»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'perl6: say 3;' or rakudo:, niecza:, std:, or /msg p6eval perl6: ... | irclog: irc.perl6.org/ | UTF-8 is our friend! | Rakudo Star Released!
Set by diakopter on 6 September 2010.
dalek p-rx: 7188f73 | pmichaud++ | src/NQP/ (2 files):
Initial version of trait_mod:is and "is pirflags" trait for subs.
00:03
p-rx: f3dabb7 | pmichaud++ | src/NQP/ (2 files):
Add traits to method definitions.
p-rx: 95acd86 | pmichaud++ | t/nqp/52-vtable.t:
Add tests for "is pirflags<:vtable(...)>".
p-rx: e528a3d | pmichaud++ | t/nqp/52-vtable.t:
Oops. Failing to plan() is like planning to fail().
sorear hello pmichaud
dalek ecza: d8c7d65 | sorear++ | lib/Kernel.cs:
Add a hook to profile name lookups
00:12
ecza: e893836 | sorear++ | / (5 files):
Tidy up MOP a bit
00:20 kcwu joined 00:24 Chillance left 00:26 kcwu left, kcwu joined
dalek ecza: 63628e6 | sorear++ | lib/ (4 files):
Generate slot mappings at compose time
00:31
00:33 jhuni joined
sorear interesting; going from the 60k to 600k test file made memory use out of control 00:33
00:38 redicaps joined 00:39 whiteknight left 00:52 [Coke] left
diakopter std: my $a = return 4 00:54
p6eval std 32123: OUTPUT«ok 00:01 118m␤»
01:01 mfollett joined 01:02 QinGW joined 01:07 tylercurtis left 01:18 TomDLux joined
mfollett the latest revision of master doesn't compile for me. 01:20
01:22 perigrin left, perigrin joined
diakopter does anyone have a graph [any kind/arrangement/projection] of the Perl 6 type system 01:23
01:24 [Coke] joined 01:25 TomDLux left 01:33 BinGOs left, BinGOs joined
dalek p-rx: 129c25b | pmichaud++ | examples/json.nqp:
Add json.nqp, an example of a JSON compiler written entirely in NQP.
01:36
01:36 yrgd left 01:39 perigrin left, TomDLux joined, perigrin joined 01:41 TomDLux left 01:44 TomDLux joined
TomDLux rakudo: my $set is KeySet; $set{$_}++ for 1..5; $set{$_}-- for 2..4; say $set.keys 01:45
p6eval rakudo 859f2d: OUTPUT«No applicable candidates found to dispatch to for 'trait_mod:<is>'. Available candidates are:␤:(Mu $child, Role $r)␤:(Routine $r, Any :default($default)!)␤:(Code $block, Any $arg?, Any :export($export)!)␤:(Mu $child, Mu $parent)␤:(Mu $type where ({ ... }), Any :rw($rw)!)␤:(Mu
..$type…
TomDLux rakudo: my %set is KeySet; %set{$_}++ for 1..5; %set{$_}-- for 2..4; say %set.keys 01:46
p6eval rakudo 859f2d: OUTPUT«No applicable candidates found to dispatch to for 'trait_mod:<is>'. Available candidates are:␤:(Mu $child, Role $r)␤:(Routine $r, Any :default($default)!)␤:(Code $block, Any $arg?, Any :export($export)!)␤:(Mu $child, Mu $parent)␤:(Mu $type where ({ ... }), Any :rw($rw)!)␤:(Mu
..$type…
01:50 BinGOs left, BinGOs joined
TomDLux I thought "is KeySet" was working, but I'm finding keys remain in existence even when they are set to false .. and p6eval totally barfs on 'is KeySet' 01:55
01:55 perigrin left, perigrin joined 01:57 TomDLux left 01:59 xinming left, am0c^ left, xinming joined 02:01 plainhao left 02:03 BinGOs left 02:04 BinGOs joined 02:07 am0c^ joined 02:24 mfollett left 02:38 [Coke] left 02:39 meppl joined 02:44 Schwern joined 02:52 justatheory left 03:07 \xF0 joined 03:16 mfollett joined 03:25 lichtkind joined
lichtkind had pugs a MOP ? 03:27
mfollett Is there a known good revision of proto that someone would recommend? 03:30
03:35 meppel joined 03:36 redicaps left
dukeleto lichtkind: what do you mean? 03:37
lichtkind: there is something called SMOP, www.perlfoundation.org/perl6/index.cgi?smop
lichtkind i know
i mean if pugs had implemented a meta object protocoll 03:38
to do its OOP stuff
SMOP has definetivly a MOP
but that was developed largly independet of CLASS::MOP 03:39
03:39 meppl left
dukeleto lichtkind: yep 03:39
03:41 meppel is now known as meppl 03:43 tylercurtis joined
dukeleto perl6: say @_[0].WHAT 03:48
p6eval rakudo 859f2d: OUTPUT«Null PMC access in can()␤ in main program body at line 1␤»
..pugs: OUTPUT«Scalar␤»
dukeleto does that rakudobug exist already ?
Util_ rosettacode.org/wiki/Reduced_row_echelon_form - Outstanding bugs fixed to existing Perl 6 solution, and 2 new math-based solutions added; procedural and OO. 04:11
04:12 QinGW1 joined 04:13 QinGW left 04:21 hatseflats left 04:22 PerlJam left, dukeleto left, Juerd left 04:32 lichtkind left 04:33 PerlJam joined 04:34 gfx joined, hatseflats joined, dukeleto joined
mfollett Am I doing something seriously wrong or does the HEAD of proto and the version of proto shipped with Rakudo * 2010-08 both not successfully install modules? I suspect I'm doing something wrong. 04:42
04:59 hlatapie joined 05:00 hlatapie left 05:02 baest joined
sorear good * #perl6 05:06
mfollett: I suspect the module you're trying to install is an alpha-only module
sorear has figured out niecza's grammar memory problem o/ 05:07
05:11 gfx left 05:13 sftp left 05:18 masak joined
masak oh hai, #perl6! 05:18
sorear hello masak
sorear entertains thoughts of butchering druids 05:20
05:20 LionMadeOfLions joined
masak butchering druids? 05:21
sorear yeah, i think a standalone Druid.exe would make a nice niecza showpiece 05:22
masak indeed.
05:22 justatheory joined
masak at the point you have all the functionality required to run Druid, I will seriously consider switching to niecza for many of my everyday tasks. 05:23
05:23 Juerd joined
sorear unmodified? :) 05:24
mfollett sorear: It appears to be a combination of me doing it wrong and what might be a bug, or me doing it even more wrong. It appears I need a projects.list and a projects.cache in my working directory.
As for what I was doing wrong, it turns out I had misconfigured my proto.conf. 05:25
masak sorear: 'unmodified' is unfortunately too idealistic right now, even on the Druid end.
sorear: the only working branch only works on alpha.
05:38 justatheory left 05:43 mfollett left, kaare joined 05:44 kaare is now known as Guest24003 05:48 Zapelius joined 05:52 mfollett joined
pmichaud what's missing for Druid to run in rakudo star, ooc? 05:53
05:54 Guest24003 is now known as kaare_
tylercurtis has managed to forget his feather password within an hour of setting it. 05:58
06:03 lue left, lue joined
sorear tylercurtis: msg me a new one 06:05
06:08 uniejo joined
masak pmichaud: Druid used a fair number of alpha-specific regex features. in particular, I had separated out a number of bare rules into a base class. those call each other and return a neat little $/ tree of results. that doesn't work at all on Rakudo master. 06:09
specifically, there's no way for bare rules to call each other. and it's not possible to mimic the old structure, because nested keys aren't implemented. 06:11
sorear "bare rules"?
masak rules outside of grammars.
sorear if that ever worked, it was a bug 06:12
masak you need my/our before them nowadays. then it works.
sorear "has" rules
masak but they still can't call each other.
sorear need to be in a subclass of Cursor
masak I know, I know. 06:13
sorear Why can't they call each other?
masak because something is not implemented, I guess.
sorear How are you trying to call them?
06:14 foodoo joined
sorear And why are they in a base class? 06:14
06:14 wtw joined
pmichaud masak: are they in a source file where I could see them, ooc? 06:14
masak pmichaud: the old ones?
sorear considers inheritance harmful in almost all cases
masak sorear: trying to call them from each other.
sorear: thanks for the advice.
sorear masak: but how
masak fetches the source URL 06:15
sorear wait, I have a druid master checkout
masak github.com/masak/druid/blob/master/...id/Base.pm
uuuh, that's not right. those changes should be restricted to the ng-compat branch. 06:16
pmichaud why can't that easily become a grammar instead of a class? Or are there other connections?
masak pmichaud: I don't know, I haven't re-thunk the design yet. 06:17
pmichaud masak: okay.
masak hm, false alarm. that's the latest Druid-for-alpha code. 06:19
pmichaud okay.
masak phew.
right. observe lines 5-7. those are bare regexes *without* my/our. that's illegal now. but note especially line 7. it calls the regexes on lines 5 and 6. that doesn't work on Rakudo master. 06:20
oh, and consequently the similar calls on lines 9-10 don't work either. 06:21
pmichaud it would probably work fine if we s/class/grammar/,
getting 9-10 to work will be... challenging, yes.
masak ok.
pmichaud probably need to switch some of them to &coords
masak I can do s/class/grammar/, that seems feasible.
