»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.
MasterDuke but i would suggest adding ', it must be inside a sub or block', or something like that, to the message 00:00
00:00 cognominal left
BenGoldberg m: { $^a++ }.say; 00:01
camelia rakudo-moar 971e14: OUTPUT«-> $a { #`(Block|68813512) ... }␤»
BenGoldberg m: { $^a++ }().say;
camelia rakudo-moar 971e14: OUTPUT«Too few positionals passed; expected 1 argument but got 0␤ in block <unit> at <tmp> line 1␤␤»
00:01 cognominal joined
BenGoldberg m: { $^a++ }.(42).say; 00:01
camelia rakudo-moar 971e14: OUTPUT«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 rw…»
masak I use "mainline" to mean something roughly equivalent to "outside any routine"
BenGoldberg m: { $^a++ }.(Scalar.new(42)).say; 00:02
camelia rakudo-moar 971e14: OUTPUT«Cannot make a Scalar object using .new␤ in block <unit> at <tmp> line 1␤␤»
BenGoldberg m: { $^a++ }.([42][0]).say; 00:03
camelia rakudo-moar 971e14: OUTPUT«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 rw…»
babydrop heh
MasterDuke i don't think "mainline" is too hard to figure out, but it's the first time/place i've ever seen that word used in all my programming experience
00:03 cyphase left
PerlJam MasterDuke: what about "Cannot use placeholder parameter $^a outside of sub or block" ? 00:03
babydrop +1
MasterDuke +1 00:04
BenGoldberg m: { $^a++ }.(Proxy.new: FETCH => { 42 }, STORE => &say );
camelia rakudo-moar 971e14: OUTPUT«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 rw…»
BenGoldberg m: { $^a = $^a + 1 }.(Proxy.new: FETCH => { 42 }, STORE => &say ); 00:05
camelia rakudo-moar 971e14: OUTPUT«Cannot assign to a readonly variable ($a) or a value␤ in block <unit> at <tmp> line 1␤␤»
babydrop BenGoldberg: it'll be read-only regardless of what you stick into the arguments
dalek ateverable: 3ac4ca3 | (Aleks-Daniel Jakimenko-Aleksejev)++ | Whateverable.pm6:
Fix occasional crashes associtated with timeouts

Just sending a SIGHUP does not mean that we can close the channel, the process is probably still alive and it will attempt to send stuff into a closed channel.
This fixes #60. Previously this was the cause of MoarVM panics (until it was fixed in Rakudo), but generally this is our fault.
However, there is still at least one bug in rakudo, which is why we did not notice this problem right away. When the command prints a lot of stuff into stdin, it will never time out. The whole thing is stuck in a Proc::Async tap, without giving any chance for Promice.in(…) to fire up.
BenGoldberg Do I need an 'is rw'?
babydrop or is copy
BenGoldberg m: ( -> $a is rw { ++$a })(Proxy.new: FETCH => { 42 }, STORE => &say );
camelia rakudo-moar 971e14: OUTPUT«4243␤»
BenGoldberg m: ( -> $a is rw { $a = 'Mu' } )(Proxy.new: FETCH => { 42 }, STORE => &say ); 00:07
camelia rakudo-moar 971e14: OUTPUT«42Mu␤»
00:07 Perl6Robot joined
Perl6Robot OHAI! 00:07
00:07 Perl6Robot left
AlexDaniel :) 00:07
00:07 Perl6Robot joined
Perl6Robot OHAI! 00:07
00:07 Perl6Robot left 00:08 Perl6Robot joined, Perl6Robot left, cyphase joined
BenGoldberg just realized that our various 6bots all use ipv6. Cool! 00:08
00:10 sena_kun left
AlexDaniel BenGoldberg: why is that surprising? These bots are running on a proper server, and all servers nowadays have decent ipv6 support 00:11
the issue is much worse on the client side, that's true. Is there anybody running a bot from home? :)
BenGoldberg To be honest, I would never have noticed if not for seeing the raw address in the irc messages: "Perl6Robot (~z@2001:41d0:2:5eb5::) has Joined #perl6" etc. 00:12
I'm guessing that either freenode doesn't yet do reverse ip6 lookup, or something is misconfigured, or perhaps our server doesn't have a name. 00:14
geekosaur I think it does them, but either usually doesn't get an answer fast enough or has a bad upstream DNS; many but not not all ipv6 hosts seem to not get resolved 00:15
for example, it usually gets the linode I'm on right 00:16
AlexDaniel BenGoldberg: whateverables are running on my server though, not hack
konobi you need PTR records
geekosaur and last time I checked it was using ipv6, although maybe I broke that when I switched over to secure connections 00:17
konobi your IP transit provider should be able to delegate the in.arpa zones to a nameserver of your choice if you want them to be reverse resolvable
AlexDaniel yea, I have to do that 00:23
00:24 zakharyas left
AlexDaniel .tell jnthn hey! Turns out there was a bug in our code, so I guess you fixed everything. However, I found another issue: gist.github.com/AlexDaniel/44e5cde...6fe04b9c1d 00:26
yoleaux AlexDaniel: I'll pass your message to jnthn.
00:26 pierre_ joined
AlexDaniel .tell is it a known issue? Somehow I feel that it is. Basically, a lot of output on stdout of the process started with Proc::Async clogs up something in rakudo (well, the tap itself), and Promise.in(1) does not fire up at all because of that 00:28
yoleaux AlexDaniel: I'll pass your message to is.
AlexDaniel .tell jnthn is it a known issue? Somehow I feel that it is. Basically, a lot of output on stdout of the process started with Proc::Async clogs up something in rakudo (well, the tap itself), and Promise.in(1) does not fire up at all because of that
yoleaux AlexDaniel: I'll pass your message to jnthn.
AlexDaniel oops. :)
poor “is” will get this message on day. I wonder what his reaction would be 00:29
00:29 AlexDaniel is now known as is
is . 00:29
yoleaux 00:28Z <AlexDaniel> is: it a known issue? Somehow I feel that it is. Basically, a lot of output on stdout of the process started with Proc::Async clogs up something in rakudo (well, the tap itself), and Promise.in(1) does not fire up at all because of that
00:29 is is now known as AlexDaniel
AlexDaniel or maybe not 00:29
00:29 canopus left 00:32 pierre__ joined 00:33 pierre_ left
AlexDaniel eval: say :35<lizmat> 00:34
00:34 skrshn joined, evalable6 joined
AlexDaniel eval: say :35<lizmat> 00:34
skrshn how do you read commands from a file in the REPL?
