1. 22
  1. 24

    Is a full blockchain and contract engine really necessary for this? Trying to understand.

    1. 6

      It’s probably a hackthon idea about blockchain/contracts that has been continued without much research on the customers.

      Good on them to open source it though :)

    2. 14

      This is pretty interesting, but I think the real test for “society-changing” applications of cryptography like this is if normal people can use it without making mistakes. In particular, the interface for restaurant owners probably needs to be basically seamless— no restaurant owner wants to have to learn the details of public key cryptography, they want to make and serve food.

      also, it doesn’t look like there’s any details on how deliveries are going to work? all I can see is a somewhat overengineered protocol for placing an order with a restaurant, for a which a great decentralized system already exists: a landline phone. How would you prevent a delivery driver from jacking your food, or someone who claims to be a delivery driver from taking food?

      1. 4

        There are countless stories of people losing money because they’ve accidentally deleted their BitCon wallet key, or they’ve sent money to someone else and it turned out that they were a fraudulent entity. Much as I like decentralisation, centralised solutions have one big advantage: accountability. In the case of fraud, my bank can reverse transactions. If they do something illegal, they have a registered address that the police can visit and a load of recorded assets that can be seized. As a result of this, ‘know your customer’ legislation can be enforced, which make it much easier to find the perpetrators of fraud and help shift the liability towards banks that enable it. These things are really hard to replicate in a decentralised system.

        These aggregators do provide a few bits of value to restaurants, only some of which are captured by this:

        • They provide a single place to browse for a load of different things, which helps discovery. I’ve tried a bunch of takeaways via Deliveroo and Just Eat that I’d never have heard of otherwise.
        • They allow restaurants to outsource delivery. Unless you’re doing a lot of delivery business, paying people to deliver for you can be expensive. If you have slack times, you’re paying them anyway, whereas outsourcing it means that other restaurants can take up some of that slack.
        • They provide a reputation system. Deliveroo’s differentiating feature at launch was that they were selective in the restaurants that they’d sign up. If a restaurant gets too many bad reviews, they’re kicked off. Even Just Eat, which accepts pretty much anyone onto the platform, tracks reviews and knows that the person leaving the review actually ordered (and paid for) the food (and it was picked up by a delivery person who wasn’t affiliated with the restaurant), which makes it much harder to scam.
        • They handle refunds. I had a pizza delivery person accelerate too hard on his scooter so that my pizza was smushed into one end of the box. Just Eat handled the refund immediately. Again, knowing that I won’t have problems with refunds increases my confidence and makes it much lower risk for a customer to try a new take-away.
        • They handle all of the payments. Most restaurants can handle credit card payments in person, but doing so online requires more infrastructure. Outsourcing this reduces costs.

        The big problems with these companies are that they’re abusive to their delivery workers (Deliveroo recently had an IPO and their share price tanked immediately, in a large part due to the fact that they’re expected to be taken to court soon and end up having to pay their riders more) and they take a disproportionate cut of the price.

        In a decentralised system, there are a bunch of other questions:

        • How is the data handled privately?
        • Who is liable in case of a GDPR violation (is it the individual restaurants who opt in?)
        • How do I know a restaurant is legitimate / of decent quality?
        • If it handles matching riders with restaurants, how does it comply with employment law?
        1. 4

          Accountability is just the flip side of power abuse. If you can prevent people from signing up, reverse transactions, delete their accounts without traces, etc. you might as well do a good thing when someone asks you kindly.

          Distributed and decentralized solutions are usually created to avoid giving some instance power, and this is one example of the downside that this position grants you, but it is to a certain degree unavoidable – unless you regress on the central principle of being distributed and/or decentralized.

          That doesn’t mean it doesn’t have to be all bad. Different approaches can be taken. Maybe you could have “trustworthiness indexes”, where some food critic you trust (or pay) publishes how good a restaurant is, comparable to block-lists for ad-blockers. Maybe you could have a gossip system where friends can recommend or advise against visiting a restaurant? It is difficult, but without a central authority, there is no “definitive” knowledge. But then again, “real life” is also a distributed network of humans and their relations that suffers from the same problem.

        2. 1

          “ …great decentralized system already exists: a landline phone…”

          That does not offer automation or elasticity/scale. Not trying trivialize a response to your assertion, but I think automation is needed by any small business with a razor-thin margin.

          “…How would you prevent a delivery driver from jacking your food, or someone who claims to be a delivery driver from taking food?”

