7 ms·

The Transaction Costs of Tokenizing Everything

133 points by thisisit 9y ago

woah 9y ago
This article is a collection of straw men. It extrapolates the failure of one dot-com era startup (Mojo Nation) into the extremely wide ranging thesis > As it turns out, bandwidth — and most computational resources — are simply too cheap to meter. This is patently false, as bandwidth and all other computational resources are metered on a huge scale today, by centralized cloud computing and ISP companies. It's even false for BitTorrent. The article describes BitTorrent with > After Mojo Nation’s demise, a former employee stripped the token incentives out of the protocol and created a simple tit-for-tat filesharing system. However, most serious BitTorrent users are active on trackers which require a certain seed:leech ratio, thereby metering the user's contribution. The fact that no money is involved has more to do with the fact that most of them are used to share files in violation of copyright, not because of some intrinsic law of computational resources being "too cheap to meter". It also references some hare-brained scheme cooked up by Enron shortly before their collapse to claim that bandwidth cannot be traded, and makes the false assertion that all peering and internet transit today is free.
- wpietri 9y ago
  > most serious BitTorrent users are active on trackers which require a certain seed:leech ratio, thereby metering the user's contribution. It's funny, but I happened to have a conversation with Bram Cohen about this very topic. He thought seed:leech ratios were a dumb idea, creating an artificial scarcity where none was needed. You can't save bandwidth for later. And I write this comment from my home Sonic gigabit fiber connection, for which I'm paying exactly the same price as I paid for my ADSL connection. My bandwidth is unmetered. Other home ISPs like Comcast do meter. But is that because they need to? Or just because it's a convenient way to figure out which of their customers are willing to pay more?
  - icebraining 9y ago
    He thought seed:leech ratios were a dumb idea, creating an artificial scarcity where none was needed. You can't save bandwidth for later. The idea is to get users to seed, rather than just leach and then shut off their client. Where does saving bandwidth come into the picture?
    - simias 9y ago
      I always thought it was a pretty silly system too. You end up in a situation where you want do download something, other people want to upload that thing but you don't do it because you don't want to ruin your ratio. This in turns means that these other people don't increase their ratios and end up in the same situation. You have an artificial 0-sum game for absolutely no good reason. Then you need freeleech periods or bonuses to make up for the ratio credit "deflation". I think it would make more sense to punish people who never seed (seeding if nobody is downloading is fine, you can't do anything about it) and incite people to seed torrents with a low seed count.
      - icebraining 9y ago
        I agree. In the only private tracker I use, you get rewarded for both the amount you've seeded and also for the number of torrents you're seeding (even if nobody downloads them).
        TeMPOraL 9y ago
        I guess I'm socially inept both in real life and on the Internet, because I never figured out, how do you get on a private tracker?
        simias 9y ago
        Depends on the trackers, most of the time you need an invite. Private trackers are often topical, either focusing on movies, video games, music, ebooks etc... So if you want an invite to these you just need to hang out with people who share these interests and then ask around. IRC is a good place to start in my experience. Some semi-private trackers also have periods where registration is open to all.
- thisisit 9y ago
  The author's point is on trying to create tokens out of things which don't really need one, specially the file storage ones like storj, sia, maidsafe - all of which are explicitly mentioned in the article. So I am wondering are you reading this as a critique of bitcoin and cryptocurrency or the so-called "storage" tokens? Without quantifying - "most serious", this is an actual straw man: > However, most serious BitTorrent users are active on trackers which require a certain seed:leech ratio, thereby metering the user's contribution. The point is majority of the BitTorrent works like that, it doesn't matter if there is a minority of "serious BitTorrent users" which require certain seed:leech ratio.
