»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.
TimToady hypers aren't that smart yet 00:00
n00b_ And I can't roll my own hypers yet, right? I know it's on the roadmpa
TimToady well, it's all written in p6/nqp already, so it's theoretically possible :) 00:01
n00b_ I suppose I could try to roll my own rank operator in nqp, but I wasn't planning on getting my hands that dirty. 00:01
TimToady but the current metaops are a bit hardwired, yes 00:02
we have the grammatical categories, but the action routines are not going to get auto-written for you, as they are with normal operators
n00b_ One meta-op which I think is worth consideration is rank: op"N expresses that the "natural cell" of op is an array of rank N, and so " breaks up an arrays > N into equal-sized shards of rank N, feeds them to op seriatim, and collects the results into a larger array (just concat) 00:03
I'm not sure if this needs to be so array-focused as in the APLs; N in Perl6 could just as easily be a Signature 00:04
timotimo hm, a bit like rotor?
TimToady we can express array shapes in declarations, but nothing is currently gonna do the pattern matching for you 00:05
n00b_ secretly, / is just a shorthand for "(natural_cell_of_op, Inf)
I don't know rotor, let me check that out.
timotimo heads out
TimToady rotor is rather more procedural than declarational though 00:06
TimToady and only handles the top dimension like other listops 00:06
n00b_ The top is also a useful choice, though I think bottom is more general (because it's easier write top in terms of bottom than vice versa) 00:07
TimToady if we decide the current hypers aren't so useful, we can always tweak the semantics of hyper within a lexical scope too
n00b_ Wouldn't that require the ability to roll one's open hypers? 00:08
Without getting into nqp and actions, that is.
TimToady well, you'd have the actions call into other support routines
n00b_ Intuitively, that feels like it might not be sufficient to express the structural walking inversion from topmost to bottommost 00:09
TimToady is quite willing for someone smarter than him to do that, as long as it doesn't force mere mortals to understand before they need it :) 00:11
n00b_ m: say (1,2,3) X, (4,5,6) 00:14
camelia ((1 4) (1 5) (1 6) (2 4) (2 5) (2 6) (3 4) (3 5) (3 6))
n00b_ m: say ZX, (1,2,3) xx 2
camelia 5===SORRY!5=== Error while compiling <tmp>
Undeclared name:
ZX used at line 1
n00b_ How can I create 2 copies of (1,2,3) and zip or insert X, between them?
It's a ridiculously expensive but syntactically cheap way to get me closer to the idea I want to express 00:15
TimToady m: say [ZX,] (1,2,3) xx 2
camelia (((1 1)) ((2 2)) ((3 3)))
TimToady m: say [ZX,] ((1,2,3) xx 2) xx 2
camelia (((1 1) (1 2) (1 3) (2 1) (2 2) (2 3) (3 1) (3 2) (3 3)) ((1 1) (1 2) (1 3) (2 1) (2 2) (2 3) (3 1) (3 2) (3 3)))
n00b_ Oh. I may not even need the Z then.
TimToady m: say [X,] (1,2,3) xx 2
camelia ((1 1) (1 2) (1 3) (2 1) (2 2) (2 3) (3 1) (3 2) (3 3))
n00b_ m: say [X,] (1,2,3) xx 2
camelia ((1 1) (1 2) (1 3) (2 1) (2 2) (2 3) (3 1) (3 2) (3 3))
TimToady "insert" means you want to reduce
n00b_ Thanks, that's a useful distinction 00:16
TimToady the main problem you'll run into there is that reducing on a single list will explode the list instead of treating it scalarly 00:17
TimToady due to the single-arg rule 00:17
so for some applications you might need a way to disable that rule
n00b_ I'm having the opposite problem. I want c_d to operator on every pair of lists, but my attempts with [] turn out to feed c_d every pair of scalars. 00:19
TimToady it's a bit like regex matching /.* foo/ 00:21
TimToady finding the next-to-last of something is always a bit fraught :) 00:22
but I can see how it might be powerful 00:23
n00b_ Is there some form of boxing I can use to turn (4,5,6) into [4,5,6] in the P5 sense?
TimToady you mean, make it an item? 00:24
n00b_ yeah
TimToady m: say $_ * 2 for 1,2,3, $(1,2,3,4) 00:25
camelia 2
4
6
8
TimToady like that?
n00b_ Yes, but I want to item-ify each list in a list of lists
m: say $$_ for ( (1,2,3) , (4,5,6) )
camelia (1 2 3)
(4 5 6)
n00b_ m: say ($$_).perl for ( (1,2,3) , (4,5,6) ) 00:26
camelia $(1, 2, 3)
$(4, 5, 6)
n00b_ nah, that didn't help
what I'm thinking is if I could "box" an array so it looked like a scalar, I might be able to access the second-to-last level with hypers 00:27
by tricking them into thinking it's the last level
TimToady mebbe
m: say ([$_,]).perl for ( (1,2,3) , (4,5,6) ) 00:28
camelia [(1, 2, 3),]
[(4, 5, 6),]
TimToady or some such
anyway, a real [] has scalars inside 00:29
m: say ($[$_]).perl for ( (1,2,3) , (4,5,6) )
camelia $[1, 2, 3]
$[4, 5, 6]
TimToady hmm
brane hertz...
n00b_ my brain doesn't have enough hertz
TimToady 'course one can go all nominal typing with objects, too 00:30
n00b_ I'm abusing the language enough as it is, I think adding objects in to force it to confess is a bridge too far 00:31
I'm not above cheating, but I'm above *icky* cheating. 00:32
TimToady part of the problem here is that we're trying to use scalars as much more than just 0-dimensional arrays :) 00:33
so we don't have this pure view of dimensionality that you get in aplish languages 00:34
n00b_ you have to; having strings be rank 1 vectors in APL is a PITA, a mistake from the 70s we carry around 00:35
TimToady Perl is much more interested in the singular/plural distinction of natural languages
so at least we didn't make that mistake :)
n00b_ yes! though I'd like an easy facility to determine what's singular for me, without going all OOP 00:36
overtyping tempts me even further to overengineering 00:37
TimToady m: say $(1,2,3).WHAT
camelia (List)
TimToady m: say $(1,2,3).VAR.WHAT
camelia (Scalar)
n00b_ BTW, Ken Iverson shared your view that taking cues from natural languages is critical, but he was too Strunk and White about it 00:38
TimToady omit needless symbols!
n00b_ Sure, I get the $() thing, but it's syntactic, which is limiting 00:39
TimToady the main problem I have with APLish languages is that all the symbols look like each other in terms of weight
n00b_ One bit of syntax APL discarded as limiting was [] for indexing. Took a long time for everyone else to cop to that.
