This has high impact on eval.
Use case example from
First table input. concerning field is showing as number(9,2), according to database. Records value in database for this column are 10.8 and 10.85.
Preview is showing 10.8 for both rows. Creating checksum gives same checksum. Also using this field as key match in join etc. gives a match althought db values are
different. Only adding a select values step after input and using 0.00 in meta-data format will give proper preview and different checksums and matches as I should
espect because db values are different.
Other use cases:
- previewing data from a database table with more digits
- Exporting data to Excel/TextFile
- Apply the default format depending on the source data type (minimum is 2)
- When it is a Number (floating point) we need to spike on how we can keep all digits in here
- Have a backward compatibility flag
- Have it documented in the Upgrade Guide
- Put all default values in the kettle.properties explicitly
- Need very clear errata on any change to default values
- Need to look at all format mask defaults for sanity (the big-number format for example)