06:22 stepnem left
masak pmichaud: will &coords still create a $<coords> entry? 06:22
pmichaud <coords=&coords>
masak ok.
pmichaud and declare the regexes as 'our', then. 06:23
(so they appear in the namespace)
sorear My 2¢ is to change Druid::Base to module Druid::Common and export my regexes 06:24
I'll look into this while beating it into working on niecza
pmichaud back tomorrow -- gotta run out to the store for some milk 06:25
masak nice! twitter.com/meraxes/status/24040386254
06:25 Zapelius left, mfollett left
masak I'll look into making Druid work again on master. 06:25
maybe even this weekend.
pmichaud masak++
let me know if you need anything :)
06:26 Zapelius joined
pmichaud I'm slowly working on the backlog of bugs 06:26
masak is slowly adding new ones :) 06:27
06:27 stepnem joined
tylercurtis Does &.foo impose a context? 06:28
sorear &.foo hasn't been defined 06:40
ditto &$foo 06:41
masak how realistic is a perl5-to-perl6 converter nowadays? would it be a selling point for Perl 5 people who hesitate about learning Perl 6? 06:42
06:42 cotto left, cotto joined
jnthn morning, #perl6 06:44
pmichaud jnthn: /o
er, o/
tylercurtis Morning, jnthn.
pmichaud (I'm tired, so my arm drooped :)
jnthn Well, I do sometimes cause people to scratch their head... :-)
pmichaud and I still need to get to the store but I'm too busy bisecting for a parrot bug 06:45
jnthn ugh
Which bug/
pmichaud this one is interesting
masak phenny: tell moritz_ s/whom can inherit/who can inherit/. when in doubt, use 'who', not 'whom'.
phenny masak: I'll pass that on when moritz_ is around.
pmichaud it appears that in recent versions of Parrot, it selects the String PMC .trans method even for Rakudo Str objects 06:46
masak huh
jnthn ugh
pmichaud just a sec and I can paste :)
jnthn Well, I can only offer than in a couple more months we won't be using the String PMC... :-)
pmichaud oh? 06:47
jnthn Doesn't make much sense.
Same for Integer PMC
pmichaud we'll have our own PMC types, then?
jnthn We'll have a repr for the native types
(I think.)
pmichaud > my $x = 'a'; say PARROT($x); $x.trans('a' => 'b'); 06:48
Perl6Scalar->Str
too few positional arguments: 2 passed, 3 (or more) expected
Yes, $x *is a Str*
jnthn :S
pmichaud I'm wondering if a method cache is being lost somewhere.
06:48 Schwern left
jnthn Possibly. 06:48
pmichaud I also thought it might be that $x isn't p6opaque... but I manually added in p6opaqueness and that didn't seem to change anything.
jnthn Well, that'd change the dispatcher that is used. 06:49
pmichaud anyway, I have a good commit and a bad commit, so it's time to bisect :) 06:50
jnthn Yes, hopefully that gives the answer. :-) 06:51
pmichaud I went through it once and it didn't seem to work, so I'm trying again.
I might've missed a step or put in a bad message
s/message/result/ 06:52
sorear masak: 5to6 now is a very interesting problem. ingy probably has some ideas, 5to6 is going to face a lot of the same issues as c'dent
masak sorear: oh, indeed. I haven't considered that.
sorear: I've looked at it from the viewpoint of TimToady's STD5 module.
sorear B or PPI are probably more solid foundations for a standalone conversion tool (as opposed to lexical use v5;) 06:54
sorear wonders how hard it would be to put together B::JSON 06:55
jnthn Wow, mberends++ made a nice ASCII art dependency diagram for 6model :-) 06:57
moritz_ good morning 07:03
phenny moritz_: 06:45Z <masak> tell moritz_ s/whom can inherit/who can inherit/. when in doubt, use 'who', not 'whom'.
masak good mo', mo' 07:04
jnthn Or just never use whom.
It's dying.
masak .oO( 6model has an ASCII art dependency? )
jnthn masak: :P 07:05
masak jnthn: I just love the accusative in all its forms. but I agree, 'whom' is dying.
jnthn Netcraft confirms it.
moritz_ well, German has accusative too, and I don't see why I shouldn't use it in the general case
masak moritz_: since German has the accusitve, I'm especially surprised that you chose to use it on the subject of a phrase. 07:06
s/it/'whom'/ 07:07
see also languagelog.ldc.upenn.edu/nll/?p=1353 :) 07:10
jnthn om nom nom chokladbollar 07:13
.oO( not a typical breakfast food... )
07:14
masak :) 07:15
dukeleto perl6: Failure + 2 07:16
p6eval pugs: OUTPUT«*** No such subroutine: "&Failure"␤ at /tmp/ZjIT54jrwf line 1, column 1 - line 2, column 1␤»
..rakudo 859f2d: OUTPUT«Method 'Bridge' not found for invocant of class ''␤ in 'infix:<+>' at line 3515:CORE.setting␤ in main program body at line 22:/tmp/kEuSouVayG␤»
dukeleto is that a known rakudobug?
sorear it's known LTA 07:18
it is, afaict, correct 07:19
but colomon++ would be the one to know
dukeleto LTA? 07:22
jnthn Less Than Wesome
*Awesome
dukeleto is enlightened 07:23
i would expect a Failure who hasn't had .Bool or .defined on it to trigger an exception, but perhaps + triggers a .defined check? 07:24
pmichaud we still have some issues with the Failure implementation 07:25
but I'm not sure how the type object should behave in this case. Probably like any other type object, though.
dukeleto i am noticing a bunch of spectests fail on rakudo master. is this expected?
perl6: ~Failure 07:26
p6eval rakudo 859f2d: ( no output )
..pugs: OUTPUT«*** No such subroutine: "&Failure"␤ at /tmp/UE7qihWVDH line 1, column 1 - line 2, column 1␤»
pmichaud dukeleto: what revision of parrot are you using?
dukeleto pmichaud: i used --gen-parrot
masak pmichaud: do you perchance recall the etymology of "strangely consistent"? because that's the name of my new blog.
pmichaud dukeleto: say $*VM<config><revision> 07:27
dukeleto pmichaud: 48768
07:28 au|irc_ left, au|irc joined, Eevee left, Eevee joined
pmichaud dukeleto: I don't know then. It should be passing spectests at that revision. 07:30
masak: I don't recall, but I'm looking through the #perl6 logs
masak pmichaud: I did that yesterday. found no real clues that way. 07:31
pmichaud: here's what I remember. it was at YAPC::EU 2009, in the big assembly hall, on the Tuesday.
pmichaud first irc log reference I can find is irclog.perlgeek.de/perl6/2009-03-31#i_1030097
masak pmichaud: either you coined the phrase, or I did.
pmichaud: yes, I found that one too. I'm not sure if that reference counts. it looks too innocent. 07:32
pmichaud agreed
masak only after YAPC::EU is the phrase used with intent.
pmichaud anyway, istr it was at yapc::eu, and I remember you being taken with the phrase. I don't remember the context in which it arose.
masak I liked it well enough to include it in the Web talk on the Wednesday.
07:33 envi^home joined
pmichaud I'm wondering if it came out of the Rakudo BOF 07:33
(which was on Tue, iirc)
masak no, it was before that on the Tuesday.
pmichaud hmmm
masak the BOF wasn't in the assembly hall.
pmichaud well, my memory is toast :)
masak no worried. just thought I'd ask.
pmichaud anyway, whatever story you come up with I'm likely to agree with :)
masak I think you came up with the phrase, and I liked it. :) 07:34
pmichaud That would be indeed strangely consistent. :)
masak today's autopun, brought to you by pmichaud. :)
07:35 jferrero joined
masak .oO( when you execute an autopun with a deadpan expression, is it called an "autopan"? or a "deadpun expression"? ) 07:35
moritz_ masak: (re accusative) I attribute it to my sleepiness 07:37
masak: and I agree it's a misuse (and corrected it) 07:38
07:38 brxue joined
masak moritz_: goodie. I liked the post, by the way. 07:38
dukeleto pmichaud: i am going to rm -rf my parrot_install dir and do a realclean, i think i forgot that. 07:39
sorear how not to allocate a bitfield: new int[bitCount] 07:40
dukeleto should #?pugs lines exist in perl6/roast.git ? 07:42
07:42 cottoo joined, cotto left
sorear yes 07:42
they're just fudges
pugs is an implementation too
07:44 kjeldahl joined
dalek ecza: 368485d | sorear++ | lib/Cursor.cs:
Store backtrack frames in a single object each
07:44
ecza: 940874c | sorear++ | lib/Cursor.cs:
Fix LexerState stupidly using 32x correct memory
dukeleto sorear: of course, just wondering 07:48
07:55 meppl left
moritz_ masak: glad you liked it. Do you agree? or do you still want protected in p6? 07:56
dukeleto: I usually leave the #?pugs lines, but I don't make much effort to maintain them
dukeleto perl6: say /42/.perl 07:57
masak sorear: [backlog] Druid uses type constraints all over the place. but usually things work without the type checks they represent.
p6eval pugs: OUTPUT«{regex}␤»
..rakudo 859f2d: OUTPUT«{ ... }␤»
dukeleto seems like Regex objects don't know how to do .perl in Rakudo
masak moritz_: I'm not sure I ever wanted 'protected' in p6. my colleague asked about it once, and that's when I started a p6l thread about it.
TimToady here's my entry for intentional FP-style obfuscation: see 2nd part of rosettacode.org/wiki/Constrained_Ra...cle#Perl_6
dukeleto TimToady: that is pretty awesome 08:01
TimToady that's probably about as close as we get to APL 08:02
at least in terms of reading everything right-to-left 08:03
funny thing is, I can grok any one bit of it as long as I don't look at the rest of it at the same time. :) 08:04
huf what's the dot in ~.map ? :)
TimToady "unary dot"
means ~ $_.map
huf oh! 08:05
TimToady
.oO(behold the power of this fully functional Death Star)
it only *look*s half constructed... 08:06
pmichaud "You may fire at will, Commander."