  - vasilipupkin 9y ago
    tokens are there because they are first and foremost a fundraising and network bootstrapping mechanism. Remove that part of it and nobody would issue any tokens.
    - vasilipupkin 9y ago
      I like how this is being downvoted. Which part of this do you disagree with ? that it's a fundraising mechanism?
- shapiro92 9y ago
  Bandwidth doesn't really matter to the average BitTorrent consumer, I believe that's his point. When did you ever wish to download a file through P2P and you were worried about bandwidth allocation? It's more of a let it run and because it's free on a resource that is already paid (internet connection) it doesn't matter. I am a but naive with crypto coins for every use. Maybe I am stuck up on my love for P2P but cannot understand why we need the complexity and the ever increasing cost of blockchain to allocate disk resources. I believe the combination of ICO craze with the hype of crypto as a gambling for this generation created a golden hammer.
  - TeMPOraL 9y ago
    I think the only real problem these storage tokens try to solve is incentivizing people to seed random stuff without active selection made by the seeders. Whether or not we need a cryptotoken for that is debatable, but I don't know of any other alternative that would keep the system entirely decentralized.
    - icebraining 9y ago
      It's also a matter of availability. Lots of people only seed when they're downloading something else. I've had torrents in the single-digit MBs take over a month to download, simply because the seeders were never online.
- evgen 9y ago
  [Putting the caveat up front here: my perspective may be a bit less abstract and neutral than a true third-party observer since I designed Mojo Nation and was the guy who hired Greg, Bryce, and Bram to help work on the problem.] While there may be problems with the original article I think that you are far wider of the mark than the author. Several times in your comment you have taken the behavior and practices of large-scale participants and tried to extrapolate these down to the small-scale players who make up the bulk of the users in these various systems. This is the sort of mistaken assumption that just earned someone a Nobel prize in economics ( * ) and it is probably one of the problems we had when building Mojo Nation. There are some things that are metered in bulk but which make no sense to do so at the smaller scale at which the consumer encounters them, and the nature and scale of these bulk goods can change quite quickly. When we started work on Mojo Nation many were still on dial-up and even those with broadband connections had large asymmetry between the upstream and downstream bandwidth; trying to take advantage of this asymmetry was the genesis of our peer-swarming idea and what we had hoped would distinguish us from systems like Napster that had problems with large media files (aka pirated movies) when you were no longer in a college LAN environment. When you are considering something like bandwidth or local storage capacity from the perspective of cloud computing providers or ISPs then it is most certainly a resource that has a cost and which is closely metered, but for individual users their point in time bandwidth is effectively unmetered. No one is checking their current monthly usage statement before firing up Netflix, nor are they worried too much about the storage costs of holding on to a download for another few days so that they can improve their seed:leech ratio. For most Americans water is effectively too cheap to meter. Electricity and fuel for heating are close to this level. Yes, before you point it out I am talking about the fat middle of the usage graph and not the extremes of either usage at one end or poverty at the other; for the vast majority there is very little thought given to cost when deciding to turn on a light or to take a shower. At various points in time people have had to pay close attention to their bandwidth usage or local storage capacity, but currently most people can ignore usage for daily tasks even when considering mobile bandwidth. Clay Shirky was quite correct in pointing out that the transaction cost, both in clearing the transaction within the system as well as the mental cost to the user in just needing to think about the cost, was going to be a problem for us. There were a whole host of other issues related to general complexity of the system, too many knobs for users to tweak and/or use to game the