AlexDaniel evalable6: say :35<lizmat> 00:35
00:35 canopus joined
AlexDaniel evalable6: come on, cooperate 00:35
MasterDuke skrshn: EVALFILE might do what you want 00:36
00:40 evalable6 left, evalable6 joined 00:41 lizmat joined
BenGoldberg thinks that yoleaux should ask for confirmation, if you .tell to a name which it hasn't ever seen before. 00:42
dalek ateverable: 6139b4f | (Aleks-Daniel Jakimenko-Aleksejev)++ | Whateverable.pm6:
OK, but now we also have to sink-try it
 00:47
00:47 evalable6 left, bisectable6 left, benchable6 left, committable6 left, unicodable6 left, benchable6 joined, bisectable6 joined, committable6 joined, unicodable6 joined, evalable6 joined, ChanServ sets mode: +vv bisectable6 committable6
AlexDaniel evalable6: say :35<lizmat> 00:47
evalable6 AlexDaniel, rakudo-moar 0c3d9e8: OUTPUT«(exit code 1) ===SORRY!=== Error while compiling /tmp/vqYCnTbRBb␤Invalid base-35 character: li<HERE>zmat␤at /tmp/vqYCnTbRBb:1␤------> say :35<lizmat><HERE><EOL>»
AlexDaniel … where? 00:49
m: say :35<lizmat>
camelia rakudo-moar 0c3d9e: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Invalid base-35 character: 3li7⏏5zmat␤at <tmp>:1␤------> 3say :35<lizmat>7⏏5<EOL>␤»
AlexDaniel ok, let's say this is OK 00:51
PerlJam z must be a really special character
AlexDaniel (though I was expecting something like “invalid base-35 character ‘z’”
)
PerlJam That would be more informative 00:52
AlexDaniel PerlJam: well, it's just not a base-35 character :)
well I was staring at “li⏏zmat” for a minute and my conclusion is that it is good enough
li→z←mat :) 00:53
00:53 seatek joined
PerlJam More informative would be a message that says what the valid characters are. 00:54
skrshn MasterDuke: thanks 00:55
AlexDaniel PerlJam: well, that's a good point 00:56
AlexDaniel submits rakudobug
01:04 pierre__ left 01:11 rindolf left 01:20 aborazmeh joined, aborazmeh left, aborazmeh joined 01:22 zamanf joined 01:24 bjz_ left 01:30 cognominal left 01:33 cognominal joined 01:38 cxreg left 01:39 mad__ left, diakopter left 01:40 pyrimidine joined 01:43 dataf3l joined 01:44 dataf3l left 01:45 pyrimidine left 01:46 diakopter joined, giraffe left
samcv .seen MadCapJake 01:53
yoleaux I saw MadcapJake 2 Dec 2016 06:23Z in #perl6: <MadcapJake> well I'm gonna fiddle with this laptop some more and then hit the hay. Glad to be back! See you this weekend!
samcv ready for 1.9 to be released now for Atom syntax highlighter
just fixed an annoying one, where regex didn't highlight anywhere except after = , ) { ~ symbols. so got around it 01:55
well. fixed the specific issue where bare regex used like => / / wouldn't highlight and also made rx/ / or m/ / anywhere, even after things it 'thinks' mean a regex is possible to follow 01:56
i've thought all day about how for it to recognize if bare / / is a regex and not the / operator. but haven't been able to come up with anything that would work 01:57
so bare ones at least work in almost all normal places at least and you can get it to highlight right 01:58
01:59 giraffe joined
skrshn I can use ^methods to get the methods of a object. How do I drill more and get the different ways to construct an object? I.e., getting more information for things like Method+{<anon|140540621714592>}.new ? 02:01
02:03 pyrimidine joined
MasterDuke m: for IO::Path.^methods.grep({.name ~~ /"new"/}) {say .name ~ .signature.perl} 02:07
camelia rakudo-moar a2ede3: OUTPUT«new:(RESTRICTED-CLASS $: | is raw)␤»
MasterDuke m: for Seq.^methods.grep({.name ~~ /"new"/}) {say .name ~ .signature.perl} 02:08
camelia rakudo-moar a2ede3: OUTPUT«new:(Seq $: Iterator:D $iter, *%_)␤new-consumed:(Seq $: *%_)␤»
02:08 pyrimidine left
MasterDuke skrshn: ^^^ something like that? 02:08
skrshn MasterDuke: perfect. thanks 02:12
Actually, no
Without reading the documentation, I wanted to know how I can create a new Date object for instance 02:13
seatek samcv: is that atom editor including your recent fixes to highlighting perl6, if you build the editor from source?
02:13 bjz joined
skrshn I wanted to get something that will tell me one option is Date.new(:year(2016), :month(12), :day(3)) 02:14
m: for Date.^methods.grep({.name ~~ /"new"/}) { say .name ~ .signature.perl } 02:15
camelia rakudo-moar a2ede3: OUTPUT«new:(Date $: | is raw)␤new-from-daycount:(Date $: $daycount, :&formatter, *%_)␤»
skrshn this does not tell me that
02:15 cibs left
skrshn Is this even possible to find apart from reading documentation? 02:16
02:17 cibs joined
seatek skrshn: new is where you pass in the attributes you'd like to defined on construction right? You're asking if there is a way to know ahead of time what attributes are necessary for object object construction? 02:19
02:20 diakopter left
skrshn seatek: yes. If I had not idea how to construct an object, say Date, I wanted to know all the different ways I can construct a new Date object 02:21
seatek so you want a list of all attributes for a class?
skrshn I want to do this just by doing meta object queries
02:22 rodarmor left
skrshn I know you can get that from ^attributes. But sometimes, the constructor may be different from that 02:22
MasterDuke interesting. if you look at github.com/rakudo/rakudo/blob/nom/...re/Date.pm there definitely are more new()s than Date.^methods shows
02:23 rodarmor joined
seatek oh i see, where someone has overridden the normal new, which slurpily takes in everything 02:24
so you want to know when someone has overriden the normal way that "stock" new exists, and what the signatures are for those overriden "new"s
02:25 pierre_ joined, pyrimidine joined 02:28 pierre_ left 02:30 cognominal left
babydrop reads the backlog and grumbles 02:30
mspo can I pass a block {} to a sub? 02:32
babydrop mspo: yes
02:33 cognominal joined
MasterDuke babydrop: any idea what we're missing about ^methods? it doesn't seem to show multies (multis? which does [Coke] prefer?) 02:35
babydrop m: for Date.^methods.grep({.name ~~ /"new"/}) { say .name ~ .signature.perl ~ .candidates } 02:36
camelia rakudo-moar a2ede3: OUTPUT«Method object coerced to string (please use .gist or .perl to do that)␤ in block at <tmp> line 1␤Method object coerced to string (please use .gist or .perl to do that)␤ in block at <tmp> line 1␤Method object coerced to string (please use .gist …»
babydrop Nah, I'm going to go back to sleep.
