Add unit test for non-standard utf8 characters, remove double validation check for tokenId as string, check for multiple blank spaces and add unit test for this case

declare tx_relay directly in the if's condition, skip the the init-statement

That fixes the warnings, but not the tests

@bot build-win64

• parentAffiliatedId set to the official ABC account from Antony
• Reverted settleAddress back to undefined otherwise the Receiving Address field will not be displayed in the widget. Whoever is managing the ABC account can just execute monthly withdrawals from the official account to the IFP address.

In D14728#331167, @emack wrote:

In D14728#331163, @Fabien wrote:

Any block height related optimization will incur an API call to get the latest blockheight as a minimum.

Not with websockets ?

Nay, the BlockConnected websocket message only contains the message type and block hash. Usually you'd pass the block hash to the chronik client's block() API to get more info but that itself is an API call.

The other websocket messages don't return height either.

In D14728#331163, @Fabien wrote:

Any block height related optimization will incur an API call to get the latest blockheight as a minimum.

Not with websockets ?

alright, green. sorry for holding this hostage for so long, but at least I didn't ask for ransom, so I guess it's not that bad

Any block height related optimization will incur an API call to get the latest blockheight as a minimum.

I must have missed something here

remove debug logging and unused const

nice wrapping, one less mutex and one less global

@bot build-tsan

In D14726#330975, @bytesofman wrote:

Is this diff in response to user complaints? what's the root cause of wanting to add this functionality?

is "fix it" here from the diff description referring to problem of not knowing the private key and having to enter it manually, or electrum is just always using the 0 key which is now not always correct?

Rebase to master

In D14729#330977, @bytesofman wrote:

are these proof editor improvements driven by user stories?

I dunno how much mileage we are going to get by improving the proof editor experience in electrum. It is already effective there. Everyone who uses the feature is not doing it on a daily basis.

If the problem is "we want to improve the UX for creating an avalanche proof" ... electrumabc is mb not the answer. It's a power user program. I'm also not sure we really want it to be easier or more convenient. It is relatively straightforward to do now, for users who also are competent to run and maintain a full node from the command line. If we are really trying to optimize the UX, something like a staking service business is probably better positioned (though I do not think ABC is the right group to run such a service).

I'm hesitant to add anything that is "caching the password more often for convenience."

Very nice ! That is a great unit tests which is easy to read, understand and review.