Hacker News Reader

Comment by 3rdAccount | Hacker News Reader

Perl5 definitely isn't the future IMO, but many will point out that neither is Perl6.

Perl5's CPAN is still very useful, the language is still available on most UNIX systems, it has some really nice OO modules like Moose, and the community still has lots of conferences and releases regular updates. I personally prefer Python in this space though.

Perl6 on the other hand is gorgeous. You can write beautiful OO, FP, imperative...whatever code. Grammars and MOP are great. The parallel support is cool and the Unicode support is top notch. You can also do really cool stuff not normally seen in scripting languages like restrict the type going into a function to be a non-negative integer in the range of 1..10 kind of like in Ada. The downside is that the language is still immature and adoption is slow. The VM is getting better, but with things like Nim, Julia, Crystal(fast languages), Perl6 is a tougher sell.

View on HN

Not sure how you see Perl6 and Nim/Crystal as being competitors...they occupy completely different niches.

Perl6 is supposed to be a new approach for dynamic scripting languages. Maybe you think stuff like gradual typing means it is trying to compete with systems languages...but I don't even think it is the case that "idiomatic" Perl6 will have a lot of type annotations and I personally don't use them much.

I think a better comparison is Racket. Racket is also trying to push the boundaries for dynamic languages. I do not consider Python3 to be a competitor because it seems quite clear that Python3 is not intended to offer up much in the way of fresh thinking, just a cleaned-up Python2 (which is perfectly fine).

They are in different niches to a degree, but when you have a language like Nim that is as easy to use as Python, nearly as expressive as Python...etc, then you might go ahead and choose Nim over something like Perl6 as it is so fast and distribution is just a binary. C++ is in a different ballpark than Perl6, but not Nim & Crystal. They overlap to the slower lang's detriment.

You are comparing these tools based on superficial syntactic characteristics and completely missing the dynamic capabilities.

You should look at a tool like Perl (5 or 6) much like you would a Lisp - a dynamic system that provides interesting and useful runtime features.

I use Perl when I want features like `eval`...or when I want to push data around while delaying evaluating its type.

This is why I suggest comparing the Perls to tools like Racket.

I will never understand why it was not possible to leave Perl5 as Perl and call Perl6 something else, maybe 'Butterfly'. They share very little except their creator.

One of the most active Perl 6 contributors made a repeat plea to Larry on the topic of its name just the other day: https://perl6.party/post/A-Request-to-Larry-Wall-to-Create-a...

And it has been mostly rejected by everyone else, just like the first time it was made.

This is a silly debate but I suppose Zoffix has contributed enough awesome code to buy himself some credit.

In any case, it is Perl6 and changing the name now would be an even bigger mistake.

Larry was just discussing this with my wife (the official(?) Perl poet) yesterday. His wife and my wife didn't like the name he proposed. I kinda liked it.

I don't feel at liberty to disclose the name, but it might be in the recent irc logs.

I'd have called it Amber. A natural gemstone, as is a pearl, but takes a lot longer to form :)

The same reason the spiritual successor to Modula was named Modula-2 by Wirth.

Reply on news.ycombinator.com