WIP: Added fallback for VELO tracks' cov matrix in DTF

Currently, the Kalman step in the DTF fails regularly for velo tracks. This will resolve it, partially. (still a good initial value for the momentum at the first state is advisable)

This should likely be ported to the 2018-patches branch as well.

@wouter

Hmm ... have you checked https://gitlab.cern.ch/lhcb/Phys/blob/master/README.md or remember my email to the DaVinci mailing list about where to commit code to ... ;-)?

Yes, I realised too late that I should've rebased to 2018-patches, and not the other way around :) Let me rebase this and reopen.

News ...?

closed

See !395 (closed) (run2-patches MR)