»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'p6: say 3;' or rakudo:, std:, or /msg camelia p6: ... | irclog: irc.perl6.org | UTF-8 is our friend! | feather will shut down permanently on 2015-03-31
Set by jnthn on 28 February 2015.
timotimo "now"? 00:03
00:04 koo6 left
andreoss how do i overload .gist method properly? 00:05
or whatever method needed to stringify class 00:06
timotimo what problem are you encountering? it should be enough to define a "method gist"
well, we have .Str, .gist and .perl
they differ in which thing will choose which
m: class Foo { method gist() { say "gist" }; method Str() { say "Str" }; method perl() { say "perl" } }; say Foo.new; print Foo.new;
camelia rakudo-moar 536c26: OUTPUT«gist␤True␤Str␤This type cannot unbox to a native string␤ in block <unit> at /tmp/LWoT1QxY8e:1␤␤»
timotimo oh my 00:07
m: print True
camelia rakudo-moar 536c26: OUTPUT«True»
timotimo m: print say "test"
camelia rakudo-moar 536c26: OUTPUT«test␤True»
timotimo that's ... weird
raydiak it's weird that say returns true? 00:08
timotimo no
andreoss may be my class needs to do Stringy?
timotimo the "cannot unbox to a native string" thing
andreoss or it would be overkill?
raydiak really ought to start pressing pageup more often or chatting in a taller window
timotimo probably not, but you can implement a method "Stringy"
zengargoyle is it weird that you're say ing instead of returning a string from the methods?
andreoss timotimo: when such method will be used?
timotimo good question :S
in coercing stuff, i suppose 00:09
00:11 laouji joined
raydiak I'd bet, like zengargoyle said, something internal expects/requires .Str to return a string 00:13
timotimo some internal things expect .Stringy returns something that does .Stringy 00:14
Sysaxed` so, I've heard that perl6 will have some unicode operators by default, any reason why ≤ and ≥ are not supported?
timotimo good question
m: say 1 ≤ 5
camelia rakudo-moar 536c26: OUTPUT«5===SORRY!5=== Error while compiling /tmp/7HnrEmEOMB␤Two terms in a row␤at /tmp/7HnrEmEOMB:1␤------> 3say 17⏏5 ≤ 5␤ expecting any of:␤ infix␤ infix stopper␤ postfix␤ statement end␤ stateme…»
timotimo m: my &infix:<≤> := &infix:«<»; say 4 ≤ 5 00:15
Sysaxed` timotimo: :)
camelia rakudo-moar 536c26: OUTPUT«True␤»
00:15 laouji left
Sysaxed` yea, sure, I can define it myself if I really want 00:15
timotimo right
that'd be something to put into the speculations :)
TimToady well, where to draw the line...
Sysaxed` but why not default?
00:15 spookah left
timotimo ————— 00:15
dalek osystem: 6e72fb5 | cygx++ | META.list:
Add DEBUG to the ecosystem
00:17
BenGoldberg m: class Foo { method Str() { True } }; say Foo.new;
camelia rakudo-moar 536c26: OUTPUT«Foo.new␤»
BenGoldberg m: class Foo { method Str() { True } }; print Foo.new;
camelia rakudo-moar 536c26: OUTPUT«This type cannot unbox to a native string␤ in block <unit> at /tmp/M9MpgZFVFi:1␤␤»
TimToady that's several lines, you really want ───── instead
timotimo right
00:18 agentzh_ left
raydiak why is there a line here? it's not like we've made any attempt to avoid using up too many different characters :) 00:18
00:18 meisl left
raydiak the grammar changes would be pretty simple, no? 00:19
00:19 laouji joined
timotimo yes 00:19
geekosaur doesn't even need a grammar change, just a setting change, I think? the question is basically how many things qualify as belonging in the core
00:19 thou left
BenGoldberg m: my &infix:<≤> := &infix:«<»; say 4 ≤ 4; 00:19
camelia rakudo-moar 536c26: OUTPUT«False␤»
TimToady yes, it's really just a decision about where to shout "bloat!" 00:20
BenGoldberg m: my &infix:<≤> := &infix:«<=»; say 4 ≤ 4;
camelia rakudo-moar 536c26: OUTPUT«True␤»
timotimo BenGoldberg: yeah, a somewhat important detail
BenGoldberg Would it slow things down too much have some sort of autoloading operators? 00:21
TimToady we added the quotes because people were starting to get interference from their clients smart-quoting things
in this case lazy loading wouldn't buy you much, since the symbol is about the only new thing, and you have to remember that to autoload it... 00:22
unless we had a general I-see-strange-Unicode lazy loader
Sysaxed` also I don't really see any obvious reason for begginers to understand why >= must be written this way rather than =>, and some people might type => when trying to reverse <=, hehe! Why not just create a possibility to throw it all away by allowing ≤ and ≥
TimToady well, they *do* have to learn what => means... 00:23
Sysaxed` who knows
TimToady unless you also give 'em ⇒ of course :) 00:24
Sysaxed` TimToady: if you're a guy who wants to crunch some numbers quickly, possibly without significant background...
