1. You go into the airport and you are only allowed 1 bag that weighs a max of 100 pounds.  (1MB limit - current scheme)
2. You go into the airport and you are only allowed 1 bag that weighs a max of 200 pounds. (2MB hard fork - proposed scheme of classic)
3. You go into the airport and you are allowed 2 bags that weigh a combined 170-300 pounds depending the items (multisig txs) you carry (segwit scheme by core)

Why would you insist on whether it's one bag or two bags? Does it matter to you?

Similarly, segwit allows 1.7MB to 3MB+ of data. Just because it is in two separate data structures, doesn't mean it's less data or that it's ...still 1MB. It isn't.

If one file in your PC saves 1MB and the other saves another 700kb, it's still 1.7MB of data in txs (and not 1MB). And you also solve the malleability problem that is there for years.

The underlined part does not match real world experience of the last 20 years:

1995: Pentium 100 / 4-16MB RAM / 800mb - 2GB disks / dialup and ISDN connections at speeds like XX kbps
2015: multicore i7s / 4-16GB RAM / 1TB-4TB disks / home connections in XX mbps+.

We are seeing 1000x in processing, ram, storage, internet - and we are really getting those incrementally, not as radical breakthroughs.

If the software can currently do something like 10tx/s (with 2MB blocks + segwit), then with 1000x in hardware (and no change in software), just on the current trendline of tech progression, we'll be seeing 10.000 tx/sec by 2035 and hundreds of thousands of tx/sec further ahead.

If you add software improvements, breakthrough advances (?), or tapping new resources (like GPU processing which is already >10x the CPU processing power) for General Purpose computing tasks like validation, real time compression etc), we might get there far earlier.

QC-resistance may be one of the factors that regresses scaling if it employs a bloated scheme, otherwise it's a certainty that massive scaling *will* happen. Just not now - it'll take time.

Perhaps at some point, we will see a new type of "trustless" and "decentralized" model, in the form of Artificial Intelligence.

The main issue with trust and centralization in currencies or transactions is that you don't know what the trusted party might do. A currency issuer might issue another 100% of a currency. An intermediate payment service can withhold money, reverse txs etc.

So you have these simple algorithms in place right now, and the blockchain which is essentially a consensus / ledger where everyone knows what everyone else did, and agrees with it as a historic record. The network also has certain parameters about what the money is, how it is distributed, what is the monetary base - stuff like that which are agreed upon.

But what if we replaced human-operated third parties (banks, paypals, cc's) with an Artificial Intelligence persona, that only operates under a certain window of parameters. In other words the AI is *trusted* to move money from person A to person B, and not keep it, because that is its design function and has no motive to do something else. The AI would be the (trusted) "robotic" manager of every single tx.

This model would probably combine several aspects of efficiency from the centralized/hierarchical model, with those aspects that make simple algorithms more trustworthy than human-operated systems.

It is simultaneously trusted (because you have to trust the AI) and trustless (because you don't expect the AI to fuck you over due to its design parameters and predetermined role). It is simultaneously centralized (as the AI does all the validation of TXs, instead of a multitude of people mining or staking) and decentralized (as the AI "lives" computationally in a P2P type cloud).

The political environment is weird and volatile right now. Every action could be used for flaming - either way an argument goes. For example:

If, say, devs make some changes that lead to increased fees "ohhh they want to exclude the small guy..."
If, say, devs make some changes that lead to fee reductions "ohhh they are fucking over the miners and the long-term security of the network"

If, say, miners process junk txs with near-zero fees, "ohhh they are including all the spam"
If, say, miners don't process junk txs, "we must take action to FORCE miners to mine txs, they can't just mine nearly empty blocks because they don't like the small fees paid"



Long-term goal is for bitcoin to be able to handle every microtx.

Short-term, while hardware, network and software aren't up there yet, is to "manage" the situation, even by excluding small micro-txs as uneconomical to be processed. This is not something that "Core" started. It was started by Satoshi himself.

Bumping the blocksize up with current software, networks, storage, cpu etc etc, is hitting a wall after a few tens of txs per second - it doesn't work for a massive payment system, and definitely not for microtxs.

