This BIP proposes redefining the bitcoin unit of account to represent the protocol’s smallest indivisible unit as “1 bitcoin,” eliminating the need for decimal-based UI conventions.
Authored by: John Carvalho (@BitcoinErrorLog)
Mailing list:
https://groups.google.com/g/bitcoindev/c/YwF-djZi1Bo/m/nIkyuClEAgAJ
0@@ -0,0 +1,149 @@
1+<pre>
2+BIP: 9999
3+Title: Redefinition of the Bitcoin Unit to the Base Denomination
4+Author: John Carvalho <bitcoinerrorlog@gmail.com>
5+Status: Draft
6+Type: Informational
7+Created: 2024-12-10
8+License: CC0-1.0
9+</pre>
Regarding the number assignment, one of the tests will always fail until a number has been assigned and it’s updated in both the table and the BIP. Don’t worry about passing the tests right now.
We usually have the preamble indented, and I added the Comments-URI that is required by BIP 2. The Created date is the date a number was assigned, so we will figure that out later.
0<pre>
1 BIP: ?
2 Title: Redefinition of the Bitcoin Unit to the Base Denomination
3 Author: John Carvalho <bitcoinerrorlog@gmail.com>
4 Comments-URI: https://github.com/bitcoin/bips/wiki/Comments:BIP-?
5 Status: Draft
6 Type: Informational
7 Created: ?
8 License: CC0-1.0
9</pre>
80+'''Perception of Supply:'''
81+While the total count of base units is roughly 2.1 quadrillion, this proposal does not alter supply in any way. The change is purely representational. Comparisons can be drawn to other currencies like the Japanese yen or Indonesian rupiah, where high unit counts are standard and not perceived as inflationary.
82+
83+===Addressing Alternative Approaches===
84+
85+== Refuting the "Bits" Proposal ([https://github.com/bitcoin/bips/blob/master/bip-0176.mediawiki BIP 176])
It will actually be easier to support the established convention of ₿ 1.000… in accounting software spreadsheets etc… One simply needs to repurpose metric unit conversion logic to implement Bitcoin accounting idioms. Allowing app developers to repurpose metric unit conversion code will actually ease adoption friction more than this proposal or the “bits” proposal. In the long run - languages/words evolve! - this whole debate seems to arise out of people’s need to control the words other people use more than anything really useful.
Imagine renaming a “penny” to a “dollar” and then trying to force everyone to adopt the new “jargon”. It is as absurd as it sounds!
NAK
Also, the file currently doesn’t render for me in the GitHub preview.—It did render when you had it as a markdown file, but since it was changed to a a mediawiki file it doesn’t. I’m not sure, yet, why. It might be related to the missing indentation in the preformatted text block.
Apologies, I was working in md and assumed I needed to submit as mediawiki, which I am unfamiliar with. Happy to adjust any formatting to be correct… with a little help.
How isn’t this still decimal?
One of the most prominent and rational feedback to the original version of this proposal was to retain stock ticker denominations. This seems rational to me, where we have BTC and XBT, or the like. Maybe this relegates the “21M” meme and culture to the finance world. I don’t know.
But otherwise, I hope to simply specify how people could display Bitcoin units as “bitcoin” and thus avoid bad thinking due to ignorance about how Bitcoin uses integers.
There are a bunch of side benefits to this, but most feedback was otherwise highly subjective or speculative.
What a terrible idea of redefining a widely used term by the different meaning.
Thank you for your feedback.
Please consider that I am merely suggesting we call the only units in Bitcoin “bitcoins” - as well as the other rationale in the document.
It will actually be easier to support the established convention of ₿ 1.000… in accounting software spreadsheets etc… One simply needs to repurpose metric unit conversion logic to implement Bitcoin accounting idioms. Allowing app developers to repurpose metric unit conversion code will actually ease adoption friction more than this proposal or the “bits” proposal. In the long run - languages/words evolve! - this whole debate seems to arise out of people’s need to control the words other people use more than anything really useful.
Imagine renaming a “penny” to a “dollar” and then trying to force everyone to adopt the new “jargon”. It is as absurd as it sounds!
NAK
Thank you for your misc opinions, but, in the end, I could only consider feedback that maintains the core intent of the proposal.
79+Adopting the smallest unit as the primary measure ensures a consistent standard that can scale smoothly as Bitcoin adoption grows.
80+
81+'''Perception of Supply:'''
82+While the total count of base units is roughly 2.1 quadrillion, this proposal does not alter supply in any way. The change is purely representational. Comparisons can be drawn to other currencies like the Japanese yen or Indonesian rupiah, where high unit counts are standard and not perceived as inflationary.
