Re: Motion M0052 Expressions: revised text
> Since NaI is not an interval, it doesn't have an interval part
So what is the set of decorated interval datum ?
a) Cartesioan product of bare interval and decoration
DT = {ill,trv,def,dac,com} x T
b) Disjoint sum of NaI and of another cartesian product
DT = {NaI} + T x {trv,def,fac,com}
I would like (b), but the draft currently says in 11.2 that (a).
Also 11.4 doesn't forbid combinations like [1,2]_ill .
Perhaps, the proof of FTDIA can help to choice between (a), (b).
Ned, can you help to clarify ?
-Dima
----- Исходное сообщение -----
От: mhack@xxxxxxx
Кому: stds-1788@xxxxxxxxxxxxxxxxx
Отправленные: Вторник, 12 Ноябрь 2013 г 17:08:29 GMT +04:00 Абу-Даби, Маскат
Тема: Re: Motion M0052 Expressions: revised text
Dmitry Nadezhin was concerned about:
"The standard requires that at Level 2, for all interval types,
operations and inputs, the interval part of a decorated interval
operation equal the corresponding bare interval operation."
> I think that it is necessary to add some disclaimer to subsection 6.4
> about [NaI].
Since NaI is not an interval, it doesn't have an interval part, so the
issue is moot! Now, at Level 2 we have to return something or throw
an exception. If that "something" does not have an exceptional value
such as NaN for numeric functions, and the environment does not support
exceptions, there is not much we can do other than document the values
actually returned, such as Empty or False. IEEE 754 faced the same
issue for boolean and integer-format results, where (in the absence of
exceptions) the best that could be done was saturation.
This is why we have the belated introduction of a third standardized
exception, for intervalPart(NaI).
Btw, it seems to me that if I have to return an interval for something
that does not exist, Entire would be more appropriate than Empty -- but
nothing really makes sense, and once the NaI propagation chain ends it
seems to me that an exception, if at all possible, should be thrown.
Michel.
---Sent: 2013-11-12 13:02:35 UTC