On 20/05/17 15:38, José Mejuto wrote: > At least in 3.0 (trunk not tested as I don't have it compiled) the first > and second random numbers are always the same value, so: That is unrelated to this thread, but has been fixed in FPC 3.0.2. Jonas