5 ms·

This is a super undeveloped idea, but I was thinking about something along these lines: 1. OSS users pay a monthly premium for insurance on the software (incl

by abiro 5y ago

This is a super undeveloped idea, but I was thinking about something along these lines: 1. OSS users pay a monthly premium for insurance on the software (incl transient dependencies) that is critical to their business. 2. Part of the cash flow from premiums is pooled into an insurance fund that pays interest to OSS developers. The rest is payed out directly to OSS devs. 3. If there is a bug in an insured OSS project, policy holders have a claim on the pool. The magnitude of the claim depends on the severity of the bug and the time it remains open. Future payouts to the OSS project with the bug are directed to the fund until the payed out insurance benefit is covered. The hardest part is coming up with a decentralized consensus mechanism for establishing the existence, severity and resolution of bugs, but there is a lot of experimentation going on in DeFi to solve similar problems. As to why it should be decentralized, it's the same reason OSS is decentralized: you don't want a single actor to control OSS funding. edit: wording

SkyMarshal 5y ago
> The hardest part is coming up with a decentralized consensus mechanism for establishing the existence, severity and resolution of bugs, but there is a lot of experimentation going on in DeFi to solve similar problems. Pretty much. Getting real world data like this onto a blockchain such that the representation of it on the blockchain can be trusted is as yet an unsolved problem. Blockchains are great at managing data created internally in a trustworthy way, but not data created externally.
perlgeek 5y ago
What what you want is a kind of business or foundation where: 1. users pay monthly fee/premium 2. part of that is kept in a pool, part is payed to the developers 3. when bugs happen, funky stuff happens with the pooled money. See, I didn't need any crypto for that. Maybe one day crypto can reduce the legal overhead, maybe not. The hard part is getting people to actually pay for things that they could have for free. Do you know what the reminds me of? RedHat. RedHat sold (and still sells, I guess) Open Source software together with maintenance contracts (pretty close to what you call insurance). They needs lots of marketing and sales staff to attract customers. Who replaces the marketing and sales people in the crypto world? Who does the convincing?
Scarblac 5y ago
All this would work much better with simple old fashioned money. OSS isn't decentralized at all, there's always at least some group of core devs who hold the commit rights and who make releases. And they will want to control their funding.
- abiro 5y ago
  Individual OSS projects aren't decentralized, but the process of building and distributing OSS mostly is (leaving GitHub is relatively low barrier). See iOS for a centralized ecosystem. If there was a central entity controlling OSS funding they could for example prevent forked projects from succeeding.
- leppr 5y ago
  > All this would work much better with simple old fashioned money. That's the thing, on the internet, simple old fashioned money is not like simple old fashioned money. Keeping accounts and transferring old fashioned money between peers on the internet is not trivial. Discounting the on-boarding problem, crypto is way closer to cash than payment processing networks like Visa, Mastercard, Paypal. Accepting crypto for any kind of work is trivial, you just generate a standard address/QR code that can be read by any other standard wallet, and receive a standard crypto that can be exchanged freely.
  - saalweachter 5y ago
    My read on this is "Patreon for OSS/corporate relations". Start a Patreon-like business where corporations/users can sign up to give a monthly donation-subscription-contribution to a slew of OSS developers, anywhere from $1 / mo to $$$. Reduce the friction of financially supporting every tiny project you make use of. Throw in a nice suite of uniform communication channels for bug requests and version announcements.
shafyy 5y ago
Sorry, but what are you talking about? How does this "insurance" solve a problem, with or without crypto?
- abiro 5y ago
  Business: I want to make sure that if there are bugs in my OSS dependencies, they're fixed quickly. OSS dev: I want to make a living from my OSS work without starting a business around it or seeking employment.
