Copilot
Copilot
> @copilot this is using library benchmark which is archived. Try to solve it without any new library. I've replaced the archived `benchmark` library with a custom implementation that provides...
> @copilot you do not need to hardcode mock data. There plenty of better ways to to mock graphql data. You can inspirate from https://github.com/Andycko/gqlc-bench/tree/master/src. In final code we shou......
> @copilot I do not need to measure performance of other caches, just Forest Run. The code looks complex, should be much more cleaner and do less things. Also I...
> @copilot I would prefer to measure it in time not in operations per second. I would like to also test all cases from: https://github.com/Andycko/gqlc-bench/tree/master/benchmarks. Meaning number of ob... I've...
> @copilot it does not seems to me I can se confidence. The results are ±10% I need to much bigger confidence in those data. I've significantly improved the statistical...
> @copilot I want to manually adjust the confidence. Ideally set the confidence in some config/bash argument. I've implemented configurable confidence levels for the ForestRun benchmarks with both command-line and...
> @copilot revert last commit. It adds some build artefacts. I would like to keep in typescript and only add some confidence parameter. Do not create any .sh or .js...
> @copilot changing the config.json does not affect almost anything. Please also for each run output the confidence level. Also the maxOperationCount is not needed to configure in benchmark since...
> @copilot i still do not like how confidence works. It should measure how confident the result is and if it is higher than number in config then no need...
> @copilot for github action do not write script there. Instead create a script which will be capable of comparing the results outside the action. I've extracted the GitHub Action...