»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'p6: say 3;' or /msg camelia p6: ... | irclog: irc.perl6.org or colabti.org/irclogger/irclogger_log/perl6 | UTF-8 is our friend! 🦋
Set by Zoffix on 25 July 2018.
00:03 gabiruh left 00:09 reach_satori_ left, reach_satori joined 00:23 reach_satori_ joined 00:24 reach_satori left 00:29 Xliff left 00:34 reach_satori_ left 00:48 wamba joined 01:04 kamog left 01:32 Sgeo__ joined 01:34 lucasb left 01:35 Sgeo_ left, whosgonna left 01:45 Xliff joined
Xliff m: say hyper (1..100000).grep( *.is-prime ) 01:45
camelia (2 3 5 7 11 13 17 19 23 29 31 37 41 43 47 53 59 61 67 71 73 79 83 89 97 101 103 107 109 113 127 131 137 139 149 151 157 163 167 173 179 181 191 193 197 199 211 223 227 229 233 239 241 251 257 263 269 271 277 281 283 293 307 311 313 317 331 337 347 349…
Xliff That's pretty fast. Why is this slower?
m: say (hyper (1..100000).grep( *.is-prime ))[* - 1] 01:46
Actually... ^^ that times out.
I don't know why.
guifa The first only calculates the ones that are needed
Xliff So why can't I get the last one calculated in a similar amount of time? 01:47
guifa the default is to only output I think the first 50 elements when you say @foo
so it only calculates the first 50
Xliff Oh! It's CHEATING!
guifa Whereas the second one, it has to go through all the 1..10000 first to figure out the last one 01:48
It’s not cheating, it’s lazy ;-)
Lazy is code for efficient :D
Xliff Right. So how can I parallelize getting the nth prime number? 01:49
01:49 camelia left 01:51 camelia joined
camelia (timeout) 01:53
Elronnd Xliff: it's not a *super* parallelizable task, although cores can help. There are algorithms on wikipedia
01:53 ChanServ sets mode: +v camelia
guifa (1..10000).grep(*.is-prime)>>.say; 11.6s on TIO 01:56
(1..10000).hyper.grep(*.is-prime)>>.say; 4.5s on TIO 01:57
Elronnd guifa: well, you wouldn't do it like that 02:00
you would use a prime number sieve or something like that
guifa Elronnd: oh I know. I was just saying how you could do a naive approach and still save some time. I assumed Xliff wasn’t as worried about the .is-prime and just didn’t think the hyper was actually doing anything 02:03
Geth_ doc: 62c718117a | holli-holzer++ | 33 files
update
 02:06
Elronnd guifa: oh, ok 02:07
02:08 StPetersburgh is now known as Doc_Holliwood
Doc_Holliwood Err, guys. I think I might have broke something. So I have local copy of the docs from a fork on github 02:10
02:10 Cabanossi left
Doc_Holliwood in order to update them i made a new clone, copied everything over and pushed it. Somehow I made a push to the master branch thoigh 02:10
Geth_ doc: 45ef5e8a71 | holli-holzer++ | 33 files
Revert accidental push to the wrong repository

This reverts commit 62c718117a20d8c93ca10b5773807c39500ebaa9.		 02:14
doc: fba598b317 | holli-holzer++ (committed using GitHub Web editor) | doc/Language/containers.pod6
Update containers.pod6

Fixing the code example about custom containers. Without the "-rw" the code dies.		 02:17
synopsebot Link: doc.perl6.org/language/containers
02:21 Cabanossi joined 02:41 Xliff left 02:45 hythm joined 02:47 reach_satori joined
hythm is it required to include `META6.json` file in modules or its recommended only? I'm asking because I came across 'Encode' module which doesn't have meta file. 02:47
actually it has `META6.info` but not `META6.json` 02:51
03:04 molaf joined 03:10 reach_satori_ joined 03:12 reach_satori left, p9s joined 03:20 AlexDani` joined 03:22 molaf left 03:25 AlexDaniel left 03:35 p9s left 03:36 p9s joined 03:44 p9s left, p9s joined 03:46 p9s left, p9s joined 03:52 p9s left, wamba left, p9s joined 03:56 p9s left, p9s joined 03:58 p9s left 03:59 p9s joined, p9s left 04:00 p9s joined 04:06 hythm left 04:28 eythian left 04:29 eythian joined 04:39 nightfrog left 04:45 rindolf left 04:55 nightfrog joined 05:04 wamba joined 05:06 AlexDani` is now known as AlexDaniel, AlexDaniel left, AlexDaniel joined 05:09 robertle left
