The test-run banner (visible on every Test-script view) now carries the
same live stats as the script head and the suite summary: elapsed so far,
the average per completed iteration and the estimated time left. The
text is built by a pure helper (testRunBannerText) so it is testable;
testRender only calls it while a run is active, and non-load runs keep
their position text (step or suite-member progress).
The existing banner-source assertion moved to the helper. Tests: the
banner text for a load suite and a plain run. 1061 Python + 1075
frontend; sw cache simple-v481. CHANGELOG/help/docs/AGENTS updated.
The script-run head and the suite summary now display, while a load run
repeats: the elapsed time so far, the average per completed iteration and
the estimated time left (testLoadTiming / testDurationText /
testIterAvgText), recomputed on every 2 s status poll. A finished run
keeps the server's final elapsed_ms/avg_ms (no estimate), and before the
first completed iteration only the elapsed time is shown - no bogus
average or ETA.
The timing is client-side over the existing load block (started, done,
iterations), so it works with the running server as well; no server
change is needed.
Tests: the timing math and the two formatters, the script head render and
the suite summary render (3 new). 1061 Python + 1074 frontend.
sw cache simple-v480; CHANGELOG [3.38.0], help EN/RU,
docs/test-scripts.md and the AGENTS notes.
Both layers: the server's _TEST_LOAD_MAX and the PWA's
TEST_ITERATIONS_MAX plus the two input max attributes. The value stays
inside 2^53-1 (the JavaScript safe-integer range), so the counters and the
progress display survive the JSON/JS round trip exactly; the practical
ceiling is the run duration (50 ms/iteration ~1.6 years at the cap).
The server's range error names the new cap automatically. The frontend
tests read the cap from index.html instead of stubbing a literal, and
clamp at it (above the cap, exactly at it, and the input max).
Docs: api.md, test-scripts.md, help EN/RU, the CHANGELOG (folded into the
unreleased 3.38.0) and the AGENTS notes. sw cache simple-v479.
1061 Python + 1071 frontend.
POST /api/test/run takes an optional iterations count (the PWA's
Iterations field): the run repeats, successful iterations are discarded
(steps/log cleared, only load.done and the timing grow), the first error
stops the loop and keeps that iteration's report, a stop ends it after
the current iteration, and a card session change between iterations is
an error. The SCP80 counter state is shared across iterations, so
packets continue from what the previous run consumed.
The status gains a load block (iteration/done/failed_iteration/
elapsed_ms/avg_ms; disabled for a normal run). The suite runner was
split into a per-iteration body (_test_suite_execute_once) + the
single-run finaliser, and two load workers own the loops.
Review fixes: a raised runner error finalises as error (no stuck
running state), a suite-internal session change is an error, a stopped
member ends the suite loop, the final status keeps a warning from any
iteration, avg_ms covers the attempted iterations, and each suite
iteration's counters_before is read fresh.
1041 Python + 1069 frontend; simple-v475.
catarrh