system, and our desire to build a persistent storage system rather than an ephemeral file sharing tool, but this transaction cost was a non-trivial problem that we put off for too long. In the end the closest we came to trying to eliminate the mental burden and lower systemic clearing costs was to use Paris Metro pricing (e.g. you paid to move to the front of the queue, but if you were willing to wait for your bits the cost was 0.) This model has obvious problems when dealing with persistent storage that we were never able to work around, but it is easy to see the leap from this system to the tit-for-tat that Bram eventually landed on for handling ephemeral file sharing. While it is possible to look at seed:leech ratios on major trackers as metering, this is a generous definition of the role of these ratios and they are much closer the the reputation-backed personal promises/tokens that Mojo Nation started with than to any actual currency. The goal of such ratio systems is incentive engineering, not strict accounting. And to end this on a minor diversion that ties directly to your offhand dismissal of Enron in all this, I will point out that at one point Enron’s bandwidth trading group had enough interest in Mojo Nation to fly me out to Houston several times for meetings. A plush corporate palace with a pimped out Harley-Davidson in the upper-floor lobby of a group supposedly trading bandwidth should have been a sign of their upcoming fall, but at the time I was young and easily impressed… ( * ) Yes, I know that the economics prize is not, strictly speaking, a Nobel prize...
- jcfrei 9y ago
  I think the author summarized his hypothesis quite well in the title: The Transaction Costs of Tokenizing Everything [... are too high compared to the costs of bandwidth, etc.]. If you introduce a bidding market for every computational resource then that itself creates new costs (keeping all the systems running that create bids, check if they have been filled, cancelling old ones, drafting bidding protocols, etc.) that might largely offset the gains in more efficient resource allocations.
  - 0xcde4c3db 9y ago
    I think the underlying problem is that being in charge of the transaction system confers so much power. The promise of that power creates a huge incentive to build such systems (and actively undermine any technical or political threat to their influence) even if it's worse for everyone else.
  - floatrock 9y ago
    And it might not. That's like going up to Eisenhower after he proposed the interstate system and saying, "Wait, the cost of paving over that much land and building that many overpasses might largely offset any gains of this supposed more efficient transportation network." All new technologies introduce costs, but ideally they also act as multipliers that bring far more benefits. The multiplier effect might not be enough, but then again, it might be. You're not really saying anything. Better to talk about under what circumstances the benefits-multiplier might not be enough to make up for the costs. For example, someone elsewhere in the thread said that yes, if humans were bidding on such tokenization, it might create too many barriers. But look at all the ad networks we've built where billions of tiny auctions happen instantaneously -- that's all by machines. So if the system is built for machines, the multiplier effect could be potentially huge.
macawfish 9y ago
This is super real. I don't want added abstractions of value if they're not necessary. To me, the brightest possible future of money is generosity combined with raw dimensional analysis, as in people having education about how to count physical resources and a willingness to work together and share in creating more abundant lives for themselves and others. To the extent that tokens help do that, cool. But if they just become another tax/rent pyramid then to put it frankly, fuck that i don't want it.
- deckar01 9y ago
  Markets that decouple price from consumption are unsustainable. Consumers who consume less than the average have an encentive not to participate.
AgentME 9y ago
The article keeps mentioning torrents as a free panacea that disproves the need for decentralized monetized hosting systems, but for some reason no one else ever helped seed the torrent of my encrypted backups when I posted it onto piratebay.
- admax88q 9y ago
  Google, Amazon or Tarsnap will all host it for you for a small fee. What value is there in decentralized hosting across lots of commodity nodes of your private encrypted data? As opposed to ”centralized" hosting with a few different providers.
