Oddbean new post about | logout

Notes by Luke Dashjr | export

 nostr:npub1lh273a4wpkup00stw8dzqjvvrqrfdrv2v3v4t8pynuezlfe5vjnsnaa9nk If you still have a #Hestia... 
 I'm not actively working on it these days, but my FreeAbode works fine 
 The OP_RETURN discussion is not new and dates back to 2014 when Bitcoin Core 0.9.0 was released with the OP_RETURN policy included which was intended to discourage more egregious forms of spam. At that time, 40 bytes was the default max datacarriersize limit across all node implementations; this was and still is sufficiently large for tying data to a transaction (32 bytes for a hash and 8 bytes for a unique identifier). Core subsequently increasing the default to 80 bytes was an entirely voluntary decision and in no way contradicts the design objective that OP_RETURN creates a provably-prunable output to minimise damage caused by data storage schemes, which have always been discouraged as abusive. There are also other good technical reasons which I have chosen to retain the lower default in Bitcoin Knots, and no justification for increasing it.

It is not my intention, nor that of my team at 
@OCEAN, to filter coinjoins. These present an innovative tool for increasing Bitcoin’s privacy and, when constructed properly, coinjoins can easily stay within the OP_RETURN limit (indeed, there is no reason for them to have *any* OP_RETURN data at all). I have some ideas on how to alleviate the recent issue where some coinjoin transactions were flagged as spam from Knots v25, and I am willing, with the full resources of my team, to work collaboratively on a solution in good faith.

Bitcoin does and always has allowed nodes to set filters based on multiple sets of criteria and Knots v25’s defaults are IMO what is best for Bitcoin at this time. Others may disagree and that is ok. They are free to (and should) run their own nodes - it is good for Bitcoin to have more people running nodes, including miners, and there should be a natural diversity in node policies. As was stated before, OCEAN is on a path to decentralization and very soon we are going to be in a position where hashers will be able to fully participate as miners and perform the intelligent parts of mining such as deciding which version of node software to run and what filters or other policies to apply to block template construction. 
 We're not censoring. They're just exceeding a limit that goes back a decade. We didn't do anything here 
 This wasn't ready at launch, and we publicly announced it when it was 
 You don't know what you're talking about 
 This is from 9 years ago with no context at all.  nostr:npub1qny3tkh0acurzla8x3zy4nhrjz5zd8l9sy9j... 
 It was a lie then too. Just because lies are old doesn't make them true 
 It was a lie then too. Just because lies are old doesn't make them true 
 Liar 
 And I've been right for a decade. 
 Does Coinjoin exceed standard 42 bytes?

Wonder why Samurai does… 🤔 
 Someone's lying to you

Core changed theirs to 80, but 40 is what the standard was.

If they don't want to comply, that's fine, but blaming me for their decision not to is dishonest. 
 Also note that having ANY extra data hurts your privacy, which seems to contradict the goals of coinjoins. Not really sure what Samourai is thinking here ... 
 42 is 40 + 2 opcodes (which are counted in the current versions)

If you think Core gets to just dictate things, get over your centralization mindset.

Knots has used 40/42 since 2013. Samourai are the ones who chose to exceed it. 
 80 isn't justifiable, and brc20 spam is at 45 
 Why bother? 
 MINING POOLS THAT CENSOR BITCOIN TRANSACTIONS ARE AN ATTACK ON BITCOIN. ANY SELF RESPECTING MINER... 
 No, this is a bug in Samourai's implementation.
Other conjoin implementations work fine afaik. 
 Nothing was false advertised  
 It's a bug if they expected it to work under the 40 byte spec.

If they knowingly exceeded it... Then what are they complaining for? That's on them 
 As far as I know, this is a Samourai only issue.
It doesn't make sense for coinjoins to dox themselves like this in the first place. 
 PSA: “Inscriptions” are exploiting a vulnerability in #Bitcoin Core to spam the blockchain. Bitcoin Core has, since 2013, allowed users to set a limit on the size of extra data in transactions they relay or mine (`-datacarriersize`). By obfuscating their data as program code, Inscriptions bypass this limit.

