5 ms·

Perl 6 in 2011 - A Retrospection

67 points by perlgeek 15y ago

spidaman 15y ago
Back in the day, I was an avid mod_perl user and Perl 5 enthusiast. I recall being optimistic that Perl 6 would come out of the oven in 12 to 18 months. Sadly, that was in 2004. My feelings about Perl 6 are like those about a Microsoft OS that is stable, secure and performant: a nice idea that has grown into a myth. Sorry to dump on Perl, I still have many friends who use Perl 5 on daily basis. I think for a lot of folks, the real "Perl 6" is... Ruby.
- chromatic 15y ago
  I think for a lot of folks, the real "Perl 6" is... Ruby. That doesn't make sense; why would a "real Perl 6" be measurably worse than Perl 5?
  - spidaman 15y ago
    A "ruby vs. perl" fight is the furthest from my mind, I'm just speaking anecdotally about what my friends who, like me, were heavy perl users are using in it's place now. If the "measurably" bit refers to how the runtimes benchmark, yes, absolutely, it doesn't make sense.
    - chromatic 15y ago
      If the "measurably" bit refers to how the runtimes benchmark... That and language stability, ecosystem, internal coherence, library availability, release process, compatibility, deployment concerns, tooling....
mjs 15y ago
So there are now three Perl runtimes/compilers/interpreters--Rakudo (+ Parrot VM), Niecza (CLR), Perlito (?)--none of which yet implements the full Perl 6 spec? (Which itself is not yet complete, although "2011 was a rather quiet year in terms of spec changes.")
- perlgeek 15y ago
  That's correct. They all approach Perl 6 from different angles, and with different goals and techniques. I'm sure you also know that the specifications of Java, C, C++, python, Haskell and other languages are still being developed too.
  - squiggly101 15y ago
    So when does work start on the one that will be angled for use in the real world, you know like Java, C, C++, python, Haskell and other languages?
    - masak 15y ago
      I'm trying to think of a way to read your question that doesn't make it dismissive and condescending. Why wouldn't the current three implementations be angled for real-world use? I'm using Perl 6 "in the real world". I use it on my spare time. I use it at work. I have production code running in Perl 6. I use Perl 5 even more heavily, but with each month I use Perl 5 a little less and Perl 6 a little more.
      - prog 15y ago
        I am very interested in Perl 6. What would be the recommended implementation to use in production? What are your experiences with using it in production?
        masak 15y ago
        I've been an active Perl 6 user since 2008; during that time, I've often had too high expectations on then-current Rakudo implementations: I've been dabbling in wiki software, etc. It was slow. Sometimes it was unstable. Seen from ten thousand meters, Rakudo 2008-2010 was about features, and Rakudo 2011 has been about performance. As moritz' retrospective states, the fruits of that work are only coming online now. That will enable more people to move into performance-sensitive fields such as web development with Perl 6. That's already starting to happen; see tadzik's Bailador project, for example. Meanwhile, my Perl 6 production code has been getting by with slightly lowered expectations on performance and memory frugality. Crashes/segfaults are no longer a problem (as they were in 2008/2009). Tradeoff example: My blog, which runs on Perl 6, is statically generated rather than dynamically. That's been working out pretty nicely. When I need more speed and/or regex/grammar features, I go with Niecza. Otherwise, I mostly develop on Rakudo. But Niecza's main developer is developing at an almost intimidating pace, and I expect to be using Niecza even more in 2012.
skyheights 15y ago
In a couple of important places Rakudo really stalled in the last half of 2011 due to the unfortunate absence of pmichaud++, its 'pumpking'. Jnthn++ and moritz++ have been able to recover some of the lost territory, but I'd say there's been almost a 50% drop-off in the rate of overall progress. I'm afraid the bus number is too low for expectations to be too high.
restofus 15y ago
The good thing about perl 6 is that lots of ideas from there are getting back into perl 5 and believe it or not Haskell got a bit of a boost because of perl 6 (when the pugs project was in full swing). Even if one of the projects implements the full perl 6 spec that exists today, there are very few perl 5 devs that will switch over ,If I were asked to guess it would be less than 5 %. Perl 6 as a language has to start at the bottom with other programming languages . The name perl6 was used because one of the reason (I am not sure of this) was to retain the name to get perl5 devs to move to the next version of the language and at that time perl5 was a dominant language. Another perspective to look at is that perl6 was announced in early 2000 clojure came around 2007 , scala appeared around 2003. There are other languages like go and dart which have google backing that have come out. So IMHO if you want to learn language design and the other things around it perl6 is a good playground you will learn and easy to get into the community but you will not get a job because of it. If you are a user of programming language who wants to use a language for solving a real world problem today (with libraries and get questions answered) or learn a language which will get you a job you will be better off with perl5 , php, ruby, python, scala, clojure or <insert your lang>

