- When background syncing, the wallet wipes the spend key
from memory and processes all new transactions. The wallet saves
all receives, spends, and "plausible" spends of receives the
wallet does not know key images for.
- When background sync disabled, the wallet processes all
background synced txs and then clears the background sync cache.
- Adding "plausible" spends to the background sync cache ensures
that the wallet does not need to query the daemon to see if any
received outputs were spent while background sync was enabled.
This would harm privacy especially for users of 3rd party daemons.
- To enable the feature in the CLI wallet, the user can set
background-sync to reuse-wallet-password or
custom-background-password and the wallet automatically syncs in
the background when the wallet locks, then processes all
background synced txs when the wallet is unlocked.
- The custom-background-password option enables the user to
open a distinct background wallet that only has a view key saved
and can be opened/closed/synced separately from the main wallet.
When the main wallet opens, it processes the background wallet's
cache.
- To enable the feature in the RPC wallet, there is a new
`/setup_background_sync` endpoint.
- HW, multsig and view-only wallets cannot background sync.
1. rewrite of the Haven variation of the Pricing Record class to support:
- versioning of the PR format
- nested supply_data and asset_data structs
- verification of the signature using a variable-length string (not R+S)
2. calculation of the slippage tallies for a block in add_block(), so that
we can work out the yield that is due to be paid out for the block.
Loads of little fixes and cleanups.
After all of the trials and tribulations, we finally have a viable prototype!
The code is working for conversions of all kinds. There are a number of
strange and annoying bugs still present, like the rings in the CLI wallet are
misbehaving at times. But we have a viable product.
Lots of refactoring still to be done in the codebase, and currently TRANSFER
method does not work - the daemon does not send the real output for some reason.
Lots still to do - please be patient.
This code is NOT ready for use. Or testing. Or anything else.
implementation of knaccc's "return address scheme" (for more information on this,
please see https://github.com/monero-project/research-lab/issues/53).
Fulmo's PROTOCOL_TX construct relies on a variation of knaccc's design to be
able to make return payments to individuals.
This code is not commercial-grade - it isn't even safe for alpha testing yet,
but it is operational under at least some circumstances, and the math is correct.
Caveat emptor.
1. The CONVERT TX is creating the necessary information.
2. The PROTOCOL TX is creating the necessary information.
3. The wallet recognises the subaddress (kind of) on incoming amounts.
At present, the PROTOCOL TX outputs are NOT spendable or included in balances.
jeffro256