BugFix: ATC error message improvement #463
Conversation
barnjamin
force-pushed
the
atc-err-improvement
branch
from
e4c4cdf
to
33ac1ac
Compare
barnjamin
changed the title
| @@ -32,7 +32,7 @@ def parse_string(text): | |||
| algod_port = 60000 | |||
| kmd_port = 60001 | |||
|
|
|||
| DEV_ACCOUNT_INITIAL_MICROALGOS: int = 10_000_000 | |||
| DEV_ACCOUNT_INITIAL_MICROALGOS: int = 100_000_000 | |||
There was a problem hiding this comment.
Why is this here too? Seems completely unrelated
There was a problem hiding this comment.
Tests have been v flakey with issues related to overspend errors. This change was an experiment to see if it fixed the overspend errors.
The tests passed on the first attempt once I included it, but I cant be sure this is the fix
There was a problem hiding this comment.
I outlined the rationale for this one liner here as well: #459
It resolves current flaky tests because rekeying will change how transient accounts are funded in a non-deterministic way
There was a problem hiding this comment.
We already do this in the JS SDK but swept it under the rug in another PR: https://github.com/algorand/js-algorand-sdk/blob/develop/tests/cucumber/steps/steps.js#L122
There was a problem hiding this comment.
The rekey changes the ordering of the kmd wallets in py-algorand-sdk (I believe)
i think the change should be to provide non-deterministic access to the main network accounts instead
closes #462