It needs much better networks/storage/processing power, or it needs much more efficient software (or a combination of both) to get there for microtxs.

In theory the network is agnostic whether you transfer 0.01 btc or 10 btc. But in the real world, when a bank or paypal charges you 20$ for a 300$ transfer, it's best if you take this through BTC - and you keep the 20$.

...and when you have, say, 30 txs per second in BTC as max capacity, you'd better be utilizing them to the best of your ability to maximize your savings versus the banking system, by saving serious money that would be paid in banking fees (10-20-100$ in traditional payment method fees) rather than wasting these 30 txs/sec with microtxs, dice, etc. Besides, there are altcoins that would be far cheaper for this type of use. When the system is able to do 300-3000-30.000 tx/s, then even microtxs will be quite viable.

Near zero-cost txs = near zero-cost abuse.

When you have systems that are highly inefficient and thus vulnerable to abuse in order to achieve being trustless and decentralized, then you have to protect these systems from abuse through economic disincentives.

This is not only relative to bitcoin: Altcoins suffer the same fate. There is no alt that can have near-zero or zero fees and not have an open attack vector of abuse.

However, what is now problematic, may not be in the future as technology progresses.

From 1995 to 2015 we went 1000x in networks, storage, cpu etc. Till 2036 we could see a similar growth pattern of 100x-1000x-10.000x - who knows.

https://bitcoinfees.21.co/

Which fee should I use?

The fastest and cheapest transaction fee is currently 40 satoshis/byte, shown in green at the top.
For the median transaction size of 258 bytes, this results in a fee of 10,320 satoshis.

....
Fees are at just 0.04 USD for urgent first block inclusion. And you can go down to 0.01-0.02 USD if you are willing to be included in the next 2-3 blocks instead of the next block.

So there are tonz of crap added every second that evidently aren't in any rush to get quickly validated. This means that whether 100KB, 1 MB or 10MB, it doesn't matter. A spammer can always fill them with junk as long as there are willing miners to process near zero cost txs.

With the fees for first block inclusion being at 0.04 USD, and +2/3 blocks at 0.01-0.02$, we have a near-zero-cost tx situation which allows near zero cost spamming / bloating etc, especially for low-urgency "transactions".

We are talking about fees, not subsidy.


As far as I'm concerned, the main problem is mining centralization through pools, not where these pools are located. Still, the protocol is vastly most decentralized compared to a bank that confiscates your money, delays your transfer by a week, etc etc. Same for the token and its issuance, relative to govt issued fiat currency.

This is not a case of freedom, it's a case of using an inefficient technology and bloating it for the lolz, at practically zero cost - which shouldn't be happening.

All abuse scenarios will typically play out if people are "free" to play them out. If you want to have a robust and resilient system, you can't allow the "Freedom" to abuse and kill the system. Was Satoshi a fool for implementing a 1MB restriction instead of allowing the "freedom" to have 10 or 20mb of txs per block? Of course not. Was Satoshi a fool for having a fee system in place? Wouldn't it be more "popular" if he said "let's do the freeTXs4all, 4evah"? Wouldn't it be more "popular" if he said "every microtx in the planet should go through BTC" instead of saying "eh, well, we are not very suited for microtxs right now"?... Wouldn't it be more "popular" if he said that the BTC blockchain can store all kind of data structures in it, instead of requiring different blockchains for different applications and admitting that one-blockchain-for-all-uses doesn't scale?


The other day I received 250$ in paypal. I hadn't used paypal in a while and, despite having mentioned what they charge quite often / knowing what they charge in theory, I *still* got blown away by the charges.

I got charged something like 10.5$ in fees for receiving the money.

I got charged ~6$ for converting to euros

I got charged ~3.5$ by my bank for the incoming wire (outgoing wire by paypal was "Free")

Of the 250$, I lost 20$ in paypal fees, conversions and bank wire. The sender of the money also lost some in currency conversion to $$$, fees, etc.

Btw, after I pressed the withdraw button in paypal, my withdrawal was put under "review" for "my security". This would delay the money to my bank account by +1 working days.

