Topic: 66 minutes between blocks? (Read 330 times)

I'd be open to considering a reduction in block time should a good argument be put forward for it, but it does not solve the issue we are discussing here. No block time will ever be fast enough for reliable point of sale transactions. You will either be waiting at least a minute or two with the risk of it being 10 minutes or more for your first transaction, which is completely useless for someone paying for a coffee or a fast food, or you will get a confirmation in seconds but that confirmation will have too high a chance of being reversed to be sufficient, so you will be stuck waiting for more anyway.

Sure, propagation and block validation is faster than it was 10 years ago, but that increase has slowed and will hit a limit. We still see empty blocks being mined (because of how long it takes every miner to receive and validate each block) and stale blocks fairly regularly, so propagation and validation are obviously still not fast enough to completely remove these issues even at a 10 minute limit.

I have zero interest in any of those altcoins so I couldn't really tell you, but I know that Ethereum uses different methods for dealing with such blocks (which they call Uncle blocks) and pays miners for them, and a quick glance at a Dogecoin explorer shows that they have mined over 200 completely empty blocks in the last 24 hours alone.

And even if you wear a seatbelt if you fall in a river after the accident and this gets blocked you die drowning.
Sorry, but this is how you sound now to me, you're clinging to an arbitrary time that was chosen in a day when the internet speed was 10 times lower, Satoshi didn't have a clue about pool mining or at least GPU mining, there was no segwit, there was no FIBRE and still think those are the best choices one could have made. Even car speed limits have changed a hundred times all over the world to reflect traffic and safety and in this case, we're talking about a simple case of throwing a body against a chunk of steal but no, god forbid we touch the magical number of 600 seconds despite the fact that block checking and propagation times has been cut in a thousand pieces. You have 10 pools with 96% of the hashrate linked between them at ms speeds but no, there will be a ton of orphaned blocks if we cut the time in by one minute. Seriously?


How often has this caused any real problems with Ethereum or Doge or Litecoin?

And with a 1 minute block time you can still pay 1 BTC/vbyte and be waiting upwards of 10 minutes for your first confirmation. Decreasing the block time will never be sufficient for a point of sale transaction, as by the time you decrease it enough to mean you'll never wait more than 30 seconds for a block, then it is so fast that stale blocks and double spends would be so common place that you'd have to wait 10+ minutes to have enough confirmations that you are sure they wouldn't suddenly all be overturned.

No, but they are rare enough that they aren't a major issue, and a chain of more than 1 is exceedingly rare. Drop the block time to 1 minute and you'll regularly see chains of 3 or 4 being overturned. Monero initially had a block time of 1 minute and had to increase it due to the excessive number of stale blocks they were experiencing.

Probably stop for a moment and start to think about my life choices.
Might sound incredible to you but in my circle of friends/workers/relatives, 90% at least of the money is already in the bank, so if I would count out Bitcoin funds I would have close to zero to deposit into my bank account from whatever cash reserves I have.
Actually, I don't know anyone around me who is still paid in cash, and the number of people who withdraw all their money from their card on payday has dwindled a lot even before the covid scare and the things this has bought.

So you see, I have no extra step, I have my wage wired to my bank account, one-click deposit money to the debit (spending card).
If I would have been paid in Bitcoin, I would still need that block to reach LN.


Appreciating the honesty.  , I knew the answer, I love when people are saying wait for a time when fees are low, use LN, etc etc but then when I ask if 1 million would be in the same situation, not just me the only answer is centralized solutions.
Anyhow, ranting time is over as we're going too far from the main subject.

And how's that a better analogy? What will you do once you run out of money in your bank account? Won't you deposit? Fiat & Bitcoin are the currencies; bank & LN are related, leave visa out. It looks more like the comfort provided by the blue wallet's LN implementation which works custodially.

The LN is more like a trustless bank. You still need a third party, but that happens non-custodially.

I wasn't planning to use this as an argument.

