ttcn3-msc-test regressions since build 1732, 17 days ago
This may have come up before, and IIRC there was some discussion that it's something you're working on, neels ? If my memory is wrong, please re-assing.
It is worrying that for 17 days we see regressions in ttcn3-msc-tests
https://jenkins.osmocom.org/jenkins/view/All%20no%20Gerrit/job/ttcn3-msc-test/ as well as the same regressions in msc-test-latest: https://jenkins.osmocom.org/jenkins/view/All%20no%20Gerrit/job/ttcn3-msc-test-latest/
Any regression is normally discussed in the weekly review and then resolved quickly. It is not our normal workflow to introduce changes into either the IUT (msc) or the test suite that knowingly cause regressions for more than a few days.
Updated by fixeria over 1 year ago
If fixing the missing IE is a lot of work, we can temporarily revert this patch:
Updated by neels over 1 year ago
It is valid to have unit tests that are expected to fail.
At least it is common practice: there is an indicator that some probem is known, and not yet fixed.
Hence I am fine with the tests failing, no matter how long it takes to fix them.
Fixing the issue itself has already happened on the osmo-msc neels/codecs branch,
but before i can submit it for review, i want to do physical testing -- our only way to ensure a proper voice stream.
I hope to be able to start physical testing next week.
It would also be possible to spend additional effort to also fix the missing IE on current master, without waiting for the branch. Currently not deemed necessary, see #5529