What is Perl 6?

chromatic writes "Perl.com has a new article entitled What is Perl 6?. It analyzes the changes to the language in light of the good and bad points of Perl 5 and provides new information about the current state of the project: Perl 6 exists, you can write code in it today, and it's more consistent and easier to use than Perl 5."

Comments

unfortunately

[caffeinemessiah]

You can never be told what Perl is.
You just have to see it for yourself.

sorry, i just had to.

Re:unfortunately

[value_added]

You can never be told what Perl is.
You just have to see it for yourself.

Is this what you saw?

What is perl?

[Anonymous Coward]

Baby don't hurt me,
Don't hurt me
No more

Re:What is Perl 6?

[Perey]

That's the thing, though. PHP is the big name (from management's perspective), at least in the P category of LAMP, right now. Not that Perl's gone away by any stretch of the imagination, but the existing Perl shops are happy to keep on doing what they're doing, while the PHP advocates crow about how many new jobs are being done in their language.

So is Perl 6 going to bring about a Perl revival, or is it (as I suspect) going to fall flat when faced with Perl 5's quietly entrenched support and PHP's proclaimed grip on new uptakers? TFA mentions the reasons for cutting backwards compatibility (or at least reducing its priority) far too often for me to be optimistic there.

I think Perl 6 will catch on, eventually... but it's going to be more of an alternative language, not an upgrade, to Perl 5 for a long time yet.

No language that I like better

[BadAnalogyGuy]

I never really understood Data Structures until I learned Perl. I was consistently and thoroughly confused in my DS class. The language used there was C++. There was simply too much baggage in the language that obfuscated the very points we were being taught. If you can't get past the template syntax, how in the world are you going to be able to understand the data structure concepts?

Then I met Perl (5.003). What a difference it made! The data structures were built in, and on top of that, it was EASY to nest structures to build complex data types. It was like having a semester of Data Structures immediately made clear.

Then I found myself back with C++ again. First I wrote my own List classes. However I soon realized that STL made available exactly the types of data structures that Perl has. Maps, Lists, Vectors. And since I understood what I was doing in Perl, it was so much easier to catch on with C++.

Perl taught me C++. Who would have thought?

Re:No language that I like better

[patio11]

Perl taught me C++

Somewhere, a maintenance programmer just slit his wrists.

Re:No language that I like better

[Lisandro]

Can you recommend a resource for OOP with Perl?

Right now i found all i needed in the Perl.org site - this OO tutorial for Perl is pretty complete. There's also this one, which is oriented to begginers.

    In fact, i always keep a browser window open to Perl.org when i'm coding Perl - the tutorials are very nice, but the function reference has been priceless to me.

Re:No language that I like better

[AuMatar]

Ignore the analogies, OOP has nothing to do with obkects at its core. OOP is basicly encapsulation and interfaces. You design your program so that its broken into parts. Each part handles one bit of functionality- networking, or database communication or the like. Each part is self contained, no 2 parts should know how the other part is coded- they only communicate with each other through a set of functions. These functions are called an interface. The functionality of each part is said to be encapsulated, because the actual design of the code is not known to the outside world.

In C, you do this with functions in a .h file and call it a library or a module. Sometimes you use a struct to pass around data. A great example is the stdio library- you have a bunch of library functions to do IO. You don't know how those are implemented. ANd since you can have open IO on more than 1 file at a time, you need a bit of data to hold the state of an IO. Thats why you pass back and forth FILE* parameters.

In C++, you put the data and the functions together in one package and call it a class. Think of a class as a C struct with a bunch of function pointers in it. So instead of calling myfunction(mystruct) you call myclass.myfunction(). Its conceptually the same (in fact, the machine code is almost identical).

By now you should be thinking "damn, that sounds like what I do in C already". It is. Good C programmers did OOP before OOP was a buzzword. languages like C and Java just add some syntactic sugar like inheretance (one of the most abused language features ever, especially in Java) and autoatic constructiors and destructors into the mix.

Re:No language that I like better

[Anonymous Brave Guy]

Another is swallowing a bullet, while a third fell on his sword. There's more than one way to do yourself in!

What is Perl 6?

[Rob_Ogilvie]

I can't define it, but I know it when I see it.