AlexDaniel tobs: discord is not the only thing out there 05:11
for example, there's a matrix bridge too
I'm also pretty sure there's nothing illegal in that and it's not against their tos 05:14
depends on how it's done though
while they were here they complained about discord, matrix and google. They're free to leave if the tools that we use are so misaligned with their views 05:24
05:41 jmerelo joined
jmerelo releasable6: status 05:41
releasable6 jmerelo, Next release in ≈26 days and ≈13 hours. R6 is down. At least 5 blockers. Unknown changelog format
jmerelo, Details: gist.github.com/dd4377af58f56f8414...2ee20d4c7f
05:41 wamba left 05:52 wamba joined 05:57 filehandle32 joined 06:05 wamba left 06:22 ravenousmoose joined 06:32 jmerelo left 06:33 wamba joined
tyil tobs: I am unable to alter the topic, so I can't do anything 06:34
06:37 domidumont joined 06:50 ufobat joined, wamba left 06:53 p9s left 07:01 wamba joined 07:08 Cabanossi left 07:15 epony left, Cabanossi joined 07:34 abraxxa joined 07:47 zakharyas joined 07:57 p9s joined 08:02 dakkar joined 08:05 sena_kun joined, scimon joined 08:08 nebuchadnezzar left 08:12 nebuchadnezzar joined 08:16 robertle joined 08:40 wamba left, pmurias joined 08:42 fingolfin40 joined 08:44 p9s left
fingolfin40 Hi. What is the shortest way to import code from file? As "include" in C? Have a nice day! 08:44
08:44 wamba joined
sena_kun fingolfin40, `use foo` if `foo` is in any path that is checked for libs 08:45
tadzik there is nothing quite like include in C though, afaik
sena_kun `need foo` if you need a runtime loading, which you hardly need
tadzik unless you want to literally eval another file :)
sena_kun tadzik, well, C include doesn't make any sense in Perl 6, as Perl 6 has concept of packages, while C is "lets dump everything into a loooong file and compile that" 08:46
tadzik sena_kun: indeed
which you could do with EVAL slurp 'file.pl' if you *really* wanted to... ;) 08:47
but that still won't work for most stuff I guess
ie variables won't magically appear in your scope I think
sena_kun tadzik, I don't think import == EVAL in the question, fingolfin40?
tadzik perhaps I treated the question a bit too literally :) 08:48
sena_kun :)
tadzik like sena_kun says, `use` is probably the way to go
fingolfin40 So, if I want just splice long program to several files, do I need use packages? 08:49
Or use?
sena_kun for example, if you have a.pm6 and b.pm6 in a single directory, you can do `use b` in a.pm6 and `perl6 -I. a.pm6`. But be careful not to run it with `-I.` from the home directory, as it'll scan everything... which will be not very nice.
fingolfin40, just create a module
give me a second... 08:50
docs.perl6.org/language/modules <- a pretty long, but covering basically everything you need
fingolfin40 Thanks 08:51
sena_kun you can use tools, e.g. `zef install Ddt` or `zef install mi6`, and then look up their readme for commands, they can create a "default" structure of a module for you to fill in the code later.
but if you are starting, it is better to not rely on quick magic, and there are not so many things to know: you have your code in `lib` directory, e.g. `Foo::Bar` is `lib/Foo/Bar.pm6`, you have your tests in `t` directory, just `00-sanity.t`, and you fill META6.json file in the root. 08:52
then you can `zef install` it and in some script: `use My::Cool::Module; call-sub-from-module(42)`. or you can create `bin` directory in your module and put scripts there, then they will be "installed" and runnable without typing `perl6 ...`. 08:54
fingolfin40, "Distributing modules" section of the documentation is what you want.
