4 ms·

This looks interesting, logistics and supply chain seems to be one of the few recurring use cases I see that looks somewhat promising.

dmitriid 5y ago
Doesn't require blockchain and doesn't solve the actual problem these companies have: that the records are correct as entered.
- 3np 5y ago
  Have you worked in these industries? Irrevocable timestamps and ordering for cryptographic signatures provide a level of audibility and traceability that wasn't there before. History can be rewritten opaquely in non-blockchain ledger systems. If all stakeholders can fully trust everyone involved (and their security), your point may hold, but that is not always the case.
  - dmitriid 5y ago
    > History can be rewritten opaquely in non-blockchain ledger systems. Ah yes. As if re-writing history was the primary major concern. Let's say I'm a malicious actor. I say I shipped 1 ton of gold bricks, and entered that on the blockchain that "provide a level of audibility and traceability that wasn't there before". Only one single brick reaches the final destination. No history was overwritten. So, how does blockchain help?
    - qayxc 5y ago
      > So, how does blockchain help? Easy: the malicious actor has zero chance of arguing that port authority XY messed up the paperwork or that the documentation said it was 1 brick all along. Sure, there is the oracle problem, but that's not what the blockchain (not a public BC by the way, it uses a fixed number of trusted nodes) tries to solve. The issue is often patchy documentation and lack of traceability. BC is only one tool of many here and the major improvement is digitalisation of the entire supply chain. I worked on a project in wholesale logistics several years ago and you wouldn't believe how many parts of the delivery system are still handled using literal paperwork. Coffee stains and manual "corrections" are daily occurrences and because of that, resolving irregularities way too often comes down to his/her words against theirs. A fully digital supply chain with every step and update permanently recorded in a cryptographically secured manner, helps immensely with these kinds of issues. BC is just one part of it and doesn't aim to solve everything, but it's a huge improvement. The distributed nature of the ledger provides an incentive for all participants to take part in it, because it reduces trust issues that might arise if all trust is to be put into a single party. I'm not knowledgeable enough to tell whether a more conventional distributed database system would've sufficed, and what the pros and cons of either approach are in detail, but BC is a viable solution.
      - Jasper_ 5y ago
        The malicious actor insists they shipped 1 ton of bricks, and that the port authority is lying or stole the remaining ones. The blockchain says they shipped 1 ton of bricks, after all.
        kazoomonger 5y ago
        I'm actually kind of bullish on blockchain for these sorts of scenarios, because yes, the malicious actor can claim that someone else in the chain screwed up, but all parties involved can verify all claims that were made in the entire chain. In other words, the malicious actor can't destroy or modify the paperwork and then claim something different.
        dmitriid 5y ago
        Malicious actors don't have to "destroy or modify paperwork". Paperwork is very rarely "destroyed or modified". It is usually already kept/entered as modified (and there's potentially an entirely separate and isolated paperwork trail for the actor's own accounting purposes, but that one by definition won't be on a blockchain).
        kazoomonger 5y ago