For each person? It's practically impossible. I'd rather see a smarter way of paying lightning-ly; maybe having one channel for each family or by using a fully custodial service. Yeah, I said it. If they don't care about that, which they mostly don't, we aren't going to force them be responsible for their own money.

But, someone who demands it should have that right, agreed?

Nope, wrong analogy. Bitcoin on chain (bank account) - LN (visa).
I can transfer money from my main account to any of my cards in a matter of seconds, so, no it's not like that. If you think of funding your bank account from an external source like cash deposits, it's the same with bitcoin, you still need those extra steps. And before you say you can do it faster at converting cash to bitcoin at BATM you can do the same at an ATM for a debit card or bank account.
Bitcoin can only rival international transfer, for national ones those days are gone.

Besides, since we're speaking of time and having LN funds available, how long it will take for the entire population of Salvador to open one channel?  . Even if they don't close one and there won't be any kind of other traffic, about a month?

The lightning network is, essentially, a time deposit. If you're left with some cents in your bank account, you'll have to deposit, which takes time. Same thing happens here; you deposit bitcoins onto your lightning “account”. So, I'd say that unless you find the bank deposit funny, the need to increase your funds in LN isn't funny.

It isn't a pattern. If the transaction that takes at least 40 minutes to get confirmed is less profitable to be included into a block compared with the other 3-4, then they'll obviously be confirmed sooner.

Well, the max time between blocks that I've witnessed is around 55 minutes so I wouldn't be surprised is a 66-minute time interval since it's 11 more minutes that may happen.



I can't say for certain but I've seen a pattern where there will be 3-4 transactions that will get confirm within 15 minutes and then there will be 1 transaction that will take at least 40 minutes to get confirmed, I've seen this pattern many times. I don't think it makes sense but you can check it out yourself.

It is not unusual, on average a block should be found in 10 minutes, however that is the average, there are instances in which several blocks are going to be found in a span of a few minutes while sometime a block needs more than an hour to be mined, this is not rare and it should not be a sign of worry as this is how the network is designed to operate, I know this can be troublesome if you want a fast confirmation but it is what it is.

I can pay 1BTC/vbyte, if no block is mined that doesn't solve anything.
Common, I understand you want to defend the decade-long stone age setting but you're grasping at straws here, putting only the things that won't be solved first and not the ones that would be.


How many times did that prove to be a problem with Ethereum?
Now, If I would apply the same negativism as you, it will mean that to be safer we would need at least 1 hour blocks as even 10 minutes block time didn't prevent orphaned blocks.
Satoshi never said a word about segwit or asics or fibre but at the same, he offered no reason for the 10 minute time, so why would some things be allowed to deviate and others not? We already messed with the propagation time because of segwit, whatever math was used for the previous value can't be applied anymore, so why stick to it?

Or just wait and plan ahead. 1 sat/B should be the default and bumping that up should only be done in emergencies.

If you let the wallets decide, then when an exchange moves some funds from cold to hot or back or a market action triggers a bunch of trading bots, fees rise and the wallets start rising fees in response causing a cascade effect that takes quite a while to cool down. As i have mentioned before, its not a Bitcoin issue, but a wallet issue, but people blame Bitcoin.

In the old days the default tx/fee was fixed, now there is a guessing game prone to manipulation which artificially inflates the fees. You can opt out by going manual again in some (good) wallets, but all the garbage custodial wallets many casual users use (from exchanges, etc) won't let you decide anything. And some people go off-chain instead (ie. altcoins) which is even worse.

Only because a wallet won't let you pick the fee, as it was originally conceived...

RBF should be available in many wallets now and is something extra you could have for emergencies as well.

Cutting the block time doesn't solve this, though. Even with block times of one minute it is easily possible to underestimate a fee, the mempool to fill up, and your transaction to be delayed. The solution to this is enabling RBF so you can bump your fee, not shorter block times.

