A Just In Time VM for Not Quite Perl

[raiph]

If you're interested in the guts and gore that involves VMs, or re-implementing Perl, or implementing other languages, or hanging out with or helping those who do, you might be interested in MoarVM which "reboots the whole VM idea for Perl, based on experience with Parrot, without many of the flaws of Parrot".

It looks like most of the current action is on the IRC channel #moarvm on freenode.net (for which logs started earlier today).

[Anonymous Monk]

Is there an explanation for those who haven't followed the evolution of parrots and all that?

[raiph]

I'm going to focus on the relevance of this news to monks interested in Perl 6. I happen to think this news is also relevant to those only interested in Perl 5, but such aspects will best reveal themselves at YAPC::NA and beyond.
• NQP (Not Quite Perl) is a Perl 6 subset, much smaller than Perl 5 or Perl 6. The first version of NQP was started and finished around 2007. NQP is designed to be ideal for writing compilers. The NQP compiler is (mostly) written in NQP. The Rakudo Perl 6 compiler is also (partly) written in NQP (most of the rest is Perl 6).

• The Parrot VM is a backend target for NQP. Unfortunately Parrot has fallen far behind the evolution of NQP (and related pieces, especially 6model) with all sorts of negative consequences.

• MoarVM, quietly written by a small team over the last 14 months or so, is a new backend target for NQP. It is specifically designed for NQP/6model.

• Today MoarVM reached the point where it is cross-compiling itself and Jonathan Worthington considered it worth revealing to the Perl 6 community, especially given that the YAPC::NA hackathon starts tomorrow. (I, in turn, thought it was worth specifically mentioning at the Monastery, in case some monk(s) wish to check it out and maybe get involved in the hackathon virtually.)

I recommend interested monks pay close attention to the news coming out of YAPC::NA 2013 that starts tonight in Austin, TX and/or join #moarvm on freenode (log).

Hth.

[Anonymous Monk]

Hang on..

You still don't give us the complete details. If the JVM effort was underway doesn't it make sense to finish it and then start on something else?

We still don't have a feature complete stable release on one VM. What is the whole point in producing half complete implementations on so many VM's. While efforts could have rather spent on producing one feature complete stable implementation on one VM.

As of now it really looks like there are too many abandoned projects, too many re writes and too many sub project time sinks preventing anything worthwhile to come out of this whole exercise

If anything adding MoarVM to picture now. The entire Perl 6 production release got delayed by another 2 years. Porting Rakudo to MoarVM looks like a 6-8 month exercise and after all that you wouldn't have added a single user facing enhancement to the whole product.

[Anonymous Monk]

is Parrot finally abandoned?