6 ms·

Even though it's not popular to say so, I really like Perl. It's a shame it got a reputation as being uncool as it's a very capable language and nice eco syste

Even though it's not popular to say so, I really like Perl. It's a shame it got a reputation as being uncool as it's a very capable language and nice eco system. I also think the criticism about unreadability is way overblow, it's what you make of it really.

lelf 13y ago
I feel it got a reputation of @#$@#(==~/s(.*)+) language that was updated last time in 1995. Not that it is uncapable of @@->$/=+s{}. But when you use Modern::Perl; use Moose; it's quite different thing. Perl6 may become cool, but it's a different story. (Moose is full-grown object system, with meta object protocol, nice syntax and implemented completely as a library).
- _delirium 13y ago
  I kind of like the non-modern parts of Perl myself, for some of the original purpose of making a superset of sed+awk+sh with better control flow constructs and better performance. I don't entirely understand why sed/awk are so popular on HN but old-style Perl isn't, since it's generally nicer and usually faster, for most things the sed/awk/sh combo would be used for.
bane 13y ago
I love Perl as well, but it's definitely gotten the reputation as the really awkward nerd among nerds. I agree with lots of people that the early stages of the Perl 6 effort caused so much distraction for so long that Perl 5 languished and lost lots of momentum for the language (and language philosophy). The Perl community, tired of waiting for a workable Perl 6 implementation, has lavished some fine attention to Perl 5 and it's in incredible shape, but the target community, also tired of waiting for Perl 6, has largely moved on. It's going to be tough to bring people back into the fold. Everybody wanted Perl 6, and it barely works now. The features in 6 are amazing (might even be enough to get folks to come back), but the implementations are all over the place. Perl 6 needs a couple of killer apps. The CFG work might have made handling XML stupid simple, but XML is quickly falling out of relevancy. The transition from Perl 5 to Perl 6 might go down as one of the great language mismanagements in history. For goodness sake, it's been in design for almost 14 years! And there are really just now workable implementations coming out, but all are still incomplete, lack features, robustness and speed. I'm also guilty of this, having been excited about Perl 6 a decade ago, but taking a wait and see approach that lasts to this day, until development settles down. I don't want to commit lots of work that will be made obsolete the next time Larry decides to change an operator's semantics. Is it finally time to sit down and learn Perl 6?
- colomon 13y ago
  As one of the Perl 6 developers, my suggestion would be to wait a couple of months. There have been some very exciting developments with the new JVM backend for Rakudo Perl 6, like new work on threads and very significant speedups on longer processes. But it's not quite yet ready for prime time -- there are definitely still some rough edges. The rate of progress on it has been phenomenal -- it essentially didn't exist at the end of March, and passes 99% of the Rakudo spec tests today. I will be very surprised if the current issues with it still exist in October. That said, there really haven't been too many major changes to the language itself (as opposed to the implementation) in the last couple of years. I have several modules I wrote 3+ years ago, and keeping them up to date with the language has only required trivial changes in that time. You could certainly start learning the language today using the Parrot implementation with the knowledge that something better will be along very shortly.
  - SimHacker 13y ago
    What is the point wasting any time on Perl 6, given it's undeniably unambiguously spectacularly awful track record, and the fact that there are so many other wonderful and better languages available to spend your time learning. There's nothing in Perl 6 that makes it any more powerful (and definitely not any more easy to learn and use and maintain) than any other language. There are no jobs programming Perl 6, and there never will be.
    - lelf 13y ago
      There's nothing in Perl 6 that makes it any more powerful You're just trolling and clearly haven't got a clue. Object system with MOP, subtyping, pattern-matching, new human readable “regexpes“, grammars etc. And for the first time in humankind it chars in string are counted by not bytes, not 2-bytes, not codepoints, not another shit, but Unicode graphemes! 2013!
    - bane 13y ago
      Grammars and proper unicode support are where my main interests lie with the language. I use Perl to "get shit done" (GSD) it's not sexy, it's not my main job, hell, it's barely reusable. But using Perl has been the difference between failure/late delivery and success/early delivery/under-budget more times than I can even remember. As late as the end of last year I has hacking Perl 5 code as a side project to GSD and took a project that went from not-existing because the scope was unfeasible for the next 3 calendar years to existing and producing measurable results 3 weeks later. That work got me a promotion and then a new better job making more money (and not doing Perl work) and overall joy and happiness because I was the guy that GSD (and Perl made that happen).
LekkoscPiwa 13y ago
I used to work with Perl 6 years. As much as I like it, going to Ruby was such a relief and joy! I never looked back. Perl is just way too @$%{}__ for me. And some stuff, like OO really, really not only looks but also feels like a dirty hack put on top of something else.
- berntb 13y ago
  Ah, so you left the Perl world just before Moose -- and don't even know about it? :-)
  - mdellabitta 13y ago
    That's a good explanation of me. I just took a quick look at the manual. The class definitions look seriously verbose. I guess with some coding templates or macros, it wouldn't be too bad, though.
    - berntb 13y ago
      Well, you have a ~ typing system for attributes, good roles, etc. And if you don't like the syntax, see e.g. http://search.cpan.org/~ether/MooseX-Method-Signatures-0.44/lib/MooseX/Method/Signatures.pm http://search.cpan.org/~ether/MooseX-Method-Signatures-0.44/... http://search.cpan.org/~flora/MooseX-Declare-0.35/lib/MooseX/Declare.pm http://search.cpan.org/~flora/MooseX-Declare-0.35/lib/MooseX... [Edit: I might add -- the OO system is mature and largely inspired by Common Lisp. It is better than the "competition" among the scripting languages (says the ones with heavier background in those environments than me).]
      - SimHacker 13y ago
        Well you should just use Common Lisp if you want an object system like Common Lisp's. It comes without all the baggage and duct tape of Perl.
        berntb 13y ago
swash 13y ago
We're using Perl for a new startup. Our team has a background in computational biology where Perl is still big. Some of us have more than 10 years of Perl experience and if you know the language well it's just insanely productive. Readability was never an issue. We follow a few common guidelines and with Perl::Tidy we get consistently formatted code. So in principle Perl is perfect for us, the only drawback is that there is less and less library support for Perl. Most services like for example Stripe have libraries for Python, Ruby but don't bother with Perl (and I don't blame them). Third party libraries from CPAN are usually not that robust and well maintained. But in general the Perl backend is the most robust part in our codebase and I would not trade it for any other language.
- SimHacker 13y ago
  Good luck finding any decent Perl programmers to hire. And if you do find them, good luck paying their exorbitant hazard pay and keeping them happy enough to stick around. And then after they leave, good luck finding more Perl programmers to willing to clean up the unmaintainable mess they left behind.
  - antitrust 13y ago
    You should come to one of our local PerlMongers meetings (http://pm.org/ http://pm.org/). I think it would change your mind about a number of things!
  - arc_of_descent 13y ago
    Whats with the luck factor? You can hire newbies and train them to learn Perl. The Learning Perl books are fantastic. If you have an experienced team and new programmers willing and eager to learn Perl, lady Luck can go for a fuck.
  - Diederich 13y ago
    I do a fair amount of Perl hiring, and here in the south bay (silicon valley), it's been pretty easy to hire good Perl talent.
antitrust 13y ago
I agree. Perl is great and often the fastest way to git 'r done, especially across multiple operating systems. Perl got a bad rep for two reasons. First, writers of boutique languages wanted a reason to feel good about themselves. Second, certain segments of the Perl community (I won't name names) insisted on writing obfuscated, baffling code that read like line noise from a modem. The Perl community counteracted this, but there were still enough of these people out there to give it a bad rap. In the meantime, if you're a Perl programmer who can write functional, efficient, clean and readable code, there's no shortage of stimulating work.