fingolfin40 I was think that using modules is too hard for my simple work, but it seems to be the only right way. 08:57
tadzik modules are super easy :) 08:58
you don't really need to make them installable and all that though
just shove them in lib/<that path>, then run your code with perl6 -Ilib
no need to install anything until you actually need to
you can also do `use lib 'lib'` to your source code before you do the `use` 08:59
then you don't need the -Ilib anymore
fingolfin40 thanks
09:05 wamba left 09:08 robertle left 09:09 robertle joined 09:10 wamba joined 09:17 [particle] joined 09:19 [particle]1 left 09:26 pat_js joined 09:38 pmurias left 09:47 Black_Ribbon left 09:48 pilne left 09:52 pat_js left 09:54 isBEKaml joined 10:09 pmurias joined 10:21 robertle left 10:25 moritz_ is now known as moritz, robertle joined 10:27 filehandle32 left 10:30 pat_js joined 10:38 wamba left 10:54 pat_js left 10:57 molaf joined
tbrowder hi, #perl6 10:58
question on redefining built-in routines: 10:59
say i want to redefine 'copy' 11:00
i write: sub copy($a,$b) { copy $a, $b } 11:01
moritz now you've invented recursion :D
tbrowder which 'copy
moritz I guess you want to wrap CORE::copy? 11:02
m: sub say($x) { say $x }
camelia ( no output )
moritz is surprised there wasn't a timeout
sena_kun it wasn't called to recurse? 11:03
m: sub say($x) { say $x }; say 42;
sena_kun is not a bad person who abuses bots 11:04
tadzik *sad violin*
11:04 Geth joined
camelia (timeout)MoarVM panic: Memory allocation failed; could not allocate 327872 bytes 11:04
tbrowder m: sub say($x) { CORE::say $x }; say "a" 11:05
camelia 5===SORRY!5=== Error while compiling <tmp>
Two terms in a row
at <tmp>:1
------> 3sub say($x) { CORE::say7⏏5 $x }; say "a"
expecting any of:
infix
infix stopper
statement end
statement modi…
sena_kun tbrowder, you want `multi sub copy($a where 42, Str $loc) { say "Copied 42 to $loc" }; copy 42, 'foo';
oops
m: multi sub copy($a where 42, Str $loc) { say "Copied 42 to $loc" }; copy 42, 'foo'; 11:06
camelia Copied 42 to foo
tadzik m: sub say($x) { CORE::say: $x }; say "a"
camelia 5===SORRY!5=== Error while compiling <tmp>
Confused
at <tmp>:1
------> 3sub say($x) { CORE::say:7⏏5 $x }; say "a"
expecting any of:
colon pair
tadzik hrm
tbrowder add semi at end?
tadzik nah, it didn't like the colon calling I think 11:07
m: sub say($x) { CORE::say($x) }; say "a"
camelia Cannot find method 'Nil' on object of type Str
in sub say at <tmp> line 1
in block <unit> at <tmp> line 1
tadzik I guess it is just for methods
sena_kun 1)you need to add multi to not shadow the existing candidate; 2)why do you need to redefine it? to handle different candidates?
or to handle the same candidates in a different way?
in the first case, `multi` is your friend, in the second case, why not write a simple wrapper? 11:08
isBEKaml m: sub say($x) { CORE::say $x }; say("a");
tbrowder i'm trying to build a File::Copy module
camelia 5===SORRY!5=== Error while compiling <tmp>
Two terms in a row
at <tmp>:1
------> 3sub say($x) { CORE::say7⏏5 $x }; say("a");
expecting any of:
infix
infix stopper
statement end
statement mo…
isBEKaml Ah, TTIAR
sena_kun instead of messing with builtins which is probably slower by definition;
m: say CORE::say; say &CORE::say;
camelia Nil
&say
isBEKaml CORE:: is in nqp namespace?