  - mbrock 9y ago
    The decentralized network is a marketplace where different providers can participate with unified payment and APIs, so it's a way to abstract over the hosts, increasing efficiency and resilience. Whether you will end up with a small number of megahosters or a more p2p like topology seems like a question for the market to decide.
  - gwbas1c 9y ago
    Might be useful as a P2P backup? For example, If I want to backup 1GB, I have to donate 3GB to the network. Considering the hardware cost, it's probably obvious why everyone just pays a 3rd party.
    - adrianN 9y ago
      I'm pretty sure that plugging a big USB stick into your router is cheaper than paying some cloud service to host your backups. The problem is that it's a lot more complicated to have distributed backups than just plugging storage into an internet-connected device in your home.
  - AgentME 9y ago
    Why should those companies get to have a monopoly on file hosting? I'd prefer to have a single system where I can put my files out there with a bit of money and have them kept by anyone who is able to host them. A system like that could offer more competitive prices, and by using such a system, I'd be encouraging a world where I or people like me could easily get paid for hosting when we find ourselves with extra capacity.
mbrock 9y ago
The good torrent sites do have token systems, though, and manage their economies in pretty interesting ways, e.g. periods of "freeleech" to stimulate engagement; stipends for new users; bounties for desirable rare content; etc. And this goes back to warez FTP sites. Maybe private trackers would be a good use case for blockchain storage tokens. Certainly the hypothetical seeding ratio I could have on a certain private tracker for art film would be actually valuable to me.
rahimnathwani 9y ago
s/Vince Cerf/Vint Cerf
gwbas1c 9y ago
I thought this was going to be an article about the resource overhead of bad serializers in network applications! Disappointed!
elevensies 9y ago
I just realized, I'm severely confused about what Filecoin is, and possibly ICOs in general. I assumed "ICO" was analogous to IPO. If a token is essentially a bearer share that is coupled to the payment system. No reason you couldn't create the coin, run it privately until you get some revenue, and then offer it, as IPOs are usually done. But when I search Filecoin's website for "revenue" and "profit", I only see references to the participants in the system, and not the token holders. https://www.google.ca/search?q=site%3Afilecoin.io+revenue https://www.google.ca/search?q=site%3Afilecoin.io+revenue https://www.google.ca/search?q=site%3Afilecoin.io+profit https://www.google.ca/search?q=site%3Afilecoin.io+profit I did a quick search of the top 5 tokens on https://coinmarketcap.com/tokens/ https://coinmarketcap.com/tokens/ and I'm not immediately finding any clear statements on how profit is made and distributed. A question for people who are better informed, is there any kind of list or tracking on tokens payouts? Something like this (https://arxiv.org/abs/1703.03779# https://arxiv.org/abs/1703.03779# ), but for non-ponzis?
- computerphage 9y ago
  Tokens do not have to be bearer shares. Filecoin isn't.
- lojack 9y ago
  ICO can mean any number of things, but typically refer to a token that is tied to something that isn't its own currency. I'm assuming the Filecoin is redeemable for data storage on their network, but I'm not super familiar with that technology specifically. To give some other examples... Tether is a token tied to fiat currency NEO Gas is used for compute time in their smart contracts. NEO also has a currency that is used to mine gas. It can be a little confusing, but one of them is a token and one is a coin. This is different than Ethereum, where the currency is also the token used to pay for compute time. I can dig deeper and give examples of how some of the other tokens work, but unfortunately I'm not super familiar with the other tokens in the top 5. And, yes, ICOs can be tied to some number of shares in a company. In these cases, the token is redeemable for a share of the company.
  - elevensies 9y ago
rfugger 9y ago
Yes, these are terrible interfaces for humans, but they are excellent interfaces for machines. Well, to be fair, they are tricky interfaces to program to, but the potential in terms of fast and efficient resource allocation in a world where more and more decisions are automated is huge...

