Nicholas Lusty wrote:
> Both give identical probabilistic definitions of SIL-3 and SIL4.
>
There seems to be a complete disagreement between subscribers to this 
list about whether or not SILs are defined probabilistically. In the 
past weeks, we have seen several unequivocal statements that 61508 does 
NOT define SILs probabilistically. How can something so fundamental to 
safety be allowed to remain unclear?

[If anyone is tempted to say that 61508 defines system-level SILs 
probabilistically whilst stating that probabilities do not apply to 
software, then I challenge them to explain how 61508 suggests that one 
can provide adequate evidence for a system-level probability of failure 
if the system contains software].

Martyn