TimToady which is why our metaops look "bigger" than normal ops 00:39
n00b_ There's a different mindset in APL. We view the egalitarian nature of symbols as a virtue 00:40
TimToady and we don't mind using larger symbols for larger or less-frequent ideas
huffman coding, and all that
n00b_ And that ops not being bigger than functions erases a bias against doing things top-down
APL also huffman-codes (no one uses {:: for example), but we have a different view on what (should be) "less frequent" 00:41
TimToady from a natural languages viewpoint, it's a foolish consistency
but yeah, the math viewpoint tends to fight linguistics
n00b_ yep 00:42
TimToady natural languages don't actually go in for much orthogonality
and the most irregular forms are the commonest ones 00:43
because that's where you need the extra redundancy
n00b_ John Wilkins, a 17th C philosopher, was an early proponent of trying to reduce argumentation to shuffling symbols, the first in a wave of pitting math against language
the whole concept is misguided, but very attractive 00:44
TimToady sure, he's one of the chief villains of The Search for the Perfect Language, by Eco
one of my favorite books
n00b_ I'll add it to my list. I've dipped into some of his work. 00:45
n00b_ For now, what's a straightforward way to generate all pairs of lists from a given list of lists? 00:48
lookatme :) 00:52
TimToady m: say (<a b c>, <d e f>, <x y z>).combinations(2) 00:57
camelia (((a b c) (d e f)) ((a b c) (x y z)) ((d e f) (x y z)))
n00b_ niiice 00:59
[Coke] yawns 01:47
samcv timotimo: we could add the strings length if we end up with a hash that is 0 01:48
so then we will always have a non-zero hash
well actually the length plus 1 01:49
though i guess we have to account for it going over 32bit signed int. though wouldn't be an issue when i make the hashes 64bit
Geth doc: 0927fc50aa | (JJ Merelo)++ | doc/Language/traps.pod6
Grammar correction
 05:59
synopsebot Link: doc.perl6.org/language/traps
doc: fb337f39c3 | (JJ Merelo)++ | doc/Language/traps.pod6
Document Nil → Any trap closes #1135
Geth doc: 92d045bee9 | (JJ Merelo)++ | doc/Language/traps.pod6
Document Nil → Any trap closes #1134

Uses also #1135 example to document it.		 06:00
Geth doc: 5e1a7a5ed8 | (JJ Merelo)++ | doc/Type/Seq.pod6
sink method documented in Seqs. Closes #2028
 06:12
synopsebot Link: doc.perl6.org/type/Seq
lookatme How can I declare a sub/function type has specify signature ? 06:16
jmerelo lookatme: what do you mean?
lookatme like C does : typedef int (*intfun_t)(void); intfun_t s; s is a function
jmerelo lookatme: pretty much in the same way docs.perl6.org/syntax/Constraining...0Callables 06:17
lookatme Hmm, I am trying to use subset but not working 06:18
m:　subset WalkCb of Callable where *.signature ~~ :(Int $a); my WalkCb $x = sub (Int $a) { return $a; };
camelia Type check failed in assignment to $x; expected WalkCb but got Sub (sub (Int $a) { #`(Sub...)
in block <unit> at <tmp> line 1
jmerelo lookatme: would you mind to post it too in StackOverflow? It's the kind of thing that belongs there too. 06:19
m:　subset WalkCb of Callable where *.signature ~~ :(Int $a); my WalkCb $x = sub (Int $b) { return $b; }; 06:20
camelia Type check failed in assignment to $x; expected WalkCb but got Sub (sub (Int $b) { #`(Sub...)
in block <unit> at <tmp> line 1
jmerelo m:　subset WalkCb of Callable where *.signature ~~ :(Int $a); my WalkCb $x = -> Int $b { return $b; };
camelia Type check failed in assignment to $x; expected WalkCb but got Block (-> Int $b { #`(Block|...)
in block <unit> at <tmp> line 1
jmerelo m:　subset WalkCb of Block where *.signature ~~ :(Int $a); my WalkCb $x = -> Int $b { return $b; }; 06:21
camelia Type check failed in assignment to $x; expected WalkCb but got Block (-> Int $b { #`(Block|...)
in block <unit> at <tmp> line 1
jmerelo m:　subset WalkCb of Block where *.signature ~~ :(Int $a); my WalkCb $x = WalkCb.new(-> Int $b { return $b; }); 06:22
camelia Default constructor for 'WalkCb' only takes named arguments
in block <unit> at <tmp> line 1
El_Che I don't think you can declare signatures as a type
lookatme Do you have any other suggestions ? 06:23
El_Che, :)
jmerelo m:　subset WalkCb of Block where *.signature ~~ :(Int $a); my WalkCb $x = (-> Int $b { return $b; }).WalkCb;
camelia No such method 'WalkCb' for invocant of type 'Block'
in block <unit> at <tmp> line 1
jmerelo m:　subset WalkCb of Block where *.signature ~~ :(Int $a); my WalkCb $x = -> Int $b { return $b; };
camelia Type check failed in assignment to $x; expected WalkCb but got Block (-> Int $b { #`(Block|...)