sena_kun just for the record, ^ is a "correct" way to do that. :) 11:09
you need to sigil things from namespaces
isBEKaml m: sub say($x) { CORE::say: $x }; say("a");
camelia 5===SORRY!5=== Error while compiling <tmp>
Confused
at <tmp>:1
------> 3sub say($x) { CORE::say:7⏏5 $x }; say("a");
expecting any of:
colon pair
tbrowder m: sub say($x) { &CORE::say $x}; say("6"); 11:12
camelia 5===SORRY!5=== Error while compiling <tmp>
Two terms in a row
at <tmp>:1
------> 3sub say($x) { &CORE::say7⏏5 $x}; say("6");
expecting any of:
infix
infix stopper
statement end
statement mo…
sena_kun m: sub say($x) { &CORE::say($x) }; say 42;
camelia 42
sena_kun m: sub say($x) { my \term := &CORE::say; term 42 }; say 42; 11:13
camelia 5===SORRY!5=== Error while compiling <tmp>
Two terms in a row
at <tmp>:1
------> 3b say($x) { my \term := &CORE::say; term7⏏5 42 }; say 42;
expecting any of:
infix
infix stopper
statement end
…
tbrowder \o/ yay
sena_kun apparently you can't do that without writing an operator
tbrowder sena_kun: ++
sena_kun tbrowder, but anyway, what's the issue you are trying to solve with overloading copy? I am somehow sure just having a wrapper `my-smart-copy` will be both cheaper and easier to understand for the reader. 11:14
maybe I am wrong, though.
11:15 zakharyas left
tbrowder well, the real issue is making the buitlin copy more usable (it's a rakudo issue i think): be able to copy a file to a dir, for example 11:16
sena_kun a multi candidate sounds nice for that 11:17
sub (IO::Path $file where *.f, IO::Path $dir where *.d) {...}
multi sub copy (IO::Path $file where *.f, IO::Path $dir where *.d) {...}
tbrowder jnthn and others want to make an enhanced copy as a module to get lots of use before considering enhancing the built-in copy 11:18
but the multi couldn't be used outside rakudo, correct? 11:20
sena_kun why?
m: multi sub say($_ where 42) { say "SUPER 42!" }; say 42; say 10; 11:21
camelia SUPER 42!
10
sena_kun e.g. you can overload operators just ok and then are part of CORE, and operators are just subs
tbrowder okay!
sena_kun: thank you, i'll try that 11:22
sena_kun tbrowder, have fun! 11:23
tbrowder i though you needed to use either one sub or all multis, not a mixture 11:24
see ya later, gator 11:25
sena_kun luckily, `say` is a multi 11:26
just as `copy`
tbrowder ah, many thanks! 11:32
11:33 robertle left 11:34 jdv79_ is now known as jdv79 11:35 robertle joined
pmurias timotimo: perl6.github.io/6pad/#2c710afec97b...31fe38d803 - I made your spaceship drawing code run almost unmodified in the browser 11:59
timotimo heck yeah!
12:04 lgtaube left 12:09 wamba joined 12:11 wamba left 12:23 rindolf joined 12:25 pmurias left 12:42 robertle left 12:44 robertle joined 12:53 Cabanossi left 12:58 lucasb joined, rindolf left 13:01 molaf left 13:06 wamba joined
Doc_Holliwood `:lang<JavaScript>`? What kind of dark wizardry is that again? 13:13
masak .oO( this is #perl6 -- we only dabble in light wizardry, not dark ) 13:15
13:18 zakharyas joined
Altreus what the good heavens is this 13:20
this is definitely deep magic 13:21
13:21 Sgeo__ left
tadzik hahah that's greart! 13:21
13:34 thowe left 13:40 rindolf joined 13:44 lgtaube joined 13:57 pat_js joined 14:07 pmurias joined
pmurias when wrapping a JS object in Perl 6 land I override some methods like .item so it can be used without jumping through hoops 14:09
what would be a good way when you want to actually use an js item method?
I'm thinking of using a named argument like maybe :native/:raw 14:10
so $foo.new would do a 'new $foo' in js land but $foo.new(:raw) would do '$foo.new' 14:11
any thoughts about that?
timotimo could try with a .^ method
pmurias you mean with a .^call_js_method('new', |@arguments)? 14:13
timotimo perhaps
pmurias that would work, but not sure if that's isn't too ugly 14:16
timotimo i'm not sure under what circumstances you need to do one vs the other
pmurias if you have a js object that has a new or item method 14:17
and you want to call that
14:17 isBEKaml left 14:18 shlomif joined
vrurg pmurias: I have basically done merging makefile templates for moar and jvm. But JS doesn't make sense until it bootstraps itself. 14:19
Altreus I don't know what you're doing, making JS objects available to P6, but I hate-love it 14:20
pmurias Altreus: I'm compiling Perl 6 to JS 14:21
vrurg: the nqp build system rarely changes right? so keeping the js stuff as is shouldn't cost us much? 14:22
vrurg pmurias: no. That's why I decided that cure would be worse than the illness in this case. :) 14:23
Altreus successfully? This is definitely magic
vrurg Just keeping you posted.
