Details
- Reviewers
deadalnix schancel - Group Reviewers
Restricted Project - Commits
- rSTAGINGe94a989a32f9: Fixed padding and tx ordering in bip68-sequence.py post-fork
rABCe94a989a32f9: Fixed padding and tx ordering in bip68-sequence.py post-fork
./test_runner.py bip68-sequence
Diff Detail
- Repository
- rABC Bitcoin ABC
- Branch
- pad-bip68
- Lint
Lint Passed - Unit
No Test Coverage - Build Status
Buildable 3666 Build 5407: Bitcoin ABC Buildbot (legacy) Build 5406: arc lint + arc unit
Event Timeline
| test/functional/bip68-sequence.py | ||
|---|---|---|
| 392 ↗ | (On Diff #5509) | Something is seriously not right here. Why is sendtoaddress generating a TxID that is not valid? |
Removed bad pad_tx call that could mask real errors in the future
| test/functional/bip68-sequence.py | ||
|---|---|---|
| 392 ↗ | (On Diff #5509) | Good catch. I added pad_tx() as a precaution rather than it failing the test. Since you brought this up, I realize that it can actually mask real errors and should *not* be applied. I've removed it. |
| test/functional/bip68-sequence.py | ||
|---|---|---|
| 30 | It doesn't seems like a great idea to stop testing for current behavior and only check for post fork behavior before the fork. | |
| test/functional/bip68-sequence.py | ||
|---|---|---|
| 30 | Is it preferable to duplicate this test and delete the pre-fork version after the fork? | |
| test/functional/bip68-sequence.py | ||
|---|---|---|
| 30 | You can pass flags through to the test. See in test_runner.py TEST_PARAMS = {
# Some test can be run with additional parameters.
# When a test is listed here, the it will be run without parameters
# as well as with additional parameters listed here.
# This:
# example "testName" : [["--param1", "--param2"] , ["--param3"]]
# will run the test 3 times:
# testName
# testName --param1 --param2
# testname --param3
"txn_doublespend.py": [["--mineblock"]],
"txn_clone.py": [["--mineblock"]]
}Then use the flags to pass through to the daemon if running in --enable-magnetic-anomaly or something. It might be worthwhile to make this pass through to all tests and pass through to the daemon? We could do this directly in the test framework. | |
Test both before and after fork activation, with minimal changes and no duplicate tests
| test/functional/test_runner.py | ||
|---|---|---|
| 77 ↗ | (On Diff #5540) | Fixed in the latest revision. |
Only set activation time when it's specified. Lo and behold it revealed a bug in the test, which is now fixed.