»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!
Set by masak on 12 May 2015.
RabidGravy I can only barely read them to be honest 00:00
uruwi why is there a ↉
00:00 FROGGS joined
AlexDaniel m: say ⅟ 00:00
camelia rakudo-moar : OUTPUT«1␤»
Hotkeys Yeah at like 12pt Ican barely read the fractions
with my laptop on my lap
AlexDaniel Time to get glasses? :) 00:01
Hotkeys sorry 10pt
consolas
AlexDaniel actually, I think that some of these fractions were added earlier than others: ½
uruwi puu.sh/lyPO1/5800660bd2.png using IceChat 9 and setting the font to VL Gothic
Hotkeys because I couldn't be bothered to get another font
that text looks strange 00:02
AlexDaniel unifont?
uruwi ^
though it's bitmap
Not the best if you can't handle aliasing
Hotkeys I do like smooth looking type 00:03
00:03 rigsby left
hahainternet the nice benefit of having these supported in p6 now 00:03
AlexDaniel Hotkeys: you might want to have it installed anyway
hahainternet is that no 'coding font' maker has any excuse anymore :)
00:03 rigsby joined, leont left
Hotkeys I'm pretty sure I've tried having a just-in-case fallback font on windows for strange unicode 00:03
and hexchat refuses to use it
AlexDaniel so that you have a fallback for some rare characters :)
uruwi 미 00:04
Hotkeys are there any better windows irc clients
uruwi I personally use IceChat
00:05 RabidGravy left
uruwi Seems to display that Hangul character fine, despite not having it in the default font 00:05
미마
Hotkeys icechat has windows in the window
what is this nonsense
Zoffix m: say 2³²
camelia rakudo-moar : OUTPUT«4294967296␤»
Zoffix w00t!
AlexDaniel m: say 2³2 00:06
camelia rakudo-moar : OUTPUT«5===SORRY!5=== Error while compiling /tmp/QAFEHhKsjS␤Two terms in a row␤at /tmp/QAFEHhKsjS:1␤------> 3say 2³7⏏052␤ expecting any of:␤ infix␤ infix stopper␤ postfix␤ statement end␤ statement modif…»
AlexDaniel m: say ²
camelia rakudo-moar : OUTPUT«2␤»
AlexDaniel m: say ³²
camelia rakudo-moar : OUTPUT«9␤»
AlexDaniel well…
uruwi Well, well, well. 00:07
AlexDaniel I don't even know what should happen in such case…
Hotkeys I mean without a number to exponentiate it's just a number on its own
uruwi Syntax error.
It'll cause confusion if they're treated like ordinary digits. 00:08
AlexDaniel or perhaps Perl 6 Obfuscated Code Competition should happen!
Hotkeys lol
using only subscript ints
uruwi So I propose reserving those superscript ints for exponentiation
psch m: say postfix:<ⁿ>(Mu, 3)
camelia rakudo-moar : OUTPUT«Cannot call postfix:<ⁿ>(Mu, Int); none of these signatures match:␤ (\a, \b)␤ in block <unit> at /tmp/CjuTah_RuJ:1␤␤»
psch m: say postfix:<ⁿ>(Any, 3)
camelia rakudo-moar : OUTPUT«Use of uninitialized value of type Any in numeric context in block <unit> at /tmp/yyBN7RCOEZ:1␤0␤»
dalek osystem: bdfb17e | (Zoffix Znet)++ | META.list:
Remove Operators::Math::Superscripts [obsoleted by addition to core]
Hotkeys heh 00:09
00:09 Upasaka left
Zoffix :) 00:09
Hotkeys m: say ²³²
camelia rakudo-moar : OUTPUT«4294967296␤»
AlexDaniel well, that's probably LHF but I'll still submit it
uruwi m: say ²⁺³
camelia rakudo-moar : OUTPUT«8␤»
Zoffix m: say ²³²²³²²³²²³²
camelia rakudo-moar : OUTPUT«0␤»
Zoffix m: say 2*(²⁺³)
camelia rakudo-moar : OUTPUT«16␤»
Zoffix
.oO( I wonder if there are superscript brackets )
 00:10
psch not sure that's particularly LHF...
Hotkeys ⁽⁾
yep
uruwi m: say ⁽²⁾
camelia rakudo-moar : OUTPUT«5===SORRY!5===␤Argument to "say" seems to be malformed␤at /tmp/r7lBW9ikH1:1␤------> 3say7⏏5 ⁽²⁾␤Bogus postfix␤at /tmp/r7lBW9ikH1:1␤------> 3say 7⏏5⁽²⁾␤ expecting any of:␤ infix␤ infix stopper␤ …»
Zoffix m: say 2*⁽²⁺³⁾
camelia rakudo-moar : OUTPUT«5===SORRY!5=== Error while compiling /tmp/P4Sz9lOJbe␤Missing required term after infix␤at /tmp/P4Sz9lOJbe:1␤------> 3say 2*7⏏5⁽²⁺³⁾␤ expecting any of:␤ prefix␤ term␤»
uruwi Don't work, thoug
Zoffix m: say 2⁽²⁺³⁾
camelia rakudo-moar : OUTPUT«5===SORRY!5=== Error while compiling /tmp/5PyjsnbmTs␤Bogus postfix␤at /tmp/5PyjsnbmTs:1␤------> 3say 27⏏5⁽²⁺³⁾␤ expecting any of:␤ infix␤ infix stopper␤ postfix␤ statement end␤ statement m…»
uruwi Let's add them as alternatives to regular brackets!
Hotkeys we can use them for the superliteral quoting
uruwi Example? 00:11
Hotkeys sorry just literals
there are string with interpretation like "Answer is $foo" 00:12
then string w just escaping using single quotes
superscript brackets could be used for literal strings
no escaping or interpretation
s/interpretation/interpolation/
00:12 raiph left 00:13 Upasaka joined
Hotkeys I think you already can as long as you also have the Q 00:13
m: say Q⁽$hello wo\rld⁾; 00:14
camelia rakudo-moar : OUTPUT«$hello wo\rld␤»
Hotkeys :D
Zoffix :-0 00:15
00:15 telex left
AlexDaniel m: my \n = 8; say 2ⁿ 00:15
camelia rakudo-moar : OUTPUT«5===SORRY!5=== Error while compiling /tmp/uLpBn7hTyV␤Confused␤at /tmp/uLpBn7hTyV:1␤------> 3my \n = 8; say 27⏏5ⁿ␤»
Hotkeys no variables just literals
AlexDaniel Hotkeys: perhaps that's what Perl 6 should say… 00:16
00:16 telex joined
Hotkeys I'm sure either a nice error will be put in or variables will be implemented 00:17
It was only just put in like an hour ago
uruwi For i and n, though.
Too bad there are no superscript dollar signs
Hotkeys yeah 00:18
uruwi and 00:19
AlexDaniel yeahᵎ
uruwi DARN NO SUPERSCRIPT NEGATIVE ONE
(said the person who tried writing a TI-Basic book)
Hotkeys m: say 2⁻¹
camelia rakudo-moar : OUTPUT«5===SORRY!5=== Error while compiling /tmp/qJu5WwCBFK␤Bogus postfix␤at /tmp/qJu5WwCBFK:1␤------> 3say 27⏏5⁻¹␤ expecting any of:␤ infix␤ infix stopper␤ postfix␤ statement end␤ statement modifier…»
Hotkeys I saw it earlier
Zoffix m: 2¯² 00:20
camelia ( no output )
Zoffix m: say 2¯²
camelia rakudo-moar : OUTPUT«0.25␤»
uruwi m: say 2¯¹
camelia rakudo-moar : OUTPUT«0.5␤»
AlexDaniel .u ⁻
yoleaux U+207B SUPERSCRIPT MINUS [Sm] (⁻)
Hotkeys m: say 2¯¹
camelia rakudo-moar : OUTPUT«0.5␤»
AlexDaniel .u ¯
yoleaux U+00AF MACRON [Sk] (¯)
uruwi .u ¯
yoleaux U+00AF MACRON [Sk] (¯)
Zoffix is SMRT
uruwi Yep, better support both of them
Zoffix .u ¯
yoleaux U+00AF MACRON [Sk] (¯)
Hotkeys how do you do the macron
Zoffix this is what I get with Compose -> ^-
AlexDaniel .u ⁻ 00:21
yoleaux U+207B SUPERSCRIPT MINUS [Sm] (⁻)
AlexDaniel .u ¯
yoleaux U+00AF MACRON [Sk] (¯)
Hotkeys weird
Zoffix CTRL+SHIFT+u -> AF also works
Hotkeys I get superscript minus
although
this is wincompose
AlexDaniel Okay, macron is Compose key, ^, -
00:21 FROGGS left, lichtkind left
Hotkeys ah 00:22
AlexDaniel but ^- is superscript minus…
Hotkeys I got it
compose + _ + _ or compose + ^ + _
uruwi ¯
Yep
Hotkeys er
the second one is _ + ^
Zoffix .u ¯
yoleaux U+00AF MACRON [Sk] (¯)
Hotkeys m: say 2¯³ 00:23
camelia rakudo-moar : OUTPUT«0.125␤»
Zoffix This is so awesome :)
m: say (2¯³)¯³
camelia rakudo-moar : OUTPUT«512␤»
Zoffix This works too :O 00:24
uruwi .tell TimToady Make the ⁿ operator take ⁻ (superscript minus) as well as ¯ (macron) for negative powers.
yoleaux uruwi: I'll pass your message to TimToady.
AlexDaniel honestly, I have no idea why macron is supported…
uruwi Maybe it's a typo
Zoffix wtf is a "macron" anyway?
Hotkeys it's a diacritic
uruwi A diacritic
Hotkeys diacritical mark*
Zoffix hm
Hotkeys like Ē 00:25
uruwi as in ā
Zoffix Right
Hotkeys or ā
hahainternet might be my fault, those are the examples i gave
AlexDaniel m: say 2⁺⁸
camelia rakudo-moar : OUTPUT«256␤»
hahainternet i used macron by mistake
AlexDaniel okay, plus works
Hotkeys lol
AlexDaniel hahainternet: I'm glad that this feature is implemented!
hahainternet: thanks!
Hotkeys .u ¯ 00:26
yoleaux ...
Hotkeys ...
hahainternet AlexDaniel: it is very cool
Hotkeys m: say "¯".uniname
camelia rakudo-moar : OUTPUT«MACRON␤»
Hotkeys yeah
hahainternet proper unicode adoption is only going to drive better unicode use
Hotkeys it's just hardcoded in there
$power := -$power if $<sign> eq '¯'
AlexDaniel .tell TimToady or just drop macron… 00:27
yoleaux AlexDaniel: I'll pass your message to TimToady.
Zoffix hahainternet, TBH Compost ^6 gives me ⁶ so when I pressed Compose ^- and got a ¯, I expected it to be a minus. I think we should still support is as such
Noooo
.tell TimToady no don't drop macron. TBH Compost ^6 gives me ⁶ so when I pressed Compose ^- and got a ¯, I expected it to be a minus. I think we should still support is as such
yoleaux Zoffix: I'll pass your message to TimToady.
Zoffix .botsnack
yoleaux :D
synbot6 om nom nom
uruwi Change your compose key configuration.
hahainternet Zoffix: that is what made me do it
Zoffix uruwi, it's default.
Hotkeys We could just support both, no?
hahainternet and uruwi it's more a matter for distributors and the like to manage
we should just support whatever makes sense
AlexDaniel Zoffix: why don't you fix your keyboard instead? 00:28
Zoffix AlexDaniel, it's not broken.
hahainternet and a diacritic on its own isn't valid either i don't think
uruwi It's a standalone diacritic.
hahainternet so i'm ambivalent
Hotkeys it technically isn't the combining macron
just a macron
uruwi It wouldn't make sense for a combining one.
Hotkeys so it can be on its own
hahainternet what i mean is, as part of a number or variable
so it remains unambiguous, i think
Hotkeys 00:29
00:29 FROGGS joined
AlexDaniel still it is not a minus… 00:29
m: say 2 ➕ 5
camelia rakudo-moar : OUTPUT«5===SORRY!5=== Error while compiling /tmp/S8UG1kkqqr␤Confused␤at /tmp/S8UG1kkqqr:1␤------> 3say 27⏏5 ➕ 5␤ expecting any of:␤ infix␤ infix stopper␤ postfix␤ statement end␤ statement modifier␤ …»
00:29 raiph joined
AlexDaniel but this is a plus sign! 00:29
uruwi m: say "t̋".chars
camelia rakudo-moar : OUTPUT«1␤»
hahainternet AlexDaniel: i see your point for sure
Hotkeys m: say "̄̄̄̄̄̄̄̄"
camelia rakudo-moar : OUTPUT«5===SORRY!5===␤Argument to "say" seems to be malformed␤at /tmp/7XwscKWzrv:1␤------> 3say7⏏5 "̄̄̄̄̄̄̄̄"␤Bogus postfix␤at /tmp/7XwscKWzrv:1␤------> 3say 7⏏5"̄̄̄̄̄̄̄̄"␤ expecting any of:␤ infix␤ inf…»
hahainternet oh good point Hotkeys
Hotkeys I tried to put combining macrons in there
kek
hahainternet e and pi also valid exponents
Zoffix Alright :) I better leave before the Great Unicode Fight starts :)
hahainternet well i'm just happy there's some support in there 00:30
00:30 Zoffix left
Hotkeys is there a superscript pi? 00:30
hahainternet what you just typed ;)
uruwi No
hahainternet but no, you just made me think
uruwi What's the point 00:31
m: say "く̥̣̯て̯̣͏͜り̃す̣̥　（ドア）".chars
camelia rakudo-moar : OUTPUT«9␤»
hahainternet uruwi: that's how things are written in mathematics?
if only it was as useful as latex :(
uruwi Who raises numbers to the power of pi
AlexDaniel only subscript one? ᵨ
.u ᵨ
yoleaux U+1D68 GREEK SUBSCRIPT SMALL LETTER RHO [Lm] (ᵨ)
Hotkeys let's implement latex parsing in perl 6
:^)
uruwi Maybe as a module
Hotkeys all our sourcecode will be pdfs 00:32
uruwi It's a bit too insane for core ;P
Hotkeys with their own source code
in plaintext latex
the former for reading and the latter for parsing etc
uruwi And don't forget the IDE that combines all these 00:33
hahainternet :D
Hotkeys yes of course
hahainternet i have to write my new CV in latex later
there's a certain level of slightly masochistic enjoyment
Hotkeys write it in the upcoming PerlTeX instead
hahainternet in writing in latex
Hotkeys :p
uruwi I actually enjoy it
Hotkeys if we can have lualatex why not perltex 00:34
Yeah I write all of my papers in latex
uruwi はい　はい　はい
Hotkeys I used to do lab reports too but that was highschool and I don't have chemistry or physics anymore
because uni
hahainternet places that accept CVs only in .doc format
are awful
the worst thing is when they reformat CVs to send to you for interviews 00:35
the formatting of the CV tells me as much about the candidate as the contents :(
AlexDaniel
.oO( la6ex )
.oO( pronounced as lasex? )
 00:36
geekosaur lasix!
AlexDaniel yeah, that's probably more appropriate
Hotkeys but the tex part of latex is pronounced like tek 00:37
uruwi lasick
lasik
👀
Hotkeys I think lasik is already taken 00:38
by the laser eye surgery thing
I like when subsequent lines match in length
uruwi You ruined it 00:39
Hotkeys I know
: )
AlexDaniel ha
uruwi :) : )
Hotkeys : )
haha
uruwi hahaha
I can! 00:40
Say 1!
More!!
Ha!!!!
AlexDaniel but yeah, it is also fun when the line is constantly growing
uruwi Than!!
You!!!
Hotkeys cheats
uruwi Cunning?
カニング
Hotkeys ouch
uruwi But, anyway, hence the eyes. 00:41
Hotkeys talking about euclids elements in philosophy class
this is my jam
00:44 uruwi is now known as uruwi[A]
TimToady oops, I thought it was just a font glitch :) 00:45
yoleaux 00:24Z <uruwi> TimToady: Make the ⁿ operator take ⁻ (superscript minus) as well as ¯ (macron) for negative powers.
00:27Z <AlexDaniel> TimToady: or just drop macron…
00:27Z <Zoffix> TimToady: no don't drop macron. TBH Compost ^6 gives me ⁶ so when I pressed Compose ^- and got a ¯, I expected it to be a minus. I think we should still support is as such
00:46 Actualeyes left
dalek kudo/nom: 05e94d7 | TimToady++ | src/Perl6/ (2 files):
allow superscript minus
 00:47
00:48 FROGGS left
AlexDaniel m: say 2¯1 00:48
camelia rakudo-moar : OUTPUT«5===SORRY!5=== Error while compiling /tmp/lWfehXHTvt␤Bogus postfix␤at /tmp/lWfehXHTvt:1␤------> 3say 27⏏5¯1␤ expecting any of:␤ infix␤ infix stopper␤ postfix␤ statement end␤ statement modifier␤…»
hahainternet TimToady: when you say 'no variables', you mean that say $x² is fine, but not say 2^$x (or equiv) right?
also did you make a decision on subscripts in variable names?
no pressure on either of those of course, just wanted to be clear in my mind 00:49
AlexDaniel m: say 2¯¹
camelia rakudo-moar : OUTPUT«0.5␤»
AlexDaniel m: say 2¯¯¹
camelia rakudo-moar : OUTPUT«5===SORRY!5=== Error while compiling /tmp/Zc6HkETUlC␤Bogus postfix␤at /tmp/Zc6HkETUlC:1␤------> 3say 27⏏5¯¯¹␤ expecting any of:␤ infix␤ infix stopper␤ postfix␤ statement end␤ statement modifier…»
00:49 khw left
AlexDaniel m: say 2¯⁺¹ 00:49
camelia rakudo-moar : OUTPUT«5===SORRY!5=== Error while compiling /tmp/WwtQif1eCL␤Bogus postfix␤at /tmp/WwtQif1eCL:1␤------> 3say 27⏏5¯⁺¹␤ expecting any of:␤ infix␤ infix stopper␤ postfix␤ statement end␤ statement modifie…»
AlexDaniel m: say 2¯
camelia rakudo-moar : OUTPUT«5===SORRY!5=== Error while compiling /tmp/DYabivpKiX␤Bogus postfix␤at /tmp/DYabivpKiX:1␤------> 3say 27⏏5¯␤ expecting any of:␤ infix␤ infix stopper␤ postfix␤ statement end␤ statement modifier␤ …»
AlexDaniel okay-okay 00:50
uruwi[A] m: say 2⁻¹
camelia rakudo-moar : OUTPUT«5===SORRY!5=== Error while compiling /tmp/IaMrJXG083␤Bogus postfix␤at /tmp/IaMrJXG083:1␤------> 3say 27⏏5⁻¹␤ expecting any of:␤ infix␤ infix stopper␤ postfix␤ statement end␤ statement modifier…»
AlexDaniel uruwi[A]: not yet
TimToady another minute or so
uruwi[A] Also, any decision on standalone superscripts? 00:51
TimToady m: say 2⁻¹
camelia rakudo-moar : OUTPUT«0.5␤»
AlexDaniel m: say 2²⁻¹
camelia rakudo-moar : OUTPUT«0.25␤»
AlexDaniel m: say 2²⁻⁺¹
camelia rakudo-moar : OUTPUT«5===SORRY!5=== Error while compiling /tmp/8p29S7P2Ot␤Bogus postfix␤at /tmp/8p29S7P2Ot:1␤------> 3say 2²7⏏5⁻⁺¹␤ expecting any of:␤ infix␤ infix stopper␤ postfix␤ statement end␤ statement modi…»
TimToady well, standalone superscripts fall under the :No rule
AlexDaniel m: say 2⁺⁺¹
camelia rakudo-moar : OUTPUT«5===SORRY!5=== Error while compiling /tmp/IS584xfYsi␤Bogus postfix␤at /tmp/IS584xfYsi:1␤------> 3say 27⏏5⁺⁺¹␤ expecting any of:␤ infix␤ infix stopper␤ postfix␤ statement end␤ statement modifi…»
TimToady only a single + or - is allowed 00:52
uruwi[A] any plans to change?
00:52 FROGGS joined
TimToady no, it's not an expression 00:52
it's just a literal
Unicode doesn't really support anything beyond that very well 00:53
we don't have raised $ for instance 00:54
it's not really a place we can be productive
hahainternet: subscripts don't tickle my fancy, but modules will happen... 00:55
hahainternet TimToady: interesting, i wouldn't have thought they'd be too contentious
if others express an interest, would it change your mind?
00:55 Hazim joined
Hazim Hello, 00:55
TimToady hahainternet: I have to say "no", so as not to encourage rabble rousing. :P 00:56
uruwi[A] Hazim, hi
Mouq hahainternet: it gets silly quick :P
hahainternet aww, well i'd like to leave it in your mind somewhere as a feature that shouldn't be too confusing, x_a is common enough notation
TimToady blech
hahainternet but obviously, no pressure :)
TimToady I have resisted the urge to turn Perl 6 into many languages, so I guess I'll have to add TeX to the list... 00:57
AlexDaniel hahainternet: yet superscripts are doing ** magic yet you want subscript to mean nothing? :)
hahainternet AlexDaniel: no i just want it to be a permitted character in variable names 00:58
so you could have $x₁ and $x₂
Hotkeys seems reasonable
hahainternet both distinct, anyway it doesn't matter
it's up to TimToady and i don't want to push
uruwi[A] I'd probably use $x1 and $x2 anyway
hahainternet although i would like to know how to get the equiv of .u inside p6
i know about uninames and unival
AlexDaniel well yeah… Still I'd say that $x₁ is $x[1] …
hahainternet is there a nice succinct unicode description option?
nice use of … there AlexDaniel :)
Hotkeys lol 00:59
imagine
Hazim Hello I lIke To Ask if there is a function to manage "picture, Sounds, Hardware" or talk to Hardware directly from perl 6
Hotkeys $x₁…$x₂
hahainternet Hotkeys: beautiful
Hotkeys imagine
TimToady Hazim: your best bet would be to use NativeCall to a C library 01:00
Hotkeys $x₁+$x₂…$xₙ
$x₁+$x₂…$xₙ₋₁+$xₙ I suppose 01:01
AlexDaniel where ₙ could mean * I guess?
Hotkeys if you want to be explicit
it'd probably be better for an array notation
@x₁ @x₂ 01:02
AlexDaniel right
Hotkeys that was you could have @xₙ
and it would make sense
not sure if TimToady wants real subscripts for array subscripts though
TimToady both these ideas belong in module space
not really interested in having them built in at this point 01:03
Hotkeys alriiiiight
Hazim <TimToday> ok then i should learn C first, sorry for interrupting, I'm not programmer yet, but I could not find answers on the web 01:04
TimToady Hazim: well, even in C, you can't usually talk directly to hardware. You have to use the operating system.
AlexDaniel Hotkeys: which at this point means that you will sacrifice more than 1 second of startup time just to have subscripts. Oh, hey, look, superscript! It works! 01:05
Hotkeys startup time shmartup time
if I want subscripts I'll implement subscripts
although I don't care enough to do it right now 01:06
maybe later
Hazim TimToday: Ok, How To Learn&Find All The Function in the Perl 6,
AlexDaniel Hotkeys: haha yeah, fork Perl 6! Let's start a revolution!
Hotkeys With blackjack and subscripts 01:07
TimToady Hazim: see doc.perl6.org/
Hotkeys Game::Blackjack anyone?
AlexDaniel Hotkeys: and ∞!
Hotkeys lol
neat ∞ is just compose 8 8 01:08
I was thinking it'd be like ✈ (&airplane)
TimToady that one will probably sneak in at some point
Hotkeys I approve of it 01:09
I will include it in Perlkeys
Perlkeys 6?
hahainternet m: say 2÷2
camelia rakudo-moar : OUTPUT«5===SORRY!5=== Error while compiling /tmp/tEpe1SAi8n␤Bogus postfix␤at /tmp/tEpe1SAi8n:1␤------> 3say 27⏏5÷2␤ expecting any of:␤ infix␤ infix stopper␤ postfix␤ statement end␤ statement modifier␤…»
AlexDaniel Hotkeys: look, they are catching up! Quick, add ≤ ≥ ≠ !
TimToady
.oO(use PERLKEY-TYPING;)
 01:10
Hotkeys those are all attractive symbols
I'm always tempted to type =/= when programming even though I know it's very usually !=
Hazim TimToday: Thanks Man, Last Thing About The Picture, Sounds,3d...! And C library" i should learn C or no, COZ i like to Make Apps With GUI, so i ask this Q 01:11
hahainternet or <> (nnngh)
AlexDaniel Hotkeys: compose = /
Hotkeys ≠
beauty
Why did I not ever get wincompose before
besides the fact I didn't know about it
AlexDaniel try ≤ ≥ while you are at it
hahainternet i didn't realise hovercar wrote it
i'll pass your thanks along
Hotkeys ≥≤
≥.≤
it's a face 01:12
hahainternet i don't actually know how to type those
TimToady Hazim: for GUI, you probably want to learn C first, then C++
hahainternet ≤≥
figured it out
Hotkeys yeah
hahainternet compose is quite intuitive
Hotkeys straightforward
hahainternet also TimToady is totally right, learn a bit of C, when you come to C++ recoil in horror and write Perl instead :D
Hotkeys haha
TimToady Hazim: in a year, Perl 6 will be better at GUI though
Hotkeys perl 6 is the universal language of the future
look out c# 01:13
and c++
uruwi[A] I would use perl6 bindings to native libraries more if they worked on Windows :S
TimToady well, our optimizer has to get a lot better to compete with C++
uruwi[A] ^
01:13 yeahnoob joined
AlexDaniel Hotkeys: no! Perl₆ is! 01:13
Hazim TimToday: Ok i will wait, but when it will be released.
Hotkeys perfect name AlexDaniel
much better than Perlkeys
uruwi[A] I think some people would avoid it just for performance reasons, even if they don't need that extra speed 01:14
01:14 uruwi[A] is now known as uruwi
hahainternet for context: en.wikipedia.org/wiki/Mathematical...tors_block 01:14
there's quite a few here i've no idea how to type
quite a few of which could be reasonably adopted
uruwi To say that all of these will eventually have a use in Perl 6...
Hazim TimToday: And What About The Web Application, how Perl 6 Go with the web app. 01:15
Hotkeys ☺
oh compose is wonderful
AlexDaniel hahainternet: set operations are already in, I guess
zengargoyle if you're on Linux-ish and really using XCompose (and not some other input method rip-off) : /usr/share/X11/locale/en_US.UTF-8/Compose
yoleaux 25 Nov 2015 16:18Z <TimToady> zengargoyle: yes, I decided to keep | where it was, and if we want a slipper at so/not precedence, call it something else, like "sl" or some such, but then its not really saving keystrokes over |(1..10)
Hotkeys also if you're using wincompose there is a gui thing for viewing 01:17
01:17 salva left
TimToady Hazim: we already have several Web modules on modules.perl6.org 01:17
AlexDaniel m: unit module Perl₆; sub foo { say ‘yeah!’ }
camelia rakudo-moar : OUTPUT«5===SORRY!5=== Error while compiling /tmp/rGr7qTCFat␤Unable to parse module definition␤at /tmp/rGr7qTCFat:1␤------> 3unit module Perl7⏏5₆; sub foo { say ‘yeah!’ }␤ expecting any of:␤ generic role␤»
AlexDaniel noooo
zengargoyle if using something like iBus/SCIM/UIM might want to look into the rfc1345 input method also.
Hotkeys Σ would be attractive for sum 01:18
zengargoyle sadly, iBus doesn't really use XCompose, just a subset (and doesn't honor .XCompose customization :( )
AlexDaniel Hotkeys: well it seems like you managed to write a non-breaking space
Hotkeys I mightve hit compose a second time 01:19
uruwi Hotkeys, people will also want to use the Greek letter Σ too, then others will complain that they can't use it for variable names
Hazim TomToday: Thanks Man I Will Look To the Modules and Doc,
Hotkeys fair enough
01:19 salva joined
hahainternet interesting zengargoyle 01:19
i have an ibus service installed
but shift + altgr is my compose
and i can't find set contains in the compose list :( 01:20
uruwi try compose I N
A ∈ B
Hotkeys ∈
you tricked me
it was compose i n
hahainternet not for me it seems on debian, with either case
zengargoyle if you use iBus, compose keys are in sqlite database: /usr/share/ibus-table/tables/compose.db
uruwi What about the other set ops?
Hotkeys ∋ 01:21
compose n i
guessed that one
hahainternet zengargoyle: nothing on mine, and damn Hotkeys now i feel left out
uruwi wow
∉ !in
hahainternet going to have to up my compose game on debian
uruwi ∌ /ni
∌ /ni 01:22
Hotkeys ∅
is compose { }
AlexDaniel ịn
uruwi {^ {u ∩ ∪
Hotkeys heh
hahainternet ∅ works for me but that's about it :(
zengargoyle Hotkeys: it's from ibus-table-compose package.
uruwi {( }) ⊂ ⊃
* {)
AlexDaniel <Multi_key> <leftshoe> <U0338> : "⊄" 01:24
Hotkeys ⊄
AlexDaniel I wonder what is “leftshoe”
Hotkeys it's the subset symbol
I assume
uruwi Probably what that symbol looks like
AlexDaniel how does it expect me to type it?
Hotkeys with compose I guess
AlexDaniel compose in compose? 01:25
Hotkeys I can do compose ! {(
AlexDaniel I've never tried that
Hotkeys I can do compose ! { (
no I mean
maybe it's a combination one
⊇ 01:26
AlexDaniel .u 0x0bda
yoleaux No characters found
AlexDaniel .u x0bda
yoleaux No characters found
AlexDaniel .u U+2282
yoleaux U+2282 SUBSET OF [Sm] (⊂)
AlexDaniel so that's a <leftshoe> 01:27
Hotkeys ah
is what I said :p
AlexDaniel weird, so it expects me to have this character on my keyboard. Not a problem, but it is a weird assumption 01:29
Hazim Goodbye and Have a nice day to all.
Hotkeys .u U+2276
yoleaux U+2276 LESS-THAN OR GREATER-THAN [Sm] (≶)
Hotkeys that's a neat sign
01:29 Hazim left
Hotkeys never actually seen that 01:30
.u U+2278
yoleaux U+2278 NEITHER LESS-THAN NOR GREATER-THAN [Sm] (≸)
Hotkeys wouldn't this be the same as '='
mathematically speaking
uruwi It would
Hotkeys just checking
uruwi Unless you're talking about JavaScript
01:30 salva left
zengargoyle snorts 01:30
Hotkeys .u U+2272 01:31
yoleaux U+2272 LESS-THAN OR EQUIVALENT TO [Sm] (≲)
uruwi .u U+2277
yoleaux U+2277 GREATER-THAN OR LESS-THAN [Sm] (≷)
uruwi .u U+2279 01:32
yoleaux U+2279 NEITHER GREATER-THAN NOR LESS-THAN [Sm] (≹)
AlexDaniel what about LESS-THAN OR GREATER-THAN AND NEITHER LESS-THAN NOR GREATER-THAN ?
uruwi What about LESS-THAN AND GREATHER-THAN
.u U+2280
yoleaux U+2280 DOES NOT PRECEDE [Sm] (⊀)
Hotkeys but
my brain hurts
uruwi .u U+2273
yoleaux U+2273 GREATER-THAN OR EQUIVALENT TO [Sm] (≳)
uruwi .u U+2274
yoleaux U+2274 NEITHER LESS-THAN NOR EQUIVALENT TO [Sm] (≴)
Hotkeys .u U+227D 01:33
yoleaux U+227D SUCCEEDS OR EQUAL TO [Sm] (≽)
uruwi Hotkeys, please do not go into JavaScript
Hotkeys does javascript have these symbols
uruwi No
But comparisons are weird
01:33 salva joined
uruwi Such a shame, because JavaScript deserves to use them 01:33
JS U+2279 Java 01:34
Hotkeys what does less than and greater than even mean
uruwi It might not be false if you use an abysmal programming language
Hotkeys I like perl comparisons 01:35
specifically the ability to chain them
perl 6*
(never used 5)
uruwi used bits of 5 for golfing
01:36 flussence left
uruwi Mainly to improve on others' answers. 01:36
01:36 cdg left
Hotkeys I did a bit of perl 6 golfing a few weeks ago 01:36
01:36 flussence joined
zengargoyle x is < and > y if x.succ < y and x.pred > y ? 01:36
Hotkeys it was never the shortest thanks to golf langs
but
but how zengargoyle
uruwi One of the shortest that isn't specifically designed for golf
Hotkeys: if some idiot designs a language that makes this sometimes true 01:37
of course
zengargoyle in some field that is not as rational numbers?
Hotkeys lol
zengargoyle there's some weird stuff out there in maths... :)
uruwi Maybe modular numbers?
Hotkeys maybe
I remember ordinal numbers blowing my mind
and the different kinds of infinite in general 01:38
er 01:39
infinity
uruwi Neither equal nor unequal 01:41
?
Mouq OTOH, if ∞ = -∞, then ∞ ≶ 0 but ∞ ≠ 0
Hotkeys they aren't equal usually are they?
uruwi Nope 01:42
Mouq (NEITHER LESS-THAN NOR GREATER-THAN, assuming I copied the right symbol)
Hotkeys yeah
uruwi Wrong symbol
Hotkeys oh
right wrong one
Mouq Wait
uruwi Which is different from the wrong right one
Mouq all the wrongs 01:43
vOv
Hotkeys ≸
Mouq nvm :)
uruwi I think the food they delivered us was laced with...
With...
Hotkeys I like this one
TimToady to me, ≸ looks like "incomparable" in a topological sort
Hotkeys ⋘
very much less than
.u ⋘ 01:44
yoleaux U+22D8 VERY MUCH LESS-THAN [Sm] (⋘)
Hotkeys :D
uruwi ⋘←—
TimToady always wonders, how much is "very"?
Hotkeys .u ≪ 01:45
yoleaux U+226A MUCH LESS-THAN [Sm] (≪)
Hotkeys more than much apparently
uruwi Let a AAAAAAAAAAAAAAAAAAAAAAAAA b = a ⋘ b ∧ a ⋙ b
Hotkeys how to do the (not? or? I forget my logic symbols)
uruwi Compose slash yen 01:46
Hotkeys yen???
uruwi And it's an AND
TimToady remembers using ⥁ and ⥀ for rotate right and left in a rosettacode entry...
uruwi Those should definitely be the symbols for rotation operators
Hotkeys I was close
apparently it's compose / \
∧
grondilu uruwi: it would not make it clear which is which 01:47
TimToady well, Japanese keyboards still confuse \ and ¥, I suspect
uruwi I said yen as a joke :P
Hotkeys oh you
okay lecture is over heading home
01:47 BenGoldberg joined
Hotkeys be back in ~45 minutes 01:48
uruwi m: say 62 ⟲ 5
camelia rakudo-moar : OUTPUT«5===SORRY!5=== Error while compiling /tmp/npPQnj2Nbv␤Confused␤at /tmp/npPQnj2Nbv:1␤------> 3say 627⏏5 ⟲ 5␤ expecting any of:␤ infix␤ infix stopper␤ postfix␤ statement end␤ statement modifier␤ …»
uruwi What about +<< and +>>
I can never remember: +< or <+
01:49 BenGoldberg left
skids m: my %a; %a.VAR.WHAT.say; my %b::a; %b::a.VAR.WHAT.say 01:49
camelia rakudo-moar : OUTPUT«(Hash)␤(Scalar)␤»
01:51 salva left 01:52 dayangkun joined 01:54 BenGoldberg joined, Zoffix joined 01:55 BenGoldberg left, BenGoldberg joined
Zoffix Is it possible to alter what a call to {now} returns? I have code that will function different based on what time it is and I want to have some easy way to test it without changing my system time. 01:55
01:56 salva joined
uruwi m: sub now {return 1;}; say now; 01:56
camelia rakudo-moar : OUTPUT«Instant:1448503031.478328␤»
uruwi Apparently now
not
grondilu maybe use the TZ env var? 01:57
Zoffix Guessing by the name it's a time zone?
grondilu m: say now; %*ENV<TZ>='Europe/Paris'; say now
camelia rakudo-moar : OUTPUT«Instant:1448503100.303290␤Instant:1448503100.308322␤»
Zoffix Won't be enough. I need to go through about 30 periods of time between Dec 31, 2015—Jan 1, 2016 01:58
grondilu well it doesn't work
zengargoyle m: sub now() { 1 }; say now;
camelia rakudo-moar : OUTPUT«Instant:1448503127.863511␤»
uruwi m: sub now {return 1;}; say now();
camelia rakudo-moar : OUTPUT«1␤»
skids m: my constant now = "OHAI"; now.say; # hrm WTH is bare "now"
camelia rakudo-moar : OUTPUT«Instant:1448503153.312523␤»
zengargoyle m: \now := 1; say now;
camelia rakudo-moar : OUTPUT«5===SORRY!5=== Error while compiling /tmp/O21LvNel5B␤Cannot use bind operator with this left-hand side␤at /tmp/O21LvNel5B:1␤------> 3\now := 17⏏5; say now;␤»
uruwi m: my \now = 1; say now 01:59
camelia rakudo-moar : OUTPUT«Instant:1448503192.080633␤»
zengargoyle m: my \term:<now> := 1; say now;
camelia rakudo-moar : OUTPUT«Instant:1448503202.164129␤»
uruwi You'll have to use a sub with a different name
zengargoyle lol
TimToady that's a bug, I think 02:00
testing a fix
zengargoyle m: my &now := sub { return 1 }; say now;
camelia rakudo-moar : OUTPUT«Instant:1448503288.789749␤»
grondilu m: $*TZ=3600; say now; $*TZ=-18000; say now 02:01
camelia rakudo-moar : OUTPUT«Cannot modify an immutable Int␤ in block <unit> at /tmp/B13ShpAEAA:1␤␤»
zengargoyle m: my &term:<now> := sub { return 1 }; say now;
camelia rakudo-moar : OUTPUT«1␤»
zengargoyle bingo!
Zoffix not really
m: my &term:<now> := sub { return 1 }; say DateTime.now; 02:02
camelia rakudo-moar : OUTPUT«2015-11-26T03:02:02+01:00␤»
Zoffix heh
zengargoyle, thanks.
I think I'll just use an env var (or one of those "dynamic variables" and in my code use a $*TEST_TIME // now
grondilu m: my $*TZ=3600; say now; my $*TZ=-18000; say now 02:03
camelia rakudo-moar : OUTPUT«Potential difficulties:␤ Redeclaration of symbol $*TZ␤ at /tmp/tmKHYytBs4:1␤ ------> 3my $*TZ=3600; say now; my $*TZ7⏏5=-18000; say now␤Instant:1448503418.146657␤Instant:1448503418.148529␤»
grondilu m: my $*TZ=3600; say now; $*TZ=-18000; say now
camelia rakudo-moar : OUTPUT«Instant:1448503425.745172␤Instant:1448503425.749895␤»
grondilu meh
02:03 rwilson_py joined, rigsby left
AlexDaniel .u ⥀⟲ 02:04
yoleaux U+27F2 ANTICLOCKWISE GAPPED CIRCLE ARROW [Sm] (⟲)
U+2940 ANTICLOCKWISE CLOSED CIRCLE ARROW [Sm] (⥀)
zengargoyle m: my &now := sub { return 1 }; say now; use MONKEY-TYPING; augment class DateTime { sub now() { return 1; } }; say DateTime.now;
camelia rakudo-moar : OUTPUT«Instant:1448503511.871252␤2015-11-26T03:04:35+01:00␤»
skids m: my $*TZ=3600; say DateTime.now; $*TZ=-18000; say DateTime.now;
camelia rakudo-moar : OUTPUT«2015-11-26T03:04:36+01:00␤2015-11-25T21:04:36-05:00␤»
AlexDaniel hahainternet: here is a crazy thought
what if subscripts are only allowed in scalar names, but forbidden in array names. Then both parties are satisfied 02:05
zengargoyle m: my &term:<now> := sub { return 1 }; say now; use MONKEY-TYPING; augment class DateTime { sub now() { return 1; } }; say DateTime.now;
camelia rakudo-moar : OUTPUT«1␤2015-11-26T03:05:42+01:00␤»
AlexDaniel it will not work on array references, but…
uruwi Too complicated now. 02:06
AlexDaniel well, not much. $x₂ and @x₂
zengargoyle m: my &term:<now> := sub { return 1 }; say now; use MONKEY-TYPING; augment class DateTime { method now() { return 1; } }; say DateTime.now; 02:07
camelia rakudo-moar : OUTPUT«5===SORRY!5=== Error while compiling /tmp/5XpR5EcaoM␤Package 'DateTime' already has a method 'now' (did you mean to declare a multi-method?)␤at /tmp/5XpR5EcaoM:1␤»
AlexDaniel both are kinda DWIM, but yeah, not pretty 02:08
zengargoyle m: my &term:<now> := sub { return 1 }; say now; use MONKEY-TYPING; augment class DateTime { multi method now() { return 1; } }; say DateTime.now;
camelia rakudo-moar : OUTPUT«===SORRY!===␤Cannot have a multi candidate for 'now' when an only method is also in the package 'DateTime'␤»
02:08 molaf left, FROGGS left
Zoffix m: say DateTime.new( $*TEST_TIME // now ).local.month 02:08
camelia rakudo-moar : OUTPUT«11␤»
Zoffix m: my $*TEST_TIME = 1428503546; say DateTime.new( $*TEST_TIME // now ).local.month
camelia rakudo-moar : OUTPUT«4␤»
Zoffix Sweet 02:09
zengargoyle m: my &term:<now> := sub { return 1 }; say now; use MONKEY-TYPING; augment class DateTime { proto method new { * }; multi method now() { return 1; } }; say DateTime.now;
camelia rakudo-moar : OUTPUT«5===SORRY!5=== Error while compiling /tmp/y5Ntcfv0g_␤Package 'DateTime' already has a method 'new' (did you mean to declare a multi-method?)␤at /tmp/y5Ntcfv0g_:1␤»
Zoffix Gonna see how painful it is to do timezone stuff with P6 :P
I never managed to Do It Right™ in P5.... always droves of bugs
Mouq m: my &term:<now> := sub { return 1 }; say now; class DateTime { method now() { return 1; } }; say DateTime.now; # :P 02:14
camelia rakudo-moar : OUTPUT«1␤1␤»
02:18 Ben_Goldberg joined 02:19 xpen joined 02:21 BenGoldberg left
hahainternet AlexDaniel: honestly i wouldn't mind either way, as long as it was something agreed upon 02:24
for now TimToady says he doesn't want it in core so it'd be down to whichever module you used :) 02:25
AlexDaniel hahainternet: yeah right, if only it didn't affect the startup time so much. 02:26
02:26 Ben_Goldberg is now known as BenGoldberg
Hotkeys Hello I made it home 02:29
02:29 rwilson_py left
Zoffix \o 02:30
m: my \x = 42; say x² - x² == x² - x²; say x*(x - x) == (x+x)*(x - x); my \z = (x - x); say x*z == (x+x)*z; # Therefore x == (x+x) or x == 2*x :) 02:33
camelia rakudo-moar : OUTPUT«True␤True␤True␤»
Zoffix hmm.. Actually, in Perl 6 it just *might* be true due to rats 02:34
uruwi Divide by zero
Zoffix m: my \x = 42; say x² - x² == x² - x²; say x*(x - x) == (x+x)*(x - x); my \z = (x - x); say x*z/0 == (x+x)*z/0;
camelia rakudo-moar : OUTPUT«True␤True␤True␤»
Zoffix lols
this is epic :) We broke math
uruwi m: say 1/0
camelia rakudo-moar : OUTPUT«Attempt to divide 1 by zero using div␤ in block <unit> at /tmp/yTpxMX7jvz:1␤␤Actually thrown at:␤ in block <unit> at /tmp/yTpxMX7jvz:1␤␤»
uruwi m: say 0/0 02:35
camelia rakudo-moar : OUTPUT«Attempt to divide by zero using div␤ in block <unit> at /tmp/0RGdULBSyZ:1␤␤Actually thrown at:␤ in block <unit> at /tmp/0RGdULBSyZ:1␤␤»
uruwi m: say 1/0 == 2/0
camelia rakudo-moar : OUTPUT«True␤»
Hotkeys 1/0.nude