ugexe m: sub foo(Block $block) { say $block() }; foo { 42 * 10 } 02:37
camelia rakudo-moar a2ede3: OUTPUT«420␤»
mspo oh just Block type
seems so obvious now ;)
MasterDuke m: for Date.^methods(:local).grep({.name ~~ /"new"/}) {for .candidates { say .name ~ .signature.perl } } 02:38
camelia rakudo-moar a2ede3: OUTPUT«new:(Date $: $year, $month, $day, :&formatter, *%_)␤new:(Date $: :$year!, :$month = 1, :$day = 1, :&formatter, *%_)␤new:(Date $: Str $date, :&formatter, *%_)␤new:(Date $: Dateish $d, :&formatter, *%_)␤new:(Date $: Instant $i, :&formatter, *%_)␤ne…»
MasterDuke skrshn: ^^^
babydrop++
seatek babydrop++ MasterDuke++ :)
skrshn MasterDuke: +1 02:42
babydrop: +1
02:44 ilbot3 left 02:45 AlexDaniel left 02:46 cognominal left 02:47 cognominal joined, ilbot3 joined
skrshn Has anyone used Julia? How does it its multi-dispath compare with Perl6? 02:56
02:58 diakopter joined
skrshn Its claim to fame is in c-like performance using a JIT and type information 02:58
I can't wait for Perl6 to get there :-). 02:59
03:01 cognominal left, cognominal joined
mspo julia embeds llvm I think 03:12
03:16 cognominal left, cognominal joined
babydrop mspo: well, the type constraint is not required. In fact, by restricting it to block, you're throwing away, Sub and WhateverCode 03:21
03:21 noganex_ joined 03:22 cxreg joined
skrshn mspo: that is correct 03:22
03:24 noganex left, thowe left 03:30 kurahaupo_ left 03:31 cognominal left, cognominal joined 03:32 kurahaupo__ joined
skrshn what is the difference between put and say? I was not able to glean that from the docs 03:32
03:33 skids joined
babydrop skrshn: say calls .gist, put calls .Str 03:34
MasterDuke doesn't say also truncate after a certain number of bytes?
babydrop Nope
.gist of List outputs only first 100 els tho 03:35
MasterDuke ahh, that's what i was thinking of
skrshn babydrop: thanks 03:36
03:37 kurahaupo left 03:38 kurahaupo joined 03:45 tadzik left, M-Illandan left, Matias[m] left, Matthew[m] left 03:46 ilmari[m] left, dp[m] left, cognominal left, cognominal joined
babydrop Oh no... the matrix crashed! 03:46
03:49 mrsolo left
dalek c: c2b2b1a | titsuki++ | doc/Type/Iterator.pod6:
Fix example of pull-one
 03:55
c: 9a8e564 | titsuki++ | doc/Type/Iterator.pod6:
Merge pull request #1029 from titsuki/fix-pull-one

Fix example of pull-one
synopsebot6 Link: doc.perl6.org/type/Iterator
samcv seatek, you have to install Perl 6 FE 04:03
the builtin ''''Perl 6 Highlighter'''' basically highlights mostly like perl 5
but version 1.8 was released a few days ago has some of my fixes. Here's the changelog for v1.8, and v1.9 which will be released whenever MadCapJake does the pull github.com/samcv/language-perl6fe/...ANGELOG.md 04:05
it's a long list!
seatek samcv, thanks! :) I'll just install through a .deb then rather than compiling. so these are separately available modules... i'm really curious how the syntax highlighting will compare to emacs 04:08
and for regex vs op on '/' - seems like it would be very very hard without knowing the context of the stuff around it 04:09
that stuff makes me crazy, because it's SO MUCH thinking for so little a thing :)
emacs highligher gets confused by the use of < or > stuff around quotes '"' 04:10
so much so that i've mostly stopped using the <> syntax for hash keys 04:12
04:14 pyrimidine left 04:18 ilmari[m] joined 04:22 M-Illandan joined, Matthew[m] joined, tadzik joined, dp[m] joined, Matias[m] joined 04:23 TEttinger left 04:24 bjz left, TEttinger joined 04:27 jeek left 04:28 jeek joined
samcv seatek, yeah basically :P 04:33
before my changes it would only work if = , ( ) ~ or `when` were the thing preceeding it. or it was at the very start of the line. so added => as another whitelisted thing. 04:35
04:35 kurahaupo__ left
samcv plus also made rx/ or m/ always highlight no matter where it was, so you can always add that in case you're having it come after something odd 04:35
04:35 kurahaupo left
samcv well in version 1.9 it'll be like that 04:36
1.8 doesn't have that change
04:36 AlexDaniel joined
AlexDaniel m: say :35<lizmat> 04:36
camelia rakudo-moar 37ead5: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Invalid base-35 character 'z': 3li7⏏5zmat. Please use one of 0..9, A..Y␤at <tmp>:1␤------> 3say :35<lizmat>7⏏5<EOL>␤»
babydrop Don't tell me there's something wrong with it :) 04:37
AlexDaniel m: say :35<lizmatz>
camelia rakudo-moar 37ead5: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Invalid base-35 character 'z': 3li7⏏5zmatz. Please use one of 0..9, A..Y␤at <tmp>:1␤------> 3say :35<lizmatz>7⏏5<EOL>␤»
AlexDaniel well, it can probably tell about all invalid characters…
babydrop idly mentions my $x = 42|72 highlights 42 as a number, but 72 as not
AlexDaniel: I rather we didn't make an already dense method denser. 04:38
AlexDaniel babydrop: I'll think about it :D
but yes, what we have now is really good, I think 04:39
babydrop++
babydrop AlexDaniel: the error is different for :35('meows') form cause that one uses val()
AlexDaniel m: say :35('lizmatz')
camelia rakudo-moar 37ead5: OUTPUT«Cannot convert string to number: malformed ':35' style radix number, expecting '>' after the body in '3:35<li⏏5zmatz>' (indicated by ⏏)␤ in block <unit> at <tmp> line 1␤␤Actually thrown at:␤ in block <unit> at <tmp> line 1␤␤»
AlexDaniel oh and that's not a compile-time error
babydrop yeah
AlexDaniel so I guess people should prefer <> where possible? 04:40
babydrop I guess
samcv m: q'test'
camelia rakudo-moar 37ead5: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Two terms in a row␤at <tmp>:1␤------> 3q'test7⏏5'␤ expecting any of:␤ infix␤ infix stopper␤ statement end␤ statement modifier␤ statement modifier l…»
samcv AlexDaniel, do you know, i think this should work?
babydrop m: say :12(':36<inseption>')
camelia rakudo-moar 37ead5: OUTPUT«52644193907927␤»
04:40 bjz joined
samcv m: say q 'test' 04:40
camelia rakudo-moar 37ead5: OUTPUT«test␤»
AlexDaniel m: say q‘test’ 04:41
camelia rakudo-moar 37ead5: OUTPUT«test␤»
AlexDaniel hm
samcv yeah i found it while i was getting all the q's working in Atom P6 FE
babydrop samcv: it's cause ' is a valid identifier character. It's not supposed to work there
AlexDaniel right!
samcv but it works with a space? ah ok
so that should be documented
AlexDaniel m: say q-test-
camelia rakudo-moar 37ead5: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Missing required term after infix␤at <tmp>:1␤------> 3say q-test-7⏏5<EOL>␤ expecting any of:␤ prefix␤ term␤»
AlexDaniel m: say q -test-
camelia rakudo-moar 37ead5: OUTPUT«test␤»
AlexDaniel same thing
samcv say q -test-
m: say q -test- 04:42
camelia rakudo-moar 37ead5: OUTPUT«test␤»
samcv ah kk
AlexDaniel well, but… that's LTA
samcv what's LTA stand for? 04:43
babydrop It's perfectly fine
AlexDaniel less than awesome
babydrop You don't call qameowa LTA.
