5 ms·

Completely anecdotal: - In my small benchmarks, Crystal is very fast - In benchmarks not my own, Crystal is very fast - Crystal has a very nice batteries inclu

Completely anecdotal: - In my small benchmarks, Crystal is very fast - In benchmarks not my own, Crystal is very fast - Crystal has a very nice batteries included stdlib. Benchmarks that aren't mine: http://lh3.github.io/2020/05/17/fast-high-level-programming-languages http://lh3.github.io/2020/05/17/fast-high-level-programming-... The only other language that is high level in its ballpark is D.

mratsim 6y ago
Benchmarking strings is often benchmarking memory allocation speed or the GC. If you manipulate strings day in day out in your workflow, the first thing you do is remove all those allocations and for what is left you use a memory pool. The speedup could easily go into the 20x in any language. It's actually one of the area where Python or JS can beat statically typed languages, they focused a lot into optimizing strings while statically typed languages often leave that as an exercise to the reader.
- throwaway894345 6y ago
  > It's actually one of the area where Python or JS can beat statically typed languages, they focused a lot into optimizing strings while statically typed languages often leave that as an exercise to the reader. I think this has more to do with writing the string handling functions in heavily-optimized C rather than in Python/JS than it does with memory tricks, but I'll happily accept correction.
  - echlebek 6y ago
    Almost certainly. Iteration is around 40x slower in Python than in a compiled language, last I checked.
  - mratsim 6y ago
    Yes low-level C + retaining memory around in the GC. The second part avoids many malloc/free that may be hidden in destructors in low-level languages.
- totalperspectiv 6y ago
  The article linked is doing much more than benchmarking string operations. I'd even argue that string allocations are about as minimal as possible with the algorithms used. The C version takes this a little farther, but each of the other implementations are pretty similar. I don't disagree that benchmarks should be taken with a mound of salt though.

mratsim 6y ago

Benchmarking strings is often benchmarking memory allocation speed or the GC. If you manipulate strings day in day out in your workflow, the first thing you do is remove all those allocations and for what is left you use a memory pool. The speedup could easily go into the 20x in any language. It's actually one of the area where Python or JS can beat statically typed languages, they focused a lot into optimizing strings while statically typed languages often leave that as an exercise to the reader.

throwaway894345 6y ago
> It's actually one of the area where Python or JS can beat statically typed languages, they focused a lot into optimizing strings while statically typed languages often leave that as an exercise to the reader. I think this has more to do with writing the string handling functions in heavily-optimized C rather than in Python/JS than it does with memory tricks, but I'll happily accept correction.
- echlebek 6y ago
  Almost certainly. Iteration is around 40x slower in Python than in a compiled language, last I checked.
- mratsim 6y ago
  Yes low-level C + retaining memory around in the GC. The second part avoids many malloc/free that may be hidden in destructors in low-level languages.
totalperspectiv 6y ago
The article linked is doing much more than benchmarking string operations. I'd even argue that string allocations are about as minimal as possible with the algorithms used. The C version takes this a little farther, but each of the other implementations are pretty similar. I don't disagree that benchmarks should be taken with a mound of salt though.