(I always wonder if that's "will" or "Will" :)
TimToady I think where there is a Will, he is away. 08:07
08:07 dalek left
pmichaud I notice there's no "Works with Rakudo XXX" line there... is that because it doesn't? ;-) 08:08
TimToady it works with HEAD, at least
at least, as of last time I compiled it 08:09
08:09 dalek joined, ChanServ sets mode: +v dalek
TimToady but for a pleasant change, I don't recall having to work around any bugs this time 08:10
08:10 azert0x joined
huf hmm, i think i have a rough understanding of that snippet 08:10
but i dont know the word for how awesome it is :)
TimToady negatives don't go that low, eh? :) 08:11
huf hey! you're not supposed to be pessimistic
TimToady it really was generated via step-by-step transformation from the previous example
huf you're saying this isnt code we should see in production? 08:12
masak TimToady: that's the most Haskell-like Perl 6 code I've ever seen. kudos.
TimToady depends on how many people you work with that have IQs of 220
huf but... a haskell programmer can write haskell in any language ;)
TimToady I was particularly happy about using given to define a variable declared in the first clause :) 08:13
sort of a cargo-culted Haskell 'where' 08:14
masak aye. never seen that idiom before.
this rosetta code thing is a gold mine. :)
TimToady++
TimToady I think I'd better go to bed before I do any more damage to the credibility of Perl 6... 08:17
or perhaps it's crudability now 08:18
moritz_ sleep well TimToady++
TimToady zzz &
08:19 Schwern joined
masak dream of cargo-culted Haskell idioms. :) 08:20
08:21 icwiener left
pmichaud 08:10 <TimToady> but for a pleasant change, I don't recall having to work around any bugs this time 08:29
that sounds like a good milestone. :)
sorear parse 60kb of JSON -> 2.95 seconds 08:30
parse 600kb -> OOM
*something* is not right here
pmichaud makes me glad that I fixed statement_modifier:<for> :)
sorear ? 08:32
pmichaud (for TimToady's code)
08:32 orafu left 08:33 orafu joined 08:37 drbean left
sorear interesting, mono is 50% faster with --gc=sgen 08:38
tylercurtis Good night, #perl6. 08:41
08:42 Alias_ joined 08:44 Alias left
smash mornin' everyone 08:45
dalek ecza: a10d7a2 | sorear++ | lib/Kernel.cs:
Elide containers for readonly Variables
08:46
ecza: d076f11 | sorear++ | src/ (2 files):
Create array descriptors in regexes once
08:47 drbean joined
smash gil.di.uminho.pt/users/smash/rakudo-bench.html updated 08:49
08:49 tylercurtis left 08:51 thebird joined
sorear is: self := 5; reasonable? 08:51
you can rebind any explicit parameter 08:52
moritz_ I'm pretty sure self has a type constraint
08:53 timbunce joined, JimmyZ joined 08:55 am0c^ left 08:56 JimmyZ_ joined 08:58 dakkar joined, JimmyZ left, JimmyZ_ is now known as JimmyZ 09:01 drbean left 09:03 drbean joined
masak two OSDC.fr talks submitted: "Hard core bug reporting" and "Perl 6: there's some madness to the method after all" 09:04
09:08 cjk101010 joined
smash cleaned up List::Utils.transpose a bit (github.com/colomon/List-Utils/commi...fa#diff-1) more suggestions welcome 09:17
huf i find people have really odd misconceptions about readability 09:18
someone smarter than me could make a good blogpost out of this
masak here's a "shot on sight" opinion that I don't understand: twitter.com/erez/status/24090023075 09:19
pmichaud I found the bug in parrot trunk that is causing Str.trans to fail.... will fix it after sleep though.
sorear I think readability is partly not a function of code length 09:20
sublinear
masak pmichaud++
sorear if I have 100 lines of PIR to do a job, or 10 lines of Perl6, there will be a less-than-10fold difference in reading time
so by a naive interpretation the PIR is "more readable" 09:21
huf i think length matters a lot. if an expression is short enough that i can see it from start to finish without moving my eyes around, it's easier to read 09:22
but what my point actually is is that readability is almost entirely subjective, yet almsot everyone likes to pretend it's an objective measurement
masak there's some point at which I'd gauge compression/golfing as hurting readability. 09:23
huf true, there's a line
but the topic's just another vi vs emacs flame, except people dont seem to realize it yet
pmichaud masak: I don't understand the "shot on sight" opinion either.
sorear compression almost never helps readability
readability is a function of essential complexity 09:24
the difference between refactoring and golfing is that, I think
09:26 risou_ left
szbalint pmichaud: maybe the person doesn't think that the P6 features backported to P5 are really backports? 09:26
huf sorear: it's rare tho that a more constrained form is longer (i mean in the sense that @a = map @b is more readable since there's less surprises that could happen than in the equivalent push @a, ... for @b)
masak huf: that sounds like the "number of primitives" measure of Arc fame. 09:28
sorear it's not just number of primitives, it's also applicability of primitives
two values which have the same linguistic representation might still be different in the reader's mind 09:29
and the cost of the resulting unwritten coercion cannot be discounted
huf hmm?
sorear like the J code for spiral matrices
turning a spiral matrix, into a row-major 1d list, then feeding that through a Burroughs-Wheeler-like transform, creates several very expensive mental coercions 09:31
even though it's just a handful of characters
09:33 jhuni left
huf sorear: that depends heavily on what kind of manipulations you consider basic 09:34
which is in turn shaped by all the languages you've used
sorear no, no
that's not what I mean at all
since you seem to be distracted by the complexity of the BWT, try x & (x - 1) 09:35
this is a very basic operation on bit vectors
but in order to understand it, you have to see it as x & integer2bitvec(bitvec2integer(x) - 1)
so it's much harder to grasp than x * (x - 1), even though it uses the same number and complexity of primitives 09:36
huf but say you use a language that has a unary operator for x & (x-1) ;)
use it for a few years and suddenly it's a basic operation in your head 09:37
and you no longer think about bitvec->int conversions at all
florz you totally do
sorear again you are missing the point
florz for that's the point of the idiom =:-)
sorear the readability of x * y does not depend on the readability of the microcode that implements addition
or multiplication for that matter 09:38
the readability of %x, likewise, has absolutely nothing to do with the readability of x & (x - 1) or whatever else it expands into
09:39 daxim joined 09:40 Ross joined
huf sorear: but if you call it ispow2(x) it's suddenly really easy and you dont have to think about any of that at all 09:43
sorear huf: That's exactly what I just said
huf then i dont get what you're saying, sorry 09:45
09:50 Ross left
masak lunch & 09:53
09:57 risou joined 10:00 JimmyZ left 10:03 QinGW1 left
brxue hello, is perl6 coming soon? 10:05
sjohnson masak: whats for lunch?
brxue: just around the corner! 10:06
the devs here have been putting enough man hours into perl6 rakudo to move a small mountain
brxue I'm so expecting it
sjohnson you can already get rakudo perl in a working state 10:07
but i imagine very soon, an official stable release will be announced
brxue thanks for all your guys' hard work
it's so exciting 10:08
sorear pmichaud says there never will be an official release
brxue any reason? 10:09
sjohnson sorear: if not that, how bout one that will fit the needs of anyone curious about a stable perl 6 implementation? 10:10
sorear brxue: because perl6 is defined by a spec and a philosophy, there is no canonical implementation 10:11
10:11 Ross joined
brxue I reviewed perl6's tutorial, I think it's wonderful 10:12
sjohnson i'm pretty excited as well
i use perl5 a ton 10:13
but am really excited about the innovations in p6
try rakudo star, you will probably enjoy it 10:14
brxue It seems Larry Wall made a presentation on O'Reilly conference about Perl6 a few months ago, but I cann't find this info from official site of perl 10:16
sjohnson he'll probably be awake in a few hours 10:17
brxue this is a URL which cover this news in chinese developer.51cto.com/art/201007/214223.htm 10:19
10:20 Ross left
brxue the theme of this presentation seems to be "status of onion" 10:20
10:20 Ross joined
sjohnson heh 10:21
www.perlfoundation.org/perl6/index...._the_onion
your last msg should be the topic of the day 10:22
:)
10:22 [Coke] joined
mathw The trouble with the wiki page is that there is no content for the last few talks 10:22
:( 10:23
But I can't really criticise, I wasn't there to do it myself
sjohnson yeah 10:24
do you mean the wiki for the url i just pasted, or wikipedia's perl6 link?
mathw I mean what you just pasted
sjohnson cause i found the later to be a burst of fresh air of refreshing facts and knowledge.com
latter* 10:25
mathw: 2009 is more recent than the others i found a few months ago, so i guess it's not too bad
:-) 10:26
mathw 2009 is a placeholder
sjohnson DOH! 10:27
i assumed that by the lack of content, people were too awe-struck with p6's work to say anything 10:28
mathw lol 10:29
well we have R* now
a shining beacon
keeps me warm at night
after the cat has abandoned me
sjohnson haha 10:30
i feel guilty for not trying it 10:31
maybe too many bad experiences with linux compilations from scratch have scarred me
but someone here told me it isn't that bad to setup
florz it's really easy indeed, on debian lenny at least
dalek ecza: af3387d | sorear++ | / (8 files):
Avoid creating throwaway containers in the binder
ecza: f9bb192 | sorear++ | src/CgOp.pm:
Implement the CPS constants optimization

I thought I did this weeks ago. Weird.
