Björn Rabenstein

There is #9641 , which appears to be abandoned. If anyone wants to pick up that work, that would be great.

@milencium Please, go ahead. PRs welcome.

Looking at this during the bug scrub again. Since there was no follow-up, we'll close this now. However, we would still love to see progress on this. Whoever wants to...

Hello from the bug scrub! @jyz0309 are you still willing to work on this? We'll happily assign this issue to you (or to others that want to help with it)....

Yeah, good question. So the source of dissonance here is that the encoders and parsers in https://github.com/prometheus/common/tree/main/expfmt are all based on https://github.com/prometheus/client_model , which in turn represents the data model...

Inspired by this issue, I had added an agenda item for the dev-summit many months ago. However, it wasn't discussed so far as other agenda items were deemed more pressing....

I would like to note that the concrete use case that served as an example here is by far not the only one where users run into the dreaded `vector...

> Unfortunately, it doesn't work in Prometheus now and will not work after the proposed idea is implemented. If the metric name is dropped as the last thing, you can...

We [discussed this topic during the last dev-summit](https://docs.google.com/document/d/11LC3wJcVk00l8w5P3oLQ-m3Y37iom6INAMEu2ZAGIIE/edit?pli=1#bookmark=kix.1yn3rd8rsq8u). I try to summarize the results here: ### About allowing `label_replace` to work on range vectors This would solve most use cases...

With v3.0 projected for 2024, the above annoyance about requiring a feature flag before 3.0 has become less bad, because 3.0 is going to happen quite soon rather than "at...