Comment by creata
1 day ago
True! But pjmlp was referring specifically to advanced JIT implementations, so I wondered which JITs he was referring to as advanced.
1 day ago
True! But pjmlp was referring specifically to advanced JIT implementations, so I wondered which JITs he was referring to as advanced.
In Ruby's case that would be RubyMotion, TruffleRubby and JRuby.
That trace back to Apple's efforts with MacRuby, or Sun's (for a while Netbeans even had Ruby support).
If you go that route, GraalPy is there too, so the argument is not as strong as it seems.
You missed my remark about PyPy feeling abandoned on the corner, well the same applies to GraalPy.
The Ruby JITs I mentioned are used in production.
While other dynamic language comunities embrace their JITs, in Python world, outside using it as a DSL for GPGPU JITs, it is pretty much let's just keep using CPythion with C and C++ extensions. Adding a JIT to CPython only became a thing after Facebook and Microsoft decided to push for its development.