In D119#3244, @jback wrote:

Yes, adding the const functions would be good, then we can simplify retrieving the resonances for the total amplitude as suggested. What is the specific git commit I should use?

• Reverted obsolete testing .json file sufixes.
• Added entries into history and readme files.
• Updated reference web sited for Photos and Tauola in readme file.

• Find Sherpa version and use it to control use of DECLARE_GETTER macro
• Implemented support for fsrGenerator as argument for testing file. Implemented modification of output and ref files if FSR is on.
• Implemented Tom's comments

• Fix problem with GCC13 compilation of EvtSherpaRandom

• Add option for saving GCC temporary files (commented out for now)
• Fix problem with GCC13 compilation of EvtSherpaRandom

• Moved internal Sherpa configurations for RNG interface into a dedicated cpp file.
• Adapt setupEvtGen.sh to new Photos and Tauola URLs
• Update CI and Doxygen configs to include Sherpa
• Disable Sherpa in HepMC2 build
• Implemented comments by John.

• Moved internal Sherpa configurations for RNG interface into a dedicated cpp file.

Thanks for the replies tlatham. I am happy with the implementations of them. Since the results of the tests look, as expected, numerically identical, except for the few cases where the changes are understood (and where the models need anyway fixing), I think that this differential is ready to be landed.

I spotted a few places where we could get rid of commented code and I also have a few questions out of curiosity. Presumably, the output of the tests will be identical with the master branch. In which case, I think that this differential is ready to be landed.

As discussed in !16, this old switch was currently still used only in examples and hence it is fine to remove it.

Closing this revision after the changes have been landed in the commit 71807eba8af5.

• Updated History.md

• Removed training space in History.md

I added a few points to consult with Sherpa experts as we discussed.

Thanks for checking! Ok, will land this differential.

• Reverted doRadCorr function to not be const for semantic correctness in Vincia and Sherpa interfaces.

This was solved in D107 and D108.

The changes look good to me.

I have search for each removed model and class via grep in the MR branch and found entries only in the history and release-note files. Therefore, I think that an explicit mention of the removed models and classes in the history file might be useful. I also checked whether the deleted models are used at Belle II or LHCb and it seems they are not using them. I remember that @kreps mentioned that PROPSLPOLE is used at LHCb, but I did not find any decay file in the DecFiles project.

Rebased master with check fix.

The changes look good to me. Ok, understood that this is just avoiding to use an user defined CMAKE_CXX_STANDARD if it is lower than our standard.

This looks good to me. I only have a few minor comments, which you find below.

In D99#2612, @tlatham wrote:

Thanks for this @abudinen. Do we want to add some cases for hadronic final states?

• Updated FSR testing scripts: removed empty histograms, improved binning.

All changes look plausible to me.

• Avoid printout at every step in the scan to find probmax. Print info only at the end of the scan.

• Changed names of some variables to make it consistent with model name.

I also checked that the numbers are consistent with the ones from the presentation. It might be a good idea to add a comment in the code that the amplitude magnitudes are obtained from the mentioned BaBar paper but corrected for normalisation.