in block <unit> at <tmp> line 1
jmerelo m:　subset WalkCb of Block where *.signature ~~ :(Int $a); my WalkCb $x = -> Int { return $_; }; 06:24
camelia Type check failed in assignment to $x; expected WalkCb but got Block (-> Int $ { #`(Block|5...)
in block <unit> at <tmp> line 1
jmerelo lookatme: definitely StackOverflow 06:24
El_Che there is one already
but I can't find it 06:25
jmerelo El_Che: if you mean the one on using signatures as variables, I don't think that one fits here...
lookatme on the stackoverflow ?
jmerelo El_Che: you mean this one? stackoverflow.com/questions/415313...-in-perl-6 06:26
El_Che yes 06:27
jmerelo m: my $x = -> Int { return $_; }; say $x.signature
camelia (Int)
jmerelo m: my $x = -> Int { return $_; }; say $x.signature ~~ Int 06:28
camelia False
jmerelo m: my $x = -> Int { return $_; }; say $x.signature ~~ "Int"
camelia False
jmerelo m: my $x = -> Int { return $_; }; say $x.signature ~~ :(Int)
camelia True
El_Che stackoverflow.com/questions/440936...1#44097741
dunno is it's applicable
jmerelo m:　subset WalkCb of Block where *.signature ~~ :(Int); my WalkCb $x = -> Int { return $_; };
camelia Type check failed in assignment to $x; expected WalkCb but got Block (-> Int $ { #`(Block|8...)
in block <unit> at <tmp> line 1
jmerelo El_Che: it kinda is, but it implies it should work... 06:30
m: my $x = -> Int { return $_; }; say $x.^name
camelia Block
lookatme Seems like there's no way to do like C does :)
jmerelo lookatme: there is definitely one way. But whoever is around here now does not know the answer... That's why it's better if you post it on StackOverflow, where the most knowledgeable Perl6 people dwell 06:31
lookatme I am not have time to post it on the StackOverflow 06:33
jmerelo lookatme: OK, I'll do it for you then
lookatme Thanks, and hope you understand me :) 06:34
El_Che I don't see in the answer where foo is called 06:36
ahah , I see it 06:37
never mind :)
if it's a solution, it's ugly
if the signatures are the same, make a common sub/method and pass the parameters by a wrapper sub/method? 06:38
the error would be less obvious for the user, but's it's only one indirection
jmerelo m:　subset WalkCb of Block where *.signature ~~ :(Int $); my WalkCb $x = -> Int { return $_; }; 06:39
camelia Type check failed in assignment to $x; expected WalkCb but got Block (-> Int $ { #`(Block|7...)
in block <unit> at <tmp> line 1
jmerelo m:　subset WalkCb of Block where *.signature ~~ :($); my WalkCb $x = -> Int { return $_; };
camelia Type check failed in assignment to $x; expected WalkCb but got Block (-> Int $ { #`(Block|7...)
in block <unit> at <tmp> line 1
jmerelo m:　subset WalkCb of Block where *.signature ~~ :(Int); my $x = -> Int { return $_; }; say $x.signature; my WalkCb $y = $x; 06:40
camelia (Int)
Type check failed in assignment to $y; expected WalkCb but got Block (-> Int $ { #`(Block|5...)
in block <unit> at <tmp> line 1
jmerelo OK, I think I know the answer 06:41
jmerelo m:　subset WalkCb of Block where *.signature ~~ :(Int, -> Int); my $x = -> Int { return $_; }; say $x.signature; my WalkCb $y = $x; 06:41
camelia 5===SORRY!5=== Error while compiling <tmp>
Malformed parameter
at <tmp>:1
------> 3b of Block where *.signature ~~ :(Int, -7⏏5> Int); my $x = -> Int { return $_; }; s
jmerelo m:　subset WalkCb of Block where *.signature ~~ :(Int, --> Int); my $x = -> Int { return $_; }; say $x.signature; my WalkCb $y = $x;
camelia (Int)
Type check failed in assignment to $y; expected WalkCb but got Block (-> Int $ { #`(Block|7...)
in block <unit> at <tmp> line 1
jmerelo m:　subset WalkCb of Block where *.signature ~~ :(Int $a --> Int); my $x = -> Int $a { return $a; }; say $x.signature; 06:42
camelia (Int $a)
jmerelo No I don't. 06:43
m:　subset WalkCb of Block where *.signature ~~ :(Int $a --> Int); my WalkCb $x = -> Int $a { return $a; };
camelia Type check failed in assignment to $x; expected WalkCb but got Block (-> Int $a { #`(Block|...)
in block <unit> at <tmp> line 1
jmerelo lookatme: Check it out in StackOverflow stackoverflow.com/questions/503848...in-general 06:44
lookatme jmerelo, I am not mean **using subset**, it's fine if we have any other way to do it :) 06:59
lookatme It's like a function pointer of C, I just thought it's looks easy to do it with subset 07:01
jmerelo lookatme: that's why it's always better if you do it yourself :-) I have edited it, though 07:02
lookatme jmerelo, haha, Never mind 07:03
Geth doc: e295315c72 | (JJ Merelo)++ | doc/Type/List.pod6
Aggregates sink, closes #1607
 07:09
synopsebot Link: doc.perl6.org/type/List
Geth doc/master: 5 commits pushed by (JJ Merelo)++ 07:49
jmerelo squashable6: status 08:07
squashable6 jmerelo, Next SQUASHathon in 15 days and ≈1 hour (2018-06-02 UTC-12⌁UTC+14). See github.com/rakudo/rakudo/wiki/Mont...Squash-Day
jmerelo .tell lookatme The solution is here stackoverflow.com/questions/503848...1#50386481 (by LizMat) 08:35
yoleaux jmerelo: I'll pass your message to lookatme.
