Re: Motion P1788/M0029.01: Level-3-interface-only --- discussion period begins
Lee Winter wrote:
So possibly SEM is good enough for purely exchanging data between P1788
implementations. But this may make it real difficult to ensure the NaI
information is then propagated correctly by using existing IEEE 754
hardware
once an actual computation is commenced.
I disagree. Your conclusion above appears to be based on the
assumption that the external encoding has some influence upon the
operations over the internal encoding. I suggest that that the
standard should explicitly deny any such influence.
I agree that making external representation of NaI information vs. internal
propagation of NaI information as separate parts of the standard may be one
feasible solution. If they can be made the same, however, I see this could
also be useful. For example, it may allow binary exchange of large arrays of
intervals between applications in a format that is also suitable as an
internal representation... so extra layers of translation and conversion
could in many common cases be avoided.
Does any of this help shed light on the issue?
Yes, but I may still be missing some factor or issue that mandates the
influence mentioned immediately above. So I am interested in further
exposition.
I didn't mean to imply there is a mandate. I'm only making the observation
that there may be some advantages to having a public interchange format that
could also be used as an internal format for what will probably be many
common implementations (i.e., where the internal format is a pair of IEEE
754 datums).
It may be useful to divide the discussion into indepedent topics, one
topic being the external representation of interval data, for which
NaN payloads need only be represented, but not propagated, and the
other topic being the propagation of of tag information during
operations on internal representations, for which NaN payloads are
merely one possible representation.
I agree that would be useful.
Note that I am opposed to the
proposal for using NaNs to encode tags, but I support the notion of
using payloads to annotate NaNs that arise during (lack of)
initialization and later processing.
My mind isn't fixed on any of these issues, so I like to hear the pros and
cons of both points of view.
Thanks for the detailed reply.
Likewise.
Nate
Lee Winter
NP Engineering
Nashua, New Hampshire
United States of America (NDY)