  - around_here 5y ago
    The OSS dev literally wants to avail themselves of public infrastructure and money, while paying nothing back in. That's all this is. All you did there was state the basic tenet of capitalist wage labour, except no taxes.
  - UncleMeat 5y ago
    OSS dev: I'm frightened of the personal risk of having bugs in my code suddenly mean that money is hoovered out of my bank account since it is extremely difficult be confident my code contains zero bugs. Insurance usually goes the other direction. Large organizations can amortize risk. Individuals need to mitigate risk. Having somebody buy insurance on my code makes me less likely to want to release it rather than more likely.
    - abiro 5y ago
      Read this: https://news.ycombinator.com/item?id=27842848 https://news.ycombinator.com/item?id=27842848 OSS devs would risk future earnings only.
      - thraxil 5y ago
        So, I've developed some software, gotten a bunch of users and developed a bit of an income stream, then there's a big bug and now I have no chance of future earnings from that software. I'd probably abandon that project and move on to something else instead; maybe get a job at $BIGCO or become a turnip farmer. So some people got an insurance payout, but the software they were using and considered important enough that they would pay for insurance on is now abandonware for them and for any other users that happened to be out there.
crispyambulance 5y ago
> thinking about something along these lines... There are some solid ideas behind cryptocurrency and perhaps your example is an illustration of how those good ideas could work. Many of the proponents of cryptocurrency describe such use-cases and it seems perfectly sound and air-tight on the surface. But the current reality is that the OVERWHELMING use-cases of crypto-currency are wild get-rich-quick shell-games, illicit trade and tax-avoidance schemes. This is showing no signs of sobering up, in fact it has become worse. People keep saying that it will "settle down" but there is NO EVIDENCE for that. The truly disturbing thing about cryptocurrency is that it reveals the possibility that "money", as a concept, is NOT necessarily a medium to exchange goods and services. It is something much more complex and fluid. Cryptocurrency, as a currency, takes all the awful and volatile hidden features of money, accelerates them and makes them overt and central to it's operation. I think many of us find that idea repulsive. If crypto (as it currently exists) becomes the money of the future, it will mean that the vast majority of the population who earn money by exchanging their labor will find themselves at the mercy of wild and random fluctuations in the value of their labor. Their labor will mean less or more not because of what they actually do, but because of completely unrelated fluctuations caused by "millionaires" and institutions acting on whims, or even joking around. You could argue that it happens now with dollar money, but that kind of fluctuation has been almost always slow. Crypto promises to make crazy fluctuations the norm. Vitalik Buterin, for example, famously donated ~1 billion "dollars" towards India COVID relief. Except that it wasn't really dollars, it was some cryptocoin that he was gifted ("shiba-shitto" or something like that). That value plummeted almost immediately. I don't know what it is now, it seems the news orgs have lost interest in the story. I do hope that India COVID relief got something out of it, but would have been better if they didn't have to hire crypto people to process that "money" into something useable when they're dealing with a pandemic.
- jazzyjackson 5y ago
  Here’s the SHIB donation, looks like it’s quoted at 350,000,000 USD now [1] and the relief fund has a transparency page linking to their wallet and bank and grant transactions [2] I agree with you btw just wanted to link to context [1] https://etherscan.io/tx/0xb65bcbb85c1633b0ab4e4886c3cd8eeaeb63edbb39cacdb9223fdcf4454fd2c7 https://etherscan.io/tx/0xb65bcbb85c1633b0ab4e4886c3cd8eeaeb... [2] https://cryptorelief.in/transparency https://cryptorelief.in/transparency
JumpCrisscross 5y ago