83+
84+===Addressing Alternative Approaches===
How is this proposal different to the already established “sats” denomination?
Also, doesn’t the LN already use “fractional satoshis”? If so, the integer-only argument wouldn’t be valid.
Someone explain to me how we moved from there are multiple ways to increase precision to sat is indivisible in less than 3 years?
In my opinion, the proposals you linked could all be classified as hard forks, and/or, hacks to support a new subordinate UoA, and/or, entirely fictional hugely controversial consensus changes.
This BIP on the other hand is generally harmless.
How is this proposal different to the already established “sats” denomination?
Also, doesn’t the LN already use “fractional satoshis”? If so, the integer-only argument wouldn’t be valid.
I will re-read the proposal with eye to make sure any “sats” dynamics are clearly addressed. Feel free to contribute!
Frankly, I don’t see any difference at the moment – other than the name. That’s why I was asking for including this discussion into the “alternatives”.
LN “millisats” do not “exist” exactly, they are just a trusted intra-channel credit feature. Mostly they just confuse people to once again misunderstand how Bitcoin (and Lightning) works. Should I add specs to remove millisats as concept to this BIP too?
I take it the definition in this BIP is about Bitcoin the currency (as opposed to Bitcoin the payment system). As such, I think this BIP should consider all existing uses of denominations. There is no point in renaming satoshi to bitcoin just for L1, but not for L2. You would also want it to be used in stock quotes and charts, for example.
Frankly, I don’t see any difference at the moment – other than the name. That’s why I was asking for including this discussion into the “alternatives”.
I retained the BTC (XBT, etc) denomination only at the request of the first round of feedback, because I could rationalize stock tickers being a separate system from Bitcoin. I’d be happy to re-remove this concept if the peers here all agree it corrupts the purpose too much.
I take it the definition in this BIP is about Bitcoin the currency (as opposed to Bitcoin the payment system). As such, I think this BIP should consider all existing uses of denominations. There is no point in renaming satoshi to bitcoin just for L1, but not for L2. You would also want it to be used in stock quotes and charts, for example.
I suppose the LN community considers their BLiPs and BOLTs and such to be the place to spec for that protocol, or they could otherwise umbrella under BIPs. This BIP, as all, is just a web page in the end, and anyone is welcome to refer to this BIP to replicate this method.
If everyone likes, I could take the time to specify further how people normally display, abbreviate, and prefix large numbers and denominations of units. I was hoping my example of “2.1M” was enough of a nod…
Also, the file currently doesn’t render for me in the GitHub preview.—It did render when you had it as a markdown file, but since it was changed to a a mediawiki file it doesn’t. I’m not sure, yet, why. It might be related to the missing indentation in the preformatted text block.
Apologies, I was working in md and assumed I needed to submit as mediawiki, which I am unfamiliar with. Happy to adjust any formatting to be correct… with a little help.
Either Markdown or Mediawiki are acceptable since #1504 got merged in 2023.
42+* Use of the ₿ symbol MAY be used to represent base-unit bitcoins but is OPTIONAL.
43+
44+Example 1:
45+
46+ * Old display: <code>0.00010000 Bitcoin<code>
47+ * New display: <code>₿10,000<code> or <code>10,000 bitcoins<code> or <code>0.00010000 BTC<code>
0 * Old display: <code>0.00010000 Bitcoin</code>
1 * New display: <code>₿10,000</code> or <code>10,000 bitcoins</code> or <code>0.00010000 BTC</code>
47+ * New display: <code>₿10,000<code> or <code>10,000 bitcoins<code> or <code>0.00010000 BTC<code>
48+
49+Example 2:
50+
51+ * Old display: <code>10.23486 Bitcoin<code>
52+ * New display: <code>₿1,023,486,000<code> or <code>1,023,486,000 bitcoins<code> or <code>10.23486 BTC<code>
0 * Old display: <code>10.23486 Bitcoin</code>
1 * New display: <code>₿1,023,486,000</code> or <code>1,023,486,000 bitcoins</code> or <code>10.23486 BTC</code>
52+ * New display: <code>₿1,023,486,000<code> or <code>1,023,486,000 bitcoins<code> or <code>10.23486 BTC<code>
53+
54+Example 3:
55+
56+ * Old display: <code>0.345 BTC<code>
57+ * New display: No changes required or <code>₿34,500,000<code> or <code>34,500,000 bitcoins<code>
0 * Old display: <code>0.345 BTC</code>
1 * New display: No changes required or <code>₿34,500,000</code> or <code>34,500,000 bitcoins</code>
59+NOTE: Traditional number display abbreviations, like <code>2.5M<code> for millions, are also optional.
