Appendix D (SegWit) confusion around SegWit P2SH address generation.

[grisha]

Right above 3deadbeef (fixed in issue #331) there is also a confusing (if not erroneous) part:

The hash of the witness program is then embedded in a P2SH script:

HASH160 3e0547268b3b19288b3adef9719ec8659f4b2b0b EQUAL

Finally, the P2SH script is converted to a P2SH bitcoin address:

37Lx99uaGn5avKBxiW26HjedQE3LrDCZru

This reads to me like HASH160 3e0547268b3b19288b3adef9719ec8659f4b2b0b EQUAL somehow becomes 37Lx99uaGn5avKBxiW26HjedQE3LrDCZru, but actually, the address is derived from the witness program mentioned three paragraphs earlier (0 ab68025513c3dbd2f7b92a94e0581f5d50f654e7):

echo '0 [ab68025513c3dbd2f7b92a94e0581f5d50f654e7]' | bx script-to-address 
37Lx99uaGn5avKBxiW26HjedQE3LrDCZru

A similar problem down below (bottom of page 339 in print):

Mohammed’s wallet creates a P2WSH witness program

0 9592d601848d04b172905e0ddb0adde59f1590f1e553ffc81ddc4b0ed927dd73

Then, the witness program itself is hashed with SHA256 and RIPEMD160, producing a new 20-byte hash, as used in traditional P2SH:

  86762607e8fe87c0c37740cddee880988b9455b2

Next, Mohammed’s wallet puts the hash into a P2SH script:

  HASH160 86762607e8fe87c0c37740cddee880988b9455b2 EQUAL

Finally, the wallet constructs a bitcoin address from this script:

3Dwz1MXhM6EfFoJChHCxh1jWHb8GQqRenG

The wallet constructs the address not from "this script", but rather from the witness program:

echo '0 [9592d601848d04b172905e0ddb0adde59f1590f1e553ffc81ddc4b0ed927dd73]' | bx script-to-address 
3Dwz1MXhM6EfFoJChHCxh1jWHb8GQqRenG

P.S. Fantastic book, BTW, thank you!

[aantonop]

Confirming...

[aantonop]

Thank you for this. You found several errors. I had mixed up the steps.

Now fixed, with bx examples throughout and correct addresses 1ddd89b

Please review and confirm, if you can!