> you don't want a single actor to control OSS funding You’re replacing a single actor you can sue in court with a dynamic, pseudo-anonymous committee one can lobby that will make its decision based on rules or whims. That sounds like a step backwards. Separately, why would anyone pay the premium? If you want a bug fixed and are willing to pay for it, you’d get the job done faster hiring the dev on freelance. Swapping this to a simple insurance model exposes the flaws. Unfortunately, there is a lot of this going on in crypto. (Though I am of the belief that people are free to lose their money, as long as it doesn’t become a political problem.)
- abiro 5y ago
  > You’re replacing a single actor you can sue in court with a dynamic, pseudo-anonymous committee one can lobby that will make its decision based on rules or whims. That sounds like a step backwards. This is exactly the problem that decentralized oracles are trying to solve. Consensus courtesy of cryptography and game theory. > Separately, why would anyone pay the premium? If you want a bug fixed and are willing to pay for it, you’d get the job done faster hiring the dev on freelance. Nothing solves bugs faster than the original dev being financially incentivized to fix it ASAP.
  - JumpCrisscross 5y ago
    > Consensus courtesy of cryptography and game theory This solves the vote tallying problem. But that’s not the real problem. Trusting the people making the decision is. Cryptographically guaranteed kangaroo courts don’t help anyone.
    - abiro 5y ago
      I can’t downvote, so I have to write: the parent comment is ignorant of how decentralized oracles work.
      - JumpCrisscross 5y ago
        > the parent comment is ignorant of how decentralized oracles work I probably am. But I get pitched a project that involves one about every day. Multiple oracles, staking...they solve a fake problem--counting votes--while leaving the real problem--who the oracle(s) is/are and how they make their decision--untouched. If you're measuring something objective, like a stock price or weather report, fine. Whatever. But a bespoke, subjective output? Requiring specialists to be done correctly? Like a decision on whether an insurance pay-out is warranted in a specific case? We could create arbitration with extra steps, which is what ChainLink and friends are doing. But again, you've replaced a centralized actor governed by law and reputation with a centralized set governed by reputation alone. (Staking doesn't solve this problem; it just dilutes and punts it.) And if you're pulling from a predictable pool of specialists, you can bet your ass they'll be lobbied.
giaour 5y ago
Any insurance that can only pay out in proportion to the premiums already paid in seems like a red flag. That's a savings account and not a genuine transfer of risk. Unless the bug that triggers a payout is extremely circumstantial, it seems likely that any single claim would trigger a flood of claims, so I don't buy that pooling premiums received makes an insolvency-inducing run less likely.
- abiro 5y ago
  Thanks for the insightful comment! Yeah, the way I presented the idea above, it's closer to a cooperative than an insurance company. I think it's more important for the buyers of the "insurance" to get their bugs resolved quickly than it is to receive monetary compensation, so the purpose of the payouts is really just to incentivize OSS developers to resolve bugs quickly by penalizing their future earnings if they fail to do so. But if real insurance with real payouts is desirable, then that's solvable by raising capital for the fund in an ICO and directing some of the cash flow to investors. If I'm not mistaken this is how you'd start an insurance company today: you'd need to meet some capital requirements before being allowed to sell insurance to the public and you'd do that by raising money from investors.
newswasboring 5y ago
> The hardest part is coming up with a decentralized consensus mechanism for establishing the existence, severity and resolution of bugs, but there is a lot of experimentation going on in DeFi to solve similar problems. What are those DeFi projects? If someone can figure out such high level consensus mechanisms that would be fascinating.
- abiro 5y ago
  I won't link to projects to avoid the perception of a shill, but the keyword is "decentralized oracle". To be clear, I don't think this is a solved problem yet, and as always with crypto, just because a project claims that they've solved a problem, doesn't mean that they have.

