Björn Rabenstein

Results 1096 comments of Björn Rabenstein

Actually, I'll first look at the design doc to check if we are on the same page there.

As you can probably already guess from my comments on the design doc: This isn't addressing yet the reverse case, using a duration literal as a number. Examples: ![image](https://github.com/prometheus/prometheus/assets/5609886/c834b112-2560-4c7e-98b6-c7848e7eebf2) ![image](https://github.com/prometheus/prometheus/assets/5609886/15e9040e-2a3f-4c67-ba06-c2c6ee8f3d74)...

Thanks for your updates. I'll review ASAP.

> The timestamp is the duration after 1970-01-01 0:00 UTC. So you could say foo @ 30y if you want to point to 2000-01-01 or something… I have to correct...

> > Do you know of any other cases where a number cannot be replaced by a duration? > > There should be no such place now, and we should...

The final outcome has to be that `NumberDurationLiteral` is used instead of `Duration` or `NumberLiteral` everywhere. Maybe those two can disappear completely, and `NumberDurationLiteral` is just defined to include both...

Not sure if it is worth the effort to distinguish those cases. `histogram_quantile(0.99, rate(rpc_latency_seconds[5m])) > 500ms` is perfectly reasonable, but to predict that, you needed to encode a lot of...

On the other hand, it's probably fine to only do the autocomplete where there is a duration for sure. You can still type things that the autocomplete didn't suggest, so...

@Maniktherana I assume you still need some answers for the parser part. @Nexucis could you help out @Maniktherana ? He spelled out his plan above and it appears he would...

Hello from the bug scrub. This sounds after all like a legitimate feature request, so we'll be open to PRs adding the feature.