Zoffix m: my \x = 42; say x² - x² == x² - x²; say x*(x - x) == (x+x)*(x - x); my \z = (x - x); say x/0 == 2*x/0;
camelia rakudo-moar : OUTPUT«True␤True␤True␤»
Hotkeys m: say 1/0.nude
camelia rakudo-moar : OUTPUT«Method 'nude' not found for invocant of class 'Int'␤ in block <unit> at /tmp/tE6LHTO_KQ:1␤␤»
uruwi m: say 1/0.nude
camelia rakudo-moar : OUTPUT«Method 'nude' not found for invocant of class 'Int'␤ in block <unit> at /tmp/DRf5zpsy_r:1␤␤»
Hotkeys m: say (1/0).nude 02:36
camelia rakudo-moar : OUTPUT«(1 0)␤»
Zoffix it's stored as denominator=>0/numerator=>0 stuff
Hotkeys I mean
the comparison isn't wrong in the context of rats
just isn't mathematically okay
:P
uruwi m: say (1/0 + 2/0).nude 02:37
camelia rakudo-moar : OUTPUT«(3 0)␤»
Hotkeys haha
Zoffix I think this came up earlier and the decision was that is is ok :)
uruwi m: say (1/0 + 2).nude
camelia rakudo-moar : OUTPUT«(1 0)␤»
Hotkeys seems fine to me
Zoffix 'cause you can do Lim(x/0) == Lim(2*x/0)
or some such
Hotkeys it's still doing the fractional addition etc. correctly
you just gotta know 02:38
m: say (1/0).denominator != 0 ?? "okay" !! "not okay" 02:39
camelia rakudo-moar : OUTPUT«not okay␤»
02:39 kaare_ joined
Zoffix m: say (⅓+⅓+⅓).nude 02:39
camelia rakudo-moar : OUTPUT«(1 1)␤»
psch Rats are kind of lazy div
Zoffix magical 02:40
02:40 pierre-vigier joined
psch i don't feel completely comfortable with that, fwiw 02:40
02:40 pierre-vigier left, pierre-vigier joined
Hotkeys m: sub ok(Rat:D $n) { $n.denominator != 0 }; say ok 1/0; 02:40
camelia rakudo-moar : OUTPUT«False␤»
psch m: sub ok(Rat:D $n) { $n.denominator != 0 }; say ok <1/0>.Rat 02:41
camelia rakudo-moar : OUTPUT«False␤»
psch m: sub ok(Rat:D $n) { $n.denominator != 0 }; say ok <1/0>
camelia rakudo-moar : OUTPUT«False␤»
psch RatStr probably coerces or somesuch vOv
m: say <1/0>.WHAT
camelia rakudo-moar : OUTPUT«(Rat)␤»
psch eh, or i messed up how to build them :/
m: say < 1/0 >.WHAT
camelia rakudo-moar : OUTPUT«(RatStr)␤»
psch there
m: sub ok(Rat:D $n) { $n.denominator != 0 }; say ok < 1/0 >
camelia rakudo-moar : OUTPUT«False␤»
Hotkeys I didn't know those existed
Zoffix m: sub ok(Rat:D $n) { $n.denominator == 0 ?? "∞" !! $n }; say [ ok 1/0, ok 42/2 ];
camelia rakudo-moar : OUTPUT«Too many positionals passed; expected 1 argument but got 2␤ in sub ok at /tmp/Wd6r9Ev7rE:1␤ in block <unit> at /tmp/Wd6r9Ev7rE:1␤␤»
psch Hotkeys: what? RatStr? 02:42
Zoffix m: sub ok(Rat:D $n) { $n.denominator == 0 ?? "∞" !! $n }; say [ ok(1/0), ok(42/2) ];
camelia rakudo-moar : OUTPUT«[∞ 21]␤»
uruwi What's a RatStr?
Hotkeys ^
yeah
psch well
Hotkeys Zoffix: that isn't true though
psch the *Str types come from a place of "on the CLI everything is a String"
Zoffix Hotkeys, prove me wrong :P
uruwi Why have both?
psch (that is IntStr, RatStr and i think ComplexStr)
Hotkeys at a stretch you could say ±∞ 02:43
but still not really
psch m: my $x = < 1/0 >; say $x.Str; say $x.Rat
camelia rakudo-moar : OUTPUT«1/0␤Attempt to divide 1 by zero using div␤ in block <unit> at /tmp/lAGOHUaQ30:1␤␤Actually thrown at:␤ in block <unit> at /tmp/lAGOHUaQ30:1␤␤»
psch m: my $x = 1/0; say $x.Str
camelia rakudo-moar : OUTPUT«Attempt to divide 1 by zero using div␤ in block <unit> at /tmp/wpklF989q1:1␤␤Actually thrown at:␤ in block <unit> at /tmp/wpklF989q1:1␤␤»
psch ^^^ that's one point where RatStr is useful
02:43 skids left
Zoffix m: sub ok(Rat:D $n) { $n.denominator == 0 ?? "∞" !! $n }; say [ ok(abs 1/0), ok(abs 42/2) ]; 02:43
camelia rakudo-moar : OUTPUT«[∞ 21]␤»
Zoffix There. Fixed it :P
psch 'cause the RatStr (and IntStr and ComplexStr) knows the Str representation *without* evaluating the Rat (or Int or Complex) representation 02:44
Zoffix might be getting delirious from perpetual lack of sleep.
Hotkeys I think you are
:p
psch m: say ComplexStr
camelia rakudo-moar : OUTPUT«(ComplexStr)␤»
psch uruwi: does that help with "why have both?" :) 02:45
Hotkeys m: say (3 + 2i).Str; say (3 + 2i).ComplexStr
camelia rakudo-moar : OUTPUT«3+2i␤Method 'ComplexStr' not found for invocant of class 'Complex'␤ in block <unit> at /tmp/b_R7w2Eh8r:1␤␤»
Zoffix m: say sqrt -1
camelia rakudo-moar : OUTPUT«NaN␤»
Zoffix :(
psch Hotkeys: no coercion type for those. as i said, they come from either the CLI or from invoking &val on a Str
Hotkeys RatStr etc. don't exist on the docs.perl6.org site it seems
ah
psch m: say val "3 + 2i" 02:46
camelia rakudo-moar : OUTPUT«3 + 2i␤»
psch m: say (val "3 + 2i").WHAT
camelia rakudo-moar : OUTPUT«(Str)␤»
psch hm, maybe not in that form
m: say (val "3+2i").WHAT
camelia rakudo-moar : OUTPUT«(ComplexStr)␤»
psch there
space are bad, cause the *Str types are for dense formes
-e
Hotkeys say (val 3 + 2i).WHAT
m: say (val 3 + 2i).WHAT
camelia rakudo-moar : OUTPUT«Value of type Complex uselessly passed to val() in block <unit> at /tmp/YfkctHCv6e:1␤(Complex)␤»
Zoffix m: say (2+1i).WHAT
camelia rakudo-moar : OUTPUT«(Complex)␤»
Hotkeys rip
Zoffix cool.. I wonder how many langs have complex numbers out of the box
Hotkeys not a ton
uruwi Python
Zoffix aw :( 02:47
Hotkeys Factor
Hotkeys plugs factor
Hotkeys likes factor
Zoffix never heard of Factor
Hotkeys It's a concatenative lang
psch "out of the box" in python means what?
'cause < print(2i) > here complains :p
Hotkeys think Forth
but not
:p
Zoffix psch, seems like cmath module is needed: docs.python.org/2/library/cmath.html 02:48
uruwi Python uses j instead of i, though
psch ah, yeah that seems to work
Zoffix m: say (0/0)+(0+1i)
camelia rakudo-moar : OUTPUT«Inf+1i␤»
psch somewhat loses meaning, though, considering j isn't i 02:49
Hotkeys Factor is fun but I never did anything serious with it
Zoffix This is kinda pushing it... (the Inf+1i)
psch ...not that i really mind that, fwiw
Hotkeys concatenative.org/wiki/view/Factor/Examples
uruwi Well, complex numbers itself exist, but you have to use cmath instead of math math
Zoffix m: say (0/0)+(0+1i) 02:50
camelia rakudo-moar : OUTPUT«Inf+1i␤»
Hotkeys psch: it is sometimes
it's the other letter for the imaginary number
Zoffix .tell masak you seem to know math well. Does this make any sense or should it be throwing due to division by zero? m: say (0/0)+(0+1i) <camelia> rakudo-moar : OUTPUT«Inf+1i␤»
yoleaux Zoffix: I'll pass your message to masak.
psch Hotkeys: i haven't encountered it, but i'm also not particularly deep into formal math
Zoffix \o night all
psch Hotkeys: usually just arithmetic to check if i can pay for groceries :P 02:51
Hotkeys psch: it's used a lot in cases where there might be ambiguity
like when i denotes electric current
psch ISTR we used a capital i for current
in school that is
maybe that's unit vs dimension 02:52
Hotkeys I've seen both
psch well, not like it really matters :)
Hotkeys even on the wiki page they use i in an image of a current and I in the formulae
yeah
you could use ɨ if you wanted 02:53
or į
Hotkeys is finding uses for compose
(just like xkcd.com/208/)
Zoffix m: say (0+i) == sqrt -1 # this is kinda disappointing 02:54
camelia rakudo-moar : OUTPUT«False␤»
Hotkeys m: sqrt -1
camelia rakudo-moar : OUTPUT«WARNINGS:␤Useless use of "sqrt -1" in expression "sqrt -1" in sink context (line 1)␤»
Hotkeys m: say sqrt -1
camelia rakudo-moar : OUTPUT«NaN␤»
02:54 khw joined
Hotkeys weird 02:54
psch m: say &sqrt.candidates>>.signatures
camelia rakudo-moar : OUTPUT«Method 'signatures' not found for invocant of class 'Sub'␤ in block <unit> at /tmp/3jv2zaTIot:1␤␤»
Hotkeys m: say sqrt -1 + 0i
camelia rakudo-moar : OUTPUT«6.12323399573677e-17+1i␤»
Hotkeys kek
psch m: say &sqrt.candidates>>.signature
camelia rakudo-moar : OUTPUT«((Numeric \x) (Cool \x) (num $a --> num))␤»
Hotkeys ComplexRat anyone?
m: say sqrt -1.Complex 02:55
camelia rakudo-moar : OUTPUT«6.12323399573677e-17-1i␤»
psch not sure an Int candidate is worth the dispatch trouble just for making sqrt -i work out
Hotkeys m: say sqrt (-1).Complex
camelia rakudo-moar : OUTPUT«6.12323399573677e-17+1i␤»
Zoffix m: say sqrt -1.0
camelia rakudo-moar : OUTPUT«NaN␤»
psch s/-i/-1/
Hotkeys looks like strong typing strikes again
or rather
yeah I guess 02:56
I'm tired
02:56 vendethiel joined
psch well, the problem is that you want to decide on the return type before evoking the sub 02:56
as in, you'd want a sqrt --> Complex
Hotkeys I know, it makes sense
for the same reason that my Rat $foo = 1; isn't okay
psch that's the path towards wantarray, fwiw :P 02:57
Hotkeys (I think?)
colomon yeah, making the sqrt of negative Reals NaN is very intentional.
Hotkeys m: my Rat $foo = 1;
camelia rakudo-moar : OUTPUT«5===SORRY!5=== Error while compiling /tmp/Er0UpQaYw8␤Cannot assign a literal of type Int (1) to a variable of type Rat. You can declare the variable to be of type Real, or try to coerce the value with 1.Rat or Rat(1), or just write the value as 1.0␤at /t…»
Hotkeys yeah okay
psch similar, yes, except Rat = 1 is more of a type system thing
Hotkeys just making sure I wasn't imagining things
same idea though
Zoffix colomon, why is it intentional?
I mean, why not return a Complex?
Hotkeys Because in the reals there is no sqrt -1
psch while "i want a Complex from sqrt here but not everywhere" is a "let me change semantics of a function by changing the callsite" on top of that 02:58
...kinda thing
colomon sqrt Real generates a Real
sqrt Complex generates a Complex
Zoffix m: say NaN.WHAT
camelia rakudo-moar : OUTPUT«(Num)␤»
psch m: say sqrt 1i
camelia rakudo-moar : OUTPUT«0.707106781186548+0.707106781186547i␤»
Zoffix m: say sqrt (-1+i) 02:59
camelia rakudo-moar : OUTPUT«0.455089860562227+1.09868411346781i␤»
Zoffix m: say sqrt (-1+0i)
camelia rakudo-moar : OUTPUT«6.12323399573677e-17+1i␤»
Hotkeys I'm sad that you can't do
m: say i²;
camelia rakudo-moar : OUTPUT«-1+1.22464679914735e-16i␤»
Hotkeys WHAT
oh waiut
uruwi Floating point
Hotkeys my local build isn't up to date
no not that
I forgot to rebuild rakudo
to get superscripts
Zoffix m: say (sqrt -1).WHAT 03:00
camelia rakudo-moar : OUTPUT«(Num)␤»
Hotkeys wtb ComplexRat type
Zoffix OK.
uruwi What kind of representation is being used for complex numbers?
Hotkeys real + imaginary
psch uruwi: nums for real and imaginary
Hotkeys oh
oops
I misunderstood 03:01
uruwi Then what's with the funky errors?
psch Hotkeys: ComplexRat seems LHF-ish, not sure how to wire it up in a least-surprise way though
uruwi: which errors?
uruwi You're not by chance storing them in the form r^(e i θ)?
Zoffix psch, what's LHF?
Hotkeys LHF?
uruwi Rounding errors
psch Low Hanging Fruit
Hotkeys is that a good thing?
psch uruwi: num is native (i.e. IEE754) float
03:01 xpen left
psch Hotkeys: it means "it should be easy to implement" :P 03:02
uruwi Yes, but why does (0+1i)**2 give -1 plus a small number i
Hotkeys ah
uruwi Shouldn't it work exactly to -1+0i?
BenGoldberg Because it's implemented in terms of sin and cos
uruwi I knew it
So you're doing intermediate conversions from rect to polar when you multiply
psch BenGoldberg++ 03:03
Zoffix BenGoldberg++ :)
psch that's beyond my readily available math :)
Zoffix Once you go i you never get real again :)
uruwi BenGoldberg++
Hotkeys m: say 0.1e0 + 0.2e0 - 0.3e0;
camelia rakudo-moar : OUTPUT«5.55111512312578e-17␤»
Hotkeys :D
yay floating points
AlexDaniel m: say 0/3 == 0/3 03:04
camelia rakudo-moar : OUTPUT«True␤»
BenGoldberg When you multiply, no conversion to polar should be done. When you exponentiate a complex, then a polar conversion does need to be done.
uruwi Fortunately you shouldn't have to resort to floating points often.
AlexDaniel m: say ↉ == 0/3
camelia rakudo-moar : OUTPUT«True␤»
Hotkeys yeah
rats make it nice
uruwi Ah BenGoldberg
That's it
Hotkeys m: say 0.1 + 0.2 - 0.3;
camelia rakudo-moar : OUTPUT«0␤»
Hotkeys :D
Zoffix m: say i * i;
camelia rakudo-moar : OUTPUT«-1+0i␤»
Zoffix w00t!
BenGoldberg 0.1 is a rat, not a float ;)
Zoffix BenGoldberg++ :)
Hotkeys nice
BenGoldberg: that's what I said
AlexDaniel m: say ⅒ == 0.1
camelia rakudo-moar : OUTPUT«True␤»
AlexDaniel m: say ⅒⅒ 03:05
camelia rakudo-moar : OUTPUT«5===SORRY!5=== Error while compiling /tmp/LPfXgrERzW␤Bogus postfix␤at /tmp/LPfXgrERzW:1␤------> 3say ⅒7⏏5⅒␤ expecting any of:␤ infix␤ infix stopper␤ postfix␤ statement end␤ statement modifier…»
uruwi m: say 0.1 == 0.1e0
camelia rakudo-moar : OUTPUT«True␤»
BenGoldberg m: ⅒.perl.say
camelia rakudo-moar : OUTPUT«0.1␤»
Hotkeys m: say 0.1 === 0.1e0;
camelia rakudo-moar : OUTPUT«False␤»
psch m: say 1e-1 == .1
camelia rakudo-moar : OUTPUT«True␤»
psch ...probably cheating somewhere
Hotkeys I mean they have the same value
AlexDaniel m: say 0.33 == 0.33e0
camelia rakudo-moar : OUTPUT«True␤»
Zoffix m: say i*i == i²; 03:06
camelia rakudo-moar : OUTPUT«False␤»
Zoffix heh
Hotkeys m: say "0.3" == 0.3 == 3e-1;
camelia rakudo-moar : OUTPUT«False␤»
Hotkeys wow
rude
uruwi Gets more annoying when you do have to actually compare two complex numbers
BenGoldberg m: say +"0.3"
camelia rakudo-moar : OUTPUT«0.3␤»
uruwi The Right Way™
AlexDaniel m: say i**2 == i²
camelia rakudo-moar : OUTPUT«True␤»
Hotkeys m: say "0.3" == 0.3;
camelia rakudo-moar : OUTPUT«True␤»
BenGoldberg m: (+"0.3").WHAT.say
camelia rakudo-moar : OUTPUT«(Rat)␤»
Hotkeys m: say "0.3" == 3e-1; 03:07
camelia rakudo-moar : OUTPUT«False␤»
psch uruwi: i have a hunch a lot of what we have with Complex is ground work, and there hasn't been a real world use-case yet
BenGoldberg m: (3e-1).WHAT.say
camelia rakudo-moar : OUTPUT«(Num)␤»
Hotkeys m: say 0.3 == 3e-1;
camelia rakudo-moar : OUTPUT«False␤»
AlexDaniel
.oO( poor camelia. Peaple are poking too much stuff into her today)
people*
Hotkeys but then why does
03:08 yqt left
Hotkeys m: say 0.1 == 1e-1; 03:08
camelia rakudo-moar : OUTPUT«True␤»
Hotkeys work?
BenGoldberg Magic?
03:08 TEttinger left
psch m: say 1e-1.fmt("%.15f") 03:08
camelia rakudo-moar : OUTPUT«0.100000000000000␤»
psch m: say 3e-1.fmt("%.15f")
camelia rakudo-moar : OUTPUT«0.300000000000000␤»
psch hrm
uruwi m: say 3e-1.fmt("%.18f")
camelia rakudo-moar : OUTPUT«0.300000000000000000␤»
Zoffix m: say 3e-1; say 1e-1
camelia rakudo-moar : OUTPUT«0.3␤0.1␤»
uruwi m: say 3e-1.fmt("%.21f")
camelia rakudo-moar : OUTPUT«0.300000000000000000000␤»
Zoffix :S
uruwi m: say 3e-1.fmt("%.99f")
camelia rakudo-moar : OUTPUT«0.300000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000␤»
psch m: my num $x = 3e-1; printf("%.21f", $x) 03:09
camelia rakudo-moar : OUTPUT«0.300000000000000000000»
uruwi m: say (1e-1+2e-1).fmt("%.99f")
camelia rakudo-moar : OUTPUT«0.300000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000␤»
uruwi What magic is this
Hotkeys haha
uruwi m: say (1e-1+2e-1-3e-1).fmt("%.99f")
camelia rakudo-moar : OUTPUT«0.000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000␤»
Zoffix m: say (3e-1-.3).fmt("%.99f")
camelia rakudo-moar : OUTPUT«0.000000000000000055511151231257800000000000000000000000000000000000000000000000000000000000000000000␤»
Hotkeys lol
Zoffix There's your problem :)
m: say (1e-1-.1).fmt("%.99f")
camelia rakudo-moar : OUTPUT«0.000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000␤»
psch m: printf("%.21f", 3e-1) 03:10
camelia rakudo-moar : OUTPUT«0.300000000000000000000»
psch m: printf("%.40f", 3e-1)
camelia rakudo-moar : OUTPUT«0.3000000000000000000000000000000000000000»
psch ...i dont get it :P
Hotkeys me either
psch shrugs and goes to sleep
Zoffix The ghost in the machine! :)
Zoffix does the same
psch hf :P
pierre-vigier m: say (0.1+0.2-0.3).fmt("%.20f")
camelia rakudo-moar : OUTPUT«0.00000000000000000000␤»
AlexDaniel m: my $x = 0.3.Num; printf("%.40f", $x)
camelia rakudo-moar : OUTPUT«0.3000000000000000000000000000000000000000»
Hotkeys I don't know enough about computer math magic to diagnose this
AlexDaniel m: my $x = 0.3.Num; printf("%.90f", $x)
camelia rakudo-moar : OUTPUT«0.300000000000000000000000000000000000000000000000000000000000000000000000000000000000000000»
AlexDaniel why does it work? It should be “broken” 03:11
Mouq I'm late to the party, but Hotkeys:
AlexDaniel m: my $x = 0.31235612.Num; printf("%.90f", $x)
camelia rakudo-moar : OUTPUT«0.312356120000000000000000000000000000000000000000000000000000000000000000000000000000000000»
Mouq m: say (-1).roots(2)
camelia rakudo-moar : OUTPUT«(6.12323399573677e-17+1i -1.83697019872103e-16-1i)␤»
AlexDaniel m: my num $x = 0.31235612.Num; printf("%.90f", $x)
camelia rakudo-moar : OUTPUT«0.312356120000000000000000000000000000000000000000000000000000000000000000000000000000000000»
Hotkeys m: say (-4).roots(2) 03:12
camelia rakudo-moar : OUTPUT«(1.22464679914735e-16+2i -3.67394039744206e-16-2i)␤»
Hotkeys those rounding errors are mighty ugly
Mouq m: say (-4).roots(2)>>.round # taadaaaa :P 03:13
camelia rakudo-moar : OUTPUT«(0+2i 0-2i)␤»
Hotkeys m: say (-4).roots(2)>>.round>>.say; 03:14
camelia rakudo-moar : OUTPUT«0+2i␤0-2i␤(True True)␤»
gfldex m: say 0.3 == 3e-1, (0.3).WHAT, 3e-1.WHAT; # apples and oranges
camelia rakudo-moar : OUTPUT«False(Rat)(Num)␤»
Hotkeys that's nicer
m: say (-4).roots(2)».round».say;
camelia rakudo-moar : OUTPUT«0+2i␤0-2i␤(True True)␤»
Hotkeys excuse to use compose
now I can use » and « in my code 03:16
this is exciting
03:17 noganex joined
Mouq m: say (-4).roots(2)>>.reals>>.Rat.map({.[0]+.[1]i}) 03:18
camelia rakudo-moar : OUTPUT«(0+2i 0-2i)␤»
03:18 vendethiel left
uruwi Is there any simple way of expressing function composition? 03:18
Mouq I think we have an "of" operator 03:19
03:19 adu joined
Mouq .u COMPOSITION 03:19
yoleaux U+2384 COMPOSITION SYMBOL [So] (⎄)
U+2A1F Z NOTATION SCHEMA COMPOSITION [Sm] (⨟)
U+2A3E Z NOTATION RELATIONAL COMPOSITION [Sm] (⨾)
Mouq hrm
.u RING 03:20
yoleaux U+00C5 LATIN CAPITAL LETTER A WITH RING ABOVE [Lu] (Å)
U+00E5 LATIN SMALL LETTER A WITH RING ABOVE [Ll] (å)
U+016E LATIN CAPITAL LETTER U WITH RING ABOVE [Lu] (Ů)
adu Mouq: are you looking for center circle?
uruwi .u circle
yoleaux U+05AF HEBREW MARK MASORA CIRCLE [Mn] (◌֯)
U+20DD COMBINING ENCLOSING CIRCLE [Me] (◌⃝)
U+20E0 COMBINING ENCLOSING CIRCLE BACKSLASH [Me] (◌⃠)
uruwi .u center circle
yoleaux No characters found
Mouq Nope, it's U+2218 RING OPERATOR
geekosaur or
03:20 noganex_ left
geekosaur .u composite 03:20
yoleaux No characters found
uruwi .u U+2218
yoleaux U+2218 RING OPERATOR [Sm] (∘)
Mouq But just o works actually
geekosaur ah, no, that is description, sigh 03:21
uruwi m: say (cos o sin)(2)
camelia rakudo-moar : OUTPUT«5===SORRY!5=== Error while compiling /tmp/_Cx44avJQs␤Unsupported use of bare "sin"; in Perl 6 please use .sin if you meant $_, or use an explicit invocant or argument, or use &sin to refer to the function as a noun␤at /tmp/_Cx44avJQs:1␤------> 3say (c…»
uruwi m: say (&cos o &sin)(2)
camelia rakudo-moar : OUTPUT«0.614300282116482␤»
Hotkeys neat
uruwi m: say cos(sin(2))
camelia rakudo-moar : OUTPUT«0.614300282116482␤»
uruwi TIL.
Hotkeys m: say cos sin 2;
camelia rakudo-moar : OUTPUT«0.614300282116482␤»
Hotkeys parens are lame 03:22
adu yeah RING OPERATOR is function composition
uruwi .u ο 03:24
yoleaux U+03BF GREEK SMALL LETTER OMICRON [Ll] (ο)
uruwi .u ꙭ
yoleaux U+A66D CYRILLIC SMALL LETTER DOUBLE MONOCULAR O [Ll] (ꙭ)
adu you know what my favorite unicode symbol is?
uruwi What
.u 🍣 03:25
yoleaux U+1F363 SUSHI [So] (🍣)
uruwi ^ my favorite
adu .u U+1F432
yoleaux U+1F432 DRAGON FACE [So] (🐲)
adu .u U+1F409
yoleaux U+1F409 DRAGON [So] (🐉)
03:25 Ben_Goldberg joined
adu .u U+1F4A9 03:26
yoleaux U+1F4A9 PILE OF POO [So] (💩)
uruwi Don't know. Can't beat sushi.
And another:
.u 👀
yoleaux U+1F440 EYES [So] (👀)
03:26 Ben_Goldberg left
pierre-vigier m: my &fct = &roots o &abs; say fct( -4 ); 03:27
camelia rakudo-moar : OUTPUT«Too few positionals passed; expected 2 arguments but got 1␤ in block <unit> at /tmp/BiBH3l5noi:1␤␤»
Hotkeys lol pile of poo
pierre-vigier why??
Hotkeys because of roots
er
no
not that
uh
uruwi Poots?
pierre-vigier m: roots(4).sayu
camelia rakudo-moar : OUTPUT«5===SORRY!5=== Error while compiling /tmp/Vq030q5DMj␤Calling roots(Int) will never work with proto signature ($,)␤at /tmp/Vq030q5DMj:1␤------> 3<BOL>7⏏5roots(4).sayu␤»
Mouq m: my &fct = &sqrt o &abs; say fct( -4 );
camelia rakudo-moar : OUTPUT«2␤»
pierre-vigier m: roots(4).say
camelia rakudo-moar : OUTPUT«5===SORRY!5=== Error while compiling /tmp/63p5i5ychG␤Calling roots(Int) will never work with proto signature ($,)␤at /tmp/63p5i5ychG:1␤------> 3<BOL>7⏏5roots(4).say␤»
Hotkeys oh
roots requires two 03:28
pierre-vigier hum, indeed
uruwi m: roots(4, 4).say
camelia rakudo-moar : OUTPUT«(1.4142135623731+0i 8.65956056235493e-17+1.4142135623731i -1.4142135623731+1.73191211247099e-16i -2.59786816870648e-16-1.4142135623731i)␤»
Hotkeys as in
m:say 4.roots(4)
m: say 4.roots(4)
camelia rakudo-moar : OUTPUT«(1.4142135623731+0i 8.65956056235493e-17+1.4142135623731i -1.4142135623731+1.73191211247099e-16i -2.59786816870648e-16-1.4142135623731i)␤»
Hotkeys or what uruwi did
pierre-vigier makes sense
03:29 Ben_Goldberg joined, xpen joined, BenGoldberg left
adu .u U+1F6BD 03:29
yoleaux U+1F6BD TOILET [So] (🚽)
pierre-vigier m: my &fct = &sqrt o &abs; say fct( -4 ); 03:30
camelia rakudo-moar : OUTPUT«2␤»
Mouq m: my &fct = &roots o {.abs,2}; say fct( -4 ); # a compromise :)
camelia rakudo-moar : OUTPUT«(2+0i -2+2.44929359829471e-16i)␤»
adu uruwi: anyways my favorite is a toss up between DRAGON and PILE OF POO
Hotkeys lol
I like pile of poo because it uses the word poo 03:31
03:31 kaare_ left
Hotkeys compared to something more official like feces 03:31
uruwi SUSHI is one of mine because I like sushi
And also
.u ɬ
yoleaux U+026C LATIN SMALL LETTER L WITH BELT [Ll] (ɬ)
adu Hotkeys: although I think the Unicode consortium is required to add STREAM OF PISS for completeness
uruwi .u STREAM
yoleaux U+1F38F CARP STREAMER [So] (🎏)
uruwi How would that be depicted 03:32
And I think they'd opt for PEE instead
.u PEE
yoleaux U+1354 ETHIOPIC SYLLABLE PEE [Lo] (ፔ)
U+15EB CANADIAN SYLLABICS CARRIER PEE [Lo] (ᗫ)
U+333B SQUARE PEEZI [So] (㌻)
uruwi .u PISS
yoleaux No characters found
uruwi .u POO
yoleaux U+1434 CANADIAN SYLLABICS POO [Lo] (ᐴ)
U+1435 CANADIAN SYLLABICS Y-CREE POO [Lo] (ᐵ)
U+A576 VAI SYLLABLE POO [Lo] (ꕶ)
uruwi .u POOP
yoleaux No characters found
Hotkeys .u URINE
yoleaux U+1F755 ALCHEMICAL SYMBOL FOR URINE [So] (🝕)
uruwi .u SHIT
yoleaux U+1201F CUNEIFORM SIGN AK TIMES SHITA PLUS GISH [Lo] (𒀟)
U+121DD CUNEIFORM SIGN LAGAB TIMES SHITA PLUS GISH PLUS ERIN2 [Lo] (𒇝)
U+121DE CUNEIFORM SIGN LAGAB TIMES SHITA PLUS GISH TENU [Lo] (𒇞)
adu uruwi: it doesn't exist, but the Unicode consortium is very concerned about completeness, that's why the next version is going to include MANTA RAY and SHARK 03:33
Hotkeys oh no Digest::MD5 still hasn't been fixed 03:34
for windows
adu Hotkeys: I know a lot about MD5, but nothing about windows
Hotkeys it's specifically due to newlines
adu oh, you mean \r\n? 03:35
Hotkeys basically because of the newline handling changes it expects a hash of a string with \n
and gets a hash of the string with \r\n
yep
adu the \r\n
are there two functions? like hash_binary and hash_string? 03:36
Hotkeys idk
I just know what's broken
adu that's how I would do it
Hotkeys I think there are though
adu have the hash_string normalize newlines, and hash_binary is the raw version
03:36 kid51 left
Hotkeys it isn't the choice of the functions 03:37
unless you specify with use newline <option>;
adu so you're saying there is a hash_string-ish function, and it's not working?
Hotkeys m: say "test\r\n"
camelia rakudo-moar : OUTPUT«test
Hotkeys er
m: say "test\r\n".perl
camelia rakudo-moar : OUTPUT«"test\r\n"␤»
Hotkeys m: say "test\n".perl
camelia rakudo-moar : OUTPUT«"test\n"␤»
Hotkeys I can't make it break right now
I'm bad at this 03:38
> say "\n".perl
"\r\n"
this is why it breaks
it tests the function with "Hello world\n" 03:39
but before it even gets to the function
it gets changed to \r\n
adu that's bad 03:40
Hotkeys well not entirely it's just that you have to expect it
and I think this was written before the handling change
you can specify that the newline is only \n
but you have to specify
with "use newline lf;"
I think that would fix this
but I'm not sure
Mouq Hotkeys: Try it and make a pull request :) 03:41
Hotkeys I guess so
Hotkeys is lazy
03:42 rorx left
Mouq is pretty sure Hotkeys spent more time trying to explain the problem than it would take to fix ;) 03:42
Hotkeys that is likely
03:42 rigsby joined 03:50 rigsby left
Hotkeys alright I pull requested 03:51
03:52 BenGoldberg_ joined
Hotkeys meanwhile I installed the updated version locally 03:53
will the be the first time I
Task::Star will build for me?
the answer is yes 03:55
woo
03:55 Ben_Goldberg left
Hotkeys I apologize for filling up everyones chats, I have a bad habit of talking in many lines 03:55
04:02 BenGoldberg_ left 04:07 eliasr left 04:10 risou left 04:13 FROGGS joined
lucs What would correspond to Perl5's \K in regexes? (searched for 'keep' in S05, but found nothing relevant) 04:15
04:19 pierre-vigier left
Mouq lucs: )> 04:19
lucs Um, not sure what you mean. 04:20
Mouq m: "abcdefg" ~~ /abc)>defg/; say $/
camelia rakudo-moar : OUTPUT«｢abc｣␤»
Mouq m: "abcdefg" ~~ /a<(bc)>defg/; say $/
camelia rakudo-moar : OUTPUT«｢bc｣␤»
lucs Aha. Thanks.
Mouq m: "abcdefg" ~~ /a<(bcdefg/; say $/
camelia rakudo-moar : OUTPUT«｢bcdefg｣␤»
04:21 Ribbit left
dalek kudo/nom: d3c278d | TimToady++ | src/Perl6/Grammar.nqp:
allow override of terms self/now/time
 04:22
uruwi kok! 04:24
04:26 risou joined
TimToady m: my \now = 1; say now 04:33
camelia rakudo-moar : OUTPUT«1␤»
TimToady unfortunatly, couldn't use that on rand, since there's a &rand function 04:34
04:34 yeahnoob left, yeahnoob joined
ShimmerFairy TimToady: somehow, a random inconsistency in the allowed overrides feels appropriate for 'rand' :P 04:44
TimToady that was a randim comment :P 04:48
04:48 FROGGS left, FROGGS joined
TimToady well, a rnddim comment, anyhoo... 04:49
04:56 rigsby joined 05:05 FROGGS left 05:13 BlackNeko joined 05:18 BlackNeko left 05:19 quester joined 05:22 rigsby left 05:25 rigsby joined 05:27 snarkyboojum joined 05:36 xpen left, xpen joined 05:38 TEttinger joined 05:41 khw left, xpen left, pierre-vigier joined 05:42 adu left 05:49 rigsby left 05:55 rigsby joined 06:05 pierre-vigier left 06:07 xpen joined
grondilu probably a typo in recent commit or something: 06:10
make: /usr/locale/bin/moar: Command not found 06:11
"locale"?
oh wait, it's my Configure.pl line that has it
06:11 pierre-vigier joined
grondilu must haved type 'e' by mistake before entering the command 06:12
06:14 raiph left 06:17 rigsby left 06:28 Actualeyes joined 06:34 doublec_ joined 06:35 doublec left 06:45 rurban joined 06:51 doublec joined 06:53 doublec_ left 06:54 pierre-vigier left 06:55 domidumont joined 06:56 |Tux| joined 07:02 bjz joined 07:06 geraud left 07:07 pierre-vigier joined, doublec_ joined, FROGGS joined 07:08 doublec left 07:12 CIAvash joined 07:13 FROGGS left, rindolf joined 07:17 domidumont left 07:35 xinming left 07:36 freeze left, ilogger2_ joined, btyler joined, freeze joined, hacst_ joined, esh_ joined, ilbot3 joined 07:37 havenwood joined, havenwood left, havenwood joined, Gruber joined, petercommand joined, bakedb joined 07:38 moritz_ joined, zostay joined, sjn_ joined, mjgardner joined 07:39 gensym joined, lucs_ joined 07:40 colomon_ joined, colomon_ is now known as colomon, psch joined 07:41 pnu left, hacst_ is now known as hacst, mjgardner is now known as Gardner, atta joined, charsbar_ joined, yeahnoob joined 07:42 yakudza joined, krakan joined 07:43 ugexe joined, kipd joined, Timbus joined 07:44 oka_ joined, BooK joined, gtodd joined 07:45 pochi joined, pierrot joined, PerlJam joined, mst joined, TimToady joined 07:46 darutoko joined, pdcawley_ joined 07:47 lizmat joined 07:48 cognominal joined
TimToady had a bit of a hiccup with freenode, apparently, and wondering if anyone else had trouble getting through over the last hour or so... 07:48
07:48 tony-o joined 07:49 flaviusb joined, bjz joined, Upasaka joined
nine TimToady: there were lots of netsplits 07:50
TimToady I couldn't even reconnect for a long time
07:50 khisanth_ joined, ShimmerFairy joined
TimToady and I noticed dalek stopped reporting 07:50
07:50 domidumont joined, mathw joined, stux|RC-only joined 07:51 dustinm` joined, hacst_ joined, obra joined, mattp__ joined, zostay_ joined 07:52 Jonis joined, olinkl joined 07:53 cgfbee joined 07:54 domidumont left, Ulti joined 07:55 novapatch joined, domidumont joined 07:56 yeahnoob left, zostay left, hacst left, btyler left, tadzik left, daxim left, ribasushi joined, hacst_ is now known as hacst, leont joined, moznion joined, Zoffix joined 07:57 rindolf joined, revdiablo joined, btyler joined, raydiak joined
moritz_ TimToady: I'm restaring the IRC component from dalek 07:57
07:58 JimmyZ joined, daxim joined, El_Che joined, Possum joined 07:59 yeahnoob joined
moritz_ Nov 26 07:59:24 hack systemd[12488]: Failed at step EXEC spawning /home/dalek/perl5/perlbrew/perls/perl-5.20.1/bin/perl: Resource temporarily unavailable 07:59
07:59 apejens joined
moritz_ whatever that means 07:59
07:59 rvchangue joined 08:00 vike joined, breinbaas joined 08:01 DrPete_ joined 08:02 domidumont left 08:03 tadzik joined
nine moritz_: seems like its running into some ulimit 08:03
moritz_ Connect to freenode failed, retry in 30 secs 08:05
if I start it manually
08:06 telex joined, tinita joined 08:07 clkao joined
moritz_ but why would it run into a ulimit now, when it worked fine all the time before? 08:08
08:09 abraxxa joined, Bucciarati joined 08:11 dalek joined, ChanServ sets mode: +v dalek
moritz_ LimitNPROC=5 in the service file seemed to have caued the error 08:11
bumping it up to 8 helped
08:11 leont left, solarbunny joined 08:12 PotatoGim joined, pnu joined, skarn joined 08:13 corbyhaas joined, frew joined, n0xff joined
moritz_ the http process runs with 5 workers; that might be a tiny bit overkill 08:13
08:15 doublec joined 08:16 rurban joined 08:20 f3ew joined 08:22 advwp_ joined, mls_ joined 08:23 kipd left, kipd joined, olinkl left, olinkl joined 08:24 clkao left, clkao joined, bowtie joined, PotatoGim left, PotatoGim joined, pnu left, pnu joined, corbyhaas left, corbyhaas joined
hahainternet moritz_: 5 workers is small time 08:25
and yes that error is as you spotted
but for a webserver, max numproc should be more like ~200, although maybe i'm just used to lots of horizontal scaling 08:26
moritz_ hahainternet: the http component receives a github web hook every few minutes or hours
hahainternet: there's simply no need for more than one or two workers
hahainternet: and since this is a very shared server, I like to limit my services very much 08:27
hahainternet oh! i assume this was for the new perl modules site or similar :)
moritz_ no :-)
hahainternet moritz_: well if the EXEC is actually a fork, you can fork as much as you like
as most of the ram is COW
unless you are one of those crazy cats that turns off overcommit
08:27 moritz_ is now known as moritz
moritz hahainternet: well, I limited the number of processes via systemd container/slice fu 08:28
08:28 n0xff left
hahainternet also, you're on my list of people to thank for writing awesome articles, but i can't remember which article 08:28
so thanks in general!
something i saw had your name on it though
08:29 RabidGravy joined, MilkmanDan joined 08:30 pnu left
moritz hahainternet: you're welcome :-) 08:30
hahainternet: usually at perlgeek.de/blog-en/ or on the perl 6 advent calendar
08:30 pierre-vigier joined
hahainternet moritz: advent has been a staple of my christmas for many years 08:30
i've been meaning since last year to get more involved, but it's hard to find time 08:31
Hotkeys I'm excited for this years advent
08:34 abraxxa1 joined 08:36 n0xff joined
hahainternet www.raspberrypi.org/wp-content/upl...g_0661.jpg 08:37
ok that's pretty cool
free computer with magazine
we're officially in the future
08:40 ely-se joined 08:41 skarn left, leont joined
Hotkeys "Today, I’m pleased to be able to announce the immediate availability of Raspberry Pi Zero, made in Wales and priced at just $5." 08:43
neat
08:43 Ven joined
hahainternet $5 is a really superb price point 08:43
that's less than i have paid for esp8266s
08:44 abraxxa left, Bucciarati left
hahainternet and they're significantly less of a computer than an rpi 08:44
08:44 solarbunny left, dalek left 08:45 sjn_ is now known as sjn 08:47 g4 joined
lizmat good *, #perl6! 08:48
nine Hi lizmat!
lizmat nine o/
Tux__ o/
Ven \o
lizmat Tux__ Ven o/ 08:49
nine I wonder what people see as requirement for merging curli
08:49 zakharyas joined 08:50 bjz left 08:51 bjz joined
lizmat personally, I think having the stuff in lib be installed using curli ? 08:52
and having panda just work (TM)
08:53 kjs_ joined, g4 left
hahainternet oh that reminds me, i have to see if i'm behind on Linenoise 08:53
cause it doesn't work so well with some unicode glyphs 08:54
lizmat hmmm... new unicode plans that may affect NFG: blog.unicode.org/2015/11/new-charac...ended.html
|Tux| No such method 'context' for invocant of type 'Inline::Perl5::Perl6Callbacks'
in method invoke-parent at /pro/3gl/CPAN/Text-CSV6/.panda-work/1448527967_1/lib/Inline/Perl5.pm6:503
in block at /pro/3gl/CPAN/Text-CSV6/.panda-work/1448527967_1/lib/Inline/Perl5.pm6:991
in block <unit> at t/v6.t:19
Inline::Perl5 fail
lizmat if I scanned these unicode plans correctly, it would mean that the base codepoint of a synthetic would no longer be always the first codepoint 08:55
hahainternet yeah linenoise still problematic, enter ² and then try and backspace over it 08:56
don't know where i'd file this
|Tux| test 50000 22.252 22.142 08:58
test-t 50000 20.041 19.931
csv-parser 50000 27.576 27.466
hahainternet github.com/antirez/linenoise/issues/25
Hotkeys works fine for me I think
in the repl
hahainternet interesting, this is in tmux
lets try in a bare term
Hotkeys I'm using windows and powershell 08:59
hahainternet ah, fails with regular gnome-terminal
this issue is still open and no sign of merged fixes, so i'm going to blame the backing library
09:06 dalek joined, ChanServ sets mode: +v dalek, skarn joined
nine lizmat: both are already done 09:06
lizmat nine++
I'm afraid I'm not going to backlog the past 2 weeks: hoping for a P6W soon to get me up to speed again 09:07
09:09 bjz left
lizmat .tell jnthn the test fail in t/spec/S32-hash/adverbs.t seems to be caused by a MMD issue that only happens when running under the t/harness 09:13
yoleaux lizmat: I'll pass your message to jnthn.
lizmat .tell jnthn line 58 to be precise 09:14
yoleaux lizmat: I'll pass your message to jnthn.
09:18 g4 joined, g4 left, g4 joined 09:24 thundergnat joined 09:27 leont left
lizmat afk for a few hours& 09:27
09:28 pnu joined, pnu left, pnu joined, solarbunny joined, leont joined, yeahnoob left 09:31 brrt joined, maddingu1 joined, baest_ joined 09:32 ribasushi left, Bucciarati joined, colomon_ joined, El_Che left, charsbar_ left 09:33 jonas1 joined, breinbaas left, btyler left, colomon left, colomon_ is now known as colomon, vytas joined, gtodd1 joined, ely-se left, RabidGravy left, g4 left, obra left, cognominal left, PerlJam left, gtodd left, esh_ left 09:34 esh joined, RabidGravy joined, zoosha joined, El_Che joined, btyler joined, bowtie left, stmuk joined, breinbaas joined 09:35 xinming joined, cognominal joined, ely-se joined 09:38 ollej joined 09:39 ribasushi joined 09:41 obra joined 09:42 PerlJam joined, leont left
Zoffix Wait what... they make $5 computers in Whales? 0.o 09:47
09:47 charsbar_ joined, mtj_ joined 09:48 FROGGS joined, iH2O joined
nine Well you can build complete GSM phones for < $12, so it's not that surprising 09:50
09:50 iH2O left 09:55 bjz joined 10:00 M-Illandan joined, M-eternaleye joined
dalek href="https://modules.perl6.org/new-db-builder:">modules.perl6.org/new-db-builder: 52edb30 | (Zoffix Znet)++ | / (2 files):
Implement --interval option

