more robust check

So, this is not an exhaustive check; it just patches one (probably the most common) error case.

Note that the feature_deterministic_chain_setup.py test only exists in Bitcoin ABC, Bitcoin Core does not have it. We added it to demonstrate that the hardcoded regtests chainparams (block hashes and txoutsetinfo), which differ from Core due to differences in the block header format, are actually correct.

In D17630#401188, @emack wrote:

Can you provide some context on how to validate this output from the script above and that it is indeed excluding the not-bought or cancelled offers?

what is the browser / device here?

Ubuntu 22.04 / Chrome 125.0.6422.112

In D17625#401056, @emack wrote:

On extension the Send button is aligned left, which is correct but the rest of the component is overflowing off to the right as per current behavior. Makes the Send button almost look like a pop up not to mention it covers the labels beneath it.

give button container background color so scroll-behind looks cleaner, line it up with app buttons on mobile

only reason to do this was potential safari behavior (that, if true, this diff would potentially not resolve)

Tested on iOS Safari 18.1.1 - after paying in cashtab.com it automatically routes back to the paybutton page with the bell and then goes back to the original wp article that triggered it. I'm not seeing the reported issue where it stays on Cashtab.com.

On extension the Send button is aligned left, which is correct but the rest of the component is overflowing off to the right as per current behavior. Makes the Send button almost look like a pop up not to mention it covers the labels beneath it.

version bump + extension bump

note the test failure is from when I first put the diff up having accidentally deleted the file

rebase

Make local acceptance match what this node would eventually vote

add diff url

include diff url

Unrelated failure, maybe fixed by https://github.com/bitcoin/bitcoin/pull/26651

bump version using npm, add link to diff

rebase
The previous CI failure is unrelated

This not-quite doubles the package size (compressed: 150.4 kB -> 245.0 kB, uncompressed: 564.2 kB -> 934.9 kB), but since we still stay below 1 MB even uncompressed this doesn't seem to be a big issue