pmurias Altreus: not everything works (and some things aren't really doable like multithreading stuff) but a large part of roast passes. The performance for now sucks 14:25
14:28 epony joined
pmurias timotimo: I'm not sure what a good named argument name would be 14:38
timotimo at the very least it should be allcaps 14:40
14:44 wamba left 14:54 Cabanossi joined, john_parr left
pmurias timotimo: :RAW? 15:06
timotimo maybe? 15:07
:DIRECT? 15:08
masak :BLUNT 15:12
15:14 pmurias left 15:17 ufobat_ joined
SmokeMachine vrurg: don't you think a Event class would make it easier to `grep` it? 15:18
15:20 ufobat left 15:22 robertle left
lizmat and another Perl 6 Weekly hits the Net: p6weekly.wordpress.com/2019/09/23/...else-itch/ 15:28
15:35 jmerelo joined
Juerd \o/ 15:36
masak lizmat++
jmerelo lizmat++ (always)
15:38 p9s joined
masak aha, so TheDamian did the for-else in the way I first _thought_ the for-else worked in Python :P 15:39
15:39 domidumont left
masak fondly remembers discussions with pmichaud and TimToady about this 15:42
Altreus still think "otherwise" would have been nicer 15:45
still could be :) 15:47
lizmat masak: could you give a gist of those discussions ?
15:50 p9s left
masak lizmat: it must've been around 2009. I had recently learned about Python's for-else (and while-else) and thought it was nifty. crucially, I had misunderstood its semantics. 15:52
(in Python, the `else` triggers not on 0 iterations, but on the absence of a `break` (p6's `last`) statement 15:53
)
Altreus what is this colon syntax? has Hidden:_ $!value;
masak: wow... that sounds like the opposite of what else means 15:54
masak Altreus: the `:_` means "can be defined or undefined" or "can be a type object or an instance"
Altreus aha ... what is default?
masak Altreus: it actually makes sense, how Python does it
Altreus: you just need the explanation to see it :)
Altreus Well
Grinnz naming is hard part 9000 15:55
masak `break` tends to signify "we found it, so no need to look further"
Altreus If I don't see the explanation I'll work on my understanding of English, which says that if I abort something (with break) then it "failed"
ehhhh
masak and so the `else` means "in case we didn't find anything"
Altreus I don't like this at all :P
masak maybe it grows on one :)
Altreus I can tell we didn't find anything because there's nothing in the relevant variable :P 15:56
Grinnz this seems more like perl5-"continue but only on the last iteration"
Altreus also that's very unlike a language that won't even let me test an assignment
15:56 zakharyas left
Altreus or did they change that? Haven't done much python 15:56
Grinnz isn't that why guido stepped down and all that? 15:57
Altreus Anyway I would never have associated "break" with "we found it"
masak Altreus: yes, they changed that in 3.8
Altreus Same way I don't associate "return" with "I succeeded"
masak Grinnz: yes, that was a direct cause
lizmat perhaps a NOTLAST phaser would then be more appropriate 15:58
to be fired if the for loop did *not* exit with a "last"
still... easily enough coded :-) 15:59
Altreus Anyway, apparently for-else has different meanings in different places. And then with the other arguments for not using "else", I still propose "otherwise" instead :)
Grinnz i think it's even more fundamental than that - it's wanting to set a flag, and run an extra block if that flag didn't get set
jnthn lizmat: Those things get awkward scoping wise...
