...Talk about a pointless distinction…
The two supposed inconsistent statements are, functionally, semantically consistent..
Original claim translates as: When you discover defects the cost of fixing the defect is higher, the later you detect it...i.e. proportional to the additional work done since the defect was inserted.
Second claim is not identical, but is a natural derivation of the first based upon what it means… i.e. IF the foregoing analysis is true, then, in post release you can predict the cost of defect repair by determining how early in development the defect occurred. The properly and pertinently derived meaning not explicit in the first statement is that Not all bugs discovered in beta testing will cost the same to fix, and you can expect cost to be higher the earlier it occurred.
Because post- release is effectively the “Later”-est you Can detect… It is entirely sound as a conclusion.
Thus there is no inconsistency, the citations are, in fact, solid and correct.
And the analogy the author presents is not even remotely reflective of the problem he imagines to exist.
the prior study may well be methodologically flawed, but on the assumption it’s conclusions we’re valid, the subsequent citations were correct interpretations of its effect on the cost of fixing bugs discovered After release.
So if you came across this story in the field of infectious diseases, you would accept the later reporting as unproblematic, and not update at all in the direction of doubting the standard public health policy?
ETA: welcome to Less Wrong, by the way! Don’t let the downvotes (neither of them mine) put you off.
They are very distinct. They intersect at one point—long-standing bugs. One addresses bugs caught earlier and earlier in the development process, and the other bugs introduced later and later in the development process.
Even as they are both live for less time, it shouldn’t be too hard to see how they could wreak different amounts of havoc. Or not! We don’t know! There’s no data on it!
...Talk about a pointless distinction… The two supposed inconsistent statements are, functionally, semantically consistent.. Original claim translates as: When you discover defects the cost of fixing the defect is higher, the later you detect it...i.e. proportional to the additional work done since the defect was inserted.
Second claim is not identical, but is a natural derivation of the first based upon what it means… i.e. IF the foregoing analysis is true, then, in post release you can predict the cost of defect repair by determining how early in development the defect occurred. The properly and pertinently derived meaning not explicit in the first statement is that Not all bugs discovered in beta testing will cost the same to fix, and you can expect cost to be higher the earlier it occurred. Because post- release is effectively the “Later”-est you Can detect… It is entirely sound as a conclusion. Thus there is no inconsistency, the citations are, in fact, solid and correct. And the analogy the author presents is not even remotely reflective of the problem he imagines to exist.
the prior study may well be methodologically flawed, but on the assumption it’s conclusions we’re valid, the subsequent citations were correct interpretations of its effect on the cost of fixing bugs discovered After release.
So if you came across this story in the field of infectious diseases, you would accept the later reporting as unproblematic, and not update at all in the direction of doubting the standard public health policy?
ETA: welcome to Less Wrong, by the way! Don’t let the downvotes (neither of them mine) put you off.
They are very distinct. They intersect at one point—long-standing bugs. One addresses bugs caught earlier and earlier in the development process, and the other bugs introduced later and later in the development process.
Even as they are both live for less time, it shouldn’t be too hard to see how they could wreak different amounts of havoc. Or not! We don’t know! There’s no data on it!