|
|
|
|
|
|
by gldalmaso
2977 days ago
|
|
|
I was at a similar conference last week and the same thing happened and the same IBM case that tracked produce supply chain was presented. The same "wow" moment was shared when they said they can now track the origin of a product in under 2 seconds where before it took them a week. I happen to work in a product that manages produce traceability, and I can confirm that it is quite within the realm of possibility to track produce in under 2 seconds with a simple web app and a relational database. We have been doing it for about 10 years. To add, the produce supply chain is full of connectivity and digital literacy issues, so I would venture a guess that a blockchain implementation would generate a lot of attrition on the operations side of things. Another point of concern is performance since this is a fast moving supply chain due to the product short shelf life. It was interesting to see how this audience thinks that blockchain somehow "solves" traceability when this has been going on in several different forms in a wide array of different industries with specific needs. They seem to also think that the right way to do it is with IoT and smart contracts, as if it were just a step away to push such a solution to an existing supply chain. It keeps me up at night that one day a top-down decision will come that says we must a have a blockchain solution because market, or new competitors will bloom in the face of absurd amount of money being thrown at blockchain solutions and take market share. I imagine that 5 years from now, the winners of today's hype battles will be the losers when they have to maintain an overly complex inflexible solution for many years more. |
|
|
But now IBM can do it in 2 seconds as well! This may create a big boon for "DB2->blockchain" conversion contracts.
We still seem to be in this weird period with blockchain stuff where it's a neat idea, but the problems people are tackling with it are mostly... mild variations of solved problems so far. This may be understandable - these are problems which are known to have value, and as new companies want to solve these problems, they may opt for a blockchain-based version vs traditional "old school" relational db. But there may also be limited value there compared to old school version.
This feels a bit like "key value store" systems vs databases. There was (is?) a lot of hype around schemaless and "lightweight" KV stores, but most of the problems are also solved by intelligent use of SQL. Ditto for the XML-databases rage from ... 12-15 years ago.