Add pico amount example to Bolt11 test vectors

[Sword-Smith]

Several libraries (at least two libraries and one web service) have problems with the p numbers representing 10^(-12) BTC. So an example with this encoding should be included in the test vectors. You can use this one if you like:

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

[Sword-Smith]

Libraries and web services unable to decode this invoice

0: btcpayserver/BTCPayServer.Lightning#25 (interprets p as 10^(-11) BTC)
1: bitcoinjs/bolt11#27
2: https://lightningdecoder.com/ (reported Amount is outside of valid range on 2019.10.31)

[NicolasDorier]

m (milli): multiply by 0.001
u (micro): multiply by 0.000001
n (nano): multiply by 0.000000001
p (pico): multiply by 0.000000000001

In BTCPay our conversion in Millisatoshi is the following:

    public enum LightMoneyUnit : ulong
    {
        BTC = 100_000_000_000,
        MilliBTC = 100_000_000,
        Bit = 100_000,
        Micro = 100_000,
        Satoshi = 1000,
        Nano = 100,
        MilliSatoshi = 1,
        Pico = 1
    }

So 1000 millisatoshi = 1000 pico = 1 satoshi. I can't see the issue.

[NicolasDorier]

ah sorry 10 milli satoshi = 1 pico ?

[NicolasDorier]

Ah got it... a pico is less than a milli satoshi, but all API deals with Milli satoshi...

[NicolasDorier]

@Sword-Smith I am wondering if I fix it as nobody is using less than MilliSatoshi.

Internally our class is doing math on a long millisatoshi whose max value is
9,223,372,036,854,770,000.

As you can see, we can represent 21 millions bitcoin in term of MilliSatoshi on a long:

02,100,000,000,000,000,000
09,223,372,036,854,770,000

But if we do our arithmetic on pico, then we can't represent the maximum amount of bitcoin on a long.

[NicolasDorier]

Sorry for the spam. I decided to not support pico for now:

btcpayserver/BTCPayServer.Lightning#25 (comment)

[NicolasDorier]

I decided to support pico, just not values that are not multiple of 10. Changing the minimum denomination below 1 millisatoshi is impossible at this point in my library without breaking lot's of things.

I think that a similar restriction should be placed in BOLT11, as sub milli satoshi is not possible at the protocol level.

[Sword-Smith]

I decided to support pico, just not values that are not multiple of 10. Changing the minimum denomination below 1 millisatoshi is impossible at this point in my library without breaking lot's of things.

I think that a similar restriction should be placed in BOLT11, as sub milli satoshi is not possible at the protocol level.

I honestly think this is the correct solution in that this is what every one else is doing. I couldn't find that requirement in the specification. If it isn't there, I fully agree that it should be.