lookatme . 08:37
tyil weekly: www.reddit.com/r/ProgrammingLangua...out_perl6/
yoleaux 08:35Z <jmerelo> lookatme: The solution is here stackoverflow.com/questions/503848...1#50386481 (by LizMat)
notable6 tyil, Noted!
lookatme jmerelo, hmm, I see thanks
lizmat tyil++ 08:39
tyil lizmat: putting it straight into notable6 is easier than poking you every time I see or comment about Perl 6 and then have you do it :p 08:40
lookatme But I wonder is this form should work fine ? `subset WalkCb of Callable where *.signature ~~ :(Int $a); `
lizmat tyil: yeah, I would suggest other people do this as well, BTW :-)
lookatme or this is just a bug
lizmat lookatme: github.com/rakudo/rakudo/issues/1836 # I just opened an issue for it
tyil lizmat: I'll take make a note of it when I see it happen to tell other people to throw it into notable6 :> 08:41
lookatme oh, I see lizmat
lizmat afk for a few hours& 08:42
El_Che lizmat is fast, when I wanted to post something on SO she already did :) 08:45
jmerelo weekly: twitter.com/raiph_mellor/status/99...8504270850 on Perl 6 and type systems. 08:58
notable6 jmerelo, Noted!
pmurias to be honest the "gradual" typing of Perl 6 sucks compared to real gradual typing as is more of language weakness than a strength 09:15
hahainternet i can't say i've ever fully understood HM 09:16
but strong typing does seem to be the right way to go whenever possible 09:17
lookatme :) I have another question now 09:22
m: use NativeCall; sub create-pointer(::T $a) is export { my CArray[T] $c = CArray[T].new; $c[0] := T.new(0); nativecast(Pointer[T], $c); }
camelia 5===SORRY!5=== Error while compiling <tmp>
An exception occurred while parameterizing CArray
at <tmp>:1
Exception details:
Type check failed in binding to parameter '<anon>'; expected Any but got T (?)
in method parameterize at /ho…
lookatme Is there a better way create a nativecall pointer to something ? 09:23
seems like CArray can not work with parameter T
lookatme I am off work now, have no time talk about it, just post question here 09:24
FROGGS jnthn: should Perl6 be listed here too? reactivex.io/languages.html 09:29
jmerelo lookatme: maybe just use CPointer? 09:41
FROGGS: most probably...
jnthn FROGGS: Probably. We liked it so much we built it in to the language's standard library. :-) 09:53
FROGGS jnthn: that's what I mean... are you goiing to inform them, or shall I? 09:55
jnthn FROGGS: If you're up for it, please do :) 09:58
jnthn has a few too many things to do at the moment
FROGGS k 10:02
lizmat pmurias: would you care to elaborate on the sucking of gradual typing in Perl 6 ? 10:20
jmerelo lizmat: right. It's not Haskell, but it's good enough and very useful for lots of things. 10:22
buggable New CPAN upload: Desktop-Notify-0.3.0.tar.gz by FRITH modules.perl6.org/dist/Desktop::Not...cpan:FRITH
lizmat m: my constant A = :(Int $); # probably related to #1836 10:43
camelia ===SORRY!===
QAST::Block with cuid 1 has not appeared
lizmat R #1836
GH #1836 10:44
hmmm
github.com/rakudo/rakudo/issues/1836
robertle .seen DrForr 12:24
yoleaux I saw DrForr 13 Apr 2018 14:25Z in #perl6: <DrForr> My head isn't in that space right now otherwise I'd look for possible examples...
El_Che it's been a while 12:29
robertle yeah, darn
I need something like TT for perl6, and saw that he has started on that. but don't know what the state is and where things are going... 12:30
lizmat fwiw, he appears to be active on FB
robertle FB?
lizmat FaceBook
timotimo is tt template toolkit?
robertle ah, that!
yeah tempalte toolkit
timotimo in such a case i'd suggest looking at the test suite
robertle I did! 12:31
so basically it looks as if the perl 6 version is some way there, but not entirely. probably about 50% of the features I need, less on the arcane ones
timotimo ah, that sucks i guess 12:32
time for Inline::Perl5 ;)
robertle I guess to contribute would be to port more of the tests from the perl5 version and then fix them :)
but I am just wondering what his thoughts on the thing are. perhaps there is a reason for not going down that route?
