The impact of nova eruptions on the long-term evolution of Cataclysmic Variables(CVs) is one of the least understood and intensively discussed topics in the field. Acrucial ingredient to improve with this would be to establish a large sample of post-novae with known properties, starting with the most easily accessible one, the orbitalperiod. Here we report new orbital periods for six faint novae: X Cir (3.71 h), ILNor (1.62 h), DY Pup (3.35 h), V363 Sgr (3.03 h), V2572 Sgr (3.75 h) and CQ Vel(2.7 h). We furthermore revise the periods for the old novae OY Ara, RS Car, V365Car, V849 Oph, V728 Sco, WY Sge, XX Tau and RW UMi. Using these new dataand critically reviewing the trustworthiness of reported orbital periods of old novae inthe literature, we establish an updated period distribution. We employ a binary-starevolution code to calculate a theoretical period distribution using both an empiricaland the classical prescription for consequential angular momentum loss. In comparisonwith the observational data we find that both models especially fail to reproduce thepeak in the 3 - 4 h range, suggesting that the angular momentum loss for CVs abovethe period gap is not totally understood.