I don't know, maybe I'm exaggerating 00:25
but when thinking "unicode operators" ≤ is the first thing that comes to my mind
geekosaur the flip side is most people still don't have fancy editors that will let them easily type ≥
Sysaxed` and ≥ is the second...
geekosaur and if you say "it should recognize >=" then all you've done is push your question into the editor instead of the language --- and the editor is probably worse equipped 00:26
Sysaxed` geekosaur: it is possible that some editors will attempt to replace stuff with unicode equivalents, giving instant feedback
geekosaur: so are you saying that we should have no unicode operators at all? 00:27
geekosaur (the language has better understanding of syntax and has semantic knowledge the editor cannot)
TimToady well, arguably ≠ is the 0th one
Sysaxed` geekosaur: not by default, at least
geekosaur you seem to be arguing that we should prioritize the unicode operators
Sysaxed` geekosaur: not at all 00:28
geekosaur i.e. it's ">=" that should be the alien
[11 20:22:48] <Sysaxed`> also I don't really see any obvious reason for begginers to understand why >= must be written this way rather than =>, and some people might type => when trying to reverse <=, hehe! Why not just create a possibility to throw it all away by allowing ≤ and ≥
this seems to imply that these operators are problematic and should not be considered primary
Sysaxed` hm
TimToady well, for sure C and math have a profound difference over the meaning of = 00:29
Sysaxed` no, I'm saying that an alternative could be there by default
well, it could be there, I'm not sure
geekosaur use UnicodeSyntax;
then it can also grow without requiring a perl6 release 00:30
timotimo the opposite of "use english" 00:31
dalek kudo/nom: a5496e0 | TimToady++ | src/core/Failure.pm:
plug a major leak of unhandled Failures
TimToady that fixes the part of forest-fire that reports unhandled Failures 00:32
timotimo oh, neat
geekosaur once it stabilizes, it can be considered for inclusion in the core
timotimo did you check with rc-forest-fire from the perl6-bench repo?
because i recall that failing during my benchmark
TimToady no, I have my own copy
but that also has the (), problem in the middle, last I looked 00:33
timotimo OK
zengargoyle TimToady: speaking of funny characters, I saw a presentation where you were playing with a keyword -> kanji sort of thing. is that code that can be seen online somewhere or a personal toy?
raydiak I'm still thinking whether those specific operators are dubbed worthy of inclusion or not, if this is a language for the "next 100 years" or whatever it was, we will sooner or later *need* to be able to grow CORE without this level of restraint and consideration required for even the tiniest, most un-bloated of additions
timotimo in 100 years, ram will be cheaper still that it is right now :P 00:34
dalek rl6-bench: 2f432a8 | TimToady++ | perl6/rc-forest-fire:
suppress Use of Nil in numeric context
geekosaur I think it's entirely sensible to handle it by "starting" stuff in the ecosystem and transplanting into the core 00:35
it works for plants...