lizmat suggests lack of tuits 12:38
El_Che lizmat: you could be a house with tuits 12:41
lizmat hehe... 12:42
but first some time afk&
ufobat timotimo, something new regarding my jit error? 14:47
MasterDuke ufobat: you have a one-liner to repro it, right? 14:59
ufobat yap
perl6 -e 'use LibCurl::HTTP; my $url= "1.eu.dl.wireshark.org/src/wireshar...1.tar.xz"; my $http = LibCurl::HTTP.new( :ssl-verifypeer(False), :ssl-verifyhost(False) ); $http.GET($url).setopt(:!failonerror).perform.say' 15:00
it just happens with "larger" http get requests 15:01
MasterDuke ufobat: which libcurl do you have? apparently i have to choose between openssl, nss, and gnutls 15:04
ufobat installed on my ubuntu is libcurl3, libcurl3-gnutls and libcurl4-oepnssl-dev 15:11
MasterDuke ufobat: i don't have any fixes, but i created github.com/MoarVM/MoarVM/issues/862 15:20
ufobat okay 15:22
mahafyi hi. I am looking to update rakudo-star-2017.10 to 2018.04? Is there an uninstallation step, or a howto guide for this? I am using debian 9. 15:37
mahafyi is it sufficient to simple install the latest version in a folder and change the $PATH, then delete the older version src folder? 15:39
MasterDuke mahafyi: how did you install 2017.10? from a repo? 15:41
mahafyi MasterDuke : no, i downloaded it and installed 15:42
perl6intro.com/#_installing_perl_6
from source 15:43
then, i added "/opt/rakudo-star-2017.10/bin:/opt/rakudo-star-2017.10/share/perl6/site/bin" in $PATH 15:44
MasterDuke yeah, what you said earlier should work 15:46
huggable: debs
huggable MasterDuke, CentOS, Debian, Fedora and Ubuntu Rakudo packages: github.com/nxadm/rakudo-pkg/releases
MasterDuke there're also those ^^^
and i think 9 actually has a rakudo package (not sure if star) in their repos 15:47
mahafyi thanks 15:53
raschipi MasterDuke: There's no star in Debian 15:57
No zef either
MasterDuke ah, thanks 15:58
buggable New CPAN upload: Test-Declare-0.0.1.tar.gz by DARRENF cpan.metacpan.org/authors/id/D/DA/...0.1.tar.gz 16:02
El_Che mahafyi: ping me if you have question about the rakudo-pkg repos and packaged 16:03
if you use debian, centos, fedora, opensuse or ubuntu the easiest way is to use the repos
you'll get updates automatically
El_Che rakudo-pkg has rakudo and zef installed and a few scripts to install zef as a user, set the path and use it in de windows sbsystem for linux 16:04
El_Che mahafyi: github.com/nxadm/rakudo-pkg#os-repositories 16:06
in case you want to use rakudo star, get it from the main rakudo page and compile the source 16:07
mahafyi El-Che : thank you 16:10
jkramer Is there a shorter way of flattening an array of arrays than @x.flatmap(*.flat)? 16:20
timotimo m: my @aoa = <hi bye>, <foo bar>, <a b c>; say @aoa.map(*.Slip).List; 16:21
camelia (hi bye foo bar a b c)
timotimo m: my @aoa = <hi bye>, <foo bar>, <a b c>; say @aoa.map(*.Slip).List.perl;
camelia ("hi", "bye", "foo", "bar", "a", "b", "c")
timotimo m: my @aoa = <hi bye>, <foo bar>, <a b c>; say @aoa.map(*.Slip).perl;
camelia ("hi", "bye", "foo", "bar", "a", "b", "c").Seq
timotimo m: my @aoa = <hi bye>, <foo bar>, <a b c>; say @aoa>>.Slip.perl;
camelia (slip("hi", "bye"), slip("foo", "bar"), slip("a", "b", "c"))
timotimo OK, the map does the magic
m: my @aoa = <hi bye>, <foo bar>, <a b c>; say @aoa>>.Slip.List.perl;
camelia (slip("hi", "bye"), slip("foo", "bar"), slip("a", "b", "c"))
timotimo yeah, i think the map(*.Slip) is the shortest
m: my @aoa = <hi bye>, <foo bar>, <a b c>; say @aoa.map(&slip).perl; 16:22
camelia ("hi", "bye", "foo", "bar", "a", "b", "c").Seq
timotimo that's one character shorter
jkramer :D
Nice, thanks 16:23
m: my @aoa = <hi bye>, <foo bar>, <a b c>; say @aoa.map(|*).perl; 16:24
camelia ("hi", "bye", "foo", "bar", "a", "b", "c").Seq
jkramer Saved 3 more chars :)
El_Che the last one is nice
Geth ecosystem: darrenf++ created pull request #396:
Add github.com/darrenf/p6-test-declare		 16:25
timotimo ah, of course 16:26
Zoffix m: my @aoa := <hi bye>, <foo bar>, <a b c>; dd @aoa.flat 16:30
camelia ("hi", "bye", "foo", "bar", "a", "b", "c").Seq
Zoffix Don't make them arrays in the first place :)
m: my @aoa = <hi bye>, <foo bar>, <a b c>; dd |«@aoa 16:31
camelia [("hi", "bye"), ("foo", "bar"), ("a", "b", "c")]
Zoffix Would've though this'd work
Zoffix m: &prefix:<|> does role { method nodal { }}; my @aoa = <a b c>, <d e f>, <g h i>; dd |«@aoa 16:33
camelia (slip("a", "b", "c"), slip("d", "e", "f"), slip("g", "h", "i"))
Zoffix m: &prefix:<|> does role { method nodal { }}; my @aoa = <a b c>, <d e f>, <g h i>; dd flat |«@aoa
camelia ("a", "b", "c", "d", "e", "f", "g", "h", "i").Seq
El_Che throw a .list somewhere :) 16:34
Geth ecosystem: 9a9109876c | (Darren Foreman)++ (committed by Zoffix Znet) | META.list
Add github.com/darrenf/p6-test-declare (#396)

  `Test::Declare` is a rudimentary way to express test scenarios in a fairly declarative sense. Assuming I understand the word 'declarative' properly.
Zoffix m: my @aoa = <hi bye>, <foo bar>, <a b c>; dd @aoa».list.flat 16:35
camelia ("hi", "bye", "foo", "bar", "a", "b", "c").Seq
Zoffix m: my @aoa = <hi bye>, <foo bar>, <a b c>; dd @aoa»<>.flat 16:37
camelia ("hi", "bye", "foo", "bar", "a", "b", "c").Seq
Zoffix
.oO( rocket operator )
 16:38
TimToady for a more general but longer solution: 16:39
yoleaux 12:22Z <lizmat> TimToady: opinions about github.com/rakudo/rakudo/commit/8ae82a558b ?
