You wanna see free? Post your new movie on The Pirate Bay, and if it's any good, there will be millions of copies worldwide within hours, at no cost to you beyond seeding it a few times.

It's not that it's ACTUALLY free, it's that everyone with broadband access is paying fifty bucks a month for bandwidth, and they're choosing to use their allotment to help move your movie. The cost per person is just a few pennies, a tiny fraction of their fixed overhead. They pay the same whether they use it or not, so they use it.

The abstraction breaks down when you start to scale large enough. When you want to move one movie, no big deal. If you want to move a hundred million of them, and you want all the bits coming from servers under your control, well.... that's gonna cost you.

Small amounts of distributed bandwidth are extremely cheap. Large amounts of centralized bandwidth can be very expensive indeed.
posted by Malor at 5:31 PM on May 27

Yeah, but where do I download the food?
posted by General Tonic at 5:40 PM on May 27

I like how pretty much everybody assumes "from each according to their ability, to each according to their needs" means that they, personally, will be worked to death and recieved nothing.
posted by Pope Guilty at 5:53 PM on May 27 [2 favorites]

"I'm not sure your example of phones and PDAs makes the point about digital anarchy that Kelly is trying to make, since in those cases capitalist entities are taking what the anarchic effort produces and packaging it up very nicely, specifically to address the shortcomings of the anarchic effort. Apple also falls into this category, with Darwin being the open source kernel but with Apple's own developers adding what's needed to make it a successful consumer-level desktop system."

As for the anarchic effort, as you call it, there are many companies which contribute money and code to Linux and related projects, specifically because their benefits are exponential. Yahoo and Hotmail (before MS) were built on FreeBSD, and even to this day, Yahoo still contributes back a lot of useful code and helps to fund the project.

I think you might misunderstand how this works. Open source is not meant to make all software free, nor is it meant to completely replace efforts like the Palm, which are very market-driven and can only be successful with a company behind it. The cell phone industry runs on a model of being locked-in to a carrier with a 2 year contract with a new phone with renewal. There is no way that a software project alone can predict or manage bi-annual, large scale hardware releases, so it's up to the industry which created this model to find a way to make open source useful for them, if it is. And it certainly can be.

There is a significant long-term advantage to developing your own refinements to an existing project rather than becoming dependent on a single vendor with a proprietary product, one whose development model and cycle may not jibe with your business goals or model. That's why Yahoo went with FreeBSD way back when. They developed a highly customized version which they used for their own purposes. You can't do that with Windows. A lot (most?) routers use specialized versions of Linux for their firmware. NetBSD is used widely in embedded applications. And that is why a lot of the commercial applications for open source go unseen, because they're internal projects or work in the background. People use it every day and aren't aware of it. All the developers ask, in general, is that people who find use for the software help to keep it going, either by contributing back code or resources or by abiding by certain license requirements which encourage the ongoing development of such valuable tools which we can all share and improve.

"I'm not sure that Firefox is a good example, either, since the Mozilla Foundation is heavily dependent upon funding from first AOL, then Google, and most of the key developers are paid by the Mozilla Foundation."

Yes, precisely. Large projects like Firefox require a core which has the resources to maintain the project adequately. That doesn't mean it becomes a commercial operation or that its sponsors have any real control over the project. If a project gets to this level, it requires a bureaucratic underpinning which will allow it to deal with costs, because at that point, there are costs to the project. And it works pretty well like this, although there certainly are valid objections to Mozilla's structure, things like that, but the project remains open, volunteer-based and free to use and modify. If 200,000,000 people want to download it, well, then you have to find some resources for the servers and so forth, and some working arrangements with the biggest users. That doesn't mean the project becomes closed or proprietary, it just means it's needed enough where it requires some sponsorship, but not ownership, and there is a significant difference.
posted by krinklyfig at 6:07 PM on May 27 [2 favorites]

I wonder if there isn't a core fallacy here that digital reproduction on a massive scale is really zero cost.... Having run some big sites in my time and seen these distribution and maintence costs on a P&L, I think we have made a big mistake assuming "free" distribution.

Well, that's a little complicated. There are a couple of things that fall under the 'digital production costs' label.

