5. Blockchain Basics & Transactions, UTXO and Script Code



MIT 15.S12 Blockchain and Money, Fall 2018 Instructor: Prof. Gary Gensler View the complete course: …


49 Comments

  1. So, as I can understand, UTXO is dust of bitcoins, let's called a few satoshis that were awarded on coinbase or fees, that are so miniscule that are not worth spent them until they have some value in the future, and this UTXO, that are transactions outputs are stored in a seperate database in the bitcoin blockchain. What else did I miss ?? I might read further later.

  2. Hash function is a computation that transforms arbitrary length data into a fixed length output (the hash). E.g. for SHA256 the output is 256 bits of length. Goes without saying that if you hash the exact same thing twice you will get exactly the same hash (it's a function after all). However what hash functions also satisfy is that a small change to the input data (e.g. changing just one bit) should change the output in an unpredictable way. In other words hashing two similar inputs will not produce two similar hashes. If you have two inputs and you compute their hashes and then compare them: 1) if hashes are different, you know for sure the inputs are different as well 2) if the hashes are equal, there is a very high chance the inputs are the same as well. For instance, in case of a 256-bit hash, there should be roughly 1/ 2^256 chance for an accidental hash conflict. A good hash function should also be irreversible, meaning it shouldn't be possible to draw any conclusions about the input based on the hash output for that input. That last property is key for proof of work. If it wasn't satisfied, it would be possible to compute solutions faster than by blind trial and error.

  3. For aviva's question can we store data as a hash in cloud. In storage there is a technology called deduplication, where hash is primiary component which helps in reduction of storage size by discarding duplicate block of data.

  4. To his question initially about centralized banking system vs bitcoin fees – wasn't there another point mentioned in the bitcoin paper about reversibility of transactions under the banking system? i.e. that inevitable banks can step in and reverse a transaction compared to blockchain always being irreversible? i felt like that was satoshi's bigger point rather than the system being more secure…or am I misunderstanding the paper

  5. IMHO, one thing they left out of sight when talking about whether btc blockchain is truly different from commercial banking system (and this is a really good point) is what happens once all coins have been mined and there is no more inherent reward. Well, my educated guess tells me that fees will skyrocket particularly given that the size of the block is fixed. So, the question is what fee would keep miners incentivized while keeping others engaged in using btc blockchain. Fixed block size is a real curse of BTC's blockchain (though satoshi had a certain solution to this problem but nobody gave a damn at the time). I think that's why every btc hodler speculates about btc being a digital gold to justify its failure to act as means of payment. I am surprised that this wasn't thoroughly contemplated cause it is a HUGE issue. One might say that LN and Strike should solve the scalability issues of btc blockchain but it only dumps deeper the whole concept of decentralization.

  6. Just my two bits on the last question related to library and hash question. What hashes do is prepare a proof for the data but you can't trace back and create data from those hashes. For example, if you have War and Peace book with you, those words in it are nothing but 0 & 1 to the computers. The computer program will take all of them and convert it into 256 unit hexadecimal number called hash. Now, when you change even one word of War and Peace, it will give you different 256 unit hexadecimal hash. Thus you have a proof of the book, which can't be tempered with now. Similarly, you can do hashes of hashes, by taking the hash of all the books in the library and convert it into a single 256 unit hexadecimal hash. And that becomes the proof for the entire book in the library. Now, if you remove one book or even one word from any book, the final hash of the library will change. Thus hash gives you just the proof of content but not the content.
    Please correct me if my understanding is not correct. Thanks.

Leave a Reply

Your email address will not be published.


*