Joshua MacDonald

> If not I would expect it to live in the exporter helper instead of its own config package. I was following `configretry`, which appears to be exporterhelper-specific but resides...

If a processor -- say batch processor -- wants to respect timeouts, won't it need a configuration? I will submit an RFC.

I don't understand how this is a breaking change? I believe this is the intended outcome stated in https://github.com/open-telemetry/opentelemetry-collector/issues/9591 The memory limiter extension is replacing the memory limiter processor: the...

@JonasKunz I am curious to know whether you think it is essential for the on-end handler being discussed in this PR to make changes that are visible to other processors....

@JonasKunz I assumed that you mean for `OnEnding()` to receive a "read/write span object" the same as is passed to `OnStart()`. I think it means that processors can read (the...

Discussed this in the context of #4062

@pellared Would you say that if #4030 is accepted, this PR should be approached differently? (To me, it seems like the answer is yes.) I've speculated about what the solution...

@axw has written drafts for the Collector which would add a similar feature directly to the Collector instead of via SDK integration. I think that is a loss for us,...

@TylerHelmuth I would like to add to the list of considerations that there are a couple of existing places where I think we need a protocol-agnostic error status. Discussed in...

I know that @bogdandrutu himself has called for making the collection interval variable on a per-instrument basis. As you say @Oberon00, I believe this is about configuration in the "Accumulator",...