1. The amount of work needed to create the first 'instance' of a particular thing (taking a picture and developing it, writing a book, writing a piece of software)


2. The amount of work needed to create each additional instance of that particular thing (printing a second copy of a picture, printing a second copy of a book, duplicating a digital file)


3. The infrastructure necessary to make each additional copy (a photo lab, a printing press, a computer)

The energy invested in figuring out HOW to create something (i.e., choosing the words, imagining a story, composing music) is pretty much a constant; the post-it note is about the biggest development in thinking and idea management I've seen in decades. Whether you're talking digital of concrete physical goods, this part also stays the same.

The work that is needed to actually produce a concrete implementation of that thing (playing the music, painting the image, recording the story in a transmittable form, whatever) is slightly more variable and depends on the kind of thing being produced. Software takes lots and lots of time, even if you properly imagine how it should work. Painting a painting, too. Some kinds of work (recording an album, making an indie movie) is getting easier, but that's because tools are getting cheaper, and digital media reduces the cost of a failed attempt to 'the time it took you to make it' rather than 'the time it took you to make it, plus the cost of lots of ruined film'.

The next category is the one that really, really changes. The iterative cost of producing the next copy of any given thing is completely changed by digital media. Software, music, video, photographs, whatever -- as long as you have space to store it (we'll get to that in a moment) the cost is effectively however many electrons you need to power the computer while it's being copied. Compared to printing physical books, that's huge. Even transmitting the copy to another location, if done digitally, has a minuscule cost compared to, say, shipping a printed book or sending a pack of photographs.

The final category is still an issue: cumulative cost of running the infrastructure that lots of people use to store and distribute their digital content. That's not quite the same, though, as those other things. It's a bit like comparing the cost of running FedEx to the cost of sending a package. They are related, but very different in nature.

Even MeFi costs $5.

Not quite. Access to the program that creates new content on metafilter costs mefi. The per-use cost of generating one page full of existing metafilter content (IE, retrieving data from a database and displaying it to one viewer over the internet) is quite low. Much lower than, say, printing the same page full of information and mailing it to someone.
posted by verb at 6:08 PM on May 27

"Small amounts of distributed bandwidth are extremely cheap. Large amounts of centralized bandwidth can be very expensive indeed."

But it's true that this is not based on the cost of delivering the bandwidth, but on rent-seeking by the companies which "own" the infrastructure, but which we all really paid for ourselves. Whether bandwidth is cheap or not depends on what people are willing to pay, but not what it's really worth compared to cost. The problem is that competition is very difficult on a large scale, creating virtual monopolies which set prices which have no reflection of the reality of the real costs or who actually paid for it. This is not how it was supposed to work, btw, but the US government apparently has little regulatory control anymore to hold the telecoms to their promises, and of late has become a puppet for the industry.
posted by krinklyfig at 6:14 PM on May 27

A couple of years ago co-workers and I worked with Sony BMG to build a re-usable web site framework for all of their artist web sites. They were sick of rolling from-scratch stuff every time someone released a new CD, and wanted cool features that were hard to produce on the tight deadlines of a single album's promotional schedule.

So we used an existing open source platform, made a bunch of enhancements to it, and they donate all of those enhancements back to the open source project. Then they hired the guy who was working on one of the key plugins for that project, and have spent several years funding his work on it because they rely on it.

A year later, Warner Brothers decided they wanted to do something similar. They used all of Sony BMG's donated code and added on new features they wanted. Which Sony subsequently used, too.

This arrangement made perfect sense for both parties: producing software was not their primary business -- software was simply an artifact of their own product promotion work. Leveraging existing work, and sharing the cost of creating, improving, and maintaining that particular tool, let them all do more than they would have been able to otherwise. And everyone else using that open source tool benefited, too.

That kind of arrangement doesn't make sense for everyone, but more and more businesses are starting to say that it makes sense. That's a really big change.
posted by verb at 6:15 PM on May 27 [2 favorites]

krinklyfig: But it's true that this is not based on the cost of delivering the bandwidth, but on rent-seeking by the companies which "own" the infrastructure, but which we all really paid for ourselves.

I'm quite wary of making that characterization in all cases, because while the fibers are fairly cheap, overall, the machinery DRIVING the fibers is enormously expensive and difficult to make. That's part of why distributed bandwidth is so cheap -- a reliable 1MBit connection is very cheap and very easy, doable with hardware that costs pennies. But terabit links are staggeringly expensive, even if you own all the hardware and all the cables. Even without any rent-seeking at all, running a terabit line across a downtown metro area would be extremely expensive.