lelf 13y ago

I feel it got a reputation of @#$@#(==~/s(.*)+) language that was updated last time in 1995. Not that it is uncapable of @@->$/=+s{}. But when you use Modern::Perl; use Moose; it's quite different thing. Perl6 may become cool, but it's a different story. (Moose is full-grown object system, with meta object protocol, nice syntax and implemented completely as a library).

_delirium 13y ago
I kind of like the non-modern parts of Perl myself, for some of the original purpose of making a superset of sed+awk+sh with better control flow constructs and better performance. I don't entirely understand why sed/awk are so popular on HN but old-style Perl isn't, since it's generally nicer and usually faster, for most things the sed/awk/sh combo would be used for.

bane 13y ago

I love Perl as well, but it's definitely gotten the reputation as the really awkward nerd among nerds. I agree with lots of people that the early stages of the Perl 6 effort caused so much distraction for so long that Perl 5 languished and lost lots of momentum for the language (and language philosophy). The Perl community, tired of waiting for a workable Perl 6 implementation, has lavished some fine attention to Perl 5 and it's in incredible shape, but the target community, also tired of waiting for Perl 6, has largely moved on. It's going to be tough to bring people back into the fold. Everybody wanted Perl 6, and it barely works now. The features in 6 are amazing (might even be enough to get folks to come back), but the implementations are all over the place. Perl 6 needs a couple of killer apps. The CFG work might have made handling XML stupid simple, but XML is quickly falling out of relevancy. The transition from Perl 5 to Perl 6 might go down as one of the great language mismanagements in history. For goodness sake, it's been in design for almost 14 years! And there are really just now workable implementations coming out, but all are still incomplete, lack features, robustness and speed. I'm also guilty of this, having been excited about Perl 6 a decade ago, but taking a wait and see approach that lasts to this day, until development settles down. I don't want to commit lots of work that will be made obsolete the next time Larry decides to change an operator's semantics. Is it finally time to sit down and learn Perl 6?