This bug was recently fixed in Bitcoin Knots v25.1. It took longer than usual due to my workflow being severely disrupted at the end of last year (v24 was skipped entirely).

Bitcoin Core is still vulnerable in the upcoming v26 release. I can only hope it will finally get fixed before v27 next year. 
 Drop in replacement 
 If you tell your node not to allow extra data beyond 40 bytes, and it then allows 40k, how is that ever not a bug? 
 Yes and no

By the byte, yes
But they also enable miners to make the blocks larger (which is a second bad effect) 
 Your node has limited impact by itself, but the more people who fix this, eventually yes 
 It's worth noting you still can configure the limit however you want in Knots. This fix just means it will actually honour your choice. 
 That's Bitcoin Core. It's only fixed in Knots so far 
 nostr:npub1qtvl2em0llpnnllffhat8zltugwwz97x79gfmxfz4qk52n6zpk3qq87dze
nostr:npub1lh273a4wpkup00st... 
 Yes, until there's 8 blocks of work done after it 
 TIDES is basically PPLNS as it was originally supposed to be, with a very long N.

PPLNS today has become something very different, so we picked a new name to avoid confusion 
 8 times the current difficulty, so each share should be on average paid in 8 blocks 
 If you were waiting for me to publish my new "master" OpenPGP key announced at #FutureOfBitcoinMining, it should now be available from keyserver.ubuntu.com along with signatures:
- master key signing codesigning key (used for Knots v25.1)
- master key signing security communication key (for email about security issues)
- codesigning key signing master key

Master key: 93CB4961F69A65082D4410802CBA8253089655C3
Codesigning key: 1A3E761F19D2CC7785C5502EA291A2C45D0C504A
Security comms key: FAC098FE8DF9975F902418813666E2B1782A18E1 
 There was a minor issue with one of the three main features that we promptly fixed. 🙄

The known-late feature (better spam filtering) was not advertised yet as it wasn't/isn't ready. 
 Congratulations to our early miners! Our first block today was a huge reward and we’re happy fo... 
 Currently it's Knots v23 with defaults except for blockmaxsize/weight (maxed out) and blockprioritysize (zero)

Testing and upgrading to v25.1 ASAP 
 We found a block already.

The generation TX wasn't correct - all the btc will go where they're s... 
 Belated. Need to get social media people an easier way to cross post to other platforms 
 nostr:npub1qtvl2em0llpnnllffhat8zltugwwz97x79gfmxfz4qk52n6zpk3qq87dze is giving out coinbase tx w... 
 Yes, it's a best practice after each new block

Otherwise miners could produce a stale block that detracts from the network security 
 We never custody, but you're right below the threshold wouldn't be paid yet.

However, a huge reward like this is going to push many more past that threshold  
 As announced at the #FutureOfBitcoinMining conference, I have relaunched my Eligius mining pool to help bring more decentralization to #Bitcoin mining.
I am calling it OCEAN @OCEAN
 https://www.prnewswire.com/news-releases/jack-dorsey-leads-seed-round-in-support-of-oceans-mission-to-decentralize-bitcoin-mining-globally---announces-launch-at-future-of-bitcoin-mining-conference-301999073.html?tc=eml_cleartime 
 Liar 
 At 300, it's about 11 days between blocks on average (likely a bit shorter in practice)

Normally a block reward is split across the past 8*difficulty work performed. But that much work hasn't been done yet, so early miners would get a higher payout for a while 
 The expected work to find 8 blocks 
 TBD but very likely 
 Only part right now. Hopefully the rest eventually 
 Currently 10 TBC (0.01048576 BTC)

