[Rivet] yaml-cpp bundling and immediate Rivet 2.1.0 release?

[Andy Buckley]

On 28/01/14 11:32, David Grellscheid wrote:
> Hi Andy, hi all,
> 
> one thing ahead: I just fixed a nasty projection ordering bug affecting
> Fastjet. All runs that have mixed analyses with/without invisibles in
> the Fastjet clustering are affected. Up to now, the projection applier
> would not take that difference into account, and potentially apply the
> wrong fastjet projection depending on analysis ordering, e.g. using
> visibles instead of invisibles!
> 
> There may be more ordering bugs to come, as I check the cuts branch
> before remerging to trunk.
> 
> Is there an easy way to split a YAML file into individual files with one
> analysis each?

Hrm, not right now. Should be easy to write a Python script to do it via
the YODA interface, though... let me know if you need help with that
(and if you think we should make it standard: the "split" feature and
pattern-matching options in the AIDA converter scripts made them complex
and I ~never used them, but maybe that stuff was necessary.)

>> I merged the yaml-cpp bundling from LHAPDF6 into Rivet's release-2-0
>> branch last night. 
> 
> Excellent! Again with its own symbol namespace Rivet_YAML to avoid clashes?

Yep.

[Nerd detail: the awkwardness is that with libtool the symbols can only
be imported from the static lib when building a "final" shared lib
rather than a static convenience lib. And when you do that, libtool
brilliantly writes "-lrivet-yaml-cpp" into the dependency_libs entry in
libRivet.la -- a dependency condition which neither needs to, nor can,
be met when compiling any program against libRivet. We got away with
this in LHAPDF because the test programs are not built using libtool,
and hence the .la is ignored. But anyone building an executable or lib
against LHAPDF6 via libtool would hit this problem: I will add the
libtool-kill-dependency_libs.sh hack that I found online to LHAPDF6 as
well.]

>> The remaining things on my 2.1.0 release TODO list are:
>>
>> * check/force correct passing of CXXFLAGS, etc. to the bundled cmake
>> (CXX is working, other vars aren't as far as I can tell)
>> * check/fix the JADE/OPAL integration boundaries as advised by Christoph
>> Pahl
>> * incorporate latest CMS analysis
>> * remove IsolationProjection and IsolationTools
>> * migrate the remaining 3 analyses from 1.9.0 which require non-trivial
>> histo code conversion
>> * check the broken (?) D0 W charge asymmetry. I saw no asymm from the
>> H++ validation events when we did the Rivet 1->2 validation.
> 
> * include the 3 bugfixes Peter made on the AIDA branch

And I also saw that you had filed a bug report about HeavyHadrons --
does that need to be solved for 2.1.0? (I guess yes, but haven't looked
at details.)

Andy

-- 
Dr Andy Buckley, Royal Society University Research Fellow
Particle Physics Expt Group, University of Glasgow / PH Dept, CERN