colomon 13y ago
As one of the Perl 6 developers, my suggestion would be to wait a couple of months. There have been some very exciting developments with the new JVM backend for Rakudo Perl 6, like new work on threads and very significant speedups on longer processes. But it's not quite yet ready for prime time -- there are definitely still some rough edges. The rate of progress on it has been phenomenal -- it essentially didn't exist at the end of March, and passes 99% of the Rakudo spec tests today. I will be very surprised if the current issues with it still exist in October. That said, there really haven't been too many major changes to the language itself (as opposed to the implementation) in the last couple of years. I have several modules I wrote 3+ years ago, and keeping them up to date with the language has only required trivial changes in that time. You could certainly start learning the language today using the Parrot implementation with the knowledge that something better will be along very shortly.
- SimHacker 13y ago
  What is the point wasting any time on Perl 6, given it's undeniably unambiguously spectacularly awful track record, and the fact that there are so many other wonderful and better languages available to spend your time learning. There's nothing in Perl 6 that makes it any more powerful (and definitely not any more easy to learn and use and maintain) than any other language. There are no jobs programming Perl 6, and there never will be.
  - lelf 13y ago
    There's nothing in Perl 6 that makes it any more powerful You're just trolling and clearly haven't got a clue. Object system with MOP, subtyping, pattern-matching, new human readable “regexpes“, grammars etc. And for the first time in humankind it chars in string are counted by not bytes, not 2-bytes, not codepoints, not another shit, but Unicode graphemes! 2013!
  - bane 13y ago
    Grammars and proper unicode support are where my main interests lie with the language. I use Perl to "get shit done" (GSD) it's not sexy, it's not my main job, hell, it's barely reusable. But using Perl has been the difference between failure/late delivery and success/early delivery/under-budget more times than I can even remember. As late as the end of last year I has hacking Perl 5 code as a side project to GSD and took a project that went from not-existing because the scope was unfeasible for the next 3 calendar years to existing and producing measurable results 3 weeks later. That work got me a promotion and then a new better job making more money (and not doing Perl work) and overall joy and happiness because I was the guy that GSD (and Perl made that happen).

LekkoscPiwa 13y ago

I used to work with Perl 6 years. As much as I like it, going to Ruby was such a relief and joy! I never looked back. Perl is just way too @$%{}__ for me. And some stuff, like OO really, really not only looks but also feels like a dirty hack put on top of something else.

berntb 13y ago
Ah, so you left the Perl world just before Moose -- and don't even know about it? :-)
- mdellabitta 13y ago
  That's a good explanation of me. I just took a quick look at the manual. The class definitions look seriously verbose. I guess with some coding templates or macros, it wouldn't be too bad, though.
  - berntb 13y ago
    Well, you have a ~ typing system for attributes, good roles, etc. And if you don't like the syntax, see e.g. http://search.cpan.org/~ether/MooseX-Method-Signatures-0.44/lib/MooseX/Method/Signatures.pm http://search.cpan.org/~ether/MooseX-Method-Signatures-0.44/... http://search.cpan.org/~flora/MooseX-Declare-0.35/lib/MooseX/Declare.pm http://search.cpan.org/~flora/MooseX-Declare-0.35/lib/MooseX... [Edit: I might add -- the OO system is mature and largely inspired by Common Lisp. It is better than the "competition" among the scripting languages (says the ones with heavier background in those environments than me).]
    - SimHacker 13y ago
      Well you should just use Common Lisp if you want an object system like Common Lisp's. It comes without all the baggage and duct tape of Perl.
      - berntb 13y ago

swash 13y ago

We're using Perl for a new startup. Our team has a background in computational biology where Perl is still big. Some of us have more than 10 years of Perl experience and if you know the language well it's just insanely productive. Readability was never an issue. We follow a few common guidelines and with Perl::Tidy we get consistently formatted code. So in principle Perl is perfect for us, the only drawback is that there is less and less library support for Perl. Most services like for example Stripe have libraries for Python, Ruby but don't bother with Perl (and I don't blame them). Third party libraries from CPAN are usually not that robust and well maintained. But in general the Perl backend is the most robust part in our codebase and I would not trade it for any other language.

SimHacker 13y ago
Good luck finding any decent Perl programmers to hire. And if you do find them, good luck paying their exorbitant hazard pay and keeping them happy enough to stick around. And then after they leave, good luck finding more Perl programmers to willing to clean up the unmaintainable mess they left behind.
- antitrust 13y ago
  You should come to one of our local PerlMongers meetings (http://pm.org/ http://pm.org/). I think it would change your mind about a number of things!
- arc_of_descent 13y ago
  Whats with the luck factor? You can hire newbies and train them to learn Perl. The Learning Perl books are fantastic. If you have an experienced team and new programmers willing and eager to learn Perl, lady Luck can go for a fuck.
- Diederich 13y ago
  I do a fair amount of Perl hiring, and here in the south bay (silicon valley), it's been pretty easy to hire good Perl talent.

antitrust 13y ago

I agree. Perl is great and often the fastest way to git 'r done, especially across multiple operating systems. Perl got a bad rep for two reasons. First, writers of boutique languages wanted a reason to feel good about themselves. Second, certain segments of the Perl community (I won't name names) insisted on writing obfuscated, baffling code that read like line noise from a modem. The Perl community counteracted this, but there were still enough of these people out there to give it a bad rap. In the meantime, if you're a Perl programmer who can write functional, efficient, clean and readable code, there's no shortage of stimulating work.