When I was working in France, we had the CM66 (Construction Metallique 1966) which advised that the buckling coefficient for diagonals and other struts was to be taken as 0.9 times the length between the theoretical centres of the trusses.
Probably due to the level of fixity that hot riveting gives you combined with the relative rigidity of the main members.
This was concerning angles or back-to back angles. Probably not applicable in other circumstances such as I-beams.