Fix Images For Normal HD Key Derivation; Mention Ancestor Key Risk

[harding]

While reading an excellent article by @vbuterin about HD wallet risks, I realized I used an incorrect key derivation formula that failed to include parent public keys in the hash. This patch for #393 fixes that error in both text and images.

I also added a short paragraph describing the risk explained in @vbuterin's article and updated the corresponding image to illustrate that risk. I would've preferred to add this new content separately after #393 gets merged, but I saw no point in updating the same image twice---once to correct my error and once to illustrate the additional risk.

Unrelated change: I removed an includes line from devref referring to a previously-deleted file; an upgraded version of Jekyll was dying because of this line.

[saivann]

@harding Thanks for your work on this! I couldn't find anything to add. I think this is better than what we have now, so:

In the absence of critical feedback, this pull request will be merged on May 19th.

[gmaxwell]

@harding This is why the hardened derivation exists, as it closes that issue. Perhaps that could be made more clear in the text?

[vbuterin]

Closes it at the expense of having a way to create a complete watch-only
wallet from the master public key. It's important to point out the
tradeoffs.

On 14-05-19 07:11 PM, Gregory Maxwell wrote:

@harding https://github.com/harding This is why the hardened
derivation exists, as it closes that issue. Perhaps that could be made
more clear in the text?

—
Reply to this email directly or view it on GitHub
bitcoin#408 (comment).

[gmaxwell]

@harding Also, can you suggest improvements to the BIP32 text that would have made it more obvious to you there?

[harding]

@gmaxwell I like making things clearer. :-) I think we could make it
pretty darn explicit by giving the Hardened Keys subsection a nice
strong first sentence, such as:

Hardened extended keys fix a potential problem with regular
extended keys.

That'll require some rewording of the rest of the opening paragraph,
so I'll work on it tomorrow and be sure to @ mention you in the
resulting pull request.

Re: BIP32, I know the exact list item from which I drew an incorrect
inference. I'll re-read it tomorrow (to make sure I didn't read it
wrong before) and submit a pull request you and @sipa can review.

Thanks!

David A. Harding

[harding]

On Mon, May 19, 2014 at 05:04:09PM -0700, vbuterin wrote:

Closes it at the expense of having a way to create a complete watch-only
wallet from the master public key. It's important to point out the
tradeoffs.

@vbuterin Hi! Thanks for your article! More importantly, thanks for your
nice code in pybitcointools. It was a great aid to me in understanding
this risk.

When I make the revision @gmaxwell suggested, I'll be sure to add a
quick note that the master extended private key is hardened by default
in a BIP32-compliant wallet.

Thanks!

David A. Harding