          From the readme, they seem to rely on strong digital identity of each service provider:

          “Identity creation for resource providers is made costly with a computational proof-of-work mechanism based on the partial preimage discovery first employed by Hashcash…. Free Food requires resource providers to supply a photographic proof of identity which includes a unique symbol which is mathematically bound to the proof-of-work associated with their public key.”

          That does not prevent a verified provider to do bad things, of course. But we can assume that the bad behavior, if it happens, at all, would happen only once. So the outcomes would not be no different than in centralized solutions.

          “… cryptography like this is if normal people can use it without making mistakes…”

          Evolution of technologies.

          Personal hardware cryptowallets, (that,also, incorporate password wallets) already exist. And, then, in the future, perhaps, incorporate more things (like health record, employment record, EDU record, asset records) – will likely to be a thing (assuming that we are allowed by governments, to have digital personalization, but without centralization).

          It should not be that difficult for this project, to approach 2 or 3 hardware cryptowallets provider and ask them to allow their solution to store the identity and authorization tokens for the libfood based systems. These, in a way, are tokens of trusts, and have value across locations, countries, decentralized networks, etc. I hope more and more such things are done. As person goes through their life, accumulating good ‘verifiable resume’ is of enormous value (regardless of the industry).

          Overall this service is looking for ways to decentralize (and therefore, if you accept the leap of faith, democratize ?) the discovery and integrated delivery part of the restaurant business.

          I can certainly see how this would work way beyond their modest goals, in many other businesses.

          1. 1

            also, it doesn’t look like there’s any details on how deliveries are going to work?

            Lots of local restaurants have, or could hire and staff, drivers, pizza places have been doing it for decades. This gives the restaurant owner/manager more choice and accountability, in comparison to grubhub type services.

            a landline phone

            I personally love the landline phone, BUT, I think the benefit over a landline phone is pretty clear to anyone who has worked a rush hour (it’s been a long time, but I remember). In this case, you’re avoiding a lot of things:

            If you have integrated payment processing (like he mentions), you’re avoiding transcription errors around card numbers. But most importantly, you’re saving time, and queueing in a different part of the system. When I call my local pizza place on a friday night (on their land line), I sometimes can’t get through until the third or fourth try (it’s good Pizza). Also, the system avoids mis-ordering or entry errors on the part of the order-taker. Finally, it gives the buyer an opportunity to double-check a placed order, or add impulse items ;). I think it’s interesting, and it’s just my perspective. Thoughts?

          2. 6

            Cool idea - but it seems too much tech, not enough customer. Seems like there would be a huge onboarding cost in time effort to understand

            1. 4

              “Identity creation for resource providers is made costly with a computational proof-of-work mechanism based on the partial preimage discovery first employed by Hashcash.”

              Dude, stop. Stop using PoW.

              1. 4

                Currently small, location-based businesses experience enormous dilution of their provider-to-client relationship value.

                Every major venue of business discovery and experience feedback interaction between a business and their client, is taken up by these aggregators, the middleman.

                When you go to your favorite search engine, and type a query about food place for specific type of food in a specific area– you will see not the actual restaurants, but the aggregators offering you ways to interact with the ‘restaurant’ (often without the restaurants even agreeing to this substitution).

                My initial read is that this system is trying to give control back to the small business owners, the service providers – by replacing the centralized aggregators/middleman with a decentralized trust system, while preserving technological advantages of a ‘standardized’ customer-engagement workflow.

                1. 2

                  Free Food uses OpenStreetMap’s Nominatim open source geocoder to convert street address to latitude and longitude.

                  IMO this makes this DOA in North America, as I haven’t had great experiences with Nominatim.

                  1. 2

                    Depends on whether it’s the only system they use. I recently used a delivery service that dropped the pin on the map from the address and asked me to move or confirm it (Uber does the same). If you’re on a mobile phone, you can just grab the GPS coordinates to start with instead. That’s generally a much better way of taking delivery drivers to exactly the right place. We moved into a new house a few years ago and it took Google Maps over a year to add the road that we were on (it was on OSM as an in-development road as soon as planning permission was granted and showed up properly shortly before we moved in), getting delivery people to the address was painful. Our postcode didn’t exist in most systems. If we’d been able to just drop a pin in the map, it would have been trivial.

                  2. 1

                    Neat! I’ve signed up for email updates.

                    How does this compare to Secure Scuttlebutt?