Status of Perl 6 | Hacker News Reader

Status of Perl 6

perl6.guide

60 pointspeteretep46 comments

> there's a usable alpha, ready to download and use right now.

Back in 2004 we had to choose the programming language for a new project. PHP5 (which was just released at the time) was the top candidate, but I really wanted to use Perl6 which a lot of people were very enthusiastic about and whose weekly mailing list updates made it look like an early partially usable version which could be used for staring the project was just around the corner.

In the end we went with PHP. 10 years later it looks like we made he right decision.

I realize that a lot of factors contributed to Perl6 being this late, including several cases of plain old bad luck, bit I still can't help being disappointed that the language I thought would be the next big thing might not get much use at all...

> Perl6 which a lot of people were very enthusiastic about and whose weekly mailing list updates made it look like an early partially usable version which could be used for staring the project was just around the corner.

If there's one thing that I would point to about why Perl is no longer a popular language, this is it. I too remember all the hubub and excitement. I think I even have a copy of the O'Reilly Perl 6 Essentials somewhere waiting for a v1.0 to come out any day now.

Perl 6 should have really been a staged release. v1.0 will have this, 2.0 will have this, etc. As languages age they add new features and this wouldn't have been viewed as odd at all.

But there are people who have gone from middle-school to millionaire in the time it's taken for a non-alpha Perl 6 to make it out the door.

You would have gone with a language that was not released yet? Two years later you would have gone with PHP6?

A more realistic approach would have been to compare PHP5 with Perl5, where the latter would offer a migration path to Perl6 if it had become popular.

(Not only had Perl5 compared favorably back then, the core of the language has been stable, deployment looks similar and there has been much fewer security issues with the language itself.)

Ground up rewrite failed? So surprised.

It's interesting how much negativity Perl 6 (or Perl, in general) generates, and how dismissive folks are of it's usefulness. The fact is that Perl is used extremely widely in a huge variety of roles. New projects are being launched in Perl (DuckDuckGo is a relatively recent thing that has been very successful, but there are many others), and a lot of very smart people work in Perl, both 5 and 6.

If Perl isn't your bag, or Perl 6 is taking too long to be "finished" for you to want to use it, that's fine. There are hundreds of languages to choose from. I understand the frustration of longtime Perl developers; Perl 6 is the Perl developer vanguard seemingly turning their backs on "pragmatic" and running somewhat haphazardly toward the theoretical and (very far) forward looking (Perl 6 contains many more novel concepts than a lot of newer languages, such as Go).

Could it, should it, have been done differently? Incrementally, with real-world testing for each new feature? Probably. Hindsight is 20/20. Is it a victim of second system effect? Possibly, we won't know until it hits some reasonable levels of production; right now it's still quite academic. Hubris is one of the Perl virtues, and it may have led to "perfect is the enemy of the good" winning out on more decisions than it should have.

But, Perl 5 is a great language today, certainly a match for Python or Ruby for conciseness, performance, capability, etc. And a vastly better language than Perl 5 was 14 years ago. Perl lost the popular mindshare war..but, it still helps people build a lot of cool software in a lot of fields.

Regardless of whether you'll use Perl 6, I'd love it if people here at HN would actually take a look at the novel ideas contained therein, rather than focusing on the superficial "it's late and I don't like it". There are concepts in Perl 6 that could improve your favorite language, and might alter the way you think about solving problems. If only you'd check it out. I'd wager most of you have read a bit about Haskell or Go or Erlang or Scala, without ever seriously intending to develop in it. I would suggest the same approach to Perl 6.

The P in PERL stands for "practical". Switching the story and saying that it's now an academic pursuit feels like abandonment to some. Programming in perl5 was part of a livelihood for many people; uncertainty and decade-long limbo is very poor leadership. (And leadership is a privilege; followers get mad when they lose faith in a leader.)

That being said, perl6 is interesting, so I think looking at it more like haskell could be valuable.

Larry Wall also says that the P can equally be validated as "Pathetic". And it's a lowercase Perl, not PERL.

Source: the Camel book.

The P in perl also comes from the P in BCPL, a language from 1966, also according to Larry Wall.

And from BCPL came, in order...

