this line causes problems in WebAssembly only on mainnet because it depends on libunbound: github.com/monero-project/monero/bl…11589/src/wallet/wallet2.cpp#L13579

You can set use_dns to 0.

There's a command line flag for that.

sorry I disconnected if there were any responses

You can set use_dns to 0.

There's a command line flag for that.

oh perfect. thank you

I try to come to a full understanding how seed words are used to calculate the private spend key in Monero

My problem is with this part of the code, where 3 words are treated as number in the 1626 number system and turned into 32 bits: github.com/monero-project/monero/bl…c/mnemonics/electrum-words.cpp#L316

I would have written this simply as: w[0] = w[1] + wordListLength * w[2] + wordListLength * wordListLength * w[3];

But now it's more complicated, as can be seen in the source code. I wonder why.

I mean, what's the math principle behind ((word_list_length - w[1]) + w[2]) % word_list_length) ?

is the mod producing the checksum? monero.stackexchange.com/questions/…68/how-are-monero-wallets-generated rbrunner

No, it's not about the checksum. It's much more basic, how words are "turned into bits". It seems needlessly complicated, but I am quite sure there is some intention behind it. To know which would be interesting.

One theory I came up in the meantime: subtracting w[1] in the expression for the second "digit" I gave negates some "rotating" or "shifting" within the words which may make the seed more "varied" i.e. using more different words.

But well, on the other hand, the bits should be completely random anyway ...

Thanks for the link, scoobybejesus. It's much more than just an explanation of the checksum alright, but I could not see this particular detail in luigi's explanations