New Perl excitement

[ChrisDolan]

What makes Perl strong, in my opinion, is the community's interest in maintaining a large and well-tested library of useful code in CPAN. Without CPAN, it's not clear that Perl would be as alive and healthy as it is today.

What Perl 6 offers is a rejuvenation of the language. Perl 5 still works great (better than ever due to new efforts to stamp out even the most obscure bugs) but this new revision is attracting some *really* smart people who are bringing interesting new ideas to the language. Audrey Tang and Luke Palmer come to mind right away.

My greatest hope, however, is not that a revitalized Perl will squash the other dynamic languages (Python, Ruby, PHP, ECMAScript, etc) but will instead bring them into a state of interoperability. I really, really want Parrot to succeed so well that the other languages decide to target it as a backend so I can trivially call Python or C libraries from Perl and vice versa.

Re:New Perl excitement

[killjoe]

Parrot is multi dispach and supports multiple inheritance. It also does not have a patent sword hanging over it like mono does.

Q. What is Perl 6?

[dazlari]

A.Two Kiwi oysters going at it.

WHOOPITUPTITUDE!

[reidman]

From TFA:

Not everyone who starts learning Perl for whipituptitude needs manipulexity right away, if ever, but having a tool that supports both is amazingly useful.

Whipituptitude?!

That is awesome. Made up words a--

Whats this? Manipulexity?

How much awesome can you cram into a single sentence?

My short experience with perl...

[rsidd]

Three years ago, I could program in C, but had never used a scripting language (except bash, for very basic stuff). I needed to do some non-trivial manipulation of text files and figured that this was a good time to learn. Since others in the group were using perl, I tried perl.

I knew what I wanted to do, but needed to learn the language. I struggled with the awful syntax for three days. The breaking point came when I wanted a list of lists and realised that Perl "flattens" nested lists. How do you write nested lists such as [[1,2],[3,4],5,[6,7,8]]? In Python, it's trivial (that's how you'd write it), but in perl, nobody I talked to could give me an answer. It flattens it, unasked, to [1,2,3,4,5,6,7,8] and, try as I might, I can't see the point. (It turns out it's possible to have nested lists, but it's yet another example of perl's horrendous syntax).

Finally, I decided to give python a try. I spent an hour reading the python tutorial, and in another three hours, I had reimplemented everything I'd done in the last three days in perl, and an hour after that I'd finished the job. Python syntax was, and still is, the cleanest I've ever seen. It's an amazing language. And it changed the way I think about programming: it gave me an appreciation of functional methods (I now use ocaml a lot) and also changed the way I write C (vastly for the better).

That was it. No more perl for me.

Re:My short experience with perl...

[cliveholloway]

Oh please.

"Especially when every perl doc I see around tells me to use curved parentheses for lists, and @ prefixes for variables that refer to them..."

How hard did you look, really? If you go to Google and type in perl list of lists, the FIRST link takes you here.

And within 1/2 a page, you see this:

# assign to our array a list of list references
@LoL = (
[ "fred", "barney" ],
[ "george", "jane", "elroy" ],
[ "homer", "marge", "bart" ],
);

print $LoL[2][2];
bart

Damn anti-Perl trolls :-)

Re:My short experience with perl...

[forty7]

Yes, that's really intuitive, thanks. Especially when every perl doc I see around tells me to use curved parentheses for lists, and @ prefixes for variables that refer to them, and I have no clue what data structure you've used above.

You're right; you do need curved parens for real lists. It may be helpful to think of the @ mark as referring to multiple values, rather than to a list specifically. It's also used for list and hash slices, like this: @list[1,3,5] and @names{'tom', 'dick', 'harry'} (Both of those expressions really just evaluate to lists, but, particularly with the hash slice, that's not necessarily obvious).

The other responders, in their haste to point out how much perlier-than-thou they are, have glossed over the fact that 'my $list = [[1,2],[3,4],5,[6,7,8]];', as you've pointed out, doesn't exactly produce a list of lists. It produces a data structure that's close enough to a list of lists that they can call it that without being entirely wrong.

(Warning: Perl content. I love Perl, so it'll probably sound like I'm rambling. My apologies.)