by abiro 5y ago

SkyMarshal 5y ago

> The hardest part is coming up with a decentralized consensus mechanism for establishing the existence, severity and resolution of bugs, but there is a lot of experimentation going on in DeFi to solve similar problems. Pretty much. Getting real world data like this onto a blockchain such that the representation of it on the blockchain can be trusted is as yet an unsolved problem. Blockchains are great at managing data created internally in a trustworthy way, but not data created externally.

perlgeek 5y ago

What what you want is a kind of business or foundation where: 1. users pay monthly fee/premium 2. part of that is kept in a pool, part is payed to the developers 3. when bugs happen, funky stuff happens with the pooled money. See, I didn't need any crypto for that. Maybe one day crypto can reduce the legal overhead, maybe not. The hard part is getting people to actually pay for things that they could have for free. Do you know what the reminds me of? RedHat. RedHat sold (and still sells, I guess) Open Source software together with maintenance contracts (pretty close to what you call insurance). They needs lots of marketing and sales staff to attract customers. Who replaces the marketing and sales people in the crypto world? Who does the convincing?

Scarblac 5y ago

All this would work much better with simple old fashioned money. OSS isn't decentralized at all, there's always at least some group of core devs who hold the commit rights and who make releases. And they will want to control their funding.

abiro 5y ago
Individual OSS projects aren't decentralized, but the process of building and distributing OSS mostly is (leaving GitHub is relatively low barrier). See iOS for a centralized ecosystem. If there was a central entity controlling OSS funding they could for example prevent forked projects from succeeding.
leppr 5y ago
> All this would work much better with simple old fashioned money. That's the thing, on the internet, simple old fashioned money is not like simple old fashioned money. Keeping accounts and transferring old fashioned money between peers on the internet is not trivial. Discounting the on-boarding problem, crypto is way closer to cash than payment processing networks like Visa, Mastercard, Paypal. Accepting crypto for any kind of work is trivial, you just generate a standard address/QR code that can be read by any other standard wallet, and receive a standard crypto that can be exchanged freely.
- saalweachter 5y ago
  My read on this is "Patreon for OSS/corporate relations". Start a Patreon-like business where corporations/users can sign up to give a monthly donation-subscription-contribution to a slew of OSS developers, anywhere from $1 / mo to $$$. Reduce the friction of financially supporting every tiny project you make use of. Throw in a nice suite of uniform communication channels for bug requests and version announcements.

shafyy 5y ago

Sorry, but what are you talking about? How does this "insurance" solve a problem, with or without crypto?

abiro 5y ago
Business: I want to make sure that if there are bugs in my OSS dependencies, they're fixed quickly. OSS dev: I want to make a living from my OSS work without starting a business around it or seeking employment.
- around_here 5y ago
  The OSS dev literally wants to avail themselves of public infrastructure and money, while paying nothing back in. That's all this is. All you did there was state the basic tenet of capitalist wage labour, except no taxes.
- UncleMeat 5y ago
  OSS dev: I'm frightened of the personal risk of having bugs in my code suddenly mean that money is hoovered out of my bank account since it is extremely difficult be confident my code contains zero bugs. Insurance usually goes the other direction. Large organizations can amortize risk. Individuals need to mitigate risk. Having somebody buy insurance on my code makes me less likely to want to release it rather than more likely.
  - abiro 5y ago
    Read this: https://news.ycombinator.com/item?id=27842848 https://news.ycombinator.com/item?id=27842848 OSS devs would risk future earnings only.
    - thraxil 5y ago
      So, I've developed some software, gotten a bunch of users and developed a bit of an income stream, then there's a big bug and now I have no chance of future earnings from that software. I'd probably abandon that project and move on to something else instead; maybe get a job at $BIGCO or become a turnip farmer. So some people got an insurance payout, but the software they were using and considered important enough that they would pay for insurance on is now abandonware for them and for any other users that happened to be out there.

crispyambulance 5y ago

> thinking about something along these lines... There are some solid ideas behind cryptocurrency and perhaps your example is an illustration of how those good ideas could work. Many of the proponents of cryptocurrency describe such use-cases and it seems perfectly sound and air-tight on the surface. But the current reality is that the OVERWHELMING use-cases of crypto-currency are wild get-rich-quick shell-games, illicit trade and tax-avoidance schemes. This is showing no signs of sobering up, in fact it has become worse. People keep saying that it will "settle down" but there is NO EVIDENCE for that. The truly disturbing thing about cryptocurrency is that it reveals the possibility that "money", as a concept, is NOT necessarily a medium to exchange goods and services. It is something much more complex and fluid. Cryptocurrency, as a currency, takes all the awful and volatile hidden features of money, accelerates them and makes them overt and central to it's operation. I think many of us find that idea repulsive. If crypto (as it currently exists) becomes the money of the future, it will mean that the vast majority of the population who earn money by exchanging their labor will find themselves at the mercy of wild and random fluctuations in the value of their labor. Their labor will mean less or more not because of what they actually do, but because of completely unrelated fluctuations caused by "millionaires" and institutions acting on whims, or even joking around. You could argue that it happens now with dollar money, but that kind of fluctuation has been almost always slow. Crypto promises to make crazy fluctuations the norm. Vitalik Buterin, for example, famously donated ~1 billion "dollars" towards India COVID relief. Except that it wasn't really dollars, it was some cryptocoin that he was gifted ("shiba-shitto" or something like that). That value plummeted almost immediately. I don't know what it is now, it seems the news orgs have lost interest in the story. I do hope that India COVID relief got something out of it, but would have been better if they didn't have to hire crypto people to process that "money" into something useable when they're dealing with a pandemic.

