Blockchain - Sicherheit auch ohne Trust Center

1. 00:00Now, let's summarize what we found in the first week of our openHPI course on Blockchain.
2. 00:08Because we've already got, as we're about to discover, a naive approach to such a blockchain, everything that's necessary.
3. 00:17So we saw, that consistency is difficult in distributed systems, there are different models.
4. 00:26That is the challenge we have according to the different models we have different errors.
5. 00:33We were particularly interested in a mistake like that, that's not easy to spot now, like a crash bug,
6. 00:38but we've come to terms with the Byzantine errors, so with all the random mistakes,
7. 00:43that always consist of doing something. not following the rules.
8. 00:51We have seen, even if we have such Byzantine models, so if we - and these were the traitors in the network -
9. 00:57so if we also have traitors in such a network, then cheaters, that we could then use a consensus algorithm like that to still be able to ensure consistency.
10. 01:09We have been successful with such a democratic approach, namely that simply the number of votes is evaluated in such a 0/1 decision.
11. 01:21You can always take that quite well as a model of thought, how to do that for more complex messages.
12. 01:28Then we saw what we could do, to attack this democratic consensus algorithm,
13. 01:37is to use such a Sybil attack to fake participants, to simply influence the voice ratio.
14. 01:45We've seen how we can stop it, and we've now seen in the last clip how we've changed the chronological order of the commands or messages sent in the network,
15. 01:58how we can determine the timing beyond any doubt without a central timestamp service, those were the linked timestamps.
16. 02:08We had also seen with the consensus algorithm, that if we have 2m+1 participants, we'll be up to m scammers, up to Byzantine errors.
17. 02:23And when we have this together, then we can think about it now,
18. 02:29that we're not going to put this on a situation like this with old generals, who want to attack a city or retreat,
19. 02:35but when we apply that to a bank like this. and wants to use the banking business, that's what it's all about,
20. 02:43that the bank, as an anchor of trust in such a system. always precisely records the exact order of the transfers
21. 02:53and anybody who wants to check it out. I have to see this bank's ledger.
22. 02:59The bank thus becomes an anchor of trust.
23. 03:03Now we had seen that this approach with Bitcoin was just that, we can't run the whole system without a bank, without a central point of confidence recognized by all
24. 03:22and we've seen or we've seen that we can do it.
25. 03:27Because if we now in this picture of the bank instead of in this picture of the old generals as messages to take the referrals,
26. 03:36then we can be in such a cryptographically linked list to record exactly the individual instructions and their sequence.
27. 03:46And we can create consensus over the head of such a linked list that all participants, even if they don't trust each other, acknowledge this flow of news.
28. 04:00Now, in order to cast an activity vote there. and prove that you're not a fake participant in such a network, we had the idea of the Proof of Works,
29. 04:15and are now simply in a situation where our Byzantine generals they're no longer at war, they're at a bank.
30. 04:25So we actually made it - of course that's a little simplified - to see how we could run such a bank, such a financial institution, without having this central, trustworthy office of a bank.
31. 04:42So we've made a trustworthy account book.
32. 04:48Our distributed bank keeps its account book on all instructions, she runs the distributed and without a coordinator.
33. 04:57So it is decentralized like in a distributed network.
34. 05:01The Byzantine mistakes that there are quite among the participants fraudsters are tolerated if there are not too many.
35. 05:10So it may not cheat more than half of the participants, and we can attack the Sybil that a participant because several other participants are faking,
36. 05:23we can fight it with a proof of work.
37. 05:27So there's no need to trust anyone, with the help of the cryptographic procedure this trustless account book is guaranteed beyond doubt.
38. 05:43And with the Linked Timestamping, so with these linked Hashlisten of the individual transfers we have a data structure,
39. 05:53so that every new link in the chain can only be inserted by a proof of work,
40. 06:00which is also unchangeable. and not allowing anyone to block this referral process.
41. 06:09And, of course, that was a huge sensation, because the banks, of course, had central positions,
42. 06:14that is also the big question in the years when the banking crisis was, as well as the confidence in the banks was shaken,
43. 06:22so that this approach of the blockchain then suddenly in all areas, in the media, was celebrated
44. 06:30and the meaning has just been recognized, that it is possible in such a distributed system, in a network of equals,
45. 06:39to keep a decentralised, trustless and unchangeable accounting book.
46. 06:45Naive we have understood everything by now how it is with the blockchain, but it's a little more complicated than that.
47. 06:57Namely, what have we achieved with our solution?
48. 07:00We've seen a lot of messages that need to be sent, every participant must to build that consensus, the voice of every other participant,
49. 07:12and he must be informed of the other participant's vote. ask another participant.
50. 07:20It was this democratic consensus algorithm.
51. 07:23So you got the message from the network subscribers, and you have inquired with other net participants, what message they got from a particular net subscriber.
52. 07:33So if we just add that up, when we have a subscriber on the network to send 2n² -2n messages, to be within this consensus algorithm.
53. 07:45And if now n is a large number, if we think of a bank with many, many participants, then of course is this square effort, which scales poorly and is impracticable for large networks.
54. 07:58So the way we've been doing this so far, it's not working out well yet. and not yet efficient and cannot be applied in practice.
55. 08:06And then this great pioneering work of Satoshi Nakamoto appeared, where the Bitcoin system was introduced, which was the same way,
56. 08:22namely not to create trust through a central instance, but in a network where there is no central trust authority, To establish trust with the help of cryptographic mathematical methods.
57. 08:39And this work then brought a way, how to make such an electronic payment system, a currency, namely Bitcoin,
58. 08:49how to actually use them in practice can have as pure a peer-to-peer network as any user, Participants in such a distributed system can participate in it.
59. 09:05There's no instance, no trust anchors, and yet financial transactions can be made.
60. 09:12So, and how this works now, Bitcoin, that's what we want on the basis of knowledge, what we acquired this week,
61. 09:18then take a closer look next week.

Щоб увімкнути запис, виберіть мову в меню налаштувань відео.