* B language

* C, and its statically-typed "C" derivatives C++, Java, C#, ObjectiveC

* P for perl, and its dynamically-typed cousins and derivatives Python, PHP, Ruby, etc

* L for whatever's next, must be the final victory of Lisp

[citation needed]

> Source: the Camel book

> [citation needed]

You didn't have a citation on your own quote when I responded to yours without a citation. Perhaps you should have written "Whoops, I neglected to add the source of my quote when you replied but have done now. I can't, however, find the source of yours -- where did you get it?".

As for my quote about the acronym of BCPL being a premonition of the languages after it, I read it somewhere sometime before 1997 which was the last time I programmed in Perl.

So, we've confirmed that:

* You maybe programmed in Perl once in '97.

* You didn't know the difference between PERL and Perl.

* You don't know what the camel book is (Programming Perl by O'Reilly), and have never even read the first chapter.

* you have no citation for your BCPL statement.

Thank you for the information, but it doesn't really detract from my point. Anyone going to perl5 around 2000 was under the impression that it was a very practical language.

I was one of the people championing a Pathetically Eclectic Rubbish Language back in 2000, and I was definitely not alone. But you win points for making a "No True Scotsman" Fallacy[1].

1. http://en.wikipedia.org/wiki/No_true_Scotsman

I feel the biggest problem for Perl 6 at least in terms of mindshare is that Perl's spiritual successor already offers most of Perl 6's promises and it's already in production. It's called Ruby.

I agree partially. When Ruby took off,it captured a lot of programmers that otherwise would have probably gravitated towards Perl (or Perl6).

But if Perl6 had been finished in 2005 or 2006, things might look quite different today.

I'm not sure. imo Perl5 developers tend to be more conservative than the average developer, which makes sense given that maintenance of large sets of Perl5 code is very difficult and requires a lot of discipline; so taking a risk is a lot more expensive for most Perl5 devs. I feel that people moving from Perl5 to Ruby only did so after Ruby became established.

Perl 6 was not written with Perl hackers in mind. It was written with programmers from other languages in mind in hopes to draw them into the Perl ecosystem and I think that was a bad idea. It basically made Perl 6 DOA.

What we really needed was a Perl 5.5. A cleaner syntax and OO that isn't shoehorned onto the language. Many of the "features" of Perl 6 aren't really features at all, and in many ways Perl 6 doesn't stand up to other languages that it might compete with (JavaScript, for example).

I think this is an excellent point. Perl excels at facilitating a wide variety of users, and as a language it really does feel like everything and several variations of a kitchen sink included. That openness to any idea is seen as inelegant by some, and accommodating by others.

Early on Perl 6 felt like a different language, and while that has moderated somewhat, it still doesn't feel very 'perlish'. It will be interesting to see where it goes.

Indeed, when I first heard about Perl 6 so long ago, it seemed to be trying to remove some of the most quirky, but also most useful features of Perl 5. I more or less immediately decided to ignore it, so I haven't followed what's been going on since apart from being secretly pleased that it has never really shown up.

Yeah, Perl 6 just doesn’t appeal to me as someone who uses Perl 5 on the regular.

Now, a cleaner syntax? Maybe it would be better to avoid the magic variables, but one thing Perl 6 did away with which I really like about Perl 5 is that the dereferencing operators ($scalar, @array, %hash, &sub, *glob) have distinct types, and “use warnings/strict” can catch their misuse. Dynamic typing is a lot more tractable for large software when you have succinct explicit typing.

As for non-shoehorned OO, maybe it would be easier for classes-and-inheritance people to learn Perl if it had a more conventional object system, but the fact that you can bless any reference with a package name and use it as an object is really simple and really powerful, and I like it a lot.

I am a particular fan of the infix operators, ala:

https://github.com/cosimo/perl6-digest-md5/blob/master/lib/D...

Some of the operators in that file make even the craziest inventions of Scala users look tame.

I think the syntax highlighter felt pretty much the same way as you do.

"Keep using Perl 5 for production."

I'd really love to see and outsider do writeup on why Perl 6 is where it's at. It'd be a great lesson in project management, time and scope boxing and feature creep. I mid-career now, and Perl 6 was announced when I was an intern. At this rate, production releases will come out when I'm deciding on early or standard retirement.

Meanwhile, Perl 5 seems to have gotten all kinds of wonderful new life breathed into it.

I'd really love to see and outsider do writeup on why Perl 6 is where it's at.

