Lets just drop the 5 from Perl 5 | Hacker News Reader

Lets just drop the 5 from Perl 5

blog.kraih.com

64 pointskraih30 comments

Open articleView on HN

I almost missed the links in that blog post to the other discussions. They are

http://blogs.perl.org/users/alberto_simoes/2011/06/perl-perl...

http://www.modernperlbooks.com/mt/2011/06/perl-perl-5-perl-6...

http://blogs.perl.org/users/mithaldu/2011/06/why-are-people-...

http://jjnapiorkowski.typepad.com/modern-perl/2011/06/perl-5...

I think John's post on the modern perl blog is the most rational dissection of this "problem" that is plaguing the perl community.

Few people outside the perl community see "Perl 5" and "Perl 6". For most people referring to the language they simply call it "Perl."

For marketing to developers specifically, I think the biggest hurdle isn't the name but the stigma attached to the language. Most developers I talk to outside of the perl community still see perl as a "write only language" that consists purely of "line noise." They don't specifically say whether they are talking about Perl 5 or Perl 6. To them there is no distinction.

I agree.

Nobody want's to invest time into learning something called "Perl 5" when there's something already out there called "Perl 6". Yes, some research into it will reveal that they are now on two separate dev paths, but there is still some mental stigma associated with investing time in learning an "old version" of something. And learning the "newer version" quickly reveals that it is not widely used/supported yet. They are basically two separate languages and should be treated as such.

These days the arena Perl competes in is extremely competitive, and these seemingly small issues will continue to be disastrous for Perl.

I can attest to this. This was my first insight into the Perl 5/6 issue. Until now, I always assumed it was along the lines of Python 2.x and Python 3.

I've been espousing this for years! If java can do it, so can perl. Go go perl 14!

And how about Perl 7? Perl 14 would feel like the Java version bump, and people would say "well Perl 5 just dropped the 5" which means nothing in fact. Perl 7 could start enforcing all "use feature" and "strict modernity" out-of-the-box and package things like Moose and sub signatures. And as a side dish, we could do a cleanup of the interpreter code for p5p mental sanity, in a "less is more" fashion.

Actually the odd numbers could become the Perl5 series and its gradual evolution. And the evens, the experimental Perl6 series of taking computer languages where no man has gone before. But that's just pointless... it would be another 20 christmas before Perl 8 would start to materialize.

Besides, Perl 7 is supposed to be "God's rewrite of Perl". So, here's where God comes in and Perl 7 saves us all.

I always liked the idea of renaming Perl 6 to Rakudo, being the major/official implementation. It's also what most other programming languages do. A major, leading implementation, as well as others following. This would also make it easier for newcomers.

Rakudo is by no means more official than any other implementation (for example currently niecza is quite strong, see https://github.com/sorear/niecza/ for the code).

Renaming Perl 6 to Rakudo makes about as much sense as renaming C to GCC to MSVC.

Rakudo, while a nice name, is also too easy to mistype - even core committers (me included) have frequently misspelled its name in commit messages etc.

> Rakudo, while a nice name, is also too easy to mistype - even core committers (me included) have frequently misspelled its name in commit messages etc.

I agree with the principle of this, but you do have to compare this to Pearl.

This seems like a change similar to the perl 5.000000000008 to 5.8 change (yes, I know it had fewer 0's in it but there were a lot).

s/Perl 6/Perl Forever/

What's Larry Wall up to these days, anyway?

Well, he has a day job, spends significant amounts of his free time on answering questions on the Perl 6 specification and changing it when necessary, and attends many perl conferences (YAPC::NA, YAPC::EU, YAPC::Asia, OSCON) (where Perl 5 and Perl 6 peacefully coexist and share both speakers and audience, fwiw).

Did I mention that he also has a family, and a local church in which he is very active?

Update: I forgot, he also helps Tom Christansen to bring out the next edition of "Programming Perl".

Why? Isn't Perl6 the "new hot" thing? You could rename Perl5 to Pearl Classic. (see Python...)

Sadly, the situation is different from Python's.

