Formerly an apolitical nerd minding my own business, got political after Bret Weinstein/Evergreen College, James Damore/Google, and BLM/defund the police.
I am just a layman. But I've spent a lot of time trying to understand this issue. Here's what I think is the most compelling technical pro-Core (increase limits on OP_RETURN) argument:
We should increase the limits on OP_RETURN from 80 bytes to L bytes. This is strictly an improvement over the previous situation, because following this change, any transactions that we see using L bytes in OP_RETURN could have also done the same prior to this change by using fake public key outputs to encode those L bytes. But at least now we're giving them the option of not polluting the UTXO space.
(I use "L" here to represent the maximum number of data that could be encoded by fake public key outputs)
And here are my two counterarguments to this, one more technical and one more contextual.
1. Encoding data in OP_RETURN is cheaper than encoding data using fake public key outputs. I relied on chatGPT for this analysis (see the table below). It claims that encoding 1kb of data using OP_RETURN would use 1036 vBytes, while the cheapest way of encoding the data using fake public key outputs would require 1364 vBytes. This is a significant cost difference, making it cheaper to embed data via OP_RETURN (and the pro-Core crowd loves to bring up incentives). So in fact, there could be "applications" that were formerly uneconomical but made economical with this change.
2. This argument assumes that we can't do anything to prevent embedding data in fake public key outputs in the first place: "we can't do anything about it anyways, let's just try to make the best out of a shitty situation".
Now, I understand that it is theoretically impossible to filter all spam. It doesn't follow that we should therefore do nothing. Here, I have to defer to people much smarter than me. For example, see this thread (which does concede that it is a cat-and-mouse game):
https://t.co/yDZICp5iCZ
Now, getting into the contextual argument:
I can't tell if the above approach will work, but I'm open to trying it, and I also know that Core has been in control while we watched the UTXO set grow from like 4GB to 12GB in less than two years. I'm sure there were all these arguments why nothing could be done. Luke and others thought otherwise (e.g. Luke's PR from 2023: https://t.co/enc97VBZHW). Core had its chance and now I want to give Luke a chance. I find @GrassFedBitcoin 's argument here compelling: they did nothing while UTXO size tripled, and now they suddenly want to make this change to the OP_RETURN limit that has been in place for almost a decade?
The pro-Core crowd in general rubs me the wrong way. I hate reading things like "Bitcoin is a database", which essentially tries to put all applications of Bitcoin on equal footing – as long as the transactor is willing to pay money, it's legit. No. Bitcoin's most important use by far is as sound money. Everything that dilutes this in a significant way can go to hell.
Even Core devs seem to have this "Bitcoin is a database" mindset. Here's a concrete example of that:
https://t.co/eerHFAxZT1
Hence I sense a fundamental philosophical disagreement with Core.
Also, all the gaslighting from the pro-Core side is super annoying, e.g. the constant strawman of "filters don't work (because they don't prevent spam from getting mined)" or "why don't you address it at the consensus level", which is obviously super impractical.
I'm ready to try something new.
Now, I do also see potential issues with Knots, but it's got nothing to do with the FUD spread by pro-Core people (e.g. the CEO of River alluding to Knots as a "poorly maintained forks of Bitcoin (with their own large flaws)" while later in the same post saying "the debate should remain objective and technical" along with a bunch of other misleading stuff [1]).
My main concern with Knots is if, in the dream scenario where the majority of nodes are Knots nodes, the network becomes too volatile/hostile to any development. What if some L2 absolutely needed a feature that Luke was unwilling to cede? For example, I'm running Knots right now on a Start9 server, and the default value for the OP_RETURN limit is 42 bytes. Core is 83 bytes. If I was a developer with an application that relied on 80 bytes of OP_RETURN, suddenly I'm getting "rugged", so to speak. Let's be real, this is messy. On the other hand, maybe the current situation with a relatively stable environment was too good to be true, given Bitcoin's distributed nature. Apparently, it was also too good to be true that Bitcoiners could just sit on their hands and watch number go up. That's probably like people expecting a democracy to work when the majority of citizens are uninformed.
Anyways, we're far from the point of Knots domination. In the meantime...running Knots.
[1] https://t.co/rqbardtiMY
At @River we will continue to run Bitcoin Core because it is the only properly maintained Bitcoin node software. The Bitcoin Core dev team is highly competent and dedicated to the long-term success of the network. The current state of the debate around OP_RETURN relay rules is the result of overdramatizing a nuanced technical decision regarding tx standardness rules. Running poorly maintained forks of Bitcoin (with their own large flaws) because of this minor technical debate is an overreaction IMO.
No Bitcoin Core dev wants the blockchain spammed with data because they are the ones forced to deal with the scaling issues and edge cases that follow. However, the reality today is that consensus rules allow transactions with large amounts of data in OP_RETURN, regardless of mempool policy! So, the broader debate here is twofold:
1. People want to put data onchain and they will do so with or without OP_RETURN limits. There is nothing stopping these people from skipping the mempool and submitting nonstandard txs directly to a mining pool today, or using even uglier approaches that bloat the UTXO set (even worse for the network). If we want to actually “fix” this, we would need a consensus fork, not a mempool rule change. The decision to change the standardness rules to allow these txs to be relayed seems reasonable and not a big deal.
2. The divergence between standardness rules and consensus rules: There are many transactions one could make that are valid by consensus rules that cannot be broadcast to the network today because they are “nonstandard”. This does not stop a miner from including the transactions in a block. High divergence between standardness rules and consensus rules leads to complexity in the Bitcoin Core software and causes problems for block relay efficiency when people start sending non-standard transactions direct to miners. Starting to move away from this divergence seems reasonable, but needs to be done carefully. We can likely never fully delete standardness rules but I understand the goal to reduce complexity here.
This debate should remain objective and technical. It’s being blown way out of proportion. Remember, the vast majority of Bitcoin Core devs deeply care about the long-term health of Bitcoin and have dedicated their careers to getting Bitcoin where it is today.
This "solution" is completely backwards. The people who should be penalized are the _designers_, not the players. The right way to handle this would be:
1. Apologize to the community. Examine current practices and implement measures (e.g. add a step in their testing/designing checklist) to prevent this from happening in the future.
2. For the long term health of the game, remove the gold from people who did this, and compensate them for it – say, something like one year's worth of WoW.
3. Take a closer look at who designed this encounter, and consider demoting or firing them.
So I slept on it. Woke up with the realization that Bitcoin is just not interesting to me. It used to be kind of interesting 5 years ago. Now it's just mostly boring. The things going on with AI is infinitely more interesting than Bitcoin.
@Coriolanus9000@AutismCapital Did you notice that the quoted study says "43.5% of athletes experienced incontinence with daily tasks"? It doesn't only happen when they are bracing.
@phinn888@sivhades See video below. Ignore differences in eye color and eyebrows, they are not carried over from the SD selections. HD Face Shape "Narrow" is the closest match, and if you compare the bridge of the nose, you'll see that they look completely different between models.
Did any big streamers cover how ugly the human HD models are? I had stupidly assumed the HD models were just more detailed renditions of the SD models.
They also reduced the amount of cleavage.
These are relatively minor changes but the fact that somebody at Blizzard thought these changes were necessary is highly annoying, and not something I want to support with my money.
As a long-time TSLA holder, I am firmly opposed to a merger with SpaceX.
We have been waiting patiently for years for FSD to materialize. And right when it's on the cusp of doing so, they want to dilute the profits into the SpaceX void? NO THANKS.
Interestingly, Grok was against the merger when I asked a few days ago, whereas now it supports it.