search

hanabi1224 / Programming-Language-Benchmarks

Yet another implementation of computer language benchmarks game
https://programming-language-benchmarks.vercel.app/
MIT License
631 stars 133 forks source link

Benchmark comparison is weird when opposite's implement doesn't exist #411

Open Eveheeero opened 1 year ago

Eveheeero commented 1 year ago

I came to this site to compare the execution speeds of Java and C++, and I saw the following: the Java source code was the fastest, and the C++ source code didn't even show up in the rankings.

I thought this was odd, so I dug deeper and realized that the algorithm implemented in Java was not implemented in C++.

https://programming-language-benchmarks.vercel.app/cpp-vs-java https://programming-language-benchmarks.vercel.app/java-vs-cpp

In the first link, you can see that C++ and Java are properly compared, but in the second link, it doesn't look right because it is compared to the unimplemented C++ code.

After seeing this, I think it shouldn't show up unless there is code implementing the algorithm in another language.

cyrusmsk commented 1 year ago

For “mainstream” languages you can check original project: https://benchmarksgame-team.pages.debian.net/benchmarksgame/fastest/gpp-java.html

Eveheeero commented 1 year ago

I've seen it, and now I know which language is faster.

I just think it's strange that the "vs" tab on that page shows the execution speed of language A even when there is no implementation of language B, giving the illusion that language A is much faster than language B.

It's not just java vs c++, it's c++ vs csharp or whatever.