133 points by thisisit 9y ago

woah 9y ago

This article is a collection of straw men. It extrapolates the failure of one dot-com era startup (Mojo Nation) into the extremely wide ranging thesis > As it turns out, bandwidth — and most computational resources — are simply too cheap to meter. This is patently false, as bandwidth and all other computational resources are metered on a huge scale today, by centralized cloud computing and ISP companies. It's even false for BitTorrent. The article describes BitTorrent with > After Mojo Nation’s demise, a former employee stripped the token incentives out of the protocol and created a simple tit-for-tat filesharing system. However, most serious BitTorrent users are active on trackers which require a certain seed:leech ratio, thereby metering the user's contribution. The fact that no money is involved has more to do with the fact that most of them are used to share files in violation of copyright, not because of some intrinsic law of computational resources being "too cheap to meter". It also references some hare-brained scheme cooked up by Enron shortly before their collapse to claim that bandwidth cannot be traded, and makes the false assertion that all peering and internet transit today is free.

wpietri 9y ago
> most serious BitTorrent users are active on trackers which require a certain seed:leech ratio, thereby metering the user's contribution. It's funny, but I happened to have a conversation with Bram Cohen about this very topic. He thought seed:leech ratios were a dumb idea, creating an artificial scarcity where none was needed. You can't save bandwidth for later. And I write this comment from my home Sonic gigabit fiber connection, for which I'm paying exactly the same price as I paid for my ADSL connection. My bandwidth is unmetered. Other home ISPs like Comcast do meter. But is that because they need to? Or just because it's a convenient way to figure out which of their customers are willing to pay more?
- icebraining 9y ago
  He thought seed:leech ratios were a dumb idea, creating an artificial scarcity where none was needed. You can't save bandwidth for later. The idea is to get users to seed, rather than just leach and then shut off their client. Where does saving bandwidth come into the picture?
  - simias 9y ago
    I always thought it was a pretty silly system too. You end up in a situation where you want do download something, other people want to upload that thing but you don't do it because you don't want to ruin your ratio. This in turns means that these other people don't increase their ratios and end up in the same situation. You have an artificial 0-sum game for absolutely no good reason. Then you need freeleech periods or bonuses to make up for the ratio credit "deflation". I think it would make more sense to punish people who never seed (seeding if nobody is downloading is fine, you can't do anything about it) and incite people to seed torrents with a low seed count.
    - icebraining 9y ago
      I agree. In the only private tracker I use, you get rewarded for both the amount you've seeded and also for the number of torrents you're seeding (even if nobody downloads them).
      - TeMPOraL 9y ago
        I guess I'm socially inept both in real life and on the Internet, because I never figured out, how do you get on a private tracker?
        simias 9y ago
        Depends on the trackers, most of the time you need an invite. Private trackers are often topical, either focusing on movies, video games, music, ebooks etc... So if you want an invite to these you just need to hang out with people who share these interests and then ask around. IRC is a good place to start in my experience. Some semi-private trackers also have periods where registration is open to all.
thisisit 9y ago
The author's point is on trying to create tokens out of things which don't really need one, specially the file storage ones like storj, sia, maidsafe - all of which are explicitly mentioned in the article. So I am wondering are you reading this as a critique of bitcoin and cryptocurrency or the so-called "storage" tokens? Without quantifying - "most serious", this is an actual straw man: > However, most serious BitTorrent users are active on trackers which require a certain seed:leech ratio, thereby metering the user's contribution. The point is majority of the BitTorrent works like that, it doesn't matter if there is a minority of "serious BitTorrent users" which require certain seed:leech ratio.
- vasilipupkin 9y ago
  tokens are there because they are first and foremost a fundraising and network bootstrapping mechanism. Remove that part of it and nobody would issue any tokens.
  - vasilipupkin 9y ago
    I like how this is being downvoted. Which part of this do you disagree with ? that it's a fundraising mechanism?
shapiro92 9y ago
Bandwidth doesn't really matter to the average BitTorrent consumer, I believe that's his point. When did you ever wish to download a file through P2P and you were worried about bandwidth allocation? It's more of a let it run and because it's free on a resource that is already paid (internet connection) it doesn't matter. I am a but naive with crypto coins for every use. Maybe I am stuck up on my love for P2P but cannot understand why we need the complexity and the ever increasing cost of blockchain to allocate disk resources. I believe the combination of ICO craze with the hype of crypto as a gambling for this generation created a golden hammer.
- TeMPOraL 9y ago
  I think the only real problem these storage tokens try to solve is incentivizing people to seed random stuff without active selection made by the seeders. Whether or not we need a cryptotoken for that is debatable, but I don't know of any other alternative that would keep the system entirely decentralized.
  - icebraining 9y ago
    It's also a matter of availability. Lots of people only seed when they're downloading something else. I've had torrents in the single-digit MBs take over a month to download, simply because the seeders were never online.
evgen 9y ago