ecza: 199cf9e | sorear++ | lib/Kernel.cs:
Optimize Kernel.Bind et al a bit more
sjohnson mathw: you're not alone. a friends cat i babysit is pretty cute 10:34
but scratches me often
doesnt like snuggling either 10:35
mathw oh mine likes to snuggle 10:36
but only when he's in the mood
and if I fall asleep, he'll be trying to wake me up within about half an hour
apparently the proper order of things is me being awake while he naps in my lap
sjohnson haha
here's a youtube link of the cat if you wanna see a cute cat video i took 10:37
www.youtube.com/watch?v=z_HQYQgsfqc
mathw I would, but youtube is really clunky at work
10:37 icwiener joined
mathw our proxy puts it at priority -infinity 10:37
sjohnson as cute as he is, he's pretty spoiled and crashes people in your lap scenario for not cuddling him as he likes
crashes => scratches*
mathw the cutest cat I know is also the most antisocial
she'll cuddle for a little bit
and then take your face off
but she's an irresistably cute little furball with the most amazing tail 10:38
it's so big that she's been known to chase its shadow
sjohnson mathw: next best thing: imgur.com/3q36n.jpg 10:39
screenshot i took just for you :)
mathw aaawww
it's that pleading expression
sjohnson yeah
he looks cute
but damn that cat is mean sometimes
you'll be petting his back
he's enjoying it for a about 5 mins
then without warning, instant slash with claws at your entire arm 10:40
for whatever reason
has happened to about 4 people so far
and i feel sorry for them all as it hurts their feelings. he's spoiled!
mathw www.flickr.com/photos/mattofwalton/...820239428/ fairly typical position for mine
sjohnson perhaps the scenario you described, is the same cat syndrome for this one
the owner jokes that he's austistic. now it doesn't seem too far off 10:41
tabby!!!
mathw well fortunately Rascal usually goes for you with claws retracted at first
then if you don't heed the warning, the claws come out
sjohnson my mom has a cat that looks exactly like that
mathw :)
sjohnson like, exactly
mathw there's one on my street who looks quite similar
sjohnson different colour eyes, but identical nonetheless
very cute
mathw yeah
Pumba is ridiculously soppy
total charmer 10:42
even won over my parents and they don't even like cats
sjohnson heh
i think i have a pic
of my parents cat on facebook
cause someone took a pic of it recently
mathw he also likes my sister
last time she came to visit he was straight in her lap 10:43
but his napping was disrupted by her giggling
apparently he tickles
sjohnson heh 10:44
thats cute
damn
just pics of his sunbathing
but i have a video of him sniffing my iphone4
so i will have to give you the url for sometime you aren't at work 10:45
and you will appreciate how similar he looks
sunbathing pic: sphotos.ak.fbcdn.net/hphotos-ak-snc...1853_n.jpg
mathw oh wow 10:47
what a cat
what a place for a cat to sunbathe
Pumba would love it 10:48
nice warm concrete slab
lots of space to run around in
I tried to take a picture of pumba nosing at my phone last night but it didn't work very well because the flash fired
so it's like glaringly bright cat nose on black
sjohnson yeah it's my parents house. very nice view in canada 10:49
but they aren't rich, and the house is very small... but they swiped a good spot in the early 80's
yeah
10:49 Schwern left
sjohnson i have also found flashes to be very anti-cute-cat face prone 10:49
i usually hold the camera with no flash in one hand 10:50
and a toy to distract them in the other
to get cute wide-eyed anime faces from them
and not blink when a camera snaps a photo at them
but you almost need a tripod for decent photos
mathw sometimes you get lucky though
sjohnson otherwise you come up with this: sphotos.ak.fbcdn.net/hphotos-ak-snc...5007_n.jpg 10:51
mathw ah
yes
not so good
sjohnson need to get some tripod action going
mathw ah here she is 10:52
www.flickr.com/photos/mattofwalton/4882653325/ <- the evil furball known as Rascal
sjohnson thats cute
mathw that's how she lures you in
she's cute, playful, violent and has an insatiable appetite for knitting yarn
sjohnson yeah
i understand completely !
mathw I enjoy playing with her 10:53
sjohnson thats how all my friends keep getting scratched
mathw but I'm glad she doesn't live with me
Pumba's much older though
sjohnson ever played hide and go seek with a cat?
mathw no
sjohnson in the way of, peeking around a corner?
then pretending to hide?
usually gets a lot of good reactions from the more active cats
probably in my top 3 fun things to do with certain cats
some cats dont go for it, other cats will sneak up on you too 10:54
then make a daring lunge at you
even if they are old lazy ass cats that hardly move an inch all day, as in the case with our previous orange tabby
not a claw lunge, but one where they can't take the suspence off peek-a-boo, then see what you are hiding behind 10:55
mathw :)
I'll have to try that with Pumba
sjohnson depending on the cat, you will be amazed at the results
just try to pretend that you are trying to hide from them as best you can, hiding behind sofas, but making a peek, then hiding again as if you dont want them to see you
its a ton of fun
making it obvious that you are peeking so they see you, but then pretending like you dont want them to see you 10:56
hours of entertainment!
anyways matt, i was good chatting with you again 10:57
i should probably hit the sack and go to work at a half-decent hour
ttyl!
mathw bye!
sjohnson maybe work be done faster and more enjoyable if i just imaging playing with cats instead of reverse engineering other people's uncommented javascript :-( 10:58
11:02 Ross left
dalek ecza: 10151ff | sorear++ | / (3 files):
Optimize cutting of subrules

Setting the committed flag is enough from an operational standpoint, but it doesn't allow the search tree to be pruned by the garbage collector. The new model makes that possible.
11:02
sorear 600kb in 8.1 sec and 23mb \o/ 11:03
11:08 Ross joined
sorear out. 11:10
11:12 yogaboy joined 11:15 yogaboy left
masak sjohnson: I had sushi for lunch. yummy. 11:15
sjohnson++ # perl 6 explanations to brxue 11:17
yay, OSDC.fr talks accepted \o/ 11:19
colomon \o/ 11:20
11:21 wtw left 11:24 wtw joined 11:28 aloha left 11:30 Ross left, bacek left 11:35 oyse joined 11:36 patrickas joined, [Coke] left
patrickas o/ 11:37
11:37 Ross joined
masak \o 11:38
rakudo: sub check($_) { my $l = 0; return 0 if ($_ eq "[" ?? ++$l !! --$l) < 0 for .comb; return 1 }; for ^5 { say "$_ is {.&check ?? "" !! "in"}valid" given (<[ ]> xx 5).pick(*).join }
p6eval rakudo 859f2d: OUTPUT«][[]][[][] is invalid␤]][]][[[][ is invalid␤]][[[][][] is invalid␤[]][[[[]]] is invalid␤][[]][][][ is invalid␤»
patrickas LOLITSMASAK!
masak dang, only invalid ones :)
guess they're that much more common...
rakudo: sub check($_) { my $l = 0; return 0 if ($_ eq "[" ?? ++$l !! --$l) < 0 for .comb; return 1 }; for ^5 { say "$_ is {.&check ?? "" !! "in"}valid" given (<[ ]> xx 5).pick(*).join } 11:39
p6eval rakudo 859f2d: OUTPUT«[][][[[]]] is valid␤][[[]][[]] is invalid␤[[][[]]]][ is invalid␤]]][[[][][ is invalid␤[][]][]][[ is invalid␤»
masak \o/
patrickas masak: you remind me of Grissum from CSI 11:40
masak unaware of that reference :) 11:41
except that CSI is probably something on TV...
rakudo: sub check($_) { -1 == none([\+] map { $_ eq "[" ?? 1 !! -1 }, .comb) }; for ^5 { say "$_ is {.&check ?? "" !! "in"}valid" given (<[ ]> xx 5).pick(*).join }
p6eval rakudo 859f2d: OUTPUT«][[[[]][]] is invalid␤][[]][[]][ is invalid␤]][[[[[]]] is invalid␤[[]][][][] is valid␤]][[[]][][ is invalid␤»
11:41 CoCol joined
masak \o/ 11:41
patrickas He's an entomologist :-p 11:42
masak I'll take that as a compliment :P
(whatever is an entomologist doing on a CSI team?) 11:43
patrickas well Forensic entomologist
masak gotta know who killed those bugs... 11:44
patrickas Well usually masak is the last person to have seen them unreported so I guess we better start by questioning him! 11:45
11:47 macroron joined
masak :) 11:47
11:50 Axius joined
takadonet morning all 11:51
masak o/ 11:52
rakudo: for ^5 { say "$_ is {none([\+] map { $_ eq "[" ?? 1 !! -1 }, .comb) < 0 ?? "" !! "in"}valid" given (<[ ]> xx 5).pick(*).join } 11:56
p6eval rakudo 859f2d: OUTPUT«]][]][][[[ is invalid␤][[[][[]]] is invalid␤[[[]]]][][ is invalid␤[]][]][[][ is invalid␤[]][]][][[ is invalid␤»
12:06 meppl joined
masak phenny: tell mberends that if he plans to submit OSDC.fr talks: deadline falls today. 12:11
phenny masak: I'll pass that on when mberends is around.