AlexDaniel babydrop: well, it would've been awesome if it said that you have to use a space there…
samcv yeah S02 seems to imply it should work. idk should i file a bug on it or no?
babydrop There's 10FFFF give or take characters available for use, we can live with a few thousand not being available.
AlexDaniel samcv: no
samcv ok so that's the correct functionality
it not working 04:44
AlexDaniel yes it's correct
samcv just the error isn't that great
AlexDaniel yes, not sure if there is a way to improve it though
samcv ok, so i'll make a PR for the docs to say you can't use docs.perl6.org/language/syntax#Identifiers any of these
04:44 m0ltar left
samcv so it's documented 04:44
babydrop commit directly
AlexDaniel samcv: yes please, though you can use them if you use a space, I think?
samcv yeah exactly
every thing i tried works if you use a space 04:45
babydrop m: say q ameowa
camelia rakudo-moar 37ead5: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Alphanumeric character is not allowed as a delimiter␤at <tmp>:1␤------> 3say q7⏏5 ameowa␤»
AlexDaniel () don't work without a space too
babydrop heh
IIRC that works in Perl 5 :)
AlexDaniel u: pile of poo 04:46
unicodable6 AlexDaniel, U+1F4A9 PILE OF POO [So] (💩)
AlexDaniel m: say q 💩meow💩
camelia rakudo-moar 37ead5: OUTPUT«meow␤»
AlexDaniel m: say q💩meow💩
camelia rakudo-moar 37ead5: OUTPUT«meow␤»
AlexDaniel with space or without, poo works
samcv i tried to get all those weird things to work
but for whatever reason it wasn't working :( 04:47
i mean i have it working for heredocs
but i have no clue why it failed to work for whatever you wanted
AlexDaniel m: say Q:w'test' 04:49
camelia rakudo-moar 37ead5: OUTPUT«5===SORRY!5===␤Unrecognized adverb: :w'test␤at <tmp>:1␤------> 3say Q:w'test7⏏5'␤Couldn't find terminator ' (corresponding ' was at line 1)␤at <tmp>:1␤------> 3say Q:w'test'7⏏5<EOL>␤ expecting any of:␤ '␤␤»
AlexDaniel babydrop: /o\ there are just too many cases to cover 04:50
is there a unicode character for /o\? 04:51
u: 🙆 04:53
unicodable6 AlexDaniel, U+1F646 FACE WITH OK GESTURE [So] (🙆)
AlexDaniel well… I'm not sure if that's ok
u: 💆 04:54
unicodable6 AlexDaniel, U+1F486 FACE MASSAGE [So] (💆)
samcv ugh literally put my regex through ruby, because that's the engine that atom/textmate/whatever uses 05:03
and it works fine
05:03 pyrimidine joined
samcv backreferences are used fine for the heredocs 05:04
05:11 pierre__ joined
babydrop star: use DBIish 05:12
camelia ( no output )
babydrop .tell stmuk_ does a R* install run module tests? If yes, may be worth ensuring this ticket is resolved prior to next release: github.com/perl6/DBIish/issues/77 05:13
yoleaux babydrop: I'll pass your message to stmuk_.
05:13 pierre__ left 05:14 pierre_ joined 05:15 BenGoldberg left
skrshn samcv: in your update, does if ( 1 <= @*ARGS.elems ) { ... } work? 05:16
samcv let me check 05:17
skrshn In the vim-perl module, it does not work
samcv i.imgur.com/OBmXG9m.png
looks fine 05:18
05:18 pierre_ left
skrshn awesome 05:18
babydrop skrshn: FWIW .elems isn't needed there
Arrays does Cool
MasterDuke neither are those () around the conditional 05:19
samcv skrshn, it seems to work for me with vim?
babydrop aye
samcv maybe it's part of a larger amount of text that's breaking it?
skrshn hmmm
samcv do you have the full text?
05:19 pierre_ joined
MasterDuke i get bad hightlighting in vim for that text 05:19
samcv filetype is perl6? 05:20
MasterDuke starting from the <
yeah 05:21
samcv this is what i get i.imgur.com/Wh7jfwy.png
skrshn This is what I get : imgur.com/a/6Mvkk
Where r u getting vim-perl from? 05:22
I am using 'vim-perl/vim-perl'
samcv i think just whatever's in vim but
github.com/vim-perl/vim-perl6
this is where they moved it to
MasterDuke i.imgur.com/RzVOkno.png
skrshn MasterDuke: that is what I am getting too 05:23
both vim-perl/vim-perl and vim-perl/vim-perl6 does not work for me
samcv hm 05:24
that's... weird
05:24 khw left
samcv skrshn, that's what i see if i set the filetype to perl 05:24
:set ft=perl6
skrshn No.. mine is set to perl6 05:25
samcv hmm
well. i see exactly what you see if i set filetype to perl
skrshn btw.. I am using vim8.0 on mac .. not sure if that has anything to do with it
MasterDuke vim8.0 on linux here 05:26
samcv this is what i see i.imgur.com/NVzXWHC.png
I'm using arch. didn't install any plugin just what vim came with. though
i think it's the same as in that repo? 05:27
i mean i made a PR to the perl6 one
i remember it being the same
sec
skrshn imgur.com/a/1qLwE
MasterDuke i'm using arch also 05:28
samcv hmm..
wtf
this is my vimrc gitlab.com/samcv/dot-files/blob/master/_.vimrc
MasterDuke my ~/.vimrc is 'set mouse-=a' 05:30
what's the md5sum of your ~/.vim/syntax/perl6.vim? 05:31
skrshn samcv: it works if I use your vimrc 05:32
samcv it highlights fine even if i remove my vimrc or run vim as root
oh?