You can't directly create a list of lists in Perl. Not possible, because they'll get flattened. But what you can do is create a list of *references* to lists. There are two ways to do this:
my @list1 = (1, 2, 3);
my @list2 = (4, 5, 6);
my @combined = (\@list1, \@list2);

Or:
my @combined = ([1, 2, 3], [4, 5, 6]);

This second syntax is called the anonymous list syntax.

The upshot is that lists are surrounded by (), as you already know, but that by replacing them with [], you get a reference to a list with no other name. The same can be done for hashes with {}, which means that all Perl data structures can be arbitrarily nested.

So, your Python example, in Perl, would create a reference to a list of references to lists. Its elements would be accessed like this:
my $list = [[1,2],[3,4],5,[6,7,8]];
print $list->[0][1] # prints 2

The -> is used for dereferencing, but Perl is clever enough to figure out that the value of '$list->[0]' is also a reference, and doesn't require you to write the -> between [0] and [1].

You could also write your example like this:
my @list = ([1,2],[3,4],5,[6,7,8]); # note the ()

print $list[0][1] # *also* prints 2!

It turns out Perl is *still* clever enough to figure out that $list[0] is a reference, and will automatically dereference it for you. No -> required.

I hope this helps clear things up a bit!

Re:My short experience with perl...

[BadAnalogyGuy]

This is one area where the Perl docs (as detailed as they may be) fail new users.

Intuitively, a new user would look at the TOC and see perldata "Perl Data Types" and think that the complete definition of the 3 main Perl data types would be described. So rsidd looks for instructions on creating multidimensional arrays, sees "List value constructors" and gets this:

LISTs do automatic interpolation of sublists. ... arrays and hashes lose their identity in a LIST... To make a list reference that does NOT interpolate, see the perlref manpage.

So they head over to perlref (an extra level of indirection) and notice in Item 2:

A reference to an anonymous array can be created using square brackets:

        $arrayref = [1, 2, ['a', 'b', 'c']];

Here we've created a reference to an anonymous array of three elements whose final element is itself a reference to another anonymous array of three elements. (The multidimensional syntax described later can be used to access this. For example, after the above, $arrayref->[2][1] would have the value ``b''.)

But this isn't really easy to understand. Why does he need an arrayref when he wants an array?

@array = [1, 2, ['a', 'b', 'c']];

That isn't the same as what he wants. In fact, it's not what you'd expect from DWIM. It's a single entry array, not a multidimensional array. It's not even a list of lists (unless you perform a little magic on it).

So finally after struggling with this and ending up with some ugly monstrosity like the following:

@array = @{[1,2,\@{['a','b','c']}]};

Now his code works, but it isn't very easy to understand, and the maintainers of this code are going to tell everyone how evil and illegible Perl is because the programmer here couldn't figure out how to make a multidimensional array.

The only FAQ entry with the term "multidimensional" in it refers to some DBM-specific topic that doesn't seem to have any relation to the problem at hand. While "list of lists" may be the preferred term in the Perl community, it would be nice to have a FAQ entry like "How do I create a multidimensional array?"

As you've mentioned, perllol has the exact syntax of how to do this. Unfortunately for our poor programmer, the link to that is buried in the See Also section alongside perldsc (which is large and contains quite a bit of irrelevant information like 'use strict' information, while at the same time not providing very detailed information about the data structures themselves). The very first 'perldoc perllol' page displayed gives the answer immediately:

An array of an array is just a regular old array @AoA that you can get
at with two subscripts, like $AoA[3][2]. Here's a declaration of the
array:

        # assign to our array, an array of array references
        @AoA = (
                      [ "fred", "barney" ],
                      [ "george", "jane", "elroy" ],
                      [ "homer", "marge", "bart" ],
        );

        print $AoA[2][2];

Why is it so hard to get to this simple explanation? Why should a neophyte have to go through two documents to finally get to perllol? The FAQ should describe the technique using "multidimensional" as a keyword.

I love Perl, and I love the depth and breadth of the Perl docs, but they are difficult to navigate for Perl neophytes.

Re:My short experience with perl...