12:17 CoCol left 12:18 perlygatekeeper1 joined, perlygatekeeper left, meinwald left, plainhao joined 12:19 meinwald joined, meinwald is now known as Guest39477, Eevee left 12:20 Eevee joined, CoCol joined 12:31 sftp joined, ruoso joined 12:34 Axius left 12:35 Axius joined 12:41 macroron left 12:52 risou left 12:56 Mowah joined 12:57 oyse_ joined, oyse left 13:00 Mowah_ joined 13:01 Mowah left, CoCol left 13:06 Mowah_ left
masak Perl 5 is not very friendly to us poor Perl 6 programmers who regularly forget the parentheses on ifs and loops :/ one gets all manner of strange unrelated errors doing that. 13:06
takadonet heh 13:07
masak in fact, I've come to generally associate a big heap of unrelated errors with the possibility that I might have forgotten such parens.
maybe I should go on p5p and ask that they make the error messages slightly less LTA in this circumstance? :) 13:08
Juerd use List::Util qw(sum); # [+]
masak Juerd: ooh! one could make a blog post of "imports you don't need to make in Perl 6" :) 13:09
Juerd I still do prefer "sum" to "[+]" by the way
rakudo: (1..5).sum.say 13:10
p6eval rakudo 859f2d: OUTPUT«Method 'sum' not found for invocant of class 'Range'␤ in main program body at line 22:/tmp/izRahoH1R3␤»
Juerd Meh :)
masak huh.
alpha: say (1, 2, 3).sum
p6eval alpha 30e0ed: OUTPUT«Method 'sum' not found for invocant of class 'List'␤in Main (file src/gen_setting.pm, line 324)␤»
jnthn It's sumwhat non-generic though. :-)
13:11 Mowah joined
masak jnthn: welpun back! we missed you :) 13:11
jnthn I've been disappointed at the lack of opportunities for bad puns of late, but I figured one would happun sooner or later. :-P 13:12
13:12 orafu left 13:13 kaare_ left 13:14 kaare_ joined, orafu joined 13:16 Patterner left 13:17 Psyche^ joined
takadonet is there a better way to capture in a regex decimal and non decimal values then: (\d+'.'\d+|\d+) for values like : 75.34 245 5.3 etc.... 13:17
13:17 Psyche^ is now known as Patterner
PerlJam takadonet: / ( \d+ [ '.' \d* ]? ) / # dunno of that's "better" 13:18
s/of/if/ 13:19
masak alpha: say ?(42.18 ~~ /<Perl6::Grammar::dec_number>/) 13:20
p6eval alpha 30e0ed: OUTPUT«1␤»
masak that's my preferred method :)
takadonet PerlJam: i have that as well.
PerlJam masak++
takadonet rakudo: say ?(42.18 ~~ /<Perl6::Grammar::dec_number>/)
p6eval rakudo 859f2d: OUTPUT«===SORRY!===␤regex assertion not terminated by angle bracket at line 22, near "::Grammar:"␤» 13:21
masak rakudo--
takadonet hmmm
all well
masak jnthn: have you ever savoured Finnegan's Wake? www.trentu.ca/faculty/jjoyce/fw-16.htm 13:22
jnthn ...no?
jnthn looks
masak it's... odd in a good way. :) 13:23
and good in an odd way.
jnthn ... 13:24
masak *lol* 13:25
jnthn It's odd alright :-)
He can prapsposterus the pil-3
lory way to Hirculos pillar
lol
:-)
masak "You tollerday donsk? N. You tolkatiff scowegian? Nn. You spigotty anglease? Nnn. You phonio saxo? Nnnn." 13:26
jnthn Wid wad for a norse like?
masak Load Allmarshy! 13:27
13:28 Ross left
jnthn In the ignorance that implies impression that knits knowledge that finds the nameform that whets the wits that convey contacts that sweeten sensation that drives desire that adheres to attachment that dogs death that bitches birth that entails the ensuance of existentiality. 13:28
masak *lol*
jnthn Sentence...win. :-)
I'm still not sure if any of this actually means anything or just sounds like it does. :-) 13:29
masak it's a very interesting book.
James Joyce doesn't use language like the rest of us do. :)
arnsholt_ IIRC it's supposed to make the same kind of sense dreams do =) 13:30
jnthn I daren't read it in case I start this way speaking unstoppably the chat or relayed internets.
masak arnsholt_: yes. it's basically livestreaming a dreamer's thoughts.
13:30 Zapelius left 13:31 arnsholt_ is now known as arnsholt
masak arnsholt_: the awesome thing is, people are pretty sure of the dreamer's *name* from the contents of the book! 13:31
arnsholt Nifty ^^
masak and of course, a whole lot of other details pertaining to the dreamer. 13:33
I do hope to take the time to read it all at some point in my life.
13:33 kaare_ left
jnthn oh noes I've got chocolate all over my multi-dispatch design notes. 13:34
masak solid or liquid? 13:35
jnthn solid but I guess it musta got liquid enough to get sticky at some point 13:36
13:36 Ross joined
jnthn ...maybe this is nature's way of telling me I have a sweet design. 13:36
PerlJam looks like you'll just have to lick them clean then 13:37
masak yes, we prefer a clean design.
jnthn: suck it up!
jnthn :P
meh, this sucks, we should know all the candies at compile time anyways... 13:38
jnthn tries again
13:44 colomon left 13:51 jferrero left 13:54 Guest39477 is now known as meinwald 13:59 Axius left 14:01 Italian_Plumber joined 14:05 Lorn joined 14:06 uniejo left 14:15 colomon joined 14:16 pkkm joined, masonkramer joined 14:18 stepnem left 14:19 icwiener left, stepnem joined 14:21 brxue left
pmichaud good morning, #perl6 14:22
14:22 meppl left
masak morning, pmichaud. 14:23
takadonet pmichaud: morning
colomon o/
14:25 CoCol joined 14:27 wtw left
mathw o/ 14:30
14:32 mfollett joined 14:35 patspam joined
jnthn morning, pmichaud 14:35
mathw takes deep breaths 14:36
Sometimes I'm very tempted to take a solemn vow never to use C++ ever again
arnsholt The C++ FQA makes me question ever learning it =) 14:37
14:37 justatheory joined
mathw Then I realise that it's not really C++ I'm angry with, it's Visual Studio 14:37
14:40 sftp left, dual left, sftp joined
pmichaud jnthn: ping 14:41
jnthn: could you (or anyone else) verify that check_invokable (src/pmc/perl6multisub.pmc:804-808) isn't actually being used anywhere? 14:42
i.e., I can remove it?
14:45 masonkramer left 14:54 Ross left 14:55 timbunce left
masak "Joyce's suggestion, like Grice's, is that we create discourse out of a vast, in principle infinite assemblage of linguistic and encyclopedic elements, and that the threshold at which we label any given discourse as incoherent is in part a function of particular communicative contexts." www.thefreelibrary.com/The+Mutt+and...a016528210 14:58
I'm soddy, bat this orl remynds me of the blag proust "Idiotic Parl Sicks" :) 14:59
und so, is in FAQ nott off tropic :) 15:00
aur at LISP, naught enteyerly. 15:02
15:04 timbunce joined, Holy_Cow joined 15:05 satyavvd joined, oyse_ left 15:07 tadzik joined
masak s/Idiotic/Audiomatic/ 15:08
15:14 satyavvd left, Ross joined
tadzik moritz_: www.reddit.com/r/programming/commen...?context=3 -- this might have been clearly said in the post 15:17
15:20 spq1 joined 15:24 mfollett left 15:26 tadzik left 15:29 bacek joined
jnthn pmichaud: Gotta decommute - will check when I get home. 15:34
pmichaud: though imo if you remove it and it compiles, then it's fine :-)
pmichaud jnthn: that's what I'm doing now.
15:34 aloha joined
pmichaud I couldn't find any other references to it in the repo (and it's static, so they'd be limited to the one file anyway) 15:35
dalek kudo: c957c9f | pmichaud++ | build/PARROT_REVISION:
Bump PARROT_REVISION to update to a known working 64-bit Parrot (passes all spectests).
15:36
15:46 Ross left
dalek kudo: c84d0d3 | pmichaud++ | build/PARROT_REVISION:
Oops. *This* is the correct Parrot revision for 64-bit builds -- fixes c957c9f.
15:54
15:57 Guest50200 joined 15:58 Guest50200 is now known as Ross, Ross left, Ross joined, zb joined 16:00 Ross left 16:04 ashleydev left, ashleydev joined 16:05 Ross joined 16:08 alksentrs joined
masak nom & 16:11
16:11 masak left 16:12 patrickas left, foodoo left
takadonet anyone have some good complex examples on how to use action methods with a grammar? (other then the spec tests) 16:12
PerlJam takadonet: I find your request odd. :-) Why do you want "good complex examples"? 16:17
16:19 patrickas joined
pmichaud takadonet: I don't know if this helps at all, but yesterday evening I wrote github.com/perl6/nqp-rx/blob/master...s/json.nqp 16:19
16:19 patspam left
takadonet PerlJam: That involves reading nested file formats such as : www.ncbi.nlm.nih.gov/Sitemap/samplerecord.html which using the action class creating an (object or complex data structure) 16:20
16:20 cdarroch joined, cdarroch left, cdarroch joined
takadonet I currently have my grammar parsing out everything nicely just need to put into a nice data structure 16:21
btw I'm not doing what is at the link but a hmmer profile file
pmichaud: that will help thanks! 16:22
16:22 rgrau_ joined
arnsholt This is odd. For some reason my TOP action method isn't called... 16:23
pmichaud takadonet: also see Ch. 10 of the Using Perl 6 book, if you haven't seen that.
16:26 dual joined
arnsholt Specifying TOP as method instead of a regex-like shouldn't suppress the call of the action method, should it? 16:28
pmichaud yes, it does.
if you write it as a method, you have to invoke the action method yourself.
automatic calls to the action method are a function of the regex, not of the dispatcher.
dalek kudo: 7517e39 | pmichaud++ | src/pmc/perl6multisub.pmc:
Remove dead check_invokable() code from perl6multisub.pmc .
16:29
arnsholt Right, I misread the NQP code then. Cheers! 16:30
takadonet pmichaud: I have seen it thanks :) 16:34
arnsholt But in that case, how does TOP in NQP's Actions.pm get called? I can't find an explicit call to it
16:38 amkrankruleuen joined, patspam joined
pmichaud arnsholt: it doesn't. 16:48
arnsholt Ah. Right =)
pmichaud since the TOP method returns the result of comp_unit directly, we get $<comp_unit>'s .ast also. 16:49
We could/should remove the TOP method from Actions.pm, I guess.