TimToady zengargoyle: still a toy, alas, but I should maybe productize it for YAPC::Asia or so...
raydiak yeah that's fine, it's the problems that need to be considered when including in core that I'm worried about
TimToady raydiak: well, the whole point of making the core an outer lexical scope is that inner scopes can define things without worrying about core overriding them later 00:36
zengargoyle i'm thinking of making a similar thing in p5 for myself.
TimToady zengargoyle: well, it's written in p5 currently
but it's a bit of a mess
zengargoyle guessing kanjidic + kradfile + radkfile based? 00:37
maybe i'll just wait and let laziness provide. :) 00:38
TimToady well, the CJK database I typed in all by myself, though there are some auxilliary definitions from elsewhere for Japanese and Chaines
Chinese
raydiak TimToady: yes that's a perfectly valid solution to many instances of this problem, but you don't think CORE growth won't eventually be a problem if we don't do some sort of lazy loading or whatever? 00:39
00:39 laouji left
TimToady as I said, laziness doesn't really help if you want to parse the symbols in order merely to install an alias 00:39
aliases don't really cost anything
it helps a lot more when there's a body that can be lazily loaded 00:40
if we could get the whole thing down to 512MB, then maybe we could run on a CHIP processor :) 00:41
timotimo TimToady: is there (or will there be) easily accessible recordings/data/whatever for the weekly blog readers to ingest from your appearance at collision? 00:42
pmichaud timotimo: I suspect that depends on the collision folks. 00:43
timotimo i know there's that thing with the app, but i don't really know how to talk to the readers about that
at osdc.no there was the hands-on perl6 with jnthn, iiuc 00:44
TimToady they certainly had a very impressive camera pointing at the stage, and great gobs of equipment off to the side
pmichaud yes, it went VERY well
(hands-on perl6 with jnthn)
timotimo mhm
raydiak TimToady: not so worried about the semantics of the solution ("lazy *or whatever*")...was more wondering if you see the same problem in the distant future that I do, or if I'm just inventing problems :)
timotimo ah, i found the "perl6" tag on tha tpage
that'll be a good link target
pmichaud I've started inquiring with the yapc::na folks if there would be a way to squeeze in a reprise of that session in SLC
(I'd be glad to run it.) 00:45
geekosaur remember that hardware advances pretty rapidly 00:46
pmichaud hmmm... my laptop computer has advanced hardly at all since I bought it. :-P
00:46 spider-mario left
pmichaud (surprisingly, I bought this laptop one day before my first trip to Oslo... which is where I am now. I wonder if Oslo will be this laptop's last trip as well.) 00:47
00:49 laouji joined
timotimo TimToady: so was it you gave a talk at collision conf and then did a q&a with that twitter-related app? 00:49
TimToady yes, though that turned out kinda silly
timotimo were there many people who tried to make fun of the perl6 project? :) 00:50
TimToady nope
pmichaud most of what I've encountered since FOSDEM is people with growing respect for it.
Sysaxed` pmichaud: I'm not sure that it is the effect of FOSDEM. Perl6 has advanced significantly during the last year, at least that's how it feels 00:51
TimToady the most "make fun" was from a well-known Perl 5er, who asked if it really coming out this Christmas, to which the only appropriate response was, "Yep, this Christmas. Ho, ho, ho!"
00:52 muraiki joined
pmichaud it was fun this last Friday to be able to challenge Simon Phipps' claim about Hurd being released before Perl 6. :) 00:52
geekosaur I do know that one, for years my travel laptop was an iBook. its hard drive died a couple years ago
pmichaud (during our panel session at OSDC.no)
geekosaur (iBooks were infamous for needing to be torn down completely to replace the HD)
pmichaud it's very much time for me to get a new traveling computer. I've been considering a hybrid laptop/tablet. 00:53
TimToady well, hopefully, even though DNF beat us out the door, we'll do better long term than they did
tony-o pmichaud: i use a surface pro 3 with debian/windows dual booted
zengargoyle how is the surface pro 3 with debian? 00:54
pmichaud tony-o: that's excellent to know! I like the surface pro 3 but wanted to make sure I'd be able to get Linux working on it
timotimo at the GPW, perl6ers helped people who wanted to get started with perl6, too, right?