67 points by perlgeek 15y ago

Back in the day, I was an avid mod_perl user and Perl 5 enthusiast. I recall being optimistic that Perl 6 would come out of the oven in 12 to 18 months. Sadly, that was in 2004. My feelings about Perl 6 are like those about a Microsoft OS that is stable, secure and performant: a nice idea that has grown into a myth. Sorry to dump on Perl, I still have many friends who use Perl 5 on daily basis. I think for a lot of folks, the real "Perl 6" is... Ruby.

chromatic 15y ago
I think for a lot of folks, the real "Perl 6" is... Ruby. That doesn't make sense; why would a "real Perl 6" be measurably worse than Perl 5?
- spidaman 15y ago
  A "ruby vs. perl" fight is the furthest from my mind, I'm just speaking anecdotally about what my friends who, like me, were heavy perl users are using in it's place now. If the "measurably" bit refers to how the runtimes benchmark, yes, absolutely, it doesn't make sense.
  - chromatic 15y ago
    If the "measurably" bit refers to how the runtimes benchmark... That and language stability, ecosystem, internal coherence, library availability, release process, compatibility, deployment concerns, tooling....

mjs 15y ago

So there are now three Perl runtimes/compilers/interpreters--Rakudo (+ Parrot VM), Niecza (CLR), Perlito (?)--none of which yet implements the full Perl 6 spec? (Which itself is not yet complete, although "2011 was a rather quiet year in terms of spec changes.")

perlgeek 15y ago
That's correct. They all approach Perl 6 from different angles, and with different goals and techniques. I'm sure you also know that the specifications of Java, C, C++, python, Haskell and other languages are still being developed too.
- squiggly101 15y ago
  So when does work start on the one that will be angled for use in the real world, you know like Java, C, C++, python, Haskell and other languages?
  - masak 15y ago
    I'm trying to think of a way to read your question that doesn't make it dismissive and condescending. Why wouldn't the current three implementations be angled for real-world use? I'm using Perl 6 "in the real world". I use it on my spare time. I use it at work. I have production code running in Perl 6. I use Perl 5 even more heavily, but with each month I use Perl 5 a little less and Perl 6 a little more.
    - prog 15y ago
      I am very interested in Perl 6. What would be the recommended implementation to use in production? What are your experiences with using it in production?
      - masak 15y ago
        I've been an active Perl 6 user since 2008; during that time, I've often had too high expectations on then-current Rakudo implementations: I've been dabbling in wiki software, etc. It was slow. Sometimes it was unstable. Seen from ten thousand meters, Rakudo 2008-2010 was about features, and Rakudo 2011 has been about performance. As moritz' retrospective states, the fruits of that work are only coming online now. That will enable more people to move into performance-sensitive fields such as web development with Perl 6. That's already starting to happen; see tadzik's Bailador project, for example. Meanwhile, my Perl 6 production code has been getting by with slightly lowered expectations on performance and memory frugality. Crashes/segfaults are no longer a problem (as they were in 2008/2009). Tradeoff example: My blog, which runs on Perl 6, is statically generated rather than dynamically. That's been working out pretty nicely. When I need more speed and/or regex/grammar features, I go with Niecza. Otherwise, I mostly develop on Rakudo. But Niecza's main developer is developing at an almost intimidating pace, and I expect to be using Niecza even more in 2012.

skyheights 15y ago

In a couple of important places Rakudo really stalled in the last half of 2011 due to the unfortunate absence of pmichaud++, its 'pumpking'. Jnthn++ and moritz++ have been able to recover some of the lost territory, but I'd say there's been almost a 50% drop-off in the rate of overall progress. I'm afraid the bus number is too low for expectations to be too high.

restofus 15y ago

The good thing about perl 6 is that lots of ideas from there are getting back into perl 5 and believe it or not Haskell got a bit of a boost because of perl 6 (when the pugs project was in full swing). Even if one of the projects implements the full perl 6 spec that exists today, there are very few perl 5 devs that will switch over ,If I were asked to guess it would be less than 5 %. Perl 6 as a language has to start at the bottom with other programming languages . The name perl6 was used because one of the reason (I am not sure of this) was to retain the name to get perl5 devs to move to the next version of the language and at that time perl5 was a dominant language. Another perspective to look at is that perl6 was announced in early 2000 clojure came around 2007 , scala appeared around 2003. There are other languages like go and dart which have google backing that have come out. So IMHO if you want to learn language design and the other things around it perl6 is a good playground you will learn and easy to get into the community but you will not get a job because of it. If you are a user of programming language who wants to use a language for solving a real world problem today (with libraries and get questions answered) or learn a language which will get you a job you will be better off with perl5 , php, ruby, python, scala, clojure or <insert your lang>