Update Bitcoinj to latest version 0.13.2

[gary-rowe]

The latest version of Bitcoinj contains many fixes that we should include. In particular a fix for Tor connectivity (see #651) in Orchid 1.2.

We should use bitcoinj 0.13 when it is released

[jim618]

Probably need to bump to use bitcoinj 0.13.2 when it comes out (in a week) as that is fixing some regressions in 0.13.1

[jim618]

There is a new update to alice: release-0.13.2-alice-0.1.3

There is a branch pushed and the artifacts are in the staging repo.

It does NOT increment the version of Tor though - it is a few bug fixes to Wallet/ Transaction etc. thus the Tor connectivity problems we've been seeing are not addressed by this bump.

It still needs all the intermediate repos (multbit-commons) etc updated. I have jammed in the dependency into the MBHD develop pom.

MBHD appears to work exactly as before - done sync, repair wallet, send/ receive etc.

Awaiting review and closing.

[gary-rowe]

@jim618 Can you build the latest MBHW as develop-SNAPSHOT and verify that it correctly generates expected addresses for a Trezor wallet, please? Ideally using a fresh wallet compared against, say, https://dcpos.github.io/bip39/.

There was an API change in Bitcoinj that moved the Utils.doubleHash() into Sha256Hash.hashTwice() which is used in the GetPublicKey of MultiBit Hardware. I'm confident it's OK due to my own testing, I just want independent verification.

[jim618]

Successfuly tried out on a Mac:

• empty and wipe existing Trezor wallet.
• create new Trezor wallet
• fund with 11 transactions from Classic
• wait for confirmation on Trezor
• check 2 receive addresses against https://dcpos.github.io/bip39/.
• send 3 transactions out from Trezor
• perform empty wallet on Trezor.

This all worked as expected.

Will install on the Windows 8 machine and try that out next

[jim618]

I have performed a similar test on the Windows 8 machine and it all worked correctly. Closing