Planning to make it configurable 
 nostr:npub1lh273a4wpkup00stw8dzqjvvrqrfdrv2v3v4t8pynuezlfe5vjnsnaa9nk you’re a fucking shitcoin... 
 OP is a lying troll apparently 
 nostr:nprofile1qqsq9k04vahllseell55m74n3047y88pzlr0z5yany32st29fapqmgspz3mhxue69uhhyetvv9ujuerpd4... 
 It's mostly a question of what's practical. Definitely don't CPU/GPU/NerdMiner 
 NerdMiners wouldn't have even been viable in 2009 
 
Looking forward to presenting at the #FutureOfBitcoinMining conference next week.
Even if you won't make it, I recommend joining via the livestream. 😉

nostr:nevent1qqsgy8mj8qwfcxcv9dswnkkx3530f32rj9knrzvv6wy7ku7lyj5mwtqpz3mhxue69uhhyetvv9ujuerpd46hxtnfdupzqqke74nkll7r88l7jn06kw97hcsuuytudu2snkvj92pdg485yrdzqvzqqqqqqyjvxusa 
 Thinking this through once again: #ordisrespector does not seem dangerous knowing that:
- bitcoin... 
 Only difference between Core and Knots fee estimates is Knots by default targets confirmation within 24 hours.

Looking at your current mempool for fee estimation is broken by design 
 #Bitcoin Knots 25.1.knots20231115 (finally) released! 🎉  

Be sure to _verify_ your download(s)!
New PGP key: 1A3E761F19D2CC7785C5502EA291A2C45D0C504A

https://bitcoinknots.org/?25.1?nostr 
 Nobody cares about BSD enough to add support for it 
 Yes, I'm not quite setup to update the PPAs yet - that might not be ready until v26 in a few weeks 
 Why? There's been open source playback code for it, for a long time 
 I wonder if the Vatican has the genealogical records from Jerusalem before it's destruction in 70... 
 Should anyone care? 🤨 
 Just signed up, So what's all of this, how does it works? Last such platform I used was weird lol... 
 What did you sign up for? 🤔 
 Apple *just* pushed out updates to all apple products (iOS, Apple Watch, Mac).

Details on vulner... 
 Sounds possibly useful for rooting?🙃 
 That's how it works when there's a security fix....🫤 
 No 
 Something big I'm working on for #Bitcoin also needs a full time #Lightning protocol developer.

Please reach out privately if you're interested. 😄

https://luke.dashjr.org/jobs/202309-Lightning-job-description.html 
 ? 
 Not working how? 
 Know what really sucks? Having a super low sex drive.
It's a lot harder than I reckon most people... 
 It does change things significantly.

Within marriage, you can't just walk away and find someone else.

(Divorce isn't real) 
 Liquid is multisig to begin with... The point is they're transparent with the security model unlike Ethereum which pretends it's something it really isn't in practice.

(Also, IIRC those backups aren't actually accessible by Blockstream without the functionary giving it to them - they're encrypted so the functionary can't use it alone: it functions as an effective 2-of-2) 
 have any of you ditched mobile phone for laptop only? 
 More inclined to do the opposite

I miss handheld/pocketable laptops 
Event not found
 They want people to buy more clothes? Why? 
Event not found
 No, you can ask the miners to give you mainchain coins, but there's no guarantee of any sort there 
Event not found
 No, a dispensation is required to marry heretics too - and even then, the marriage would still be invalid if the non-Catholic intends to convert the Catholic or their children to another religion. 
 They do, but it's a lie in their case 
 That's not correct. It states the disparity is an impediment, which would make it null unless dispensed. 
 Some of them do, but really the only thing Protestants all have in common is their denial of the papacy 
 Believing in baptism alone isn't sufficient either 
Event not found
 Birth certificates certify birth, nothing else. They are not agreements. 
Event not found
 Feels like there's more to the story they aren't telling you 
Event not found
 Don't forget to verify it got there with your own full node