        > Malicious actors don't have to "destroy or modify paperwork" Sure, they don't have to, but it eliminates one possible source of malicious behavior. I'm not saying that using a blockchain will eliminate all malicious behavior, so I'm not sure what you're getting at.
        dmitriid 5y ago
        > but it eliminates one possible source of malicious behavior. Why is it that every single discussion about blockchain starts with "it opens up unprecedented opportunities never seen before and will disrupt eveyrthing" and ends with "well, it will maybe perhaps help in some minor way"? In this discussion: How it started: "irrevocable timestamps and ordering for cryptographic signatures provide a level of audibility and traceability that wasn't there before." How it's going: "it eliminates one possible source of malicious behavior" Color me not impressed.
    - 2cb 5y ago
      > So, how does blockchain help? Because the data is entered into the blockchain by automated systems and can be verified with cryptographic keys. You seem to have invented a strawman system in which a blockchain replaces a spreadsheet and a human can type whatever they like into it. This is not the suggested use case for blockchains at all, not in enterprise and not anywhere else. The allure is blockchains is in automated smart contracts that don't rely on human honesty and aren't as prone to human error. (I say as prone because, obviously, as with any code, humans still have to write the smart contracts. Security audits are extremely important here especially for enterprise critical applications. But the end result is a publicly accountable ledger and that includes the smart contracts themselves too. So your hypothetical would be impossible unless you hired a black hat. Even in the worst case scenario of a malicious smart contract, however, you'd have permanent public logs of exactly what was done, how, and when, and when its code written by an employee of yours, that makes it quite easy to work out who.)
      - dmitriid 5y ago
        > This is not the suggested use case for blockchains at all, not in enterprise and not anywhere else. The allure is blockchains is in automated smart contracts that don't rely on human honesty and aren't as prone to human error. This is a bunch of words that have no relation to reality. What is the connection between a smart contract and an actual physical good that is being transported? How does a "smart contract" verify any of the following: - that the goods are what they are (eco bananas, not green slime) - that the goods are in the quantity described (1 ton of gold bricks, not 1 kilo) - that the actual provenance of the goods as stated (that country of origin labelling required in the EU is accurate and reflects the actual country of origin) - that the goods have actually been transported as required (refrigerated, heated, on time, within a time range etc.) At every single point in this process you will either have people entering data manually or dozens, if not hundreds, of different systems scanning and entering data.
        qayxc 5y ago
        Smart contracts are perhaps the most oversold part of the whole block chain concept. What works for digital products (service contracts, financial transactions, data delivery) cannot always be mapped to the physical world. There are some interesting concepts that deal with trust and verifiability of data entry, scanning, and sensors, though. They just don't have anything to do with smart contracts or block chain per se, though they borrow some high-level ideas from it. But this an area where block chain in general has no place and other mechanisms (simple rule-based plausibility checking, automated anomaly-detection, etc.) are much more helpful. The physical world requires different approaches.