16:50 tylercurtis joined 16:51 kjeldahl left
arnsholt Yeah, I think I get it now. Thanks for your help and patience 16:53
16:54 patspam left 16:55 kjeldahl joined 16:57 dakkar left, daxim left 17:08 thebird left
TimToady rakudo: say -> Int $x { $x < 2 ?? 1 !! $x * &?BLOCK($x - 1); }.(10); 17:08
p6eval rakudo 7517e3: OUTPUT«Could not find sub &?BLOCK␤ in <anon> at line 22:/tmp/yBcmPGYKSE␤ in main program body at line 22:/tmp/yBcmPGYKSE␤»
TimToady is that a known bug?
it seems to be counting the pointy as a routine ("sub") 17:09
actually, this one looks borken too: 17:10
rakudo: say (sub (Int $x) { $x < 2 ?? 1 !! $x * &?ROUTINE($x - 1); }).(10);
p6eval rakudo 7517e3: OUTPUT«Could not find sub &?ROUTINE␤ in <anon> at line 22:/tmp/55M0HpCZoq␤ in main program body at line 22:/tmp/55M0HpCZoq␤»
TimToady alpha: say (sub (Int $x) { $x < 2 ?? 1 !! $x * &?ROUTINE($x - 1); }).(10); 17:11
p6eval alpha 30e0ed: OUTPUT«invoke() not implemented in class 'Undef'␤in Main (file <unknown>, line <unknown>)␤»
TimToady pugs: say (sub (Int $x) { $x < 2 ?? 1 !! $x * &?ROUTINE($x - 1); }).(10);
p6eval pugs: OUTPUT«3628800␤»
TimToady pugs: say -> Int $x { $x < 2 ?? 1 !! $x * &?BLOCK($x - 1); }.(10); 17:12
p6eval pugs: OUTPUT«<SubPointy(<anon>)>␤*** Cannot cast from VBool True to Pugs.AST.Types.VCode (VCode)␤ at /tmp/Pgi8cOHGi6 line 1, column 1-60␤»
17:15 patrickas left 17:18 rgrau_ left
TimToady seems like those oughta BLHF 17:20
had to note non-functionality at rosettacode.org/wiki/Y_combinator#Perl_6 17:21
17:22 pkkm left
TimToady rakudo: say .(10) given { $_ < 2 ?? 1 !! $_ * &?BLOCK($_ - 1) } 17:24
p6eval rakudo 7517e3: OUTPUT«Could not find sub &?BLOCK␤ in <anon> at line 22:/tmp/M95UNzHzlo␤ in main program body at line 22:/tmp/M95UNzHzlo␤»
17:25 patspam joined
moritz_ is pretty sure that it worked in pugs at some point 17:25
TimToady there's an idiom waiting to happen
17:33 jaldhar left, smash left 17:36 timbunce left
pmichaud afaik, &?BLOCK and &?ROUTINE are nyi 17:37
they're not exactly LHF, either.
moritz_ I just thought about how to implement them 17:38
pmichaud (although that may be due to insufficient imagination)
17:38 Axius joined
moritz_ and I think the biggest problem is getting a reference to an anonymous routine that might not have been compiled yet 17:38
pmichaud oh! 17:39
no.
moritz_ (sub f() { &?ROUTINE() }).() 17:40
(sub () { &?ROUTINE() }).()
pmichaud one of the challenges is that &?BLOCK and &?ROUTINE need to reference the Code object, not the Parrot Sub
moritz_ where in the compilation does the wrapping happen? 17:41
pmichaud block_create_closure and block_create_code 17:42
however, if we're allowed to assume that &?BLOCK and &?ROUTINE are indeed compile-time constants -- i.e., they refer to the static code object and not the dynamic instance, it might be much simpler.
moritz_ that's how I understand the spec, yes 17:43
pmichaud oh, then it's not nearly so hard.
moritz_ oh, you thought about going through getinterp ?
pmichaud that's one possibility
17:43 rgrau_ joined 17:44 stepnem left
moritz_ we need the parrot sub -> Code transition for callframe() at some point anyway 17:44
TimToady callframe(0).my.<__CANDIDATE_LIST__> doesn't quite get there
but static is more better anyway
pmichaud moritz_: it can also be done as a compile-time item
moritz_ pmichaud: I don't understand how I can get a reference to sub that hasn't been compiled yet 17:45
pmichaud but I need to settle on a generic framework for handling all compile-time vars, and not a bunch of special cases
moritz_: all I need is a reference to its PAST::Block
we actually do it in several places in Actions.pm
(we have to, because of order-of-initialization issues)
moritz_ pmichaud: and then compile it again? or is there a way to get the reference of the compiled sub?
pmichaud PAST::Var.new( :value($block_past) ) # iirc 17:46
moritz_ ah
pmichaud no, wait
PAST::Val.new( :value($block_past) )
moritz_ finds that too much knowledge about rakudo internals is in #perl6, instead of in source code comments
TimToady wonders which is wrong more often... 17:47
pmichaud returns a compile-time constant reference to the Parrot Sub corresponding to $block_past, without re-generating $block_past
17:47 stepnem joined
moritz_ rakudo: my $x = <-> $a { } 17:49
p6eval rakudo 7517e3: OUTPUT«===SORRY!===␤Confused at line 22, near "my $x = <-"␤»
pmichaud Anyway, &?BLOCK could likely be created as a "simple" make block_code(@BLOCK[0]); 17:50
moritz_ and &?ROUTINE would go through @BLOCK and search for the first routine?
pmichaud yes. 17:51
and it might need to be block_closure_ref(...)
TimToady wonders if @BLOCK is a space leak
pmichaud It's a stack.
dukeleto i get a bunch of "Missing test file" when I run "make spectest" on rakudo master. is that expected?
pmichaud It's going to eventually be @*BLOCK, at the least.
Or more likely we'll switch to simply $BLOCK with a chain of <OUTER> references.
er, $*BLOCK 17:52
moritz_ dukeleto: not expected
TimToady dynamic variable doesn't seem right either
moritz_ am I right in assuming that each ?-twigil var will need a special case in the compiler? 17:53
pmichaud moritz_: I don't think that's right, or desirable. 17:54
TimToady rakudo: constant $?foo = 42;
p6eval rakudo 7517e3: OUTPUT«===SORRY!===␤Constant type declarator not yet implemented at line 22, near "= 42;"␤»
pmichaud I'd certainly like to avoid too much special-case-ness in Actions.pm
dukeleto moritz_: this is what I see when I type "make spectest" in rakudo: gist.github.com/574069
TimToady some special cases are needed, but most constants can live in the normal symbol table seen by constant declarators 17:55
if they were yi
dukeleto moritz_: i am getting a bunch of failures too, but something must be wrong on my machine. I blew away every version of parrot I had on my machine, and I am using --gen-parrot, but something is still borked
TimToady or would that be syi?
moritz_ thought about things like $?LINE, $?FILE, &?BLOCK etc.
TimToady $?LINE has to be special, methinks, but $?FILE doesn't 17:56
17:56 tadzik joined
moritz_ dukeleto: that's weird... try to rm -rf t/spec and 'make testable' again? 17:56
or maybe your rakudo copy is stale?
anyway, afk
pmichaud TimToady: is something like CALLER::<?$LINE> valid? I've been assuming "yes" 17:57
dukeleto moritz_: will try, thanks.
pmichaud er, CALLER::<$?LINE>
TimToady I think I said that one wasn't necessary, we can do caller.line or some such
pmichaud okay. 17:58
TimToady rakudo: say callframe().line
p6eval rakudo 7517e3: OUTPUT«22␤»
17:58 mfollett joined
pmichaud oh, and it even works. Shiny. 17:58
18:00 mberends joined 18:01 mfollett left 18:05 dju left, dju joined
colomon rakudo: say callframe().file 18:07
p6eval rakudo 7517e3: OUTPUT«/tmp/K8zTKm_UYf␤»
colomon I presume there's a way to walk back up the stack? 18:08
pmichaud callframe(1)
callframe(2)
callframe(Routine)
etc.
See S06.
TimToady S06:2350 in fact 18:09
colomon sweet
TimToady ignore DYNAMIC::<$?LINE> though :)
pmichaud submits specbug :-P 18:10
18:12 Ross left 18:13 mikehh joined 18:14 meppl joined
shortcircuit It occurs to me, watching P6 development in here, that P6 is what happens when you get a bunch of hackers together collaboratively building a language out of features they like. 18:15
TimToady that's more or less what we announced in 2000 18:16
shortcircuit I've seen several langauges where they were designed to include the parts the author likes. Some even hit production as part of commercial entities, but P6 may be the only one I've seen that throws hackers and collaboration into the formula. 18:17
18:17 Ross joined
shortcircuit is failing at English sentence structuring today. 18:17
TimToady we still needed to unify the design though; the original proposed feature set was completely incoherent 18:18
shortcircuit TimToady: I remember that announcement. I think I still have the relevant Linux Magazine issue somewhere. :)
dukeleto moritz++ # rm -rf t/spec; make testable seems to have worked 18:19
i think the way that rakudo updates t/spec may need some tweaking
seems like a "git pull --rebase" would be better than a "git pull", and perhaps an explicit "git checkout master" 18:20
TimToady walk &
18:20 Chillance joined
dukeleto otherwise, if someone is working on a branch in t/spec, "git pull" will commit a merge to whatever branch it is one 18:20
s/one/on/
18:22 Axius left 18:24 alksentrs left
dukeleto is (1,1) .. (2,1) supposed to be an infinite list ? 18:26
perl6: (1,1) .. (2,1)
p6eval pugs, rakudo 7517e3: ( no output )
pmichaud rakudo: say (1,1) cmp (2,1) 18:29
p6eval rakudo 7517e3: OUTPUT«-1␤»
pmichaud rakudo: say (1,1).succ 18:30
p6eval rakudo 7517e3: OUTPUT«1 2␤»
pmichaud hmmmmm
I don't quite understand that one. 18:31
rakudo: say ~(1,1).^methods
p6eval rakudo 7517e3: OUTPUT«item ACCEPTS perl hash Capture of elems at_pos sort Str iterator Bool elems fmt item Num Str Numeric Int list acotanh trim-trailing from-radians capitalize pred atanh cosec exp lc p5chop chars sinh to-radians acotan lcfirst roots lines sec trim asec trim-leading cotanh cis
..log log…
18:31 cjk101010 left
pmichaud rakudo: say Parcel.^methods 18:31
p6eval rakudo 7517e3:
..OUTPUT«ACCEPTSperlhashCaptureofelemsat_possortStriteratorBoolitemStrNumericIntlistelemsfmtitemNumcotanatan2lcfirstucseccislogsubstrfrom-radianssprintftranssinIntpredacosecsubstsinhcosechacosechabschompindexceilingtrimasectrim-leadingunpolartrim-trailingsechordrandtruncateRatrindexsqrtasin…
pmichaud rakudo: say ~Parcel.^methods
p6eval rakudo 7517e3: OUTPUT«Bool item ACCEPTS perl Capture hash of elems at_pos sort Str iterator list elems fmt item Num Str Numeric Int trim asec trim-leading cotanh ucfirst tan sech cos rand truncate fmt trans sqrt asin cosh succ subst sign Numeric chomp index uc chr acotanh samecase substr
..trim-trailing …
pmichaud ah, Parcel ~~ Iterable and Iterable ~~ Cool 18:32
not sure about that second one.
