Repository navigation
city: reconcile GitHub's commit count for repo and org cities, in numbers - #12
Merged
Merged
Conversation
…bers A repo city's panel said "Buildings show every non-merge commit on the default branch since 2025-10-13", and for golang/go drew 3,491 where GitHub counts 3,748 since that day. A reader takes "since" as "landed since", which would be 3,716; the city goes by the day a commit was written. An audit of golang/go against GitHub's own history accounts for every commit: 3,748 = 3,491 non-merge commits written since the window start (the buildings, exactly) + 32 merges + 225 commits written earlier and landed since (Gerrit changes sit in review, one since 2016). The finish step already has GitHub's count for each repo (discovery's history(since), by landing day), and each record holds every commit with its merge flag and author minute. For repo and org cities the payload now carries totals.landed: GitHub's count for the drawn repos and the merges written in the window. The panel reads: GitHub counts 3,748 commits on the default branch since 2025-10-13, by the day each landed. Buildings show the 3,491 written since then. Not drawn: 32 merges and 225 commits written before 2025-10-13. The split is given only where it adds up exactly: every drawn repo has GitHub's count, no record is approximate or cut at its 20,000-commit cap, and the count is not below the buildings and merges. Otherwise, and for payloads from before this or from the Python CLI, the panel says what it draws without the ambiguity: every non-merge commit written since the window start.
|
The latest updates on your projects. Learn more about Vercel for GitHub.
|
azank1
marked this pull request as ready for review
October 13, 2026 13:07
This branch was successfully deployed
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
A developer who checks a repo city against GitHub finds a gap the page never explains. For golang/go, the city draws 3,491 commits where GitHub's history counts 3,748 since the same day. The panel says "Buildings show every non-merge commit on the default branch since 2025-10-13". A reader takes "since" to mean landed since, which would be 3,716, but the city goes by the day a commit was written.
Audit (golang/go against GitHub's GraphQL history, every commit fetched):
history(since: 2025-10-13), by landing dayChange. The finish step already has GitHub's count for each repo: discovery's
history(since), by landing day. Each record holds every commit with its merge flag and author minute. For repo and org cities, the payload now carriestotals.landed: GitHub's count for the drawn repos, and the merges written in the window. The panel then reads (rendered from golang/go's production payload with the count this change computes):The split appears only where it adds up exactly:
Otherwise the panel says what it draws, without the ambiguity: "every non-merge commit written since …". The same holds for older payloads and for the Python CLI, which doesn't compute the count (a follow-up if wanted).
Tests.
bundle()gets a real replayed record with one commit made a merge and one written before the window; it checks the split and each case where it is withheld. The finish step asserts the org count. The Python parity test is unchanged and passes; inputs without the count produce identical payloads. Suite: 50 tests, 44 pass, 6 live tests skipped, 0 fail; typecheck clean. Viewer rebuilt:app.021aac7758.js, live bundle unchanged.After merge. Repo and org cities show the reconciliation once rebuilt; the gallery rebuild after #10 and #11 covers that.