We all know benchmarks are just ultra-scoped iterations that must of the time say nothing about real life, but it is nice to see how dart is a great language and some time we don’t appreciate it.
It’s a nonsense benchmark that does not really mean anything in isolation and especially not when you compare across many languages. It’s like comparing books by the number of pages in them.
I would be more interested in someone doing similar breakdown as @mraleph did for dart, for every language to try and push the limits of each language. I think it could be a lot closer then
In a sufficiently large group of developers you will find opinions from the entire range of possibilities - vocal supporters of such benchmarks who will use it to confirm their biases, people who just shrug ¯\_(ツ)_/¯ when see it, and ones who detest such ways of comparing languages. I wouldn’t say that “community” has been super interested in this particular benchmark, at least compared to other online bubbles.
I think we should take the opportunity to educate developers about all the shortcomings and simplifications of such benchmarks, and maybe crave for more publicity for actual useful tests e.g. pushing json parsing to the limits or improving complex layout performance.
In some scenarios, Dart can be slower than certain languages that appear lower on various performance rankings. Don’t get me wrong—I love Dart—but its speed really depends on the specific situation you’re testing.
For example, I dislike using Dart for cryptographic operations because it’s noticeably slower compared to Rust. That’s why I rely on rust_dart packages for those tasks.
Overall, the Dart team and community contributors have done an excellent job. However, they can’t stop improving, as there are still areas where the language can be optimized. And yes, sometimes it’s not the language itself but the way certain packages are implemented that causes slower performance.
Flutter and the related logo are trademarks of Google LLC. We are not endorsed by or affiliated with Google LLC.
Using contents of this forum for the purposes of training proprietary AI models is forbidden. Only if your AI model is free & open source, go ahead and scrape.