Re^4: A "Perl-7" that I could actually USE right now

[Jenda]

I hadn't assumed a thing. I was pointed to those three repositories. A little earlier I was pointed to the repositories linked from the Perl6 module list. Both were supposed to prove that Perl6 is being actively developed (I never doubted that) and that there is a sizable community around it (which they did not). In summary, as you write, there had been 20-30 regulars, 50-70 occasionally returning irregulars (here I think you are exaggerating a little, yes occasionally they return) and (and here's the main disconnect) "a trickle of new folk". No. There's a trickle of people that come, stay for a week and go.

I don't doubt there are a few people that work their asses off. I just don't see people flocking to tear the results out of their busy hands. I don't see people waiting eagerly for Perl6 to get into a (for them) useful state.

Within the echo chamber of #perl6 it certainly looks like there's a lot going on, like there's a lot of life in the project. Outside, strange noises are heard from time to time. Perl Incorporated built a shiny new hall, right next to the main gate, built a highway to ship the planned project, made press releases and superbowl ads ... and failed to deliver anything. The hall is no longer shiny, the highway was partly reused and partly fell into pieces, but in the furthest corner of the huge building, there are three offices and a kitchen still inhabited and from time to time a bunch of school kids comes for an excursion.

Yep, the groundwork may end up being useful (though the choice of JVM will probably never cease to surprise me). Hopefully.

[raiph]

"a trickle of new folk". No. There's a trickle of people that come, stay for a week and go.

The following folk are examples of the trickle that you don't see that I do. These are from yesterday's #perl6 log. In order of appearance:

• timotimo first spoke on #perl6 Jan 2012. He began making commits this year. Some of these are visible via a search of #perl6 for dalek tracking of Timo Paulssen commits to selected repos.

• labster first spoke on #perl6 December 2012. He's been landing commits since the same month. Some of these are visible via a search of #perl6 for dalek tracking of Brent Labs commits to selected repos.

• FROGGS first spoke on #perl6 August 2012. He began making commits in September. Some of these are visible via a search of #perl6 dalek for tracking of Tobias Leich commits to selected repos.

• lizmat first spoke on #perl6 August 2012. She's been landing commits since the same month. Some of these are visible via a search of #perl6 for dalek tracking of Elizabeth Mattijsen commits to selected repos.

• not_gerd first spoke on #perl6 November 2011. He's been landing commits since March 2012. Some of these are visible via a search of #perl6 for dalek tracking of Gerhard R commits to selected repos.

Searching through other days will reveal more such folk. Like I said, a trickle. But not zero.

I just don't see people flocking to tear the results out of their busy hands. I don't see people waiting eagerly for Perl6 to get into a (for them) useful state.

Coders are, quite rightly, not interested in using P6 until it's ready for them, and it isn't ready for most coders, even adventurous, early adopter types. Similarly, the P6 project is, quite rightly, not making any attempt to attract users because it isn't ready for most coders.

Aiui there is still $100K waiting in a bank account (part of the Hague Grant) to be spent when the time is right for raising money to promote P6. My guess is that that process will start once Larry declares a 6.0.0 spec.

It's very hard to predict when that will be, but I've been consistent about my prediction (late 2014) for the last 4 years and it still looks about right to me.

the choice of JVM will probably never cease to surprise me

Are you aware the design choice was made in the late 90s, before P6 was conceived? Do you know what the rationale for implementing it now is?

[Jenda]

It might have made sense in late 90s when Java was hip and looked like it's gonna evolve. IMnsHO at this time MONO would be a much better target. Even if it takes Google a few more months to make the switch.

Anyway some pointers to the current rationale would be appreciated, thanks!

[raiph]

Don't forget there's something like a billion Android (JVM) devices and (rapidly) counting.

For a piece of the rationale for the JVM port, see the answer to the question "Was the move of porting Rakudo to the JVM planned, or was it driven by Parrot’s deficiencies?" in All about Perl 6 – interview of Jonathan Worthington (Part 1 of 3). There's more to it than that but it covers some key pieces.

jnthn has spoken about his position on a CLR backend but I didn't find it when I quickly looked. In a nutshell, it's "Sure, that would be great, and we know it'll work well, but I'm not doing it, someone else would have to run with it, and they may not even get much of my time to help them figure stuff out". There's a similar answer for the JS backend that's being progressed as a GSOC project by pmurias, fperrad's LuaJit backend, and so on.

Volunteers will do what they want, but I'd say development of the JVM and MoarVM backends now have, respectively, tactical and strategic primacy over development of other backends for the next 12 months. And they have the momentum. jnthn and sorear have been rocketing through the JVM port. (Here's jnthn's recent update on the JVM port.) And MoarVM's attracting uber hackers too such as Chip Salzenberg.