/sci/ - Science & Math

>>9670472
That's a nice clusterfuck there. Do people look at (n,m)-sheaves etc. (2 and 1 here) because it gets too complicated if you take n to infinity, or do you get something out of it that you don't in the infinite case. On the nLab, or when I peek into Luries stuff, it seems they really push for the infinite case.
And why would you need anything etale here? If I had the time, I'd get familiar with combinatorial species and I think that's where the applications of cat framework end, pretty much, for anything resembling real world data bases.
Granted, I don't believe you for a second, but in case this thought crossed your mind before you posted, what would there be to classify when it comes to the blockchain transactions? The protocols are compact and self-contained and predictable (only metadata really being the timestamp). And most "transactions" are not even on the chain, all the exchanges obviously write their own database and only do transactions if somebody wants to withdraw to another address. Pic related, the biggest wallets are the exchanges and they keep it mostly on a few addresses. When you trade with someone there, nothing moves. You'd not get too much out of studying the chain. Besides, if you were to pull something like that off, restricting you to BTC would be quite the mistake I think. Finally, most of the interesting action on the chain will be invocations of dApps.
On that note, I've deployed what I think is the first and only Turing complete smart contract (the contract itself is programmable and it's not "just" the virtual machines that are part of the nodes)

https://youtu.be/CAUo5aNmvz8