Transcript for #bitcoin-dev 2017/12/31

00:07 phantomcircuit SmashShock, a utxo can be spent a total of 1 time
00:07 phantomcircuit there's no balance
00:07 phantomcircuit simple yes or no
00:08 SmashShock How so? If a UTXO has a value of 25btc, and I only want to spend 10btc, how does this split?
00:08 SmashShock Legitimately curious. Haha
00:10 SmashShock Or is a transaction spent in full, and there can be multiple outputs?
00:11 SmashShock Oh god, that's what it is, isn't it
00:11 SmashShock So the input is a UTXO, and the output is two UTXO's, one for the recipient, one for change?
00:25 SmashShock Okay, well here's a hypothetical
00:25 SmashShock Would there be anything wrong with a system that has UTXO's that can be spent multiple times?
00:26 SmashShock And once they hit 0, they are pruned from the UTXO db
00:26 SmashShock (but are still in the BLOCKDB, of course)
00:26 SmashShock phantomcircuit:
00:27 SmashShock And then a client just keeps a list of the UTXOs that belong to them, and they can spend them accordingly
00:32 phantomcircuit SmashShock, you dont split, you spend the full 25 and then have an output which pays to yourself
00:33 phantomcircuit an unspent transaction output can be spent a total of 1 time
00:33 phantomcircuit SmashShock, there are privacy issues with a system like that as well as practical issues
00:34 phantomcircuit (not to say privacy isn't a practical issue, but you get what im saying i think)
00:35 SmashShock Yeah, I suppose, if you don't have a system for multiple inputs, "coin chunks" can only get smaller, and eventually 25000000 coins in one transaction would become 25000000 different UTXOs
00:35 phantomcircuit SmashShock, but of course the system does allow for multiple inputs
00:35 phantomcircuit cause well
00:36 phantomcircuit otherwise things fail quite rapidly
00:37 SmashShock Hmm. Okay. I've been writing a cryptocurrency from the ground up, just to learn how they work
00:37 SmashShock Switching to that transaction model will take quite the rewrite. Haha
00:38 SmashShock But if it's necessary, I suppose I'll give it a shot!
08:03 drTrololo good morning
08:06 drTrololo happy to be here. im grapfics designer, blockchain researcher, coder and my english not good.
08:08 drTrololo how i can help and made something useful for bitcoin? which best step for noobie?
21:34 parco AndBobsYourUncle: eevblog fan?
21:36 parco Is there any source of data on this earth that is truly random, while also being tamper resistant?
21:37 parco and able to be verified globally
21:42 arubi cosmic microwave background?
21:58 parco arubi: is that data available to everyone around the world
21:58 parco and can't be forged/tampered with?
22:00 arubi it's coming from the edge of the universe so it's pretty hard to mess with.. I was only half joking about using that :)
22:00 parco i'm half serious about this idea
22:01 parco is it possible to have a PoW algorithm, that takes a random salt, that would require a completely new ASIC every time that salt changes?
22:02 arubi not sure about that. I don't really know what you're trying to solve
22:02 parco Miner centralization
22:02 arubi how would it help to have a convoluted mining algorithm?
22:03 arubi isn't a very easy and straight forward algo is better? one's that very cheap and easy to implement on asics
22:03 parco it takes millions of dollars to spin up an ASIC chip for a specific algorithm. If that algorithm changes every blocks, it would discourage anybody from making ASICs
22:04 arubi they'll just make an asic for all possible algorithms
22:04 parco is that possible?
22:04 arubi sure why not?
22:04 parco All possible algo's are known?
22:04 arubi can you program something unknown? maybe I'm missing something
22:05 parco if it was created from scratch, and agreed upon
22:06 arubi parco, whatever you come up with, folks can build a chip that does the work
22:06 parco you might be right, any laws that prove this so I can save some time
22:07 parco I felt like a true source of randomness would be the savior here
22:07 arubi randomness isn't the issue
22:07 arubi the issue is that some mining corporations have an edge on other miners
22:08 arubi so no matter what the algo is, centralization is better for small miners
22:09 parco I see. Do you know of any existing efforts being done to mitigate miner centralization?
22:10 parco I feel it's a serious problem at the moment, do you agree?
22:10 arubi I agree. I don't know what can be done about it really