Grinnz it doesn't have to be tied to break/last
Altreus lizmat: but that could also mean "Not the last iteration", with also NOTFIRST
for ... {} finally {}
masak lizmat: I remember the discussions had some consequences on the spec, so might be able to find logs for it. 16:00
16:00 wamba joined
lizmat yeah... although I kinda like "otherwise" 16:00
Altreus finally would run only after the final iteration of the original list
lizmat for { } otherwise { }
Altreus Therefore if it was last'd, it wouldn't run
Grinnz i think otherwise is a great idea for that (apparently) different use case
Altreus otherwise would run if the original list was empty
masak lizmat: at least for a while, one could use the _value_ of a `for` loop to determine if it iterated at all
Altreus for @list { stuff } finally { I didn't find it } otherwise { empty list happened } 16:01
Grinnz but in try/finally, the finally runs regardless of exception
Altreus yes, that one is a difficult word to find 16:02
a difficult case to find a word for
Doc_Holliwood how about `bare`? 16:11
16:11 robertle joined
Doc_Holliwood or just `empty` 16:12
Altreus werl
you could probably use a phaser for the "last iteration" jobby, so at the end of the last iteration you could do something else
for @list { do stuff; LASTLY { oh I didn't find it } } otherwise { there was nothing in the list } 16:13
FIRSTLY { this is the first iteration }
but does LASTLY imply it should run when you do 'last' ?
Doc_Holliwood it pretty much implies at least some work has been done, not none 16:14
Altreus it was meant to imply *all* the work has been done 16:15
jnthn m: do for 1 { } or say "nope"
camelia ( no output )
jnthn m: do for () { } or say "nope"
camelia nope
Altreus it was argued that the do/or construct is a bit overly separated 16:16
jnthn m: do for 1 { if False { } } or say "nope" # alas...
camelia nope
jnthn What it actually tells you is not if you iterated anything :)
Altreus ah yes and it has to actually reutrn something
something true 16:17
jnthn No, that's not the case
It must produce something that's not Empty :)
Altreus aha, so many new concepts
jnthn m: do for 1 { False } or say "nope"
camelia ( no output )
Altreus I guess the other thing I have to ask is, are people actually going to use 'for' like the python people do? 16:18
Surely we have better ways of hunting for things in lists than for/last
Which means we don't even *have* the assumption that "The list ran out" is likely to mean "I didn't find it"
and anyway we could use the word "exhausted" somehow 16:19
masak having spent some time with Pyhon's for-else (regardless of the actual semantics of it), I can say that I don't wish strongly for a for-else construct in the languages I use 16:28
I'm happy with doing the same with a boolean variable, or some slight repetition 16:29
I mean, TheDamian writes compelling blog posts, but... :)
in the end, I haven't found for-else or any other similar construct to be an improvement on clarity and at-sight understanding 16:30
masak .oO( it's not a part of the Structured Programming dream )
Altreus I have :O 16:32
and I think TT2 has one, or something like that
16:32 scimon left
Altreus It's super useful for that sort of thing anyway 16:32
16:36 dakkar left 16:39 pat_js left
Doc_Holliwood Isn't monkey typing supposed to work with methods AND attributes? 16:59
In my toy code here it works for methods, but I cannot access the attribute 17:00
pastebin.com/XxiAjhyr
discord6 <Aearnus> masak: for-else is useful for website templating
<Aearnus> not how python does it though
17:02 jmerelo left, veesh joined 17:03 dogbert11 joined
jnthn Doc_Holliwood: You cannot add attributes to a class after composition time 17:04
Doc_Holliwood yaaarrrgh. 17:06
Any other ideas of how to do this? 17:07
subclassing `Proxy` didn't work either stackoverflow.com/questions/580553...rectly-tie 17:08
17:08 Cabanossi left
Doc_Holliwood And the docs say "With augment, you can add attributes and methods to existing classes" 17:09
[Coke] jnthn: you can trigger another compose, though, yes? 17:10
17:10 gabiruh joined 17:11 Itaipu joined 17:15 Cabanossi joined
jnthn [Coke]: Yes, perhaps more clear would be: only the first compose will compose attributes. 17:17
Doc_Holliwood: Then that'd be a doc bug; we've never supported adding attributes with augment 17:18
Probably Proxy should become possible to subclass 17:19
Doc_Holliwood jnthn: allright, I will remove that from the docs then, yes? 17:21
jnthn Doc_Holliwood: Yes, please 17:25
home time & 17:28
Geth_ doc: cf60fc2860 | holli-holzer++ (committed using GitHub Web editor) | doc/Language/variables.pod6
fixed: Attributes actually cannot be augment-ed.