Python 3 is well-defined, with a stable and usable implementation. You can port your code to it today, and even big projects like Django are getting ready to shift over to it.

Perl 6's definition continues to change, with no clear spec, and the leading implementation (Rakudo) doesn't have half of the features built. Even if there were a feature-complete implementation of it I wouldn't dream of starting to port my code over to Perl 6 yet.

Perl 6 is really a different language, and should be named as such. With that out of the way, Perl 5 can then get rid of its "5" and move on to release 16 in a year.

>>Sadly, the situation is different from Python's.

>>Python 3 is well-defined, with a stable and usable implementation.

It is hard to compare.

The main similarity between Perl 6 and Python 3 is probably the age. The main difference is that Perl 6 is really, really ambitious.

Because Perl 6 is an incomplete language still in development that, in all honestly, has no chance of ever replacing Perl 5. It's been in development since 2000 and is a spec with no "official" implementation. Perl 6 is so different that it might as well be considered a different language. Hence, "real Perl" is Perl 5 and Perl 6 should be called something else.

If they'd called it "Perl NG", then we wouldn't be talking about Perl 5.16. We'd be talking about Perl 21 vs Perl NG.

That's exactly the attitude those proposing name changes want to dispel. "People ask for name changes so they can show the rest of the world just how alive Perl 5 really is."

Good luck with that. In 1999 every programmer I knew used Perl. Now I don't know anybody that hasn't moved on to Ruby or Python.

Oddly enough, despite Ruby's origins, Perl is still very widely used in Japan.

Strange. In what kinds of applications?

web development mostly. Big Japanese companies use perl for their backend and you can find many blog entries about perl in Japanese.

And many seem to have rolled their own frameworks long ago - and are reluctant to move on when things are working.

On computers, mostly.

That's disingenuous and doesn't clear up anything. Better: Perl 5 Version 16.

I'm having trouble seeing how any change to something as completely arbitrary as version number schemes could be described as "disingenuous".

How would the Perl community be attempting to "deceive people via a false impression of honesty" by dropping the long-since-atrophied major version number?

What possible point of honesty is served by gluing that 5 to the language in perpetuity, just to continue waiting for Godot to show up?

What people don't realize is that Perl means something. (And it doesn't mean specifically Perl 5.) I'd wager nobody advocating silly name changes (for either 5 or 6) had been writing Perl 10 years ago. In short: it's a marketing lie, and a poor one at that.

For the sake of (spiteful) example I submit Perl 4 should be now known as Perl, and Perl 5 should deal with succession syndrome. Oh but Perl 6 is disruptive and confusing because it's so different you might complain, adding that 5 wasn't so different from 4. You wouldn't have dealt with the frustratingly subtle differences between Perl 4 and Perl 5 then. Or reams of Perl 5 documentation written as Perl 4 code with errata.

Nah, you could just do a release number and it could just be Perl. Perl 2011.6.27. No biggie. And no, I don't advocate it either.

The problem with perls (5, 6, N) is that most of the competent, intelligent committers were either (a) subverted by O'Reilly contracts into being ardent defenders of the status quo, (b) fed up with waiting for years in order to get unintelligible exegeses from a chief committer who apparently lost all interest and was unable or unwilling to delegate, (c) driven into bafflement by the rivening of p5p into scads of unmanaged sublists and lost interest, (d) discovered thriving and engaged communities in languages not riddled top to bottom by political, organizational and leadership problems, or (e) fell into the black leather-winged embrace of crufty corporate nonsense that is python.

The numbering system merely hastened the demise by further making sure that any sensible interested committer would first check to see what was going on with the latest branch, discover it was the same godawful non-navigable incompetent mess of splintered crazy version wreckage floating abandoned on a lake of ennui that you can trivially see before you today, and run like fuck in the opposite direction.

The principal problem with perl today is that it still exists, and that last tenuous skeleton crew of 2-3 smart people who still struggle in vain to breathe life into its long dessicated corpse is still capering about claiming that there's still! some! path! to! relevance!, rather than doing the noble thing and finding employ as dishwashers or carpet cleaning fluid salesmen.

Reply on news.ycombinator.com