Release 0.2.125

[danking]

Want to release no later than 2023-10-20.

We have a number of fixes we need to get out to everyone.

• [query] rare Google Cloud Storage error #13721
• [query] automatic retry code in Java is broken #13704
• [query] pipeline triggers type inference failure #13699
• [query] show and checkpoint are very slow in Jupyter Notebooks; they are not slow outside Jupyter #13690
• [query] hl.maximal_independent_set cannot accept strings as node names #13633
• [query] when sending large literals from Python to Scala, Query should use a compact binary format #13757
• [query] Don't use py4j for Backend operations #13756
• [qob] Since 0.2.117 vds.filter_intervals unnecessarily uses a lot of RAM #13748
• [query] ensure nest_asyncio is applied, but only when necessary #13899
• [batch] Update gcsfuse version to 1.2.0 #13728
• [gradle] exclude scala-reflect #13894

After release, we will ask GVS team to verify #13712 is resolved.

[jmarshall]

I would also suggest the following as necessary for an upcoming release:

• [batch] Update gcsfuse version to 1.2.0 #13728

Google's gcsfuse APT repository currently produces 502 Bad Gateway errors when accessed via http, which shows no sign of being resolved any time soon. I've commented on #13728 noting how this PR can work around the problem. At present (since early October), the batch_worker_image job always fails with 502 during a hail deployment.

[danking]

I've added it to the list!