public inbox for bitcoindev@googlegroups.com
 help / color / mirror / Atom feed
From: Gordon Mohr <gojomo@gmail.com>
Cc: Bitcoin Development <bitcoin-development@lists.sourceforge.net>
Subject: Re: [Bitcoin-development] "bits": Unit of account
Date: Mon, 05 May 2014 15:33:37 -0700	[thread overview]
Message-ID: <536811C1.1060706@gmail.com> (raw)
In-Reply-To: <CACq0ZD7s8tp8GvJhEhZx4T7xMpeZ+tz5HNKQK-p=f=R10NaCmA@mail.gmail.com>

On 5/2/14, 10:41 PM, Aaron Voisine wrote:
> I have to agree with Mike. Human language is surprisingly tolerant of
> overloading and inference from context. Neurotypical people have no
> problem with it and perceive a software engineer's aversion to it as
> being pedantic and strange. Note that "bits" was a term for a unit of
> money long before the invention of digital computers.

Of course people *can* manage, when they need to; natural language is 
full of such overloading. But the clashes are not costless, they add 
mental load for first-time learners and low-context users.

So the concern is, when there's a free choice, why not bootstrap words 
that are less fragile and context-dependent? Why add extra comprehension 
gotchas into what is already a challenging domain?

And it's exactly the aspect that makes 'bit' attractive – "it's right 
there in the name _Bit_coin!" – that equally presents the clash – 
because the sense of 'bit' honored in the "Bitcoin" name, and central to 
the systems' essential properties, is the binary digit.

It's like intentionally introducing a 'false friend' word-correlation 
between the vernacular of the casual Bitcoin user, and the language of 
Bitcoin experts. And the word pair is nearly auto-antonymic in some 
essential dimensions: indivisible vs. divisible, base-2 vs. base-10, 
composed-geometrically vs. composed-arithmetically.

https://en.wikipedia.org/wiki/False_friend - interferes w/ lang learning
https://en.wikipedia.org/wiki/Auto-antonym

In naming amounts, there's no desperate need to stay in the shallow 
crowded pool of words just derived from "bit" and "coin". Real 
currencies have many names for their units, including subunits with 
highly-unrelated sounds. The contrasting words help create more shades 
of meaning for different purposes. Some examples:

  dollars/bucks - bits (1/8ths) - dimes (1/10ths) - cents (1/100ths)
  pounds/quid - shillings (1/20ths) - pence (1/100th; formerly 1/240th)
  yuan/kuai - jiao (1/10ths) - fen (1/100ths)

Regarding the cute example of contextual disambiguation...

On 5/3/14, 11:15 PM, Aaron Voisine wrote:
> Bit by bit, it's become clear that it's a bit much to worry even a
> little bit that overloading the word "bit" would be every bit as bad
> as a two bit horse with the bit between it's teeth that bit the hand
> that feeds it, or a drill bit broken to bits after just a bit of use.

That there are many existing definitions doesn't reassure that one 
*extra* definition will still be costless, especially for 
low-literacy/low-context/low-numeracy users or learners.

Note that this example *doesn't* showcase the new proposed 
'100-satoshi-value' usage, nor activate the 'binary digit' meaning. (It 
does activate the "part/quantity, usually small or imprecise" sense of 
'bit', 7 times depending on how you classify the idioms.)

Try instead this mash of concepts, which someone deciding whether to 
trust 'bits' will face:

"Bitcoin uses the digital science of bits, the indivisible 1s and 0s of 
computer logic, to create a networked money measured in bits, which 
split into 100 indivisible cents called satoshis. Bit amounts are 
represented as 64-bit integer counts of 1/100th of a bit, so 64-bit 
integers can represent any balance from the smallest positive bit total, 
0.01 bits (integer 1), up to a number over 92 quintillion bits (2^63-1, 
integer 9,223,372,036,854,775,808).

That max value won't be needed, though, because a crucial bit of the 
original Satoshi design is a maximum issuance of 21,000,000,000,000.00 
bits (21 trillion bits, 21 terabits). These new bits are awarded to 
computers racing to complete a digital verification task on an 
algorithmic schedule: currently 25 million bits (25 megabits) arrive 
about every 10 minutes. That is, the total number of Bitcoin bits is 
increasing at 6.33 Kbps, though that arrival slows to 0 bps by around 
the year 2140.

The most important bit to remember is that your ability to spend bits is 
controlled by secret 256-bit numbers, called private keys, bits of info 
that only you know. The fact that these keys are 256-bits long is what 
makes them practically unguessable, even if someone had a computing 
budget of all the bits in the world, or built a computer out of all the 
bits in the universe. (That is, even though the network can create 25 
million bits every 10 minutes, it can't guess your secret 256 bits in 
the lifetime of the universe!)

Watch out, though: human-chosen passwords and 4-8 word phrases typically 
provide much less than 128 bits of security, far too little to create a 
256-bit key. And in the math of bits, having half as many bits doesn't 
mean half the security, it means the square-root as much security. (For 
a 128-bit shortfall, that's 2^128 or 340 billion billion billion billion 
times less strength.)

