Re: Is it too early to get started using Perl 6?

[particle]

it's never too early to start learning perl 6.

yes, the design phase of perl 6 has settled down quite a bit. as larry likes to say, "we're done with the first 80%, and well into the second 80%." at this point, language design changes usually occur in response to issues encountered when implementing the language; for example earlier this year some revisions were made to regexes (S05) which vastly simplified implementation by removing a number of special cases, and removing the need for the 'return' statetment in the regex engine.

the synopses are living documents, whereas apocalypses and exegeses are historical. the synopses are canonical; we base our tests and implementations on these documents, so they're a good resource for learning perl 6. there are other resources too, including tutorials, articles, wiki pages, and documentation, and i'll try to get some links for you soon.

building parrot is simple. releases are available on cpan, but we recommend getting the latest code from subversion since the pace of development is rapid. it's as easy as 'perl Configure.pl && make all test'. we don't recommend installing parrot at this time, but would appreciate any help we can get in making that process work better. building perl6 on parrot is just as easy: 'cd languages/perl6 && make all test'.

parrot has a glossary for many of these terms ('docs/glossary.pod',) but it's not up-to-date to include all those you mentioned.

[flexo]

we don't recommend installing parrot at this time

Why don't you recommend installing Parrot after building it? Do Parrot developers run it right from the build directory or something?

Can users build then just install into their own ~/opt?

[particle]

yes, parrot developers run it from the build directory. as far as i know, parrot install works okay, but when installing a new parrot release on top of an old one, it may run into trouble with previous library files lying around. i program on windows, and out of the build dir, so i'm not very familiar with the problem. there are (or were) rpms and we're getting the debian package back up to date.

but as i said earlier, development pace is rapid. in the past two days, i've added named, optional, and slurpy parameter passing to nqp (not quite perl--a subset of perl 6), the language perl6 uses to transform its parse tree into an abstract syntax tree. in another day or so (holidays excluded) i should have that working in perl6, too.