This is a proposal to standardize indication of an incomplete data set. This was discussed a bit in 573. Common cases we have for incomplete data include:
a limit was reached (for example we cap hisRead to 10K by default)
a timeout was reached (for example processing took longer than 1min)
Typically in user facing applications you want to visually indicate the data set is not complete.
A parallel to this is error grids. In that case we use grid meta for all the error data. But in the case of incomplete data, I don't think we want to take over all the grid meta since there is application data too.
So I propose to standardize using a incomplete tag in grid meta where the value is nested dict with the following tags:
dis: Str message to display to the user about why data is incomplete
limit: Number if incomplete because of threshold limit
timeout: Number duration if complete because of timeout
Brian Frank Tue 1 Oct 2019
This is a proposal to standardize indication of an incomplete data set. This was discussed a bit in 573. Common cases we have for incomplete data include:
Typically in user facing applications you want to visually indicate the data set is not complete.
A parallel to this is error grids. In that case we use grid meta for all the error data. But in the case of incomplete data, I don't think we want to take over all the grid meta since there is application data too.
So I propose to standardize using a
incompletetag in grid meta where the value is nested dict with the following tags:Example in Zinc:
ver:"3.0" incomplete:{dis:"Limit hit!" limit:1000} colA, colB, colC ....Comments?
Michael Whalen Thu 3 Oct 2019
+1