  jnthn | Doc_Holliwood: Then that'd be a doc bug; we've never supported adding attributes with augment		 17:37
synopsebot Link: doc.perl6.org/language/variables
17:37 cpan-p6 left 17:38 ChoHag left, cpan-p6 joined, cpan-p6 left, cpan-p6 joined, ChoHag joined 17:44 mowcat joined 17:47 nekomune left 17:52 aborazmeh joined, aborazmeh left, aborazmeh joined 17:53 kybr left 17:55 nekomune joined 17:59 kybr joined 18:03 rindolf left, shlomif left 18:10 domidumont joined, rindolf joined, domidumont left, shlomif joined 18:11 domidumont joined
guifa wonders if he could do evil. And rewrite all of the keywords in a slang in Spanish, etc. 18:21
18:21 domidumont left 18:29 El_Che left, El_Che joined 18:31 ChoHag left
Kaiepi releasable6, status 18:31
releasable6 Kaiepi, Next release in ≈26 days and ≈0 hours. R6 is down. At least 5 blockers. Unknown changelog format
Kaiepi, Details: gist.github.com/57ec2d3230d4ac5b5a...1640357e73
18:32 ChoHag joined 18:33 wildtrees joined 18:34 Ven`` joined
Doc_Holliwood What will we do when we reach 6.z? 18:41
[Coke] I'm sure we'll decide on that before 6.O 18:43
18:44 Maylay_ left 18:45 Maylay joined 18:48 khisanth_ left 18:53 aborazmeh left
vrurg Doc_Holliwood: Greek alphabet? Anyway, it's a problem to think about in a little bit over 40yrs from now. 18:56
19:02 khisanth_ joined 19:24 Xliff joined
Xliff . 19:25
\o
19:25 wamba left 19:26 Itaipu left 19:28 aborazmeh joined, aborazmeh left, aborazmeh joined 19:29 Itaipu joined
Kaiepi i'm curious to see how people did the second task of last week's perl 6 challenge, subclassing Proxy doesn't work atm, mixing in roles to Scalar etc. didn't work when i tried it, and i couldn't think of a good way to override &infix:<=> 19:39
oh wait that's this week's one 19:42
19:45 pmurias joined
Doc_Holliwood Kaiepi, my thoughts exactly 19:55
I also tried monkey-patching Proxy 19:56
Kaiepi the last way's probably the best bet, but i'm not sure how you'd keep track of which variable's which at compile time since you'd need a slang to do it 19:57
Doc_Holliwood Well you can add methods but no attributes. You would have to keep a hash with adresses as keys or something, but you can have that with a bare closure 19:58
19:58 thundergnat joined
Doc_Holliwood It's a conspiracy to make Raku look bad 19:58
The question in the weekly 19:59
thundergnat Kaiepi: Take a look at rosettacode.org/wiki/History_variables#Perl_6 . The slightly unintuitive part is that Rakudo doesn't implement or allow a STORE method on a Scalar variable (for performance reasons) so it needs to be a sigiless var. 20:01
20:01 molaf joined
thundergnat That took me a while to figure out. 20:02
20:02 thundergnat left 20:05 wamba joined
Kaiepi does handles make it so calling one of the method names you give it on the instance of the class itself calls it on that attribute instead? 20:08
timotimo yes 20:09
Kaiepi neat 20:10
Doc_Holliwood but not type safe
20:13 Black_Ribbon joined
Kaiepi oh, it's not something i'd use then. everything i write is pretty strictly typed 20:14
20:15 fingolfin40 left
timotimo i'm not sure what that means, though? 20:16
20:19 aborazmeh left 20:35 pmurias left 20:38 pmurias joined
Xliff Doc_Holliwood: What do you mean by "not type safe"? 20:40
class A { method a (Int $b) { $b }; }; class C { has A $.d handles<b>; }; 20:41
That replaces:
class A { method a (Int $b) { $b }; }; class C { has A $.d; method a (Int $e) { A.a }; }; 20:42
20:42 thowe joined
Doc_Holliwood for instance you cannot pass it into a function that expects an Int. Of course, it's a HistoryVar. I would like my (or the) solution more transparent 20:44
timotimo "it"?