tony-o i actually love the sp3 with debian on it
i use it when i travel for work, it's compact and plenty powerful for work stuff (i do data warehouse consulting) 00:55
raydiak geekosaur: I dunno, not as impressed with raw performance hardware advancements these days, especially compared to perl 6's growth (I hope)...but hopefully my estimation is wrong, or it'll speed back up, or we'll slow down on features or something; guess I'll get back to more immediate concerns since nobody else seems worried :) 00:56
geekosaur I don't expect perl6's core to grow so fast that it will be a short term issue
in general, you want the core to be small
tony-o pmichaud: you can definitely get linux on there, it needs to be a flavor that can do secure boot uefi
pmichaud I typically do kubuntu 00:57
tony-o so your '98 version of mandrake is probably not going to work :-) - i have debian jessie + xfce
pmichaud (kubuntu is debian based, of course)
tony-o kubuntu is just ubuntu with kde as the default DE ? 00:58
zengargoyle did jessie install support uefi without pain?
pmichaud tony-o: yes. 00:59
00:59 BenGoldberg left 01:00 BenGoldberg joined, spider-mario joined
tony-o yea - i originally tried stable but had to recompile the kernel with a bunch of drivers, jessie installed well the first time - the wifi drivers had to be installed separately but it was pretty painless 01:00
01:00 BenGoldberg left 01:01 BenGoldberg joined
raydiak I just worry when we're talking about whether or not to include something so small and simple, makes me wonder if there isn't an unreasonable amount of resistance from whatever the performance issue is...but I guess since you want a small core, that resistance is probably good anyway 01:01
01:03 aborazmeh joined, aborazmeh left, aborazmeh joined 01:04 b2gills left, colomon left
pmichaud adopting things into the core should have some amount of resistance and reflection, yes. 01:05
geekosaur raydiak, in this case it's more a matter of figuring out what we want and then doing it all at once instead of one-line additions every 3 days, I think 01:07
01:07 colomon joined
raydiak that makes sense, that'd make a mess in short order, I do that all the time :) 01:07
in case it wasn't obvious since nobody here actually knows me, when I go on these tangents, I'm just trying to understand, and to reconcile what I'm learning with what I thought I knew before...sometimes people think I like to argue or something which is absolutely not my intent (except when it is) 01:09
01:11 Sysaxed` left
timotimo were jnthn's talks updated a lot from the last time he gave them? 01:12
pmichaud timotimo: not that I noticed.
timotimo OK 01:13
sadly, the pages don't have slides links ;(
pmichaud I think he just hasn't uploaded/linked them yet.
jdv79 is there a way to introspect applied roles?
on a class or instance for example 01:14
pmichaud jdv79: do you mean as in "give me a list of roles applied to this class/instance?"
jdv79 yes
pmichaud one can certainly do $foo ~~ Role :)
(that doesn't give a list, but it lets you know if Role has been applied somewhere.) 01:15
raydiak m: Str.^roles 01:16
camelia ( no output )
raydiak m: say Str.^roles
camelia rakudo-moar a5496e: OUTPUT«(Stringy)␤»
pmichaud m: say List.^roles
camelia rakudo-moar a5496e: OUTPUT«(Positional)␤»
pmichaud m: say Hash.^roles
camelia rakudo-moar a5496e: OUTPUT«(Associative)␤»
timotimo m: say (Hash but Str).^roles
camelia rakudo-moar a5496e: OUTPUT«Cannot mix in non-composable type Str into object of type Hash␤ in block <unit> at /tmp/68hm4YY_xv:1␤␤»
jdv79 oh, duh. should have just tried that. i was searching docs instead:(
timotimo m: say (Hash but IO::Socket::INET).^roles 01:17
camelia rakudo-moar a5496e: OUTPUT«Cannot mix in non-composable type IO::Socket::INET into object of type Hash␤ in block <unit> at /tmp/VOPbnZJ08C:1␤␤»
timotimo oh
those are not roles m)
raydiak m: say Buf.^roles
camelia rakudo-moar a5496e: OUTPUT«Method 'roles' not found for invocant of class 'Perl6::Metamodel::ParametricRoleGroupHOW'␤ in block <unit> at /tmp/uvtM0AI4rI:1␤␤»
timotimo it seems my brain is trying to fall asleep already
raydiak heh, sure, you can get the list as long as it's only one item :P
pmichaud say FlatSupply.^roles 01:18
m: say FlatSupply.^roles
camelia rakudo-moar a5496e: OUTPUT«5===SORRY!5=== Error while compiling /tmp/0Son8xVEGd␤Undeclared name:␤ FlatSupply used at line 1␤␤»
01:18 andreoss left, araujo left, oetiker joined 01:21 araujo joined, araujo left, araujo joined, yqt left 01:23 rba_ left, rba_ joined 01:25 airdisa left 01:30 andreoss joined
andreoss fpaste.org/220855/14313942/ 01:31
what's wrong with this code?
