Fix "Using a single OP_CHECKSIGADD-based script" #890

1. CHECKSIG / CHECKSIGADD is confused

Only the first OP-code for the first public key should be "CHECKSIG" and the following (second to n-th) OP-codes should be "CHECKSIGADD". It is confusing because it is only specified the first and last OP-codes, so I specified the second OP-code clearly. (I recommend to describe why only the first OP-code should be "CHECKSIG", not "CHECKSIGADD".)

2. Order of the signatures in witness

In the original sentence, the stack status after the all witness elements are pushed will be

and then, the first element of the script, "<pubkey_1>" will be pushed to the stack | pubkey_1 |

so the "pubkey_1" and "w_n" won't match.

The order of either "pubkey_i"s or "w_i"s should be inverted.

ACK 1. It is kind of confusing if you don't have enough context.

for 2. IIRC the order of the signatures shouldn't matter since the intermediate state of the interpreter doesn't matter unless a VERIFY operation etc. terminates processing... After all the signatures are added with the pubkeys, it should come out with a number of valid signatures for comparison.

ACK Good catch.

ACK