But, yeah, I'd like to see more government control in this area, more focus on allowing innovation but preventing monopoly abuse. As you correctly point out, the public paid for a very large fraction of that infrastructure, and we therefore have some property rights to it.
posted by Malor at 6:27 PM on May 27

This is the subject for the Monthly Metafilter / Wired Magazine post? I would have preferred the Mark Weber Tobias article. But I can thank my lucky stars it's not the Jimmy Fallon one.
posted by jabberjaw at 6:28 PM on May 27

"But, yeah, I'd like to see more government control in this area, more focus on allowing innovation but preventing monopoly abuse. As you correctly point out, the public paid for a very large fraction of that infrastructure, and we therefore have some property rights to it."

That's really the point. We built a highway system in the US back in the '50s as a national engine for commerce with taxpayer money. Although occasionally I do hear grumblings about how we should just go to all toll roads, most people don't really object to taxpayer money being used to build and maintain this infrastructure. The current Internet infrastructure was designed and built with taxpayer money, but we have little control over it. When you talk about things like the cost of delivery, this has to be taken into account. When you think about it, we're really paying for it twice; once by our tax money, and the second time by paying to line the pockets of companies which now claim complete control over it and prevent real competition. It's as if we had a handful of companies which owned all the highways, but with the same history of us paying for it to get it in place. The initial outlays for large networks are so high that it doesn't really make sense to expect entrepreneurs to tackle the problem entirely with investor money.

It does make a lot more sense to take the infrastructure public with private competition over it, and it would be magnitudes cheaper than what it's cost us already, with a goal of complete, true high-speed coverage rather than ignoring areas with a poor ROI from a business POV or dribbling out bandwidth. That's why we mandated rural phone coverage long ago backed with tax money, because otherwise the cost would prevent it, and millions of people would be essentially cut off from the most important communication development of the century. And the competition over the network would drive the continuing development, as their needs change and evolve, rather than having a single public-private company which is not really accountable to the customers, like the old Ma Bell, and much like the few large telcos in their current form. But funding the infrastructure of the phone system and the highway system made sense, and it paid us back millions of times over. I think a pretty good investment, but we now have to approach problems like resources and energy to pay us back again down the road, because the highway system also tied deeply into an unsustainable resource which has caused a lot of problems. But the network we all use is a great communication tool, and we all benefit by making sure we all have the opportunity to benefit by it. It's just that we trusted private companies too much with this resource and we lack the ability to rein them in, so it's time to make it our own. And there is a possibility we could get there, but for now we're still pumping money into the private sector and I guess hoping for the best.
posted by krinklyfig at 6:52 PM on May 27

"The modern Internet has virtually everyone paying rent to the owners of the core infrastructure, but that rent is so low relative to the perceived benefit of the service that there's not a huge incentive to create alternatives."

Whether rent is "low" or not depends on where you live. Quite a few communities are limited to nothing but dialup, and where I live high speed connections (>1Mb/s) range from $40 or so to around $70 or higher, depending on whether you live in town or too far from the DSLAM. An hour away and further out there are some towns which have no broadband, but of course you can get satellite, although it's not really adequate. The only connections over 3Mb are right in the middle of town near Qwest's CO. And, frankly, Qwest sucks rocks, but so does Verizon and AT&T, and all the large cable companies. They all have artificial monopolies.
posted by krinklyfig at 7:12 PM on May 27 [1 favorite]

I don't dispute that digital distribution is cheaper, I only argue that it is not free. Business plans that do not acccout for this difference will fall apart at scale. The pirate bay may be able to distribute my video, but it isn't reliable. The item you seek may be there, but it can also vanish. The post production elements of the digital supply chain are not nearly as simple or cheap as has been suggested. Even google costs several dollars per user per year to run.
posted by humanfont at 7:12 PM on May 27

Business plans that do not acccout for this difference will fall apart at scale. The pirate bay may be able to distribute my video, but it isn't reliable. The item you seek may be there, but it can also vanish. The post production elements of the digital supply chain are not nearly as simple or cheap as has been suggested.

The cost of making a copy of something digitally is minuscule. It falls into the 'rounding error' category. And, yes, even running a marginally successful podcast can quickly educate someone that bandwidth isn't free. However, it is orders of magnitude cheaper than it was decades ago, when duplication was a matter of printing up physical copies of a thing, or printing LP's, or renting time on a radio station -- or many radio stations around the country.

