Beefy Boxes and Bandwidth Generously Provided by pair Networks
Think about Loose Coupling

Re: Perl 6 was released years ago

by davido (Archbishop)
on Oct 31, 2004 at 16:36 UTC ( #404146=note: print w/replies, xml ) Need Help??

in reply to Perl 6 was released years ago

...Unlike the transition from Perl 4 to Perl 5, Perl 6 wasn't intended to be a replacement for its predecessor, but rather, a new and modernized tool. Perl 5 development for the 5.6.x line ended in 2006, but 5.9.x and eventually 5.10.x continues today, though admittedly popularity has waned somewhat in the face of what has become a superior tool. New Perl learners seem to be drawn mostly to Perl 6, and for good reason. It has progressed significantly since its early quirky days. Perl 5 posts on popular websites such as PerlMonks have become far less frequent in favor of Perl 6 topics, and the main focus of the development community has shifted toward Perl 6. Amazingly, Parrot (the back end of Perl 6) has now been adopted by Python and Ruby as their foundation of choice, and probably not so coincidentally, Microsoft has begun development on a Parrot-like backside for its much anticipated .WET development tool.

In 2011, Larry pulled himself and a small team away from the p6p group to begin concentrating their bleading-edge efforts on the much anticipated Perl 7. At this point details on Perl 7 are sketchy, but those in the know seem to all light up with a knowing and smug grin whenever the project is mentioned. The development community is abuzz with the few tidbits that have been overtly discussed, and seems to look forward to the omnipresent mantra of TIMTOWTDI being combined with the next generation linguistic parser that Larry discussed in his state of the onion talk in 2009.


Replies are listed 'Best First'.
Re^2: Perl 6 was released years ago
by Mr. Muskrat (Canon) on Oct 31, 2004 at 20:00 UTC

    I think you are forgetting one important thing. Perl6 is being designed to be a twenty (or even fifty) year language. There should not be a need (nor want) for a Perl 7.

    Update: Although, going back to Apocolypse 1, I see that there will be a Perl 7 and that Perl6 is the prototype. So instead my last sentence should say "There should not be a need (nor want) for a Perl 7 for a very long time".

      Perl5 was also supposed to be the last rewrite of Perl. It was designed to be extendible, without a frequent need to update the core. Development of perl6 started less than 6 years after the release of 5.000. So, I guess the main reason to not expect development of perl7 to have started in 2012 is that it's unlikely that perl6 will be released before 2006. ;-)

        Was Perl 5 really supposed to be the last rewrite?

        Was it designed, at all? ;)

        Juerd # { site => '', plp_site => '', do_not_use => 'spamtrap' }

Re^2: Perl 6 was released years ago
by bunnyman (Hermit) on Nov 01, 2004 at 00:58 UTC

    To suppose that Perl 5 will be actively developed 8 years from now is to greatly underestimate the importance of an innocuous little thing called the version number.

    I cannot imagine anybody seriously using an 8 year old obsolete version of anything and not being in the tiniest of minorities. Who uses Windows 95 anymore? Linux 2.0 or older? Java 1.0? PHP 3? Perl 4? Yes, old versions do get used, but only in special circumstances.

    Sure, Perl 4 was good enough for what it was, but then along came Perl 5 which was also good enough for that, and more. And today Perl 5 is good enough for a great number of things, and Perl 6 will be good enough for all them, and even more again.

    If Perl 6 were renamed to something other than Perl, then I could see a future for Perl 5, but if there's a bigger number available, it will get all the attention.

      You'd be surprised...

      I had a query the other day from someone using perl 5.000 (yes, that's right) and sybperl 2a7 (that's an alpha release), all 10 year old code, and running on Solaris 8....

      I very strongly advised them to upgrade (I can't count the number of bugs fixed in sybperl, let alone perl in the last 10 years!)...


        Maybe I'm sticking my head in the sand, but I want to call that a "special situation."


      I cannot imagine anybody seriously using an 8 year old obsolete version of anything and not being in the tiniest of minorities.

      FORTRAN 77. That's all I have to say.

      Well, I'll say a tiny bit more: YES, people do still use FORTRAN. YES, there are newer versions than the (19)77 version. HOWEVER, FORTRAN 77 is still in a great deal of use. Probably the majority of FORTRAN code in use today is FORTRAN 77. Certainly not a small minority.

      Oh, and just because I like to point it out: FORTRAN is the last major language, before Python, to incorporate significant leading whitespace as a "feature" of the language. :-P

      ------------ :Wq Not an editor command: Wq

        That's not a good example, because Fortran 90 compilers will (mostly) accept code written for 77. There's nothing holding you back from upgrading and still using the old stuff. Which sounds like what Perl 6 will be like, too.

        And because I like to point this out: In Fortran, spaces between columns 7 and 72 do not matter unless they are between quotes.

        For example:

        if( foo .eq. bar ) then

        is equivalent to

        i    f(f    oo.e     q.b    ar)t     hen

        And people have the nerve to call Perl unreadable!

        Oh, and just because I like to point it out: FORTRAN is the last major language, before Python, to incorporate significant leading whitespace as a "feature" of the language. Another reason perl6 is great. It will have its own set of significant whitespace rules. It will now be: FORTRAN, Python, Perl6.
        Perl does the same thing slightly differently: trailing, non-line ending white space on the end of HEREDOC lines are significant.
      I cannot imagine anybody seriously using an 8 year old obsolete version of anything and not being in the tiniest of minorities.

      There are lots of airplanes flying around that are older than 8 years. Trains are build to last 40 or more years (and they do).

      What I can't believe is that people upgrade for the sake of upgrading. There's an old saying "if it ain't broken, don't fix it". New versions of software *do* break things, whether intentionally or not. And I have to say, many open source authors don't consider backwards compatibility as important as they should (if they'd paid more attention to it, it would help in getting corperations to use more open source software). Perl tries it best to be backwards compatible, but it isn't perfect.

      Not everyone seems to realize the costs upgrading takes. Upgrading a single developer box isn't hard. But for numerous organizations, installing new, or upgrading existing software is a big deal. Banks, hospitals are air flight controllers don't just upgrade. They have long and rigorous testing procedures, including regression tests. Anything that breaks is a show-stopper until resolved. People working with the new or upgraded software may need re-training. This all costs money.

      I think the computing world would benefit if eight year old software wasn't "obsolete" and would just run without problems.

      Yah, who uses K&R C anymore?

Log In?

What's my password?
Create A New User
Node Status?
node history
Node Type: note [id://404146]
and all is quiet...

How do I use this? | Other CB clients
Other Users?
Others contemplating the Monastery: (6)
As of 2018-01-22 09:19 GMT
Find Nodes?
    Voting Booth?
    How did you see in the new year?

    Results (233 votes). Check out past polls.