TimToady m: my @aoa = <hi bye>, <foo bar>, <a b c>; say gather @aoa.deepmap(*.take)
camelia (hi bye foo bar a b c)
TimToady m: my @aoa = <hi bye>, <foo bar>, <a b c>; say gather @aoa».take 16:40
camelia (hi bye foo bar a b c)
TimToady I guess that works too
m: my @aoa = <hi bye>, <foo bar>, <a b c>; say (gather @aoa».take).perl
camelia ("hi", "bye", "foo", "bar", "a", "b", "c").Seq
Zoffix TimToady: the .take destroys Pair objects 16:42
um
Yeah, Pairs of Hashes
m: my @aoa = <hi bye>, <foo bar>, %(:42foo, :100bar); say (gather @aoa».take).perl
camelia ("hi", "bye", "foo", "bar", 100, 42).Seq
Zoffix m: my @aoa = <hi bye>, <foo bar>, %(:42foo, :100bar); dd @aoa»<>.flat 16:43
camelia ("hi", "bye", "foo", "bar", :foo(42), :bar(100)).Seq
Zoffix but the rocket operator does more or less the Right Thing™
(the .take stuff isn't a bug, but a WAT side of the DWIM for hypering stuff over hashes): 16:45
m: dd %(:42foo, :100bar)».flip
camelia Hash % = {:bar("001"), :foo("24")}
Zoffix m: sub postfix:<🚀> { $^v»<> }; my @aoa = <hi bye>, <foo bar>, %(:42foo, :100bar); dd @aoa🚀.flat 16:49
camelia ("hi", "bye", "foo", "bar", :bar(100), :foo(42)).Seq
Zoffix ^_^
El_Che it doesn't take long for Zoffix to end with emojis 16:50
he never disappoints :)
TimToady wonders what emotion is indicated by a rocket... 16:50
TimToady imminent burnout, perhaps... 16:52
Zoffix excitement 16:54
😺🌟😺🚀😻🖖🚀😸🚀 16:55
b2gills .tell tyil The problem is actually that ~~ doesn't participate in WhateverCode lambdas see: stackoverflow.com/a/50395077/1337 17:07
yoleaux b2gills: I'll pass your message to tyil.
b2gills m: subset WalkCb of Callable where .signature ~~ :(Int $); my WalkCb $x = sub (Int $a) { return $a }; say $x(42)
camelia 42
tyil b2gills: I am missing context 17:08
yoleaux 17:07Z <b2gills> tyil: The problem is actually that ~~ doesn't participate in WhateverCode lambdas see: stackoverflow.com/a/50395077/1337
tyil b2gills: I don't recall having issues with this 17:09
b2gills m: my &f = *.signature 17:12
camelia ( no output )
b2gills m: my &f = *.signature ~~ :(;; $); # not a lambda so it fails
camelia Type check failed in assignment to &f; expected Callable but got Bool (Bool::False)
in block <unit> at <tmp> line 1
b2gills m: say *.signature ~~ :(;; $ is raw); # not a lambda so it fails 17:14
camelia False
b2gills m: say [*.signature][0].signature ~~ :(;; $ is raw); 17:17
camelia True
Zoffix Look what I got in mail: i.imgur.com/xT2LSW4.jpg 17:51
Zoffix collects a 100 nerd points, reaching a new level
Juerd Zoffix: Nice! 17:53
Zoffix :) 17:55
TimToady I hope that's a non-breaking space... 17:58
Zoffix :D
MasterDuke cool 17:59
El_Che Zoffix: Oh, did we forget to tell you about the renamingM 18:09
M
? 18:10
Zoffix I can always order another set of plates :P
And it's not a rename, it's a creation of alias :)
El_Che Yeah, everyone know Carlos the Jackal by his real name Ilich Ramírez Sánchez 18:11
great geeky plates by the way 18:13
TimToady Our tiobe index might go up if we renamed it "Trump"... 18:19
El_Che tiobe index will go up if you rename it BlockChain6 18:21
timotimo TimToady: i imagine if we called it "TV" there'd be loads and loads of "tv programming" searches and results 18:22
El_Che lol 18:25
Zoffix rofl 18:26
jmerelo O/ 18:31
Kaiepi \o
jmerelo A new question to stackoverflow, just in cae you can help stackoverflow.com/questions/503987...tion-calls 18:58
Geth doc: 27b3eb7044 | (JJ Merelo)++ | doc/Language/contexts.pod6
Adds indexing for contexts
 19:09
synopsebot Link: doc.perl6.org/language/contexts
ingy can someone tell me how to do a string replace ($x ~~ s///) where rhs is a function? 19:49
ingy is glad laziness is a virtue here... 19:50
moritz m: $x = 42; $x ~~ s[\d+] = 2 * $/; say $x 19:51
camelia 5===SORRY!5=== Error while compiling <tmp>
Variable '$x' is not declared
at <tmp>:1
------> 3<BOL>7⏏5$x = 42; $x ~~ s[\d+] = 2 * $/; say $x
moritz m: my $x = 42; $x ~~ s[\d+] = 2 * $/; say $x
camelia 84
moritz ingy: ^^
Zoffix m: my $x = 42; say $x.subst: /\d+/, 2 * *
camelia 84
ingy docs.perl6.org/routine/subst ?
moritz using a method when you could have used syntax is a boring cheat :-) 19:52
Zoffix ingy: that's the method form of the op, yeah. The op is also available. Documented here: docs.perl6.org/syntax/s$SOLIDUS$SOLIDUS$SOLIDUS
ingy moritz: can that usage use a set of statements?
Zoffix ingy: you can use do {} 19:53
ingy cool
moritz: agreed
Zoffix m: my $x = 42; $x ~~ s/\d+/{say "Looks like you want me to work with $0"; $0 × 2}/; say $x 19:53
camelia Use of Nil in string context
Looks like you want me to work with
0
in block at <tmp> line 1
Use of Nil in numeric context
in block at <tmp> line 1
Zoffix m: my $x = 42; $x ~~ s/\d+/{say "Looks like you want me to work with $/"; $/ × 2}/; say $x 19:54
camelia Looks like you want me to work with 42
84
ingy :) 19:54
Zoffix m: my $x = "42 by 3"; $x ~~ s/(\d+).+(\d+)/{say "Looks like you want me work with $0 and $1"; $0 × $1}/; say $x
camelia Looks like you want me work with 42 and 3
126
ingy how do I match [\-\w] in p6? 20:04
ingy ie p5 /[\-\w]/ to p6 20:07
rjbs eval: ("fo-o bar" ~~ /(\w | "-")+/).gist.say 20:09
evalable6 ｢fo-o｣
0 => ｢f｣
0 => ｢o｣
0 => ｢-｣
0 => ｢o｣
rjbs probably there's a better way
ingy rjbs: thanks. I agree but in a hurry... 20:11
rjbs: do you happen to know how to make it non-capturing? 20:12
MasterDuke [] are non-capturing groups
rjbs eval: ("fo-o bar" ~~ /[\w|"-"]+/).gist.say 20:13
evalable6 ｢fo-o｣
ingy well that's a mindf**k!