The reason not to do so is the big increase in stale blocks you would get with a one minute block time, which would result in more potential double spends and therefore mean you would be waiting for more than a couple of confirmations anyway. Not to mention all the increase in wasted work.

What if I don't want s* like this to happen?  


You send a payment and for one hour and a half no blocks are mined, mempool floods and surprise, when I come back I see this.
It's not just coffee or food, maybe you want to send someone some coins, maybe you want to fund your exchange account and you have to stay glued to your screen waiting for blocks no matter what fees you choose. Oh, use LN. What if you don't have enough funds in LN? You need...to...wait for the next block! Admit it's funny!

Common, let's be honest, there is NO way to have something slower when it could be done faster! I'm not even demanding the same blocks space to not enter a new debate about hardware requirement, cut the time in ten, and the space by the same. There is no reason not to do so!
Oh, and since I'm in the complaining mode, two weeks difficulty adjustment?

God was merciful that Bitcoin wasn't invented in 2001 or we would have 10kb blocks every 3600 seconds.

None, and that's the point. None of this matters to the customer - it's entirely up to the merchant if they want to accept zero confirmation transactions or not. But from the point of view of actually being able to spend the money and knowing that the money isn't going to be unilaterally removed from their account without their consent, bitcoin wins hands down.

I don't see the reason to change it. If you drop it to 5 minutes or even 1 minute, then it is still too slow to get confirmations for immediate sales likes coffee or fast food, and there is still the risk of a block interval of >10 minutes. If you want immediate confirmations, then no block time will ever be short enough, so just use Lightning. If you want complete finality, then you will still be waiting the same amount since it would take 60 confirmations with a 1 minute interval to reach the same finality as 6 confirmations with a 10 minute interval.

And which customer do you think gives a fuck for how long afterward is the money still inaccessible for the merchant?
The customer doesn't care, he swipes his card, sees the message grabs his products and he's out of the door, that's what counts for him, if you would be required after a card payment to wait 5 days in the shop before being allowed to take leave then nobody would use, even if the time would be 10 minutes, most would simply drop card payments and go back to cash.


Yeah, I bet a lot of people who bought mining gear via bitcoin would love to have that option, tough luck, transactions can be reversed, scammers got away with all the money. Of course, for a merchant, it's a dream to keep the money no matter what but for a client that has been cheated, received a fake product, didn't get anything, that doesn't sound too good, isn't it?

There is no perfect system, each has flaws, but I really don't see any reason to stick to the 10 minutes rule, hundred of coins are working flawlessly with shorter intervals.

The stigma against unconfirmed transaction has decreased dramatically over the years. The minimum number of confirmations required by most exchanges has dropped from 6 to 3 to now 1 in some places. The problem isn't with me so much as the people who won't give me what I want w/o somewhere between 1-6 confirmations.

But also, when you're trying to demonstrate to somebody the miracle of bitcoin and you're both sitting there waiting for the confirmation to roll in as the final part of your presentation and 15 then 20 minutes go by and you're like what the fuck. Try explaining everything you just told me to somebody who doesn't know much about bitcoin but wants peace of mind that its impossible for me to stop the payment.

Ideally LN will someday be used to cover such transactions, but its still not the easiest idea to get onboard for most people.

I still think that we all think about this wrong.

We are waiting for confirmations to ensure a transaction is irreversible, right? The only payment method which is instantly irreversible is cash. The most common - credit cards - can be reversed for between 90 and 180 days after the payment is made. Similar with debit cards, checks, PayPal, bank transfers, etc. All the payee has to do is phone their bank or payment provider and dispute the payment, claim they were hacked/scammed/phished/whatever, claim the goods or services were not provided as promised, or a dozen other reasons, and they payment can be reversed.

A credit card transaction you see instantly, the money is available for the merchant to spend in 3 to 5 working days, and can be reversed for up to 180 days.
A bitcoin transaction you see instantly, the money is available for the merchant to spend instantly, and can be reversed (with much more difficulty) for up to an hour, maybe.

