Should IP address transactions be depreciated entirely?
RHorning starts a discussion: Should IP address transactions be depreciated entirely?
RHorning, author of the "bring it on" post on WikiLeaks that Satoshi rejected

At 22:17 UTC on December 4, 2010, the BitcoinTalk user RHorning wrote a long post in a thread asking for WikiLeaks contact information. PayPal had suspended the WikiLeaks account the day before. RHorning came down in favor of encouraging WikiLeaks to put a Bitcoin donation address on its website, pointed out that he was posting under his real name and not a pseudonym, and ended the post as follows.
RRHorning (BitcoinTalk post, December 04, 2010, 10:17:44 PM UTC)Basically, bring it on. Let’s encourage Wikileaks to use Bitcoins and I’m willing to face any risk or fallout from that act.
He signed the post “Robert S. Horning”, with “Logan, Utah” on the line below. Ten hours and fifty minutes later Satoshi Nakamoto quoted that closing paragraph and replied.
SNsatoshi (BitcoinTalk post, December 05, 2010, 9:08:08 AM UTC)No, don’t “bring it on”.
The project needs to grow gradually so the software can be strengthened along the way.
The post that drew this reply is one of 462 that RHorning left on the forum. Four months earlier the same man had begun writing a specification of the protocol, and eight days later he started a thread about an attack by miners acting together.
RHorning registered on BitcoinTalk at 16:30 UTC on July 12, 2010. Four and a half hours later he had started his first thread. He had read the paper, the FAQ and the source code, and still did not understand how coins were generated. Why was the algorithm supposed to be so difficult? Would a machine built only to generate coins not have a huge advantage over an ordinary client? His own interest lay elsewhere: he wondered whether the same mechanism could run a distributed MMORPG.
Gavin Andresen answered 28 minutes later. Specialized hardware could be built, but if coins were generated faster the network would raise the difficulty.
GAgavinandresen (BitcoinTalk post, July 12, 2010, 9:30:21 PM UTC)If Bitcoins become even more wildly popular, then maybe people will think they can profit by building or renting specialized hardware. That would be a very good problem for Bitcoin to have.
RHorning’s reply was an estimate. He mentioned that his family had owned a gold mine in Oregon for a while, and wrote that with the engineers he knew he could put together a dedicated FPGA system for about a dozen dollars per processor and perhaps a couple of thousand dollars in all. Then he accepted Andresen’s explanation: even a concerted effort with dedicated hardware would eventually become unprofitable.
Five days later he compared generation to a lottery now that the difficulty had risen. A computer picks a random sequence, and wins if it happens to pick the right one. The odds on a Bitcoin block, he wrote, were for now worse than those of any ordinary lottery.
On July 25 RHorning could not register on the wiki. The confirmation e-mail had not arrived after four days. The next day he switched to an address at a different ISP and the registration finally went through.
Three days after that he started a thread titled “Bitcoin draft specification document (v. 0.01 draft)”. He had begun writing a formal specification of Bitcoin on the wiki. It would record how clients interact with one another, and in time hold everything needed to implement Bitcoin again from scratch. He had complained before that no such document existed, he wrote, and was now putting his money where his mouth was. For now he was working from the source code.
On August 8 he asked whether Bitcoin should adopt a bug tracking system such as Bugzilla. The development forum had its uses, but it was “a bit of a kludge” for dealing with specific issues. The post was written a week before the value overflow incident.
In November RHorning went on asking. On the 19th he wrote that he did not understand how fees worked. What stops a miner from charging a 95% fee? Is there a way to refuse a fee and wait for a cheaper miner? On the 25th, in a thread about bundling the block chain with the release, he replied that part of the problem was that SourceForge’s bandwidth was perceived as a free good.
On the 28th he misread the payout rule for cooperative mining and Satoshi corrected him. RHorning conceded the error. He added that he was still not entirely convinced that hashes computed for a block the pool failed to win should go unrewarded.
The post of December 4 begins as an answer to a joke by Hal Finney: if Bitcoin became known as the WikiLeaks currency and was attacked by governments all over the world, at least it would get its Wikipedia page back. An hour and a half later RHorning took the joke seriously. In that case Bitcoin would be on the front page of most newspapers.
From there the post turns to himself. Police had already come into his house without permission. He was connected to enough people politically that it would be noticed if something happened to him. Supporting WikiLeaks was the morally right thing to do, and he was not afraid of whatever the US government might do to him. An investigation of Bitcoin by governments was coming sooner or later, so why fight the inevitable? Bitcoin could not be killed as a project. His one concern, he wrote, was whether the protocol as it stood was robust enough.
Satoshi’s reply went straight to that concern. Bitcoin was a small beta community, and the heat WikiLeaks would bring could destroy it at that stage. WikiLeaks began accepting Bitcoin donations in June of the following year. For those six months, see WikiLeaks Begins Accepting Bitcoin Donations. For how this exchange fits into Satoshi’s withdrawal, see the page on Mike Hearn’s 2025 retrospective.
Five and a half hours before Satoshi replied, RHorning had already started another thread: should IP address transactions be deprecated entirely? Three protocol messages, checkorder, submitorder and reply, were used only for sending to an IP address. Encouraging other implementations to build a back door nobody used seemed a bad idea. Only someone working through the protocol’s messages one by one, in order to write them down, would ask that question.
On December 8 RHorning expressed a reservation about version 0.3.18. It was the first update he had not rushed to download. He regarded the new check, which rejected non-standard transactions, as a political change to the network, and wrote that he would like Satoshi to express his opinion.
Four days later, at 18:09 UTC on December 12, he started the thread “Mining cartel attack”. For now, anyone who puts CPU cycles into generating can create a block and have it recognized by the network. RHorning began by confirming that assumption, and went on.
RRHorning (BitcoinTalk post, December 12, 2010, 6:09:12 PM UTC)Instead, in a “mining cartel attack”, I’m proposing that a substantial number of “miners” who possess a substantial fraction of the computing power of the network, but not necessarily 50% of the network, could form into a cartel that would only recognize blocks generated by each other.
The cartel recognizes its members’ blocks and now and then rejects those of other miners. The rejected miners find mining no longer worth the effort and drop out, and the cartel’s share grows. RHorning wrote that he saw no easy way to stop it. Thirteen minutes later Satoshi made his last post on the forum, in a different thread. He did not post in the cartel thread.
It was ByteCoin who took the argument further. In a post on December 14 he described cartel miners keeping a newly found block among themselves, unpublished, and mining the next block on top of it. The same day he posted simulation results and wrote that a cartel with no preferential network access could be profitable with 33% of the network’s computing power. RHorning answered within the day. He agreed that blocks saved up in reserve could be released as needed, and turned to how such an attack might be detected: an unusually large number of competing chains, and block timestamps that were regularly out of order.
Withholding blocks so that honest miners waste their work was later named selfish mining. On November 5, 2013, after two years and ten months without a post, someone added a link to the thread. It pointed to “Bitcoin Is Broken”, an article dated the day before in which Eyal and Sirer introduced their selfish-mining paper.
The thread of January 7, 2011 is titled “End the Confusion - Organizing the Protocol development effort”. Besides the draft RHorning had begun six months earlier, four other efforts to document the protocol now existed: the bitcoin.it wiki, the notes of a Python reimplementation, Andresen’s notes, and a specification on GitHub. RHorning proposed making one of them “semi-official” for the moment, and then looked further ahead. The web has the W3C, and network protocols have the IETF. How should the Bitcoin protocol be governed? “This is not really a technical issue so much as a political issue,” he wrote.
Andresen answered about three hours later. The new wiki was the right spot, and he would put a pointer to it at the top of his own notes. Finney added that the rules a peer should follow ought to be documented as well as the on-the-wire protocol. Four days later RHorning thanked everyone and wrote that he would move the applicable parts of his draft to the new wiki. He had held back until then because he did not want to seem to be intruding on somebody else’s project.
The 2013 selfish-mining paper lists that wiki’s “Protocol specification” page as reference 5.
RHorning’s BitcoinTalk profile shows 462 posts. Almost all of them were written in the seven months from July 12, 2010 to February 12, 2011, and thirty are collected here. In February 2017 the same account posted five times in threads that advertised services and sold accounts. Those five posts are about something else entirely, and they are left out of this record.
RHorning’s posts are long, and most of them begin with a question. Why is generation difficult? Who sets the fees? When it turned out that the answers existed only in the source code, he began writing a specification. The “bring it on” post belongs to the same series. If an investigation by governments cannot be avoided, why fight it? That was the one question Satoshi answered with a no.
On December 8 RHorning wrote that he would like Satoshi to express his opinion. Satoshi’s last post on the forum came four days later. On January 7 RHorning was no longer asking Satoshi. He was asking the whole forum how the protocol should be governed.
32 entries / 10 threads
RHorning starts a discussion: Should IP address transactions be depreciated entirely?
RHorning starts a discussion: Mining cartel attack.
RHorning starts a discussion: End the Confusion - Organizing the Protocol development effort.
RHorning starts a discussion: Bitcoin Generation and Alternative Applications.
RHorning starts a discussion: Bitcoin draft specification document (v. 0.01 draft).
RHorning starts a discussion: Bugzilla & other bug tracking systems for Bitcoins.
Context post by RHorning in BitcoinTalk topic 1976. after msg25119, quotes Satoshi.
Context post by RHorning in BitcoinTalk topic 2162. after msg28302.
Context post by RHorning in BitcoinTalk topic 383. after msg3305.
Context post by RHorning in BitcoinTalk topic 43. after msg3565.
Context post by RHorning in BitcoinTalk topic 1850. before msg22952.
Quoted post by RHorning in BitcoinTalk topic 1790.
Context post by RHorning in BitcoinTalk topic 338. before msg6307.
Context post by RHorning in BitcoinTalk topic 1931. before msg24438.
Context post by RHorning in BitcoinTalk topic 1735. quoted by msg26999.
RHorning advocates for UI tools to adjust fee schedules and mining policies, noting that free transactions currently work but may not be sustainable long-term.
Mike Hearn ↔ Satoshi Nakamoto, Robert Horning
In a CoinGeek interview, Mike Hearn offers rare reflections on Satoshi's personality — observing he enjoyed experimenting but appeared to grow frustrated when evangelical personalities arrived.