This is the bullshit reality of our money transfer industry. Money taking ages to get transferred, having third parties controlling your money, charging you insane amounts for fees, etc etc. There is a huge market for BTC whether at 0.03 USD, 0.3 USD or 3 USD fee. I'm not saying we need 3 USD fees, I'm just saying that there are A LOT of payment scenarios which even if you pay 3 USD for BTC fees, you'll still do it waaaaaay cheaper than through banks, paypal etc etc - so there is a market even if the fees were 100x or 200x what they currently are.



The fallacy of this line of thinking is twofold:

1) It presupposes we have 1mb forevah, when this is not the case

2) It presupposes a 0/1 binary thinking where you go from being able to transact in an affordable manner, to not being able to transact due to very high fees. It's not like 1 cent or 100 USD in fees as our only 2 options. It could be 10 cents, 20 cents, 50 cents, 1 USD etc. This is not an on/off switch. It's a potentiometer / an analog progressive switch.


If users are going to be put off by, say, 10-20 cents in fees, then perhaps they can buy plenty of lube and go get fucked by the banks and the paypals instead. Having 1-2-3 cents in fees and complaining that it's the end of the world if we go upwards and that it's crippling adoption, when RL banking payments are charging the hell out of us (and they are being used everyday by billions of people) and we get fucked by them constantly, is somewhat hypocritic.

Besides the main selling point of BTC shouldn't be an addiction to near-zero cost txs (which we know that is not sustainable in the long run), but rather its decentralized nature and its advantages over central banks (as a token/currency) and commercial banks in terms of controlling your money, paying without trusted 3rd parties which could withhold or confiscate your money, the non reversibility of (confirmed) txs etc etc.

It was a known fact that as we go forward subsidy reduction will be compensated by increased tx fee revenue for the miners. We are nearing the second halving and we are at around 3/4ths of the monetary base already being distributed - leaving just 1/4 left. So where exactly is the increased fee revenue?

What are the fees? 1-2-3 cents? Are we serious? And you are blaming "cripplecore" etc etc, like if it was 2mb then something would change (fees would actually be lower because first block inclusion would be assured even at 10 satoshi per byte).

And forkers went out and said "ohhh we can't have a fee market, we must go to 8MB and 20MB urgently because the rising fees are unacceptable, they will kill bitcoin adoption" etc etc. They pretended all txs <1MB are all legit and that the lack of more space is preventing ...scaling real btc txs, instead of simply crowding out the spam.

If you make a precedent in favoring near-zero cost txs through hard forks, because you somehow have an ideological opposition to one of bitcoin's first principles (that as we go forward fees per block will rise) then you can make it again by increasing inflation instead of fees. You used to say that you don't want to pay fees because you bought your right to transact when you bought your coins and now you are saying "ohhh with lightning there will be less fees for the miners etc etc". It seems you spin it any way you like just to stir stuff.

By the way, in your next messages you are attacking a model that doesn't exist: The 1MBforevah. Core certainly doesn't support 1MBforevah or 3tx/s forevah. So you are projecting an irrationality to then deconstruct it but that irrationality doesn't really exist.

It's arbitrary really. Satoshi actually gave more room than necessary, but then again the fees paid back then were like 0.01.

An argument like that will never be presented "we want more coins and high inflation for the lolz".

It will be presented as "we want cheap txs".... Arguments like "People want cheap txs, who are you to stop that".... Populist bullshit like "bigger blocks" and the "dangers" of "fullblockalypse".

A forker can troll us with an argument of the following style:

"High tx fees reduce adoption, so we need low tx fees to increase adoption. If that doesn't happen, BTC is dead. So it's preferable to have more than 21mn coins through constant inflation rather than raising tx fees".

They could say "you are crippling BTC's potential through those high fees, we need much lower fees and in order to do that we need subsidy / more coins".

If people insist on having low-fee txs, then the end-game is a ...DOGEcoin (infinite inflation).

Higher fees solve both the inflation issue and also address the sane block utilization issue.

Fees at 1 to 3 cents are ridiculous anyway.

