It seems that it's better to wait until Decimal256 implementation will be
merged in master, before starting to implement low-bit Decimals. Decimal256
should provide a reference for how we should define new Decimal types at C++
part.
Also, as it was mentioned in this [1] comment, the introductio
Hi Dmitry,
Thanks for volunteering to contribute. Note that we are in the process of
implementing Decimal256 support already, which is currently on a separate
branch [1] but I'm hoping to have PR sometime early next week. If we are
proposing adding support for these lower bit-widths, I think we
Based in what I've seen in other systems (like Apache Kudu) that
support at least 32/64-bit decimal, representing them with a single
integer value is probably the best thing (in terms of computing
performance, consistency with other implementations)
I added you as a contributor on Jira so you can
Hi everyone, I would like to work on this JIRA ticket:
https://issues.apache.org/jira/browse/ARROW-9404
([C++] Add support for Decimal16, Decimal32 and Decimal64)
This will be my first experience with contributing to Arrow, so I want to ask
advice what approach should I use.
As far as I know, cur