\o/
MasterDuke mine is 683772507d529bcf195b549d9152326b
samcv /usr/share/vim/vim80/syntax/perl6.vim mine's here
3dd289aa3adbdd2f3a47fda089fb637a
skrshn imgur.com/a/qW9g0 05:33
samcv skrshn, using my vimrc?
skrshn yes
samcv lemme do a diff
between git perl6 vim and the one i have
MasterDuke ah, i removed my ~/.vim/syntax/perl6.vim and now it highlights that line fine 05:34
samcv heh
heh cry.nu/p/x20b this is a diff 05:35
i think they're mostly the same one's just been preprocessed or something 05:36
idk. submitted a PR to vim-perl. a while later they tell me to do the PR in vim-perl6
and it's been months?
no reply
skrshn so I removed both vim-perl and vim-perl6.. only used the default one that came with vim8.0 -- works! 05:38
samcv sweet 05:39
yeah i think the one in that repo isn't 'preprocessed'
so i think that's causing the issues
for more complex things or something? no clue
skrshn or there might have been a regression 05:40
05:40 darutoko joined 05:41 pierre__ joined
skrshn I also see this imgur.com/a/pR8XH 05:41
05:42 pierre_ left
skrshn samcv: do you know of anything in the atom/perl6fe version that works better than what is there in vim8.0? 05:44
samcv uhm
skrshn I have not gone through the changelogs in detail
you don't have to be exhaustive.. just few or many .. I ssh into machine so atom is not an option for me 05:45
samcv uh
heh looks like they have the same issue with Q[] quoting forms highlighting characters that shouldn't be escaped 05:47
well looking at it here it doesn't allow q{{}} q(()) though that's not that big 05:48
skrshn, you should check the changelog github.com/samcv/language-perl6fe/...ANGELOG.md
skrshn yes
samcv well it only is there for 1.8 and the soon to be released 1.9, whenever madcapjake takes my pull
it's not thatttt long a list
idk i haven't used vim too heavily for a while. i know it was mostly good when i was using it. but i wasn't doing as many things with it then 05:49
oh it can do heredocs
doesn't seem to work on vim 05:50
skrshn Ok
good to know
samcv heh even got q:to/📝🔚/ working recently 05:51
well today actually
let me check if vim lets you nest “ ” or q[ [] ]
05:52 rburkholder left
samcv ok it doesn't support fancy quotes at all 05:52
uhhh it doesn't support multiline comments 05:53
seatek does atom do multiple file split vertically?
samcv yes
seatek ooo!
and horizontally to then yes? 05:54
samcv like #`( #`(( #`( #`｢ idk not gonna list them all
yeah seatek
seatek and perl6 syntax highlight?
samcv yeah
seatek emacs keybindings?
samcv you can get a plugin for that
pretty sure
seatek i'm feeling... happiness?
samcv aww vim doesn't support multi line comments at all. that's pretty lame 05:55
seatek emacs does awesome at multiline comments. even helps you out with them
skrshn yes it is
samcv can even nest #`( ( ) ) them in atom now too
skrshn didn't notice yet
samcv and it'll match them
seatek so can you easily program search/replace macros pretty easily in atom? 05:56
such as... like in emacs i frequently "record" keystrokes... and play them back
samcv prolly
there's so many plugins
seatek hmmm 05:57
samcv and really nice to search all the functions you can do
seatek does it have a sense of classes and methods in perl6?
samcv ctrl + shift + p for me, and search for like every thing
yeah
seatek ok that's a step up then 05:58
samcv does vim not highlight that?
05:58 Cabanossi left
seatek i mean logically know about classes... like you can highlight a method in one file, and it can take you to where it was defined in the class 05:59
IDE'ish type stufff ;)
05:59 Cabanossi joined
samcv oh 05:59
can emacs do that for perl 6? 06:00
seatek i've never seen anything be able to do that with perl... except Eclipse did a bit ...
06:00 rurban joined, rurban left
samcv there's some perl 6 tool thing that supposed to let you click on things to get the docs 06:00
seatek you could do it by running a separate type of ctags i think
samcv but it doesn't work
seatek yeah
samcv so i disabled that plugin sadly
it just errors, idk how old it is
06:01 djbkd joined
seatek i wonder if things like that will be easier with perl6 being more standardized, if it is.. 06:01
skrshn what would be cool would be to have syntax highlighting based on the parse tree !!!
samcv heh 06:02
if somebody made a perl 6 based editor that would be great
seatek i would hate to load a large file for editing with that...
samcv Perl 6's killed app!
will be a perl 6 editor!
haha
AlexDaniel seatek: I am pretty sure spacemacs can do that by default 06:03
seatek padre needs to update
ooo really?
samcv with perl 6?
AlexDaniel maybe with perl 6 too, I remember it did something smart
seatek "sophisticated and polished". Emacs advanced kit focused on Evil 06:04
AlexDaniel I mean, maybe it was giving false positives sometimes, but it was better than nothing 06:05
babydrop m: use nqp; my $x = (1,2); say $x.^attributes; say nqp::getattr(nqp::decont($x), List, '$!reified')
camelia rakudo-moar acbbc8: OUTPUT«(List $!reified Mu $!todo)␤(1 2)␤»
babydrop I don't get it... A List has an attribute $!reified of... type List?
AlexDaniel seatek: spoiler: I stopped using it after a year or so 06:06
seatek why??
AlexDaniel well, generally I think it is easier to build upon something simple than trying remove features from something complex… and there is definitely some shit you want removed :) 06:07
skrshn :-) 06:08
babydrop m: dd (lazy (1,2)).Array
camelia rakudo-moar acbbc8: OUTPUT«(1, 2, Any, Any, Any, Any, Any, Any, Any, Any... (lazy list)␤»
seatek so irritating stuff would sneak in sometimes during your normal workings, that you'd have to find a way to work around?
oh they've changed around stuff 06:09
making it more vim-ish in some ways
AlexDaniel also, there are two modes I think
one is vim-ish and another one is emacs-ish
I've been trying the vim-ish one… maybe if I didn't, it would've been different, not sure right now 06:10
seatek wow i would have never had the guts or craziness to dive into emacs innards like they must have had to for this
AlexDaniel well, not really… they've just thrown many packages into a huge pile and that's it 06:11
you can install evil and a bunch of other stuff and it will work just as well 06:12
seatek "Spacemacs was initially intended to be used by Vim users who want to go to the next level by using Emacs"
you went backwards AlexDaniel !