it segfaults for me
TimToady try changing to 'has Block $.code;' 01:32
by default you can't initialize private attributes through the public interface 01:33
geekosaur probably shouldn't segfault in any case
TimToady you can still call it with $!code
though that should probably be $!code(|@a) 01:34
or method exec(|a) { $!code(|a) } for the most generality 01:35
'course if it's a public attribute, you can just say $f.code()(1,2,3) too 01:36
I agree it shouldn't segfault
01:38 dayangkun joined 01:39 aborazmeh left
TimToady we also need to find a better way of warning that the code arg isn't used 01:40
andreoss i also receive errors about &!code, even if &.code is used 01:42
fpaste.org/220861/39513014/ 01:45
why the dereferencing is needed here? 01:46
probably wrong choice of sigil 01:48
m: my &x = -> $a {$a.perl}; say &(&x)(1); 01:49
camelia rakudo-moar a5496e: OUTPUT«1␤»
tony-o gist.github.com/tony-o/2b309e70ed16867e0849
andreoss: ^
m: $x = -> $a { $a.say; }; $x(5); 01:51
camelia rakudo-moar a5496e: OUTPUT«5===SORRY!5=== Error while compiling /tmp/cBaEeaxX2D␤Variable '$x' is not declared␤at /tmp/cBaEeaxX2D:1␤------> 3<BOL>7⏏5$x = -> $a { $a.say; }; $x(5);␤»
tony-o m: my $x = -> $a { $a.say; }; $x(5);
camelia rakudo-moar a5496e: OUTPUT«5␤»
dalek ast: e476c69 | (Justin DeVuyst)++ | S06-currying/assuming-and-mmd.t:
Tests for RT #77520 & RT #125155.
synbot6 Link: rt.perl.org/rt3/Public/Bug/Display...l?id=77520
TimToady you can say &!code()(1,2,3) as well
the first one is just because &!code and &.code are really method syntax, so the first () is passing zero args to that 01:52
(&!code)(1,2,3) would also work
timotimo "ignoremark" support is in branches right now, right?
tony-o m: class A { has $.a; method go(*@a) { &$.a(@); }; }; A.new(:a(-> *@R { @R.perl.say; })).go(1,2,3); 01:53
camelia rakudo-moar a5496e: OUTPUT«Too many positionals passed; expected 1 argument but got 2␤ in method go at /tmp/1mmGqQugIm:1␤ in block <unit> at /tmp/1mmGqQugIm:1␤␤»
andreoss $.code.(...) what does the second dot do?