[forty7]

Dang it, copy-and-paste missed a couple of sentences at the end. Here they are, in context:

It turns out Perl is *still* clever enough to figure out that $list[0] is a reference, and will automatically dereference it for you. No -> required. The beauty of this automatic dereferencing is that it allows Perl to DWIM (Do What I Mean): Perl is perfectly content to let you pretend that a 'list of *references* to lists' is actually just a 'list of lists'. The downside of the automatic dereferencing is that it turns the easy-to-miss difference between '((1,2),(3,4))' and '([1,2],[3,4])' into Just Another Bit Of Fiddly Syntax To Remember.

Re:My short experience with perl...

[Russellkhan]

That's the first time I've ever seen or heard anyone say that Lisp didn't have enough parentheses!

Re:My short experience with perl...

[rsidd]

You put it very well. It's possible that if I had seen this doc to begin with, I'd have stayed with perl. But perhaps I would merely have switched later. This was just the last straw -- I'd already spent three days trying to figure out perl syntax (and meanwhile, one colleague was strongly urging me to try python -- everyone else used perl). And I'd already read ESR's article on his first exposure to Python, so I had a favourable impression of it.

Having used python for about 3 years now, I'm yet to find something that I can't easily do in it, that I can in other languages. Except for speed/numbercrunching issues. And then I use ocaml (for new code) or C (for existing code). Even if I liked perl syntax and hated python's whitespace-significance (I don't and I don't), I imagine I'd have migrated to ruby by now.

Actually, the really cool feature of Python, one that I use all the time now, is "list comprehensions" -- an idea stolen from Haskell, and as far as I know, no other language has it. You can't do anything with it that you couldn't with map() and filter() but it's a much more elegant way of writing things -- just the way mathematicians would with sets.

10 Years Overdue

[dorpus]

People in the mid-1990s spoke of "overnight obsolescence", that Perl 6 would replace everything in a few weeks, and that you had better learn a new programming language every month. Over 10 years later, perl 6 is still in beta mode.

Re:10 Years Overdue

[chromatic]

Over 10 years later, perl 6 is still in beta mode.

Did you read the same article I wrote or is your post from the mysterious future? Larry announced Perl 6 in the summer of 2000.

Hiring here.

[Tei]

Hello.

We need 5 years experience Perl 6 programmers for 3D game. Reference: P6DNF.

PDL

[LordMyren]

The article talks about some of the defining features for perl. Well, one of the defining features in my perl experience has been Perl Data Language, pdl. PDL _is_ whipitupitude. Its a wonderful wonderful matrix library. And it comes with the best perl shell I know.

I had to break down a equation into a sequence of linear equations. So I hacked up some PDL in like 2 hours to do that. Couldn't have been easier, even though I'd never used PDL or its perldl perl shell; I just started typing in the interactive shell until it worked as expected and until I knew what I was doing. Then I needed the results in interger, so I rounded everything down, built a permuter and sorted the permuted results for each individual segment. That took three hours, but only because I kept botching the matrix multiplication. Even with huge datasets, generating hundreds of thousands of linear equations, each spanning dozens of datapoints, permuting the linear equations, sorting them and selecting the optimal, PDL would run it all my slow arse 800mhz crusoe laptop in seconds. Matlab couldnt touch it.

Thats the other really truly thing about PDL; the performance. If someone else would chime in and do it better justice, but my crude understanding is that it generates some kind of extremely optimized machine code on first use and runs whatever equations you've thrown at it like silk from that point on.

Little late and a little off topic, but PDL really is just a masterpiece of perl hackability. The PDL perl shell is truly spectacular; get some symbolic integrators and differential equation solving packages in there and I wouldn't need to break open Mathematica or Matlab ever again. Ok, long way away, pdl is really just about matricies, but it is really really sweet, and its shell is good for anyone who just wants to try something out really quickly in fully interactive perl.

That being said, I really cant wait to see where the perl6 VM is going.

G'night!
Myren

PUGS

[putko]

Here http://www.pugscode.org/ is something on the PUGS project, which is making an implementation of Perl 6 in Haskell, conformant to the spec.

Apparently they are having a lot of fun.

Meditations on Perl 6

[starX]

If the release of Perl 5 was any indication, Perl 6 is the single magic bullet that will kill all of my (Perl) code.