It will always be possible to make a business model that is doomed to fail. That doesn't mean that the profound collapse of 'duplication cost' in a digital world isn't a game-changer, an enabler for new ways of approaching certain kinds of problems. It also doesn't mean that the collapse of duplication cost doesn't undermine older business models that padded the duplication cost to cover sunk infrastructure cost.
posted by verb at 7:33 PM on May 27

Disclaimer: I work for Mozilla, but I am not speaking for Mozilla in an official capacity here.

fatbird: I'm not sure that Firefox is a good example, either, since the Mozilla Foundation is heavily dependent upon funding from first AOL, then Google, and most of the key developers are paid by the Mozilla Foundation.

The Mozilla Foundation was started by money from AOL (and Mitch Kapor), but that was 10 years ago and there have been no ties to AOL since the initial funding grant.

Mozilla does have a partnership with Google, but we do so with many other search engines around the world. These partnerships do help to fund the operations that support the development of Mozilla products. Mozilla's financial documents are completely open for anyone to review, of course. That is not the case with any other browser provider.

However, it is important to note that 40% of Firefox is built by volunteers (a number that has stayed constant for over 5 years.) Without those 40% contributions from the open source community that supports Mozilla, Firefox would not be the browser that it is. Some of the developers are paid, many are not. It is an important distinction for us.

Another important metric is that Firefox 3.5 will be available in over 75 languages, of which Mozilla provides only English. All of the other localizations of Firefox are developed by volunteer localizers. I don't know about the other browsers but I'm pretty sure that Mozilla's open-source nature enables us to provide Firefox in more languages than a commercial browser provider would.

I'm happy to take other questions via MeMail should anyone have further questions.
posted by gen at 8:10 PM on May 27

krinklyfig: Large projects like Firefox require a core which has the resources to maintain the project adequately. That doesn't mean it becomes a commercial operation or that its sponsors have any real control over the project. If a project gets to this level, it requires a bureaucratic underpinning which will allow it to deal with costs, because at that point, there are costs to the project. And it works pretty well like this, although there certainly are valid objections to Mozilla's structure, things like that, but the project remains open, volunteer-based and free to use and modify.

krinklyfig is right on here.

Many prominent software products use Mozilla code (either Gecko itself or Mozilla software libraries.)

The full list of software projects using Mozilla code is quite large (and includes Google's Chrome, which uses Mozilla's NSS and NSPR libraries.)
posted by gen at 8:30 PM on May 27

Twang: "So did the phone companies, which is why they fought tooth-and-nail to stop stuff like municipally-owned WiFi."

Muni WiFi isn't mesh networking, or really anything close to it. Mesh networking is decentralized and peer-to-peer; municipal WiFi just makes the local government an ISP.

It's still centralized and users are still 'renting' access, they're just paying for it via tax dollars instead of their monthly bill. (Theoretically — and I'm a fan of muni internet so I happen to believe this, at least to a certain extent — the rent is less in the municipal case than in the private-sector case, because the good is a natural monopoly. But the rent is still there.)

If I could make a vehicle analogy, private ISPs are like early toll roads, set up along routes that their investors think will generate profits, charging whatever they think people are willing to pay. Muni WiFi and municipal Internet (like the system they have down in Wilson, NC that was the subject of a FPP last week) are like the currently-dominant system of municipal roads. They might appear "free" at first glance, but the costs are paid for via taxes on everyone, whether they use the roads or not, on the assumption that everyone benefits from their existence. Mesh networking, in the early-90s Doctorowian sense, would be like…personal jump-jets? Hitchhiking? I don't really know that there's a good analogy there. It's something where everyone pays a one-time fee for their own equipment, and that equipment both acts as an endpoint but also acts as part of the network itself, passing other messages along. There's no ongoing cost for access, it really is free. Unfortunately, I don't see the technical problems being overcome except for narrowband emergency communications, when nothing else is available at any price.

krinklyfig: "Whether rent is "low" or not depends on where you live. Quite a few communities are limited to nothing but dialup, and where I live high speed connections (>1Mb/s) range from $40 or so to around $70 or higher […]"

Sure, I believe that. I pay $50 for an allegedly 3Mb, actually 0.5Mb, connection from Comcast, who aren't exactly noted for inspiring warm fuzzies in their customers. Still, I think it's pretty cheap for what it is: access to the greatest information system that's ever been created, and one which gets bigger and more functional by the day. I thought the Internet was a good deal back when it was billed by the hour, dog slow, and all you could really do was IRC, Usenet, email, and FTP; although it irks me to pay half a C-note to Comcast every month because I despise them as a company, the product they deliver is spectacular for the price.