tony-o m: class A { has $.a; method go(*@a) { $.a.(@); }; }; A.new(:a(-> *@R { @R.perl.say; })).go(1,2,3);
camelia rakudo-moar a5496e: OUTPUT«[]<>␤»
tony-o m: class A { has $.a; method go(*@a) { $.a.(@a); }; }; A.new(:a(-> *@R { @R.perl.say; })).go(1,2,3);
camelia rakudo-moar a5496e: OUTPUT«[1, 2, 3]<>␤»
TimToady oh, that's just another way to defeat the $.foo() syntax
tony-o yea
andreoss m: my $x = -> $a {$a.perl}; say $x(1); #no dot here 01:54
camelia rakudo-moar a5496e: OUTPUT«1␤»
andreoss m: my $x = -> $a {$a.perl}; say $x.(1); #
camelia rakudo-moar a5496e: OUTPUT«1␤»
tony-o it's telling it that $.a() isn't a call to a self.method() and it's basically a nicer looking &($.code)(args)
TimToady $.foo() is a special form, but $foo() isn't
tony-o m: class A { has $.a; method go(*@a) { &($.a)(@a); }; }; A.new(:a(-> *@R { @R.perl.say; })).go(1,2,3); 01:55
camelia rakudo-moar a5496e: OUTPUT«[1, 2, 3]<>␤»
TimToady most of the time people just want a virtual $.foo and don't really think about the fact that it's calling an accessor
but you've found the WAT corresponding to the DWIM :)
andreoss $.foo.bar() ~~ self.bar($.foo) ? 01:56
jdv79 isn't #105848 a dup of #77350?
synbot6 Link: rt.perl.org/rt3/Public/Bug/Display...?id=105848
jdv79 only one link? :( 01:57
timotimo heads home 01:58
tony-o $.foo() means call self.foo() so doing $.code(<something>) is trying to call this.code(<something>) where code is a method of the class instead of the callable you stored in code
andreoss ok, now i get it
thanks 01:59
tony-o m: class A { has $.a; method go(*@a) { $.a(@a); }; method a(*@a) { "here".say; $.a.(@a); }; A.new(:a(-> *@R { @R.perl.say; })).go(1,2,3);
camelia rakudo-moar a5496e: OUTPUT«5===SORRY!5=== Error while compiling /tmp/ZmJrK_vPeF␤Missing block␤at /tmp/ZmJrK_vPeF:1␤------> 3(:a(-> *@R { @R.perl.say; })).go(1,2,3);7⏏5<EOL>␤»
tony-o m: class A { has $.a; method go(*@a) { $.a(@a); }; method a(*@a) { "here".say; $.a.(@a); }; }; A.new(:a(-> *@R { @R.perl.say; })).go(1,2,3);
camelia rakudo-moar a5496e: OUTPUT«(timeout)here␤here␤here␤here␤here␤here␤here␤here␤here␤here␤here␤here␤here␤here␤here␤here␤here␤here␤here␤here␤here␤here␤here␤here␤here␤here␤here␤here␤here␤here␤here␤here␤here␤here␤here␤he…»
andreoss it explains why it complained about extra arguments
02:00 laouji_ joined
tony-o m: class A { has $!a; method go(*@a) { $!a(@a); }; }; A.new(:a(-> *@R { @R.perl.say; })).go(1,2,3); 02:00
camelia rakudo-moar a5496e: OUTPUT«Cannot find method 'Any'␤ in method go at /tmp/PSOmB_z98f:1␤ in block <unit> at /tmp/PSOmB_z98f:1␤␤»
02:04 laouji_ left 02:08 noganex_ joined 02:29 rmgk left 02:30 rmgk joined 02:33 muraiki left
timotimo published the weekly before he went home 02:40
raydiak look forward to reading it the next time I take a break :) timotimo++ 02:43
timotimo bah. i've had it with gnome3 02:46
02:46 BenGoldberg left
timotimo i wanted to give it a chance. i really did. but it's doing some things that are just unacceptable 02:46
02:50 BenGoldberg joined
tony-o i used fluxbox up until trying xfce a couple of months ago 02:55
raydiak used to love fluxbox, should give it another go, haven't in over half a decade...what are your deal-breakers with gnome 3 timotimo? 02:56
timotimo i have xfce4 on my laptop and used to have xfce4 on my desktop before i re-installed fedora
sometimes it will not let any mouse clicks or keyboard events through to applications unless i alt-tab around a bit
raydiak ew
timotimo the first thing i thought was that i OOM'd somehow
which would be consistent with the way perl6-bench is behaving right now in that terminal i had just sent a command to 02:57
.tell japhb i'm experiencing an extreme delay in both the compare and history subcommands of perl6-bench right now
yoleaux timotimo: I'll pass your message to japhb.
