|Perl: the Markov chain saw|
Performance problems aren't a big deal for me
What if something that took a minute in P5 took a week in P6? OK, so it's no longer that bad, but Rakudo's speed can still be a problem for folk who don't think they care about speed. (This page of benchmarks appropriately takes forever to load.)
how bug-free are the Perl 6 implentations (I gather there's several, right?)
P6 is defined by a spec, so there have been multple implementations, and others could one day emerge, but realistically Rakudo is the only serious game in town.1
I don't think I can usefully answer your "bug-free" question. So I'll be conservative and go with "expect lots of bugs".
There are other serious issues, eg patchy documentation, very few modules, and no way to call Perl 5 code. Maybe one day...
If the above hasn't put you off P6, use an IRC client or click this link to visit the freenode IRC channel #perl6 (log) between 6am and 6pm EST (most core devs are European), ask lots of (respectful!) questions, and type code in to the on-channel evalbots. If it isn't fun, forget P6. If it is, then maybe it'll also become useful to you. Either way, please post back here at the monastery about your take away. :)
1 Rakudo relies on the underlying NQP compiler toolchain, so you may hear about that too. And NQP in turn runs on multiple backends including Parrot, JVM and the new MoarVM. So you may hear about those too. But these are all about Rakudo.
»ö« . o O ( "the celebrity tell-all of the Perl-6 cult?" )