Github user fhueske commented on the issue:
https://github.com/apache/flink/pull/3641
Sorry, I think did not explain the change correctly.
I did not mean to group multiple record together.
We still emit one row per input row, but given the following input
```
(proc-time, id, a)
(1, a, 1)
(2, b, 2)
(3, c, 3)
(3, d, 4)
(4, e, 5)
```
and the query
```
SELECT id, sum(a) OVER (ORDER BY proctime RANGE BETWEEN 2 PRECEDING AND
CURRENT ROW) FROM x
```
the result would not be
```
(a, 1), (b, 3), (c, 6), (d, 10), (e, 14)
```
and not
```
(a, 1), (b, 3), (c, 10), (d, 10), (e, 14)
```
Because `c` and `d` arrived at the same time, their aggregations need to be
the same.
This is actually also the implementation for event-time OVER RANGE windows.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---
