Nick Vatamaniuc

> I'd prefer the {view_row, #{}} approach myself, it will help us clearly lay out clauses in receive code in future, unless we will also add some guaranteed key/value in...

> Originally, the problem what this PR wants to solve is about the case when streaming of view rows is "abruptly" terminated in response to the stop in the callback...

> I would not call those cases "abruptly terminated" because there is a chance to gracefully send out all the required messages. Yeah definitely. In fact, I am not sure...

> Mind we are switching to the discussion of the Mango rolling execution statistics here, which is only a possible application. Note that the execution_stats message is only used by...

> Right. But is this a real problem here? The folding of Mango execution statistics hinges upon addition, which is a commutative and non-associative operation, which even lends itself to...

I noticed this CI failure in a full CI run. It seems to be in the test code in the replicator: ``` [2024-03-15T00:36:32.126Z] couch_replicator_scheduler_docs_tests:70: scheduler_docs_test_prefixed_db_test_ (t_scheduler_docs_total_rows)...test/eunit/couch_replicator_scheduler_docs_tests.erl:137:: Error = [2024-03-15T00:36:32.126Z] test/eunit/couch_replicator_scheduler_docs_tests.erl:138::...

It might be just exposing a flaky test. I'll re-run the CI and we'll see if we get the same error on the same os/arch combination. The difference from the...

Indeed it seems to be a flaky test. After re-running the CI I got the same failure on ubuntu 22.04 https://ci-couchdb.apache.org/blue/organizations/jenkins/jenkins-cm1%2FFullPlatformMatrix/detail/main/1042/pipeline/233 I wonder if it's worth reverting the behavior to...

Based on its shape there is a chance it's coming from a remote worker node, so it's an error that serialized and sent back as a string.

+1 to check the length and maybe also ensure it has only alphanumeric characters, no escapes, slashes, semicolons, etc.