- Wallet.make_unsigned_transaction takes a FeePolicy parameter
- fee sliders act on a FeePolicy instead of config
- different fee policies may be used for different purposes
- do not detect dust outputs in lnsweep, delegate that to lnwatcher
- Separation between SwapManager and its transport:
Legacy transpport uses http, Nostr uses websockets
- The transport uses a context to open/close connections.
This context is not async, because it needs to be called
from the GUI
- Swapserver fees values are initialized to None instead
of 0, so that any attempt to use them before the swap
manager is initialized will raise an exception.
- Remove swapserver fees disk caching (swap_pairs file)
- Regtests use http transport
- Android uses http transport (until QML is ready)
Somewhat a follow-up to 649ce979ab.
This adds some safety belts so we don't accidentally sign a tx that
contains a dummy address.
Specifically we check that tx does not contain output for dummy addr:
- in wallet.sign_transaction
- in network.broadcast_transaction
The second one is perhaps redundant, but I think it does not hurt.
Note that HTLCs must not be cancelled after the funding transaction
has been broadcast. If one want to cancel a swap once the funding
transaction is in mempool, one should double spend the transaction.
script destinations.
This is mainly to support OP_RETURN outputs, which typically have a zero amount output value,
but as we don't special case OP_RETURN, this is currently done for all non-address scripts
Also, it's probably good to add a warning popup for OP_RETURN outputs with a non-zero output value, but this
would also need special casing for OP_RETURN.
Saving of script output payment identifiers is disabled for now, as reading the script from the stored invoice
back into human-readable form is currently not implemented, and currently only lightning invoices or address output
is supported.
Fix '!' amount expand while PaymentIdentifier not set due to editor pushback timer.
Fix '!' amount expand for BIP21 without a specified payment amount.
```
229.18 | E | gui.qt.main_window.[test_segwit_2] | on_error
Traceback (most recent call last):
File "...\electrum\gui\qt\util.py", line 917, in run
result = task.task()
File "...\electrum\gui\qt\send_tab.py", line 681, in broadcast_thread
if self.payto_e.payment_identifier.has_expired():
AttributeError: 'NoneType' object has no attribute 'has_expired'
```
In SendTab.broadcast_transaction.broadcast_thread, self.payto_e.payment_identifier was referenced -
but do_clear() has already cleared it by then.
E.g. consider SendTab.pay_onchain_dialog: it calls save_pending_invoice(), which calls do_clear(),
and later (in sign_done), it calls window.broadcast_or_show, which will call SendTab.broadcast_transaction().
As there might be multiple independent transaction dialogs open simultaneously, the single shared state
send_tab.payto_e.payment_identifier approach was problematic -- I think it is conceptually nicer to
pass around the payment_identifiers as needed, as done with this change.
However, this change is not a full proper fix, as it still somewhat relies on
send_tab.payto_e.payment_identifier (e.g. in pay_onchain_dialog). Hence, e.g. when using
the invoice_list context menu "Pay..." item, as payto_e.payment_identifier is not set,
payment_identifier will be None in broadcast_transaction.
but at least we handle PI being None gracefully -- before this change, broadcast_transaction
expected PI to be set, and it was never set to the correct thing (as do_clear() already ran by then):
depending on timing it was either None or a new empty PI. In the former case, producing the above
traceback and hard failing (not only for bip70 stuff!), and in the latter, silently ignoring the logic bug.