rjbs MasterDuke: thanks :)
ingy mostly like regexes except for the rest of the time :) 20:14
MasterDuke m: say "fo-o bar" ~~ /<[-\w]>+/
camelia ｢fo-o｣
ingy MasterDuke: thanks. and thanks for forking my mind
MasterDuke np
ingy s:g/// messes up my vim hilighting 20:17
s:g/// #: fixes it
lizmat decommute& 20:23
rindolf ingy: hi, sup? 20:43
rindolf ingy: did you see github.com/ingydotnet/spork-pm/issues/1 ? 20:45
ingy rindolf: at my current rate, one of my digital grandchildren will get to that circa 2112 AD 20:48
rindolf ingy: ah :| 20:49
Zoffix Got my plates registered to may car and attached :) 21:07
I now officially own the coolest car in town: temp.perl6.party/perl6mobile-1.jpg
:D :D :D
(though I think I'll get better plate frames; these are too thick and get too close to the "♥" and the "6") 21:08
AlexDaniel this is awesome
MasterDuke i've heard good things about the civic si, do you like it? 21:09
Zoffix I love it
It's standard transmission and got a turbo. 21:10
MasterDuke oh, i thought they were NA. when did they start coming with a turbo? 21:11
Zoffix Looks like since 2017
MasterDuke i do like turbos though, my car is also standard and has a turbo (a WRX)
Zoffix (mine's 2018, but no changes from 2017 model)
I like the sucking sound from the turbo when you release the throttle :)
Or the spinning down sound. Whatever that sound is it's cool :) 21:12
bartolin indeed, very nice 21:12
lucasb nice ride :) 21:18
tbrowder_ Perl 6 on the town on fancy wheels! 21:22
timotimo how do we do gaussian distributions in perl6? 21:22
timotimo rosettacode has my back 21:24
thundergnat Hey, I've been testing Rakudo HEAD against modules to find breakage before the release and I keep getting test failures from HTTP::UserAgent 21:34
Has anyone else tried installing that on a recent commit and succeeded? 21:35
It _looks_ like it may be hash oder related, but I may just be overly sensitive.
timotimo did you look at the toaster results from the other day? 21:36
thundergnat gist of the test failure: gist.github.com/thundergnat/13f731...661cb6226f 21:37
I didn't show up in the toaster run, but that doesn't necessarily mean it was good, just that it didn't fail then.
*It 21:38
retupmoca do non-CPAN module updates get pulled into p6weekly automatically?
thundergnat: looks like github.com/sergot/http-useragent/issues/197
thundergnat retupmoca: Indeed. 21:39
Ok, aready noted.
gah. already 21:40
comborico1611 What does the 20 in 2018.20 mean on P6Weekly? 21:52
20th week?
thundergnat Hmm. Looks like the only problem with HTTP::UserAgent is some test routines need to change from is() to is-deeply(). 21:56
Zoffix comborico1611: yes 22:03
Zoffix clear plate frames much better: i.imgur.com/Y2AmBYv.jpg 22:03
timotimo i rewrote that gradient descent example where python, numpy, and nim are compared 22:04
it's not terribly fast
Zoffix timotimo: how much slower are we?
timotimo the fastest i could get it so far is 15s for 1000, whereas python does it in 35 seconds (comparing my machine against the blog author's machine, though, so that could be off, too)
i'm sure there's a bit more to be gained by cleverly putting nqp ops in there. my first attempt to make the inner loops nqp::while loops made it about 75% slower instead 22:05
gist.github.com/timo/2c0962ddb7e18...f59c9f1ab0
wanna give it a go?
Zoffix I'm on my phone 22:06
timotimo: 15s vs 35? that makes it faster than python
15m?
timotimo oh
sorry, that's for 1/10th the workload
Zoffix ah
timotimo so it'd be 150s vs 35s
Zoffix m: say 150/35 22:07
camelia 4.285714
Zoffix not horribly bad :)
timotimo the absolutely naive version was almost 2x slower than what i made it with nqp ops
so about 8.5x slower 22:08
could be worse
OK, on my machine it's 39.7 seconds 22:10
so the factor isn't as bad
timotimo also, the result seems to be wrong? 22:11
thundergnat seen sergot
.seen sergot
yoleaux I saw sergot 14 Mar 2018 21:46Z in #perl6: <sergot> ufobat_: ping
timotimo there's a repository with the code in it so you don't have to copypaste the bits together 22:13
timotimo github.com/henryiii/framework_comp...regression 22:13
OK, the code from that library (purepython.py) gives me not only the right results, but also just 30 seconds instead of 35. or maybe that's the difference between py2 and py3? 22:14
timotimo yup, that's what was wrong 22:15
maybe it was doing int division somewhere or so
comborico1611 zoffix: Thanks! 22:16
comborico1611 Does rakudo.org Debian prereq "libssl-dev" look like "libss1-dev", or is that just me? 22:22
Zoffix copy-paste it somewhere with different font / or eq it 22:30
timotimo Zoffix: should i look forward to your improvement, or are you leaving this one be? :)
El_Che comborico1611: what do you mean?
comborico1611 Yeah, I found out it was "L" after command not working, but wanted to say something for folks more newb than myself being confused by that.
El_Che ah the font? 22:31
comborico1611 El_Che: in "libssl" the second "L" looks bigger or different on my screen.
