Topic: Now that we've reached the 250kb soft limit... (Read 3646 times)

Compare the following case:
1. News articles start appearing in the media pointing 7tps hard transaction limit of bitcoin blockchain
2. News articles start appearing in the media pointing a hard fork happened inside bitcoin which created two different bitcoin network: Bitcoin legacy and Bitcoin 2.0, and there will be 3 types of coin: Pre-fork coin, after-fork coin 2.0 and after-fork coin legacy. And possibly in the future there will be another new type of coin together with another fork... with each fork another 11 million of coin were created immediately

Which one do you think will more likely to cause people abandon bitcoin  

Dead puppies aren't much fun.

This is interesting, but can you just clarify the term again, please?

Some people might be interested in some analysis of recent blocks.

There seems to be a substantial rise in high fee transactions (fees as high at 0.005 BTC/KB) out-competing other transactions.  I've been calling these transaction dead-puppies due to a incomprehensible made on IRC. In any case, running a script to identify the DP-involved transactions on the recent blocks show. A transaction is considered DP involved if it pays to an identified DP address or if any of its inputs paid to an identified DP address.

Height  Size    Amount of DP-involved.
224737 163012 82.0755%
224736 498888 94.9111%
224735 163124 27.1875%
224734 249140 93.4021%
224733 351883  6.77507%
224732 498991 85.5789%
224731 249033 67.8657%
224730 169849 55.9459%
224729 487840 76.87%
224728 249091 80.2395%
224727 249204 64.2512%
224726 187733 69.8947%
224725 121400 49.0196%
224724 248878 69.899%
224723 248923 72.5989%

The couple blocks with lower than typical DP concentrations appear to be miners that have also noticed these transactions an are apparently depriortizing them, though they don't get all of them because they appear to identify them only on the basis of output addresses rather than the transaction history.

Of note, the concentration was similar to somewhat lower 24 hours ago when there were fewer >250KB blocks.  Because the supply of these transactions seems to be basically unbounded it seems likely that adjustments to block target sizes are unlikely to produce faster confirmations at this time.

+10 internets verify

Simply leaving the max blocksize limit at 1mb and walking away is such a huge mistake, I really can't believe so many people don't see it.

If the bitcoin community doesn't deliver on small transactions, someone else will.  You know who that someone else will be?  Paypal.  Banks.  Government.   By leaving the blocksize limit in place, you're handing the power right back to the corporate and government interests that you just pulled off a fucking miracle taking it away from.  All this nonsense about "a market will emerge" is just bullshit, Paypal is what happens when you just sit back and let someone else do it.  People have no idea what's possible until you show them.  If you just let the market handle it, the people will just get what they expect.

Bitcoin has gotten as far as it has precisely because it's not controlled by corporate or government interests.  Don't stop now, the finish line is still a ways away.  

I'm not saying the blockchain necessarily needs to store all these tiny transactions.  But the solution needs to come from right here.  

I agree, I don't give a shit about anyone else's alt chains, bonded banking startup, or ripple investments. I love Bitcoin. Fuck anyone who doesn't.

thanks for the explanation

I'm not sure if that's true. In some cases I'm more inclined to believe malice to be more likely than ignorance.

Some people don't want Bitcoin to be successful.

The 250KB "limit" is simply the default maximum block size created by the reference implementation that most miners use. Nodes will still accept up to the hard maximum block size, which is 1MB.

For instance, on testnet there have been two 1MB blocks created: 00000000476781c04b82b3ea91af1a86f3a863e1c9312b50302ffa01b7bdf960 and 0000000010bf4453b170a6756d911e207734ae181af6c8c02b42787d5885b333

how can be block 224711 more than 250kB?

http://blockchain.info/it/block-index/355118/000000000000003a49c4828ebf1b042f48058f4bbec461b5e3648a30e9e30d4d

This is so fucking naive.