Perl 6 may be more akin to a divine programming language, which makes the implimentation of complex data structures simple and sublime. Then again, it could all be a nasty trick to lead us away from the true path of enlightenment.

Perl 6 is not .NET.

Perl 6 is not controlled by any major corporation; I haven't decided whether this is advantageous or not yet.

If I were to have a child, would it be written in Perl 6?

Can Perl 6 be used to unlock the secret mysteries of the Bible code to reveal the end times?

Is Perl 6 really being developed by the descendants of Jesus Christ? Is the Pope trying to cover it up? Does the Pope know what Perl is? If so, is using Perl 5 a sin? How about Perl 6?

I bought a preview book on Perl 6 a few years ago. Is it still useful? Can I have my money back?

If Ruby was an upgrade to Perl, and Perl 6 is a an upgrade to Perl and Ruby, will Ruby need to changes their name in such a way as to play off of Ruby Tuesdays?

If I enter the Perl 6, can I change my mind later?

If Perl 6 is brillian, but no one uses it, is it still brilliant? What if it's awful and everyone uses it?

So very tired....

Re:What is Perl 6?

[Scarblac]

But if you need to do a longer project, pick up Programming Perl and read it (from front to end). Without that book, Perl can just be extremely weird, after reading the book it at least makes sense in its own world, no matter how unconnected that world seems to be to the rest of the universe :-)

About 3 years too late

[Colonel Panic]

Ok, Perl6 does indeed look cool. Lots of interesting things there. Sure, you can apparently write Perl6 code today and run it on PUGS (a Haskell implementation of Perl 6; that's gotta be speedy, eh?). But as is mentioned in the article, Perl6 was announced at OSCON 2000; that's 5.5 years ago. It's now become the posterchild for vaporware in the open source world, hence this article to keep the faithful hopeful (and to keep'em from sneaking off to Ruby, Python or even Io). Really, it just looks like the purpose of the article is to say "yes, we're still here working on Perl 6. We're working hard, we really are. Please, don't lose hope. This is hard work. It'll be here one day and it'll be great", while a lot of Perl folks who yearned for something better have already moved on to Ruby or Python.

I really hope that Perl 6 arrives one day. I'm pretty deep into using Ruby these days having left Perl 5 behind long ago (the part of the article about what's wrong with Perl 5 was really superfluous; maybe it was intended to convince the remainingn Perl folks who are happy with 5 to check out 6), but I'll give Perl 6 a look when it arrives. The grammar support alone looks pretty awesome; it'd be great to have a viable lex/yacc alternative. In the meantime I want to learn some languages that have a bit more immediate promise like Io. It seems that maybe the plans for Perl 6 were just too ambitious. Yes, it's great to start with a clean slate and try to revolutionize, but often it's evolution that wins out.

Re:Perl 6 is evolving the language into awesome!

[Anonymous Coward]

I like the use of "." for concatenation

"+" as an operator sugguests that the order of the items being operated on has no effect on the answer...

eg. 2 + 3 == 3 + 2

but
"two" + "three" != "three" + "two"

Re:The funny thing is...

[LarsWestergren]

... that they all spell that "whipituptitude" differently. Look for yourself:

        * In TFA1: whipituptitude
        * In TFA2 referenced from TFA1: whipuptitude


With Perl, there is always more than one way to spell it.

Re:Perl 6 is evolving the language into awesome!

[Viol8]

"sugguests that the order of the items being operated on has no effect on the answer..."

And a dot suggests a decimal point. Plus or double bar are used as concat
almost everywhere else. Using a dot was not very logical. But I guess that
follows the general philosphy of perl syntax anyway.

Re:Perl 6 is

[TheP0et]

Not just another virtual machine nobody needs. Perl6 is also a consequent implentation of language development paradigms. Java started some of it with its so called "platform independent" VM. Microsoft jumped on that train with DotNet and its intermediate language for the CRL. Perl6 bundles those ideas together, gets rid of hierarchy constraints inherited from the old OO drafts, and adds complete costumizability to syntax and grammar while trying to keep the number of built-in functions as small as possible.

