(qt gui thread vs asyncio thread race)
Traceback (most recent call last):
File "/home/user/wspace/electrum/electrum/gui/qt/main_window.py", line 1685, in do_send
self.pay_lightning_invoice(self.payto_e.lightning_invoice)
File "/home/user/wspace/electrum/electrum/gui/qt/main_window.py", line 1667, in pay_lightning_invoice
self.invoice_list.update()
File "/home/user/wspace/electrum/electrum/gui/qt/invoice_list.py", line 73, in update
_list = self.parent.wallet.get_invoices()
File "/home/user/wspace/electrum/electrum/wallet.py", line 525, in get_invoices
out += self.lnworker.get_invoices()
File "/home/user/wspace/electrum/electrum/util.py", line 401, in
return lambda *args, **kw_args: do_profile(args, kw_args)
File "/home/user/wspace/electrum/electrum/util.py", line 397, in do_profile
o = func(*args, **kw_args)
File "/home/user/wspace/electrum/electrum/lnworker.py", line 1007, in get_invoices
for key, (invoice, direction, status) in self.invoices.items():
RuntimeError: dictionary changed size during iteration
Previously commands did not run on the asyncio thread but now they do.
"restore" was polling like "while 1: time.sleep()", blocking the event loop.
Now "restore" does not sync the wallet; which is actually cleaner
as previously this wallet would never get unloaded from the daemon (syncing forever).
This is also symmetric with the "create" cmd which also does not try to sync with the network.
However now it became difficult to write a script that restores a wallet and wants to wait
until it gets synced. Workaround for now is to poll with "list_wallets" whether it's synced.
We could create a new command that blocks until the loaded wallet gets synced.
e.g. when interacting with hw wallets (e.g. signmessage)
it does not make sense to time out
also, str(e) of some exceptions such as TimeoutError is ""...
SomberNight
ThomasV