[Putting the caveat up front here: my perspective may be a bit less abstract and neutral than a true third-party observer since I designed Mojo Nation and was the guy who hired Greg, Bryce, and Bram to help work on the problem.] While there may be problems with the original article I think that you are far wider of the mark than the author. Several times in your comment you have taken the behavior and practices of large-scale participants and tried to extrapolate these down to the small-scale players who make up the bulk of the users in these various systems. This is the sort of mistaken assumption that just earned someone a Nobel prize in economics ( * ) and it is probably one of the problems we had when building Mojo Nation. There are some things that are metered in bulk but which make no sense to do so at the smaller scale at which the consumer encounters them, and the nature and scale of these bulk goods can change quite quickly. When we started work on Mojo Nation many were still on dial-up and even those with broadband connections had large asymmetry between the upstream and downstream bandwidth; trying to take advantage of this asymmetry was the genesis of our peer-swarming idea and what we had hoped would distinguish us from systems like Napster that had problems with large media files (aka pirated movies) when you were no longer in a college LAN environment. When you are considering something like bandwidth or local storage capacity from the perspective of cloud computing providers or ISPs then it is most certainly a resource that has a cost and which is closely metered, but for individual users their point in time bandwidth is effectively unmetered. No one is checking their current monthly usage statement before firing up Netflix, nor are they worried too much about the storage costs of holding on to a download for another few days so that they can improve their seed:leech ratio. For most Americans water is effectively too cheap to meter. Electricity and fuel for heating are close to this level. Yes, before you point it out I am talking about the fat middle of the usage graph and not the extremes of either usage at one end or poverty at the other; for the vast majority there is very little thought given to cost when deciding to turn on a light or to take a shower. At various points in time people have had to pay close attention to their bandwidth usage or local storage capacity, but currently most people can ignore usage for daily tasks even when considering mobile bandwidth. Clay Shirky was quite correct in pointing out that the transaction cost, both in clearing the transaction within the system as well as the mental cost to the user in just needing to think about the cost, was going to be a problem for us. There were a whole host of other issues related to general complexity of the system, too many knobs for users to tweak and/or use to game the system, and our desire to build a persistent storage system rather than an ephemeral file sharing tool, but this transaction cost was a non-trivial problem that we put off for too long. In the end the closest we came to trying to eliminate the mental burden and lower systemic clearing costs was to use Paris Metro pricing (e.g. you paid to move to the front of the queue, but if you were willing to wait for your bits the cost was 0.) This model has obvious problems when dealing with persistent storage that we were never able to work around, but it is easy to see the leap from this system to the tit-for-tat that Bram eventually landed on for handling ephemeral file sharing. While it is possible to look at seed:leech ratios on major trackers as metering, this is a generous definition of the role of these ratios and they are much closer the the reputation-backed personal promises/tokens that Mojo Nation started with than to any actual currency. The goal of such ratio systems is incentive engineering, not strict accounting. And to end this on a minor diversion that ties directly to your offhand dismissal of Enron in all this, I will point out that at one point Enron’s bandwidth trading group had enough interest in Mojo Nation to fly me out to Houston several times for meetings. A plush corporate palace with a pimped out Harley-Davidson in the upper-floor lobby of a group supposedly trading bandwidth should have been a sign of their upcoming fall, but at the time I was young and easily impressed… ( * ) Yes, I know that the economics prize is not, strictly speaking, a Nobel prize...
jcfrei 9y ago
I think the author summarized his hypothesis quite well in the title: The Transaction Costs of Tokenizing Everything [... are too high compared to the costs of bandwidth, etc.]. If you introduce a bidding market for every computational resource then that itself creates new costs (keeping all the systems running that create bids, check if they have been filled, cancelling old ones, drafting bidding protocols, etc.) that might largely offset the gains in more efficient resource allocations.
- 0xcde4c3db 9y ago
  I think the underlying problem is that being in charge of the transaction system confers so much power. The promise of that power creates a huge incentive to build such systems (and actively undermine any technical or political threat to their influence) even if it's worse for everyone else.
- floatrock 9y ago
  And it might not. That's like going up to Eisenhower after he proposed the interstate system and saying, "Wait, the cost of paving over that much land and building that many overpasses might largely offset any gains of this supposed more efficient transportation network." All new technologies introduce costs, but ideally they also act as multipliers that bring far more benefits. The multiplier effect might not be enough, but then again, it might be. You're not really saying anything. Better to talk about under what circumstances the benefits-multiplier might not be enough to make up for the costs. For example, someone elsewhere in the thread said that yes, if humans were bidding on such tokenization, it might create too many barriers. But look at all the ad networks we've built where billions of tiny auctions happen instantaneously -- that's all by machines. So if the system is built for machines, the multiplier effect could be potentially huge.