JWZ explained it more than a decade ago: http://www.jwz.org/doc/cadt.html

I have never heard that the Perl 6 developers are teenagers. However, the truth is that Perl 6 inspiration had a lot of elements that sound... strange.

http://www.theregister.co.uk/2008/12/16/larry_wall_perl_6/

And then there's the first moment of inspiration (in 2000, that is, 14 years ago):

http://www.developer.com/lang/perl/article.php/3076571/Perl-...

"An hour into the discussion, around the time most people nod off in any meeting, Jon Orwant (the reserved, universally respected editor of the Perl Journal) stepped quietly into the room and snapped everyone to attention with an entirely uncharacteristic and well-planned gesture. Smash! A coffee mug hit the wall. “We are *@$!-ed (Crash!) unless we can come up with something that will excite the community (Pow!), because everyone´s getting bored and going off and doing other things! (Bam!)” (...)

Within a few hours the group settled on Perl 6, a complete rewrite of Perl. The plan wasn´t just a language change, just an implementation change,or just a social change. It was a paradigm shift. Perl 6 would be the community´s rewrite of Perl, and the community´s rewrite of itself."

I have never heard that the Perl 6 developers are teenagers.

The important point of JWZ's article isn't "teenagers". It's "rewriting everything from scratch... happens, over and over again". For example, Rakudo's current "Great List Refactor": http://pmthium.com/2014/10/apw2014/

I've read it, it's not "rewriting" for the sake of it, they actually significantly change the behaviour too. What I never understood was why people believed that the Parrot VM had to be made.

Apart from maybe Guile, there was no really solid, cross-platform VM with a compatible license suitable for Unix interoperability and the dynamic features Perl needs. Lua would only later be a reasonable contender.

> Meanwhile, Perl 5 seems to have gotten all kinds of wonderful new life breathed into it.

I agree, and I am glad about it. One way or another, I use Perl almost daily. Well, at least two out of five workdays every week.

I have been more interested in Parrot VM rather than Perl 6 itself. I like the idea of a unified VM, though it looks unlikely to take off. The ability to share code/libraries between languages would be awesome.

Besides the JVM and CLR pointed in sibling threads.

What about P-Code (1978), M-Code(1978), OS/400 TIMI (1988), or the granddaddy UNCOL (1958)?

Parrot was hardly a new idea.

do you mean just like the CLR?

http://en.wikipedia.org/wiki/Common_Language_Runtime

Or the JVM. Many languages compile to it, even some with their own interpreters (one example: Ruby). Or the Erlang VM (Beam). There is at least Elixir that compiles to its bytecode.

So, more or less the same as it has been...

Is perl 6 now officially the Duke Nukem Forever of lightweight languages?

Getting there. DNF took 15 years. Perl 6 will surpass that if we don't get a real release in 2015.

But the real champions for delayed development are LaTeX 3 at something like 20 years & counting, and the GNU Hurd at 24 years & counting.

The most protracted piece of vaporware in history is actually Project Xanadu, at a whopping 54 years.

Is someone actually still working on that?

Yes. In fact, OpenXanadu was unveiled this year: http://xanadu.com/xanademos/MoeJusteOrigins.html

That said, it's mostly just a web reimplementation of the incomplete 1998 Udanax client.

Could be! I recall it being unofficially "Duke Nukem Forever" as early as 2010: http://stackoverflow.com/q/3367226/240438

I'll switch when Python 3, IPv6, and IE11 are the most-popular.

What's interesting, is that you called out three technologies that are all definitely going to be market leaders in their particularly category over time. If we consider IE 11 to be the "Category of Microsoft Browsers", Python 3 to be the Category of Python Languages, and IPv6 to be the category of IP.

Here are the dates on which each will be category leaders:

  o IE 11 - 2014
  o Python 3 - 2016
  o IPv6 - 2022

I find those years reasonably believable, but where did they come from?

IE - I browsed for Market Share. Python 3 - several years of carefully watching mindshare, assessing the challenges with going to Python 3 (mostly the more cognitively complex (if somewhat more consistent) challenges with dealing with Unicode/Strings/Bytes, plus my experience shifting from Python 2 to Python 3, cross referenced with watching new application uptake of Python 3 + 3rd party Library Support).

And IPv6 is my full time job. I teach it to a lot of people, and see it deployed all over the world - so I have a pretty good sense of whether it's going to happen (it will), and how long it will take (quite a while, Carrier Grade NAT could actually keep IPv4 going forever)

What's popularity got to do with it?

Reply on news.ycombinator.com