Of course, the value of the product — the Internet — isn't really being generated by Comcast, or any of the other ISPs. Comcast's contribution to the content on the Internet is pretty meager. I wouldn't give them two squirts for their portal page and other 'exclusive content,' so it's pretty clear that they're just using its monopoly position to extract a toll for access to things that other people have built. Classic rentierism. As a result, even though I think their value proposition isn't a bad one, we should still feel free to use the bludgeon of public policy unsparingly to keep their profits low and prices as close to the marginal cost of service delivery as possible. Not because the content isn't worth the cost they're charging, but because they shouldn't be the ones charging for and profiting from it.
posted by Kadin2048 at 12:07 AM on May 28

Malor - ...the fibers are fairly cheap, overall, the machinery DRIVING the fibers is enormously expensive and difficult to make. That's part of why distributed bandwidth is so cheap -- a reliable 1MBit connection is very cheap and very easy, doable with hardware that costs pennies. But terabit links are staggeringly expensive, even if you own all the hardware and all the cables. Even without any rent-seeking at all, running a terabit line across a downtown metro area would be extremely expensive.

I'm very network-tarded, so I ask in all seriousness, why is terabit wire so expensive compared to MB?
posted by saysthis at 8:48 AM on May 28

saysthis: "I'm very network-tarded, so I ask in all seriousness, why is terabit wire so expensive compared to MB?"

It's not the wire, it's the machinery on either end of the wire. Although there are different grades and types of fiber, and some can carry more bandwidth than others or go for greater distances without repeaters, that's not really the key consideration when you're talking about data-carrying capacity. It's all about the routers or switches that you have on either end of the link.

Right now, terabit switching equipment is up in the medium-size-car price range. (It doesn't have a single "terabit" interface, either, just multiple 10Gb ports; I assume to get a terabit you'd bond several interfaces together and then multiplex them over the physical fiber. That seems to be what they're doing here.) The equipment probably makes use of a lot of custom ASICs to push packets that quickly.

In contrast, gigabit equipment is a whole lot cheaper. You can get gigabit UTP-to-fiber media converters for multimode fiber for under a hundred bucks, IIRC. Probably more for long-distance single-mode fiber converters, but basically commodity gear these days. And the routing/switching equipment is nothing special either; most cheap gigabit routers are standard-design embedded systems with the hard work done in software.

In a few years, the price of 10Gb and terabit gear will have sunk, and it'll be possible to upgrade the endpoint equipment and get higher speeds out of existing fiber links.
posted by Kadin2048 at 11:34 AM on May 28

saysthis: I'm very network-tarded, so I ask in all seriousness, why is terabit wire so expensive compared to MB?

Kadin mostly got that, but I guess I wasn't very clear -- it's not the fiber that's expensive, it's the machinery to encode and transmit the photons down the fiber. There's tons of dark fiber laid, because it doesn't cost that much. The expense is in digging everything up and paying people to lay the cable, not the actual cable itself. It's almost the same price to run a hundred fibers than to run just one. But all that dark fiber stays dark, for the moment, because the machines to light it up are very, very expensive.

They're also having to deal with Shannon's Law, which is the measurement of the total bandwidth that's possible with a given carrier wave. It's related to the frequency of the carrier. To go past a terabit, the frequencies in the light will get so energetic that fibers will apparently become miniature death rays if they're unplugged, so that's starting to look like a hard-ish limit on individual links. We'll end up aggregating many terabit links, most likely, instead of going much faster down any individual fiber.
posted by Malor at 11:54 AM on May 28

Thanks!

It's interesting, because it kind of tells you where telcos are taking their money. I'll have some more time later this weekend to put together a fuller post.

I had no idea we were getting close to the hard limit on fiber-optic cables. Lately I've been hearing about CPU limits, data density limits, and now data speed limits. Dammit.
posted by saysthis at 9:47 PM on May 29