| Thread Links | Date Links | ||||
|---|---|---|---|---|---|
| Thread Prev | Thread Next | Thread Index | Date Prev | Date Next | Date Index |
Michel Hack wrote:
Finally, I disagree with the intent to drop bare decorations. Those are useful when there is no point in continuing to compute with the interval portion. For ill-formed it probably makes little difference because the same test is (almost always) needed before every operation -- but if some of the other decorations (unbounded, possibly undefined etc.) are deemed to pollute all derived results anyway, it would be nice if the rest of the computation could be skipped entirely, just propagating the decoration.
Exactly.If P1788 abolishes bare decorations, this basically takes us back to the days of pre-Motion 7. In that case, most of the work done on decorations will have been in vain, since branch-and-bound algorithms require the decorations to propagate in the NaI payload to function properly and efficiently.
This is also why it is important that a construct like bare(NaI) = Empty should not exist. Otherwise the bare decorations propagating through a branch-and-bound algorithm can get silently "absorbed". So it would basically open a way for users to (unintentionally) circumnavigate the whole exception-handling system and cause the branch-and-bound algorithm to fail in unexpected or catastrophic ways. This is why Motion 8 specifically underwent a revision before voting to make sure this could not be allowed. I'd be very unhappy to see P1788 change its mind and reverse all of this. =(
Nate