Beefy Boxes and Bandwidth Generously Provided by pair Networks
Perl: the Markov chain saw
 
PerlMonks  

Re^3: Perl 6 is going to get a lot faster in 2013

by DrHyde (Prior)
on Feb 04, 2013 at 11:46 UTC ( #1016937=note: print w/ replies, xml ) Need Help??


in reply to Re^2: Perl 6 is going to get a lot faster in 2013
in thread Perl 6 is going to get a lot faster in 2013

Larry can declare that speed is the number one blocker of adoption as much as he likes. The real blocker to adoption is that perl 6 ain't yet complete.


Comment on Re^3: Perl 6 is going to get a lot faster in 2013
Re^4: Perl 6 is going to get a lot faster in 2013
by raiph (Friar) on Feb 04, 2013 at 16:34 UTC
    Your view is prevalent. If you could answer a few questions it might become clearer whether your view is based on something I'm not taking into account, or misunderstanding, or something else:

    • Are you looking forward to using Perl 6 when it is "complete"?
    • When did you last use Perl 6, and which compiler?
    • What do you think is missing?

    Folk adopted Perl 5 en masse even though it was and has always been "incomplete". For example, from a talk submitted to YAPC::NA yesterday:

    Perl 5.2 introduced prototypes to the language, and in the documentation has noted from the very beginning "Alphanumerics have been intentionally left out of prototypes for the express purpose of someday in the future adding named, formal parameters." That was almost 18 years ago, and we still don't have named, formal parameters in core.
      Folk adopted Perl 5 en masse even though it was and has always been "incomplete".

      Hey look, a false equivalence!

        I feel a little sad to read comments like in this thread from chromatic. I've learned many from you, a very famouse guy in perl community, your book, your blog etc. And I do hope to always see your comments for helping perl improvement than some prejudice biasely.

        I can understand your feeling about parrot, but, you have to admit by far, JVM seems be a better platform for perl6 than parrot, and whatever it be slow or not, it should not become a reason which you doubt others efforts, like the comments you leave here, like your blog in several days ago. Before pouring cold water, we, I mean guys who love perl, should at least provide our praise and encouragement.





        I am trying to improve my English skills, if you see a mistake please feel free to reply or /msg me a correction

      Yes, I'm looking forward to trying it out. I've never used it.

      I agree that perl has always been somewhat incomplete, in the sense that it's always possible to make it better. But that's a silly way of interpreting completeness. I've given an example (which took me approximately 30 seconds to find using google) elsewhere in this thread of a core feature that doesn't yet exist. I understand that object introspection and macros are also still not implemented.

      If I didn't already have a set of excellent tools available, then I might consider using perl 6 even in its current state - just like I did back with early perl 5.

        Yes, I'm looking forward to trying it out. I've never used it.

        If you decide to have a go this year, start by visiting the IRC channel #perl6 on freenode between about 10am and 10pm GMT (most Perl 6 hackers are based in Europe) and asking questions. The channel has a very friendly and productive vibe -- and folk like Nicholas Clark and Stevan Little have begun to hang out there in recent weeks which is fun too.

        I've given an example (which took me approximately 30 seconds to find using google) elsewhere in this thread of a core feature that doesn't yet exist.

        Hmm. I don't see it. (I see two posts to this thread using the DrHyde nick, the one that was general and this one which I'm replying to.)

        I understand that object introspection and macros are also still not implemented.

        Object introspection was implemented early in Rakudo's life (2008?). It's a fairly mature feature at this point. Fwiw here's a search of the #perl6 logs for "introspection".

        Carl Mäsak first dreamed of macros in Perl 6 around 2005/2006 iirc, and began fleshing out a design for Rakudo in 2011. He landed an initial implementation of hygienic AST macros (ie Lisp style, not C style) in late 2012.

        If I didn't already have a set of excellent tools available, then I might consider using perl 6 even in its current state - just like I did back with early perl 5.

        Imo it's about 18 months too early for those whose focus is getting stuff done rather than having fun helping productize Perl 6.

        In contrast I think that Perl 6 is attractive to those who want to have fun trying to find things Perl 6 is particularly good at, pushing against its rough edges, and contributing to its improvement.

Re^4: Perl 6 is going to get a lot faster in 2013
by moritz (Cardinal) on Feb 04, 2013 at 17:27 UTC
    The real blocker to adoption is that perl 6 ain't yet complete.

    What makes you think so? Do you have any data that supports your statement?

    I ask because when we released the first version of the Rakudo Star distribution, we got loads of feedback. And most of it was along the lines of "the features are very nice, but it's too slow". Which indicates the the majority of potential users doesn't think that he lack of completeness is the main problem.

      I agree with this. In my attempts to get on board it was speed and the dearth of documentation (and corresponding weakness in diagnostic feedback for misuse) that kept putting me back in the drink. I never considered "code completeness" for a moment; only "code what can I do with thisness."

        Documentation is still a weak point though a number of things were significantly improved in 2012. Most notably the new doc.perl6.org project has established the right architecture for a Perl 6 equivalent of Perl 5's perldoc.

        At least four big parts of the diagnostic feedback problem have been addressed in the last couple years:

        • The level of completion and stability of Rakudo. If there's an error, it's a lot more likely that the error is in the user's code not the compiler.
        • Error messages. Rakudo has finally caught up with Larry Wall's canonical STD parser. There's color coding, accurate pinpointing of where an error is, uncannily clever suggestions about what you meant to do, and so on. (Niecza was based on STD from the start.)
        • #perl6. The best way to learn Perl 6 is to get on the #perl6 freenode IRC channel and combine use of the on channel evalbots with questions. In watching #perl6 for the last year and a half I've seen a lot of newbies come ask questions and they invariably encounter patience and friendliness. Another great source for figuring Perl 6 out is to read the #perl6 log archive.
        • Rakudo source code debugger. A nice UI, debugging of regexes/grammars, debugging of exceptions, the works. And eminently hackable, too!
        Slowness has been the major factor stopping me for getting into Perl 6 also.

Log In?
Username:
Password:

What's my password?
Create A New User
Node Status?
node history
Node Type: note [id://1016937]
help
Chatterbox?
and the web crawler heard nothing...

How do I use this? | Other CB clients
Other Users?
Others examining the Monastery: (12)
As of 2014-08-01 16:27 GMT
Sections?
Information?
Find Nodes?
Leftovers?
    Voting Booth?

    Who would be the most fun to work for?















    Results (28 votes), past polls