fix(performance): count each initial request URL and soften the CI wording

Deduplicate initial resources by request URL rather than by file, so a
file referenced twice with different query strings is counted twice, as
the browser fetches it; fragments are still ignored. Each resource now
carries both its file path and its request URL.

Describe the CI ratchet as the goal of the loop rather than as existing
enforcement until the baseline and job PRs merge.

Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
This commit is contained in:
4grayandClaude Fable 5.1 committed 2026-09-26 10:29:36 +02:00
1 parent 3cd032510e
commit 5cb69b1ec1
4 files changed
+50 -26

No files matched your search

+6 -4
View File
@@ -2,10 +2,12 @@
IPTVnator measures performance through a small set of everyday user journeys.
Each journey has deterministic counters that are asserted exactly, and
wall-clock timings that are recorded as evidence. Counters are ratcheted in CI:
a committed baseline may only be lowered, and only with the measured output as
evidence. This document is the contract for that loop; `tools/performance/`
holds the scripts.
wall-clock timings that are recorded as evidence. Counters are meant to be
ratcheted in CI: a committed baseline that may only be lowered, and only with
the measured output as evidence. This document is the contract for that loop;
`tools/performance/` holds the scripts. The measurement script lands first;
the baseline file and the CI job follow in their own PRs (#1693, #1694), so
until they merge the reported number is informational, not enforced.
## Journeys