On 2020-03-22 at 07:51:40, Jeff King wrote: > When t3419 was originally written, it was designed to run a smaller test > for correctness, and then the same test with a larger number of patches > for performance. But it seems unlikely the latter was helping us: > > - it was marked with EXPENSIVE, so hardly anybody ran it anyway > > - there's no indication that it was more likely to find bugs than the > smaller case (the commit message isn't very helpful, but the original > cover letter describes it as: "The first patch adds correctness and > (optional) performance tests". > > - the timing results are shown only via test_debug(). So also not run > unless the user says "-d", and then not provided in any > machine-readable form. > > If we're interested in performance regressions, a script in t/perf would > be more appropriate. I didn't add one here, because it's not at all > clear to me that what the script is timing is even all that interesting. > > Let's simplify the script by dropping the EXPENSIVE run. That in turn > lets us drop the do_tests() wrapper, which lets us consistently use > single-quotes for our test snippets. And we can drop the useless > test_debug() timings, as well as their run() helper. And finally, while > we're here, we can replace the count() helper with the standard > test_seq(). I'm also fine with this solution. As long as this test doesn't fail with EXPENSIVE, I'm happy. -- brian m. carlson: Houston, Texas, US OpenPGP: https://keybase.io/bk2204