WIP: Port fusing from 2.2.x to master
- Port weight handling to !256 (merged)
- Test exclusion of decay b-quarks
- YAML read-in of HOOK parameters
- On rel, skipped Sudakov veto (NLO) is implemented by Qcut -> \infty (should be 0, but for technical reasons achieved with \infty). Have to check on master, whether we can use Qcut -> 0, and still get correct skipped Sudakov veto (NLO).
Dire currently provides ME cluster amplitude natively (from perturbative interface, is only extended by this MR. CSS is built from scratch. Should unify?
- Dire and CSS currently order their cluster amplitudes (already at ME level) reversely, which is taken into account in L92 in fragmentation hook. Should this be unified?
- Document 3 new MEPS cluster modes: 512 (for ttbb direct to cluster back to 2->2 even though no ME), 1024 2048
- Validation in Zbb with CSS: currently [email protected] doesn't look very good to start with, but fusing looks consistent.
- Validation in Zbb with Dire: already Zbb 4F doesn't work, many spikes, to be debugged.