Topic: 195.200.253.240 is a real jerk (Read 4224 times)

Less than 10 minutes if he has more hashing power than the network had at the time and if he is hashing the rewrite offline.

Nope, he owns 15% of some flats but ONLY rents them to people who don't actually plan on living there. This way he gets his fixed income (50 BTC/block) but not the extra tips/gifts people give to their landlord at christmas or so (transaction fees). On the other hand he'll never have the hassle of having to fix broken pipes, deal with angy people in his flats etc.

Also to put it into perspective: it now takes 15% longer to get a (dead cheap) transaction in the system - 11.5 minutes instead of 10. If you wait for 6 blocks for confirmation, that's 90 seconds more to 1 hour of waiting, meaning 3690 instead of 3600 seconds. On the other hand it is 15% more difficult to attack the Bitcoin blockchain, which surely is also worth something!

Some solutions proposed here might make it even harder to scale bitcoin up to a relevant level (I call 1 transaction every ~5 seconds not really relevant...) and also will further disincentivize people from mining. There is already the drop to 25 coins per block in a bit less than 9 months which will make mining MUCH harder, if the price for Bitcoins doesn't double or the security (=difficulty) of the block chain doesn't get cut in half too because so many are quitting mining.

TX fees don't have to be raised at all - BUT if someone doesn't choose to pay a fair price (which TX fees currently really aren't), why should it even be forced upon miners to include more transactions that they feel comfortable with? Also this kind of stuff can lead to more branching of blocks, as it will (hopefully) only be a feature of the satoshi client.

Food for thought:
If someone already owns a botnet that can hash as fast as 15% of all miners together (if using CPU that's a LOT of computers!), don't you think it might also be possible to deploy a stripped down "dumb" client there, that relays blocks to as many satoshi clients as possible? The few ten thousand satoshi clients then would delay blocks between each other - but it wouldn't really matter, as they would nearly instantaneously anyways get the blocks from the "blockrelay" client.

What DeathAndTaxes wants is akin to an employee requesting a raise, and won't work until he gets that raise. Yet he demands full salary for idling and making noise at the office, and will claim that it adds to security of the business.  Now it's his right to demand a raise, but it is also the right of the employer to give him the boot.

To use another analogy, imagine a market where an individual monopolizes 15% of stalls without selling any products, yet he still wants to paid.

Raising TX fees might also affect users that don't have many BTC.  I think TX fees are good as they are.  We shouldn't placate greedy miners that lack vision.

A block is compiled into a merkle tree by the miner's bitcoind from transactions flying around the net in realtime. The tree is hashed¹, and once a hash is found that matches difficulty, the entire tree is stamped with the hash and sent out as a valid block. If the miner doesn't bother to gather the transactions that are whizzing around the network and compile them into a tree and hash it, only the generation transaction will be included, making the block "empty".

¹ Not all the data is hashed, but that is a bit deep to explain.

Funny thing is nobody would know about him/it if it weren't for the abnormal blocks.

No.  It simply has no transactions. 
It can't be seen in advance.
It isn't a block already mined.  


It is simply a miner who has chosen to not include any transactions (well technically he includes the coinbase = 50 BTC payment to himself as that is required by the protocol).

A 1 tx block:
https://blockchain.info/block-index/197553/00000000000004eb01a7063c81066ec47523516c2bcaa3308c1351c0e6176e68

A "normal" block (this one has 146 txs):
https://blockchain.info/block-index/197562/00000000000001b824bbb053f94a7f1af4b9265f4e09c580ef1c95e6137a6187

How does one know in advance that a block is empty?  Has it been previously mined by the network?

A miner (called "mystery" in the other 3 or 4 threads) is mining "empty blocks". 

Blocks which contain no other transactions (other than the reward one to him/her).

Some people have their panties in a bunch because it is (pick one or more):
a) "not fair"
b) evil
c) terrorism
d) going to kill Bitcoin

Currently tx fees as so low that miner is only "losing" 0.06% of potential revenue by mining empty blocks.
Higher tx fees would increase the amount of BTC the miner is losing and possibly encourage a change.

Can someone explain this to me in layman's terms?  Preferably in mono-syllabic words

thanks steve, d&t, ribuck... i'm not really a miner at all so sometimes i just talk out my ass when it comes to how these things work.

How about merging this thread with the mystery miner one? it's getting harder to keep up with this.

It's quite legitimate to not include transactions in your block if the fees aren't high enough (which currently I guess they aren't!) to make it worth your while. This is the dynamic that encourages users to pay fees and it's a normal part of the bitcoin system.

Heh, strange logic that.

He is mining, obviously because 50BTC is a sweet reward. Whereas fees are negligible, so he doesn't give a flying one about them.

If 50 BTC or 25 BTC reward are not enough of an incentive, then they shouldn't be mining. Botnets are bad for Bitcoin's image anyway.

Again profit might not be the goal here.

Adding to network security?  Really?   I think we are better off without his "contribution" to network security.

Why a Nazi, why not DeathAndTaxes?

You seem awfully defensive.

The marginal cost of adding a transaction to a block is pretty small.

There is the cost of hashing the transaction; then recalculation of log(n) hashes where (n) is the number of transactions already in the block.

This is because changing a node in the merkle tree requires recalculation of all the parent hashes too.

While the marginal cost is small; the total cost might be off-putting.  In order to add a transaction you have to be able to check its validity.  To do that you need access to a correctly indexed copy of the block chain.  That requires a gig (and growing) of space.  Once you have that maintained, adding a transaction isn't a huge amount of computational work.

I would guess that that is the reason this miner isn't including any transactions.  The work needed to go from adding zero transactions to one transaction is enormous.  Once you can include one; including two, three or four is easy.

If I were writing a botnet, I wouldn't bother including transactions either.  After all; you're already an ass-hat as you are stealing your hashing power; what will you care for the bitcoin network.

The solution is time.  As the block reward reduces and the transaction fees increase, including transactions will become the way to earn from mining.  This problem will then automatically go away.

You are all ignoring the development effort required to make a collaborative implementation. If the incentives are not sufficient, people will not relay transactions, it's a simple as that. Especially if they have zero variable costs (botnet) and high fixed costs (developing exploits and the custom mining software).
The critical path to profit is a non-collaborative miner. Once I have that, maybe, just maybe, I will start working on transaction relay and Merkle tree updates. If it looks like allot of work and the fee revenue seems modest, I'll pass.

Terrorist?  Really?  Following the protocol rules and adding to network security = terrorism.   

Are you feeling terror right now?  Not frustration or anger but honestly terror.  As in his actions are paralyzing you in fear?

Whats next he is also a Nazi and/or responsible for genocide?

That only if his motive is profit, but I doubt it.  And if that were the case, one should never give in to terrorists.

No, that's not how it works. The only things that count are the difficulty and the number of blocks.

Someone might win a block after spending 30 minutes hashing, or they might get lucky and win the block after spending 5 minutes hashing. But the strength of the blocks are the same.

A bad guy rewriting the block chain will still take on average 10 minutes to find a replacement block, no matter how long it took the winner of the original block.