defaulting to 5 seconds allows us to stay under GitHub"s rate limit of 5,000reqs/hr when builder is run continuously (Closes #36)		 10:01
10:03 espadrine joined
dalek href="https://modules.perl6.org/new-db-builder:">modules.perl6.org/new-db-builder: 1c800f7 | (Zoffix Znet)++ | build-project-list.pl:
--github-token-file option must take an argument and not be a boolean
 10:03
10:10 boegel joined 10:14 g4 joined, g4 left, g4 joined 10:15 ollej left 10:27 Ven left, Ven joined 10:32 bowtie joined 10:37 jesper_ joined 10:39 pierre-vigier left, bjz left 10:40 Ven left, scopesoft joined
scopesoft Just downloaded the latest Perl6 using rakudobrew and wrote a simple program that pushed a million floating point numbers (which where calculated), onto a list and then re-read the list checking the values. With perl6 this took 24s, perl5 took 0.2s, with D 0.04s. Taking away the floating point calculation and simply pushing integers on the list the perl6 time was reduced to 4s. I'm hoping performance will improve before the release at Christmas. 10:41
Zoffix quotes self from another channel... 10:43
nine scopesoft: very unlikely. Until the release it's all about making things correct so people may depend on the behavior. I guess pretty much everyone is very much looking forward to working on performance after the release :)
Zoffix "I wouldn't hold breath for Christmas. There are lots of more important things to take care. Optimization will come later on :)"
10:44 pecastro joined
scopesoft Given that there will probably be a large interest in this release, not only in the Perl domain it could be the final nail in the Perl 6 coffin, even before it's got off the ground. 10:45
arnsholt Shit being flat out wrong would be a nail in the coffin 10:47
Making something fast before it's right is completely pointless 10:48
mrf_ would much prefer a reliable language that can get faster that a fast language that will always change
Zoffix scopesoft, we'll put a sticker on it "It's still slow. Patches welcome" :)
scopesoft, and I think allusions to coffins are overly pessimistic. It's a brand new language. It's a bit silly comparing it to Perl 5 that had more than a decade to optimize itself :) 10:49
scopesoft That would be true in normal situations, but we've been waiting years for Perl 6, and when we finally get a date for a release, it's not a teeny bit slower than Perl 5 but light years slower
Zoffix scopesoft, the above statement ignores the fact that Perl 6 is no longer an "update" to Perl 5. It's a brand new language now. 10:50
scopesoft But comparisons will be made not only with Perl 5, but Ruby, Python, etc.
Zoffix scopesoft, sure. I don't see a problem with that. 10:51
scopesoft, those langs are 20 years old. Perl 6 doesn't have a "stable" release yet.
scopesoft, we have a FAQ entry on speed, if you'd like to peruse it: doc.perl6.org/language/faq#Is_Perl_..._for_me%3F
scopesoft, you could also give a helping hand with optimization, if you can: github.com/perl6/nqp/ github.com/rakudo/rakudo/ 10:53
scopesoft Zoffix: Thanks for the link.
Zoffix We use a modern compiler, so optimization is definitely possible. We just haven't gotten to it, yet.
10:55 pecastro left
DrForr_ Welcome to instant gratification. It's about to be released, so of course it must be perfect. Oh, no, it's not? Forget about it. 10:57
Zoffix I also want it to make pancakes! 10:58
scopesoft The reason I've brought this up, is due to a conversation I had with a mate this morning who writes a lot of Perl, but he said to he's not looking at Perl 6 at all, stay away from it, look at other languages such as Javascript. The company he works at are moving from Perl 5 to PHP !!! A Christmas release of Perl 6 which Larry has pre-announced with great fan-fare, with this sought of simple test will mean there is no chance it will be ever considered as an a
Zoffix :S javascript? :S
DrForr_ So the company is moving to PHP. Ergo he isn't interested in Perl 6. WHAT A SHOCKER. 10:59
scopesoft Recently I've done some work at the BBC and every time Perl or Perl 6 was mentioned it was just ridiculed.
Zoffix scopesoft, well, first, I think PHP will lose quite a bit of its popularity since Wordpress is moving to node.js, so many companies who decide to use that objectively horrible language will have fewer reasons to do it.
scopesoft, and so far you offered few reasons to think Perl 6 is ill-fated from the start. Your Perl 5-coder mate told you not to look at it? C'mon. 11:00
scopesoft DrForr: he's supporting the Perl legacy code at the moment
Zoffix scopesoft, Perl 6 has one of the most advanced object models and concurrency support among modern langs. It's a bit silly to say it's ill-fated just because your mate says so.
Plus, this works: 11:01
m: say ⅓²
camelia rakudo-moar : OUTPUT«0.111111␤»
Zoffix ^_^
DrForr_ I could have used that during a talk a few months ago.
Zoffix DrForr_, the powers are brand new. Added just last night :)
DrForr_ I know, saw on FB. 11:02
arnsholt That's more of a novelty thing. More useful, IMO, is this:
scopesoft Given my experience of management in companies they will simply look at such results and base the decision on this. Other languages have OO, are functional, etc Even D is a serious contender
arnsholt m: say 3*0.3 == 0.9 ?? "equal" !! "not equal"
camelia rakudo-moar : OUTPUT«equal␤»
DrForr_ scopesoft: perl5 was slow when it first came out too, as was Ruby, Python and whatever languages you choose to name. If they're that short-sighted then they're .. well, typical management. 11:04
Zoffix m: sleep 1 for ^4 .hyper(:batch(1)); say now - INIT now
camelia rakudo-moar : OUTPUT«1.0103883␤»
DrForr_ Perl5 has had 20 years of optimization to the point where it beats C in some cases. Perl 6 hasn't had 20 years of production yet.
Zoffix scopesoft, ^ four concurrent 1-second sleeps finish in 1 second
mrf m: say ⅓² * 9
camelia rakudo-moar : OUTPUT«1␤»
mrf ^ Only modern language to get this right 11:06
DrForr_ In other words, do *not* confuse release with production.
11:06 ely-se left
arnsholt mrf: Well, Common Lisp gets it mostly right too, to be fair =) 11:06
DrForr_ *cough* modern :)
mrf mostly isn't really good enough. 0.999999 is mostly right 11:07
Zoffix heh
11:07 pierre-vigier joined
arnsholt mrf: Mostly right in that decimal literals are still floating point arithmetic, not rats 11:07
Division creates rats though
DrForr_: If we're talking features, CL is pretty damn modern though (except the syntax, of course, which is atrocious =) 11:08
mrf arnsholt: thats fair.
The Biologist in me would have killed for a language that could have done accurate maths
11:09 pecastro joined
DrForr_ arnsholt: I know, but perception... 11:09
arnsholt DrForr_: I know, I know =)
But CL and Perl 6 are actually very similar, IMO
Woodi hallo today :) 11:11
Zoffix \o
mrf Hello yesterday
Woodi mrf: can you give few examples/requirements ?
dalek href="https://modules.perl6.org/new-db-builder:">modules.perl6.org/new-db-builder: a41ca4e | (Zoffix Znet)++ | lib/ModulesPerl6/DbBuilder/Dist (5 files):
Implement TravisCI PostProcessor
mrf Woodi: ? 11:12
Woodi ...of precise math
mrf m: say ⅓² * 9
camelia rakudo-moar : OUTPUT«1␤»
11:12 pierre-vigier left
Woodi looks exact... 11:12
mrf yes which is awesome. when I was doing BioInfo at uni most languages would give 0.9999 for that which is really annoying 11:13
(including calculators in some cases)
Woodi but do biology curently lack high performance & prcise programming language ? 11:14
mrf Most biologists use either Perl5, Python or Go. 11:15
There are hacks and libraries in all of these to make things mostly work but its a PITA
11:18 ab6tract joined
ab6tract o/ #perl6 11:18
Woodi mrf: sorry for torturin you more but any idea for something better ? like some high level/specialized data types or other things ?
dalek href="https://modules.perl6.org/new-db-builder:">modules.perl6.org/new-db-builder: 3c23291 | (Zoffix Znet)++ | lib/ModulesPerl6/DbBuilder/Dist/Source.pm:
Handle strings in "authors" meta (closes #41)
Zoffix m: say 0/0 + i 11:20
camelia rakudo-moar : OUTPUT«Inf+1i␤»
Zoffix :)
grondilu_ m: say 0/0 11:21
camelia rakudo-moar : OUTPUT«Attempt to divide by zero using div␤ in block <unit> at /tmp/AxNT1KvhAd:1␤␤Actually thrown at:␤ in block <unit> at /tmp/AxNT1KvhAd:1␤␤»
mrf Woodi: Its not something I have been involved in since Uni. As I understand it the maths in perl 6 is ideal as it gives accurate values and can thus be relied upon. I certainly would recommend a better outreach to the biological community in particular though 11:22
Coor vectors would be really handy for atom mapping in 3d space
DrForr_ We have quaternions available :) 11:23
11:23 grondilu joined
grondilu (and geometric algebra) 11:24
Woodi I saw lastly discusions about Rakudo 6.c guessed performance... so, first: in "official" communication there is no need to mention "slowness", concentrate on good parts like correctness or niceness, etc.
11:25 brrt left
mrf will investigate protein modeling in p6 11:25
Woodi mrf: biology and atoms and 3D, wow :)
11:27 rurban left
ab6tract .tell jnthn i had a color to add to the supply/distribute naming .. what about receive / transmit ? 11:27
yoleaux ab6tract: I'll pass your message to jnthn.
mrf Woodi: predictioncenter.org/ 11:28
ab6tract .tell jnthn it has the advantage of conjuring to mind networks / radios / communication vs economics :)
yoleaux ab6tract: I'll pass your message to jnthn.
11:29 pierre-vigier joined, pmurias joined, jesper_ left
ab6tract scopesoft: regarding performance.. you should really consider the historical performance trends for rakudo 11:30
when you can see all that has happened with performance *without even focusing one it*, i think you will get a better picture of where perl 6 can go after the language is finalized
*on it 11:31
Woodi second: Rakudo is not very slow anymore. 30 x perl5 2 months ago (without jit) is "modern" :) and if you check benchmarks then perl5 is 90+ times slower then C# (Mono), sometimes perl5 wins 4 times...
Zoffix Woodi, I disagree. It's the most straightforward comparison you can do to other languages, so I'd expect people to do that right away. And *being surprised* by the slowness might lead to wrong reaction. Sure, it makes sense for first release to be not-optimized, but people—especially language zealots—will take the announcement as "Perl 6 is ready to replace your favourite language X". We don't need to say "it's slow" but we should mention that opti
mization will be the next step.
ab6tract we should also probably compile a list of "known slows"
Zoffix 1) html parsing 11:32
90s vs ~0.8s in P5 on a test pagew
ab6tract m: my $start=now; my int $x = 0; while $x < 5_000_000 { $x += 1}; say "{now - $start} seconds" 11:33
camelia rakudo-moar : OUTPUT«3.7304304 seconds␤»
Woodi Zoffix: I was talking there about official communications/texts. this is like candidating for president position - you not need to make trobles to yourself :)
ab6tract hm.... i had a nested set of native int while loops which would finish more than 2 or 3 times faster than perl 5
crux m: say (-1+0i)**0.5 11:34
camelia rakudo-moar : OUTPUT«6.12323399573677e-17+1i␤»
ab6tract ah, crap, that's containerizing, right?
m: my $start=now; my int $x = 0; while $x < 5_000_000 { $x = $x + 1}; say "{now - $start} seconds" 11:35
camelia rakudo-moar : OUTPUT«0.0311954 seconds␤»
Woodi mrf: looking
ab6tract yeah, there we go
mrf Woodi: no need to really. Just an example of how important protein folding prediction is in BioInfo
ab6tract scopesoft: start doing some while loops with native ints and you can see that the performance "bottom" for perl 6 can really beat perl 5 significantly. the core devs just can't finalize the interior decorating *and* rewire the engine at the same time when there is a deadline looming 11:36
dalek c: ba5c2f0 | (Steve Mynott)++ | doc/Language/faq.pod:
move immutable in p5 example ... oops
 11:37
ab6tract (in my opinion this native int while loop examples should be used as a counterpoint in our "we know it's slow but we can make it better" ... it's visceral, *extant* proof that the gradual typing thing can deliver 11:38
Zoffix has no idea what use those native ints have 11:39
ab6tract Zoffix: that's not the point.
the point is the same language is capable of doing something 1000 times as fast
the same thing, from the programmer's perspective 11:40
so the only reason Int $x doesn't do it as fast as int does is because that hasn't been wired into the optimizer
most benchmarks are inanely arbitrary
sorry, 100x 11:41
11:41 weihan joined
Woodi mrf: nice pictures I can "understand" ;) but also found something about machine learning, do we have something about this in modules ? 11:42
11:43 xpen joined
mrf No that I am aware of. It has however been at least 5 years since I last looked at anything like this. 11:43
(There is usually at least a $5k prize each time for the best prediction model)
Zoffix $5k? 11:44
Zoffix sets out to release Bio::Machine
arnsholt The only ML module I'm aware of is my currently-broken Viterbi decoder for HMMs 11:45
11:45 bjz joined
Zoffix Interesting this topic came up. Toronto Perl Mongers just had a "Machine Learning" meeting last night. I'm yet to watch the vid 11:45
mrf P5 has/had BioPerl which was popular at one time. NumPy is fairly big for lots of things
Woodi have book about machine learning but it was boring becouse only one topic there was statistical function regression...
arnsholt Woodi: That's literally what machine learning is all about though 11:46
11:46 Ven joined
arnsholt There are a couple of exceptions (decision trees, for example), but apart from that it's all regression 11:46
mrf DrForr_: did you get anywhere with social. Still not seen anything! 11:48
Woodi arnsholt: this not sound very complicated but applications like hand writing recognition are very futuristic :) or maybe not any more... 11:49
DrForr_ Um, I still have to go through emails. I was buried doing proposals the night before last and just needed to crash last night. 11:50
And again, given that LPW is happening practically next week I'm not *too* concerned.
mrf DrForr_: fair 11:51
DrForr_ Plus I do have some semiserious shopping to do anyhoo :) 11:53
mrf Thats cool. just planing my weekend as I don't live in London so need to know if I will be travelling 11:54
If not I can go hack in brighton
11:56 zakharyas left
DrForr_ No interest in the social apparently, which is cool. 11:57
12:02 pierre-vigier left 12:03 kjs_ left 12:05 xpen left, zakharyas joined 12:10 Skarsnik joined 12:14 kjs_ joined 12:15 ely-se joined, pierre-vigier joined, Actualeyes joined
dalek kudo/shaped_nonnative_typed_arrays: 131367c | timotimo++ | src/core/Array.pm:
teach TypedArray about shapes
 12:16
Woodi ok, deep learning is more complicated :) 12:18
moritz is a shallow learner 12:19
12:20 |Tux| joined
jnthn afternoon, #perl6 12:24
yoleaux 09:13Z <lizmat> jnthn: the test fail in t/spec/S32-hash/adverbs.t seems to be caused by a MMD issue that only happens when running under the t/harness
09:14Z <lizmat> jnthn: line 58 to be precise
11:27Z <ab6tract> jnthn: i had a color to add to the supply/distribute naming .. what about receive / transmit ?
11:28Z <ab6tract> jnthn: it has the advantage of conjuring to mind networks / radios / communication vs economics :)
jnthn You don't actively receive in a push model though. Stuff just happens to you. :) 12:25
lizmat: There's a test fail in adverbs.t? :)
jnthn didn't know that :)
12:25 kaare_ joined
timotimo RIP hard drive ?!?! 12:28
nope. unplugging and re-plugging helped 12:30
lizmat hi jnthn
nine jnthn: adverbs.t has been failing for weeks here 12:31
lizmat the thing is, it only fails inside the harness. since it is a subtest, I'm not sure *what* fails there
12:31 baest_ is now known as baest
jnthn OK. I've not seen it fail at all here 12:31
lizmat m: my %h = a => 42; say %h<a>:zop
camelia rakudo-moar : OUTPUT«Unexpected named parameter 'zop' passed␤ in block <unit> at /tmp/H4M6nAaRM3:1␤␤»
lizmat that;s the expected failure...
since the test is a throws-like on Exception 12:32
it would seem to me code *doesn't* fail
*the
jnthn Maybe valgrind it.
lizmat not really an option on OSX afaik 12:33
jnthn Nor Windows ;)
12:33 ab6tract left
lizmat then I saw RT #126655 12:33
jnthn lizmat: Will you be about still in 20 mins or so?
lizmat yup
jnthn wants lunch, but wants to ask something about Test::TAP too :)
ok, cool
bbs then
12:35 mls_ is now known as mls 12:36 abraxxa1 is now known as abraxxa, mst left, mst joined 12:38 maddingu1 is now known as maddingue 12:42 molaf joined 12:45 Actualeyes left
timotimo lizmat: i think that comes from having to recompile the .c portions of rakudo 12:47
wait 12:48
i totally misread that
lizmat I was wondering :-)
timotimo anyway. there's a test about typed shaped arrays that doesn't seem quite right
it expects to be able to push and pop a my @foo[5] of Int
lizmat well, this is a simpler test:
12:49 rurban joined
lizmat m: use Test; my %h = a => 42; throws-like '%h<a>:zop', Exception 12:49
camelia rakudo-moar : OUTPUT« 1..2␤ ok 1 - '%h<a>:zop' died␤ ok 2 - right exception type (Exception)␤ok 1 - did we throws-like Exception?␤»
lizmat so this can only fail if the code didn't fail, or threw something other than Exception
both are weird 12:50
12:51 kjs_ left
timotimo fixed his desktop somewhat 12:53
m: my @foos[5] of Int = 1, 2, 3, 4, 5; say @foos.pop; say @foos.perl 12:58
camelia rakudo-moar : OUTPUT«5␤Array[Int].new(1, 2, 3, 4)␤»
timotimo ^- this only worked because the shape was ignored
hmm 12:59
where is that test that was blowing up with my changes?
jnthn You can't .pop a shaped array
Not until we support * in a shape, which we won't in 6.c
timotimo right, "jagged arrays"
jnthn So the best is le wrong.
timotimo good to know! :) 13:00
jnthn uh, test
dalek href="https://modules.perl6.org/new-db-builder:">modules.perl6.org/new-db-builder: 8ba0fa8 | (Zoffix Znet)++ | lib-db-builder/P6Project (5 files):
Sayōnara old db builder classes
 13:05
jnthn lizmat: So, in Test::Tap there's this line: 13:07
for ^$timeout { last if $done or $s.done; sleep .1 }
I can't figure out what $s.done is about there.
It's a method not found error now.
But before it would have been...well, I don't even really know. :)
lizmat ah, I think a thinko
jnthn Whether we'd tapped done?
OK, but just removing it gives me a bunch of hangs 13:08
lizmat no, whether the tap was done
jnthn Ah
lizmat sorry, whether the supply was done
as in: had a .done on it
so, this is just bad, I think
jnthn But since we .tap it with &done always?
OK
13:08 bjz left
jnthn The other thing is 13:08
We used .append in various places
lizmat hmmm....
jnthn Which made supplies magically flatten things 13:09
I've changed them for .push
lizmat fwiw, I have no memory of changing them to append
jnthn So I think some test regressions may be about that too
I don't think you did
dalek href="https://modules.perl6.org/new-db-builder:">modules.perl6.org/new-db-builder: 8a1be68 | (Zoffix Znet)++ | lib/ModulesPerl6/DbBuilder.pm:
Handle fatal errors for each dist
jnthn I think TimToady steamrollered through CORE.setting doing so :)
lizmat so, yes... change them as you see fit :-) 13:10
I mean, they were originally per-GLR pushes
*pre
afaict
jnthn I can't remember what pre-GLR push actually did ;)
But yeah, post-GLR having emit @foo magically flatten feels like a bad idea 13:11
lizmat to come back to $s.done in Test::Tap: that is intended to check whether the supply is done *without* using the normal ways of checking
perhaps we should just use await with a timeout Promise
or the awaiterator :-)
13:12 bjz joined
jnthn Supply.from-list( [1,2],[3,4,5] ).map( {.flat} ) 13:14
This test wanted the flattening behavior for example
lizmat probably not a very good test 13:15
:-(
ilmari m: say 5²³
camelia rakudo-moar : OUTPUT«11920928955078125␤»
ilmari TimToady++
jnthn eek, the Supply.map tests are also often testing map on lists taht are fed to from-list... o.O 13:19
dalek href="https://modules.perl6.org/new-db-builder:">modules.perl6.org/new-db-builder: e5b9e0e | (Zoffix Znet)++ | lib/ModulesPerl6/DbBuilder.pm:
On Win32, exec instead of system because daemon does not daemonize itself
13:19 ely-se left
jnthn No wonder they didn't do what I expect :) 13:19
lizmat
.oO( maybe having no tests would have been better :-( )
jnthn The *majority* of the test files actually worked out pretty fine 13:20
Seems that .from-list never got updated post-GLR, and was doing *@foo, when it probably should follow the single arg rule
lizmat ah, ok... :-) 13:21
dalek kudo/resupply: 1bce55f | jnthn++ | src/core/Supply.pm:
Supply.from-list should follow 1-arg rule.
 13:22
ast/resupply: 6f1e17e | jnthn++ | S17-supply/map.t:
Revise Supply.map tests.
jnthn OK, that's map.t off the list :) 13:23
13:23 ely-se joined, doublec_ joined
dalek href="https://modules.perl6.org/new-db-builder:">modules.perl6.org/new-db-builder: f7d2e86 | (Zoffix Znet)++ | lib/ModulesPerl6/DbBuilder (2 files):
Infer author_id from GitHub username if META does not declare it (Closes #42)
 13:25
13:25 doublec left 13:26 andreoss joined 13:31 colomon left
jnthn Urgh, flat.t shows up something nasty going on with .close closing wrong taps :( 13:31
FROGGS IS THAT ABOUT BEER OR JUST PROGRAMMING??!!11 13:34
arnsholt Why not both? =)
FROGGS AAAAAAAHHHHH 13:35
13:37 kjs_ joined, pierre-vigier left 13:38 lucasb joined
dalek href="https://modules.perl6.org/new-db-builder:">modules.perl6.org/new-db-builder: 685d801 | (Zoffix Znet)++ | lib/ModulesPerl6/DbBuilder.pm:
Check for and delete any logotype files for dists no longer in db
 13:38
13:41 kid51 joined
dalek href="https://modules.perl6.org/new-db-builder:">modules.perl6.org/new-db-builder: 1cbc13c | (Zoffix Znet)++ | lib/ModulesPerl6/DbBuilder.pm:
Separate dist build log output from finishing messages
 13:41
13:42 pierre-vigier joined
tadzik if it's not beer-driven development, let it be at least beer-inspired :) 13:45
dalek href="https://modules.perl6.org/new-db-builder:">modules.perl6.org/new-db-builder: 525b1d1 | (Zoffix Znet)++ | lib/ModulesPerl6 (7 files):
Fix typoes / add stopwords for spellchecker
 13:50
href="https://modules.perl6.org/new-db-builder:">modules.perl6.org/new-db-builder: 6e3a6d8 | (Zoffix Znet)++ | lib/ModulesPerl6/DbBuilder/Log.pm:
We do not need 5.020 here
 13:51
href="https://modules.perl6.org/new-db-builder:">modules.perl6.org/new-db-builder: c65e1dc | (Zoffix Znet)++ | / (2 files):
Our code needs at least perl 5.014
 13:53
13:53 kid51 left
jnthn The further I golf the bug, the weirder it gets 13:53
Woodi btw. about performance: if some simple script executes (from command line to comman line) faster then a blink then it is quite good, whatever a numbers are :) IMO 13:56
13:56 domidumont joined 13:58 Actualeyes joined
dalek href="https://modules.perl6.org/new-db-builder:">modules.perl6.org/new-db-builder: 3abc10c | (Zoffix Znet)++ | build-project-list.pl:
Do not bury constant logos dir in the code
 13:59
13:59 kjs_ left 14:03 colomon joined
dalek href="https://modules.perl6.org/new-db-builder:">modules.perl6.org/new-db-builder: 7c11262 | (Zoffix Znet)++ | / (4 files):
Move all executables into /bin/
 14:07
14:14 doublec joined, skids joined
jnthn D'oh, found it, I think. 14:15
14:16 doublec_ left 14:20 pmurias_ joined 14:21 Actualeyes left, weihan left 14:23 pmurias left 14:25 Actualeyes joined 14:30 sno joined
lizmat jnthn: sanity check: is .WHERE of an object guaranteed to be constant for its lifetime? 14:30
jnthn: or can a gc change it ?
timotimo jnthn: github.com/perl6/roast/blob/master....t#L61-L74 - so i can basically throw this out? 14:31
jnthn lizmat: .WHERE can (will :-)) change
lizmat: nqp::objectid is the stable thing
(Carries some cost to maek that promise) 14:32
dalek ast/resupply: fc63260 | jnthn++ | S17-supply/grep.t:
Fix now from-list follows 1-arg rule.
lizmat ok, that means that we have a problem in the way that gistseen/perlseen work
jnthn Oh, it's not using .WHICH? 14:33
lizmat it *is* for the key, but not for the representation in .perl and .gist
m: my $a = [42]; $a[1] = $a; say $a.perl
camelia rakudo-moar : OUTPUT«(my \Array_140146618956656 = $[42, Array_140146618956656])␤»
jnthn oh 14:34
yeah, I'd use objectid there
lizmat ok, will fix that
dalek kudo/resupply: 7f8030b | jnthn++ | src/core/Supply.pm:
Fix tap closing scoping bugs.

Could lead to closing of the wrong taps.
href="https://modules.perl6.org/new-db-builder:">modules.perl6.org/new-db-builder: f3e1638 | (Zoffix Znet)++ | / (4 files):
Improve coverage tests
 14:37
14:37 kjs_ joined
timotimo those numbers look like unix timestamps to me for no particular reason 14:37
lizmat jnthn: is there a reason to say 'my int = 0' when we know it's going to be 0 ?
jnthn timotimo: If they are, I think they're from a good way into the future ;) 14:38
lizmat: No more than "I like the explicitness" 14:39
lizmat jnthn: ok
arnsholt I was about to suggest "explicit is better than implicit" =)
dalek ast/resupply: d2fccdf | jnthn++ | S17-supply/ (4 files):
More fixes for .from-list fallout.
 14:40
lizmat hmmm... oddly enough, explicite in this case seems to be faster as well
jnthn Down to 9 S17-supply tests with issues now 14:41
Most due to the Test::TAP done thing
Uh, Tap 14:42
heh
That's an unfortunate confusion :)
lizmat bikeshed away, I would say :) 14:43
jnthn Can't be bothered :)
jnthn removes the $s.done and sees what happens 14:44
lizmat jnthn: the $s.done test is to prevent hanging 14:45
if a test fails for some reason
so if everything is ok, you won't see a difference 14:46
if not, you might see a hang
dalek p: ad87dc9 | (Ronald Schmidt)++ | tools/build/Makefile-common.in:
Update Makefile-common.in to work with mingw

Fix /bin/sh: C:Strawberryperlbinperl.exe: command not found errors with mingw build without (hopefully) breaking other builds.		 14:47
p: 71c6fd4 | (Jimmy Zhuo)++ | tools/build/Makefile-common.in:
Merge pull request #259 from ronaldxs/master

Update Makefile-common.in to work with mingw
jnthn lizmat: ah, ok 14:48
14:48 gtodd1 is now known as gtidd
jnthn Hm, I busted zip and zip-latest, it seems 14:48
14:49 gtidd is now known as gtodd
timotimo at least zip-latest has always been problematic 14:50
it kept oversharing values from earlier :| 14:51
gfldex m: sub range_rand(\min, \max){ (max - min).rand + min }; sub range_roll(\min, \max, \times){ range_rand(min,max) xx times }; say range_roll(0.1,0.3,10); #126664
camelia rakudo-moar : OUTPUT«(0.114553204786117 0.282474299596918 0.199445260961036 0.24923524175062 0.193473323053126 0.131453337177036 0.299215762973047 0.193697580865753 0.169199240435266 0.214324695974161)␤»
gfldex lizmat: ^^^ that may be sensible too (re: #126664) 14:52
jnthn duh, what a silly mistake...
14:53 zacts joined
lizmat is curious 14:53
14:54 xpen joined
zacts yo perl6 hackers 14:55
and nerds and geeks et al...
14:55 bjz left
tadzik hello zacts 14:55
zacts lo 14:56
dalek kudo/resupply: 39db5e3 | jnthn++ | src/core/Supply.pm:
whenever over correct supplies in zip.
jnthn So, zip is happy now 14:57
hmm, I think the zip-latest fail is a thinko in the test 14:58
it does a .done and then a .emit on the same supply
DrForr_ wr0d.
jnthn Pretty sure that's a thinko. :) 14:59
14:59 xfix joined
lizmat I'll buy that 14:59
dalek ast/resupply: 5eb47dc | jnthn++ | packages/Test/Tap.pm:
Remove hang check that no longer works.

Shouldn't affect the timeout, though.		 15:00
ast/resupply: cafea57 | jnthn++ | S17-supply/zip-latest.t:
Fix zip-latest test thinko.

It likely meant to write to the not-yet-done supply; we now enforce protocol.
15:00 doublec_ joined 15:01 doublec left 15:04 g4 left 15:15 doublec joined 15:16 doublec_ left, khw joined 15:21 doublec_ joined, doublec left 15:27 MadcapJake joined 15:37 pierre-vigier left
dalek href="https://modules.perl6.org/new-db-builder:">modules.perl6.org/new-db-builder: 9d01ad3 | (Zoffix Znet)++ | / (4 files):
Improve ::Root controller test coverage to 100%
 15:38
ast: 5f938a6 | timotimo++ | S02-types/array-shapes.t:
kick out wrong tests
kudo/nom: 131367c | timotimo++ | src/core/Array.pm:
teach TypedArray about shapes
 15:39
kudo/nom: 8618f41 | timotimo++ | src/core/Array.pm:
Merge branch 'shaped_nonnative_typed_arrays' into nom

my Int @foo[10;20] and friends now work.
jnthn \o/ 15:41
timotimo++ 15:42
15:45 jonas1 left
dalek ast/resupply: 8bd01a8 | jnthn++ | S17-supply/c (2 files):
Mark "done" at right point in some tests.
 15:48
kudo/resupply: d6ce4a4 | jnthn++ | src/core/Supply.pm:
Add Supplier::Preserving.

If the supply has not been tapped yet, then preserves any emitted events (emit/done/quit) until a tap happens, then feeds them all to that tap. Follow-up taps continue to get the live supply.
kudo/resupply: abb5da2 | jnthn++ | src/core/Supply.pm:
Use Supplier::Preserving in classify/categorize.
 15:49
kudo/resupply: 420a716 | jnthn++ | src/core/Supply.pm:
Fix arity bug in .tap's emit default.
kudo/resupply: b50f48c | jnthn++ | src/core/Supply.pm:
Don't try and clean up a tap we don't have.
ast/resupply: f5c1886 | jnthn++ | S17-supply/produce.t:
Add tests for Supply.produce.

Which does what Supply.reduce used to, since produce is the "triangle" form on lists, so it should be on supplies too.		 15:53
ast/resupply: af3d634 | jnthn++ | S17-supply/reduce.t:
Update Supply.reduce tests to match semantics.
 15:55
lucasb m: say '<--'
camelia rakudo-moar : OUTPUT«<--␤»
lucasb It took me some days to notice that camelia is not printing the revision number
jnthn hmm
15:55 FROGGS left
jnthn Phew, down to 3 supply test files needing attention 15:56
lizmat jnthn++
afk for a few hours&
lucasb maybe this commit changed something: github.com/perl6/evalbot/commit/19...6e91b7be74
15:57 xpen left 15:58 xpen joined
dalek ast/resupply: 569072c | jnthn++ | S17-supply/unique.t:
Missing flat, .done calls.
 15:59
jnthn Now it's 2. 16:01
16:03 sjn_phone joined
jnthn Darn, I've got to understand throttle 16:04
16:04 samb1 joined
lizmat is not away after all 16:05
can I help you with that ?
jnthn: ^^^
16:05 jonas1 joined
lizmat doc.perl6.org/routine/throttle # jnthn 16:07
hmmm... needs a little cleanup
jnthn lizmat: I think I can re-work it not to use "on"
16:07 zakharyas1 joined 16:08 n0xff left, zakharyas left
dalek c: 2bc558b | lizmat++ | doc/Type/ (2 files):
Cleanup some =over and =back
 16:09
16:14 BenGoldberg joined 16:15 sjn_phone left 16:18 cschwenz joined, cschwenz left
jnthn lizmat: A couple of usage examples would be really great in the docs ;) 16:19
jnthn compiles the refactored-to-use-whenever version 16:20
lizmat hmmm... I could have sworn I had done some
will do after dinner 16:21
jnthn Darn typo...
:)
jnthn tries again
I suspect I'll have to re-work the tests a bit
'cus of the supply/supplier distinction we now have 16:22
DrForr_ YAPC::EU Aug 24-26. Booking my flights now :)
jnthn :)
16:22 sQuEE` is now known as sQuEE
grondilu wonders if it would be reasonable to set [+] 1/2, 1/4, 1/8 ... * to be 1 16:23
(or more generally to use the explicit formula for the sum of a geometric sequence) 16:24
(I mean since it's been done for arithmetic sequences...)
lizmat hmmm... another use for ^1 :) 16:25
16:26 ely-se left
timotimo grondilu: it's only useful if we have a prover for any given sequence, otherwise it'd really only be that literal sequence summed up that'd result in 1 16:27
lizmat m: my \a = [23,42,a]; dd a # this feels incorrect
camelia rakudo-moar : OUTPUT«Array $var = $[23, 42, Mu]␤»
lizmat dinner& 16:28
timotimo there's no container we can put in there to have the array later, eh? 16:29
16:30 xpen left 16:31 baest_ joined
dalek p/js-cps: dd4f0de | (Pawel Murias)++ | src/vm/js/bin/run_tests.pl:
[js] Add test 87 to the list of run tests.
 16:31
p/js-cps: 3dffa81 | (Pawel Murias)++ | src/vm/js/ (3 files):
[js] Use the :ctx named parameter for ops that take a ctx.
p/js-cps: b272c80 | (Pawel Murias)++ | src/vm/js/ (4 files):
[js] Start of work on compiling code in CPS transformed way.
16:32 kjs__ joined 16:33 jesper_ joined, kjs__ left 16:34 Ven left, moznion_ joined 16:35 apejens_ joined, Zoffix_ joined, btyler_ joined, yubimusubi joined, perlpilot joined, M-Illandan left, zakharyas1 left, moznion left, vytas left, baest left, raydiak left, rindolf left, apejens left, Zoffix left, PerlJam left, btyler left, Possum left, Actualeyes left, Actualeyes joined 16:36 kjs_ left
dalek href="https://modules.perl6.org/new-db-builder:">modules.perl6.org/new-db-builder: df6be01 | (Zoffix Znet)++ | / (3 files):
Increase test coverage to 100% for ModulesPerl6.pm
 16:36
16:36 rindolf joined, vytas joined
grondilu timotimo: all geometric sequence have an exact formula for their sum, and infix:<...> can recognize geometric sequences. 16:37
16:37 M-Illandan joined
lucasb Interesting, --target=ast can produce some errors that normal perl6 doesn't seem to suffer 16:38
perl6 --target=ast -e 1 > /dev/null #=> ===SORRY!=== Failed to flush filehandle: invalid argument
perl6 --target=ast -e 'sub f {}' #=> Sub object coerced to string ... in any dump_node_list at gen/moar/stage2/QASTNode.nqp:81
dalek kudo/resupply: 0a48b97 | jnthn++ | src/core/Supply.pm:
Re-work throttle to use whenever.

At least, so far as the tests test it.
timotimo grondilu: and then the user relies on that, accidentally uses real instead of rat and wonders why there's an infinite loop in their program? :)
16:39 ribasushi left
grondilu not sure what you mean. The formula remains true with reals. 16:40
jnthn lucasb: Still outputs the AST in the second case though :)
And the first is useless anyway
grondilu as long as the geometric sequence is recognized, you're fine.
lucasb jnthn: 'useless' is such a strong word :)
16:41 colomon left 16:43 domidumont left 16:44 ribasushi joined
stmuk is anyone seeing an odd error with ecosystem modules and prove being "unable to fork"? 16:47
16:48 ZoffixW joined
ZoffixW lucasb, yup, you're right. Fix incoming 16:49
lucasb ZoffixW: yeah, tumuch shell escaping, thanks :)
dalek albot: 0bc6e6f | (Zoffix Znet)++ | evalbot.pl:
Give get_revisions_from_file proper path since now we are escaping shell stuff, so ~ is not special
 16:50
ZoffixW ^ that should do it. If anyone can pull and restart
RabidGravy stmuk, nope, though that's not to say it's not happening :) 16:55
16:56 camelia joined
lucasb m: say 'hi' 16:56
16:58 ChanServ sets mode: +v camelia
camelia rakudo-moar 8618f4: OUTPUT«hi␤» 16:58
arnsholt jnthn: I've managed to segfault Moar with NativeCall. Investigating now, but looks like CStruct.c's initialize() gets called with an STable whose REPR data is 0x0
ZoffixW \o/
16:58 sQuEE is now known as sQuEE`
lucasb ZoffixW: :D 16:58
17:03 rigsby joined
dalek ast/resupply: 6cde112 | jnthn++ | S17-supply/throttle.t:
Updates for supply/supplier distinction.
 17:03
ast/resupply: a819aa1 | jnthn++ | packages/Test/Tap.pm:
A (over-simple) fake scheduler.

To allow us to write tests using virtualized time, rather than sleep.
ast/resupply: 1f11585 | jnthn++ | S17-supply/interval.t:
Use virtual time in Supply.interval tests.
jnthn .tell lizmat If you're looking for something to hack on with supplies, then re-working some of the supply tests that sleep to use virtual time would be good. Feel free to advance the scheduler a bit too; it's the simplest thing that could possibly work. :) 17:04
yoleaux jnthn: I'll pass your message to lizmat.
17:04 colomon joined
jnthn At some point we'll want a good virtual time scheduler in the ecosystem too, so everyone can write decent tests :) 17:05
17:05 pmurias_ left, AlexDaniel joined 17:07 rigsby left, rigsby joined
dalek ast/resupply: bcb3200 | jnthn++ | S16-io/supply.t:
IO::Handle.Supply is not live.
 17:13
ast/resupply: 2408bce | jnthn++ | S02-types/WHICH.t:
Fix WHICH.t: remove unknown type, add new one.
c: cf18697 | (Zoffix Znet)++ | CONTRIBUTING.md:
Update CONTRIBUTING.md

Link to en.wikipedia.org/wiki/Active_voice for those who're not familiar with the concept		 17:14
17:14 kjs_ joined 17:15 doublec_ left
dalek ast/resupply: c82b01a | jnthn++ | integration/advent2013-day19.t:
Update for supply/supplier distinction.
 17:16
jnthn
.oO( The active voice was encouraged. )
17:17 doublec joined
ZoffixW :D 17:17
dalek c: f94185e | (Zoffix Znet)++ | doc/Language/io.pod:
Reword / typo
 17:18
17:18 emdashcomma joined 17:21 telex left, gtodd left 17:22 telex joined, pat_js joined
dalek c: 8feb8de | (Zoffix Znet)++ | doc/Language/io.pod:
Reword
 17:24
17:25 kjs_ left
TimToady m: my (\d); d = 42; dd d 17:27
camelia rakudo-moar 8618f4: OUTPUT«Int $var = 42␤»
TimToady lizmat: this might be related to the $ that happens to the array 17:28
my (\d) should probably fail without a bindee
17:30 abraxxa left
dalek c: 94e045f | (Zoffix Znet)++ | doc/Language/io.pod:
Reword
 17:31
c: 2ca1886 | (Zoffix Znet)++ | doc/Language/io.pod:
Do not explicitly suggest to check for file"s existence before reading

It"s not a suitable check due to race conditions and errors when opening for read aren"t as evil as the current wording suggests
lizmat .botsnack 17:32
yoleaux 17:04Z <jnthn> lizmat: If you're looking for something to hack on with supplies, then re-working some of the supply tests that sleep to use virtual time would be good. Feel free to advance the scheduler a bit too; it's the simplest thing that could possibly work. :)
:D
jnthn m: module foo; sub foo() { say "OH HAI" }; foo; foo()
camelia rakudo-moar 8618f4: OUTPUT«5===SORRY!5=== Error while compiling /tmp/I3bZzMcWtU␤Semicolon form of 'module' without 'unit' is illegal. You probably want to use 'unit module'␤at /tmp/I3bZzMcWtU:1␤------> 3module foo;7⏏5 sub foo() { say "OH HAI" }; foo; foo()␤»
jnthn m: module foo { }; sub foo() { say "OH HAI" }; foo; foo()
camelia ( no output )
jnthn TimToady: In ^^, foo() ends up being a coercion type literal 17:33
TimToady: Not quite sure what we should do, if anything, here
That's rt.perl.org/Ticket/Display.html?id=115608. And related is rt.perl.org/Ticket/Display.html?id=113892.
m: module A { enum Day is export <Mon Tue>; sub Day is export { 'sub Day' } }; import A; say Day(0); 17:34
camelia rakudo-moar 8618f4: OUTPUT«Mon␤»
jnthn There's no symbol table collision in either case, 'cus subs go in as &foo and types as foo
lizmat m: my %hash = a => 1; say item(%hash).perl # I think this should be "${:a(1)}" 17:35
camelia rakudo-moar 8618f4: OUTPUT«{:a(1)}␤»
lizmat ^^^ TimToady jnthn do you agree ?
17:35 pat_js left
dalek c: b59e3f3 | (Zoffix Znet)++ | doc/Language/io.pod:
Reword / typos
 17:36
jnthn lizmat: I'd say so
lizmat multi sub item(\x) { my $ = x } # the appropriate item() candidate
jnthn Yeah, suspect the .perl is wrong
lizmat ok, then I'll fix the test :-)
yeah, I'm reworking the perlseen bits, and after that this test failed 17:37
17:37 rigsby left
lizmat so it seems I also fixed a bug :-) 17:37
zengargoyle good * #perl6
lizmat zengargoyle o/
dalek c: 8ecbc76 | (Zoffix Znet)++ | doc/Language/io.pod:
Typo
ZoffixW \o
dalek c: 5c45e67 | (Wenzel P. P. Peppmeyer)++ | doc/Type/IO/Handle.pod:
clarify IO::Handle.getc
 17:39
c: 5473320 | (Zoffix Znet)++ | doc/Type/IO/Handle.pod:
Merge pull request #214 from gfldex/master

clarify IO::Handle.getc
17:39 vendethiel joined
dalek kudo/nom: 1c36fb7 | lizmat++ | src/core/ (7 files):
Generalize .perlseen/.gistseen handling some more
 17:40
ast: 6a16a67 | lizmat++ | S02-types/hash.t:
Fix item %hash test
 17:41
17:42 lucasb left 17:43 sufrostico joined 17:44 jesper_ left
dalek href="https://modules.perl6.org/new-db-builder:">modules.perl6.org/new-db-builder: b84aa37 | (Zoffix Znet)++ | lib/ModulesPerl6/DbBuilder/Dist.pm:
Handle case when Dist Source fails and returns undef
 17:46
17:47 khisanth_ is now known as Khisanth, sufrostico left 17:59 BenGoldberg left 18:00 rindolf left
TimToady jnthn: re foo vs &foo, it almost doesn't matter which way we do it, since you can work around with &foo() the one direction, and ().foo the other direction; a middle way would be to allow 'undeclaring' a typename in an inner scope, which we can't do right now, but the status quo seems okay to me, where type name declarations always take precedence 18:02
m: module foo { }; { sub foo() { say "OH HAI" }; my \foo = &foo; foo; &foo() } 18:03
camelia rakudo-moar 1c36fb: OUTPUT«OH HAI␤»
TimToady m: module foo { }; { sub foo() { "OH HAI" }; my \foo = &foo; say foo; say &foo() }
camelia rakudo-moar 1c36fb: OUTPUT«sub foo () { #`(Sub|59510816) ... }␤OH HAI␤»
dalek kudo/resupply: 986b6e9 | jnthn++ | src/core/Supply.pm:
Rename last -> tail, align semantics.
kudo/resupply: d5ab9bf | jnthn++ | t/spectest.data:
Chase test file renaming.
TimToady well, that almost works as a type override... 18:04
m: module foo { }; { sub foo() { "OH HAI" }; my \foo = &foo; say foo; say foo() }
camelia rakudo-moar 1c36fb: OUTPUT«sub foo () { #`(Sub|67694576) ... }␤OH HAI␤»
TimToady lets you drop the & anyway
dalek ast/resupply: 108337d | jnthn++ | S17-supply/ (2 files):
Chase Supply.last -> Supply.tail.
jnthn OK, thanks.
Those'll be two very RTs to resolve them :) 18:05
Just tests for current behavior and close 'em.
TimToady so I'd say Not A Bug
jnthn *very easy
18:05 rindolf joined
TimToady now imagines 'anon class foo {}' as a way to hide the outer definition of foo 18:06
dalek ast/resupply: a1eb506 | jnthn++ | S02-types/WHICH.t:
Toss two types that are gone.
 18:07
TimToady you-don't-see-me class foo {}
jnthn :)
Gonna merge resupply after this spectest run
May as well have the upheavel sooner rather than later 18:08
18:08 TEttinger joined
zengargoyle m: my \now = 1; say DateTime.now; 18:09
camelia rakudo-moar 1c36fb: OUTPUT«2015-11-26T19:09:10+01:00␤»
dalek kudo/resupply: df06a24 | jnthn++ | t/spectest.data:
Toss S17-supply/on.t, add S17-supply/produce.t.
nine jnthn: speaking of merging: what do you see as requirements for merging curli?
18:09 leont joined
jnthn nine: Spectests in shape, Panda still installs modules (Task::Star is a good check) 18:10
18:10 kjs_ joined
jnthn If there has to be a Panda patch too that's also fine 18:10
zengargoyle guess totally fooling the concept of now is a different matter. 18:12
dalek href="https://modules.perl6.org/new-db-builder:">modules.perl6.org/new-db-builder: 6cd62dd | (Zoffix Znet)++ | lib/ModulesPerl6/DbBuilder/Dist/Source.pm:
Use "url" throughout
 18:16
jnthn nine: I'd generally prefer "sooner rather than later" 18:17
dalek href="https://modules.perl6.org/new-db-builder:">modules.perl6.org/new-db-builder: 29ec6ff | (Zoffix Znet)++ | lib/ModulesPerl6/DbBuilder/Dist/Source.pm:
Fix bug where dists can override private build info via rogue META keys
jnthn nine: So once you feel comfortable with the state of it, then I'm fine. 18:18
dalek Heuristic branch merge: pushed 21 commits to rakudo/nom by jnthn 18:20
Heuristic branch merge: pushed 19 commits to roast by jnthn
nine jnthn: ok :)
jnthn resupply merged :)
RabidGravy boom!
nine Seems like there are modules in Task::Star that modify @*INC (at least in tests) 18:21
ZoffixW \o/ jnthn++
Pluggable relies on it
(probably not in Task::Star tho)
jnthn Ah, OK
Well, we can't do much about that I guess.
ZoffixW Yeah 18:22
jnthn Better to break things sooner, so they can be fixed sooner.
RabidGravy (breaking things)++ 18:23
18:23 domidumont joined, AlexDaniel left, kjs_ left
jnthn At least folks know that in less than a month, the churn will be over. :) 18:23
nine You mean I don't get to merge it the day before the release and vanish for a couple of weeks vacation?
jnthn :P 18:24
ZoffixW :P
jnthn I'd rather we merge it soon, but vanishing for vacation after the release is quite sane. :)
jnthn sure as heck will be
OK, off to cook :) 18:25
18:25 vendethiel- joined
dalek href="https://modules.perl6.org/new-db-builder:">modules.perl6.org/new-db-builder: f6c3592 | (Zoffix Znet)++ | lib/ModulesPerl6/DbBuilder/Dist/Source.pm:
Delete all rogue META keys that might interfere with our build code
 18:26
18:32 lucasb joined, rigsby joined 18:33 adu joined
lucasb merge all the things /o/ 18:33
and let's not forget to branch perl 7 development after xmas :D
18:33 Bucciarati left 18:34 vendethiel left 18:37 kjs_ joined 18:38 raydiak joined 18:39 emdashcomma left 18:41 emdashcomma joined 18:46 emdashcomma left
Quom_ Did anyone else notice this issue: 18:47
m: sub f($a) {$a**2}; say [+] (^100).hyper.map(&f)
camelia rakudo-moar 51d695: OUTPUT«328350␤»
Quom_ m: multi f($a) {$a**2}; say [+] (^100).hyper.map(&f)
camelia rakudo-moar 51d695: OUTPUT«0␤»
Quom_ m: multi f($a) {$a**2}; say [+] (^100).map(&f)
camelia rakudo-moar 51d695: OUTPUT«328350␤»
TimToady I think we noticed something like it the other day, dunno if anybody rakudobugged it 18:48
18:49 bowtie left 18:51 adu left 18:53 charsbar__ joined 18:56 zoosha_ joined
Quom_ RT #126597 18:58
18:58 krunen_ joined
Quom_ I thought we had a bot for that? 18:58
19:00 Timbus_ joined 19:01 skaji_ joined
Quom_ TimToady: Looks like we encountered a race condition 19:02
19:02 rurban left 19:04 raydiak left, lucasb left, rindolf left, ribasushi left, vytas left, moznion_ left, zacts left, skids left, andreoss left, molaf left, espadrine left, charsbar_ left, zoosha left, maddingue left, vike left, Ulti left, mattp__ left, Khisanth left, Upasaka left, Timbus left, krakan left, atta left, bakedb left, avar left, confuseAcat left, burnersk left, jordman_ left, yoleaux left, Jonis left, solarbunny left, pnu left, ZoffixW left, MilkmanDan left, corbyhaas left, frew left, vendethiel- left 19:05 M-Illandan left, jonas1 left, mtj_ left, breinbaas left, stmuk left, f3ew left, tadzik left, daxim left, novapatch left, stux|RC-only left, lizmat left, pdcawley_ left, Gruber left, grondilu left, leont left, telex left, baest_ left, samb1 left, pecastro left, boegel left, skarn left, JimmyZ left, olinkl left, dustinm` left, kipd left, clkao left, revdiablo left, zostay_ left, ShimmerFairy left, flaviusb left, tony-o left, TimToady left, oka_ left 19:06 yakudza left, gensym left, Gardner left, ilbot3 left, freeze left, khw left, M-eternaleye left, obra left, esh left, dalek left, PotatoGim left, DrPete_ left, cgfbee left, hacst left, mathw left, pierrot left, ugexe left, sjn left, havenwood left, advwp_ left 19:07 TEttinger left, btyler_ left, colomon left, Zoffix_ left, MadcapJake left, xfix left, Skarsnik left, scopesoft left, El_Che left, thundergnat left, tinita left, darutoko left, pochi left, lucs_ left, moritz left, mls left, yeltzooo left, saaki left, kjs_ left, domidumont left, doublec left, bitmap joined 19:08 Peter_R joined, rjbs joined, Humbedooh joined, ribasushi joined, rindolf joined, jordman joined, vytas joined, diakopter joined, brabo joined, CAAA5FQ joined, Jonis joined, Celelibi joined, Util joined, mspo_ joined, jantore_ joined, emdashcomma joined, krakan_ joined, Hor|zon_ joined, atta_ joined, adhoc_ joined, cxreg2 joined, broquain1 joined, Sgeo__ joined, Ulti_ joined, btyler joined, adu joined, leont joined, telex joined, ZoffixW joined, M-Illandan joined, baest_ joined, jonas1 joined, samb1 joined, pecastro joined, boegel joined, mtj_ joined, breinbaas joined, stmuk joined, solarbunny joined, pnu joined, skarn joined, MilkmanDan joined, f3ew joined, frew joined, corbyhaas joined, clkao joined, tadzik joined, daxim joined, JimmyZ joined, revdiablo joined, novapatch joined, olinkl joined, zostay_ joined, dustinm` joined, stux|RC-only joined, ShimmerFairy joined, flaviusb joined, tony-o joined, lizmat joined, pdcawley_ joined, TimToady joined, oka_ joined, kipd joined, yakudza joined, gensym joined, Gardner joined, Gruber joined, ilbot3 joined, freeze joined, sftp joined, Woodi joined, Gothmog_ joined, bonsaikitten joined, huf joined, tudorcon1tantin joined, b2gills joined, uruwi_ joined, retupmoc_ joined, DrForr_ joined, grondilu joined, leedo joined, autarch joined, garu_ joined, sitaram_ joined, domm joined, CurtisOvidPoe joined, noganex joined, flussence joined, Hotkeys joined, hobbs joined, DarthGandalf joined, [Tux] joined, royalkiran007 joined, sammers joined, ir2ivps4 joined, gypsydave5 joined, geekosaur joined, simcop2387 joined, ycaymanbo joined, shamu_ joined, Praise joined, erdic joined, gfldex joined, ab5tract joined, japhb joined, Guest51288 joined, DrParis joined, cibs joined, zhmylove_ joined, isacloud joined, riatre_ joined, pyrimidine joined, Psyche^_ joined, arnsholt joined, [Coke] joined, BinGOs joined, lsm-desktop joined, silug joined, Spot__ joined, ponbiki joined, ggherdov joined, SmokeMachine_ joined, xdbr joined, hoelzro joined, robinsmidsrod joined
TimToady anyway, I reported the HyperSeq bug, should have an RT shortly 19:09
19:09 xenu joined, vendethiel joined, zacts joined, camelia left, emdashcomma_ joined
TimToady camelia, Excess Flood? 19:09
19:09 molaf joined 19:10 go|dfish joined, abaugher joined, sunnavy_ joined, Upasaka_ joined, dpk joined, ssflying_ joined, KotH_ joined, bakedb_ joined, kjs_ joined, domidumont joined, doublec joined, ilbelkyr joined, eiro joined
Quom_ TimToady: As I said before the netsplit ended, there's already a ticket in rt.perl.org/Ticket/Display.html?id=126597 19:10
19:10 go|dfish left, go|dfish joined, ilbelkyr left, skids joined
TimToady ah, well, mine's a dup then 19:11
19:11 xfix joined, dgl joined, nine_ joined, rhr_ joined, moznion joined, Brock joined, maddingu1 joined, jnthn_ joined, jojotus_ joined, vike1 joined, zengargo1le joined, smash_ joined, notostraca joined, Zoffix_ joined, MadcapJake joined, Skarsnik joined, scopesoft joined, El_Che joined, thundergnat joined, tinita joined, darutoko joined, pochi joined, lucs_ joined, moritz joined, mrf joined, mls joined, jevin_ joined, salva joined, johan joined, nowan joined, larion joined, bbarker` joined, bhm joined, woodruffw joined, hahainternet joined, SHODAN joined, _sri joined, dbohdan joined, nebuchadnezzar joined, jdv79 joined, Axord joined, diegok joined, ashleydev joined, ingy joined, Exodist joined, yeltzooo joined, saaki joined, sergot joined
TimToady better too many going for the ball than too few :) 19:11
Quom_ Hah, true :)
19:11 ilbelkyr joined
Quom_ afk 19:11
19:11 lucasb joined 19:12 kjs_ left, Khisanth joined 19:13 notostraca is now known as TEttinger, Amnez777 joined, zengargo1le is now known as zengargoyle 19:14 MadcapJake left, raydiak joined, MadcapJake joined, khw joined, M-eternaleye joined, obra joined, esh joined, dalek joined, PotatoGim joined, DrPete_ joined, cgfbee joined, hacst joined, mathw joined, pierrot joined, ugexe joined, sjn joined, havenwood joined, advwp_ joined, luis joined, masak joined, cfloare joined, chewie_2082 joined, risou joined, avalenn joined, jme` joined, Juerd joined, mephinet joined, roguelazer joined, jsimonet joined, mrsolo joined, ilmari joined, integral joined, zemmihates joined, chansen_ joined, timotimo joined, [ptc] joined, xnrand joined, dylanwh joined, orwell.freenode.net sets mode: +v dalek 19:15 PotatoGim left, cgfbee left, Amnez777 left, Amnez777 joined, doublec_ joined
ZoffixW Man, the search on irclog.perlgeek.de/perl6/today really sucks. It can't find what I want 70% of the time, even though I 100% know the terms I search for were said. 19:16
19:16 Sgeo_ joined, KotH_ is now known as KotH, camelia joined
ZoffixW Anyone remember who suggested "Koality" instead of "Kwalitee"? 19:16
19:16 doublec left
TEttinger like koala bears? 19:17
ZoffixW Yeah :)
19:17 lucasb_ joined
TEttinger they're cute lil fellers 19:17
19:18 sufrostico joined
ZoffixW cdn.meme.am/instances/59188417.jpg 19:18
skids ZoffixW: that search has its issues but considering the state of the art has been steadily regressing for the last decade, it is ahead by not getting worse. :-)
ZoffixW heh
19:19 molaf_ joined, zacts_ joined, cxreg joined
arnsholt m: say "alive?" 19:19
Oh well 19:20
ZoffixW Careful! Last time you did that we got a giant netsplit :P
19:23 ponbiki_ joined 19:24 PotatoGim joined, emdashco- joined, mspo joined 19:25 jordman is now known as 16WAASGSD, rjbs is now known as 18WAA8QTX 19:26 Bucciarati joined, vendethiel- joined, emdashco| joined
RabidGravy OOPS! travis-ci.org/sergot/http-useragen...s/93413230 19:26
19:26 emdashcomma_ left, emdashcomma left
lucasb_ ZoffixW: I sometimes use google to search the irclog: www.google.com/search?q=site:irclo...1+someword 19:26
19:26 lucasb_ left, ribasushi left, vendethiel left, 18WAA8QTX left, broquain1 left, mspo_ left, jantore_ left, adhoc_ left 19:27 jantore joined, rjbs joined, literal joined, jordman joined, avar joined, wtw joined, mattp__ joined, confuseAcat joined
ZoffixW lucasb, I tried that, but did not find what I was looking for 19:27
19:27 Actualeyes left, Khisanth left, lucasb left, molaf left, zacts left, vytas left, cxreg2 left, Sgeo__ left
RabidGravy anyone got an insight as to what that "MVMArray: Index out of bounds" might mean before I chase it down? 19:27
19:27 wtw left
TimToady looks like camelia didn't get voiced... 19:27
19:27 ribasushi joined 19:28 ChanServ sets mode: +v camelia
ZoffixW m: say ¼² 19:28
19:28 Peter_R left
TimToady but maybe that's not the problemo 19:29
19:29 xenu left, rindolf left, Hor|zon_ left, atta_ left, Ulti_ left, btyler left 19:30 Util left, krakan_ left, leont left, telex left, baest_ left, pecastro left, mprelude joined 19:31 Hor|zon joined, vytas joined, sufrosti1o joined
Skarsnik someone know why DBI do/execute return '0E0' instead of just ... 0 ? 19:31
19:31 jordman_ joined, Util joined
RabidGravy 0 but true 19:31
19:31 krakan joined, telex joined 19:32 adhoc joined
nine_ jnthn_: from a user's perspective the most annoying thing about curli is that it won't let me install an already installed dist. Like the CORE dist that's installed as part of the build process. Or when panda's module data is out of sync with the really installed dists... 19:32
19:32 sjohnson joined
RabidGravy Skarsnik, so it succeeded but returned 0 19:33
dalek kudo/curli: 7c948ef | (Stefan Seifert)++ | src/core/CompUnit/PrecompilationRepository.pm:
Fixup error message when precompilation failed
Skarsnik why not returning True?
ZoffixW Skarsnik, because it's indicating that 0 things were affected
Skarsnik, 0 and false would mean an error happened. 19:34
19:34 nwc10 joined
ZoffixW And I'm suspecting this might be an artifact from P5's implementation, if there's a saner way to do that sort of thing in P6 19:34
nwc10 jnthn_: you're aware that t/spec/S17-supply/categorize.t
now aborts (for me)
19:34 broquaint joined
nwc10 #2 0x00007ffff66ab9d1 in uv_mutex_destroy (mutex=0x604000528cd0) at 3rdparty/libuv/src/unix/thread.c:114 19:34
#3 0x00007ffff6530a09 in gc_free (tc=0x61600000fc80, obj=0x7ffff2637ad8) at src/6model/reprs/ReentrantMutex.c:43
19:35 rjbs- joined, cgfbee joined
nine_ ZoffixW: well in Perl 6 we could have a 0 but True. However the same problem was discussed at length regarding Str.index and the solution was to add syntax as replacement for if 19:35
nwc10 jnthn_: paste.scsys.co.uk/501988
19:35 burnersk joined, literal left, literal joined, jordman left, wtw joined, mattp__ left, rjbs left, wtw left, wtw joined
nwc10 no ASAN barfage 19:35
19:36 sufrosti1o left, literal left, avar left, literal joined, Peter_R joined, sufrostico left, btyler joined 19:37 avar joined, avar left, avar joined, zacts_ is now known as zacts, vendethiel- left, Celelibi_ joined 19:38 raiph joined, sufrostico joined, sufrostico left 19:39 camelia left, Sgeo_ left, Humbedooh left, 16WAASGSD left, Celelibi left, adu left, ponbiki left
dalek c: 2f802c5 | (Zoffix Znet)++ | / (2 files):
Hide home page logo on mobile (Closes #204)

I"m choosing to hide the logo entirely, since on screens that small, the logo in the header looks large enough to serve the role for the larger homepage logo as well		 19:39
Skarsnik lol
the split
dalek href="https://modules.perl6.org/new-db-builder:">modules.perl6.org/new-db-builder: a88b0bb | (Zoffix Znet)++ | lib/ModulesPerl6 (2 files):
Write docs for ModulesPerl6::DbBuilder::Dist::Source subclass
Skarsnik nine_, what was the solution for .index?
19:40 baest joined, Khisanth joined, telex left 19:41 Celelibi_ is now known as Celelibi, Ulti joined
nine_ Skarsnik: I think it's when "foo".index("f") -> $index { say $index }; 19:41
m: when "foo".index("f") -> $index { say $index };
19:42 yqt joined
Skarsnik so if "foo".index('f') will be false? 19:42
nine_ yes
if checks boolean value, when checks definedness 19:43
19:43 rindolf joined
Skarsnik I really don't like the '0E0' it's ugly. especially I used select statement to see if an id was already in the DB and having to test against '0E0' irkked me x) 19:44
19:44 emdashco- left 19:45 atta joined, pecastro joined
ZoffixW Skarsnik, why do you test agains '0E0'? 19:46
19:46 crux joined, sufrostico joined, confuseAcat left 19:47 wtw left, burnersk left, wtw joined, Peter_R left, sufrostico left, sufrostico joined, wtw left, wtw joined
ZoffixW Skarsnik, perl6 -e 'my $ret_succ = "0E0"; my $ret_err = "0"; say "Success on zero rows!" if $ret_succ; say "Error!!" if $ret_err' 19:47
19:47 confuseAcat joined
ZoffixW well.. s/zero rows/possibly zero rows/; 19:47
moritz 0 but True
ZoffixW And I might be wrong
Skarsnik funny thing in DBIsh nothing return 0 x) 19:48
ZoffixW 'cause I recall P6 treats string "0" as true.
19:48 leont joined, burnersk joined 19:49 sufrostico left
ZoffixW Yeah, my example above is wrong. 19:49
19:49 telex joined, sufrostico joined
ZoffixW So I don't think this type of "zero but true" works in P6. 19:50
19:50 camelia joined
psch m: with "foo".index("f") { .say } 19:50
ZoffixW I get 0 in the terminal 19:51
19:51 ChanServ sets mode: +v camelia
camelia rakudo-moar 51d695: OUTPUT«0␤» 19:51
TimToady that's correct
Skarsnik I mean it's the right moment to break thing in DBIish x)
ZoffixW Skarsnik, go for it :)
m: say so "0"
camelia rakudo-moar 51d695: OUTPUT«True␤»
psch hm, downarrow on search results on doc.perl6.org doesn't work..? 19:52
ZoffixW m: say so 0E0
camelia rakudo-moar 51d695: OUTPUT«False␤»
19:52 jnthn_ is now known as jnthn
ZoffixW
.oO( here we go again )
 19:52
psch as in, it directly selects the first option and colapses the drop-down
TimToady restarted camelia, and it seems to have fixed whatever indigestion it had connecting to the server
ZoffixW That search box is evil :D
psch, yeah and I know why.
psch yes, just remove it and put a custom google search :P
raiph m: sub postfix:<.now> ($) {say "" }; DateTime.now # for zengargoyle
19:52 atta left, rindolf left, Celelibi left, mprelude left, rjbs- left, jantore left, diakopte1 joined
camelia rakudo-moar 51d695: OUTPUT«␤» 19:52
psch just kidding, actually :S
ZoffixW TOO LATE! I"M GONNA FIX IT! :D 19:53
19:53 rindolf joined
psch uh, yes, but not with a google custom search i hope? :P 19:53
19:54 emdashcomma joined, Peter_R joined
raiph m: with "foo".index("f") { .say }; 19:54
camelia rakudo-moar 51d695: OUTPUT«0␤»
psch m: when "foo".index("f") { .say }
camelia rakudo-moar 51d695: OUTPUT«Use of uninitialized value of type Any in numeric context in block <unit> at /tmp/95n96BQsX3:1␤(Any)␤»
19:54 sufrosti1o joined
ZoffixW psch, not yet, with the down arrow... which I'm surprised is broken, based on the code. 19:54
Skarsnik any though on? github.com/perl6/DBIish/issues/22 I currently fooling around with that x)
TimToady nine_: yeah, 'when' is the wrong word that begins with 'w', but I didn't see anything better than 'with'
ZoffixW psch, but as for google search... I kinda wish when "No results" is shown there were a button I could click that'd give me google search results for the term.
psch m: $_ = "foo"; when .index("f") { .say }
camelia rakudo-moar 51d695: OUTPUT«Cannot convert string to number: base-10 number must begin with valid digits or '.' in '7⏏5foo' (indicated by 7⏏)␤ in block <unit> at /tmp/hzdp_NFsJz:1␤␤Actually thrown at:␤ in block <unit> at /tmp/hzdp_NFsJz:1␤␤»
TimToady don't use 'when'!!! 19:55
19:55 Celelibi joined
TimToady when is smartmatching, you want 'with' 19:55
psch TimToady: i'm aware of this, i'm just trying to bend it until 'when' works... :S
TimToady .index("f") is going to return 0, and the try "foo" == 0
*then
psch huh 19:56
TimToady m: $_ = 0; when .index("0") { .say }
camelia rakudo-moar 51d695: OUTPUT«0␤»
psch is that a case where chaining ~~ kinda makes sense, maybe?
19:56 jantore joined, mprelude joined, wtw left, sufrostico left, burnersk left, wtw joined, burnersk joined, wtw left, wtw joined
psch 'cause the implicit smartmatch in when could chain with an e.g. infix:<==> 19:56
19:56 atta joined
jnthn nine_: So if you do "make" and then change something and "make" again, it complains about reinstalling the CORE dist? Or? 19:57
19:57 sufrostico joined
jnthn nwc10: No, it ran cleanly here... :S 19:57
psch like < $_ = "foo"; when .index("f") == 0 { .say } > i suppose
Skarsnik don't want to install/config mysql to test stuff
19:57 rjbs joined, avar left, Jonis left, diakopter left, brabo left, sufrostico is now known as 5EXAALNQG, confuseAcat left
psch eh, no, that's probably too magical too :/ 19:58
jnthn nwc10: Any idea *why* it aborts?
19:58 avar joined, confuseAcat joined
nwc10 no. that's probably not the only test 19:58
psch as in, it could maybe sensibly work with < when 0 == .index("f") >, but wouldn't or shouldn't the other way around
nwc10 I didn't dig into the other failures, but guessed that all concurrency related new failures were the same cause
I'm building without ASAN to see if it doesn't abort() 19:59
TimToady is unmotivated to bend smartmatching into a with-shaped pretzel
psch ah, no, i'm confusing the sides :/
nwc10 ie if it's a timing issue exposed by ASAN making stuff slow
jnthn OK. I've no idea why it'd do that.
categorize.t ran fine for me on repeated spectest runs
psch TimToady: well, RT #81336 exists 20:00
20:00 wtw left
psch and i've been wondering about use-cases for chaining infix:<~~> 20:00
20:00 wtw joined
dalek c: 42f7864 | (Zoffix Znet)++ | html/js/main.js:
Unbreak arrow navigation of search results psch++
 20:00
20:01 mattp__ joined, burnersk left, mprelude is now known as 5EXAALNCQ 20:02 avar left, 5EXAALNCQ left
Skarsnik oooh is deprecated is not implemented? 20:03
20:03 mprelude joined
Skarsnik m: sub is deprecated('why') foo { say "Hello"}; foo(); 20:03
camelia rakudo-moar 51d695: OUTPUT«5===SORRY!5=== Error while compiling /tmp/lmRxPw1gPe␤Missing block␤at /tmp/lmRxPw1gPe:1␤------> 3sub is7⏏5 deprecated('why') foo { say "Hello"}; f␤ expecting any of:␤ new name to be defined␤»
Skarsnik m: sub foo is deprecated('why') { say "Hello"}; foo(); 20:04
camelia rakudo-moar 51d695: OUTPUT«5===SORRY!5=== Error while compiling /tmp/3TsWHI1tHE␤Can't use unknown trait 'is deprecated' in a sub declaration.␤at /tmp/3TsWHI1tHE:1␤ expecting any of:␤ rw raw hidden-from-backtrace hidden-from-USAGE␤ cached pure default D…»
20:04 burnersk joined, avar joined, lichtkind joined
psch m: sub f is DEPRECATED("old") { say "hello" }; f 20:04
camelia rakudo-moar 51d695: OUTPUT«hello␤»
psch Skarsnik: the trait is shouty 20:05
Skarsnik Ok thx
20:05 zengargoyle left, zengargo1le joined 20:07 kjs_ joined, vendethiel joined, ponbiki_ is now known as ponbiki 20:09 zengargo1le is now known as zengargoyle 20:11 darutoko left
dalek href="https://modules.perl6.org/new-db-builder:">modules.perl6.org/new-db-builder: 9a5a991 | (Zoffix Znet)++ | lib/ModulesPerl6/DbBuilder/Dist/Source.pm:
Fix pod errors and typos
 20:15
20:16 brabo joined
ZoffixW m: class Foo { method f is DEPRECATED("old") { say "hello" }; }; Foo.new.f 20:16
camelia rakudo-moar 51d695: OUTPUT«hello␤»
ZoffixW What does it do? I'm assuming it'd cry when deprecated methods are used, but it seems to work as normal.
Skarsnik aw 1 occurrence of deprecated code. 20:21
================================================================================
Method fetchrow_array (from DBDish::Role::StatementHandle) seen at:
testtyped.pl, line 10
Please use please use row() instead.
--------------------------------------------------------------------------------
ZoffixW I see.
20:22 5EXAALNQG left, sufrosti1o left
MadcapJake is feeling dizzy from the irregularity of error messages that `perl6 -c` prints 20:22
autarch is there no way to delete keys from a hash in Perl6? 20:23
MadcapJake gonna have to write a whole bunch of javascript regexes X_X
flussence m: my %foo = a => 1, b => 2; say %foo<a>:delete; dd %foo;
camelia rakudo-moar 51d695: OUTPUT«1␤Hash $var = ${:b(2)}␤»
vendethiel autarch: %a<b>:delete
autarch why a colon and not a dot? 20:24
Skarsnik hm, I don't see how to implement for $sth.row -> @row{} github.com/perl6/DBIish/issues/22 (last moritz comment) should I mark this as operator or something?
*iterator
20:25 lucasb joined
moritz Skarsnik: method row { gather { while my @row = self.fetchrow { take @row } } 20:25
leont Because it's a special case of accessing the element
nwc10 jnthn: also aborts for a regular debugging build. valgrind finds no problems before the abort 20:26
Skarsnik moritz, Na I mean, you don't want to gather the whole rows, only at each loop iteration
moritz Skarsnik: well, it's lazy
20:26 emdashcomma left
moritz Skarsnik: it only generates the rows as they are accessed 20:26
20:27 sufrostico joined
MadcapJake does `perl6 -c` only catch X::Syntax exceptions or will it catch others too? 20:27
moritz my: my \rows = gather { for ^10 { say "generating $_"; take $_ } }; rows[2]
m: my: my \rows = gather { for ^10 { say "generating $_"; take $_ } }; rows[2]
camelia rakudo-moar 51d695: OUTPUT«generating 0␤generating 1␤generating 2␤»
moritz MadcapJake: it doesn't catch exceptions 20:28
MadcapJake s/catch/print to stderr/
moritz MadcapJake: I mean, not more or less than normal perl6
MadcapJake: all of them
Skarsnik so I put gather around the whole code of the method?
MadcapJake All that can be caught at compile time?
moritz MadcapJake: try -c -e 'BEGIN die "oh noez"'
MadcapJake: it's just normal perl 6, except that it exits after compilation 20:29
nine_ jnthn: yes, we currently try to install the CORE dist on make install and it fails if it's already there
moritz so it runs all BEGIN blocks, use etc.
MadcapJake hmm, ok :P 20:30
20:31 kjs_ left
MadcapJake I'm trying to wrap the output of `perl6 -c` to provide errors to Atom's linter. Looks like I'm going to have to write a separate rule for each error then xD 20:32
20:33 sufrosti1o joined 20:35 domidumont left
MadcapJake moritz: is there any way to limit the scope of `perl6 -c` to just syntax checking? 20:36
20:36 rigsby left
moritz MadcapJake: no. You can't syntax-check without compiling 20:37
20:38 zakharyas joined
Skarsnik moritz, hm, I still get only oen result 20:38
20:39 geraud joined
Skarsnik ho you really need the while x) 20:40
20:40 rigsby joined 20:41 zengargo1le joined 20:42 Sgeo joined, [dpk] joined
MadcapJake i wish that all error messages were single-line :( 20:43
20:44 raydiak_ joined, krakan_ joined
dalek href="https://modules.perl6.org/new-db-builder:">modules.perl6.org/new-db-builder: 6681d54 | (Zoffix Znet)++ | / (4 files):
Improve test coverage for ModulesPerl6::DbBuilder::Dist::Source
 20:44
20:44 leont_ joined 20:45 doublec joined, [dpk] is now known as {dpk}, ilbelkyr_ joined, xfix_ joined 20:46 mspo_ joined, lucasb_ joined 20:47 bakedb joined, KotH_ joined, uruwi_ left, telex left, dpk left, {dpk} is now known as dpk, telex joined, ssflying joined 20:48 mrsolo left, emdashcomma joined, jonas1 left, ribasushi_ joined
jnthn nwc10: Maybe helgrind has an idea? 20:48
20:49 ZoffixW left, shlomif joined 20:50 dpk is now known as {dpk}, rindolf left, shlomif is now known as rindolf 20:51 rindolf left 20:52 Upasaka joined 20:53 {dpk} is now known as dpk
nine_ jnthn: I think we need to have Repository::Installation precompile on load in addition to on install 20:53
20:54 sunnavy joined
dalek c: c8ec94a | lizmat++ | doc/Type/Supply.pod:
Slight Supply.throttle reorganization
 20:55
c: 229ec9b | lizmat++ | doc/Type/Supply.pod:
Add simple example of Supply.throttle (code)
20:56 Mhoram joined, sftp left 20:57 zengargoyle left, ilbelkyr left, raydiak left, Khisanth left, krakan left, leont left, doublec_ left, bakedb_ left, xfix left, sunnavy_ left, ribasushi left, KotH left, emdashco| left, go|dfish left, lucasb left, mspo left, ssflying_ left, lichtkind left, abaugher left, atta left, eiro left, Upasaka_ left, Khisanth joined, KotH_ is now known as KotH 20:58 lichtkind joined 20:59 kaare_ left 21:00 b2gills left, yoleaux joined, ChanServ sets mode: +v yoleaux 21:01 sftp joined, eiro joined, leont_ is now known as leont 21:03 raiph left 21:04 rigsby left, sufrostico left, sufrosti1o left
dalek kudo/curli: a51f6c7 | (Stefan Seifert)++ | src/core/CompUnit/Repository/Installation.pm:
Precompile on-demand in Repository::Installation, too.

CompUnit::Repository::Installation needs to precompile on-demand (when loading a module) in addition to on installation. Consider: use lib "lib"; # prepends a Repository::FileSystem to the repo chain use Test; # installed in site, therefore not in lib's precomp store
   # therefore loaded from source
use Test::Mock; # loaded from lib/ therefore precompiled on-demand.
Since Test::Mock uses Test, the latter will be precompiled, too when precompiling Test::Mock. But Test was already loaded from source before precompilation of Test::Mock so when we finally load the latter, we get a conflict.
If you precompile on-demand, you have to do it everywhere.		 21:11
21:12 atta joined, go|dfish joined
nine_ I think this ^^^ was the source of the fleeting precompilation related failures I've seen from time to time. They appeared only on the first try since on the second the precompiled version of Test was already loaded from the start. 21:12
21:13 tphilipp joined 21:14 b2gills joined
jnthn nine_: No, I don't think installation should do that... 21:17
nine_: When we "make install" a new Rakudo version then it should go and produce precomp'd versions of installed modules at that point.
(When, since we're in make install, it certainly has permissions to be doing such things)
21:18 espadrine joined 21:19 tphilipp left 21:21 zakharyas left
jnthn nine_: Well, or depends on the active pre-comp manager 21:21
nine_ jnthn: the on-load precomp is for the case where we have a FileSystem repo before the Installation repo in the chain. That also means that we use the FileSystem repo's precomp store.
jnthn Yeah, I see
OK, then it makes sense.
21:22 eliasr joined 21:23 doublec left
nine_ This fixed all strange errors of Task::Star modules. The remaining errors are all about @*INC in tests 21:23
jnthn OK :)
21:24 eliasr left 21:25 xfix_ left, doublec joined 21:27 espadrine left, n0xff joined 21:28 espadrine joined, AlexDaniel joined, n0xff left
dalek ast: 3966e1b | jnthn++ | S17-supply/basic.t:
A bunch more supply protocol tests.

Including making sure that an observer will never see an emit after seeing a done or a quit - at least, not unless somebody goes out of their way to make this happen. Covers RT #123477.		 21:33
21:39 lucasb_ left 21:40 lukasz joined 21:41 lukasz left, domidumont joined
dalek ast: d430474 | jnthn++ | S02-names-vars/names.t:
Tests for RT #113892 and RT #115608.
 21:42
21:46 espadrine_ joined, espadrine left
lizmat jnthn: is it correct that a number of supply related tests fail ? 21:47
jnthn lizmat: It's platform specific.
21:47 bjz joined
jnthn lizmat: They're fine on Windows. 21:47
I'd not have merged otherwise.
lizmat :-(
jnthn Probbly some stupid posixism
lizmat ok, will check up on them on OSX
jnthn I did enough debugging today.
And libuv just does an abort(). Presumably there is an error string that comes back from pthreads. 21:48
lizmat well, time to take it from you then
jnthn OK :) 21:49
lizmat argh, another harness related failure, it seems :-(
jnthn Note that a few tests still having timing-related issues that can fail under stress
But nwc10 reported something that's different to that.
I suspect helgrind will provide good diagnostics on that case.
The tests for batch, elems, and unique (at least) could do with using the FakeScheduler that I used for interval.t. 21:50
I can appreciate we want to test that ThreadPoolScheduler does, in fact, do something vaguely right with regard to wallclock time
But we don't need to test that over and over again in the supply/promise tests.
So I'm fairly comfortable writing those tests in "virtual time" 21:51
21:53 raiph joined
lizmat ok, will look at that tomorrow, don't think I'll be doing much more today 21:53
21:54 lsm-desktop left
jnthn Yeah, me either. 21:54
21:54 domidumont left, rindolf joined
jnthn TimToady: Next little ruling request: rt.perl.org/Ticket/Display.html?id=120916 :) 21:54
21:54 ilbelkyr_ is now known as ilbelkyr
lizmat jnthn: Supply.tail defaults to 1, List.tail enforces the number 21:55
jnthn: which one will I adapt ?
jnthn TimToady: Ticket wants "has $.foo = ..." to mean...something...I don't even... :)
TimToady: I don't actually see what it sensibly can mean. There's no way to actual provide that initializer in the class body. 21:56
TimToady: So I'm happy enough with "current behavior is fine, if you put stub code there it'll just behave like stub code anywhere else".
21:57 Zoffix_ is now known as Zoffix
jnthn (With the exception of method bodies in roles, which become requirements I guess...but there's a way to provide those...) 21:57
lizmat jnthn: should there be a "done" in a LAST block ?
21:57 Zoffix left
lizmat Supply.elems has one, Supply.tail doesn't ? 21:57
jnthn lizmat: I think defaulting to 1 is nicely consistent with head... 21:58
lizmat ok
and wrt to "done"
?
arnsholt jnthn: I have a super-weird segfault issue on MoarVM, but seemingly weird changes (like changing the name of the compilation unit) make it go away. Hopefully I'll find a reliable testcase
jnthn lizmat: "It's complicated."
21:59 rurban joined
jnthn lizmat: Well, not really, but... :) 21:59
lizmat: supply blocks try to have good defaults. The rule is: when you reach 0 active whenever blocks, then it automatically does "done" for you.
arnsholt Somehow the REPR data of a CStruct type is NULL when initialize is called. I'm suspecting an uncomposed type, or some such
jnthn lizmat: So elems probably doesn't need it either.
lizmat: It doesn't hurt to have it, of course. 22:00
lizmat well, if it isn't needed, then it shouldn't be there
jnthn lizmat: The place you *need* it is if you want to issue "done" when other whenever blocks remain active.
lizmat there's enough cargo cultiing in the world already :-)
jnthn Yeah, I ported them closely from the "on" syntax with an eye to keeping them working :-)
But elems is a fairly clear "nah, don't need it" :) 22:01
lizmat gotcha :-)
22:01 colomon joined
lizmat makes the LAST easier as well: 22:01
LAST emit($elems) if $elems != $last_elems;
jnthn Yeah, but put a { ... } around it 22:02
I think there's a bug lingering :(
lizmat ok, will check 22:03
22:03 zengargo1le is now known as zengargoyle 22:04 hankache joined
dalek kudo/nom: 05bd97c | lizmat++ | src/core/Supply.pm:
Model Supply.tail like List.tail
 22:04
22:04 noganex left 22:05 lsm-desktop joined 22:08 rigsby joined
hankache good evening 22:10
lizmat hankache o/
hankache ohai lizmat 22:11
22:11 doublec left
timotimo │ src/6model/reprs/ReentrantMutex.c:43 22:12
22:12 llfourn joined
jnthn timotimo: ? :) 22:12
jnthn can reliably reproduce RT #125515, and will work on that tomorrow
22:13 doublec joined
dalek kudo/nom: 7d636cf | lizmat++ | src/core/Supply.pm:
Streamline Supply.elems
 22:13
22:17 Zoffix joined
dalek kudo/nom: 327cbf0 | lizmat++ | src/core/Any-iterable-methods.pm:
Allow for List.tail(), aligning with Supply.tail
 22:17
ast: b599905 | lizmat++ | S32-list/tail.t:
Add tests for List.tail()
 22:18
arnsholt Ooooooooh. 22:20
22:21 Zoffix left
arnsholt jnthn: Having a class named Parameter breaks all the things. Bug? 22:21
Skarsnik lol
22:21 rindolf left
jnthn arnsholt: We resolving a bunch of built-in names lexically 22:21
So if you shadow them...
lizmat m: say Parameter.WHAT
camelia rakudo-moar 05bd97: OUTPUT«(Parameter)␤»
jnthn Same with Code, fwiw 22:22
arnsholt Right, right. Just wanted to make sure
jnthn We could choose to always look directly in the setting for such symbols.
But when I've floated it in the past I've never got a clear "yes" from TimToady
arnsholt That's more or less how Python works, FWIW 22:23
jnthn Probably 'cus he can argue it both ways, like I can :P
lizmat prefix this built-in with an R ?
jnthn ?
lizmat if the builtin Parameter would be called RParameter, there wouldn'r have been a conflict 22:24
jnthn Yeah but that's ugly :P
lizmat agree
:-)
but having bad things happen with common names such as Parameter and Code, is maybe worse 22:25
jnthn Basically we'd just be forcing people to write a custom setting if they wanted to replace certain things
Skarsnik but if you want to have a RParameter class!
lizmat Skarsnik: less likely, but yeah, get your point :-)
jnthn Who'd create a project that just puts an R at the start of something? :P
22:25 espadrine_ left
Skarsnik why not just warn/error when it's done? 22:26
m: class Parameter {};
camelia ( no output )
jnthn 'cus we don't want to totally prevent it
I think for certain things having to write a custom setting to override them is fine though
Skarsnik I mean a warning like when you write my $foo; my $foo
jnthn And we'll skip looking through lexical scopes we're in and go straight to the setting scope. 22:27
Because this situation isn't like tht
It's like my $foo; { my $foo; }
Because the built-ins are considered an outer scope to your program.
I already made exception resolution do this 'cus people were writing a class X and hosing error reporting :) 22:28
(For compiler-generated exceptions, that is)
timotimo excuse the cat on the keyboard please
dalek kudo/nom: 4a5e549 | lizmat++ | src/core/Any-iterable-methods.pm:
Allow List.head() to align with List.tail()
jnthn timotimo: I don't mind the cat being on the keyboard, but can't he fix a bug or two while there? 22:29
dalek ast: 76eac4d | lizmat++ | S32-list/head.t:
Add tests for List.head()
timotimo :)
jnthn notes we're down to 31 xmas RTs 22:30
Should get under 30 tomorrow :)
m: say "Need to fix {31 / 28} a day if we're to do 'em all by 24th :P" 22:31
camelia rakudo-moar 05bd97: OUTPUT«Need to fix 1.107143 a day if we're to do 'em all by 24th :P␤»
22:31 hankache left
dalek c: 15957e1 | lizmat++ | doc/Type/List.pod:
Update doc on List.head/tail
 22:32
jnthn 'night, #perl6 22:34
Skarsnik good night
22:34 raiph left, vendethiel left
lizmat gnight jnthn 22:37
22:42 TimToady left 22:43 TimToady joined
dalek ast: 397ddee | lizmat++ | S17-supply/head.t:
Add tests for Supply.head
 22:45
22:46 sufrostico joined
timotimo │ src/6model/reprs/ReentrantMutex.c:43 22:48
... ...
lizmat .tell jnthn supply test fails on OS X seem to be related to setting $*SCHEDULER
yoleaux lizmat: I'll pass your message to jnthn.
Skarsnik You can can copy paste? x)
*Your cat
stmuk "staring like a cat at a perl advent calender" 22:49
22:49 Khisanth left
lizmat .tell jnthn could it be some missing decont on $*SCHEDULER maybe ? 22:49
yoleaux lizmat: I'll pass your message to jnthn.
22:51 isacloud left, DrParis left, bowtie joined 22:53 sufrosti1o joined, leont left 22:54 Khisanth joined
dalek kudo/nom: 9baad52 | lizmat++ | / (2 files):
Add Supply.head + run tests
 22:56
href="https://modules.perl6.org/new-db-builder:">modules.perl6.org/new-db-builder: 7586865 | (Zoffix Znet)++ | t/03-app/00-startup-variants.t:
Fix incorrect test coverage
 22:57
href="https://modules.perl6.org/new-db-builder:">modules.perl6.org/new-db-builder: 220a336 | (Zoffix Znet)++ | lib/ModulesPerl6/DbBuilder/Dist/Source.pm:
Remove useless code
 23:00
lizmat .tell jnthn question: should you be able to last out of a whenever block without further consequences ? 23:07
yoleaux lizmat: I'll pass your message to jnthn.
dalek kudo/nom: 39869f1 | lizmat++ | src/core/Supply.pm:
We don't need an empty whenever block
 23:08
pink_mist lizmat: "further consequences"? the whenever block should stop doing things 23:10
lizmat well, it appears that the Supply.head tests run ok in the test harness if there is *no* last out of the whenever block 23:11
dalek c: a87a9b0 | lizmat++ | doc/Type/Supply.pod:
Add doc stubs for Supply.head/tail
 23:13
23:14 Skarsnik left, snarkyboojum joined
lizmat good night, #perl6! 23:14
23:17 isacloud joined 23:20 DrParis joined 23:21 DrParis is now known as Guest84367
AlexDaniel m: say 2**⁸ 23:23
camelia rakudo-moar 9baad5: OUTPUT«256␤»
dalek href="https://modules.perl6.org/new-db-builder:">modules.perl6.org/new-db-builder: c4ce087 | (Zoffix Znet)++ | lib/ModulesPerl6/DbBuilder/Dist/Source/GitHub.pm:
Do not use GitHub"s repo desc as distro description
AlexDaniel m: my $y = 2; say ++$y²³;
camelia rakudo-moar 9baad5: OUTPUT«5===SORRY!5=== Error while compiling /tmp/86Kdt5oAbM␤Operators '++' and '²³' are non-associative and require parentheses␤at /tmp/86Kdt5oAbM:1␤------> 3my $y = 2; say ++$y²³7⏏5;␤ expecting any of:␤ postfix␤»
AlexDaniel m: my $y = 2; say $y++⁸; 23:24
camelia rakudo-moar 9baad5: OUTPUT«256␤»
23:27 yqt left 23:30 RabidGravy left 23:35 Khisanth left 23:36 mrsolo joined 23:37 noganex joined, Khisanth joined 23:43 n0xff joined, n0xff left 23:44 lichtkind left 23:45 stmuk_ joined 23:46 mattp_ joined
dalek href="https://modules.perl6.org/new-db-builder:">modules.perl6.org/new-db-builder: 69d956b | (Zoffix Znet)++ | / (3 files):
Improve test coverage for Dist Source baseclass (JSON parse errors and other)
 23:47
23:47 krakan joined
dalek href="https://modules.perl6.org/new-db-builder:">modules.perl6.org/new-db-builder: 9147746 | (Zoffix Znet)++ | lib/ModulesPerl6/DbBuilder/Dist/Source/GitHub.pm:
Remove debug code
 23:48
23:48 stmuk left 23:50 telex left, telex joined
gfldex .tell moritz i would like to book a day in advent, but don't know how! 23:51
yoleaux gfldex: I'll pass your message to moritz.
dalek rl6-roast-data: 33fe841 | coke++ | / (9 files):
today (automated commit)
 23:53
kudo-star-daily: 578d8c1 | coke++ | log/ (2 files):
today (automated commit)
kudo-star-daily: 5fd76ef | coke++ | log/ (2 files):
today (automated commit)
23:54 isacloud left, go|dfish left, ilbelkyr left, krakan_ left, mattp__ left, diakopte1 left 23:56 telex left 23:58 telex joined
[Coke] waves from a car /j 23:58

Powered by Cro and the Raku® Programming Language.

Please report any issues / comments / feature requests as an issue on App::Raku::Log.

Thank you!