dmitriid 5y ago

Doesn't require blockchain and doesn't solve the actual problem these companies have: that the records are correct as entered.

3np 5y ago
Have you worked in these industries? Irrevocable timestamps and ordering for cryptographic signatures provide a level of audibility and traceability that wasn't there before. History can be rewritten opaquely in non-blockchain ledger systems. If all stakeholders can fully trust everyone involved (and their security), your point may hold, but that is not always the case.
- dmitriid 5y ago
  > History can be rewritten opaquely in non-blockchain ledger systems. Ah yes. As if re-writing history was the primary major concern. Let's say I'm a malicious actor. I say I shipped 1 ton of gold bricks, and entered that on the blockchain that "provide a level of audibility and traceability that wasn't there before". Only one single brick reaches the final destination. No history was overwritten. So, how does blockchain help?
  - qayxc 5y ago
    > So, how does blockchain help? Easy: the malicious actor has zero chance of arguing that port authority XY messed up the paperwork or that the documentation said it was 1 brick all along. Sure, there is the oracle problem, but that's not what the blockchain (not a public BC by the way, it uses a fixed number of trusted nodes) tries to solve. The issue is often patchy documentation and lack of traceability. BC is only one tool of many here and the major improvement is digitalisation of the entire supply chain. I worked on a project in wholesale logistics several years ago and you wouldn't believe how many parts of the delivery system are still handled using literal paperwork. Coffee stains and manual "corrections" are daily occurrences and because of that, resolving irregularities way too often comes down to his/her words against theirs. A fully digital supply chain with every step and update permanently recorded in a cryptographically secured manner, helps immensely with these kinds of issues. BC is just one part of it and doesn't aim to solve everything, but it's a huge improvement. The distributed nature of the ledger provides an incentive for all participants to take part in it, because it reduces trust issues that might arise if all trust is to be put into a single party. I'm not knowledgeable enough to tell whether a more conventional distributed database system would've sufficed, and what the pros and cons of either approach are in detail, but BC is a viable solution.
    - Jasper_ 5y ago
      The malicious actor insists they shipped 1 ton of bricks, and that the port authority is lying or stole the remaining ones. The blockchain says they shipped 1 ton of bricks, after all.
      - kazoomonger 5y ago
        I'm actually kind of bullish on blockchain for these sorts of scenarios, because yes, the malicious actor can claim that someone else in the chain screwed up, but all parties involved can verify all claims that were made in the entire chain. In other words, the malicious actor can't destroy or modify the paperwork and then claim something different.
        dmitriid 5y ago
        Malicious actors don't have to "destroy or modify paperwork". Paperwork is very rarely "destroyed or modified". It is usually already kept/entered as modified (and there's potentially an entirely separate and isolated paperwork trail for the actor's own accounting purposes, but that one by definition won't be on a blockchain).
        kazoomonger 5y ago
        > Malicious actors don't have to "destroy or modify paperwork" Sure, they don't have to, but it eliminates one possible source of malicious behavior. I'm not saying that using a blockchain will eliminate all malicious behavior, so I'm not sure what you're getting at.
        dmitriid 5y ago
        > but it eliminates one possible source of malicious behavior. Why is it that every single discussion about blockchain starts with "it opens up unprecedented opportunities never seen before and will disrupt eveyrthing" and ends with "well, it will maybe perhaps help in some minor way"? In this discussion: How it started: "irrevocable timestamps and ordering for cryptographic signatures provide a level of audibility and traceability that wasn't there before." How it's going: "it eliminates one possible source of malicious behavior" Color me not impressed.
  - 2cb 5y ago
    > So, how does blockchain help? Because the data is entered into the blockchain by automated systems and can be verified with cryptographic keys. You seem to have invented a strawman system in which a blockchain replaces a spreadsheet and a human can type whatever they like into it. This is not the suggested use case for blockchains at all, not in enterprise and not anywhere else. The allure is blockchains is in automated smart contracts that don't rely on human honesty and aren't as prone to human error. (I say as prone because, obviously, as with any code, humans still have to write the smart contracts. Security audits are extremely important here especially for enterprise critical applications. But the end result is a publicly accountable ledger and that includes the smart contracts themselves too. So your hypothetical would be impossible unless you hired a black hat. Even in the worst case scenario of a malicious smart contract, however, you'd have permanent public logs of exactly what was done, how, and when, and when its code written by an employee of yours, that makes it quite easy to work out who.)
    - dmitriid 5y ago
      > This is not the suggested use case for blockchains at all, not in enterprise and not anywhere else. The allure is blockchains is in automated smart contracts that don't rely on human honesty and aren't as prone to human error. This is a bunch of words that have no relation to reality. What is the connection between a smart contract and an actual physical good that is being transported? How does a "smart contract" verify any of the following: - that the goods are what they are (eco bananas, not green slime) - that the goods are in the quantity described (1 ton of gold bricks, not 1 kilo) - that the actual provenance of the goods as stated (that country of origin labelling required in the EU is accurate and reflects the actual country of origin) - that the goods have actually been transported as required (refrigerated, heated, on time, within a time range etc.) At every single point in this process you will either have people entering data manually or dozens, if not hundreds, of different systems scanning and entering data.
      - qayxc 5y ago
        Smart contracts are perhaps the most oversold part of the whole block chain concept. What works for digital products (service contracts, financial transactions, data delivery) cannot always be mapped to the physical world. There are some interesting concepts that deal with trust and verifiability of data entry, scanning, and sensors, though. They just don't have anything to do with smart contracts or block chain per se, though they borrow some high-level ideas from it. But this an area where block chain in general has no place and other mechanisms (simple rule-based plausibility checking, automated anomaly-detection, etc.) are much more helpful. The physical world requires different approaches.