random thoughts

[petjal]

I'm really nervous about making sure I have a good random number generator.

openssl

https://wiki.openssl.org/index.php/Random_Numbers

raspberry pi, /dev/hwrng

https://raspberrypi.stackexchange.com/questions/61888/how-to-know-if-dev-hwrng-is-working

sudo apt-get install rng-tools
sudo kill -USR1 `cat /var/run/rngd.pid`
sudo grep rngd /var/log/daemon.log
An occasional failure of the FIPS-140 tests is nothing to worry about.

https://scruss.com/blog/2013/06/07/well-that-was-unexpected-the-raspberry-pis-hardware-random-number-generator/

sudo cat /dev/hwrng  | rawtoppm -rgb 256 256 | pnmtopng > random.png
cat random.png | rngtest -c 1000
sudo apt install ent
ent random.png
dieharder -a < random.png  # takes approximately an hour or more?

rng-tools

https://wiki.archlinux.org/title/Rng-tools#Testing_and_usage

pi@raspberrypi:~ $ dd if=/dev/random of=/dev/null bs=4096 count=1 iflag=fullblocktime rngtest -c 4096 </dev/random
1+0 records in
1+0 records out
4096 bytes (4.1 kB, 4.0 KiB) copied, 0.0789904 s, 51.9 kB/s

Without rngd, the above command will take lots of time to run. With rngd working properly, the result should be almost instantaneous:

time rngtest -c 4096 </dev/random
rngtest 2-unofficial-mt.14
Copyright (c) 2004 by Henrique de Moraes Holschuh
This is free software; see the source for copying conditions.  There is NO warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.

rngtest: starting FIPS tests...
rngtest: bits received from input: 81920032
rngtest: FIPS 140-2 successes: 4091
rngtest: FIPS 140-2 failures: 5
rngtest: FIPS 140-2(2001-10-10) Monobit: 1
rngtest: FIPS 140-2(2001-10-10) Poker: 2
rngtest: FIPS 140-2(2001-10-10) Runs: 2
rngtest: FIPS 140-2(2001-10-10) Long run: 0
rngtest: FIPS 140-2(2001-10-10) Continuous run: 0
rngtest: input channel speed: (min=181.377; avg=415.077; max=1330.739)Kibits/s
rngtest: FIPS tests speed: (min=1.621; avg=10.777; max=11.064)Mibits/s
rngtest: Program run time: 199996961 microseconds

real    3m20.010s
user    0m6.867s
sys     0m28.323s

It is normal for any random number generator to fail in a small number of tests in 1000 passes, however if the number of failures is too great (like 10), probably there is something wrong.

macos

TODO

windows, wsl/ubuntu terminal

TODO

[grondilu]

I'm really nervous about making sure I have a good random number generator.

Technically what is important is not so much the RNG, as its source of entropy, a.k.a its seed.

With openssl you can specify your own seed with the -rand option, and in fact in early versions of bitcoin-bash-tools I was generating random bytes with this command :

openssl rand -rand <(date +%s%N; ps -ef) -hex 32

This is a method I had seen used long time a ago in a Unix manual or something, though I'm not sure as I can't quite recall. I did not invent it anyway.

I gave up on it as the openssl man page states that it is not needed anymore :

Prior to OpenSSL 1.1.1, it was common for applications to store information about the state of
the random-number generator in a file that was loaded at startup and rewritten upon exit. On
modern operating systems, this is generally no longer necessary as OpenSSL will seed itself
from a trusted entropy source provided by the operating system. The -rand and -writerand
flags are still supported for special platforms or circumstances that might require them.

You could still use it I think. You can throw anything you want in addition to date and ps, such as a ping output, the latest block hash, a personal picture...

Or you can flip a coin. I'm currently doing it, so far I have produced 24 bytes that I have memorized with a combination of the major system, the PEG system and the method of loci.

[petjal]

Thanks! Love those mnemonic systems! Just need a mnemonic to remember how to spell mnemonic.

