Björn Rabenstein

Yeah, it's complicated… If we looked purely at the state of Prometheus 2.x and OpenMetrics, then it would appear quite obvious that any future version of client_golang should completely ignore...

And coming back to the actual point: Once we know protobuf will stay a thing in the Prometheus universe, clearly we shouldn't switch everything over to an unmaintained protobuf implementation....

Could the zero value of `time.Time` be used as a marker for “do not use”? (Cf. https://golang.org/pkg/time/#Time.IsZero .) I might be completely wrong. Not an expert of the API client....

Oh right, I now even remember that we had discussed this problem in a different context already (i.e. a valid Prometheus sample might happen to have the time.Time zero value...

So yeah, disregard my comment, please.

Approaches like this were considered. However, I couldn't find a good general purpose interface for that. Very few users will actually need this kind of speedup. I don't want to...

These are great pointers. Thanks a lot. I'm especially glad that finally some movement seems to happen on the topic. I'll leave this issue open and change the title to...

First of all, I'd like to see how relevant use cases are where we need the hack. As you have already demonstrated, you can work around with "buffering" increments in...

I mostly assigned this to myself as a “default” because I used to be the maintainer of this repo. Therefore, I'm un-assigning this from myself no. New maintainers @bwplotka &...

The behavior _is_ configurable via https://godoc.org/github.com/prometheus/client_golang/prometheus/promhttp#HandlerOpts . The gzipping in the client library is very old. It might very well be that no common middleware for gzipping existed back then....