Intro    Chapter:   0. History   1. Design   2. Basic   3. Var   4. Op   5. IO   6. Block   7. Sub   8. OOP   9. Regex   10. Meta
Overview   Appendix:   A. Index   B. Grouped   C. Classtree   D. Delta   E. Exciting   F. FAQ   G. Glossary   H. Href 

"You should not take this picture to mean that linguistics is the opposite of common sense. But there are many computer scientists who have complained that it should be more like engineering and less like an art. And I disagree, but that's okay." -- Larry Wall

Main Goal

Enjoy life and enjoy the freedom of expressing yourself. Which means: support you as much as possible to get fast but solid results and not force you in any way. Perl 6 tries to be even better than Perl 5 as your humble servant for translating your thoughts into executable code (if possible).


Every style and level of expertise has to be supported. And even more difficult: all of them have to play well together. The default often imitates a natural language (e.g., English). The following oft-repeated terms encapsulate strategies and concepts for this.



Perl's everlasting slogan: "There is more then one way to do it," meaning your way is valid, too. Just be aware of logical consequences.

Visual Metaphor

People like to see their code as a picture at first glance. Therefore names and operators should be distinctive, similar things should look similar, different things different. Larry likes to add: weird things should also look weird. And sometimes even characters can be ideograms; like the Z meta-operator, which shows visually how it works (first left, then right, then next left, etc.).


This also has to do with natural attention flow. If a topic or first word of a sentence sets an understandable context for what follows, the meaning is built up word for word. This is much easier to grasp, than if the last word of a block resolves the block's entire meaning (like regex modifiers do in Perl 5). In particular, for and given are called topicalizers, because they set $_, the current topic variable.

Context sensitive

The same term or symbol might do different things in different contexts, although we have done our very best to consistently represent the same general idea. Maybe the best example is the * aka Whatever. But every context is its own domain, optimized to solve the problems of that domain.

Maximum reuse

Not only is syntax reused to minimize the stuff you have to remember, but concepts are reused also, wherever possible. For example, Pairs are used to build hashes or named parameters in routines calls. There are plenty more examples. Maybe the biggest one: the regex engine that parses your code is the same engine you use in your program.

No special rules

Most of the special rules that plagued Perl 5 are gone or generalized.

One pass parsing

Perl 6 is much easier to understand, both for people and also for software parsers. The language is designed to be understood by machines in one read, dividing the source in smaller and smaller units. As a nice side effect: syntax errors can only damage the understanding of small units.

Keep easy things easy and hard things possible

Beside TIMTOWTDI, the other ancient slogan is "Make easy things easy and hard things possible." Even if nearly every cool idea is crammed into Perl 6, we don't lose track of the basics. Features are made as accessible and free of artificial prerequisites as possible. For example, printing a variable as a line without explicit quotes and newline (say $var), or reading a file without a file handle (slurp("filename").lines). Rule of thumb: if you can think of a concept as trivial, then it should be. On the other hand, if you can think of a computable solution, it should at least be doable in Perl (no artificial walls). Huffman coding: This is a refinement of the last rule - common, often used things have short, straightforward names. More verbose names indicate more rare, harder to grasp concepts, or things that overwrite the default rules ("weird things should look weird"). It also implies that the defaults lean toward what a beginner (non-Perl 5 programmer) might expect ("hide the fancy stuff").

Water bed theory of linguistic complexity

This philosophy underlies the two previous terms. All Perl words should reflect logical units which can be freely combined. Making it more complicated than necessary is obviously not desirable. But oversimplification will create artificial complexity somewhere else. Like a water bed, where you can press down in one place but the water will go up in another. All your paradigms are belong to us. We learn, take and adapt from every source possible to offer the best of all breeds. You should be able to keep your style when switching to Perl. Perl 6 supports the structured, object oriented, aspect oriented, functional, design by contract, declarative and logic programming paradigms.

Second system syndrome

We are very aware that many ambitious software projects that are rebuilt from scratch fail. Especially those which aim to be "even better in all ways" successors of already successful creations. In fact, this happens so often that the term "second system syndrome" was coined. We chose this path for Perl 6 anyways, because we believe it's the only way to get a significantly better language. (Gradual improvements are flowing into Perl 5 all the time, sometimes inspired by Perl 6.) That's why the unofficial motto of Perl 6 is "Second system syndrome done right".



As always, Perl is optimized for freedom and richness of expression.

General Purpose

Perl 5 still has a lot of builtins, showing for which task the language was created. But to become more general purpose, the core was enhanced and some other parts got removed. File IO, Network and IPC, Formats, and finally that long deprecated pre P5 DB-binding (dbmopen) went out of the core or at least into sensible module namespace. The optional type system enables compiler optimisations to make Perl even usable in cases where you had to switch to C before.

Less Unix Centric

Perl 1 was a child of the Unix world, borrowing a lot of folklore and words. As Perl 6 had to regularize the Regex and reform some other parts, this heritage has become less visible.


Per default all Perl 6 code is Unicode and you can use any weird character for any part of the syntax.

Simple English

In Perl 5 we had my, our, use, require, bless, can, ISA (is a), chop and also of course if, else, goto and all the loops. Perl 6 adds given, when, gather, take, loop, WHERE, HOW, WHO and lot more everyday English, which should make it easy to read and understand.

Hidden Complexity

Even simple Perl 6 looks pretty much like older Perl, it has a tremendous complexity inside, which only have to show up if you really need it.

Functional Programming

High order functions, currying, you name it. All that goodies are there, as Perl 6 aims to a full support of functional programming.

Object Oriented

OOP syntax is now more simplified, compact, knows about encapsulation and has many new features, including dynamic class composition.

Like in Smalltalk or Ruby every "thing" (values, vars, blocks, routines) is an object and ops are "just" methods. It helps to have internally everything in its place. But programmer still can pretend it's not the case.

Full Introspection

Because every object knows his attributes and methods, you can ask the compiler nearly anything about the state of your program.


And because classes are made to get extended or changed, Perl 6 has not only full metaprogramming capabilities, beyond your wildest dreams. It is a metalanguage with a beginner friendly default that aims to become a family of languages. Rule of thumb: everything is allowed, as long as you declare it.

One Spec - Many Implementations

Unlike every other Perl so far, Perl 6 has a precise specification, written in Perl 6. This allows to have several implementations, even syntactical compatible ones.

Design Process

The initial source of the Perl 6 design came from 361 RFC (Requests For Comments) submissions by people from all over the world. Larry Wall sorted and commented on them in several documents called the Apocalypses (meaning "revelations"). Damian Conway wrote, for each Apocalypse, an Exegesis ("explanation"), that expounded on Larry's ideas. These are now all historic documents that not have changed since, because they were much harder to maintain than the Synopses ("abstracts" or "comparisons" (with Perl 5)). The Synopsis documents, along with the test suite, form the official specification of Perl 6. More details on that can be found in the history section.

The Synopses are continually edited because the more we think about the syntax and semantics, the more detailed the picture gets. The other important input comes from the implementors. They uncover cases that worked only in theory, and important corner cases not thought about before. This mutual influence of specification and implementation we call the whirlpool model of software development, which is an intended pun on the well-known waterfall model.

Design changes are usually discussed in the p6l mailing list and the #perl6 channel on the freenode IRC network.

Intro    Chapter:   0. History   1. Design   2. Basic   3. Var   4. Op   5. IO   6. Block   7. Sub   8. OOP   9. Regex   10. Meta
Overview   Appendix:   A. Index   B. Grouped   C. Classtree   D. Delta   E. Exciting   F. FAQ   G. Glossary   H. Href