Shunping Huang

I see. It was red for the last two weeks and flaky before that too.

Great. Thanks @liferoad

The task `:runners:flink:1.18:runQuickstartJavaFlinkLocal` has failed to run due to this commit, which leads to the recent failure of our github action "https://github.com/apache/beam/actions/workflows/beam_PostRelease_NightlySnapshot.yml". Below is the latest error from the log:...

> @je-ik PostRelease test is different from other tests. It downloads the latest Nightly Snapshot and run some validation pipelines. It does not build beam on the current branch, and...

Looks like someone have reported this issue on Flink 1.18 before. - https://www.mail-archive.com/[email protected]/msg52035.html - https://www.mail-archive.com/[email protected]/msg52467.html

Hi @cpovirk, thanks for reporting and fixing the issue. I will get back to you on the review shortly today.

By running [code search in github](https://github.com/search?q=repo%3Aapache%2Fbeam+%2Fnew+DateTime%5C%28.%2B%5C%29%2F&type=code&p=1), I found that there are a few more places where DateTime(Object) is called (Note that the code path will be triggered if we pass...

LGTM for the change here. We will have a follow-up PR to change the other occurrence in the repo.

> Have you confirm that DateTime(instant.getMillis()) supposed to return the same result as DataTime(Instant) for any locale? Joda.Instant is merely a wrapper of a long value which represents the number...

> My test suggests otherwise: > > ``` > public static void main(String[] args) { > Instant ist = Instant.now(); > DateTime dt1 = new DateTime(ist.getMillis()); > DateTime dt2 =...