60+
61+'''Conversion:'''
62+
63+* Ledger and consensus rules remain unchanged.
64+* <code>BTC<code> as a currency code remains unchanged (1 BTC = 100,000,000 base units)
0* <code>BTC</code> as a currency code remains unchanged (1 BTC = 100,000,000 base units)
105+===Backward Compatibility===
106+
107+No consensus rules are altered, and on-chain data remains unchanged. Differences arise solely in display formats:
108+
109+* '''For Developers:'''
110+ Update GUIs, APIs, and documentation to present values as integers. Remove references to fractional Bitcoin. <code>BTC<code> units remain unchanged.
0 Update GUIs, APIs, and documentation to present values as integers. Remove references to fractional Bitcoin. <code>BTC</code> units remain unchanged.
128+===Test Vectors===
129+
130+* Old: <code>1.00000000 Bitcoin<code> → New: ₿100,000,000 (or 100,000,000 bitcoins)
131+* Old: <code>0.00010000 Bitcoin<code> → New: ₿10,000 (or 10,000 bitcoins)
132+* Old: <code>0.00500000 Bitcoin<code> → New: ₿500,000 (or 500,000 bitcoins)
133+* Old: <code>0.005 BTC<code> → New: <code>0.005 BTC<code> (or ₿500,000 or 500,000 bitcoins)
0* Old: <code>1.00000000 Bitcoin</code> → New: ₿100,000,000 (or 100,000,000 bitcoins)
1* Old: <code>0.00010000 Bitcoin</code> → New: ₿10,000 (or 10,000 bitcoins)
2* Old: <code>0.00500000 Bitcoin</code> → New: ₿500,000 (or 500,000 bitcoins)
3* Old: <code>0.005 BTC</code> → New: <code>0.005 BTC</code> (or ₿500,000 or 500,000 bitcoins)
54+Example 3:
55+
56+ * Old display: <code>0.345 BTC<code>
57+ * New display: No changes required or <code>₿34,500,000<code> or <code>34,500,000 bitcoins<code>
58+
59+NOTE: Traditional number display abbreviations, like <code>2.5M<code> for millions, are also optional.
0NOTE: Traditional number display abbreviations, like <code>2.5M</code> for millions, are also optional.
It seems to me that the motivation could be strengthened and that the document should address why presenting user funds denominated in “sats” does not achieve the intended goal. However, speaking as an editor, this proposal appears to be compliant with the formatting requirements at this point.
Assessing the mailing list thread and the comments on this proposal, all commenters so far appear to be skeptical. At this time, I do not perceive this proposal to have sufficient community interest to assign a number. I am happy to revisit this assessment when the proposal picks up more momentum.
Could you clarify what the threshold is for getting a number?
I don’t have a specific threshold in mind, but a start would be some replies expressing enthusiasm or any commentary in support of this proposal.
0@@ -0,0 +1,151 @@
1+<pre>
2+ BIP: ?
3+ Title: Redefinition of the Bitcoin Unit to the Base Denomination
Assessing the situation further, I have been convinced otherwise.
Let’s call this BIP 177.
My question would be what happens if a BIP is passed in the future for lesser unit denominations that are soft-fork compatible for finer control as the Bitcoin purchasing power rises, such as from 1.00000001 to 1.0000000001? Would this necessitate another BIP correction?
afaik, there is no meaningful/enforceable way to express decimals in Bitcoin. To truly get more precision, we would need to multiply the units. There are various edge considerations in such a change, like how it affects mining rewards into the future, etc. Any such change would likely be a hard fork, and even if someone designs something reasonable as a soft fork, it would still be controversial and face headwinds. There is no prominent design proposal I know of for any of that.
In the end, I cannot design for all hypothetical Bitcoin futures (or forks), so I focus on the one we have! :)
This pull request should be used to help to review and improve the document. Please identify areas that need clarification, suggest improvements, or point out any missing aspects. Merely stating disagreement with the idea itself is not constructive. Thanks.
I’ll keep that in mind and keep my comments over on other social channels from now on. 🤙🏻
Due to the controversy behind this and the overwhelming ratio of thumbs downs to thumbs up, I propose modifying the status to “Closed” as per BIP-3 terminology before it’s merged.
Not at all that I am in favor of this proposal, I am not; but I can’t see how a reaction of a fake account can be considered any decision-making. I see maybe more danger in considering fake account serious voting than in writing a BIP against 15 years of community history.
