00:40 lizmat_ joined, TempIRCLogger__ joined 00:41 RakuIRCLogger left 00:42 TempIRCLogger left, lizmat left 00:44 TempIRCLogger__ left, TempIRCLogger joined 03:52 frost joined 04:56 frost left 08:00 Manifest0 joined 08:08 lizmat_ left, lizmat joined 08:51 dakkar joined 11:44 MasterDuke left
lizmat and yet another Rakudo Weekly News hits the Net: rakudoweekly.blog/2022/01/17/2022-...kucon-how/ 13:01
Nemokosch so I have sets of pairs 13:04
what exactly happens when I substitute one from the other?
the first result is okay, the new includes the old one 13:05
but then the other difference should contain 1760 elements, right? 13:06
13:22 MasterDuke joined 13:33 [Coke] left 13:42 [Coke] joined 14:37 qorg11 left 14:40 qorg11 joined 14:56 [Coke]_ joined 14:57 [Coke] left
Nahita Not necessarily 15:55
my @old = 12.comb;
my @new = 124444666666.comb;
say @old (-) @new;
say @new (-) @old;
m: `​`` 15:56
my @old = 12.comb;
my @new = 124444666666.comb;
say @old (-) @new;
say @new (-) @old;
Nemokosch I mean... okay, this wasn't the case
I can tell you it wasn't all duplicates but one pair
Nahita Then we need an MRE!
m:​``` 16:08
my @old = 12.comb;
my @new = 124444666666.comb;
say @old (-) @new;
say @new (-) @old;
(for completeness nothing else :))
17:14 [Coke] joined 17:15 [Coke]_ left
Nemokosch m: say (2-1*sqrt 5)**(1/3) 17:20
why is that?
m: say (2-1*sqrt 5) 17:21
m: say -0.2360679774997898**(1/3)
why don't I get this number but NaN instead?
I mean, okay, I get that the unary minus actually has lower precedence 17:35
m: say (-0.2360679774997898)**(1/3)
but this doesn't seem okay at all
17:39 dakkar left
Nahita Seems to have adopted the first point here <math.stackexchange.com/a/317546>. Like C, Java, Perl and JS but not Python. 18:49
Nemokosch it seems to me the best built-in way is using sign and abs 19:02
stevied I got some busted links on: raku.land/zef:sdondley/Karabiner::...dGenerator 19:22
I filed a report on raku.land repo for the broken USAGE links. not sure if that's the right place
this is also showing up on the page: ` [Configuration File](#Configuration File) `
in my original pod, I have this: `L<Configuration File|#Configuration File>` 19:23
am I doing anything wrong there?
20:58 aitron joined 21:49 aitron left, aitron joined 22:05 [Coke]_ joined 22:07 [Coke] left
tailgate is there a way to get a default value from a hash if a key isn't there, or is this the best way? 22:26
if %capitals{'Europe'}:exists { return %capitals{'Europe'} ;} else {return "";}
stevied `%d{'x}:exists || ''; 22:27
tailgate thanks! 22:28
stevied `%d{'x}.exists || '';
no need for "return" if it's the last line in the function
`%d{'x'}.exists || '';
oh, wait, sorry 22:29
`%d{'x'}.exists ?? %d{'x'} !! '';` 22:30
hold on, I'm new at this, too 22:32
lakmatiol You could do `%d{'x'}//''`
stevied `%d{'x'}:exists ?? %d{'x'} !! '';`
but doesn't that check if the value is True? 22:34
or defined, I mean
Nemokosch defined 22:43
isn't that exactly the wanted behavior?
also, if you want, you can define a default value for hashes, I believe 22:45
m: my %capitals is default(''); dd %capitals{'Europe'};
tailgate thanks! 22:49
stevied can I key exist with an undefined value, though? 22:51
can a key exist with an undefined value, though?
Nemokosch what does that mean? 22:53
stevied it seems `%d{'x'}//''` will test if the value is defined. he wants to test if the key exists
Nemokosch I don't think there is a conceptual difference
I mean... why would you assign not definite values to something? 22:57
23:28 [Coke]_ left 23:31 jaguart joined, [Coke] joined
jaguart Hello, old Perlie, Raku absolute beginner, trying to get @INC path and class/file naming. Have class A::B::C in ./A/B/lib/C.rakumod and that isa A::D::C which is in ./A/D/lib/C.rakumod but can't seem to get raku to tie it together. 23:38
gfldex @Nemokosch#9980 I wrote a script that tried to find methods that where missing in the docs. To do so, I had to query methods on the build in types. Those happen to be type object, and as such quite undefined. As soon as you do meta-programming you will find ample reason to store undefined values in Hashes and otherwise. 23:49
Nemokosch That doesn't seem like intended usage to me, although it works 23:51
gfldex The MOP is very intentional.
Nemokosch what is MOP? 23:52
[Coke] meta object protocol
jaguart: You need to specify -I. to specifically look starting in the local dir. 23:53
gfldex meta object protocol
Nemokosch btw I can vaguely recall that defined values of a type do have everything that the default value i.e the type object could provide
[Coke] you can also use RAKULIB
jaguart I already have the /lib/ folders in my use-lib pragma
[Coke] (first is command line arg, second is an env var.)
jaguart: ok, but you never mentioned lib in your example.
if it's still not working, can you show me your use lib, your use of the module, and the path to the module? 23:54
you did mention lib but it's buried near the bottom? 23:55
if you have ./A/B/C.rakumod, you "use A::B::C" and you specify a lib dir of .
you don't use lib near the end of the path.
jaguart ok - let me fix that and retry 23:56
[Coke] When I read your example at first, I literally tuned out after A/B, sorry . :)
gfldex @Nemokosch#9980 Note quite. Here is a practical example: gfldex.wordpress.com/2021/08/17/most-fancy/ 23:57
jaguart removing /lib and raku doesnt find the first A::B::C 23:58
[Coke] please share your use statement, path to the module, fully qualified name of the module in the file, and your -I/use lib/RAKULIB 23:59
Nemokosch well, I don't see the contradiction
[Coke] (and the error message)
gfldex To be fair, undefinedness is a fairly large field in Raku. Where other languages need a paragraph in a book, we need a whole section.