AlexDaniel yea I know. In retrospect it was a bad idea 06:13
seatek well, i do like that it's purple
great bottom display 06:14
i should try it for real
06:14 aborazmeh left
seatek i search for "spacemacs perl6" and the 10th result is "Is Assange alive?" 06:17
skrshn samcv: is it easier to write perl6fe for atom than vim script? 06:18
AlexDaniel seatek: well, the first result should actually be this: github.com/syl20bnr/spacemacs/issues/5019
babydrop: by the way, it is quite interesting what they do in spacemacs project to keep the ticket list clean 06:22
babydrop: for example, see this: github.com/syl20bnr/spacemacs/issues/7334 . They are basically drowning in all the tickets :)
samcv skaji, probably. idk though i wish i could have more logic than just regex to do it 06:24
but it's easier
babydrop AlexDaniel: so random people get random issues?
samcv but it's textmate style grammars which use en.wikipedia.org/wiki/Oniguruma and it's the same grammar github uses to highlight perl 6 code 06:25
fyi
06:25 bjz_ joined, bjz left, shadowpaste left
babydrop AlexDaniel: I like rust's method. A bot randomly assigns a PR to a contributor to review and merge (dunno if they do that for issues too) 06:25
AlexDaniel babydrop: at the moment we have no problem with PRs, I think 06:26
but yeah, that's an interesting idea for issues…
skrshn samcv: thanks
babydrop wants to have a problem with PRs 06:28
AlexDaniel babydrop: the idea there is to clean the list of issues, not to force people to actually do stuff
you already know that reviewing one thousand tickets is a bit of a problem :)
06:29 telex left
AlexDaniel we can actually try something like this for RT queue as an experiment 06:30
though I guess not many will be interested
skrshn m: my $a = "foo"; say "Hello $($a // "bar")"; 06:33
camelia rakudo-moar 5635c5: OUTPUT«Hello foo␤»
skrshn m: my $a = "foo"; say "Hello {$a // "bar"}";
camelia rakudo-moar 5635c5: OUTPUT«Hello foo␤»
skrshn whats the difference?
seatek AlexDaniel: i'm liking very much a better way to work with buffers 06:34
AlexDaniel seatek: ?
06:34 MilkmanDan left
seatek just read a thing where the buffer list can be a constant thing there for you, for easier switching 06:35
06:35 pierre__ left
AlexDaniel seatek: huh? 06:35
seatek: you mean ido?
seatek buffers... emacs buffers... like files... when you're working with 10 or 20 files at once
babydrop skrshn: one creates a new block the other doesn't
seatek i've never found a very smooth and satisfying way to work with them 06:36
AlexDaniel seatek: yea, but what is different in spacemacs?
skrshn babydrop: thanks. where can I find in docs? been searching
babydrop m: {say "Hello ${$^a // "bar"}"}("foo") 06:37
camelia rakudo-moar 5635c5: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Unsupported use of ${$^a // "bar"}; in Perl 6 please use $($^a // "bar") for hard ref or $::($^a // "bar") for symbolic ref␤at <tmp>:1␤------> 3{say "Hello ${$^a // "bar"}7⏏5"}("foo")␤»
babydrop o.O
seatek a b hotkey that provides a defined area, instead of replacing existing editing buffers in front of your face
babydrop skrshn: no idea, sorry.
skrshn babydrop: np
AlexDaniel skrshn: there is a high chance that it is not, please file a doc issue
skrshn AlexDaniel: sure 06:38
AlexDaniel skrshn: and in case it is documented somewhere, include “I looked here, here and here” in your issue, so that we can add appropriate links
06:38 telex joined
AlexDaniel skrshn: github.com/perl6/doc/issues 06:38
06:38 djbkd left 06:39 xinming_ joined
AlexDaniel skrshn: I still don't get it, sorry :| 06:39
06:39 MilkmanDan joined
skrshn AlexDaniel: what are you talking about? 06:40
AlexDaniel oops
babydrop hahaha
AlexDaniel seatek: ↑
skrshn AlexDaniel: :-)
06:42 xinming left 06:47 nightfrog left
skrshn For general info on string expansion where should I look in the docs? I don't see anything here docs.perl6.org/language.html 06:48
06:49 nightfrog joined
AlexDaniel skrshn: docs.perl6.org/language/quoting maybe? 06:49
skrshn cool tx 06:50
seatek i blame a customer for distracting me :) 06:53
06:56 pierre_ joined
babydrop What would be the negation of a wordboundary? ala \B in Perl 5? 06:56
AlexDaniel babydrop: <!before «> ? :X I don't know 06:59
skrshn samcv: one thing that I miss in vim-perl is the ability to type \tau<TAB> and it get replaced with the τ symbol. This works very well in the julia-vim
babydrop m: say "meow foo" ~~ m:g/<!before «>/
camelia rakudo-moar 5635c5: OUTPUT«(｢｣ ｢｣ ｢｣ ｢｣ ｢｣ ｢｣ ｢｣)␤»
skrshn is it there in atom? 07:00
babydrop $ perl -wlE 'say "meow foo" =~ /(\B.)/g'
samcv is that just a vim feature skrshn
babydrop eowoo
m: say "meow foo" ~~ m:g/<!before «>./
camelia rakudo-moar 5635c5: OUTPUT«(｢e｣ ｢o｣ ｢w｣ ｢ ｣ ｢o｣ ｢o｣)␤»
babydrop m: say "meow foo" ~~ m:g/<!before [«|»]>./
camelia rakudo-moar 5635c5: OUTPUT«(｢e｣ ｢o｣ ｢w｣ ｢o｣ ｢o｣)␤»
skrshn no.. it is special to julia-vim
babydrop AlexDaniel++
samcv skrshn, there's probably ways to do that maybe
skrshn its probably easy to do by looking at how it is implemented in julia-vim. I need to up my vim script skills before I can do it myself 07:01
AlexDaniel m: say "meow foo" ~~ m:g/ <!|w>. / 07:02
camelia rakudo-moar 5635c5: OUTPUT«(｢e｣ ｢o｣ ｢w｣ ｢o｣ ｢o｣)␤»
AlexDaniel babydrop: ↑ ?
skrshn it makes a lot of sense -- they have a good mapping between latex symbols so you could do \sqrt<TAB> and it expands to √
samcv idk skrshn i would just add it to my compose key :\ 07:03
tbh
speaking of which i should add \n and \t to my compose key list
growing every week that thing is
babydrop AlexDaniel: first time I see the |w thing :/
samcv u: new line 07:04
unicodable6 samcv, U+2424 SYMBOL FOR NEWLINE [So] (␤)
samcv, U+2B92 NEWLINE LEFT [So] (⮒)
samcv, U+2B93 NEWLINE RIGHT [So] (⮓)
AlexDaniel babydrop: oh, it turns out these are actually mentioned in synopses
“Perl 5's \b is replaced by a <|w> "word boundary" assertion, while \B becomes <!|w>”
babydrop Oh, awesome.
07:04 CIAvash joined 07:05 CIAvash left
samcv also docs.perl6.org/language/regexes#&l...d_boundary 07:05
07:05 ZzZombo left
samcv << and >> 07:05
do those convert to <|w> deep down though AlexDaniel ?