113+
114+===Backward Compatibility===
115+
116+No consensus rules are altered, and on-chain data remains unchanged. Differences arise solely in display formats:
117+
118+* '''For Developers:''' Update GUIs, APIs, and documentation to present values as integers. Remove references to fractional Bitcoin. <code>BTC</code> units remain unchanged.
12+
13+==Introduction==
14+
15+===Abstract===
16+
17+This BIP proposes redefining the commonly recognized "bitcoin" unit so that what was previously known as the smallest indivisible unit becomes the primary reference unit. Under this proposal, one bitcoin is defined as that smallest unit, eliminating the need for decimal places. By making the integral unit the standard measure, this BIP aims to simplify user comprehension, reduce confusion, and align on-chain values directly with their displayed representation.
Nit: This document uses multiple different terms to refer to the same concept, e.g.:
While it doesn’t strike me as confusing per se, it may help to introduce a single term at the top of the document in the context of all the relevant characteristics and then use that term consistently for the rest of the document.
Gave this document another read. It seems reasonably complete for being merged in Draft status. Review of the text rather than commentary on the likelihood of adoption has slowed down.
Do you mind if we squash the commits in this PR, since most of the commit messages are not providing additional context? @BitcoinErrorLog, were you planning any more changes?
Due to the controversy behind this and the overwhelming ratio of thumbs downs to thumbs up, I propose modifying the status to “Closed” as per BIP-3 terminology before it’s merged.
The BIPs process is currently governed by BIP 2, not BIP 3, but either way, the likelihood of adoption is not part of the criteria BIP Editors are supposed to consider when evaluating submissions to the repository.
Do you mind if we squash the commits in this PR, since most of the commit messages are not providing additional context? @BitcoinErrorLog, were you planning any more changes?
Please feel free to squash and proceed. I do not have any planned changes. Please let me know your best next recommended step for me after that, thank you!
- Redefine bitcoin base unit to smallest unit
- Propose BIP 21Q: Redefine bitcoin base unit to smallest indivisible unit
- Adds comments acknowledging and handling sats and satoshis
- Make use of "base unit" and variations more consistent and intentional
- Make "bitcoin" v "Bitcoin" consistent
- Made "bitcoin" v "Bitcoin" consistent by using Bitcoin for the protocol and idea, and bitcoin for the units, which I believe is conventional style.
Thanks, nice improvements. Squashed the commits and merged as discussed.
Thank you, sir. Could you confirm for me what my best next-step is for advancing this proposal formally?
@BitcoinErrorLog Could you clarify the motivation for why referring to the base unit as “bitcoin” is preferable over “sats”? You mentioned you added some comments addressing this, but the only thing I could find in the draft is “The informal terms ‘satoshi’ or ‘sat’ are deprecated”.
It is addressed further in Handling the Terms “satoshi” and “sat” – Generally, the proposal aims to call the only actual units of Bitcoin as bitcoins, which, aside from cultural attachments to “sats”, is a pretty basic and obvious intent.
I could add a section railing on all the subjective arguments against use of “satoshi” and why some people don’t like it at all… but I think the goal here isn’t to refute colloquial terms so much as describe a way to convey Bitcoin clearly and accurately while preventing misconceptions about its integral basis and removing the fake decimal.
Ah, thanks. Unfortunately I don’t find the argument very convincing.
Redefining Sathoshi’s original definition of the Bitcoin unit doesn’t strike me as basic or obvious at all. Just because the decimal is social consensus rather than code doesn’t mean it’s “fake”.
The decimal is quite literally fake, purely aesthetic graphics. You can find quotes from Satoshi explaining how it could be moved (or removed) anytime. Bitcoin does not actually support floating point numbering, it uses whole integers. The inspiration of this BIP is how often I encounter the misconception that we could “add more decimal places.” Bitcoin is not divisible.
The integer is a technical implementation detail. It seems misguided to make educating non-technical users about this implementation detail a priority of any naming conventions.
You have this backwards. The integer basis is a PROTOCOL detail, enforceable. The synthetic decimal place is an implementation detail, not enforceable.
Using the base unit for all amounts is probably not the clearest or most error-proof for UI. I’d argue that one unit for small amounts and another unit for large amounts (such as bits vs bitcoin as proposed in BIP176) is much clearer.
My argument against BIP 176 is already in this document.
