v3.0.0

   🚨 Breaking Changes

• Drop unmaintained node 16 support  -  by @jerome-benoit in #116 (5e08d)
• Remove useless benchmark todos  -  by @jerome-benoit in #118 (434d0)
• Remove *Concurrently methods  -  by @jerome-benoit in #135 (1297a)
• Fold warmup step into benchmark run  -  by @jerome-benoit in #150 (4fcd4)

     API simplification details:

• Bench concurrency setting is now controlled only at runtime via concurreny property. Any subsequent call to run() will apply it
• Bench warmup is now done only at run() and controlled via benchmark options and/or warmup property at runtime

   🚀 Features

• Compute and display median and median absolute deviation  -  by @jerome-benoit in #89 (46fbd)
• Latency and throughput advanced statistics  -  by @jerome-benoit in #119 (70bc3)
• Add name to Bench options  -  by @jerome-benoit in #140 (eedc1)

   🐞 Bug Fixes

• Silence tsup exports warning  -  by @jerome-benoit in #114 (b1cbf)
• Accurate Student t-distribution table  -  by @jerome-benoit in #90 (1e749)
• Add accurate percentiles approximation computation  -  by @jerome-benoit in #87 (ab467)
• Properly handle async function at all benchmark steps  -  by @jerome-benoit and Jérôme Benoit in #133 (50d00)
• Avoid duplicate call to benchmark related callbacks  -  by @jerome-benoit and Jérôme Benoit in #134 (581c2)
• Refine Fn type definition  -  by @jerome-benoit (c6062)
• Run simple-git-hooks at prepare  -  by @jerome-benoit (9a43f)
• Ensure error events are sent  -  by @jerome-benoit in #167 (bde99)
• dependabot: Add examples directory  -  by @jerome-benoit (29e44)
• test: Use proper expectation type  -  by @jerome-benoit (347a5)

   🏎 Performance

• Bind timestamping function to their modules  -  by @jerome-benoit in #149 (d9a23)

    View changes on GitHub

v3.0.0-rc.8

   🚨 Breaking Changes

• Drop unmaintained node 16 support  -  by @jerome-benoit in #116 (5e08d)
• Remove useless benchmark todos  -  by @jerome-benoit in #118 (434d0)
• Remove *Concurrently methods  -  by @jerome-benoit in #135 (1297a)
• Fold warmup step into benchmark run  -  by @jerome-benoit in #150 (4fcd4)

     API simplification details:

• Bench concurrency setting is now controlled only at runtime via concurreny property. Any subsequent call to run() will apply it
• Bench warmup is now done only at run() and controlled via benchmark options and/or warmup property at runtime

   🚀 Features

• Compute and display median and median absolute deviation  -  by @jerome-benoit in #89 (46fbd)
• Latency and throughput advanced statistics  -  by @jerome-benoit in #119 (70bc3)
• Add name to Bench options  -  by @jerome-benoit in #140 (eedc1)

   🐞 Bug Fixes

• Silence tsup exports warning  -  by @jerome-benoit in #114 (b1cbf)
• Accurate Student t-distribution table  -  by @jerome-benoit in #90 (1e749)
• Add accurate percentiles approximation computation  -  by @jerome-benoit in #87 (ab467)
• Properly handle async function at all benchmark steps  -  by @jerome-benoit and Jérôme Benoit in #133 (50d00)
• Avoid duplicate call to benchmark related callbacks  -  by @jerome-benoit and Jérôme Benoit in #134 (581c2)
• Refine Fn type definition  -  by @jerome-benoit (c6062)
• Run simple-git-hooks at prepare  -  by @jerome-benoit (9a43f)
• dependabot: Add examples directory  -  by @jerome-benoit (29e44)
• test: Use proper expectation type  -  by @jerome-benoit (347a5)

   🏎 Performance

• Bind timestamping function to their modules  -  by @jerome-benoit in #149 (d9a23)

    View changes on GitHub

v2.9.0

   🚀 Features

• Add pkg.pr.new  -  by @AmirSa12 in #80 (3a41b)

   🐞 Bug Fixes

• Build correct types for CommonJS users  -  by @kanadgupta in #85 (c58ac)

    View changes on GitHub

v2.8.0

   🚀 Features

• Concurrency mode  -  by @Aslemammad in #77 (86866)

    View changes on GitHub

v2.7.0

No significant changes

    View changes on GitHub

v2.6.0

   🚀 Features

• throws  -  by @Aslemammad in #51 (ab950)

   🐞 Bug Fixes

• Remove DOM-related types to make tinybench environment agnostic  -  by @sheremet-va in #58 (d78e6)

   🏎 Performance

• Use samplesLength instead of multiple use samples.length  -  by @Dunqing in #59 (f52f0)

    View changes on GitHub

v2.5.1

   🐞 Bug Fixes

• Accurate results and hooks fix  -  by @Aslemammad in #43 (a0e3c)
• Correct variance calculation  -  by @bagdonas in #45 (4d2c0)

    View changes on GitHub

v2.5.0

   🚀 Features

• Bench table  -  by @Sajad-Sharhani and @Aslemammad in #38 (35439)
• Add bench.todo  -  by @kamiloox in #39 (76461)

   🐞 Bug Fixes

• Await the this.bench.setup call during warmup  -  by @HlustikP in #37 (cafe0)

    View changes on GitHub

v2.4.0

   🚀 Features

• Add per-task untimed lifecycle hooks  -  by @achingbrain and @Aslemammad in #33 (0c6a3)

    View changes on GitHub

v2.3.1

   🐞 Bug Fixes

• Rewrite sort function  -  by @Aslemammad (04b2c)
• Example eslint config and task time calculate logic  -  by @kainstar (9ceac)
• Add internal async function checker, adjust divide number in hz metric compute  -  by @kainstar (bd15c)
• Calculate task time out of try-catch block  -  by @kainstar (74c5c)
• Remove dirtyAsyncCheck option and implement  -  by @kainstar (6bcb3)

    View changes on GitHub