2015.06 In the wake of FOSDEM

Although this Perl 6 Weekly post only covers about 5 days worth of events, it was pretty busy. And of course I missed a few things in the last report, so I’ll mention them this week.

Interview with TimToady

Miyagawa++ held an interview with TimToady after his presentation at FOSDEM. The entire podcast can be downloaded, the transcript is partly paywalled.

Don’t you just love Rats?

CurtisOvidPoe expressed his love for Perl 6 and COBOL at just about the same time last week’s Perl 6 Weekly was published.

Sprouting Up in Brussels

Although not strictly about Perl 6, mdk++ published a very nice impression about his first visit to FOSDEM. And there are also pictures.

All Perl 6 Modules in a Box

moritz++ has created an experimental perl6-all-modules repository, which he mentions in a recent blog post. He used it to find out how many modules in the ecosystem were using Rakudo/NQP specific nqp::xxx() calls. It turned out to be about 8% of the modules: these will need a use nqp; command in the future.

Stats / Directions for doc.perl6.org

moritz++ has been very busy indeed. Apart from his tireless work in the background for doc.perl6.org (among many other things), he has also blogged about it. Of course, help in documenting Perl 6 stays welcome as ever! In this, I should also mention [ptc]++, for being an excellent editor (in the newspaper sense).

Rakudo Star 2015.01

Did I mention that moritz++ has been very busy? He released yet another Rakudo Star! And mj41++ made a Docker image!

Rakudo todo list

colomon++ made his own todo list for Perl 6.0. I can’t help but agree that currently seen precompilation issues, as well as instability when stressing parallel execution, are high on my list as well.

Parameter Coercers work!

jnthn++ has been very busy behind the scenes in Moar and nqp, in the “6pe” (for parametric extensions) branch. As jnthn explains:

6pe in Moar/NQP is a very general mechanism for type parameterization and (once I get to it) interning them across compilation units. The primary motivation was typed data structures, but it turns out the exact same mechanism is also a great way to implement coercion types and cache mixins. The reason I did coercion types and the mixin cache first is ‘cus (a) they’re useful, and (b) they’re a gentler first test for the 6pe implementation.

This branch was merged by moritz++, which made things like
     sub a(Str(Any) $foo) { }
(take Any, but coerce to Str) work, and make mixins such as:
     C.new does R
about 10x as fast, and make the core setting about 700KB smaller through re-use. More goodies will come from this work soon, but I’m going to defer telling you about them until next week.

Movement on the JavaScript backend for nqp front

hoelzro++ took all of the work that pmurias has done on the JS backend into a branch in the nqp repository. A big step forward into getting a truly functional JS backend for nqp, and in turn for Rakudo Perl 6!

Development Allsorts

psch++ fixed a problem with [R~], found by avuserow++. FROGGS++ fixed some issues with nativecast on the JVM. FROGGS++ also fixed some warnings regarding anonymous variables. Mouq++ improved the = meta-op, for cases such as $a >== $b, and enabled lexical term definitions like
     my \term:<∞> = Inf
lizmat implemented the nqp::readlink function on MoarVM/NQP with a lot of help from jnthn++ and FROGGS++. For now, it is only used in the newio branch.

Signing off for this week

As always, please let me know of any errors, omissions, additions!


Got something to note?

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s