The proposed redefinition risks undermining Bitcoin’s core value proposition of scarcity and its established perception in the global financial ecosystem. Representing balances in billions or quadrillions of “bitcoins” could diminish the psychological impact of Bitcoin’s fixed supply of 21 million, which has been a cornerstone of its value. Furthermore, retaining the currency code “BTC” while altering its definition creates inconsistencies in financial markets and regulatory frameworks, introducing unnecessary complexity and confusion for both current and future adopters.
Your feedback is made of speculation and paranoia.
People already represent Bitcoin using base units and the world still turns.
This BIP does not alter the definition of BTC (this usage is carved out in the proposal!), nor does it change the scarcity of bitcoin in the market.
Everything is gonna be alright – this outcome is more likely to be inevitable than a threat, imo.
So today I learned what sats are. My opinion: Seems like there is pushback for adopting sats (or renaming sats as bitcoin). This can only come from those bitcoin users that are not interested in bitcoin replacing regular money. My view is that it should be useable for all possible purposes, including everyday purchases. Some people are mainly or only interested in bitcoin for speculation with the BTC exchange rate and not so, or not at all, interested in having bitcoin replace all money. Offcourse if bitcoin would replace all money there is nothing to speculate with anymore as the bitcoin exchange rate would become meaningless if all other money is abondoned in favor of bitcoin.
This BIP proposes redefining the bitcoin unit of account to represent the protocol’s smallest indivisible unit as “1 bitcoin,” eliminating the need for decimal-based UI conventions.
- Type: Informational
- No changes to consensus rules or supply
- Matches Bitcoin’s native integer representation
- Includes multi-phase rollout recommendations
Authored by: John Carvalho (@BitcoinErrorLog)
Mailing list:
https://groups.google.com/g/bitcoindev/c/YwF-djZi1Bo/m/nIkyuClEAgAJ
Bad idea. Replacing an organically adopted term with a top-down redefinition just adds to the confusion. Changing a 14 year old definition without a clear need adds to confusion. Changing a unit already embedded in 100s of apps creates work for a lot of people with no clear upside. Attempting to change the meaning of a term that has already been used a hundred million times in print with one clearly understood meaning is just plain hubris.
Thankfully what people will call fractions of bitcoin will be the result of emergent order, not of some devs ego. Language is even more decentralised than Bitcoin.
Thankfully what people will call fractions of bitcoin will be the result of emergent order, not of some devs ego. Language is even more decentralised than Bitcoin.
Yeah, basically this. I’m hopeful that “bits” will catch on by the time bitcoin hits $1 million, because 1 bit = $1 will be a cool milestone.
OP, it’s pretty clear that there is little support for a proposal that changes the definition of Bitcoin, and you haven’t really engaged with any of the valid concerns regarding confusion etc. I suggest focusing our efforts on BIP176, or another proposal that doesn’t involve redefining Bitcoin.
For people who actually use Bitcoin every day BIP-177 is a step backwards in my opinion. nobody asked for a change
“Sat” is already the spoken, typed, and printed unit. Wallets, LN invoices, PoS screens, and memes all revolve around sats. Re-labeling them as “bitcoins” scraps a vocabulary that’s working fine. The Lightning spec, LNURL, BOLT11 invoices, and most on-chain UIs ask for sats. Rename that unit and you force every wallet, POS terminal, and QR code generator to choose between lying to the user or violating the protocol label. Either way, chaos.
Decimals aren’t scary; people comfortably use cents, millilitres, milligrams - and, historically, Britain’s far more complex £/s/d system. If the average person could understand £3 17s 9d (and they did), they can understand sats without a problem.
Good luck explaining this change to tech illiterate users that just got into bitcoin recently without them panicking about suddenly being “multi-bitcoin millionaires”. I wouldn’t be able to do it.
Decimals aren’t scary; people comfortably use cents, millilitres, milligrams—and, historically, Britain’s far more complex £/s/d system. If they could understand £3 17s 9d, they can understand sats.
To you, and everyone else here, I have no intention of speculating, defending or refuting subjective/external narratives about this BIP. Your complaints should address the BIP’s motivations and design, not your paranoia.
More importantly, you should actually read the entire thing, in earnest, before attempting to provide constructive criticism on its contents.
Thanks!
If we are going to play silly games with the unit debate, I want to play too!
Just for fun, here’s my proposal for a few multiple and fractional bitcoin unit names.
Disclaimer: It’s really stupid to try to plan these things top down, people will end up making up all kinds of local, jargon, fun names and use whatever they want no matter what devs think. You can’t plan language! Emergent order…
Please take wider conceptual discussion to the mailing list thread, linked to in the pull description. Thank you.
https://groups.google.com/g/bitcoindev/c/YwF-djZi1Bo/m/nIkyuClEAgAJ