Review of open issues for compatibility with other prometheus clients in other languages.

[jdmarshall]

For a 1.0 release it would be useful to get a notion of which of the 80 open issues on client_js are asking for features that exist in sibling clients and put some effort into prioritizing those.

Further, any features that we have been on the fence about, that are in direct opposition to decisions made in the other clients, are probably safe to close as Will Not Fix, since we already haven't invested in them and they will just take us farther away from feature parity.

[krajorama]

Sounds like a good idea, what did you have in mind?

One way to do it would be to have a meeting where we triage - go through all.
With 80 issues, it would be at least 2 hours I think - so you're looking at two 1h meetings at least I guess.
We could ask for someone from client_golang / client_java maintainers to join in.

I definitely want to raise more issues 😄 (implement content negotiation, have an issue for supporting OpenMetrics 2.0). I actually tried to contribute some of this when I realized that the old repo was in flux.

[jdmarshall]

My suspicion is that there are PRs and issues in the backlog that are anathema to the Prometheus Way and I'm just not fluent enough to
be telling people no without corroboration from others.

I've been going through old issues and PRs and I've closed out about 20% of them but I suspect there's much more to trim there. I just don't have the depth of knowledge in this particular user base to . My experience is split more evenly between statsd and otel, and coming to help out with prometheus was a Vote With Your Feet situation, after watching casual users struggle to understand the otel codebase and architecture.

I was coming off of some very fruitful performance tuning work on a production backend and saw a similar class of issues in the prometheus code base, and seeing issues complaining about stats throughput versus Otel, and now here we are 18 months later.