| Thread Links | Date Links | ||||
|---|---|---|---|---|---|
| Thread Prev | Thread Next | Thread Index | Date Prev | Date Next | Date Index |
Arnold Neumaier wrote:
Nate Hayes wrote:Arnold Neumaier wrote:Nate Hayes wrote:That's all sensible, but I don't see it constitutes an advantage: it simply trades one shortcoming for another.I would support adding a new decoration, i.e., expanding the total number of decorations from five to six:D4 = safe (defined, continuous, bounded) D3 = everywhere defined D2 = somewhere defined, somewhere undefined D1 = everywhere undefined on real inputs (e.g., 1/0) D0 = everywhere undefined on empty input (e.g., 1/Empty) DX = ill-formed (bug)Newbie-friendly functions such as isUndefined() can simply return TRUE if the decoration is D1 or D0, and DX is then reserved as you and Jurgen would like as special case for things such as invalid constructions and/or uninitialized data, etc.My proposal allows to split decorations; but the splitting should not be made mandatory. So the standard should not distinguish between 1/0 and 1/Empty, but implementations that care for finer information are free to split it.Thus both you and I can be satisfied.Well, our proposal allows splitting of decorations, as well, so I could say the same to you. But something tells me you wouldn't be satisfied with that for the same reasons I wouldn't be satisfied with what you just mentioned. ;-)The problem is that your old proposal was incompatible with mine, whereas your new one is a refinement of mine. Thus in the old wersion, coexistence was not possible, while in the new version, coexistence is possible with requiring mine (which has the advantage of greater simpklicity and a uniformly defined semantics) and allowing yours.
I agree standardizing the six decorations would bring the two proposals much closer together. Mine is a work in progress and continues to change, as I have seen yours change over the past few months in our private discussions as well. I have always hoped by the end of the discussion process there would no longer be any differences.
But to answer your point: if the six decorations are not standardized, then no applications that rely on the relevant information are portable. I don't think any user with such an application would be satisfied with that, including me. So this is why they should be standardized.
As some of these recent discussions show: refining the decorations is a difficult and controversial task. This is why P1788 should do the hard work ONCE so that everyone else in the world won't be forced to re-invent this wheel over and over, likely with mistakes and incompatible semantics and implementations.
At least, that is my NEW position. :-) Nate