schema maximum value for samples_start, sample_count, others is awkward

[gmabey]

I'm working to use https://github.com/json-schema-org/json-schema-spec to validate SigMF metadata.
I'm finding that the schema-dictated maximum value for several properties to be very problematic, and I wonder if there's a good rationale for the current value:

18446744073709552000 = 2^64 + 384

I would like to suggest that the value 2**63-1 be used instead, as this makes it much easier for libraries like json-schema-spec to implement such a constraint.
This would cause a breakage for all those files out there with values > 9223372036854775807, it's true. But, given the number of elementary particles in the universe, I don't think that's going to be a problem in practice.
Like, it's one thing to enforce such a limit in python; it's something else to do so in generic C++. The implementation of json-schema-spec uses long to hold such a constraint, on a JSON element of type unsigned int (that's a nlohmann/json thing).

What is the origin of 18446744073709552000 anyway?

[gmabey]

In the interest of two witnesses, I used this on-line tool to generate some fictitious .sigmf-meta data: https://www.liquid-technologies.com/online-schema-to-json-converter
When I pasted the contents of sigmf-schema.json into the text box, I get an error: Specified cast is not valid..
BUT, when I pasted the contents of the #335 it was able to produce something.

[Teque5]

Merge when ready.