If your secret has enough bits, though, you can be confident that you 
can put millions of dollars into bits, because of the cryptographic 
power of hundreds of bits. The current value of a bit is 1/20th of a US 
cent, so 256 bits has the purchasing power of about 11¢.

If you say you've got a bit of bitcoin in your wallet, I'll need you to 
be a bit more specific. If you've got one bit of bitcoin, you've got 
1/20th of a cent worth, an insignificant bit. But if you've got one 
bitcoin, you've got about $440 worth, quite a bit! Now you know a bit 
about Bitcoin, where your highly-valuable bits are protected by the 
science of bits. Get some bits now, they're small and still cheap, what 
is there to lose?"

Now, most people may never need to understand binary digits and 
information science. But if a future of widespread cryptocurrency 
success comes to pass, more people than ever before will want (and need) 
to understand the basics, and we can help them with clear terminology.

I'd also point out to those who've built their brands around the word 
'bit' that while it may seem grand to have the currency's most-common 
unit in your name/logo, you might wind up snake-bit. How?

*If* 'bit' is effectively bootstrapped to mean a tangible amount that 
people understand intuitively for everyday commerce, that amount is 
currently tiny (1/20th of a USD cent) and would still be tiny (10¢) if 
bitcoin appreciates in value over 200x to USD$100K per bitcoin.

So "bitpay" starts reading as "penny-pay" – and under an optimistic 
appreciation scenario might someday read as "dime-pay".

"bitsofproof" reads as "pennies-of-proof" or someday "dimes-of-proof".

- Gordon




  parent reply	other threads:[~2014-05-05 22:33 UTC|newest]

Thread overview: 56+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2014-05-01 22:35 [Bitcoin-development] "bits": Unit of account Aaron Voisine
2014-05-03  2:06 ` Gordon Mohr
2014-05-03  5:41   ` Aaron Voisine
2014-05-03 15:48     ` Christophe Biocca
2014-05-03 16:02       ` slush
2014-05-03 16:10         ` Tamas Blummer
2014-05-03 16:27       ` Mike Caldwell
2014-05-04  1:04         ` Chris Pacia
2014-05-04  5:18           ` Drak
2014-05-04  6:15             ` Aaron Voisine
2014-05-04  6:23               ` Un Ix
2014-05-04  6:27               ` Wladimir
2014-05-04  6:36                 ` Tamas Blummer
2014-05-04  6:59                   ` Wladimir
2014-05-04 14:42               ` Mike Caldwell
2014-05-05 22:33     ` Gordon Mohr [this message]
  -- strict thread matches above, loose matches on Subject: below --
2014-04-20 12:35 Mike Gehl
2014-04-20 13:15 ` Rob Golding
2014-04-20 14:28   ` Tamas Blummer
2014-04-20 14:52     ` Christophe Biocca
2014-04-21  8:52       ` Thomas Voegtlin
2014-04-21  9:34         ` Tamas Blummer
2014-04-22 14:55           ` Natanael
2014-04-20 14:53     ` Pieter Wuille
2014-04-20 15:05       ` Tamas Blummer
2014-04-20 15:50         ` Alan Reiner
2014-04-20 16:19           ` Chris Pacia
2014-04-20 16:27             ` Wladimir
2014-04-20 16:30               ` Chris Pacia
2014-04-22 13:51               ` Aaron Axvig
2014-04-23  9:44                 ` Danny Hamilton
2014-04-23  9:56                   ` Tamas Blummer
2014-04-23 11:48                     ` Chris D'Costa
2014-04-20 16:23           ` Erik Garrison
2014-04-20 16:30             ` Alan Reiner
2014-04-20 16:56               ` Mike Caldwell
2014-04-20 17:47                 ` Jannis Froese
2014-04-20 18:10                 ` Pavol Rusnak
2014-04-20 17:42 ` Arne Brutschy
2014-04-20 18:11   ` Mike Caldwell
2014-04-20 18:22     ` Arne Brutschy
2014-04-20 18:34       ` Mike Caldwell
2014-04-20 18:43     ` Oliver Egginger
2014-04-20 19:19       ` Christophe Biocca
2014-04-20 19:32         ` Gmail
2014-04-20 20:28         ` Mike Caldwell
2014-04-21  0:16           ` Justin A
2014-04-21  1:18             ` Mike Caldwell
2014-04-21  1:33               ` Un Ix
2014-04-21  3:34                 ` Mike Caldwell
2014-04-21  4:08                   ` Christopher Paika
2014-04-21  5:41                 ` Pieter Wuille
2014-04-21  5:51                   ` Tamas Blummer
2014-04-21  6:21                   ` Tamas Blummer
2014-04-21 12:14                     ` Un Ix
2014-04-21 12:24                       ` Tamas Blummer

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=536811C1.1060706@gmail.com \
    --to=gojomo@gmail.com \
    --cc=bitcoin-development@lists.sourceforge.net \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox