Has anyone got any evidence that a spam attack actually happened?
I wasn't doing any analysis, so I can't throw any numbers at you, but what made me think about spam was a very unusual behavior seen on the blockchain.
There are some interesting charts available if somebody would like to analyse this.
On December 6 there was only ~50k transactions in the mempool, and it included no transactions with an over 200 satoshi fee. In less than 12 hours this number reached 220k unconfirmed transactions, so it got over 4 times higher within hours and included 30k transactions with a fee of over 200 sat/B. Then just 2 days later on December 10 it suddenly decreased by 130k transactions (over a half) and went back to 90k, with all those 200-300 sat/b transactions disappearing. Call it what you want, a spam or not a spam, a mixer activity or whatever, I have nothing to prove you wrong, but it looks suspicious at the very least.
I am not convinced that those 200sat/B transactions were spam. They seemed to be transactions by people desperate enough to pay whatever was required to have their transaction confirmed in reasonable time.
What I saw (but I don't have the charts) and found suspicious were gaps in
https://btc.com/stats/unconfirmed-tx fee charts. For instance, there would be a large number of 400sat/B transactions, then nothing, and then some say 160sat/B transactions. I coudn't have explained them at that time, thought they were some transactions by institutions moving large quantities of bitcoins and not caring what they pay as a fee. But I didn't investigate further.
It's not really fix since most cryptocurrency already "charge" more fees when there are more input/output since the transaction size got bigger. While i think the idea isn't bad, it could be troublesome for exchange/services who's doing batch transaction (unless they charge the user for the batch transaction). If the spammer simply want to spam network, there are other ways such as 100 transaction with 1 input & 2 output rather than 1 transaction with 1 input and 100 input
Also, we don't have to pay high fees for bitcoin transaction, usually $0.2 (or even less) are good enough these days.
The current situation is good, but it is only temporary, I am afraid. We should be looking for the permanent fix. The fix suggested by Litecoin would make it costly for a spammer in both scenarios: 100 transaction with 1 input & 2 outputs would be same as 1 transaction with 1 input and 200 outputs, since they would be charged per byte, but required minimal fee per output. The point is, the transaction not meeting the minimal fee requirements wouldn't be even accepted as valid.
With these transactions that are many small outputs in one huge transaction. Is there any real case where these would occur?
If there is then I think you have the answer right there as to why the change can't be made. The fees would be increased for people trying to use bitcoin in a fair and normal way.
Sure there is - for instance the exchanges may emmit
bulk transactions to realise many of their users' payments all at once. But ask yourself this, what is better, that the minimal fee grows proportionally to the number of outputs, for exchanges and for spammers too, or that the network is unusable completely AND the fee per
byte is at the level of hundreds time higher than the reasonable fee for
regular users - you and me.
Then it doesn't seem like a great strategy. There have to be better ways to do things that doesn't unfairly increase costs for some people and uses. Short term it might be a viable solution but a better solution would be to have a network that can even handle these spam attacks. Fortunately for now it seems Ver and his loyal troll brigade have crawled back under their rock under the bridge.
If we just do nothing than we must brace for the next spam wave when bitcoin becomes susceptible again (=when the number of people transacting increases). The fix suggested by Litecoin doesn't seem to be unfair for anybody. Rules are very clear and same for everybody.
That's not a fix, but a pure form of usage exclusion, which is not what Bitcoin is about.
I can't emphasise this point enough. We all want to prevent spam but centralised control of what can and cannot be transacted is the antithesis of what Bitcoin is for.
Has anyone got any evidence that a spam attack actually happened?
Back in July last year I saw transactions like this one:
https://btc.com/70aea3503f5a636ff244a86f6be70a3520ec344d0a4e1c8bad60983778e8b1911 input and 749 outputs of 1 sat each. That's clearly spam as the recipient can't spend the dust. Has anyone seen anything like this from last November to December?
I think you provided proof of spam answering your own question. In a scenario from OP (with minimal fee per output requirement), that transaction wouldn't be possibe because it wouldn't be meeting any reasonable minimal fee per output requirements. Even considering fee rate per byte it was very cheap, indead much cheaper than 1sat/B.
Anyway, it is clearly impossible to tell spam from real use case, as the discriminator would be
the intent which is clearly subjective and cannot be measured objectively. Even the transaction from your post might have been a transaction from some ARG or whatever, there might have been a real purpose to it. Although I doubt it.
