Camelia, the Perl 6 bug

IRC log for #p6p5, 2012-09-19

| Channels | #p6p5 index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
01:09 rurban_mobile joined #p6p5
01:10 rurban_mobile ingy is now having his little presentation of his p6p5 ideas in austin (live on google hangout, later on youtube)
01:11 sorear austin?  in sep2012?
01:11 sorear some unannounced perl event?
01:12 rurban_mobile a first austin compiler meeting
01:12 rurban_mobile today, and next week also.
01:12 sorear what can I do to be informed of things like this?
01:13 rurban_mobile The YAPC::US Austin 2013 goal will be an optimizing compiler for perl5.
01:13 rurban_mobile And the idea with VM integration came up.
01:14 rurban_mobile We are only two people here in Austin, and discussing some ideas.
01:14 rurban_mobile ingy and me.
01:14 rurban_mobile lets see what ingy will come up with. he'll present it later publicly
01:15 sorear right, I was planning to come to Austin in *2013*
01:15 rurban_mobile super
01:16 sorear I am extremely annoyed that I did not find out sooner about this current event
01:16 rurban_mobile its not really an event. its a meeting
01:16 rurban_mobile with the austin group.
01:16 rurban_mobile but we got several ideas we want to present later.
01:17 sorear oh, ok
01:20 rurban_mobile So he came with perl11 (5+6=11)
01:21 rurban_mobile both calling each other
01:21 rurban_mobile with no shared state
01:22 sorear wait, rurban_mobile, you live in Austin now?
01:23 rurban_mobile austin is just 3 hrs from houston away
01:23 rurban_mobile i live in houston, tx
01:24 stevan__ austin > houston
01:24 stevan__ no offense
01:24 sorear aha
01:24 sorear I thought you were still in .de, oops
01:25 rurban_mobile i am austrian who married a german
01:25 rurban_mobile and my mother is german
01:25 stevan__ and moved to texas, so he can be a cowboy!
01:26 rurban_mobile Ingy stared by saying: dont be afraid to start slow. real slow
01:26 stevan__ rurban_mobile: is this all going to be related to NQP at all?
01:26 rurban_mobile not. yet
01:26 stevan__ k
01:26 rurban_mobile llvm as backend for nqp yes.
01:26 stevan__ we need to be careful of competing efforts
01:26 rurban_mobile but also as backend for perl5
01:26 stevan__ right
01:27 rurban_mobile but first: Pegex (a parser)
01:27 rurban_mobile pegex-pgx self-hosting parser
01:28 stevan__ yes, he presented that in NYC, very interesting
01:28 rurban_mobile parses perl, python, ruby, node
01:28 rurban_mobile coffescript
01:28 rurban_mobile and so on...
01:28 stevan__ coffescript is an abomination, but yeah
01:29 stevan__ ok, time to watch bad psuedo horror tv with $kid[0]
01:29 * stevan__ will eventually get her to watch Evil Dead, but we have to start with somewhere
01:29 * stevan__ waves
01:30 rurban_mobile evil dead for a 11 year old?
01:33 castaway joined #p6p5
05:11 cognominal joined #p6p5
05:52 cognominal joined #p6p5
06:22 rurban_mobile joined #p6p5
06:27 cognominal joined #p6p5
07:08 cognominal joined #p6p5
07:20 [Sno] joined #p6p5
07:45 andrefs joined #p6p5
07:46 slavorg joined #p6p5
08:27 * leont would love to see nqp on LLVM
08:37 moritz LLVM is a bit low level for that
08:37 moritz you need at least a GC and some runtime support like dynamic arrays and hashes
08:38 castaway nqp?
08:39 moritz NQP = Not Quite Perl, a subset of Perl 6 that can easily and efficiently be compiled
08:39 moritz large parts of Rakudo are written in NQP
08:41 castaway ah, neat
08:42 castaway is there a spec for that?
08:42 moritz no
08:43 moritz NQP also contains the grammar engine and the compiler toolkit that Rakudo uses
08:53 sciurius joined #p6p5
08:56 andrefs joined #p6p5
10:35 andrefs joined #p6p5
11:18 tobyink joined #p6p5
11:52 sawyer joined #p6p5
12:02 cognominal joined #p6p5
12:14 fglock joined #p6p5
12:20 cognominal joined #p6p5
12:24 fglock we were discussing running perl5 in alternate vms a few hours ago (ingy, rurban and others)
12:25 fglock I was looking at how we could specify vm-specific options
12:29 fglock much like
12:29 fglock use mro 'c3';
12:31 fglock and
12:31 fglock my $can_use_threads = eval 'use threads; 1';
12:32 fglock so we can have things like:
12:32 fglock no refcount;
12:38 sawyer what about: use cflag '-refcount';
12:39 sawyer use cflags '-refcount', 'optimize'; etc.
12:50 tobyink use cflags qw(optimize); no cflags qw(refcount);
13:07 fglock I can't find a reference for cflags - what is it?
13:09 fglock sawyer: cflags would be a pragma, like 'feature', right?
13:09 cognominal joined #p6p5
13:11 sawyer fglock, yeah
13:11 sawyer and all interpreters would implement that pragma on the compiler level
13:11 sawyer so each compiler could provide features it supports (or compiler flags)
13:12 sawyer you could have an interface in interpreter level, and different compiler flags available (and their corresponding implementations)
13:12 sawyer then again, i've never learned compilers and you wrote one! so, you might have a much better idea than this :)
13:40 timbunce joined #p6p5
14:43 tobyink joined #p6p5
16:03 danaj joined #p6p5
16:15 sawyer_ joined #p6p5
16:28 tobyink joined #p6p5
16:55 sivoais joined #p6p5
17:08 sivoais joined #p6p5
17:20 andrefs joined #p6p5
17:34 rurban_mobile joined #p6p5
17:35 rurban_mobile wouldn't that be a vm feature? use vm -norefcnt; or use vm -gc;
17:36 sawyer_ joined #p6p5
18:41 timbunce_ joined #p6p5
18:50 tobyink joined #p6p5
18:56 andrefs joined #p6p5
19:01 sawyer_ joined #p6p5
19:42 rurban_mobile joined #p6p5
19:59 timbunce_ joined #p6p5
21:27 cognominal joined #p6p5
23:10 timbunce_ joined #p6p5

| Channels | #p6p5 index | Today | | Search | Google Search | Plain-Text | summary