macawfish 9y ago

This is super real. I don't want added abstractions of value if they're not necessary. To me, the brightest possible future of money is generosity combined with raw dimensional analysis, as in people having education about how to count physical resources and a willingness to work together and share in creating more abundant lives for themselves and others. To the extent that tokens help do that, cool. But if they just become another tax/rent pyramid then to put it frankly, fuck that i don't want it.

deckar01 9y ago
Markets that decouple price from consumption are unsustainable. Consumers who consume less than the average have an encentive not to participate.

AgentME 9y ago

The article keeps mentioning torrents as a free panacea that disproves the need for decentralized monetized hosting systems, but for some reason no one else ever helped seed the torrent of my encrypted backups when I posted it onto piratebay.

admax88q 9y ago
Google, Amazon or Tarsnap will all host it for you for a small fee. What value is there in decentralized hosting across lots of commodity nodes of your private encrypted data? As opposed to ”centralized" hosting with a few different providers.
- mbrock 9y ago
  The decentralized network is a marketplace where different providers can participate with unified payment and APIs, so it's a way to abstract over the hosts, increasing efficiency and resilience. Whether you will end up with a small number of megahosters or a more p2p like topology seems like a question for the market to decide.
- gwbas1c 9y ago
  Might be useful as a P2P backup? For example, If I want to backup 1GB, I have to donate 3GB to the network. Considering the hardware cost, it's probably obvious why everyone just pays a 3rd party.
  - adrianN 9y ago
    I'm pretty sure that plugging a big USB stick into your router is cheaper than paying some cloud service to host your backups. The problem is that it's a lot more complicated to have distributed backups than just plugging storage into an internet-connected device in your home.
- AgentME 9y ago
  Why should those companies get to have a monopoly on file hosting? I'd prefer to have a single system where I can put my files out there with a bit of money and have them kept by anyone who is able to host them. A system like that could offer more competitive prices, and by using such a system, I'd be encouraging a world where I or people like me could easily get paid for hosting when we find ourselves with extra capacity.

mbrock 9y ago

The good torrent sites do have token systems, though, and manage their economies in pretty interesting ways, e.g. periods of "freeleech" to stimulate engagement; stipends for new users; bounties for desirable rare content; etc. And this goes back to warez FTP sites. Maybe private trackers would be a good use case for blockchain storage tokens. Certainly the hypothetical seeding ratio I could have on a certain private tracker for art film would be actually valuable to me.

rahimnathwani 9y ago

s/Vince Cerf/Vint Cerf

gwbas1c 9y ago

I thought this was going to be an article about the resource overhead of bad serializers in network applications! Disappointed!

elevensies 9y ago

I just realized, I'm severely confused about what Filecoin is, and possibly ICOs in general. I assumed "ICO" was analogous to IPO. If a token is essentially a bearer share that is coupled to the payment system. No reason you couldn't create the coin, run it privately until you get some revenue, and then offer it, as IPOs are usually done. But when I search Filecoin's website for "revenue" and "profit", I only see references to the participants in the system, and not the token holders. https://www.google.ca/search?q=site%3Afilecoin.io+revenue https://www.google.ca/search?q=site%3Afilecoin.io+revenue https://www.google.ca/search?q=site%3Afilecoin.io+profit https://www.google.ca/search?q=site%3Afilecoin.io+profit I did a quick search of the top 5 tokens on https://coinmarketcap.com/tokens/ https://coinmarketcap.com/tokens/ and I'm not immediately finding any clear statements on how profit is made and distributed. A question for people who are better informed, is there any kind of list or tracking on tokens payouts? Something like this (https://arxiv.org/abs/1703.03779# https://arxiv.org/abs/1703.03779# ), but for non-ponzis?

computerphage 9y ago
Tokens do not have to be bearer shares. Filecoin isn't.
lojack 9y ago
ICO can mean any number of things, but typically refer to a token that is tied to something that isn't its own currency. I'm assuming the Filecoin is redeemable for data storage on their network, but I'm not super familiar with that technology specifically. To give some other examples... Tether is a token tied to fiat currency NEO Gas is used for compute time in their smart contracts. NEO also has a currency that is used to mine gas. It can be a little confusing, but one of them is a token and one is a coin. This is different than Ethereum, where the currency is also the token used to pay for compute time. I can dig deeper and give examples of how some of the other tokens work, but unfortunately I'm not super familiar with the other tokens in the top 5. And, yes, ICOs can be tied to some number of shares in a company. In these cases, the token is redeemable for a share of the company.
- elevensies 9y ago

rfugger 9y ago

Yes, these are terrible interfaces for humans, but they are excellent interfaces for machines. Well, to be fair, they are tricky interfaces to program to, but the potential in terms of fast and efficient resource allocation in a world where more and more decisions are automated is huge...