So, the reason I'm here is to get my family and friends stacking sats as reasonably safely, simply, and quickly as possible. I was thinking I could help them with something like:

1. generate a very, very good private key
2. generate a receiving address
3. back both of those strings up
4. sign up for DCA at SwanBitcoin, with autowithdrawal to the receiving address
5. study, adjust, plan succession later

I need to learn more about seeds in this context...off to google...
https://en.bitcoin.it/wiki/Seed_phrase

[petjal]

From a recent run of the Synopsis commands,
#30
are these the three example strings my family and friends would need?
Protect and backup the xpub and xprv like they were family heirlooms, and auto-withdraw their sats to the bc1 public receiving address?

xpub661MyMwAqRbcFstkfbizC5LcEG2psva5BzaVVLFEW45EZnNzh584uo7zDUJRE1YKAWZDnKn9Cj68jiLXCY6kqimPBET75Ukv3UmvHxcdqNr
xprv9zJeGmMZbuSoxLLC1YUdhu8LJducTuYXNbNYwXS2suhcBPwZFHrqxSDTeRaFrKeS8NQiCyYRF3W99QuxZjSkH1XFzKPHpmp1rn5GfDRn717
bc1qw508d6qejxtdg4y5r3zarvary0c5xw7kv8f3t4

[petjal]

I'm missing something...

~/dev/bitcoin/git $ myseed="$(openssl rand -hex 24)" ; echo $myseed
4b8dcb82a30c86f8646b204a3b3da0a3cb1738aab56a70d6             <---- (try to) remember this

~/dev/bitcoin/git $ openssl rand -rand <(echo $myseed) -hex 32
bfc9d1dfc7c287b513dc9915473426e6db9c654f3b56e3bcd69d933115d90602

~/dev/bitcoin/git $ openssl rand -rand <(echo $myseed) -hex 32
4cbbbb14435b34fb38ede94a332fca23215f4b51ae568ee9eeb1540551d603bf

~/dev/bitcoin/git $ openssl rand -rand <(echo $myseed) -hex 32
e438e93fb06e5b89ef9c003ce4dc2879c845c16af9eb63de59e2d11416683caf

[grondilu]

I'm missing something...

~/dev/bitcoin/git $ myseed="$(openssl rand -hex 24)" ; echo $myseed
4b8dcb82a30c86f8646b204a3b3da0a3cb1738aab56a70d6             <---- (try to) remember this

~/dev/bitcoin/git $ openssl rand -rand <(echo $myseed) -hex 32
bfc9d1dfc7c287b513dc9915473426e6db9c654f3b56e3bcd69d933115d90602

~/dev/bitcoin/git $ openssl rand -rand <(echo $myseed) -hex 32
4cbbbb14435b34fb38ede94a332fca23215f4b51ae568ee9eeb1540551d603bf

~/dev/bitcoin/git $ openssl rand -rand <(echo $myseed) -hex 32
e438e93fb06e5b89ef9c003ce4dc2879c845c16af9eb63de59e2d11416683caf

This is surprising indeed. I suppose openssl uses the provided file, but not exclusively. The point anyway is to improve the quality of your RNG. Making it deterministically depend on user input and on user input only would almost certainly make it worse, not better.

Clearly this means that if you plan on memorizing a seed, you must absolutely not use it as a seed for openssl's rand. Only use it with a deterministic function such as sha512 hmac as in bip32.

You can still use the -rand option, but only in case you want to use openssl to generate the seed you will memorize, not the other way around.

[petjal]

True?:

Remember, backup, and succession-plan the first string (your private seed).
Backup and succession-plan the xprv string.

Your private seed here replaces the seed phrase; see https://github.com/search?q=org%3Abitcoin+seed+phrase

~/.../git/bitcoin-bash-tools $ myseed="$(openssl rand -hex 24)" ; echo $myseed ; time for i in {1..1000} ; do bip32 -s $myseed ; done | sort | uniq -c | sort -k1nr
f78759a7abbe8b10f80ebcc15158fdc4ce046dc43896475a      <---- your  private seed
   1000 xprv9s21ZrQH143K45GUmKpnT3RahjxsphWQVWvz2p4UtWgtmCVeAxe2kHQNTS3FVeyrriTZd7XSbJunzyxAVYz6DCxhvuspjzRfg9T5Gdz9zHj

real    0m52.705s
user    0m40.996s
sys     0m29.488s

next step: make a receiving address to which to (auto)withdraw periodically from the exchange...

[grondilu]

you actually have several choices for what you want to memorize.

The most common one is to memorize the words generated according to bip39. These words are what people usually call a seed phrase, I think. If you do that, the random bytes you used to generate these words can be discarded/forgotten. These words will be turned into a seed that will feed bip32 to generate the master key.

An alternative is to bypass bip39 altogether and to memorize the seed you plan to feed to bip32. This is more difficult but I think it's doable. Risky, of course.

