Re: A motion on I/O
This is maybe a bit on the side of this motion, but obviously needs
discussion.
Probably (one of the) primary reason(s) of this motion in interchange of
data, especially if we have the binary interchange format. As I see it,
up until now we more or less agreed on being silent on an
implementation's level 3 representation (like [inf, sup] vs.
[-inf,sup]), which is perfectly alright with the abstraction layer in
this motion.
This being silent is probably also alright with the exception mechanism;
Motion 8 merely recommends to limit the number of trits to support using
NaN-payloads for the storage. Now comes the question: What do we do with
the---to be decided upon---trits (or octagits or what have you) in IO?
We obviously need a standard way to represent them in IO. But we
probably have to wait for an explicit motion on this until the actual
form of the decorations is decided...
Cheers,
Christian