No other language does make it so easy by now to adapt the language itself to the needs of the special environment. You could for example write a module that forces strict inheritance to all of your objects, or implicitly puts all variables and values into a protected memory range, or dump your current program state onto a disk, everything within the blink of an eye.

Other implementations of introspection, like Java's Reflection API, provide the needed self-awareness of the code and the means for dynamic object handling up to some point, but that point reached, things tend to get rather ugly. Perl5 provided some "dirty hacks" for a part of that, but couldn't cover everything without breaking compatibility and, most of all, performance.

Still Perl6 won't be the sole answer to every question - as allways TMTOWTDI. But it's certainly interesting enough to keep an eye on it.

Perl 6 = COBOL 9x

[scottsk]

Perl 6 reminds me of the super-revamped, object-oriented COBOL that came out in the 90s -- by the time this perfect language was created (and it was a decent upgrade, although the OO stuff was so verbose it wasn't funny), COBOL 85 had such a huge installed base that COBOL 9x was irrelevant. The amount of code written to the COBOL 85 standard was immense, and most new development had moved on to other languages. Maybe Perl 6 can escape the Fortran 9x and COBOL 9x trap of being really great languages about a decade too late. The biggest hurdle Perl 6 is going to face is its own installed base. It will have to be 100% compatible with Perl 5 to get people with a code base of Perl 5 code to even think about using it. Plus, it needs to have a compelling story to tell. I was excited about Perl 6 five or six years ago.

Re:Too many changes all at once

[Lost Found]

Whenever I hear people saying things like this I think that they really lack visibility or understanding into the language's design process. Granted, the coming-to-life of Perl 6 hasn't been the quickest miracle we've seen, but I'm a hundred thousand percent positive that its being done in the best possible manner.

Think about it. Larry Wall accepted numerous RFCs from programmers of all walks, discussing Perl's problems / desires for new features / suggestions for new implementations / ideas how to change the syntax. He commented on each one, indicating whether (1) he agreed with the problem, (2) whether he agreed with the solution, (3) what, if anything, he thought should be done about it.

In the mean time, a radically new language glue system is introduced - Parrot. Perl had such wild success with XS - granted, Parrot isn't just about making language A talk to language B, but it's certainly an example of natural evolution.

As for Pugs, it's been fantastic. It's allowed lots of people to write real and working Perl 6 code (including lots of tests) to evaluate all aspects of the Perl 6 design before it goes into production.

Now, I'm not addressing you directly with this last part; rather, a greater community of Slashdot trolls. If you don't feel like Perl is for you, or if you feel like Perl is no longer for you, fine. Find your way to Ruby, Python, Java or whatever floats your boat.

But please, it's getting really goddamned irritating to have to sift through the comments of a handful of armchair morons that sit at home, interfacing with something called "comments.pl", eating doritos and talking about how the greater Perl community should just drop everything and go to language X, or repeating a tired meme about how the language is making no progress at all (when all they need to do to see the massive progress is read Audrey Tang's blog or visit pugscode.org). And then, there are some mods that feel it appropriate to mark clueless jabs as "insightful".

I am thankful of one thing - Perl's momentum. While everyone else is barking about how (name my scripting language) is great this week for doing web pages or some nonsense, there is still a huge community of devoted, bleeding edge language researchers and smart people, chisel in hand, forming Perl 6 from the rocks.

And while the naysayers are switching languages once a week as they make incremental advantages over eachother -- while they're totally clueless that so many of the 'advancements' in their own languages over the years have been 'borrowed' from or 'inspired' by perl, the aforementioned language scientists are preparing to do once again what Larry Wall did, intentionally or not, when he released Perl on the world - bring about a revolution.

Re:Perl 6 is evolving the language into awesome!

[TheLink]

Because in Perl
"1"."2" eq "12"
  and
1 + 2 == 3

If a loosely typed language is using + for concatenation, it's poorly designed (you'd end up typing more to specify what you want done).

You need to know that the concatenation of two variables is not the same as adding them together.

Slightly relieved that Perl 6 switched from using underscore to tilde for concat - underscore is overloaded with so many other tasks already. Unfortunately ~ still requires shift to be pressed on my keyboards, but I guess they are running out of symbols, and at least I think ~ won't require you to keep putting spaces around it to disambiguate it from other meanings.