|
|
|
|
|
by PommeDeTerre
4768 days ago
|
|
What exactly is the achievement here? I'm not certain that yet another mostly-incomplete virtual machine will really help the Perl 6 community much. They've already got that with Parrot. The whole emphasis on targeting virtual machines, whether it's Parrot, or the JVM, or now this MoarVM, has harmed the ability of Perl 6 to be implemented. We've seen one partial implementation after another, for years on end. Unfortunately, we can't actually use any of these implementations for anything serious, like we can with Perl 5, Python, and Ruby. This makes Perl 6 unusable, which is quite a shame. |
|
That's the second most disappointing part. The Parrot developers had a strong push between two and three years ago to redesign the internals to be a better fit for Rakudo along these lines (better runcore, easier JITting, improvements to the core object model, representation polymorphism), but the reaction from the Rakudo developers was a mixture of "No, thank you!", "Not yet.", and "If you have to, but don't change any APIs or behavior that Rakudo counts on."
It's a shame this couldn't have all been done at the same time. The available volunteer knowledge and interest might have saved two or three years of Perl 6 floundering around in the netherworld of big promises and very modest realities.
The most disappointing part is realizing that this sets back the potential delivery date of a usable Perl 6 by at least another couple of years. I guess that shouldn't be a surprise by now though.