What is Perl 6?

Since this is an introductory post to this, a blog dedicated to Perl 6, I figure I should establish a few things about what Perl 6 is, before just 'jumping in.' Perl 6 is more than a language, it's a community and a development effort by those willing to be involved.

To start, I'll go ahead and list of a few things from the 'Perl 6 jargon file', so you're not caught by suprise when these are referenced to later:

1. Apocalpyse: Larry wall, BDFL of the Perl project, designed Perl 6 to be 'the community's rewrite of perl,' where Perl 5 was his rewrite of Perl. To do this, people from all around the globe submit RFC's to the perl6-language mailing list (short form: p6l which is how I'll refer to the list from now on,) talking about what they believe what changes should take place in the language. Wall manually answers these RFCs and describes whether he accepted or rejected the change, and why. An apocalypse is a collection of these RFC's and their answers. Every apocalypse directly relates to a chapter of the Camel book.
   
   Note, that Apocalypses, after their initial release, can be revised. It is not uncommon for an Apocalypse to soon change after it's publication; if you have code that isn't working, but only 'proved to be correct, yet not tried,' [sic] check the accompanying Apocalypse: the syntax and specifications may very well have changed.
   
2. Synopsis: After an Apocalypse, a Synopsis is written. The synopsis essentially sums up the changes taking place described in a single Apocalypse. The name is just what it sounds like. Synopsis are generally more approachable and less lengthy than their accompanying Apocalypse.
   
3. Exegesis: The Exegesis are practical explanations and code samples of the changes taking place as described by both the Synopsis and Apocalypse. These are your examples of the specifications the Apocalypses describe.
   
4. Pugs: Pugs is an implementation of Perl 6, written in Haskell. Pugs intends to bootstrap the perl 6 compiler by writing it in Haskell. The objective is that once all the pugs milestones are done, that if needed, Pugs will be rewritten in Perl 6, to develop a fully fledged Perl 6 compiler, in Perl 6. You can find more on the wiki entry. Commit access to pugs is handed out quite liberally, and anybody is free to work on whatever they feel needs to be worked on. Just the way Perl 6 is a community effort and a rapidly changing landscape, so is Pugs.
   
   Note that Pugs is not the only implementation of Perl 6, it is merely the most complete and established. Via the use of the v6.pm module, it's possible to write limited amounts of Perl 6 in Perl 5 (you may think it as either 'perl-on-perl' action or 'perl-on-haskell' action.)
   In my future posts, I will generally stick to Pugs for the implementation, as it is more feature complete and more accurately reflects the current state of Perl 6.
   
5. Parrot: Parrot is a virtual machine, designed to run Perl 6. The virtual machine is designed for fast (one of it's primary goals) execution of dynamic languages such as, but not limited to, Perl 6. Multiple languages can run on Parrot (languages for parrot can be constructed via the use of a very interesting compiler-toolchain, designed for Parrot), and the architecture is modern and well done. Parrot is discussed primarily on the perl6-internals mailing list (p6i.)
   
   Please keep in mind, development of Parrot in relation to the actual Perl 6 language is nigh orthogonal: Parrot's only obligation is to support Perl 6 and it's semantics, but the language designers have no say in how it's done or why. Parrot is, as well, the only virtual machine for Perl 6. Communication between these two branches is done, however, they are two seperate entities by all practical means.

So now that the lingo is out of the way, what's the real deal with Perl 6? How can you get involved?
Getting involved is essentially as simple as joining the mailing lists (and I would recommend joining IRC as well: irc.freenode.org #perl6) and posting. That's it. Perl 6 is a community effort in its entirety: you, the people, say what stays and what goes. Submit your RFCs, comment on others, the power to change the language is yours.


Aside from the community, Perl 6 has changed a lot technically. Niklaus Wirth said that if we add things to a language, we must throw things away. You are still writing perl, and you will still be backwards compatible, but a lot of things are changing from Perl 5 to Perl 6. Things are getting added, things are getting thrown away.

Quick examples are:

• Sigil invariance. When referencing an array itself (or any other type, for that matter), or an element of the array, you always give it the original sigil you declared it with, i.e. you are going from $arr[1]; to @arr[1];. Note, context in Perl 6 still exists, but sigils do not vary.
  Example:
  
  pugs> my @a = 1..5;
pugs> @a[0]
1
pugs> $a[0]
Internal error while running expression:
***
Unexpected "["
expecting "::"
Variable "$a" requires predeclaration or explicit package name
at line 1, column 3
pugs>
• Object orientation is full blown now. Classes and object-orientation is greatly improved, as you now have classes, [multiple] inheritence, private/public pieces of data, constructors and destructors; object orientation is much more complete.
  
• Regex's no longer exist. You're using Rules, now (this change in lexicon is due to the fact that perl's "regular expressions" are so far off the mark of a traditional, formal regular expression, that they were renamed.) One of the greater things about Rules and Grammar's in the new Perl 6 engine is that Grammars are essentially 'classes for rules,' a grammar just inherits from the Rule class itself.
• Static types. Perl 5 was a dynamic language, Perl 6 adds static typing capabilities. These are simply strictly optional annotations to your already existing variables to explicitly define your type. Either style of programming can be used.
  It's interesting to note that as you can specify types to a variable, you can specify two kinds: a type refering to what the variable holds (strings, int's, etc.) and a type referring to what the variable actually is; a hash, an array, ex:
  
  pugs> my Int $i is scalar;
pugs> $i = 1;
pugs> say $i
1
Bool::True
pugs>
  
  This is not limited to the built in types, any may do, including classes.
  
• Laziness. Lists and the like are now implicitly lazy, meaning you can do things such as my @a = 1...;
• Various syntax changes. Parenthesis in a large majority of constructs are strictly optional. Array's may be constructed in a 'stream of words' fashion using the < > operators, rather than qw(). I.e., from my @a = qw(a b c d); you go to my @a = <a b c d>;. Hash values can be referenced the same way, i.e. %hash<some_key>; Prototypes for subs can simply be done via sub name { ... };. There are more various little changes around, but none of these are hugely significant ones.

Those are just a few of the changes. For more, I suggest you read these.

Perl 6 is a big change to an already very popular programming language. One isn't going out the door for another to come in the same day: the adoption will be slow, but it is a needed change and a good one for both the Perl community, and Perl itself. If you would like to contribute, I can only suggest you sign onto IRC/mailing lists, and you speak your voice and do what you think you need, or are able, to do.

It's your language, it should do what you want it to; shouldn't it?