1) The network hashrate will not be maximized in a global context. Restricted number of transactions equals a supply ceiling. A supply ceiling prevents the supply / demand curve from being maximized. A non maximized curve means that total transaction revenue per block will ultimately be lower. This is also ignoring that BTC to local currency rates will be lower, further reducing economic incentive to invest in mining resources

2) Bitcoin will be less valuable as a store of value. Less total transaction revenue per block, fewer users, less overall commerce done using bitcoins, smaller economy represented in bitcoins, BTC to [your currency here] rate is lower, fewer mining resources, LOWER TOTAL HASH RATE. Lower total hash rate equals a less secure store of value

Now pay attention to this last one

3) A declining BTC to other currency price means a terrible long term store of value. Declining types of potential use cases (resulting from more expensive INDIVIDUAL transactions), means smaller total economy represented in bitcoins, which then means a stagnant or declining BTC price. A bitcoin never grows into the $1000s without Bitcoin representing a significant portion of the global economy. A transaction limit never allows Bitcoin to do that.  

i stand corrected.  i wasn't aware SD was paying those kinds of fees. 

That's totally wrong, the negative consequences for leaving that block size alone are that Bitcoin fails to adequately scale, fees for accepting bitcoins grow prohibitably large (larger than other payment methods because Bitcoin usage becomes dominated by fee insensitive transfers like gambling), merchants stop accepting bitcoins, and it ultimately fails and is replaced.

You act like a switch can just be flipped in the future, it may end up being too late.

"pending transaction memory size" is going to 1) be different for every node and 2) be vulnerable to manipulation.

Any scheme for adjusting the block size needs to be based purely on the information found in the block chain.

Why the limit is not in some real time proportion with the pending transaction memory size? Then confirmation of found block would have to accept block size witch are not bigger than the peak pending transaction memory size since last found block.

thx

This is a separate problem, that should be solved in some other way. I wouldn't mix that issue with the issue of the block size in general. We should be able to implement more extreme spamming rules that simply make it impossible to send that kind of transactions, or something like that.

I don't understand why the gambling is a problem. I really don't. It's an absolutely valid use of the Blockchain. They pay 0.0005 for every transaction which is not something that even all regular users do. It is QUITE ENOUGH to pay the miners for the END OF TIME.

Have you guys calculated this properly? If SatoshiDice made 1 million transactions per day right now, and payed 0.0005 for each, that is _500_ bitcoins in fees per day by SatoshiDice alone. That would be 3.472 BTC in fees per block. If that is not enough for 1M tx (which is 20 times what we have now), then how about a 0.001 fee (which is still in the limits of being currently acceptable), it is approximately 7 bitcoins per block in fees.

How much do the miners need? We already suffered a very significant drop from 50+fees in reward to 25+fees reward for the miners, and the network is currently seen as extremely secure. What is the amount where we deem it "not secure enough"? Or should we limit our system in purpose and just pay the maximum amount that people are willing to pay (which would lead to 99% of Bitcoin users becoming non-Bitcoin-users).

How about 5 million transactions? Let's check out the numbers. It's approximately the amount PayPal has per day. That is 60 tx/s, someone said Bitcoin is capable of around 7 tx/s. So if we raise the block size to 10MB, which gives Bitcoin PayPal level transaction amount, and each of them involves a 0.0005 BTC to transact.. let's see. That is 2500 BTC in fees PER DAY. Which is 17.36 BTC per block.

So HOW EXACTLY is a 10MB block size, with 5 million transactions per day, at a reward of 17 BTC per block (with our current fee!), a disaster? What exactly does the reward need to be? This is what needs to be thought about.

Edit: It's of course good to remember that we're still going to have some kind of fixed reward for the miners for a very long time, even though it halves every 4 years. Here I looked at the fee reward structure alone.

That's right, once Bitcoin transaction volume rises sufficiently the gambling transactions will need to pay higher fees to get reasonable confirmation times.

However, the problem with these gambling transactions is not their frequency (since they are not currently crowding out more legitimate transactions) but rather that they produce unspendable outputs. I'm talking specifically about SatoshiDICE's "lose confirmation" transactions which send 1 satoshi back. They can't be pruned from the block chain but also cannot be spent (well, it wouldn't make economic sense to spend them).

clear thinking.  the market will seek out a solution.  i think it's too soon to be changing things.

the normal response to those advocating change would be to say, "well go start your own alt chain".  but that would be difficult since no miners would follow you b/c of the higher fee structure of the current Bitcoin system.  higher fees will increase miner participation and thus higher security and encourage more investment.  more miners will be encouraged to mine thus increasing competition and keeping fees under control.

free spamming of the network by gambling tx's have the potential to clog the network.  they should be forced to contribute some fees towards miners costs.  and by forced, i don't mean by any of us; by leaving block size the same the economics of the situation will do the forcing.