NA values and Msgpack Timestamp type

[traversc]

No description provided.

[eddelbuettel]

Love it, as I love time series support.

Have a look though at RcppCCTZ. I do something similar there with nanoseconds. In function parseDouble() in file src/utilities I think started with the round() approach but later whent to truncation as is done now.

We can iron that out in another round.

[traversc]

Hi Dirk, from the parseDouble function, I see this:

        auto nanoseconds = tp.time_since_epoch().count();
        auto secs = nanoseconds / 1000000000;
        auto nanos = nanoseconds - secs * 1000000000;

How would we adopt that here?

[eddelbuettel]

I think so. That is what I had in mind, and there may be old cases / demos / ... in the issues of the previous approach with round() failing. I think this using integer math only and is more robust?