raydiak your input devices run on mana? :)
heh sorry couldn't help it, talking about fluxbox reminded me of my WoW days too 02:58
timotimo .tell japhb perl with the analyze command is now using 720 mb res, which i gather is more than usual? it's at 6 minutes cpu time now for 10 checkouts of nqp-moar
yoleaux timotimo: I'll pass your message to japhb.
timotimo ah, "out of mana", now i get it :) 02:59
03:00 kyclark joined
raydiak fits better than I realized...when it occurs, you can keep clicking but nothing happens 03:01
timotimo .tell japhb on the other hand, it may have something to do with the fact that all those timing files are about 140 megabytes in size each
yoleaux timotimo: I'll pass your message to japhb.
timotimo hehehe
spawn more overlords!
raydiak but my hive will have bled to death by the time another overlord spawns! 03:02
timotimo that doesn't seem plausible to me 03:03
buildings bleed to death if they're off creep, but hive generate their own creep
raydiak my starcraft days are longer ago than my WoW days
timotimo OK :) 03:04
i've played starcraft 2 on ranked 2v2 ladder with a friend for a while
raydiak I can be forgiven if the last time I played nobody had heard of SC2 yet? :)
timotimo but i never played the original
of course 03:05
raydiak didn't play against people much, I was kinda terrible at strategy games, didn't understand anything about balance until the last few years
timotimo i didn't play very well, either 03:06
but i do have a steady interest in Zero-K
raydiak I've seen you mention it a few times, thought I might try it out the next time I get a random inclination to be a gamer for a few days 03:07
timotimo it's not necessarily very easy to get into. there's a youtuber by the username of "shadowfury333" who casts matches regularly and he's got a few tutorial videos up as well 03:08
incidentally, he also casts achron matches, which is a thing i expect perl6 people to appreciate - at least the concept
raydiak what is it? 03:09
timotimo "meta-time strategy game"
basically you're a time-traveling AI and you can give your units orders in the future, past or present 03:10
raydiak oh that is awesome
timotimo but so's the enemy - at least in player-vs-player matches and a part of the story missions
it has its problems, though
raydiak time travel always does 03:11
timotimo no, in this case it's unit stupidity
the time traveling is pulled off pretty much flawlessly
there's a campaign mission that involves narrow corridors and the units i had were so ridiculously stupid ... they stepped on each others toes and through each "incarnation" of the time between me telling them to go somewhere and them reaching it, the order in which the units arrived was completely different, if they arrived at nearly the same time at all 03:12
incidentally, rakudo-moar's historical scores show that it went from 100 points at 2014-08-22 to 184.8 at 2015-05-11
that seems nice 03:13
but it was at 184.3 already at 2015-03-20
raydiak link?
03:14 laouji left
timotimo it'll take some time to generate an html report 03:14
raydiak ah, n/m
timotimo i want the html report anyway 03:15
03:17 Humbedooh left, danstoner left, profan left, danstoner joined
raydiak has thought we really need some kind of perl 6 programming game, like one of those ones where you write an AI to run in a VM against other programs, or control battling robots, or something 03:18
03:18 tadzik joined
timotimo t.h8.lv/p6bench/2015-05-12-rakudo_history.html 03:20
may i suggest "rocket, paper, spacegoo"? 03:21
03:21 laouji joined, b2gills joined
timotimo bitbucket.org/dividuum/rocket-scissor-spacegoo 03:21
raydiak interesting 03:24
timotimo there's a nice web frontend that animates fights between AIs
this is basically what i wrote JSON::Fast for :P 03:25
raydiak hehe
timotimo seriously, the speed of my bot was ridiculously bad 03:26
and json parsing was indeed the biggest chunk of run-time by far
i really wonder what made string-escape take such a drastic penalty 03:27
and rat_harmonic ...?!