Doc_Holliwood The instance of HistoryVar
timotimo well, of course you can't 20:45
that would be not type safe
it *is* type safe when it rejects a HistoryVar when it expects an Int
Doc_Holliwood fair. i used a wrong term. 20:46
timotimo you can have many "handles" in your class 20:47
21:05 Ven`` left, Ven`` joined 21:06 Ven`` left, Black_Ribbon left, sno left 21:07 Black_Ribbon joined, sno joined 21:10 pmurias left 21:11 pmurias joined 21:22 sena_kun left 21:27 MasterDuke left
Xliff Would this work? --> class A { multi method a (Int $b) { $b }; }; class C { has A $.d handles <a>; multi method a (Str $e) { $e }; }; 21:31
m: class A { multi method a (Int $b) { "A: $b" }; }; class C { has A $.d handles <a>; multi method a (Str $e) { $e }; }; say C.new.a(1); C.new.a('Bleah') 21:32
camelia 5===SORRY!5=== Error while compiling <tmp>
Package 'C' already has a method 'a' (did you mean to declare a multi-method?)
at <tmp>:1
Xliff Well... that answers that. 21:33
Kaiepi m: my Int:D @foo = (1,2,3); say @foo 21:34
camelia [1 2 3]
Kaiepi m: sub foo(Int:D @foo) { say @foo }; foo (1,2,3)
camelia Type check failed in binding to parameter '@foo'; expected Positional[Int] but got List ($(1, 2, 3))
in sub foo at <tmp> line 1
in block <unit> at <tmp> line 1
thowe stuff like this saddens me: github.com/nunorc/p6-Pekyll 21:35
Kaiepi why don't types for lists and hashes work the same in assignments vs function parameters?
Xliff thowe: Why?
thowe "Under heavy development, things will change.. you have been warned!" Latest commit 4 years ago. 21:36
Xliff Oh. Yeah. That. 21:37
:(
21:37 Kaiepi left 21:38 Kaiepi joined
Xliff Looks like he switched to python. 21:38
I think that would have been around the time of the Parrot vs Moar thingy.
Kaiepi which repo is this?
Xliff Kaiepi: Linked above.
Kaiepi oh, didn't see it 21:39
i wonder if there's a way you could make it so cro can handle it so you don't have to implement as much yourself 21:40
Xliff I would think Cro would be big help, there. 21:41
Kaiepi i kinda wish more advanced stuff with cro like using pipelines was documented better though, its unit tests don't always show how you would be able to use stuff like that in your own code 21:43
there is the option to get training though 21:44
tbrowder hey #perl6 /o/ 21:45
thowe hey tbrowder 21:46
tbrowder some good writer needs to put a good article on cofeproject.com. last ar
thowe: HOWdy 21:47
21:47 go|dfish left 21:48 grumble left
Doc_Holliwood that name does not resolve. did you mean covfefeproject.com? 21:48
tbrowder that's codeproject.com, could excite some people (i think they're mostly windows types.
21:49 TreyHarris left
tbrowder the last perl articles i found on a search was looong ago 21:49
21:50 grumble joined
Geth_ ¦ problem-solving: Xliff assigned to jnthn Issue Delegation does not handle multi-methods. github.com/perl6/problem-solving/issues/107 21:57
22:01 go|dfish joined
jnthn Kaiepi: Because function parameters are bound, not assigned 22:02
Kaiepi ah, makes sense
22:04 pmurias left 22:05 pmurias joined
Kaiepi i'm giving haskell a shot for the second time and it's so much easier to understand now after having used perl 6 22:06
i like it a lot so far 22:07
22:08 wamba left, wamba1 joined 22:10 thowe left
Xliff Kaiepi: That's good! As long as we don't lose you to The Dark Side, it's all good! :) 22:24
Kaiepi i'll always use perl 6 for scripting 22:25
Xliff :>
Elronnd Kaiepi: I have a haskell book a friend gave me, keep meaning to learn it 22:36
wen tthrough the first chapter or so
22:40 TreyHarris joined, Sgeo joined 22:45 |oLa| left 22:54 pmurias left 23:05 wamba1 left
wildtrees Kaiepi, #haskell on freenode is pretty nice and there is #haskell-beginners as well 23:37
Elronnd wildtrees: we're *on* freenode 23:39
wildtrees just being very specific
Kaiepi yeah, i already joined #haskell 23:40
didn't know about #haskell-beginners though, will join that as well
23:45 mowcat left 23:51 Xliff left

Powered by Cro and the Raku® Programming Language.

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

Thank you!