Jan Fajerski

> Sorry for confusion. An almost similar PR was created recently, see #14958, and some discussion happened over there. The discussion over there is relevant, but the changes should happen...

@ansh-devs yeah I think you can contribute to this. Please reach out to @SuperQ, he might be working on this already. Join us on slack for a quicker discussion.

yes this would be great to have in 3.0. Ideally along with the fixes @machine424 has proposed for #12858 and #12436.

@machine424 Is there anything else to do here? Iiuc this issue can be closed, right?

@pawarpranav83 Could uou resolve the conflicts here, then we should be good to go.

@alexgreenbank will shepherd this.

Seems like https://github.com/prometheus/prometheus/issues/12631 is a prerequisite for this? Could use case specific documentation and respective example configs be a path forward here?

Thanks for the update. I'll remove this from the Prometheus 3 project then.

We should also consider the behavior of the `@` modifier. Based on a quick test this also accepts future timestamps and returns data as described. Going by the documentation its...

> How would someone figure out what change is required? This PR introduces and additional job that `needs` the build_all job and it fails or succeeds based on the result...