.o( and for_push ?? )
.o( and while_int2str_native? )
er 03:28
.o( and while_int2str_concat_native? )
raydiak insertion-sort is the cute one, if you ask me 03:29
timotimo yeah, well ... 03:30
t.h8.lv/p6bench/2015-05-12-rakudo_c...rison.html
if you can make out any details in there, congratulations :P
03:30 kyclark left
timotimo i shall activate my swap partition for the next steps ... 03:31
raydiak wonders if we could change the palette so newer perls are progressively brighter or something 03:33
timotimo go ahead :)
javascript basically means "source code included"
raydiak though what I'd really like to see is a 3D surface so we have an actual extra exis to play with :)
timotimo it'd even be possible to put a palette selector in
but metadata about versions may be missing that'd let you sort them by date reliably 03:34
raydiak yeah that JS especially, I already had to tear into the guts of the plotting lib to get done some of the work I did before
s/exis/axis/
timotimo okay, another 8gb of swap should be enough for this task :P
and i'm very thankful for your work, don't forget about that :)
raydiak thanks :) I don't forget, but it's always nice to hear :) 03:35
timotimo are you aware of a library that'd give us 3d surface rendering?
or should we consider using gnuplot?
is there already some kind of "gnuplot killer" out there?
matplotlib doesn't really fit the exact same niche 03:36
D3 is allegedly really awesome, but also not 100% the same kind of thing
oh, i really ought to get some shuteye 03:37
talk to you again soon :)
raydiak nah tbh I don't know much about what's going on these days
g'night, sleep well, thanks again for the weekly :)
timotimo that's why we need the 24-year-old javascript hipster demographic on our side ;) 03:38
you're welcome :)
03:39 agentzh_ joined
raydiak heh I'm often mistaken for a 24-year-old, maybe I should study up and see if I can pass enough as a hipster to bring the light of Perl 6 to a few local college guys :) 03:44
ugexe .tell lizmat github.com/rakudo/rakudo/blob/nom/...it.pm#L138 If you don't pass in $out, this will always be true. If you do pass in $out and $out !~~ $?FILE.moarvm (like somewhere in blib/lib), then has-precomp will always be false (even if it precompiled) and .precomp-path will give you the wrong result. Should method precomp-path be an attribute instead and then have a $.precomp-path = $
yoleaux ugexe: I'll pass your message to lizmat.
ugexe out somewhere in there?
03:45 BenGoldberg left, agentzh_ left 03:46 laouji left 04:01 agentzh_ joined 04:10 agentzh_ left 04:11 skids left 04:16 davido_ joined 04:25 Alina-malina left 04:26 laouji joined 04:28 Alina-malina joined 04:40 berekuk_ joined 04:49 laouji_ joined 04:50 kaare__ joined 04:54 laouji_ left 04:55 thou joined 04:56 quester joined 05:00 ][Sno][ left 05:03 agentzh_ joined 05:09 silug left 05:14 fhelmberger joined 05:15 telex left 05:16 telex joined 05:19 fhelmberger left 05:22 silug joined 05:23 andreoss left 05:24 laouji left, rindolf joined 05:27 andreoss joined 05:32 raiph left 05:34 FROGGS[mobile] joined
FROGGS[mobile] timotimo: ignoremark is in master/nom 05:35
05:35 rba_ left 05:37 laouji joined 05:39 lolisa left 05:42 [Sno] joined 05:43 domidumont joined 05:47 domidumont left 05:48 domidumont joined 05:49 andreoss` joined 05:51 andreoss left 05:55 domidumont left, andreoss` left 06:04 domidumont joined
[Tux] tony-o, au contraire. it slowed down :( 06:06
wild guess: about 5% 06:07
06:07 VAAC108 joined 06:09 espadrine_ joined 06:25 thou left 06:30 bjz joined 06:33 _mg_ joined, davido__ joined 06:34 rindolf left 06:36 berekuk_ left 06:37 davido_ left 06:38 laouji_ joined
grondilu I rewrote rosettacode.org/wiki/Mandelbrot_set#Perl_6 to be parallel (taking inspiration from github.com/colomon/mandelbrot/blob...allel.pl). It works but it's unconsistant from one run to an other (sometimes one line is missing). 06:40