Hey Jeyhun,
- No, it wont disable the grouping/ordering by the column names. One
"disadvantage" is that once the flag is enabled, "GROUP BY 1" will have
a
different meaning and query result.
Ahh I see. Thanks for explaining, it was not obvious for me from the PR
and discussions.
Best,
Muh
>I think so. But this is orthogonal to this issue [1] (deciding whether to
>provide the proposed feature or not), no?
thanks for your response
yep, I looked into the issue however still not clear what problem is
solved with this approach
from my point of view the query there is good to highlight t
Hi Sergey,
Thanks for your comments.
Could you please elaborate more on use cases of this feature?
IMHO, the main use-cases can be
- simplifying some SQL queries (e.g., with queries including many/long
column names)
- having consistency when refactoring (e.g., columns are renamed)
- handling ca
Hi Muhammet,
Thanks for your comment.
Personally, I find this feature confusing, it feels always natural to use
> column names.
- I also have the similar experience w.r.t. using column ordinals.
But it seems counterintuitive to enable flag for this feature. Enabling
it, should not disable grou
Hey Jeyhun,
Thanks for starting the discussion.
Could you please elaborate more on use cases of this feature?
The one that I see in FLINK-34366[1] is to simplify referencing to
aliases in SELECT from GROUP BY
(also potentially ORDER BY and HAVING). I wonder whether there is some
other use cases w
Hey Jeyhun,
Thanks for bringing it up! +-1 from my side.
Personally, I find this feature confusing, it feels always natural to
use
column names. SQL power users will ask for it, I have seen it used in
automated complex queries also.
But it seems counterintuitive to enable flag for this featur