AlexDaniel I have no idea 07:06
no I don't think so, because << >> are left/right ?
babydrop samcv: doubt it, since they are directional
07:09 domidumont joined 07:10 CIAvash joined 07:12 bjz joined 07:13 domidumont left, bjz_ left 07:14 domidumont joined, Khisanth left
AlexDaniel m: say "meow foo" ~~ m:g/ <!wb>. / 07:22
camelia rakudo-moar 5635c5: OUTPUT«(｢e｣ ｢o｣ ｢w｣ ｢o｣ ｢o｣)␤»
AlexDaniel babydrop: ↑ this too 07:23
babydrop What's the difference beteen the two?
the <!|w> and <!wb>
AlexDaniel “As mentioned above, the \b and \B word boundary assertions are gone, and are replaced with <|w> (or <wb>) and <!|w> (or <!wb>) zero-width assertions.” 07:24
so I guess no difference
07:28 Khisanth joined, domidumont left 07:31 skrshn left
samcv m: my $var ≔ 4; 07:31
camelia rakudo-moar 5635c5: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Confused␤at <tmp>:1␤------> 3my $var7⏏5 ≔ 4;␤ expecting any of:␤ infix␤ infix stopper␤ statement end␤ statement modifier␤ statement modifier loop␤»
AlexDaniel huggable: texas
huggable AlexDaniel, All of Perl 6's fancy Unicode operators have traditional ASCII symbol alternatives (aka Texas Variants): See doc.perl6.org/language/unicode_texas
AlexDaniel meh
samcv u: ≔ 07:32
unicodable6 samcv, U+2254 COLON EQUALS [Sm] (≔)
AlexDaniel huggable: save me from Texas :is: gist.github.com/AlexDaniel/c89bd2786f9b63f31e4c
huggable AlexDaniel, Added save me from Texas as gist.github.com/AlexDaniel/c89bd2786f9b63f31e4c
AlexDaniel samcv: ↑ ;)
samcv heh AlexDaniel 07:34
oh AlexDaniel ∧ would be && i believe en.wikipedia.org/wiki/Logical_conjunction 07:37
AlexDaniel samcv: ? isn't it already there? 07:38
samcv is it?
seatek i've been having to dig through php the last hour or so. and probably a few hours more. sorry off topic. i just need someone to know. 07:39
AlexDaniel gist.github.com/AlexDaniel/c89bd27...4c#-and--1
samcv m: oh it is
camelia rakudo-moar 5635c5: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Undeclared routines:␤ is used at line 1␤ it used at line 1␤ oh used at line 1␤␤»
samcv oops
yeah it is AlexDaniel . at the bottom :p
AlexDaniel samcv: you think it should be higher up? Why don't you submit a pull request along with an RFC ticket? ;) 07:41
samcv can you do a PR for gists?
babydrop nope
AlexDaniel I meant PR for rakudo… :) 07:42
samcv k
babydrop u: rx_syntax
unicodable6 babydrop, U+0072 LATIN SMALL LETTER R [Ll] (r)
babydrop, U+005F LOW LINE [Pc] (_)
babydrop, U+0078 LATIN SMALL LETTER X [Ll] (x)
babydrop, gist.github.com/c581129eb70ee76b4e...332ebb1328
babydrop ugh
AlexDaniel … what did you mean?
babydrop Stupid terminal doesn't let me copy-paste a single char 07:43
u: ∧
unicodable6 babydrop, U+2227 LOGICAL AND [Sm] (∧)
babydrop u: LOGAL OR
AlexDaniel u: ʌ
unicodable6 babydrop, Found nothing!
AlexDaniel, U+028C LATIN SMALL LETTER TURNED V [Ll] (ʌ)
babydrop u: LOGICAL OR
unicodable6 babydrop, U+2228 LOGICAL OR [Sm] (∨)
babydrop, U+22C1 N-ARY LOGICAL OR [Sm] (⋁)
babydrop, U+22CE CURLY LOGICAL OR [Sm] (⋎)
babydrop, gist.github.com/aeb33d8deb642f5033...53cb8c5c5e
AlexDaniel MORPHOLOGICAL :D 07:44
samcv yeah i'll do that AlexDaniel
thanks for the list tho. that is very helpful
babydrop would expect a PR that adds ∧ would also add one for 'OR', for consistency.
samcv yeah exactly 07:45
AlexDaniel u: defined
unicodable6 AlexDaniel, Found nothing!
07:47 djbkd joined
samcv adding || for || may not be right. since || is boolean and & is boolean in maths. and ∧ ∨ are like && and || in perl 6 07:48
moritz lol I blug: perlgeek.de/blog-en/perl-6/2016-bo...tting.html
samcv whereas || in math is more like | and & in math is like & in perl 6
AlexDaniel o_O
moritz: substr $sudoku, $line-number * 9, 9; # Uh! 07:49
moritz: why not rotor?
ah, I see 07:50
07:54 domidumont joined 07:59 setty1 joined 08:09 bjz left, bjz_ joined 08:10 wamba joined 08:11 setty1 left 08:12 setty1 joined
babydrop New Advent Post: perl6advent.wordpress.com/2016/12/...ret-santa/ 08:14
08:14 skids left 08:21 Tonik joined 08:31 RabidGravy joined 08:32 cognominal left 08:35 firstdayonthejob joined 08:37 CIAvash left 08:40 djbkd left 08:45 cognominal joined 08:50 cibs left 08:51 cibs joined 08:52 nchambers is now known as God 08:53 pyrimidine left 08:55 AlexDaniel left 09:00 pyrimidine joined
dalek c: 2f5bf36 | (Zoffix Znet)++ | doc/Type/Mu.pod6:
Document Mu.emit

Added to Rakudo in github.com/rakudo/rakudo/commit/4e76827bb0 Tests: github.com/perl6/roast/commit/e3c8d5a6e5 RT#128968: rt.perl.org/Ticket/Display.html?id=128968		 09:01
synopsebot6 Link: doc.perl6.org/type/Mu
synopsebot6 Link: rt.perl.org/rt3//Public/Bug/Displa...?id=128968
09:01 God is now known as nchambers
babydrop notes that &emit doesn't seem to be documented 09:01
09:03 rurban joined 09:04 rurban left 09:07 domidumont left 09:08 kurahaupo__ joined 09:10 pierre_ left
babydrop 0.o this hangs: ./perl6 --target=parse -e 'my ($, $ where 42)' 09:11
Even tho without --target it works :\
09:12 pierre_ joined
babydrop Golfed ./perl6 --target=parse -e 'my ($, $)' 09:14
09:29 CIAvash joined, CIAvash left 09:30 CIAvash joined 09:31 CIAvash left 09:32 CIAvash joined, CIAvash left
stmuk_ babydrop: yes R* runs module tests and intrusive tests in upstream modules are a problem (eg. network access) 09:34
yoleaux 05:13Z <babydrop> stmuk_: does a R* install run module tests? If yes, may be worth ensuring this ticket is resolved prior to next release: github.com/perl6/DBIish/issues/77
stmuk_ github.com/rakudo/star/issues/77 09:35
RabidGravy postgres is kind of strange in that it doesn't install a test database that can be used (and you can bail if it doesn't exist,) I think all of Oracle, Sybase, Sql Server, Informix do 09:38
stmuk_ I've seen tests attempt to fire up temp postgres servers on high ports (but its very fragile) 09:39
RabidGravy yeah, though that report does raise a few questions about the reporters situation, what kind of application is broken by a surprise table and given that is a possibility what kind of person installs un-tested software against a live instance of it 09:42
stmuk_ the test could rollback a transaction 09:43
RabidGravy does postgresql rollback DDL? a lot of engines don't 09:44
stmuk_ yes it does unlike mysql 09:45
github.com/perl6/DBIish/blob/maste...pg-array.t 09:46
actually maybe the problem is using a standard env var to set the db 09:47
09:48 CIAvash joined
