Remove globalTimeOffset getter and GlobalTimeOffset property from GaudiCommon

GaudiCommon adds a property GlobalTimeOffset and a corresponding getter globalTimeOffset to every algorithm/tool which uses GaudiCommon, regardless on whether this property is used.

As there is no code anymore that actually uses globalTimeOffset, this patch removes this accessor and the corresponding property.

Fixes GAUDI-1122.

Validation started with lhcb-gaudi-merge#22

Reassigned to @clemenci

I've updated some of the LHCb code that was propagating globalTimeOffset: https://svnweb.cern.ch/trac/lhcb/changeset/196988 https://svnweb.cern.ch/trac/lhcb/changeset/196989 https://svnweb.cern.ch/trac/lhcb/changeset/196990 I'm sure there is at least one more, namely HltLine ;-) If you consider this code, you'll realize the amount of code duplication there is in order to 'properly' create (sub)algorithms that 'inherit' some property of their parent... (as together with the 2 in Gaudi, the count is now at 5 times 'the same')

As Boole (and maybe unknown online applications not in the nightly build) still uses globalTimeOffset, pending the discussion on GAUDI-1122, I have, for the time being, reverted the three changes above...

Validation started with lhcb-gaudi-merge#28

Added 767 commits:

• 2d73809e...28a2de5c - 766 commits from branch gaudi:master
• a931169f - Merge branch 'master' of ssh://gitlab.cern.ch:7999/gaudi/Gaudi into GAUDI-1122

Validation build lhcb-gaudi-merge#28 seems to have failed due to a merge failure, eg.

error: expected primary-expression before '<<' token
 <<<<<<< HEAD

Presumably this was due to the fact that the patch didn't apply cleanly anymore. I've thus merged this branch with the current master branch, and I've also just committed the changes to Boole so that it no longer uses globalTimeOffset().

So I propose to start a new validation build...

Validation started with lhcb-gaudi-merge#30

Added 1 commit:

• f14718ca - remove globalTimeOffset from GaudiParallelizer

Validation started with lhcb-gaudi-merge#32

I've now also updated Kernel/SelectionLine

@lhcbsoft: Given that there have been a number of updates & fixes to the LHCb stack, could you please start another validation build?

Validation started with lhcb-gaudi-merge#33

Added 1 commit:

• 7581426a - update reference files

Added 3 commits:

• 7581426a...8119d23e - 2 commits from branch gaudi:master
• 83a6bf82 - Merge remote branch 'Gaudi/master' into GAUDI-1122

please note that the LHCb code, in various places which create subAlgorithms, doesn't propagate the value of globalTimeOffset to the newly created subAlgorithm -- but the accessor to globalTimeOffset still exists. I've removed all existing and known usage of globalTimeOffset from the LHCb code, but the current situation is somewhat brittle, as any new code, or LHCb code not in an LHCb repository using globalTimeOffset may get a stale value. Hence I'd very much like to see this patched merged, as that closes this loophole (or time bomb, depending on your level of paranoia)

Milestone changed to v27r1

mentioned in commit e2c585c4

Status changed to merged