colomon rakudo: say Failure + 4
p6eval rakudo 7517e3: OUTPUT«Method 'Bridge' not found for invocant of class ''␤ in 'infix:<+>' at line 3515:CORE.setting␤ in main program body at line 22:/tmp/6N0cR37pkx␤»
colomon rakudo: say (+Failure) + 4 18:34
p6eval rakudo 7517e3: OUTPUT«Use of uninitialized value in numeric context␤4␤»
dukeleto colomon: i ran into that Failure failure yesterday :) 18:35
colomon dukeleto: seems like we should have a more graceful / useful way of handling it.
dukeleto how do I run one spectest file verbosely?
tadzik perl6 /t/spec/stuff I believe 18:36
s:/t/:./t/:
that's how I do it
dukeleto tadzik++ 18:37
duh :)
dukeleto just submitted this as a rakudo bug: gist.github.com/574122 18:39
one spec test coredumps for me, the other doesn't parse. I hope I included all necessary info, please let me know if I didn't 18:40
i just added the backtrace to the coredump to the gist 18:41
colomon: is Failure + 2 a general issue with attempting to use Roles in a numeric context ?
perl6: Role + 2 18:42
p6eval pugs, rakudo 7517e3: ( no output )
colomon dukeleto: It's an error using anything that's non-numeric in a numeric context, I think.
dukeleto perl6: say (Role + 2)
jnthn pmichaud: Maybe Iterable wants to be a role? 18:43
p6eval pugs, rakudo 7517e3: OUTPUT«2␤»
dukeleto colomon: Role seems to be fine
colomon rakudo: class Blueberry { has $.skin; }; say Blueberry.new + 4
p6eval rakudo 7517e3: OUTPUT«Can't take numeric value for object of type Blueberry␤ in 'Any::Numeric' at line 1386:CORE.setting␤ in 'infix:<+>' at line 7072:CORE.setting␤ in main program body at line 22:/tmp/yl3noLOJFq␤»
colomon I guess that is a better message.
oh, now I remember!
the Failure issue is caused by
rakudo: say Failure ~~ Numeric
p6eval rakudo 7517e3: OUTPUT«1␤» 18:44
18:44 tbalazs joined
colomon I guess the real question, is what should Failure + 4 do? 18:44
PerlJam fail.
colomon Well, then all that we have in this case is a poor error message. 18:45
dukeleto colomon: yes, I agree. Failure + 2 fails, but not with the correct error 18:46
colomon Does that imply that +Failure is also wrong (because it returns 0 with a warning)?
dukeleto colomon: do you have any failing spectests on master at the moment? do you know what gist.github.com/574122 is about?
colomon I haven't seen those, but I haven't spectested since I checked in yesterday. 18:47
t/spec/S19-command-line/dash-e.t works fine for me. 18:49
errr... actually, I'm not quite up to date. five minutes, please...
dukeleto colomon: i worry that the things i am seeing only fail for rakudo on 64bit linux 18:50
colomon: which may be parrot issues, not rakudo bugs
colomon I can try that too, if you'd like. :)
18:51 mfollett joined
dukeleto colomon: it would help me to be sure that I am not crazy :) 18:51
colomon or prove that you are! ;)
PerlJam I think people that hang out on #perl6 are crazy by definition. 18:52
colomon PerlJam: hard to argue with that... 18:53
18:54 estrabd is now known as jhfieo099, jhfieo099 is now known as jhfie, jhfie is now known as estrabd, estrabd is now known as jhfieo099 18:55 jhfieo099 is now known as Ev1Lh4xr, Ev1Lh4xr is now known as estrabd
diakopter estrabd: ? 18:57
18:57 envi^home left
colomon dukeleto: very latest still works on 32-bit OS X 18:57
18:58 CoCol left
colomon dukeleto: 64-bit linux compile on-going 18:58
18:59 tylercurtis left
dukeleto colomon: cool, thanks for checking! 18:59
would anybody mind if I added *.rakudo to .gitignore in roast.git ? 19:00
pmichaud jnthn: Iterable works far better as a type, I think. I'm just not sure it should be Cool. 19:01
s/type/class/
I only stuck it in as Cool in June (while trying to explore the iterator/iterable details) because I figured it allow me to avoid some early blockers 19:02
waitaminnit
that can't be the issue anyway, come to think of it.
rakudo: say ~Cool.^methods 19:03
p6eval rakudo 7517e3: OUTPUT«acos fmt trans eval chop rindex tanh subst Numeric chomp index IO chr acotanh trim-trailing from-radians capitalize pred atanh exp cosec lc p5chop chars to-radians sinh acotan lcfirst roots lines sec trim asec trim-leading cotanh cis log sech log10 rand sprintf truncate bytes
..sqrt …
pmichaud why does Cool have a .pred/.succ ?
that doesn't seem right.
19:03 Mowah left
pmichaud augment class Cool { multi method succ() { (~self).succ } multi method pred() { (~self).pred } 19:03
}
I disagree. 19:04
pmichaud removes those and sees what happens.
19:07 tbalazs left
colomon dukeleto: I get a segmentation fault on ./perl6 t/spec/S19-command-line/dash-e.t on my 64-bit box 19:08
dukeleto colomon: i'm not crazy! 19:09
colomon so it looks like we do have some sort of issue.
dukeleto: your error can be duplicated. I'm not willing to make any further diagnosis of your mental health. ;)
dukeleto colomon: my backtrace is in the gist. looks like the parrot fast runcore is corrupting memory somehow
colomon: touche ;) 19:10
colomon: i submitted it as a rakudobug
colomon dukeleto: I think this is the point where we look to pmichaud...
dukeleto yes, pmichaud recently had to deal with some funny business regarding rakudo+parrot+64bit linux
colomon: #77778 19:13
19:14 Holy_Cow left
dukeleto colomon: i added my backtrace to the ticket 19:15
19:16 b0miva joined
colomon dukeleto++ 19:18
dukeleto colomon: i asked for help on parrot-dev as well 19:19
19:20 M_o_C joined
dukeleto looks like roast.git commits don't get announced in here, eh? 19:22
19:25 b0miva left 19:28 b0miva joined, M_o_C left 19:36 b0miva left, b0miva joined 19:40 plainhao left
x3nU i wonder, must be rakudo star installer MSI? nsis would be easier imho 19:42
19:44 masak joined
masak ahoy! 19:44
tadzik ohai!
masak sleep will pull me offline quite soon. but I thought I'd stop by :) 19:45
19:47 patspam left 19:49 patspam joined, Mowah joined 19:51 snearch joined
pmichaud x3nU: it depends solely on who creates the file. :) 19:51
masak in the newly spec'd 'dir' function: "Only items that smart-match against this test are returned." is it the file names that are smartmatched, or IO objects representing the files?
pmichaud x3nU: if you (or someone) wants to publish how-to-create-nsis or the nsis version, that'd be great. :) 19:52
19:58 lichtkind joined 19:59 Italian_Plumber left
dalek kudo: 43c27a9 | pmichaud++ | src/core/Cool.pm:
Remove .succ and .pred from Cool.
