It tends to forbid code duplication, which is what the lungs are. Redundant implementations are two implementations which work a different way, which no, are not forbidden by DRY.
> It tends to forbid code duplication, which is what the lungs are.
Lungs are not code duplication any more than running two instances of an application to provide greater peak performance or hot-standby is code duplication. You are confusing duplication of elements of the real system with duplication of specification (programming is executable specification, not the real system.)
You should've suggested (in the DRY sense) removing extra copies of sequences in DNA--excepting, you know, that I don't think that we know that is safe (those may be there for redundancy as well).
Perhaps you work somewhere where the concept of redundancy is important?
Stop being obtuse; you clearly seem to see the point I was getting at.
I certainly see the argument you were trying to make. I also think the way you tried to make it is totally invalid and casts no light on whether the TSA should or shouldn't use the processes they are planning to adopt.
Principle zero is that the system should correctly perform the work required by its owners. If what you're saying is that an additional system should be injected that has access to all the required information, and the TSA should query that system, then I won't argue. That would be a balancing act of coordination costs versus the cost of the technical debt added to the TSA, but I don't feel I have sufficient insight to analyze that balance where I currently am. Likewise if you are arguing any other rearrangement of information owners such that some process can access information and inform the TSA's decisions.
If what you're saying is that the TSA should not perform some work because it's not simple to do that work in accordance with other system design principles, then I fear you have fundamentally misunderstood what those principles are for.
I withdraw my claim that the TSA should not seek to optimize according to design principles. I still believe that you are applying them too broadly, but it isn't a position I care to defend at the moment.
