Bertrand Renuart

AFAIK yes. I noticed ApacheHttp and OkHttp configurations seem to make use of `RequestSpecificRetryHandler` too... but I have not investigated any further.

I understand it is becoming low priority since the move to Spring Retry. However some of us are currently stuck on using Netflix's RestClient until they are able to move...

A few comments already... Lumberjack is not only about the data format, it is also a protocol with window and ACK (like TCP). The ultimate objective is to provide a...

> This means that if the appender needs to reconnect for whatever reason, and acks have not been received, those un-acked events could be lost. I believe the appender should...

Hi @elefeint The visibility of `getFieldValue()` has indeed been reduced - it wasn't meant to be part of the API and should ideally remain internal to the class hierarchy. As...

I like the idea of a LifecycleManager to keep track of which components had to be started. However, the more I think about it, the more it seems to me...

> After thinking more about this, I think this is ok. Give me until tomorrow to think about it once again (it is 1am already here)... Ok?

As proposed in this [discussion](https://github.com/logfellow/logstash-logback-encoder/discussions/762), an option is to decorate the JsonGenerator to make it filter empty fields. This option would be controlled by a new `omitEmptyField` configuration property added...

Returning the input value (that is a _string_) does not make much sense for operations whose purpose is to convert to a type **other** than string. Instead, we could provide...

I made a mistake when describing the behaviour of the `:-` modifier. It is supposed to be an alternative to the `try..{}` variants. It should therefore _always_ return the input...