"We would like to say thank you to @ShenNoether for finding the flaw, and the bounty will be paid in SDC at the SDC price before the flaw was revealed to the public. This is the reason we set up the bounty program in the first place, to improve shadow's privacy, usability, etc."

So if the price was 0.00025 and is now 0.00014, then he gets paid like 20.000 SDC @ 0.00025 = 5 BTC which if dumped now, cost 20k SDC x 0.00014 (which it might go down to 0.00010 due to the dumping) = 2.8 to 2 BTC actual reward.

Thats minus -2.2 to -3 BTC for the reward.

Now, I understand that some level of responsibility must be demonstrated in these things, but on the other hand if such information is first received by a core team which may also be SDC holders then this could result in SDC dumps by the core-team holders first to preempt the public investor dumping. In a sense the bounty, in that case, is the cost of an inside info on when to dump.

Plus if a bug finder has to lose money if he reports a bug, then that's a disincentive to report it and more incentive to sell it in a black market... "hey, SDC deanonymization key only 10 BTC for the whole chain". That should not be the first option of a bug finder.

If you had 10 MB blocks tomorrow morning, you'd still need 0.8$ in fees, per tx, to replenish lost subsidy income.

People are currently paying 0$ to 0.03$ (for first block inclusion) while "blocks are full", under the current 1 MB scheme. With the current fees, even with 100 MB blocks, and them being full, you'd still be unable to pay the subsidy loss (you'd need 0.08$ per tx). So you'd need something like 250MB blocks with the current fees. This makes it pretty obvious and simple: Fees have to rise.

If fees don't rise, then miners bail out and the network loses security. OR, some new "forkers" will come and say something retarded like "we want free txs forevah, so let us inflate the money quantity instead... why limit btc to 21mn coins? That's ...crippling to BTC... yeah, let's uncripple it by issuing a few hundred million more... we don't want to pay high tx fees you know... we prefer inflation".

The size of the transaction is not relevant to the btcs you are sending. You could be sending 0.01 btc or 10 btc and the tx size could be the same.

However if you had an address that has its 10 btcs as a result of gathering, say, 1000 donations of 0.01 btc there, then while trying to send these 10 btcs you would essentially be transferring 1000 x 0.01 btc, so at that point you'd be looking at a pretty large tx in terms of bytes/kilobytes.

Sure: https://bitcoinfees.21.co/

Which fee should I use?

The fastest and cheapest transaction fee is currently 30 satoshis/byte, shown in green at the top.
For the median transaction size of 327 bytes, this results in a fee of 9,810 satoshis.

9810 Satoshi / 100.000.000 x 398$ = 3.9 cents for a 327 byte tx and 2.98 cents for a 250 byte tx. Again that's for first block inclusion - calculated with 30 satoshi per byte. And there are others paying 1-10 or 11-20 etc (you can see the breakdown).


Edit: Uploading photo... right now fees for first block are even lower at 20 satoshi/byte:

The fees currently paid, even for first block inclusion while 'blocks are full' are near zero.

Urgent first block inclusion right now costs 0.03 USD and if you want to go +1-3 blocks, you can drop that to ~1cent. Spamming the blockchain is, for all practical intents and purposes, a very low cost activity.

https://bitcoinfees.21.co/

The fastest and cheapest transaction fee is currently 30 satoshis/byte, shown in green at the top.
For the median transaction size of 306 bytes, this results in a fee of 9,180 satoshis.

9180 satoshis = 0.03 USD.

3 cents for "urgent" first block inclusion = nearly zero fees.

Or implementation problem. I mean BTC can't just turn into a masternode / mixed-mining rewards structure as it is already established.

A smaller developing coin can experiment in ways that a large established coin cannot because the economy is more "fixed" in the later. And that's not necessarily good or bad, it's just the way it is.

If someone wanted to wreck havoc*, wouldn't it be better to just make a list of all the unmasked transactions and upload it somewhere, saying "SDC deanonymized", without even leaving a clue as to how it was done - with the vulnerabilities speculation in the air being more corrosive than the known-bug situation as it is right now?

* Not that the monero people aren't enjoying this, sure they do.