Convolution with MRST2004QED via lhapdf gives zero

[Christian Speckner]

All said in description. Although I have classified this issue as WHIZARD defect, the problem arguably lies more on the LHAPDF side than on the WHIZARD side - there is a set of undocumented functions (evolvePDFphoton, evolvePDFphotonM and has_photon) which _has_ to be used for PDFs which include the photon as a parton.

I will fix this in any case in a branch (as I'd like to crosscheck the builtin version), and I propose to merge the fix into the mainline, but I'd like to discuss this first here as this will raise the bar for LHPDF to version 5.7.1 (previous version do not offer has_photon, making a general fix impossible). So, what's your opinion on this?

[Juergen Reuter]

We discussed this also on phone (WK,JRR,CS). As we anyhow check for LHAPDF version we can disable corresponding PDFs for photons for older versions, and check for the has_photon explicitly.

[Christian Speckner]

Done and merged in r2907. As a sidenote, it turns out that we cannot use LHAPDF versions earlier than 5.3.0 anyway since they do not support getxmin and friends, so I am catching this in configure.