Merge #805: logs: edit logs/2016-05-zurich-meeting-notes.txt for valid html tag

998ceaf logs: edit for valid html tag (randymcmillan)

Pull request description:

ACKs for top commit:
  kanzure:
    oh was that invalid? ACK 998ceaf
  harding:
    ACK  998ceaf(FWIW, the normal version of HTML proofer we use doesn't catch this; you must be using a different version)

Tree-SHA512: d0fdf269311fabd438193eae67056fe41effbba5aa3e6268268657463c101329e57c8e0c77efa400d82a8414a29b533fc44c8aded8f74cdeeb9fb072773ce442

Author: harding

logs/2016-05-zurich-meeting-notes.txt

@@ -506,7 +506,7 @@ Instead of SPV clients accepting miner assertion, why don't they accept miner do
 
 Bundle committed transactions with a hard-fork, and this different reduced SPV model. Luke disagrees that they implement SPV because they don't even do fraud proofs. 
 
-<a name="lightning-network-key-derivation" />
+<a name="lightning-network-key-derivation" >
 ### Lightning Network key derivation function
 
 Say you have 5 channels. Yo uwant to be pretty clear that they are actually yours. So you can build network graphs. If you just have a bunch of edges, you can't sort of link them all together into a vertex, it doesn't help. But if you say this edge and this edge are all connected, even if I'm not using the same public key for these 2-of-2 multisig, I must have had a key in common to generate all of these.