A mixed method is to memorize the bytes you use to generate the seed phrase and memorize the seed phrase too, for redundancy/backup.

I personally am not sure which method I will end up using. For the moment, I am learning and applying mnemonic techniques to memorize bytes I'm producing with a coin, but I haven't decided how I will end up using them yet.

Regarding succession, I personally don't care about that at all. Besides, I don't think there is any solution that wouldn't compromise the secrecy of your key.

[petjal]

(ps: stupid bash trick of sudo chmod 600 /dev/*random didn't work for the openssl -rand)

[petjal]

~/.../git/bitcoin-bash-tools $ myseed="$(openssl rand -rand <(date +%s%N ; ps) -hex 24)" ; echo "my private seed: $myseed" ; echo -n "my private xprv: " ; bip32 -s $myseed
my private seed: e0fdb490a3e1ae44c1c130520c0cf52e39398c99b69c9ee2
my private xprv: xprv9s21ZrQH143K2EydHRS3C7wDysDWzdTr8CnHrKZRUbZwNg5bcyNhfSFHPY4HtxAiK7AxP5jjpUiqpx5m3GUv49FWTZftZ1eoyrWkjNRi8Gd

[petjal]

A few questions, if I may.

1. Why two xprv? (One in $m, and one generated by bip32 $m/0h/5/7.)
2. I should replace the "0279be667ef9dcbbac55a06295ce870b07029bfcdb2dce28d959f2815b16f81798" with a variable from elsewhere in the script, correct?
3. How would I generate a second segwitAddress (for example, to withdraw from a second exchange)? (Maybe resolved by question 2 above?)

~/.../git/bitcoin-bash-tools $ my_private_seed=fa750dd1d0bfbcc3a61765bc10f7340c9bfc8ef518434381____ ; . bitcoin.sh ; echo newBitcoinKey... ; newBitcoinKey ; echo ; m="$(bip32 <<<$(echo $my_private_seed))" ; echo m=$m ; echo bip32 \$m/N... ;  bip32 $m/N ; echo bip32 \$m/0h/5/7... ; bip32 $m/0h/5/7 ; . bip-0173.sh ; echo ; echo segwitAddress... ; segwitAddress -p 0279be667ef9dcbbac55a06295ce870b07029bfcdb2dce28d959f2815b16f81798
newBitcoinKey...
KymHbWDFxiYgz1C8qWQ1nzdy96L1jfqnyB1reDnzBLCXRJquXdKF
1FAwPdJPCQFpaB9DQkoyghX2JVorRuC5Ta
read EC key
EC Key valid.
writing EC key
-----BEGIN EC PRIVATE KEY-----
MFQCAQEEIEvrMKp2dZcuu9n054+UCAXePE+d+FiFDnbb53SzFvSBoAcGBSuBBAAK
oSQDIgACZwux8wfbU1EvTVeyATrO/oCy6teDpU07D5go3YrrXfQ=
-----END EC PRIVATE KEY-----

m=xprv9s21ZrQH143K34VgsRRmcq73NXXYahEhuegMRJJfixX4PAhNJGVAhWSRHonSGpJW5qoBvcG6PUwnqnvwJaFs2GxTuFJ8nD7NZuXjqhsMWgF
bip32 $m/N...
xpub661MyMwAqRbcFYa9ySxmyy3mvZN2z9xZGsbxDgiHHJ43Fy2WqooRFJku96C8QjaL85AsicuWEt3unA6dn5hDvoXZqT3Dqdsq223Z564t1jh
bip32 $m/0h/5/7...
xprv9xvtzpQX3H4L3Z6Ex8voPHfaUTnh9DojejARWcgDX9v8Dc6mueumh4G7BdLRyS8x8XwWkCn8knDWTC9MqXSz8mfmNpvCQhQB1EJt9Tb6Ymi

segwitAddress...
bc1qw508d6qejxtdg4y5r3zarvary0c5xw7kv8f3t4

[grondilu]

There are not just two keys. The synopsis only shows one example of how to create a master key and a derived one.

I suspect you are not familiarized enough with extended keys as described in bip32. Don't get me wrong I don't judge as I wasn't either just few months ago.

I suggest you read the bip32 if you haven't already, or the relevant chapter in the bitcoin book :
https://github.com/bitcoinbook/bitcoinbook/blob/develop/ch05.asciidoc