20:08
20:08 amkrankruleuen left
masak rakudo: for^5 {say "$_ is {all([\+]1,map {92-.&ord},.comb)??""!!"in"}valid"given(<[ ]>xx 5).pick(*).join} 20:12
p6eval rakudo 7517e3: OUTPUT«][[]][[][] is invalid␤[]][][][[] is invalid␤[]]]]][[[[ is invalid␤[[][[]]][] is valid␤]][[][[]][ is invalid␤»
masak that's almost too short :)
tadzik rakudo: my @list = 1, 2, 3; say .pred«@list # how do I do this correctly? 20:13
p6eval rakudo 7517e3: OUTPUT«===SORRY!===␤Confused at line 22, near "say .pred\x{ab}"␤»
masak tadzik: @list>>.pred
tadzik masak: well, I wanted the example of the other-side-» :)
(other than 2-arg operators)
masak tadzik: method call is a postfix operator.
tadzik oh, so » is for postfixes and « is for prefixes? 20:14
masak aye.
pmichaud rakudo: my @list = 1,2,3; .pred.say for @list;
p6eval rakudo 7517e3: OUTPUT«0␤1␤2␤»
20:15 aloha left 20:16 nadim left 20:17 [Coke] joined, bacek left
dalek kudo: ce565f3 | moritz++ | src/Perl6/Actions.pm:
enable :r and :ratchet outside of regexes
20:20
lichtkind is each official? 20:23
i dont see it in S03
masak lichtkind: it's in S09.
lichtkind thanks 20:24
masak lichtkind: and it's conjectural.
which I guess means that it's hesitatingly official.
hm, erez++ just pointed me to the original "enable Perl 6 knock offs\nuse 5.010;" code snippet: perlmonks.org/?node_id=859596 20:25
I can see how someone closer to Perl 5 than to Perl 6 might be upset by such a comment. it's not fair to Perl 5.10 to reduce it to being about "Perl 6 knock offs". 20:26
20:26 tadzik left
pmichaud does 'each' become official when yapsi implements it? ;-) 20:27
masak on the other hand, I don't believe in shooting people on sight, not even people who write thoughtless comments. twitter.com/erez/status/24090023075
20:27 jaffa4 joined
masak jaffa4! \o/ 20:28
jaffa4 hi
masak pmichaud: we'll just have to try it and see :)
pmichaud masak: maybe he meant shoot as in "with a camera" :) 20:29
in that context, I often shoot people on sight. :)
masak "You're cute. You ought to be shot on sight." hm.
there are probably worse pick-up lines, but I can't think of any right now. 20:30
jnthn I can think of worse ones. :-) 20:31
pmichaud "You had me at 'Go away.'" :-)
jnthn lol
masak that's a followup :P
pmichaud is it still a followup if it's the first thing you say? ;-)
masak ooh, interesting.
as a conversation starter, even? :) 20:32
"You had me at 'Go away.'" -- "What?"
jnthn It...could just work. :-)
But I'll let somebody else do the research. :P
masak I can think of worse ones. :)
pmichaud I know a guy who started the conversation with his future wife with "How am I doing so far?" 20:33
20:34 tadzik joined
masak Pretty well, apparently. 20:34
PerlJam masak: could be pmichaud left out some information. "... future ex-wife" ;-) 20:35
pmichaud I have no information on that.
20:36 jaffa4 left
masak jaffa4's exit message is intriguing, evocative, and wrong. :) 20:37
20:37 b0miva left
[particle] it becomes a movie called "Inception". 20:38
masak :D
20:38 b0m1ro joined
pmichaud yes, I was just thinking how truly frightening that message could be if true. :) 20:38
jnthn It's all the more scary given I can never remember my dreams after I wake up from them. :-)
masak pmichaud: it might be true, but "the same dream" never occurs since dreams have a unique object identity. :) 20:39
20:39 patrickas joined
jnthn struct Dream { ... } 20:39
;-)
masak but Occam called, and he wants his razor back.
jnthn I think I figured out how to implement most of the lexical multi-dispatch stuff efficiently today. 20:40
However, I don't trust myself to implement it correctly tonight. :-)
masak nice!
masak decides to go to sleep
see you tomorrah, #perl6!
jnthn decides to have a beer
night, masak
masak night. 20:41
20:41 masak left
[particle] the lesson of the day from jnthn is, "trust, but beerify." 20:44
i think reagan would approve.
pmichaud code.google.com/p/chromium/issues/d...l?id=54028 # a true bug report (might be old news here) 20:46
jnthn In Soviet Russia, multiples dispatch you.
"Perhaps the watch is a Microsoft product - have you tried taking it off and putting it back on again? 20:48
"
lol
pmichaud: I hadn't seen this one. :-)
:-)
pmichaud "Hate to bring it up, but Oracle holds patents..."
jnthn yes, just got to that one :-) 20:49
systemet++ # stocking lots of German beer to celebrate Oktoberfest 20:50
systemet-- # not at German prices
jnthn should actually to go that festival one day... 20:51
20:52 literal left 20:54 Italian_Plumber joined
dalek odel: 6516ba1 | jnthn++ | dotnet/runtime/Runtime/Lexpad.cs:
Mysteriously forgot to commit Lexpad.cs at some point in the past. :/
20:57
20:58 literal joined, whiteknight joined 21:01 Italian_Plumber left, literal_ joined, literal left 21:08 stepnem left 21:12 stepnem joined, mfollett left 21:14 patspam left
diakopter as an exercise in church camp one summer, we were given bibles to place in the campfire, ostensibly to distinguish Jesus as the self/Bible-proclaimed "Word of God" (in the flesh) from the much more abstract "Word of God" (used in the Old Testament), that referred to the direct and indirect verbal communications from God, including the scriptures. 21:17
tadzik are you referring to the Koran burning case? 21:18
diakopter obliquely
21:19 b0m1ro left 21:20 rainerschuster joined
sjohnson hehe 21:21
thanks for sharing that interesting fact
tadzik insightful indeed 21:22
21:22 mfollett joined
tadzik btw, anyone interested in a module installer working on windows? 21:23
jnthn Yes
tadzik jnthn: I'm thinking about making neutro work on windows, so it could possibly make it into R* 21:24
jnthn Nice
tadzik unfortunately, I have no windows anywhere and my virtualbox is dead for some time
also, I have very little experience with it
jnthn: so, to start, is 'make' a problem? 21:25
I heard 'no' some day, but I'm thinking about someone who has just installed R* on his/her fresh Windows machine 21:26
jnthn There's no "make" by default on Windows. There are various makes that run there, one being nmake
On a fresh machine, no.
tadzik alright, pls approach then
jnthn approaching
oh, wait, it wasn't a request...
:-)
tadzik :) how about module install path? It's a fixed path in R*, isn't it?
diakopter jnthn: probably best to start with the one that comes with strawberry perl - dmake
tadzik rakudo: say $*VM<CONFIG><OS> # this way? 21:27
p6eval rakudo ce565f: OUTPUT«Any()␤»
jnthn diakopter: Maybe, apart from I don't use that... :S
diakopter: And probably not anyone building with MS VC++
diakopter yeah but, I bet almost no rakudo user will also use visual studio
tadzik why not?
jnthn It's probably best to just not expect there to be a make program, imho. 21:28
diakopter of course. I was suggesting supplying dmake the same way strawberry perl does.
tadzik yeah, that's a good think imho
rakudo: say $*VM<config><osname> 21:29
p6eval rakudo ce565f: OUTPUT«linux␤» 21:30
21:30 Ross left
tadzik looks like I'm about to steal a part of pls 21:30
diakopter rakudo: say $*VM<config><>
p6eval rakudo ce565f: OUTPUT«12/home/p6eval/rakudo/parrot818"$(PERL) -MExtUtils::Command -e cp/home/p6eval//p2/share/doc181-lm -L/usr/lib -licuuc -licudata -lm__attribute__ ((noinline))%.15g.so$(PERL) -MExtUtils::Command -e rm_f/home/p6eval//p2/etcamd64definet/compilers/pct/*.t t/compilers/pge/*.t
..t/compilers…
diakopter erm
tadzik jnthn: so where should a module be installed to on windows?
oh crap, the git issue 21:32
well, not bad. Do git commands work as on unixes if git is installed?
diakopter tadzik: well... you'd have to pick a "standardized" location (to which to install modules)
jnthn tadzik: Yes
tadzik diakopter: I think R* alredy has one it puts the modules in
diakopter same as on unix, yes, if the person puts them in their path
tadzik diakopter: so I think it'd be a nice place 21:33
diakopter tadzik: ok then :) why'd you ask
tadzik as I said, I don't know much about windowses
back in the days I used it, I was mostly playing games
diakopter Also, the burning of an English bible was used to remind us how English bibles are a-dime-a-dozen, but in lots of other languages are much more precious/rare/nonexistent. 21:35
21:38 snearch left 21:44 patspam joined, patspam left 21:49 Ross joined
patrickas rakudo: say (1, 2 , {last if $_>=5; $_+1} ... *) 21:56
p6eval rakudo ce565f: OUTPUT«12345␤»
21:57 literal_ is now known as literal 22:01 b0m1ro joined 22:02 ruoso left 22:03 kjeldahl left 22:04 Ross left, rainerschuster left 22:08 bacek joined 22:09 ash_ joined 22:10 aloha joined
ash_ you know what would be nice, in terms of documentation anyway 22:10
php does this on their official site, just do php.net/some_function and you get linked directly to the docs for that function, it might be nice to setup something like that for rakudo or the spec 22:11
22:12 timbunce joined
ash_ like, rakudo.org/grep or docs.rakudo.org/hyper-oprators or something, i duno just thinking outloud 22:12
pmichaud ash_: Well volunteered, sir! :-) 22:13
tadzik :D 22:14
22:16 tadzik left 22:19 hercynium joined
ash_ maybe when try.rakudo.org is done :P 22:24
22:24 jaldhar joined
ash_ i'd like to be able to do: help grep and have that integrated into try.rakudo.org (as a command for the online terminal) 22:24
maybe perldoc grep, i haven't decided on syntax yet 22:25
doc grep? /shrug
jnthn "wtf is grep" :-) 22:29
22:30 mfollett left
patrickas pull request sent for series new spec patch, review / comments welcome :-) 22:30
pmichaud: I did eventually solve the laziness issue of the LHS you had commented about a couple of weeks ago. With this new patch we can finally have infinite lists on the lhs of the series op. 22:34
22:34 nadim joined 22:38 nimiezko joined 22:39 timbunce left
nimiezko hello 22:41
patrickas hello nimiezko 22:44
22:47 spq1 left 22:54 jaldhar left, jhuni joined
sjohnson grep! 23:17
grep and map, perl5's favourite cousins
23:17 patrickas left 23:21 meppl left 23:28 frooh left, frew joined
sorear good * #perl6 23:28
sorear is struggling to implement rule a { b { $<X> = 1; } c } and rule a { <.b> }; method b($¢) { $¢.<X> = 1; $¢ } 23:37
both patterns are used in STD, a lot
erm. method b { self.<X> = 1; self } 23:38
23:40 justatheory left, justatheory joined 23:43 justatheory left 23:54 mberends left 23:56 cdarroch left