jazzyjackson 5y ago
Here’s the SHIB donation, looks like it’s quoted at 350,000,000 USD now [1] and the relief fund has a transparency page linking to their wallet and bank and grant transactions [2] I agree with you btw just wanted to link to context [1] https://etherscan.io/tx/0xb65bcbb85c1633b0ab4e4886c3cd8eeaeb63edbb39cacdb9223fdcf4454fd2c7 https://etherscan.io/tx/0xb65bcbb85c1633b0ab4e4886c3cd8eeaeb... [2] https://cryptorelief.in/transparency https://cryptorelief.in/transparency

JumpCrisscross 5y ago

> you don't want a single actor to control OSS funding You’re replacing a single actor you can sue in court with a dynamic, pseudo-anonymous committee one can lobby that will make its decision based on rules or whims. That sounds like a step backwards. Separately, why would anyone pay the premium? If you want a bug fixed and are willing to pay for it, you’d get the job done faster hiring the dev on freelance. Swapping this to a simple insurance model exposes the flaws. Unfortunately, there is a lot of this going on in crypto. (Though I am of the belief that people are free to lose their money, as long as it doesn’t become a political problem.)

abiro 5y ago
> You’re replacing a single actor you can sue in court with a dynamic, pseudo-anonymous committee one can lobby that will make its decision based on rules or whims. That sounds like a step backwards. This is exactly the problem that decentralized oracles are trying to solve. Consensus courtesy of cryptography and game theory. > Separately, why would anyone pay the premium? If you want a bug fixed and are willing to pay for it, you’d get the job done faster hiring the dev on freelance. Nothing solves bugs faster than the original dev being financially incentivized to fix it ASAP.
- JumpCrisscross 5y ago
  > Consensus courtesy of cryptography and game theory This solves the vote tallying problem. But that’s not the real problem. Trusting the people making the decision is. Cryptographically guaranteed kangaroo courts don’t help anyone.
  - abiro 5y ago
    I can’t downvote, so I have to write: the parent comment is ignorant of how decentralized oracles work.
    - JumpCrisscross 5y ago
      > the parent comment is ignorant of how decentralized oracles work I probably am. But I get pitched a project that involves one about every day. Multiple oracles, staking...they solve a fake problem--counting votes--while leaving the real problem--who the oracle(s) is/are and how they make their decision--untouched. If you're measuring something objective, like a stock price or weather report, fine. Whatever. But a bespoke, subjective output? Requiring specialists to be done correctly? Like a decision on whether an insurance pay-out is warranted in a specific case? We could create arbitration with extra steps, which is what ChainLink and friends are doing. But again, you've replaced a centralized actor governed by law and reputation with a centralized set governed by reputation alone. (Staking doesn't solve this problem; it just dilutes and punts it.) And if you're pulling from a predictable pool of specialists, you can bet your ass they'll be lobbied.

giaour 5y ago

Any insurance that can only pay out in proportion to the premiums already paid in seems like a red flag. That's a savings account and not a genuine transfer of risk. Unless the bug that triggers a payout is extremely circumstantial, it seems likely that any single claim would trigger a flood of claims, so I don't buy that pooling premiums received makes an insolvency-inducing run less likely.

abiro 5y ago
Thanks for the insightful comment! Yeah, the way I presented the idea above, it's closer to a cooperative than an insurance company. I think it's more important for the buyers of the "insurance" to get their bugs resolved quickly than it is to receive monetary compensation, so the purpose of the payouts is really just to incentivize OSS developers to resolve bugs quickly by penalizing their future earnings if they fail to do so. But if real insurance with real payouts is desirable, then that's solvable by raising capital for the fund in an ICO and directing some of the cash flow to investors. If I'm not mistaken this is how you'd start an insurance company today: you'd need to meet some capital requirements before being allowed to sell insurance to the public and you'd do that by raising money from investors.

newswasboring 5y ago

> The hardest part is coming up with a decentralized consensus mechanism for establishing the existence, severity and resolution of bugs, but there is a lot of experimentation going on in DeFi to solve similar problems. What are those DeFi projects? If someone can figure out such high level consensus mechanisms that would be fascinating.

abiro 5y ago
I won't link to projects to avoid the perception of a shill, but the keyword is "decentralized oracle". To be clear, I don't think this is a solved problem yet, and as always with crypto, just because a project claims that they've solved a problem, doesn't mean that they have.