Gives the impression of a "1".
timotimo hm, ligatures?
comborico1611 Are you talking to me? 22:32
El_Che :)
comborico1611 is installing Rakudo 2018.04 for the first time, which is his 5th time to install using tar. So good newb feedback possible. 22:33
One thing I didn't understand about the make test for Rakudo... There were some failures on my system, but it didn't seem to matter. 22:36
...previously.
El_Che some tests are sadly enough floppers
they fails once in a while
comborico1611 So it just seems as if the testing is not important. 22:37
Does failing some test actually stop the installation?
El_Che the testing is to make sure everthing is ok
it does not change the installation
comborico1611 I see. 22:38
Thanks!
El_Che (my packages always have all tests passing, if one fail, I run the build again. if it keeps falling, I check with the devs)
comborico1611 Interesting!
That's good to know. I'm just a budding 6er, so it's not crucial for me. But now I wish I did the tests -- just to see what the results were. 22:40
El_Che most of the time they are ok
there are only a few that a 22:41
lay fail 1/10 or less
comborico1611 Maybe one day my name will be on one of these lines for the Rakudo installation. Rakudo 2020 or something. Haha. 22:42
El_Che there are lots of low hanging fruit 22:43
comborico1611 Actually, I forgot, its 2018... So 2030.
Interesting!
El_Che comborico1611: if you want to see failing tests, have a look at travis builds of rakudo-pkg travis-ci.org/nxadm/rakudo-pkg/builds 22:44
comborico1611 Installed. About 20 minutes. 22:45
Maybe 15.
Now I need to remember how I get the PATH thing going... Good thing I've done this recently. 22:46
El_Che export PATH=/the/place/where/you/installed/rakudo/bin:$PATH
comborico1611 Thanks! 22:47
I put it in my .profile file...
I don't know if that's bad, but its the only place I know where to put it.
El_Che the .profile is the right place for lots of shells 22:49
bash, sh, korn, etc
in case you forget: github.com/nxadm/rakudo-pkg/blob/m...l6-to-path (replace the rakudo-pkg paths with the ones from your installation) 22:53
Zoffix timotimo: leaving to be :) 22:54
comborico1611 Thanks! Yup I think I got it going. I just need to restart the system to see if it took.
El_Che comborico1611: source .profile 22:57
comborico1611 El_Che: Huh?
is that a bash commnd?
raschipi yes
it's so common there's a shortcut for it: "." 22:58
comborico1611 Ahh! Very good. (I need to get a book on Bash.)
Ahh again!
./
Yup! It is up and running. 22:59
lucasb type us a "perl6 -v" 23:00
comborico1611 lucasb: This is Rakudo Star version 2018.04.1 23:02
El_Che going to bd
bye!
comborico1611 Thanks for help!
Goodnight. 23:03
timotimo oh, rakudo presents itself as actually rakudo star if you've got a star installed?
lucasb timotimo: I was intrigued by that too
comborico1611 This is Rakudo Star version 2018.04.1 built on MoarVM version 2018.04.1
implementing Perl 6.c.
lucasb that is strange. does the star distribution patches rakudo sources? :) 23:06
lucasb funny thing is, there isn't a Rakudo Star version 2018.04.1 23:06
tarball name is rakudo-star-2018.04.tar.gz 23:07
comborico1611 Okay, so what's going on here? hastebin.com/ocemufupuv.pl 23:10
timotimo true, that's the rakudo version, which is a little weird
comborico1611: In Perl 6, please use "elsif' instead of "else if"
comborico1611 I did use "elseif". 23:11
timotimo there's a bug in there, too
it's supposed to say "instead of elseif"
timotimo i see where it went wrong 23:13
comborico1611 OHH. ELSif... Weird. Okay. 23:14
MasterDuke timotimo: for your grad descent code a profile shows lots of time in AT-POS, but that it isn't being inlined. i turned on the inline log and there are three identical entries `Can inline AT-POS (5755) with bytecode size 88 into (4)` 23:40
timotimo i rewrote the postcircumfix:<[ ]> into AT-POS and it didn't make things faster. in fact, it made it significantly slower instead
MasterDuke that...doesn't make sense? 23:42
timotimo it can
it didn't look like it prevented jitting, which was what i first suspected
hm, perhaps time will show that it's wasting lots of time in spesh?
i.e. OSR not working like i found in that other case
MasterDuke 114.87ms in dynamic optimization (unmodified from your gist) 23:44
just turned the AT-POS in to [] and now the log says `can inline AT-POS into postcircumfix:<[]>` and then `can inline postcircumfix:<[]> into ` 23:45
timotimo oh, wait, what i had there was already with at-pos? hm.
not sure what exactly i saw
comborico1611 enters VIM for the second time in history. First time was like 17 years ago. First command :e ~/perl6/caesar-cipher.p6. Eww ahh -- colorful! 23:47
timotimo :syn off :)
comborico1611 Have a good evening! 23:48
timotimo same to you
comborico1611 Oh, i will. Going to feed the chickens soon. It was so hot today. It will be nice playing with my dog in a bit. 23:49
timotimo it's nice of you to play with your dog before you feed him to the chickens 23:50
wait did i misunderstand?
comborico1611 Heh 23:53
lucasb m: multi f($) {}; f(Mu) # ok, cannot resolve 23:57
camelia Cannot resolve caller f(Mu); none of these signatures match:
($)
in block <unit> at <tmp> line 1
Geth doc: briandfoy++ created pull request #2029:
Fix a pod directive in .comb
lucasb m: proto f($) {}; multi f($) {}; f(Mu) 23:59
camelia Type check failed in binding to parameter '<anon>'; expected Any but got Mu (Mu)
in sub f at <tmp> line 1
in block <unit> at <tmp> line 1

Powered by Cro and the Raku® Programming Language.

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

Thank you!