Now, obviously if you are transacting significant amounts of money then I would never dream of accepting zero confirmations, but if you are making a small payment (cup of coffee, fast food, etc.), then I think it is entirely reasonable for the merchant to accept a zero confirmation transaction, provided it pays a reasonable fee, is not flagged as RBF, and has no unconfirmed parents.

Unfortunately, this is completely true, and anyone who has been in that situation has certainly not felt comfortable - especially if they wanted to show someone how easy it is to use Bitcoin to pay, and then the transaction is not confirmed for more than 20+ minutes. I recently had such a situation, not only was the time between blocks more than 45 minutes, but for some reason, the seller asked for more than 2 confirmations which extended the whole thing to 1 hour.

On-chain payments will simply never be able to work in direct payments - so for now the only choice is LN.

The relative unpredictably of block times is one of its greatest weaknesses as a currency, in my experience. Not receiving a confirmation within ~20 minutes or so can be embarrassing for an in-person payment, and a 30+ wait can actually be costly -- for one party or both.

As far as emulating money is concerned, I think Litecoin is far more practical for payments and use it whenever I can. I used to like to place live bets on sports -- Dogecoin was ideal for that.

That's right, we can't really determine the time very accurately. We can only measure average time. We can't confirm the time with a 10 minute block.  It depends on the speed of the transaction. Sometimes it can be more sometimes it can be less.

The chance of a block not being found within a specific time frame is given by the equation:

e^(-x/600)

Where x is equal to the number of seconds. If we make x = 3600 for 1 hour, then the chance of a block not being found in 1 hour works out to just under 0.25%. Given that, we would expect to see a 1 hour or more block time every 400 blocks on average, or little bit more frequently than once every 3 days.

This of course assumes that the difficulty and hashrate are constant. If the hashrate has fallen since the last difficulty readjustment, then the chances of longer block times increases.

Also, please note that the timestamp is not the actual time. Changing the timestamp is another way to vary the block hash. The timestamp must be later than the median of the last eleven blocks (roughly in the last hour) and not more than two hours greater than the actual time.

Finding the next block is like a random game of chance or probability. In Simple terms, it's like if you are six players (miners) playing a die game and the first person to roll the dice to number 6 wins the game (Mine a new block)



You will find a situation where players will take so long to roll a number six after very many attempts, while at other times it will take any of them just seconds to roll a six.

It is why it is called an AVERAGE block time.  It would be impossible to set an exact time of 10 minutes per block because there is something miners have to solve and it may take more or less time depending on a few factors such as hashrate or difficulty.  If I gave you a math problem, it would be impossible in the same way to make you solve it in exactly 10 minutes.  The average time however is 10 min and you find out it is as soon as you start using Bitcoin often.

My transactions typically get one confirmation in at least 5 minutes and at most 15, but it is usually around 10 that it happens.  Should the hashrate increase or decrease substantially before the difficulty adjustment, block times change together with it.

-
Regards,
PrivacyG

this is a usual situation that can happen depending on the miners/pools, the longest block time that ever happened is 5 days and 8 hours from the block 1
the 3 longest blocks time from this year are :
block 689301   02h19m
block 679786   02h02m
block 687143   01h59m

The "10 minute" blocktime is the average interval between blocks which is managed by the mining difficulty, adjusted every two weeks (approximately) to keep the interval uniform.
It can be shorter than that or longer than that depending upon how fast the next cryptographic puzzle is solved.

It didn't happen for a hash to be found with the given conditions. I wouldn't say it's unusual; there's a 10 minutes interval on average between the blocks, but they may take much more than expected. They can also be found in much less time than expected. For example, block 699511 was found 3 minutes right after the previous one.

El Salvador doesn't have to do with this.

Today is "El Salvador day," so there's lots of news and lots of trading.

According to blockchain.info, block 699509 was timestamped 17:00 UTC and 699510 was timestamped 18:06 UTC -- a 66-minute interval. Isn't that quite unusual? If so, what might have been the cause?