-
Notifications
You must be signed in to change notification settings - Fork 161
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Cleaning up grpauto.tst and other tests #964
Comments
In #763 I manually changed the I am wondering, though, would it be possible to have an automatic and visible test for without packages for a PR? Then one could see immediately if some of the code breaks tests without the packages. Before I never even thought about checking my code without packages, because I thought travis is taking care of all tests. (And btw, why exactly do we test for without packages, if GAP is shipped with some default packages, anyway?) |
@hungaborhorvath : I've looked at the line which prints Testing GAP with/without packages is important due to complicated dependencies and not all packages working everywhere. There are no fundamental obstacles to do it as you suggest on Travis - the limitation of Travis tests is due to its restricted resources, and our willing for a faster turnover: you will have to wait twice longer for the test to complete if you will run each configuration with/without packages on Travis... |
@hungaborhorvath P.S. and of course we need a way to check that packages does not break the behaviour of the core system, so we need a reference run of the tests without packages. |
I have removed some more heavy tests in cf1c2fb, dcaa4b2 - now This means that we're able to merge and test further PRs, and I am going to merge #763 by @hungaborhorvath now. However, the "alternating recognition message" is still there when GAP is started with I've removed the dependency of this issue on benchmarks, as those may be independently continued in #966. |
I'm not fighting for the level of My recommendation would be to rather have the tests deliberately turn InfoPerformance (maybe even InfoWarning) to zero to ensure they do have to bother with such information. |
@hulpke - we have mechanism for that: I suggested already to use START_TEST and STOP_TEST to play with InfoLevel elsewhere, but you haven't commented on that. For now, only |
|
Nightly tests in the master branch are failing after #896. I am creating new issue to continue discussion here instead of commenting on a merged PR:
testmanuals
with default packages : done in 5ca402f#I Alternating recognition needed!
messages whentestinstall/teststandard
run with no packagesgrpauto.tst
run by simplifying tests or moving hard cases to benchmarks or other tests: Minor cleanup for group isomorphism #965, cf1c2fb, dcaa4b2The text was updated successfully, but these errors were encountered: