»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'p6: say 3;' or rakudo:, or /msg camelia p6: ... | irclog: irc.perl6.org or colabti.org/irclogger/irclogger_logs/perl6 | UTF-8 is our friend!
Set by moritz on 22 December 2015.
00:10 wamba left 00:12 BenGoldberg joined 00:15 Cabanossi left, nadim left 00:17 Cabanossi joined 00:39 pierre_ joined 00:53 r3m left, r3m joined 00:57 Actualeyes left 01:00 dct left 01:03 mcmillhj joined 01:08 mcmillhj left 01:14 mcmillhj joined 01:18 mcmillhj left 01:24 k-man joined 01:29 mcmillhj joined 01:36 mcmillhj left
Geth ecosystem: CurtTilmes++ created pull request #336:
Add DBI::Async to ecosystem		 01:37
01:38 xfix left 01:40 xfix joined 01:42 labster left 01:46 ilbot3 left 01:47 mcmillhj joined 01:48 ilbot3 joined, ChanServ sets mode: +v ilbot3 01:52 aborazmeh joined, aborazmeh left, aborazmeh joined 01:54 mcmillhj left 02:00 labster joined 02:01 mcmillhj joined 02:04 itaipu left 02:06 mcmillhj left 02:15 Cabanossi left 02:17 mcmillhj joined, Cabanossi joined
grondilu m: class { has Num method foo { rand } }.new.foo.say 02:22
camelia 0.157018793505413
grondilu ^TIL new method syntax apparently 02:23
02:26 mcmillhj left 02:30 kurahaupo joined 02:35 kurahaupo left 02:36 mcmillhj joined 02:41 mcmillhj left 02:44 dj_goku left 02:53 mcmillhj joined 02:58 mcmillhj left 03:00 stmuk joined 03:01 stmuk_ left 03:02 wamba joined 03:05 dj_goku joined 03:09 stmuk_ joined 03:10 stmuk left 03:25 aborazmeh left 03:27 pierre_ left 03:28 pierre_ joined 03:29 Cabanossi left 03:32 Cabanossi joined 03:33 pierre_ left 03:45 pierre_ joined 03:53 Actualeyes joined, kurahaupo joined 03:55 Actualeyes left 03:56 kurahaupo_ joined 03:58 pierre_ left 03:59 pierre_ joined 04:01 kurahaupo__ joined, Cabanossi left 04:02 Cabanossi joined 04:03 pierre_ left 04:04 Resol joined, kurahaupo left 04:05 kurahaupo joined 04:06 kurahaupo_ left 04:07 kurahaupo left, kurahaupo__ left 04:08 kurahaupo joined, wamba left 04:19 skids left 04:22 BenGoldberg left 04:28 tharkun left 04:35 stmuk joined 04:37 stmuk_ left 04:41 tharkun joined 04:52 curan joined 05:08 kurahaupo_ joined
Geth ecosystem: a88b802059 | (Curt Tilmes)++ | META.list
Add DBI::Async to ecosystem

See github.com/CurtTilmes/perl6-dbi-async		 05:11
ecosystem: 33b8477a85 | (Juan Julián Merelo Guervós)++ (committed using GitHub Web editor) | META.list
Merge pull request #336 from CurtTilmes/master

Add DBI::Async to ecosystem Thanks!
05:11 kurahaupo left 05:12 pierre_ joined 05:14 xtreak joined 05:17 wamba joined 05:31 mcsnolte left, bioexpress joined
bioexpress Hello, how could I remove a module from the ecosystem? 05:32
gfldex bioexpress: you send a PR to the ecosystem with the line in question removed
bioexpress Thx! 05:33
What's a PR? 05:34
gfldex pullrequest 05:35
05:35 anton_p joined
bioexpress Thx 05:37
samcv wow... 05:41
person on libtommath github issue says that i should fork the project and change the LICENSE file
and show my lawyers my fork with me having changed the license. 05:42
gfldex please don't feed the troll
samcv yeah cause... that's how things work in the real world. you can just lie about softwares licenses by editing the LICENSE file
they are a 'Member' :(
and they think the WTFPL is better than CC0 because it uses less words 05:43
github.com/libtom/libtommath/issue...-298826328 :\
wow i'm reading the complaints of the maker of WTFPL about CC0 and they're pretty much what the person just posted 05:48
that how did it take them 1000 words to write that they person could use the software however they want. seems sam hocevar misses the point of licenses 05:49
05:49 wamba left
samcv "Your honor, this license is really short. therefore i can't possibly be infringing on any copyright here. this one sentence obviously cannot be interpreted any other way. if the license fits into a small number of words you must acquit"" 05:51
said by nobody ever.
well acutalyl probably not. but. maybe a few peolpe. but they' probably didn't win their cases
can't say nobody would say that since sam hocevar obviously seems to say exactly that
Geth ecosystem: kuerbis++ created pull request #337:
Update META.list		 05:52
05:53 CIAvash joined 05:55 bioexpress left 05:57 xinming_ left 06:00 CIAvash left, Cabanossi left 06:02 Cabanossi joined 06:04 ufobat joined, xinming joined
u-ou what does this mean? Cannot resolve caller Numeric(Player+{Wolf}: ); none of these signatures match: 06:05
Geth ecosystem: 9c7a681e0a | kuerbis++ (committed using GitHub Web editor) | META.list
Update META.list

Removed Term-Form-p6
ecosystem: d4225a208f | (Juan Julián Merelo Guervós)++ (committed using GitHub Web editor) | META.list
Merge pull request #337 from kuerbis/master

Update META.list Thanks for the submission. Why did you remove it?
samcv u-ou, can paste full error? 06:06
gfldex u-ou: it means that neither Player nor Wolf got a Numeric method to turn it into a number. You likely used an operator that requires a number.
u-ou there's only one more line: (Mu:U \v: *%_) 06:07
06:09 domidumont joined
u-ou that's weird. it happens when I compare an array of them's .elems with < 06:09
but that shouldn't try to turn any of them into a number
06:13 xinming left 06:14 domidumont left 06:15 domidumont joined 06:17 xinming joined 06:19 xtreak left, xtreak joined 06:20 xtreak left 06:21 xtreak joined
samcv so it's $wold.elems < 10 06:22
or something like that?
could always add a Numeric method \-/ shrug
06:23 itaipu joined
sammers hi all 06:26
06:32 xtreak left
u-ou it happens without that line, too, so it seems to be happening in the background. I will post it on github sometime to show you if I can't work it out. 06:32
06:33 xtreak joined
u-ou the program is across many files 06:33
06:37 xtreak left 06:42 lowbro joined, lowbro left, lowbro joined 06:55 lizmat joined 06:56 wamba joined 07:00 xtreak joined 07:07 wamba left 07:10 xtreak left 07:14 xtreak joined 07:19 espadrine_ joined 07:27 espadrine_ left 07:29 nadim joined 07:30 xtreak left 07:34 pierre_ left 07:45 labster left 07:46 pierre_ joined, xtreak joined 07:50 labster joined 07:52 robertle left 07:53 darutoko joined 07:55 parv joined 08:00 itaipu left 08:02 lizmat left 08:03 wamba joined 08:07 zakharyas joined 08:08 nadim left, domidumont left, nadim joined 08:09 domidumont joined 08:11 mcmillhj joined 08:15 bjz joined 08:16 Cabanossi left, mcmillhj left 08:17 Cabanossi joined 08:20 dakkar joined 08:27 mcmillhj joined 08:31 jonas1 joined 08:32 mcmillhj left 08:46 Cabanossi left 08:47 Cabanossi joined 08:48 pierre_ left 08:56 rindolf joined 08:59 kurahaupo_ left 09:00 xtreak left 09:02 pierre_ joined 09:08 rindolf left 09:13 Ven joined, Ven is now known as Guest98555 09:19 labster left 09:23 john51 left, john51_ joined, rindolf joined 09:25 kurahaupo_ joined 09:39 xtreak joined 09:45 AlexDani` joined, Cabanossi left 09:46 Actualeyes joined 09:47 Cabanossi joined 09:48 pierre_ left 09:49 AlexDaniel left 09:52 pierre_ joined, Guest98555 left 09:55 Ven joined 09:56 Ven is now known as Guest51495 09:57 AlexDani` left, AlexDani` joined 10:02 kurahaupo` joined 10:05 kurahaupo_ left 10:06 kurahaupo` left 10:07 salva joined 10:08 Guest51495 left 10:09 domidumont left 10:11 Ven_ joined 10:17 bjz_ joined 10:18 anton_p left, bjz left 10:30 AlexDani` is now known as AlexDaniel 10:31 rindolf left 10:34 ChoHag joined
nadim good morning, a few questions about named arguments. s(:argument) would set argument to True, how is is set to false? named arguments can be :argument(values), is it possible to write argument:value someway? 10:37
moritz nadim: s(:!argument) sets it to False
argument:value isn't supported, argument => value is
10:38 pierre_ left
nadim moritz: thanks, I always forget about :! . Just to satisfy my curiosity, apart the fact that we had fat commas in P5, do you know if and why argument:value was a candidate? 10:43
moritz nadim: no, my historic knowledge doesn't stretch that far back
10:44 ChoHag left
timotimo speaking from a "right now" standpoint, argument:value doesn't work because there could be a sub named &argument:value 10:45
m: sub argument:value { say "not what you wanted" }; sub takes-named-argument(*@, :$argument) { say "$argument" }; takes-named-argument(argument:value)
camelia not what you wanted
Use of uninitialized value $argument of type Any in string context.
Methods .^name, .perl, .gist, or .say can be used to stringify it to something meaningful.
in sub takes-named-argument at <tmp> line 1
10:52 parv left 10:56 nadim left 11:00 nadim joined 11:01 Cabanossi left 11:02 Cabanossi joined 11:11 bjz joined 11:12 bjz_ left 11:13 TEttinger left 11:21 xiaomiao left 11:23 xiaomiao joined 11:27 xinming left 11:28 xinming joined 11:45 Cabanossi left 11:47 Cabanossi joined 11:57 abraxxa joined
AlexDaniel freaking pulseaudio… why is everything so clicky 11:59
timotimo it's not getting enough cpu time? 12:00
AlexDaniel it happens whenever something changes
e.g. if I change the volume, or I start something else
timotimo when i play synthesia and have qsynth on, i need to give qsynth and pulseaudio a lower nice value otherwise it'll go into crackly-staticy mode every minute or two 12:01
(synthesia on wine is a performance trainwreck)
AlexDaniel while I can ignore these occasional clicks, I can't understand why the hell would I experience this in 2017
timotimo it's got a little animation in the mode select menu where little circles go 'round a button and it chews through cpu like nobodys business
AlexDaniel thing is, I can just uninstall pulseaudio and things will go back to normal 12:02
except that I'd have to find a different web browser
fucking hell…
12:04 wamba left 12:05 ChoHag joined 12:06 daxim joined, itaipu joined 12:08 Ven_ left 12:11 Ven joined 12:12 Ven is now known as Guest31589 12:21 rindolf joined 12:23 Guest31589 left 12:24 raschipi joined
Ulti found a little bug in the docs.perl6.org it doesnt URL encode/decode search strings so you can get something like docs.perl6.org/routine/%% as a bad request 12:25
what repo has that stuff in?
is it just perl6/doc 12:26
AlexDaniel Ulti: yes
12:27 Ven_ joined
Ulti k k 12:27
will take a look
12:28 Ven_ left 12:29 Ven_ joined 12:34 wamba joined 12:38 rindolf left
Voldenet I've got some perl5 code, what's the simplest way (modules would be nice) to make it perl6-friendly: ix.io/szS 12:41
I'm thinking of using HTTP::Server::Async
timotimo yeah, your code currently looks like a cgi script
Voldenet Yeah, it's a pretty much a trivial cgi script. 12:42
s/a pretty much/pretty much/
timotimo painful how much boilerplate it has, but other than the boilerplate it uses no modules at all
Voldenet welp, that's cgi for you :P
timotimo still better than "use CGI" 12:43
Voldenet Yeah, sadly
Ulti probably narrowed it down to github.com/perl6/doc/blob/eaa4d3dd...fy.p6#L765 needing to replace % with %25 not just the \ to %5c really just put in a full percent encode maybe
12:43 haxmeister joined 12:45 mcmillhj joined
Ulti weird URI escape should be doing all this 12:45
hmmm
12:45 pierre_ joined
Ulti aha! %2525 12:46
heh
Voldenet Hm, I guess using perl6 with cgi is a really bad idea, because of script compilation time
timotimo yeah, there's way better ways to go about this 12:48
Voldenet I could just use fcgi or http
12:48 domidumont joined
Voldenet Well, http has less code smell, that's for sure 12:48
but I wonder how does HTTP::Server::Async compare to nginx 12:49
timotimo i usually install services behind a mod_proxy-alike
Voldenet mod_proxy http or mod_proxy fcgi? 12:51
timotimo http
we do have SCGI or something, though? 12:52
12:55 domidumont left
Voldenet hm, gonna do some ab on the http to compare 12:55
12:56 ChoHag left
Voldenet hmm nginx: Requests per second: 38372.99 [#/sec] (mean), HTTP::Server::Async: Requests per second: 36.97 [#/sec] 12:59
tyil thats 13:00
Voldenet :(
tyil a slight difference
13:00 Cabanossi left
timotimo yup 13:00
moritz in both cases talking to a database likely exceeds the offset from the web server 13:01
Voldenet Unless you're serving static content as well
moritz right, but you'd configure your web server to do that in any realistic production scenario 13:02
timotimo yeah, always use a proper file server to serve static content
13:02 Cabanossi joined
timotimo er 13:02
web server
Voldenet Well, it makes sense to proxy only dynamic api requests to the dynamic server.
moritz that's the advice even if using a rather mature app server, like p5/mojolicious/hypnotoad
Voldenet but then, nginx + cgi: Requests per second: 261.31 [#/sec] (mean) 13:03
hmmm, 40 rq/s on dedicated http server does not sound very realistic 13:04
haxmeister I have a question regarding this statement 13:06
By separating identifiers with double colons, the right most name is inserted into existing or automatically created packages.
does that mean if I use the double colon syntax when there are no packages other than the main file.. it will create a namespace/package that matches my identifier? 13:07
Voldenet Okay, I've used HTTP::Server::Tiny and it does 260rq/s - it's better 13:09
haxmeister: I think it means that you don't have to create 3 folders and 3 files per folder, just 3 folders and the module when it should be 13:11
so A::B::C::D would require only A/B/C/D.pm6
and the A A::B A::B::C would be autogenerated
haxmeister ugh.. I'm lost in that.. I need to read more, I don't seem to understand the usage and intent of addressing external files/modules this way 13:12
Ulti haxmeister: what language have you come from?
haxmeister last language I studied was lisp 13:13
13:13 eliasr joined
Ulti ahh I dont even know how name spacing of things works in lisp :S 13:13
haxmeister before that scheme, object pascal, C++ and C..
but only intermediate skill level.. I just like studying languages, with a particular affinity for perl
Ulti well C is similar since the include is just / for the same idea less the strict name spacing it also implies 13:14
Voldenet haxmeister: it's similar to what C++ would do
#include <boost/regex.hpp> -> use boost::regex
haxmeister so the colons are directory markers? 13:15
Voldenet I'm not 100% sure, but they sure are used that way in modules
haxmeister ok that is helpful
Ulti haxmeister: doesnt have to be
timotimo if you have a meta6.info, you can have the files organized in any way you like and only the "provides" hash is used
haxmeister oh I'm not ready for that timotimo ..lol
timotimo no problem
Voldenet Basically "you can arrange that differently, but most people don't" :)
haxmeister this tut gave the example my Int $Foo::Bar::buzz = 42; 13:16
Ulti haxmeister: you could put it all in a single file if you really wanted to since you have to fully qualify the namespace anyway even with the directory structure
13:16 cdg joined
haxmeister in that example we are creating a variable named what in what namespace location? 13:17
13:17 Ven_ left, yht joined, Ven joined
Ulti $buzz in Foo::Bar is being set to 42 ? 13:17
isnt this literally the example from the docs on namespacing that tells you what thats doing... 13:18
13:18 Ven is now known as Guest80741
Ulti haxmeister: docs.perl6.org/language/packages is this what you were reading? 13:19
haxmeister also unless you know why you are writing something like $Foo::Bar::buzz = 42 its *very* unlikely you want to be 13:20
13:25 raschipi left
haxmeister I guess it confused me because the $Foo is on the left, when I thought of Bar::buzz as the package/namespace... seemed like it should be Bar::buzz::$foo 13:35
I suppose I need to think of it as "from" instead of "in".. then the phrase kinda reads right 13:36
timotimo you're allowed to put the $ there, too, i believe
haxmeister so Bar::buzz::$Foo would be the same as $Foo::Bar::buzz ? 13:37
timotimo try it out
(not for declaration, though)
13:40 curan left
haxmeister Ulti: I'm perusing docs.perl6.org/language/syntax right this second.. I've written a small program to recurse through some folders for certain files in perl6 but I need to learn to use more features like understanding the colon syntaxes that are used in several constructs 13:42
13:42 anton_p joined 13:46 titsuki joined 13:47 titsuki_ left
moritz haxmeister: design.perl6.org/S02.html#Adverbial_Pair_forms has a summary of the diferent colonpair forms 13:55
13:56 bjz left
haxmeister ty moritz 14:03
trying to make some functions that have type restraints right this second
my first try must show my ignorance sub SFM(Num $rpm, Num $dia )
perl6 will not upgrade an int to a num.. 14:04
Ulti Bar::buzz::$Foo <--- does that not resolve $Foo for the string value to resolve as the actual symbol?
timotimo yeah, you'll want "Real" instead of "Num" 14:05
that'll give you any real number
if you also accept complex numbers, use "Numeric"
haxmeister so to get floating point numbers I'll use Real not Num?
Voldenet I suggest using `Rat` instead 14:06
but welp Real is a role, right? 14:07
14:07 xtreak left
haxmeister lol.. that's why it's not clear to me, the documentation for types suffers the same lack of clarity 14:07
timotimo Real is a role that is done by all real numbers. Int, Num, Rat.
m: say Int ~~ Real; say Num ~~ Real; say Rat ~~ Real; say Complex ~~ Real;
camelia True
True
True
False
Voldenet haxmeister: look at the bottom, there's a type graph docs.perl6.org/type/Numeric 14:08
haxmeister ok
moritz Real is a good type constraint when work with real numbers :-)
haxmeister that picture is loosing this noob.. 14:09
timotimo that shows you what is considered Rational, Real, or Numeric 14:10
Rational is just Rat and FatRat
Real is anything that's Rational, Instant, Duration, Int, or Num
and Numeric is anything that's Real, or Complex
haxmeister ok.. smart way to pass to literal ints 313 and 5 to a sub that only accepts a float.(which is not a type I see in perl6).. and have that upgrade happen 14:12
s/pass to/pass two
timotimo you can 313.Num 14:13
haxmeister sub SFM(Num $rpm, Num $dia ) 14:14
SFM(318, 3);
timotimo no, SFM(318.Num, 3.Num)
haxmeister I misunderstood Num for one thing
timotimo if they are literals, just write them SFM(318e0, 3e0)
haxmeister these functions will be part of a program that prompts user input 14:15
14:16 pierre_ left
haxmeister so need to learn to have the SFM function type protected and user input type checked.. that's my learning goal 14:16
Voldenet I guess 'Numeric' role would work here
but you can use Numeric, Real or Rational roles depending on what you want to achieve 14:17
haxmeister sheesh.. I guess I need understand "role" types a little before I try to do typed math
MasterDuke_ haxmeister: coercers might also be of interest
haxmeister lol.. you guys 14:18
Voldenet haxmeister: if you did some C++ (i see you did), then roles in this case are pretty much similar to abstract classes
MasterDuke_ m: sub a(Num() $a, Num() $b){say "$a, $b"}(1.Int, 2.Int)
camelia 1, 2
MasterDuke_ the .Int was just to make it obvious that the Num() coercer was doing something
haxmeister yah MasterDuke_ I have no idea how to read or use a sub with all that at this point 14:19
Voldenet huh, any docs on coercers? 14:20
MasterDuke_ i think the only thing a little out of the ordinary there are the `Num()`. putting the parens on the end essentially calls .Num on whatever is called
docs.perl6.org/syntax/Coercion%20Type has something 14:21
Voldenet m: sub test($x){ "Coerced $x" }; sub test-two(test() $y){ say $y; }; test-two(2);
camelia 5===SORRY!5=== Error while compiling <tmp>
Invalid typename 'test' in parameter declaration.
at <tmp>:1
------> 3t($x){ "Coerced $x" }; sub test-two(test7⏏5() $y){ say $y; }; test-two(2);
Voldenet figured it's not that magical
haxmeister I get the coercion.. I don't understand properly what numbers should be typed as Num compared to Rat..etc
it's a lot of number based types... 14:22
Voldenet haxmeister: understanding the type graph is a key... :)
haxmeister: basically, Numeric is Real + Complex
moritz haxmeister: Num are floating-point numbers
haxmeister: and Rat are fractions
Voldenet Real is Rational + Int + Num + Duration + Instant
moritz haxmeister: in Perl 6, if you write 2.1, that's a Rat
Voldenet and so on
haxmeister 2.1 is a fraction? 14:24
because it looks like a float to me.. :/
huf it's 21/10, surely 14:25
haxmeister thank you all for being patient.. I really like perl6 but it's a big package coming in
moritz haxmeister: only with scientific notation (2.1e0) is it a Num/float in Perl 6
haxmeister ok
I have to think about that for a minute
timotimo aye, perl6 is trying to prevent you from shooting yourself in the foot
Voldenet contrary to perl5
;-)
timotimo m: my $num = 0.0; while $num != 1.0 { .say; $num += 0.1 } 14:26
camelia (Any)
(Any)
(Any)
(Any)
(Any)
(Any)
(Any)
(Any)
(Any)
(Any)
timotimo hah
m: my $num = 0.0; while $num != 1.0 { $num.say; $num += 0.1 }
camelia 0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
timotimo m: my $num = 0.0e0; while $num != 1.0e0 { $num.say; $num += 0.1e0 }
camelia (signal XFSZ)0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
1.9
2
2.1
2.2
2.3
2.4
2.5
2.6
2.7
2.8
2.9
3
3.1
3.2
3.3
3.4
3.5
3.6
3.7
3.8
3.9
4
4.1…
Voldenet m: my $num = 0.0e0; while $num 1.0e0 ~~ 1.0e0 { $num.say; $num += 0.1e0 } 14:27
camelia 5===SORRY!5=== Error while compiling <tmp>
Missing block
at <tmp>:1
------> 3my $num = 0.0e0; while $num7⏏5 1.0e0 ~~ 1.0e0 { $num.say; $num += 0.1e
expecting any of:
block or pointy block
infix
infix st…
Voldenet m: my $num = 0.0e0; while $num ~~ 1.0e0 { $num.say; $num += 0.1e0 } 14:28
camelia ( no output )
Voldenet m: my $num = 0.0e0; while $num !~~ 1.0e0 { $num.say; $num += 0.1e0 }
timotimo m: my $num = 0.0e0; while $num =~= 1.0e0 { $num.say; $num += 0.1e0 }
camelia (signal XFSZ)0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
1.9
2
2.1
2.2
2.3
2.4
2.5
2.6
2.7
2.8
2.9
3
3.1
3.2
3.3
3.4
3.5
3.6
3.7
3.8
3.9
4
4.1…
( no output )
timotimo m: my $num = 0.0e0; while $num !=~= 1.0e0 { $num.say; $num += 0.1e0 }
camelia 0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
Voldenet whoa !=~=
what a nice operator
m: $*TOLERANCE = 1e-30; my $num = 0.0e0; while $num !=~= 1.0e0 { $num.say; $num += 0.1e0 } 14:29
camelia Cannot modify an immutable Num
in block <unit> at <tmp> line 1
Voldenet ah ;)
timotimo you want "my" there
Voldenet m: my $*TOLERANCE = 1e-30; my $num = 0.0e0; while $num !=~= 1.0e0 { $num.say; $num += 0.1e0 }
So it's still possible to shoot yourself in the foot 14:30
timotimo of course
Voldenet Nice!
haxmeister you guys play entirely too much with that chat bot..lol
camelia (timeout)0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
1.9
2
2.1
2.2
2.3
2.4
2.5
2.6
2.7
2.8
2.9
3
3.1
3.2
3.3
3.4
3.5
3.6
3.7
3.8
3.9
4
4.1
4.2…
Voldenet Well, it's nice, because you can actually read some perl6 syntax and features while reading irc 14:31
along with all those 'funky features' you wouldn't know that exist 14:32
14:32 lowbro left 14:33 mcmillhj left 14:34 mcmillhj joined
haxmeister so 313.Num is the same as Num(313) ? we have Num as a method of several types including integer and string.. and we have Num a free standing function all available is that what I'm seeing? 14:37
313.Num being a method call on the Int type object.. and Num(313) being a call to a standard free standing function 14:38
timotimo no, not a method call on the int type object 14:39
it's a method call on an Int instance
haxmeister oh well. yah
really?
timotimo 313.Num wouldn't be able to give you 313e0 if it were a call on the type instance
because the type instance carries no value
haxmeister damn timotimo 14:40
is 318 and object?
timotimo of course
haxmeister and what class is that obect? 14:41
timotimo m: say 318.WHAT.perl
camelia Int
moritz m: say 318.^name
camelia Int
14:41 wamba left
haxmeister so 318 is an object of the class Int 14:41
timotimo right
AlexDaniel haxmeister: not just this chat bot. We have many bots here. We play with all of them ( ͡° ͜ʖ ͡°) 14:42
haxmeister therefore 318.Num is calling the method Num() from the Int class object yah?
timotimo yeah
haxmeister AlexDaniel: it's ok.. the playfulness makes for a happy environment
so what I said before is true
timotimo nah, "on" is a wrong word there 14:43
haxmeister lol
timotimo m: Int.Num
camelia Invocant of method 'Num' must be an object instance of type 'Int', not a type object of type 'Int'. Did you forget a '.new'?
in block <unit> at <tmp> line 1
timotimo ^- this is a method call on the Int type object
haxmeister so I've got myself confused with the notation we use a bit and what it means 14:44
timotimo yeah, that's fine 14:45
haxmeister I'm glad your patient
timotimo patient is easy :)
well, mostly
14:46 khw joined 14:47 g4 left
haxmeister so if I say 313 literally I am not instantiating an object of the Int class? 14:47
14:48 nadim left
timotimo you are, though it happens at compile time, which could make a difference to you 14:49
14:49 pat_js joined
haxmeister ah.. ok. that clarifies a bit 14:49
an example of explicitly instantiating an Int object with the value 313 would look how? 14:50
AlexDaniel m: say Int.new(‘313’) # maybe this? :| 14:51
camelia 313
AlexDaniel but honestly it doesn't really make sense to do something like this
m: say 313 # this is just fine…
camelia 313
timotimo m: my $try = 0; while $try != 313 { $try++ }; say $try 14:52
camelia 313
haxmeister no it doesn't.. but to understand how perl6 is using dot notation it helps me
because 313.Num is a thing
just like "somestring".say 14:53
timotimo right
haxmeister using object notation on a literal is a different thing for me
timotimo what is "object notation" to you? 14:54
haxmeister so that's why I ask about whether it's an object or not.. and then further how to use normal object instantiation method for it
14:54 titsuki left
timotimo you need to know that "method new" is nothing special, either 14:54
haxmeister are all anonymous objects created at compile time? 14:56
moritz ponders whether his next book should be "Perl 6 Complete: Object Orientation"
haxmeister: no
haxmeister moritz: that question makes sense to you?
moritz my $x = {}; # anonymous hash created at compile time
erm sorry, at run time
moritz confused
even classes can be created at run time
haxmeister lol 14:57
so much badassery in perl6..lmao
timotimo i can't actually find the code that makes Num(123) tick
14:57 azawawi joined
moritz m: my $class := Metamodel.ClassHOW.new_type(name => 'funny'); 14:58
camelia No such method 'ClassHOW' for invocant of type 'Metamodel'
in block <unit> at <tmp> line 1
14:58 lizmat joined
moritz m: my $class := Metamodel::ClassHOW.new_type(name => 'funny'); 14:58
camelia ( no output )
haxmeister moritz: to be honest.. I think for this language it is almost better to start a tut with the OO stuff and work down 14:59
azawawi hi 15:00
haxmeister since basic typing is connected to OO syntactically and actually
moritz haxmeister: sure, start simple, then get down to the nitty gritty
hi azawawi
jnthn Integer literals are a little bit interesting, because - at least for those that are small enough - the compiler is happy to treat them as both an Int object or a native int. In fact, it compiles them - in the immediate - into both, and lets the surrounding context decide which is wanted.
azawawi github.com/azawawi/perl6-terminal-...-world.pl6 # Terminal::Caca method chaining. Feedback is welcomed 15:01
haxmeister damn
that's hardcore
jnthn So if you say 42.WHAT then you're absolutely talking about the Int because .WHAT is a thing that's done to objects
But if writing `my int $answer = 42` then it can see there's no point generating code that takes a Int and then takes the int out from it, it can just generate code with the native 42 integer value right away. :) 15:02
moritz azawawi: nice
jnthn In general, though, Perl 6 lets you consider everything an object. Even if you have a native integer, and try to use it in an object-y way, it'll box it into an Int for you automatically 15:03
moritz so currently I'm improving my manuscript ("Perl 6 Fundamentals") in response to feedback
and in the back of my mind, I'm considering what to do next
and one of my ideas was to explore a Perl 6 topic in more depth
maybe in the form of a short book 15:04
jnthn Maybe some day I should write "Perl 6 from the bottom up", to steal the "Git from the bottom up" idea. :P
moritz jnthn: I'd buy that :-)
and the topics I've come up so far that I could cover are: routines and signatues, OO, Regex and grammars
MasterDuke_ jnthn: i've mentioned this to timotimo as an idea, but i would love a blog post (or something more substantial) that goes into some detail about exactly what happns when you `perl6 -e 'say "hi"'` 15:05
moritz the other I had was to ask MJD what he thinks about a Perl 6 version of HOP
MasterDuke_ i.e., what's happening in rakudo, what in nqp, what in moarvm 15:06
jnthn MasterDuke_: omg so much :-)
moritz or a Friedl if he would like a Perl 6 edition of "Mastering Regular Expressions"
El_Che moritz: do grammars
moritz or maybe I should stop writing for a bit, and explore podcasting
or... dunno
just floating ideas around
MasterDuke_ jnthn: heh, you're the one who said "book" 15:07
moritz El_Che: vote registered :-)
El_Che moritz: als a good (deep and wide) multithreading chapter would be nice
also
MasterDuke_ oops, implied "book" maybe
moritz El_Che: that would be nice indeed, but I'm the wrong person to write that :-)
lizmat .
yoleaux 14:39Z <Zoffix> lizmat: should List.sink actually do things? I would've figured it'd do self.iterator.sink-all, since a List could be made from a Seq, for example. And we do sink-all in Seqs, but not lists: irclog.perlgeek.de/perl6-dev/2017-...i_14525429
14:49Z <Zoffix> lizmat: Never mind. jnthn++ answered
jnthn Well, a Perl 6 guts-y book would certainly have to go in the queue behind the first one I really want to do :-) 15:08
moritz jnthn: tell me more about the first one :-)
MasterDuke_ ooo, an async/multi-threading one?
jnthn MasterDuke_ guessed :) 15:09
MasterDuke_ i would buy that
moritz too 15:10
15:10 Guest80741 left 15:11 KDr2 joined, araraloren joined 15:12 Ven joined, Ven is now known as Guest37161 15:14 robertle joined
haxmeister that is very very insightful jnthn i'm glad my difficulty was understood.. with that information the syntax is far less confusing 15:17
15:19 azawawi left
haxmeister having perl6 allow you to use what might be a literal in the context of object notation certainly sets it above many other language.. it is a higher higher level language in that sense 15:19
[Coke] jnthn++ # I'd buy that book 15:20
haxmeister though that does make it difficult to look up documentation when you don't know if you should be looking for a class method of a particular objects type or a built-in function 15:21
15:22 skids joined 15:26 mr-foobar left
haxmeister so only the interpreter deals with actual literals.. everything is an object and all subs/functions/methods are accepting objects as their arguments 15:26
though usually they are anonymous objects 15:27
15:29 keix joined, KDr2 left 15:31 yht left 15:34 mr-foobar joined 15:36 Sgeo joined 15:37 Sgeo_ left
[Coke] docs should /wni 8 15:42
aha, I fooled the bot.
haxmeister oh.. block syntax.. very awesome.. 15:44
could right some outrageous macros type stuff with that
15:45 wamba joined, Cabanossi left
haxmeister can you pass perl6 code to a function or even block for execution? 15:46
15:47 domidumont joined, Cabanossi joined
timotimo functions are objects, you can pass objects as arguments 15:47
haxmeister all arguments are objects.. but didn't think about functions being objects thank you
timotimo m: my $foo = -> $a, $b { say "$a and $b" }; sub call-it($it) { $it("one", "two") }; call-it($foo)
camelia one and two
haxmeister interesting.. perl6 can learn and modify itself then 15:49
timotimo that's not it
just passing code objects around doesn't give you self-modification yet
but you can EVAL stuff and that can do what you want, probably
15:50 itsame joined
haxmeister EVAL sounds familiar 15:50
MasterDuke_ and you can modify the grammar at runtime
haxmeister if I can take a string that is valid perl6 code and then evaluate that string as code.. that's all it takes
timotimo MasterDuke_: i wouldn't call that runtime, though? 15:51
haxmeister can we create data structures by evaluating strings of perl6 code that create a data structure? 15:52
timotimo yeah, you have to go through EVAL for that
because code is code 15:53
MasterDuke_ timotimo: you wouldn't?
15:53 Guest37161 left, ChoHag joined
timotimo MasterDuke_: you pull in changes to the grammar during compile time, no? 15:53
MasterDuke_ m: sub prefix:<!>($n) { [*] 1..$n }; say !5 15:54
camelia 120
timotimo at run time, you can no longer change the grammar that was used to parse the code 15:55
jnthn That's a compile time change, not runtime. Otherwise we don't know how to parse the added operators when we encounter them in later code :)
postfix:<!> is probably a better example 'cus that doesn't exist in standard Perl 6 :) 15:56
MasterDuke_ ah, right
jnthn m: sub postfix:<!>($n) { [*] 1..$n }; say 5!
camelia 120
jnthn m: say 5!; sub postfix:<!>($n) { [*] 1..$n }; # also notice order matters
camelia 5===SORRY!5=== Error while compiling <tmp>
Negation metaoperator not followed by valid infix
at <tmp>:1
------> 3say 5!7⏏5; sub postfix:<!>($n) { [*] 1..$n }; # a
expecting any of:
infix
infix stopper
jnthn Can't use the operator before it's defined
haxmeister but with EVAl though.. have we completely taken the heart of lisp and put it in perl or is it more restrictive?
jnthn 'cus it's not even in the pare tables
*parse
MasterDuke_ m: say !5; sub prefix:<!>($n) { [*] 1..$n };
camelia 120
timotimo perl6 code isn't isomorphic or what the name is
MasterDuke_ that's a little confusing 15:57
but i guess if you're creating custom operators you should know what's happing there 15:59
haxmeister so eval can't work on perl6 code but it can work on perl5?
it just has the option "perl"
MasterDuke_ m: EVAL q|say "hi"| 16:00
camelia hi
jnthn The Perl 6 EVAL function assumes code is Perl 6 by default
I think if you're using Inline::Perl5 you can use pass :lang<perl5> as an option or something like that 16:01
haxmeister if I my Str $foo = "my $bar =\"hello world\"" now $bar is available to me?
timotimo no
haxmeister because I can eval $foo and have $bar available?
timotimo nope
haxmeister kk 16:02
timotimo lexical pads are a compile-time thing and are fixed as soon as the compiler's done
so you can't put in new lexical variables by calling EVAL at run time
haxmeister oh I gotcha.. there's the difference
can't access them or can't create them?.. 16:04
16:04 ChoHag left
haxmeister if I EVAL a string of code that includes functions and vars etc.. will it evaluate or say no 16:04
jnthn They will be compiled as if in a nested scope 16:05
So you can put functions and vars in there and call them inside of the EVAL, they just won't be available outside
haxmeister awesome 16:06
jnthn If you want that you can either (for one function) return it as the result of the EVAL, or make it escape in some other way (assign it to a lexical outside of the EVAL, stick it in a package, assign it to a dynamic variable...basically anything non-lexical :))
m: my &func; EVAL '&func = sub { say 42 }'; func()
camelia 42
jnthn That works for example 'cus &func is declared outside of the EVAL 16:07
haxmeister sheesh I need to back up.. I have fundamentals to get used to in perl6
had to go here docs.perl6.org/language/5to6-nutshell to see why the space isn't allowed between a function call and it's parenthesized arguments 16:11
that explanation might should be on the doc page about subs/functions.. 16:13
in addition to this page
TreyHarris Design opinions sought: for a portable interface to a system resource (in this case, a PID, process ID) better to have a portable MyApp::PID with a has-a that will dispatch to a OS-specific object the user will never see (even if they do a .WHAT), or create a factory method as the entry point so the user gets a MyApp::PID::Linux or MyApp::PID::Darwin, both is-a to an uninstantiable MyApp::PID etc. depending on 16:14
what OS they're on? (Presumably the factory in the latter case would live in MyApp::PID.)
I should just clone the entire ecosystem so I can ack around for possible conventions already emerg(ed|ing) for things like this. How large is it, I assume the hundreds of megs at most? 16:17
(Anybody already written a CPAN Mirror-style ecosystem downloader? Simple enough to iterate through perl6/ecosystem/META.list, but there might be unusual cases and why reinvent wheels?) 16:18
16:19 garas joined
MasterDuke_ huggable: all modules 16:19
huggable MasterDuke_, github.com/moritz/perl6-all-modules
MasterDuke_ TreyHarris: ^^^
TreyHarris MasterDuke_: ty!
garas what is the perl6 equivalent of perldoc ? 16:20
TreyHarris garas: p6doc or perl --doc depending on usage
sorry, perl6 --doc
16:21 Geth left
garas ===> Building: p6doc No such method 'abspath' for invocant of type 'IO::Path' 16:22
TreyHarris garas: with rakudo, you'll need to run 'p6doc-index build' first
16:22 rindolf joined
garas I think it is broken 16:22
16:22 Geth joined, ChanServ sets mode: +v Geth
timotimo yeah, it's "absolute" instead of "abspath 16:22
"
but isn't it supposed to just give a deprecation warning? 16:23
16:25 mcsnolte joined
garas seems easy to fix 16:25
TreyHarris hemms and haws about whether to clone moritz/perl6-all-modules in ~/src/perl6/ALL-MODULES or something for lower cognitive load and better tabbability, or under ~/src/perl6-all-modules to preserve ~/src/perl6 having only the project modules inside for a 1:1 relationship to the GitHub path 16:26
MasterDuke_ do both, disk is cheap! 16:28
and why not ~/src/perl6/modules/all also 16:29
timotimo just symlink it :P
one thing that annoys me to almost no end is that i tend to perf record while in my ~, but there's also a "perl6" folder where all my perl6 stuff is in
and perf creates "perf.data" 16:30
TreyHarris hah! the size of all modules is 236 MB. I was guessing based on pure gut instinct about a quarter gig. 16:31
timotimo that's not only the size of modules, but also their history
TreyHarris yeah, just noticed that. modules-only is 171M
timotimo and there might be a few .so files in there
garas p6doc has been fixed already on github, but zef downloaded me the broken version..
TreyHarris timotimo: not according to find 16:32
timotimo OK
TreyHarris garas: p6doc is distributed with rakudo, not separately, I believe?
timotimo TreyHarris: it's in rakudo star, not in rakudo itself
garas using rakudobrew
16:33 itsame left
TreyHarris timotimo: sorry, I thinko'd 16:33
timotimo :) 16:34
16:34 ChoHag joined, Zoffix joined
Zoffix garas: how come you're using rakudobrew? 16:35
TreyHarris Well, that's useless... not a single class in the ecosystem has "Linux" or "Darwin" in its name, believe it or not. (Actually, one has a "Darwin", but it's referring to the Australian city, not the OS.)
garas Zoffix: I have no idea..
Zoffix garas: hah
garas I use homebrew on the mac, and got used to it
Zoffix garas: well, it's not really for end users. For example, you need to run `rakudobrew rehash` after installing binaries like `p6doc` or stuff won't work right. Try running that now and see if that fixes your issue
TreyHarris s/class/namespace/ 16:36
garas thanks
Zoffix timotimo: nope, no deprecation for .abspath
timotimo OK. in that case i don't know why IO::Path would error out like that for a call to its .abspath
Zoffix .slurp-rest will be deprecated in 6.d... Don't remember offhand what else
timotimo: oh, no, .abspath just got removed, without any deprecation. 16:37
It wasn't part of the language, IIRC
timotimo oh!
that's how you meant it
Zoffix Yeah
buggable: eco doc 16:38
buggable Zoffix, Found 10 results: Pod::Parser, p6doc, XML::Parser::Tiny, HTML::Parser::XML, PerlStore. See modules.perl6.org/#q=doc
Zoffix buggable: eco p6doc
buggable Zoffix, p6doc 'Perl 6 documentation (tools and docs)': github.com/perl6/doc
TreyHarris Are there performance-based reasons to prefer the factory-is-a dispatch over the has-a dispatch, which would tend to reduce method dispatch by one for ever operation? Or is that optimized and/or negligible? (I can use 'handles' in some, but not all, cases, if there's an optimization for that.)
Zoffix garas: oh, I see why. We got dumb version in p6doc 16:39
TreyHarris (that was a misplaced modifier; the fatory is-a would tend to reduce method calls by one per operation, not the has-a)
Geth doc: 0893cdcdd5 | (Zoffix Znet)++ (committed using GitHub Web editor) | META6.json
User proper version

So that modules update stuff when we make new versions
Zoffix garas: zef --force install p6doc; rakudobrew rehash;
That should fix the issue
16:40 dakkar left
Zoffix IMO a META version of "*" should mean "install, always" to a package manager. 16:41
m: Version.new(*) before Version.new(*)
camelia WARNINGS for <tmp>:
Useless use of "before" in expression ".new(*) before Version.new(*)" in sink context (line 1)
Zoffix m: say Version.new(*) before Version.new(*)
camelia False
Zoffix hmmm
Zoffix shrugs
m: say Version.new(*) after Version.new('1.001001') 16:42
camelia False
Zoffix m: say Version.new(*) before Version.new('1.001001')
camelia True
Zoffix Good.
garas: also, p6doc is a far cry from perldoc at the moment. Still needs a bunch of work. Easier to just use docs.perl6.org 16:43
TreyHarris I'm leaning towards a factory in MyApp::PID returning MyApp::PID::$<OSNAME>. So, last question if anyone has an opinion: should such a factory be invoked on .new since it will just locate the right subclass and re-dispatch, or is it bad hygiene for .new to return anything other than an instance of exactly the class explicitly named and not a subclass?
garas Zoffix: agreed, I also like the website, but I am curious 16:45
Zoffix garas: did the fix work for you?
garas it installed it
--force was really important 16:46
also
Zoffix Sweet. Do you still have the .abspath problem?
TreyHarris Hmm, github.com/perl6/modules exists, hasn't been touched since 2009, and has a single module in it, Digest (using old syntax that probably won't work (right) anymore
garas no abspath problem
Zoffix Awesome.
16:48 keix left 16:49 zakharyas left
garas It seems to work decently :) 16:50
araraloren TreyHarris, a factory new is a good idea, I prefer that way. 16:51
Zoffix awwaiid: my feedback on your Terminal::Caca fluent interface is it's terrible for errors. For example this chain's death on line 25 is reported to be on line 9, because that's where the statement starts: gist.github.com/zoffixznet/2e8bd72...0dd6b75192 There's no reason to write that code in Perl 6, since you can use `given $o { .method1; .method2; .method3; }` etc 16:52
TreyHarris araraloren: I agree, but I just didn't know if it violated least surprise to call X.new() and get an instance of Y, even if Y is a X. 16:53
Zoffix awwaiid: though, as far as chains go, I have you beat with my juicy Perl 5 test module: metacpan.org/pod/Test::Mojo::Role:...r#SYNOPSIS :) 16:55
16:56 cdg left 16:57 Zoffix left
araraloren Em, I have seen that way in some Perl 5 Module. In Perl 6 , new is just a construct method. 16:59
17:00 Cabanossi left
araraloren TreyHarris, Just write a clear document for your module. It's no problem . 17:01
TreyHarris araraloren: well, maybe it should be a role then. 17:02
17:02 Cabanossi joined
TreyHarris nope, can't be a role, it has behavior ("class methods") I need to call on it even before OS-dispatch. 17:02
araraloren Em, Night ..I gotta sleep .. 17:03
Or maybe you can make a other name for it. Such as `method from-os(..)` 17:05
17:05 araraloren left
TreyHarris m: role Dock { method bark { say "quack!" } }; Dock.bark; class Dog does Dock { }; my $d = Dog.new; $d.bark 17:08
camelia quack!
quack!
TreyHarris Hmm, guess you can call methods on roles... is that accidental? 17:09
timotimo that's what's called "punnign"
"punning"
it does the equivalent of class anonymous does TheRole { } and calls the method on that
TreyHarris ....right! 17:10
forgot about that.
17:13 cdg joined
TreyHarris it's a PITA to rewrite .new in this case, because if you do it stupidly (making the role or superclass dispatch to the class's .new) you go into an infinite loop. better to name the factory something else I guess, unless this is a case for using bless, and I think it's not. 17:17
I've always like the Recognizer pattern (where a class looks at its input and returns an object of the appropriate isomorphic type to deal with the input, with the instance classes themselves being polled "can you take this input?" and letting them reply "yes", "no", or "maybe, but come back to me if nobody else says "yes"), and multis make it a lot easier to write them. but they still aren't exactly concise 17:21
in any OO language I've encountered.
timotimo having roles participate in the build dance is difficult, aye 17:28
normally you have a submethod BUILD and/or TWEAK, but submethods, like methods, get flattened into the class body that you're applying to
17:29 alimon left, cdg left
TreyHarris Easier just to separate out the build function into a factory and putting any recognition logic into "class methods" (is that an official term in Perl 6, or do we say "undefined invocant methods?"), letting the ordinary BUILD & TWEAK process work as normal. This also lets you put the participating methods into a class tree if you so desire, letting slightly-different classes capitalize on one another's 17:31
implementations.
timotimo hm, i wonder if a submethod BUILD will also invalidate the default values for attributes in roles you merge into the class 17:33
m: role WithDefault { has $.foo = rand }; class WithBuild does WithDefault { submethod BUILD(:$lol) { say $lol } }; WithBuild.new(:99lol).foo.say 17:34
camelia 99
0.977413050909244
timotimo cool, that works
17:34 alimon joined 17:36 azawawi joined
azawawi hi 17:36
17:37 anop joined
anop what is "do for" in perl 6 17:37
timotimo it lets you use a for loop in an expression context 17:38
to get its value
well, values
anop thanks but why we use "do " with for ? 17:39
[Coke] anop: docs.perl6.org/syntax/do 17:40
azawawi Zoffix: thx for your feedback to my synonym nickname :)
TreyHarris timotimo: can you think of any examples where both a TWEAK and a BUILD would be used aside from as an academic exercise? It seems to me like the rule of thumb is, "use TWEAK if you can, BUILD otherwise, and new only if you must" (which I guess kind of blows up the whole CAPS-vs-lowercase thing)
anop thnks timotimo 17:41
" (.=not for @doors[0, $_ ... 100]) for 1..100;" in this code, could you explan what is " .=not " ? 17:44
timotimo it's a mutating method call on $_ 17:45
(when an object before the method call is left out, but the . is there, it is a call on $_)
.=not is equivalent to $_ = $_.not 17:46
17:46 Cabanossi left
timotimo TreyHarris: i don't think there's a good reason to have both 17:46
anop thnks u so much timotimo.....
17:47 Cabanossi joined 17:48 cdg_ joined 17:50 araujo left 17:54 labster joined, skids left, raschipi joined
garas Do you think it would be nice if perl6 had a help() function in the REPL, calling p6doc? 17:56
17:58 anop left
raschipi garas: Better done with two terminals or terminal tabs. 18:01
alphah m: my $u = $*USER; say $u.perl; say $u[0]
camelia IntStr.new(1012, "camelia")
camelia
alphah m: my $u = $*USER; say $u.perl; say $u[1] 18:02
camelia IntStr.new(1012, "camelia")
Index out of range. Is: 1, should be in 0..0
in block <unit> at <tmp> line 1

Actually thrown at:
in block <unit> at <tmp> line 1
alphah How to get the id number, not the user id?
MasterDuke_ m: my $u = $*USER; say $u.perl; say +$u 18:03
camelia IntStr.new(1012, "camelia")
1012
alphah aaah
Thanks
MasterDuke_ m: my $u = $*USER; say $u.perl; say ~$u
camelia IntStr.new(1012, "camelia")
camelia
raschipi m: my $u = $*USER; say $u ~~ 1012; 18:07
camelia True
raschipi m: my $u = $*USER; say $u ~~ 'camelia';
camelia True
18:11 setty1 joined 18:15 Cabanossi left 18:17 Cabanossi joined, azawawi left 18:18 itaylor57 joined 18:20 Ven joined
raschipi m: my $u = $*USER; say $u + 2; 18:20
camelia 1014
18:20 Ven is now known as Guest48770
MasterDuke_ docs.perl6.org/language/glossary#Allomorph 18:20
18:22 garas left
alphah m: say $*USER.Str; say $*USER.Num; 18:28
camelia camelia
1012
raschipi I'm aware.
m: sub f(Numeric $x){say $x}; my $y=<2>; f($y) 18:30
camelia 2
raschipi m: my $y=<2>; $y.WHAT
camelia ( no output )
raschipi m: my $y=<2>; say $y.WHAT
camelia (IntStr)
raschipi m: sub f(Numeric $x){say $x.WHAT}; my $y=<2>; f($y) 18:31
camelia (IntStr)
raschipi m: sub f(Int() $x){say $x.WHAT}; my $y=<2>; f($y)
camelia (IntStr)
18:31 cdg_ left
raschipi m: sub f(Int $x){say $x.WHAT}; my $y=<2>; f($y) 18:31
camelia (IntStr)
raschipi m: say IntStr ~~ Int; 18:32
camelia True
raschipi m: sub f(Any $x where 2){say $x.WHAT}; my $y=<2>; f($y) 18:35
camelia (IntStr)
raschipi m: sub f($x){say $x.WHAT}; my $y=<+2>; f($y) 18:36
camelia (IntStr)
raschipi m: sub f($x){say $x.WHAT}; my $y=<2+0>; f($y) 18:37
camelia (Str)
raschipi The note in the glossary is confusing.
18:40 Zoffix joined
Zoffix raschipi: which part? 18:40
The "Note:…"
?
raschipi Yes.
Zoffix How can it be better? 18:41
Basically angle brackets are dual-purpose: a list of stuff or a way to make Rat and Complex literals
And which you get depends on whether there's stuff other than the number in the brackets 18:42
raschipi It should say it doesn't work for Ints.
Zoffix Why Ints specifically?
raschipi m: say <+2>.WHAT
camelia (IntStr)
raschipi m: say <2+0>.WHAT
camelia (Str)
AlexDaniel oh please not <> again…
Zoffix <> is awesome. You're just not ready to receive all it's got to offer :P 18:43
raschipi I like it, very handy (especially for testing). My problem is with the note.
AlexDaniel for me it would've been better if there was a distinct syntax for every single thing you can get out of it 18:44
moritz m: say <2+3i>.^name
camelia Complex
AlexDaniel like I don't know, i<> or ~<> or whatever
18:46 Cabanossi left
Geth doc: cc1b62a0ce | (Zoffix Znet)++ (committed using GitHub Web editor) | doc/Language/glossary.pod6
Clarify allomorph vs literal numeric note
 18:47
Zoffix raschipi: ^ what about now?
18:47 Cabanossi joined
Zoffix
.oO( that entry is way to big for a Glossary entry )
 18:48
raschipi Now I understand how it works.
Zoffix Sweet. 18:49
18:49 Zoffix left 18:51 jonas1 left, eliasr left 18:55 jonas1 joined 18:58 labster left 19:04 darutoko left 19:08 domidumont left 19:19 TEttinger joined 19:21 jonas1 left 19:32 r3m left, espadrine_ joined 19:33 r3m joined 19:41 itaipu left 19:44 itaipu joined 19:45 Cabanossi left 19:47 Cabanossi joined 19:50 dct joined 19:55 itaipu left 19:56 labster joined 19:59 cdg joined 20:01 skids joined 20:05 raschipi left
Geth ecosystem: 941b77ba6a | (David Warring)++ | META.list
rename repos
 20:22
20:24 azawawi joined
TreyHarris I'm confused by this behavior I've made into a short test case. I'm following the documentation example so closely here that I'm sure I'm just making a typo or something, but I retyped this example from scratch and still getting the same thing: gist.github.com/treyharris/f26bb1a...cfd970b14b 20:24
azawawi github.com/azawawi/perl6-terminal-...n-sink.pl6 # Perl 6 random lines, boxes, circles, ellipses and triangles (also fill-) primitives :) 20:26
timotimo TreyHarris: that example is wrong 20:27
you're giving the proxy a method and it's trying to access $!x, but it'l installed on the proxy object
but Proxy doesn't have a $!x
TreyHarris timotimo: oh? I'll fix the example if you tell me what's wrong with it
timotimo please try "sub" instead of "method" in there
TreyHarris ok
moritz but if you use sub, you need a placeholder argument for the proxy 20:28
timotimo ah, yes
20:28 ggoebel left
timotimo first positional parameter 20:28
moritz so FETCH => sub ($) { $!x * 2 }
20:28 Perl6Noob joined
timotimo yup, and add a $ for the STORE sub, too 20:28
in front
moritz STORE => sub ($, $new) { $!x = $new / 2 }
timotimo that makes things work properly 20:29
moritz also, the tests assume that if you assign 3, you get a 6 out
but you're assinging 3/2 to $!x, and then fetch 3/2 * 2, so 3
ugexe META version of "*" cant mean install always to a package manager without also adding logic to ignore dependencies and a bunch of other things. Instead, a META version of * is treated just like how rakudo treats it: as version 0 20:30
moritz erm wait, you're storing 3*2, and then retrieving 3*2/2
TreyHarris huh? the tests pass now and I just changed "method (" to "sub ($,"
Perl6Noob ugexe, why ignore dependency? 20:31
Trey
moritz TreyHarris: oh, because you assign to x, not to double-x
TreyHarris: never mind my ramblings
ugexe because dependencies can overlap versions
Perl6Noob TreyHarrys, because sub closes over $!x, while when it was a method it was trying to get it from elsewhere 20:32
ugexe now you have to introduce special logic for when versions overlap and you only want one or the other and not both
TreyHarris moritz: oh, yes, I wanted to simplify down to bare bones so I used a name for the proxy that wasn't overloaded at all
ugexe there are tons of details like this
declaring version * in META6.json should just be barred
TreyHarris Perl6Noob: excuse me? ignore what dependency?
Perl6Noob ugexe, ok
TreyHarris, nah, that was to ugexe 20:33
20:33 dct left
TreyHarris In the docs there are examples of proxy with "STORE => method" and "FETCH => method" where it's a code snippet, not a runnable example... when (if/ever) is "method" appropriate/ 20:35
?
There are no examples with FETCH/STORE => sub, so I feel like I should change one of them to that at least so they won't spend 2 hours like I just did trying all sorts of things but never thinking to make that particular change 20:36
Perl6Noob TreyHarris, when it's self-contained or closes over just normal variables. It really just saves you from adding an invocant to parameters list. And it failing the way it does when you closed over $!x may be a bug... Like if you do it in regexes you get a proper error about it
20:36 rindolf left
Perl6Noob IMO a block is the best thing to use there 20:38
least typing
TreyHarris Okay, but in what context would you close it over normal variables when creating an "is rw" accessor method? That's what led me to the example in the first place, docs saying to use a Proxy to achieve complex read/write accessor behavior 20:39
Perl6Noob: could you illustrate? I'm not sure what you mean. oh, you mean just "STORE => { $!x * 2 }" and "FETCH => { $!x = $^x / 2 }"? 20:40
timotimo i don't think that is enough, because of what it'll be called as
TreyHarris the second one wouldn't get $^x as the invocant? 20:41
timotimo oh
sorry, that's correct
but the STORE one should get two arguments
20:41 ggoebel joined
TreyHarris it does, I just tested it as above, except I flipped STORE and FETCH, sorry, thinko 20:43
20:43 Perl6Noob left
TreyHarris So how do you make a placeholder that matches the second positional and throws away the first? 20:43
oh, he's gone
.seen Perl6Noob 20:46
yoleaux I saw Perl6Noob 20:38Z in #perl6: <Perl6Noob> least typing
TreyHarris duh, I meant before that. :-) 20:48
20:49 Perl6Noob joined
TreyHarris Perl6Noob: I'm just wondering if you were alluding to some other "shorter way" using blocks, like perhaps avoiding the Proxy constructor completely? 20:50
Because "STORE => { $!x = $^x / 2 }" doesn't work because you need two positionals and that only takes one. (a placeholder-less block will take any arguments and discard them, so FETCH is okay) 20:51
Perl6Noob TreyHarris: just blocks. For second one, I'd use -> $, $v { $!x = ½*$v }, but you can also write as { $!x = ½*@_[1] }
Geth perl6.org: b3660e91d0 | (Steve Mynott)++ | source/downloads/index.html
Mention current Strawberry Perl dep for MSI
 20:52
Perl6Noob TreyHarris: also, you can shorten bare blocks and write them as :STORE{ $!x = ½*$^x }
m: say '_' before 'a'
camelia True
Perl6Noob m: say { $^_; say ½×$^x }(42, 10) 20:53
camelia 5===SORRY!5=== Error while compiling <tmp>
Redeclaration of symbol '$^_' as a placeholder parameter
at <tmp>:1
------> 3say { $^_7⏏5; say ½×$^x }(42, 10)
Perl6Noob Oh right.
Oh huh
m: say { $^z; dd $_; say ½×$^x }(42, 10)
camelia Any $_ = Any
21
True
Perl6Noob oh ok
TreyHarris Perl6Noob: how with :STORE? :FETCH{ ... } works, because, again, you don't need the argument
Perl6Noob m: say { $^a; say ½×$^x }(42, 10)
camelia 5
True
Perl6Noob If you really wanted to do stuff
TreyHarris: oh sorry, I meant with :FETCH{}
TreyHarris I think it looks odd to use two different syntaxes there. 20:54
Perl6Noob :FETCH(* × 2) 20:56
Never mind :)
TreyHarris m: say { $^z; dd $_; say ×$^x}(42) 20:58
camelia 5===SORRY!5=== Error while compiling <tmp>
Preceding context expects a term, but found infix × instead
at <tmp>:1
------> 3say { $^z; dd $_; say ×7⏏5$^x}(42)
Perl6Noob TreyHarris: method double-x { Proxy.new: :FETCH{$!x × 2} :STORE(-> $, $y { $!x = ½×$y }) } 20:59
That's the shortest I can think of
TreyHarris I'm not quite sure what was supposed to happen no matter how many args it got
Perl6Noob TreyHarris: I think your system is missing fonts or something? There supposed to be a ½ before multiplication 21:00
as in:
u: ½
unicodable6 Perl6Noob, U+00BD VULGAR FRACTION ONE HALF [No] (½)
Perl6Noob And that code wouldn't compile even if you added it, 'cause the sig is for two args, but you gave only one
TreyHarris wait, how does "$!x = ×$y" work? What does unary × do?
Oh, I see 21:01
Perl6Noob Hah
TreyHarris Yes, I can't read that one half
Perl6Noob m: say { ½×$^x}(42)
camelia 21
TreyHarris I see "...ONE HALF [No] ()"
Perl6Noob Look! Unary × divides by 2!
m: say { ¼×$^x}(42)
camelia 10.5
Perl6Noob And now it divides by 4. It's time-sensitive operator :P 21:02
21:02 setty1 left
TreyHarris m: say ×6 21:02
camelia 5===SORRY!5=== Error while compiling <tmp>
Preceding context expects a term, but found infix × instead
at <tmp>:1
------> 3say ×7⏏056
Perl6Noob hah
TreyHarris m: say(×6)
camelia 5===SORRY!5=== Error while compiling <tmp>
Preceding context expects a term, but found infix × instead
at <tmp>:1
------> 3say(×7⏏056)
TreyHarris m: my $x = ×6; $x.say
camelia 5===SORRY!5=== Error while compiling <tmp>
Preceding context expects a term, but found infix × instead
at <tmp>:1
------> 3my $x = ×7⏏056; $x.say
TreyHarris hmm
Perl6Noob m: sub infix:<⁣> { $^a + $^b }; say 2⁣2 21:03
camelia 4
Perl6Noob ehehe
21:03 bjz joined
Perl6Noob TreyHarris: There's no prefix ×. This is how it's all supposed to look like: i.imgur.com/fqmr3VT.png 21:04
TreyHarris Sorry, I was looking at how cross-product works on scalars because I couldn't remember the differences between scalar cross-product and multiplication, and so I missed your joke
Perl6Noob scalar cross-product is for vectors and stuff. 21:06
I even invented a shorter trick to calculate it than what they teach at school.
True Story!
21:06 Perl6Noob left
TreyHarris I know, but I thought maybe somehow promoted to a vector and then de-promoted back to a scalar, somehow you were getting a one-half operation 21:06
21:07 itsame joined, itsame left 21:08 itsame joined
TreyHarris m: (1 + q++).say 21:09
camelia 1
TreyHarris m: my \q = 0; (1 + q++).say
camelia Cannot resolve caller postfix:<++>(Int); the following candidates
match the type but require mutable arguments:
(Mu:D $a is rw)
(Int:D $a is rw)

The following do not match for other reasons:
(Bool:D $a is rw)
(Bool:U $a is …
TreyHarris Crazy stuff like that :-P
21:09 labster left
timotimo yeah, \q doesn't have a container to it 21:10
you can add one, though 21:12
21:12 mcmillhj left
timotimo m: my \q := my $ = 0; (1 + q++).say 21:12
camelia 1
TreyHarris you should probably get at least a warning if you attempt to declare a siggilless variable with the same name as a quoting construct
timotimo oh, now i get why you brought this up 21:13
TreyHarris m: q++ === "" 21:15
camelia WARNINGS for <tmp>:
Useless use of "===" in expression "q++ === \"\"" in sink context (line 1)
TreyHarris m: say so q++ === ""
camelia True
timotimo that's a hilarious way to trip up people
21:16 Cabanossi left
TreyHarris "q++" to get the empty string is a stable of obfuscated Perl entrants. :-) 21:16
timotimo q++, qq++, Q++, either of those
21:17 Cabanossi joined 21:18 perlpilot joined
TreyHarris m: my @f = qw;;; @f.say 21:18
camelia []
21:19 labster joined
timotimo m: say s+++ 21:19
camelia 5===SORRY!5=== Error while compiling <tmp>
Null regex not allowed
at <tmp>:1
------> 3say s+7⏏5++
timotimo m: say s+a++
camelia Use of uninitialized value of type Any in string context.
Methods .^name, .perl, .gist, or .say can be used to stringify it to something meaningful.
in block <unit> at <tmp> line 1
Nil
timotimo m: s+a++ 21:20
camelia Use of uninitialized value of type Any in string context.
Methods .^name, .perl, .gist, or .say can be used to stringify it to something meaningful.
in block <unit> at <tmp> line 1
timotimo bleh %)
m: $_ ="waaaaaat"; s+a++; .say
camelia waaaaat
timotimo m: $_ ="at"; s+a++; .say
camelia t
timotimo m: $_ ="wat"; s+a++; .say
camelia wt
TreyHarris ach, you beat me
timotimo that's wat i meant
azawawi timotimo: github.com/azawawi/perl6-terminal-...n-sink.pl6 # .polyline primitive :) 21:25
TreyHarris now, I just need to remember WHY I wanted a read-write attribute method in the first place 3 hours ago ;-)
I learned lots of random tidbits along the way, though. Like, putting .resume in a CATCH when clause that can match both resumable and unresumable exceptions can make your program go into an infinite recursion.... 21:29
21:38 mcmillhj joined 21:44 mcmillhj left, mcmillhj joined 21:46 Cabanossi left 21:47 Cabanossi joined 21:49 ChoHag left, mcmillhj left 21:54 anton_p left 21:56 azawawi left 21:59 itsame left 22:00 mcmillhj joined 22:05 mcmillhj left
ugexe it'll stop eventually... :) 22:06
22:10 Sound joined, Sound left, Sound joined 22:15 wamba left 22:16 mcmillhj joined 22:23 aindilis` joined 22:24 kyclark joined, Praise- joined, Praise- left, Praise- joined, smash left 22:25 Mithaldu left, mcmillhj left, sergot_ joined, El_Che_ joined 22:26 integral joined, integral left, integral joined, petercom1and joined, rblackwe_ joined, haxmeist1r joined, tinita_ joined, sivoais_ joined 22:27 Mithaldu joined 22:28 ggoebel_ joined, ufobat left, petercommand left, ggoebel left, aindilis left, Praise left, haxmeister left 22:29 Sound left, Praise- is now known as Praise 22:30 smash joined 22:36 mcmillhj joined 22:42 mcmillhj left 22:43 tinita_ is now known as tinita
kyclark I would love to understand Perl 6 Grammars better, so I’ve started off with a simple one to parse commented lines (starting with “#”) from not: lpaste.net/355166 22:47
It does not work, so I would love some help fixing this. 22:48
22:55 cdg left 22:56 mcmillhj joined
raiph kyclark: looking now. have you used Grammar::Tracer? 22:57
gfldex kyclark: your comments can only have one '#' without anything following it 22:58
kyclark Ah, I remember someone pointing me to this before. How do I look at the docs for this? I don’t see it on modules.perl6.org
skids kyclark: your comment rule doesn't eat anything but the #
gfldex `zef install Grammar::Debugger` 22:59
github.com/jnthn/grammar-debugger
23:00 mcmillhj left
raiph kyclark: in a P6 regex matches any character, including a newline 23:01
er, . in a P6 regex
so <data> gobbles the rest of the file in one go, and (by default) tokens don't backtrack 23:02
skids kyclark: also, your TOP rule asks for at least one "line" (with no newlines needed) then followed by any number of newlines
raiph so the final \n+ fails
kyclark I was grasping at straws over how to write TOP. I found the brackets “[]” somewhere that made me think “many of these” but then there’s also “<line>+” which I think also means “one or more <line>s" 23:03
What do the “[]” mean? 23:04
raiph kyclark: it just brackets some atoms together
without capturing (storing what matched)
so you could, for example, write [ foo \n ]+ to match one or more lines (consisting of just foo and a newline) 23:05
if you wrote ( foo \n )+ you'd do the same matching but you'd also store the result 23:06
m: / < foo \n > / # presumably a compile error? 23:08
camelia ( no output )
raiph m: say / < foo \n > / for "\n", 'foo', 42 ; # ah? 23:10
camelia / < foo \n > /
/ < foo \n > /
/ < foo \n > /
raiph m: say m/ < foo \n > / for "\n", 'foo', 42 ; # ah?
camelia Nil
｢foo｣
Nil
raiph m: say m/ < foo \n > / for '\n', 'foo', 42 ; # ah?
camelia ｢\n｣
｢foo｣
Nil
23:12 mcmillhj joined 23:14 espadrine_ left 23:17 mcmillhj left
kyclark Are there any docs on grammars I should read? 23:22
23:22 bjz left
kyclark E.g., I think I have a grammar that’s working, but how do I get the parsed data out? 23:22
raiph m: grammar g { token TOP { [ '#'? \N+ \n ]+ } }; say g.parse: "foo\n#bar\n" 23:24
camelia ｢foo
#bar
｣
raiph m: grammar g { token TOP { [ [ <comment> | \N+ ] \n ]+ }; token comment { '#' \N+ } }; say g.parse: "foo\n#bar\n"
camelia ｢foo
#bar
｣
comment => ｢#bar｣
raiph m: grammar g { token TOP { [ [ <comment> | \N+ ] \n ]+ }; token comment { '#' \N+ } }; say (g.parse: "foo\n#bar\n")<comment>
camelia [｢#bar｣]
23:25 Actualeyes left
kyclark What does “^^” mean in a token? 23:25
raiph m: grammar g { token TOP { [ [ <comment> | \N+ ] \n ]+ }; token comment { '#' \N+ } }; say (g.parse: "foo\n#bar\nbaz\n#quux\n")<comment>
camelia [｢#bar｣ ｢#quux｣]
raiph m: grammar g { token TOP { [ [ <comment> | \N+ ] \n ]+ }; token comment { '#' \N+ } }; say (g.parse: "foo\n#bar\nbaz\n#quux\n")<comment>[1]
camelia ｢#quux｣
raiph kyclark: start of a line; ^ means start of a string (which may contain multiple lines) 23:26
23:28 mcmillhj joined
raiph kyclark: docs.perl6.org/language/regexes#^^...nd_of_Line 23:28
timotimo it'd be kinda cool if we could put the url of a gist or other kind of paste and follow it up with some extra code 23:30
23:32 mcmillhj left
kyclark It’s works if I say “token TOP { [ [ <comment> | <data> ] \n ]+ }” 23:33
But not if I say “token TOP { [ <line> \n ]+ }"
Where “token line { <comment> | <data> }"
lpaste.net/355166 23:34
23:35 mcmillhj joined
kyclark It fails to match “line” after it’s done parsing all the data, so it ends with a FAIL. 23:38
raiph i'm getting failure in *both* cases 23:40
i think...
23:40 mcmillhj left
raiph yeah 23:41
23:41 BenGoldberg joined 23:42 Guest48770 left
raiph token TOP { [ <line> \n ]+ } will fail because the last (empty) line won't match 23:43
token TOP { [ [ <comment> | <data> ] \n ]+ } fails for same reason
both the comment and the data tokens require at least one character 23:44
(btw, the comment perhaps ought to be '#' \N* rather than '#' \N+) 23:45
timotimo you should check out the % operator, which you can put after a quantifier and it'll mean "and in between the things that are quantified, match this"
and %% is "optionally, match the in-between thing at the end once"
raiph timotimo: do you mention that because of the very specific scenario of lines of two types (comments and not-comments)? 23:46
timotimo well, lines that end in \n, except at the end of the file where it may not have a \n in between
raiph ah 23:47
kyclark This works: lpaste.net/355166
Dang, I have to go home now. Maybe I can call again sometime soon? 23:48
raiph m: grammar g { token TOP { [ <comment> | <data> ]+ %% \n+ } }
camelia ( no output )
raiph kyclark: hope you do
23:49 kyclark left
raiph m: grammar g { token TOP { [ [ '#' \N* ] | .+ ]+ %% \n+ } }; say g.parse: "a\n#b" 23:50
camelia ｢a
#b｣
raiph m: grammar g { token TOP { [ [ '#' \N* ] | \N+ ] \n ]+ } }; say g.parse: "a\n#b" 23:51
camelia 5===SORRY!5=== Error while compiling <tmp>
Malformed regex
at <tmp>:1
------> 3 { token TOP { [ [ '#' \N* ] | \N+ ] \n 7⏏5]+ } }; say g.parse: "a\n#b"
raiph m: grammar g { token TOP { [ [ '#' \N* | \N+ ] \n? ]+ } }; say g.parse: "a\n#b" 23:52
camelia ｢a
#b｣
raiph m: grammar g { token TOP { [ [ '#' \N* ] | \N+ ]+ %% \n? } }; say g.parse: "a\n#b" 23:53
camelia ｢a
#b｣
23:56 kurahaupo joined

Powered by Cro and the Raku® Programming Language.

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

Thank you!