On the roadmap. Today a report is about one author. The question people actually bring to this tool is often comparative: two candidates, or one candidate against the repository's median, and running the tool twice and eyeballing two reports is not the same thing, because the figures share denominators that only mean something when they are computed once.
Done when
- One command produces one report holding two authors, with each figure computed from the same pass over the repository.
- The canonical-JSON hash covers the comparison as a whole, so
verify works on it exactly as it does on a single report.
- Every figure that cannot honestly be compared says so instead of printing a number. Not all of them can: think carefully about what tenure does to a cadence figure before you put two of them next to each other.
- The README gains a short section that includes the sentence a reader needs most, which is what a difference between two authors does not mean.
Out of scope. Ranking, scoring, or any single number that says one author is better. The README already refuses that and this must not smuggle it back in through a sorted table.
This one carries a risk worth naming. A side-by-side view invites exactly the misreading this tool was built to prevent. A pull request that makes comparison easy and misreading easy will not be merged; the framing is as much of the work as the code.
Taking this on: comment and it is yours, so two people do not write the same patch. No design document needed unless the issue asks for one.
What gets a pull request rejected here: a change with no test; a new runtime dependency; or a number written into a README or a document that did not come from a run somebody else can repeat. CONTRIBUTING.md has the rest.
On the roadmap. Today a report is about one author. The question people actually bring to this tool is often comparative: two candidates, or one candidate against the repository's median, and running the tool twice and eyeballing two reports is not the same thing, because the figures share denominators that only mean something when they are computed once.
Done when
verifyworks on it exactly as it does on a single report.Out of scope. Ranking, scoring, or any single number that says one author is better. The README already refuses that and this must not smuggle it back in through a sorted table.
This one carries a risk worth naming. A side-by-side view invites exactly the misreading this tool was built to prevent. A pull request that makes comparison easy and misreading easy will not be merged; the framing is as much of the work as the code.
Taking this on: comment and it is yours, so two people do not write the same patch. No design document needed unless the issue asks for one.
What gets a pull request rejected here: a change with no test; a new runtime dependency; or a number written into a README or a document that did not come from a run somebody else can repeat.
CONTRIBUTING.mdhas the rest.