[AVRO-4043] Add time-nanos logical type

[glywk]

AVRO-4043

What is the purpose of the change

This pull request describes the time of day specification with a nano seconds precision as requested in AVRO-4043

Verifying this change

This change is a specification description.

Documentation

• Does this pull request introduce a new feature? (yes)
• If yes, how is the feature documented? (docs)

[opwvhk]

The consistency is good (minor nitpick: the anchor should no longer have the _ms suffix) . Thank you.

However, increasing the granularity of a timestamp a 1000/million-fold also reduces its range by the same factor. As such, I'd like us to document the possible maximum timestamp values as well.

[martin-g]

Changes to the specification should be discussed at [email protected] first.

[glywk]

Changes to the specification should be discussed at [email protected] first.

Sorry, I ignore this aspect. How to register to [email protected] to submit this proposal?

[glywk]

The consistency is good (minor nitpick: the anchor should no longer have the _ms suffix) . Thank you.

However, increasing the granularity of a timestamp a 1000/million-fold also reduces its range by the same factor. As such, I'd like us to document the possible maximum timestamp values as well.

I have kept 'time_ms' link to preserve backward compatibility. But I will remove it.

[martin-g]

How to register to [email protected] to submit this proposal?

You can click on "Subscribe to mailing list" at https://lists.apache.org/[email protected]
Or directly send a dummy email to [email protected] and then just Reply on the confirmation email.

[glywk]

Changes to the specification should be discussed at [email protected] first.

Hi @martin-g, I'll let you discuss among yourselves whether you want all time scales to support nanoseconds or not.
Time is just a timestamp subset. As request, I have also add timestamp limit for each scale. This PR can help you in your decision.

There is also the duration logical type but it involves a breaking change.

Regards

[martin-g]

Hi @martin-g, I'll let you discuss among yourselves

There is no such thing! Everything is open at the ASF and should be discussed in the mailing lists where everyone could participate!
I have no personal interest in this feature, so I won't lead a discussion. If you need it then you need to explain why it is useful and and convince others that the specification needs to be updated.

[martin-g]

I am going to set all of your PRs related to time-nanos to Draft state, so they don't get merged accidentally before the discussion.