Updated references for S28, S28r1 and S28r1p1

Updated references for S28, S28r1 and S28r1p1, added S28 to Phys/StrippingCache/CMakeLists.txt. The tests passed locally.

@erodrigu Let's hope this runs smoothly in the nightlies. I will need to update the refs again when the new AppConfig will be deployed.

Indeed let's hope! Just one thing, though: the error message spit out to stderr for at least one of the jobs seems to come from an issue in the configuration of the B23h line. I do not see any code change addressing that in this new commit. Are you sure there is no need to touch the line? Or in fact you did, in some other commit elsewhere I am not being alerted about? Thanks.

If you mean Adlene's line in S28, this was fixed by adding in S28 into Phys/StrippingCache/CMakelists.txt.

Yes that's what I meant. Thanks a lot for the clarification.

• [2017-11-23 00:14] Validation started with lhcb-2016-patches#582
• [2017-11-24 00:14] Validation started with lhcb-2016-patches#583
• [2017-11-25 00:13] Validation started with lhcb-2016-patches#584
• [2017-11-26 00:12] Validation started with lhcb-2016-patches#585

I think things are ready to go from checking the failures in the nightlies. Do you agree @sreicher?

I mean I can then merge?

Yes, please. I'm sorry for the late reply but I was travelling back to CERN today.

@erodrigu We've decided to include two more changes we received this morning, which means updating the refs again.

OK, thanks for the info. Then I stay on hold ... BTW, this morning the refs were not perfect either and 2 of the tests contained a trivial difference.

Yes, this is due to a symlink in the nightlies configuration.

added 1 commit

• ece47007 - Updated ref for S28r1p1 after final fixes

Compare with previous version

@erodrigu The failures of the tests of S28 and S28r1p1 are expected (symlink to TMVAWeights v1r999 points to v1r9). For S28r1, the test ran with the latest version of AppConfig v3r344, which removed the MCwNoPID lines from B2OC, and this is also expected. I will update the reference today but would not have managed to do this before the start of the nightles yesterday. AppConfig was deployed yesterday night, which I didn't expect to be honest.

added 1 commit

• aa445f41 - Updated S28r1 tests with AppConfig v3r344, modified TMVAWeights version to match tests

Compare with previous version

Shall I go ahead and merge? Or shall I wait till tomorrow?

Please wait until tomorrow. We received yet another bugfix, and will add also S31 and S31r1 (separate merge request).

OK.

added 1 commit

• 5c1c4f86 - Final update for S28r1p1 refs

Compare with previous version

added 1 commit

• 7886a679 - Final update for S28r1p1 refs including change of TMVAWeights version

Compare with previous version

This is now really final. Sorry for the mess with the thousand bugfixes.

The tests for S28, S28r1 and S28r1p1 have all passed in the nightlies.

mentioned in commit 18817070

merged