Library
Reading rooms
Rooms for the 2015 P2PKH catalog: the discovery thread and funding transaction, then cases, method, public papers and tools, and a first-sitting path.
Primary record
The 2015 catalog was noticed in public on BitcoinTalk. Authorship is a forum statement. Funding is a transaction. Read those first, then the atlas pages that cite them.
BitcoinTalk discovery thread
Thread #1306983 is the public discovery record, opened by Bulista. The opening-post display is 28 December 2015, 04:12:47 PM on the forum clock; treat calendar-day claims as unverified until timezone is documented. Precursor thread #1305887 is related, not a replacement.
saatoshi_rising authorship post
On 27 April 2017 the BitcoinTalk account saatoshi_rising wrote that they were the creator and described the series as a measuring instrument. Thread #5201811 quotes that post. A forum claim is not a cryptographic proof of identity.
Genesis funding transaction
The 160-address set sits inside a larger 15 January 2015 funding. Explorer read 2026-09-11: transaction in block 339085 at 2015-01-15 18:07:14 UTC, 256 outputs, 32.896 BTC. That is the on-chain object this catalog maps. It is not a later OP_RETURN story.
Cases and people
Famous spends first. Finder names only with a source label. This atlas lists handles, not legal identities, and it is not a ranking table.
Famous spends
#66, #69, #120, #125, #130, and #135. On-chain facts stay on the cases page. On-chain facts, published scalars, and rumor are split there on purpose.
People index
saatoshi_rising, Bulista, amaclin, RetiredCoder, Kowala. Each row is labeled Forum statement, Self-attribution, or Community report. A handle is not an audited identity.
Method in prose
Why an address-only target is a HASH160 scan, and why a known compressed public key plus a bounded interval is a discrete-log search. This room is reading, not a calculator.
Two search classes
Address-only targets have no usable public key in the usual leak series, so search is over HASH160 of candidate P2PKH encodings. Kangaroo and BSGS need a compressed public key P and an interval. #71 is the first class. #140 is the second. Higher address-only ranges are not practical as brute force.
How a P2PKH spend publishes the public key
In a legacy P2PKH spend, scriptSig carries the signature and the public key as separate items. That is a publication channel, not the only possible one. An unsolved puzzle may already have a public key. It is also the claim-race surface: observers see P before confirmation.
The first keys as a lesson
Published scalars #1=1, #2=3, #3=7, #4=8 look like a staircase. Then the staircase dies. That is not a followable pattern for later ranges, and it is not a method for #71.
Glossary
N, window, containment, HASH160, WIF, P2PKH, kangaroo, BSGS, scriptSig.
Catalog briefings
Two folios and the scale page repay a first sitting. Open the live catalog for status. Do not freeze a solved-count in prose.
#71 briefing
Lowest address-only target in this catalog. The interval [270, 271 - 1] contains 270 scalars. No public key from the 2019 every-fifth leak. Lottery, not inventory.
#140 briefing
Primary kangaroo-class briefing. Interval width 2139. A 96-bit window contains the key with probability 2−43, about 1 in 8.80e12 (short scale). Reaching a kangaroo work budget does not prove every key in the window was enumerated.
Work scale
Unsolved bit lengths on a log2 axis. Each increment doubles the candidate interval. #72 is not in the same world as #160.
2019 leak map
16 May 2019: 21 outputs of 1,000 sat in block 576251, puzzle #1 plus the 20 every-fifth addresses #65-#160. A receive does not leak the public key. 1 June 2019: 21-input spend in block 578732. Those 20 leak addresses plus extra input 1PvaqLqRAivje7CactLR55xQBYvBeaDrXN, which is not #1 and not in the 2015 256-output set. Remaining leak-series kangaroo-class in the 14 Sep 2026 catalog re-check: #140, #145, #150, #155, #160.
Literature and public tools
Papers and public repositories this atlas can name with a URL. Implementation constants stay attributed to their README. Private forks are not listed.
Interval discrete logarithms
Galbraith, Pollard, and Ruprai, Computing Discrete Logarithms in an Interval, IACR ePrint 2010/617. The two-kangaroo interval method has heuristic average-case running time (2+o(1)) sqrt(N) group operations. The paper also analyses faster variants. Cite the paper for the 2*sqrt(N) heuristic. Do not import an unnamed classical Pollard-rho constant as if it were the interval cost.
RCKangaroo
Public GPU kangaroo for secp256k1 interval DLP, GPL-3.0. The README documents a -max stop bound of 1.15 * sqrt(range). That is an implementation constant, not a textbook law.
BitCrack
Public CUDA/OpenCL HASH160-range scanner. The README states its main purpose is contributing to the Bitcoin puzzle transaction. MIT license. A tool, not a map, and not this atlas.
Keyhunt
Public CPU hunter for secp256k1 with address, RIPEMD-160, and BSGS modes. MIT license. A tool, not a catalog. A BitcoinTalk announcement permalink was not verified for this card.
Orientation
If you arrived from BitcoinTalk and do not want myths first, start here. Other objects that share the word puzzle live on their own page.
Start here
A seven-step reading path: what the 2015 object is, which records to open, how the two search classes differ, and where claim risk lives. No engine onboarding.
This piece versus others
Genesis-block OP_RETURN stories, brainwallets, art contests, collision pools, and vanity prefixes are different objects. The other-puzzles page keeps that boundary honest, including gaps.
Sources
Threads and transactions this atlas actually cites. Verify on an explorer. A key dump is a dump, not a source label.
Global search
Puzzle number, address, theory, myth, glossary term, person, case. Search indexes public prose, not secret scalars.
One-page sheet
Census tables, #71, #140, and the claim warning, kept at hand. Not marketing.
Practice challenges
Thirteen small drills with worked answers: ranges, containment, units, the 2019 leak, and what is not a method.
Quick answers
Short honest answers to the questions every newcomer asks, including whether any of this is a good bet.
About this atlas
What the atlas is and is not, who it is for, and where its figures come from.
Full catalog
All 160 puzzle records in one filterable index, with exact intervals and evidence badges.