Hi — apologies for arriving via the issue tracker. This is a feedback request rather than a bug report, and "not interested" is a completely fine answer; I'll close it myself if so.
I maintain Flecto, a small open-source tool that reports what a config change actually did, in plain English, as a pull request comment. I've been running it against real repositories rather than my own fixtures, because fixtures tell you nothing about whether the output is useful.
I ran it read-only against your PR #1663 ("remove placeholder values from monitoring chart"):
✅ Check passing — 2 changes in 1 file — 2 changed, 0 added, 0 removed.
Policy: no findings.
| changed | prometheus.server.alertmanagers[0].relabel_configs[0].regex | "<namespace>" | "monitoring" |
| changed | prometheus.server.alertmanagers[0].relabel_configs[1].regex | "<release-name>" | "monitoring" |
The reason I picked this PR: those two values sit five levels down inside
relabel_configs, and in a raw YAML diff they are two lines that look like every other two lines. Naming the full path and the before/after is most of what I am trying to test — whether that framing actually helps a reviewer, or whether it is just the same information rearranged.
Two questions, neither of which needs you to install anything:
- When a PR touches
charts/*/values.yaml in this repo, what do you look at today? Rendered output, the raw diff, or mostly trust plus CI?
- With 1600+ PRs, would a comment naming changed paths be signal or noise here?
If the honest answer to 2 is "noise at our volume", that is genuinely the more useful answer for me, and I would rather hear it.
Happy to close this right away if issue-tracker outreach isn't welcome — just say so.
Hi — apologies for arriving via the issue tracker. This is a feedback request rather than a bug report, and "not interested" is a completely fine answer; I'll close it myself if so.
I maintain Flecto, a small open-source tool that reports what a config change actually did, in plain English, as a pull request comment. I've been running it against real repositories rather than my own fixtures, because fixtures tell you nothing about whether the output is useful.
I ran it read-only against your PR #1663 ("remove placeholder values from monitoring chart"):
The reason I picked this PR: those two values sit five levels down inside
relabel_configs, and in a raw YAML diff they are two lines that look like every other two lines. Naming the full path and the before/after is most of what I am trying to test — whether that framing actually helps a reviewer, or whether it is just the same information rearranged.Two questions, neither of which needs you to install anything:
charts/*/values.yamlin this repo, what do you look at today? Rendered output, the raw diff, or mostly trust plus CI?If the honest answer to 2 is "noise at our volume", that is genuinely the more useful answer for me, and I would rather hear it.
Happy to close this right away if issue-tracker outreach isn't welcome — just say so.