RabidGravy yeah, it should probably require some environment to set the DB and skip all the active tests if it isn't set 09:48
CIAvash babydrop: maybe it's the same bug as rt.perl.org/Public/Bug/Display.html?id=129262 09:50
RabidGravy I wan't isn't-deeply in a test:) 09:58
having performed an ugly filthy hack to work around the scope of 'state' variables 09:59
and the latest advent post comes to the rescue by inspiring "ok $workflow2.states.map(*.WHICH).none ⊆ $workflow.states.map(*.WHICH).any, "differ"" 10:04
10:06 rindolf joined, cxreg left
seatek i've never noticed DBIish installing a test table. then again i don't leave credentials in my environment variables either.. i can't imagine dbiish would rely on environment variables for tests. unless it was an author test 10:07
can't imaging dbiish using a table called "sal_emp" either for a test table ;) 10:08
RabidGravy a lot of libraries that interact with other software use environment variables
moritz what else should it use? 10:10
RabidGravy but yeah it does "/t/36-pg-array.t: CREATE TABLE sal_emp ("
stmuk_ it should look for an existing sal_emp and add entries ;)
seatek oh my gosh
well, it could never do it on mine because i don't allow environment variables to be used for such things -- because i'm SANE 10:11
we should get rid of that
RabidGravy it should just use different environment variables
seatek users would need to know about those ahead of time 10:12
RabidGravy i.e. DBISH_PGDATABASE rather than PGDATABASE
well that's what documentation is for
seatek yeah shifts the blame onto them then 10:13
as long as we make the environment variables non-standard enough
RabidGravy you can only stop people from shooting themselves in the foot if they aren't determined to do it
10:13 geekosaur left, domidumont joined
seatek it's character building :) 10:13
10:14 geekosaur joined
seatek speaking of which, back to PHP now :( :( 10:14
Ulti PHP 7? 10:23
seatek nope 5... 10:25
it makes me sad
10:26 cyphase left
Ulti samcv: re: perl6 editor in perl6 github.com/azawawi/farabi6 10:26
seatek: the last PHP I did was 4.2 4.4 transition so feel lucky 10:27
at a time when PHP 5 existed as only a blissful new hope no one used
also Drupal :'( never dev Drupal stuff
ever
just write your own CMS from scratch 10:28
seatek that's like 7's promise.
yeah i had to play with drupal for a bit too
you might as well from scratch
considering how long it takes to figure out what other people were thinking
Ulti though I imagine the average age of module maintainers isn't 17 anymore
seatek these old things just keep coming back to haunt me 10:29
Ulti :(
seatek php makes me feel like i'm using someone else's functions they created haphazardly, and trying to make them all tie together into something that works 10:30
Ulti its because that is what early PHP was
seatek that would make sense
Ulti yeah the main one there is a slight mismatch of names and parameter orderings in the core 10:31
timotimo don't forget function names in php were assigned so that as few as possible had the same strlen, because that was the hash function used for lookup :P
10:31 cyphase joined
seatek no way 10:31
that might explain why so many are crazy long
Ulti sometimes its thing(needle, haystack) other times its thing(haystack, needle) which is just annoying
moritz hayneedle, stack 10:32
seatek :)
Ulti steedle, hack
10:33 pierre_ left 10:34 lukaramu joined
seatek preg_match. i remember when i first saw that i was so happy. but that happiness lasted a very short time 10:34
Ulti PHP and JavaScript are great languages for demonstrating how important tooling and optimisation is to the survival of any language
timotimo news.php.net/php.internals/70691
seatek oh wow. :) 10:35
gotta hand it to him though for admitting it! :) that's pretty cool actually
timotimo clearly it hasn't been like that for a long, long time 10:36
but apparently about 100 of the functions php has are set up like that?
seatek i like some of the stuff like nl2br
10:36 pmurias joined
timotimo how do you feel about hyphens in function names in perl6? 10:36
seatek i love it when it works
but sometimes it confuses it 10:37
like method item-42() {}
pmurias Ulti: wasn't it more ease of install and getting started with PHP (and know it has market share)?
timotimo those are explicitly forbidden :)
seatek hehe
timotimo after the hyphen has the same constraints as the start of a name
moritz seatek: the rules are that there can't be numbers on eitehr side, to avoid possible confusion with - as subtraction
timotimo we don't allow function names to start with a number, either 10:38
moritz: no, numbers on the left of a hyphen are allowed
seatek yeah i noticed that when i was trying to do method code-404() ;)
Ulti pmurias: thats how it started but not how it keeps going
timotimo m: sub test4-twenty() { }
camelia ( no output )
10:38 cxreg joined
seatek i'm like, wow, yeah, perl can have dashes.. WHY IS THIS FAILING!? oh yeah, i didn't ready every single letter of every single doc yet for all the exceptional cases 10:39
Ulti things lik PHPStorm is almost certainly a big reason PHP is still trucking
seatek yeah it's super easy for web work
quick and dirty
Ulti seatek: from what I recall there was a big drive to switch to dashes after enough kebab casers demanded more kebabs
seatek i think it's so much clearer to read 10:40
Ulti I like that it makes methods stand out from anything else
seatek <> are also problemmatic in hashes sometimes
also
timotimo buts subs are also allowed to have dashes, Ulti :)
oh, ulti, could you re-run the jit log for me, please? :3 10:41
Ulti sure so are variables, but no one really uses it there is what I meant
because of things like $my-thing-$other-thing
timotimo i added two ops to moar's jit, which might be able to improve jittedness of your frames a little bit
10:41 pecastro left
timotimo but i'd need a measurement to actually see if it does anything noteworthy 10:41
Ulti erm I cant do it on the same computer :(
well not until tomorrow
timotimo oh, no need for the timings to be better
i think i just need the jitlog :)
Ulti k I can get that from here 10:42
timotimo yay
don't forget to build newest moar, too
Ulti timotimo: do you want my moar to be on HEAD then rather than whatever rakudo pulls in? 10:43
k
:(
that means me doing a new setup
timotimo oh, i'm sorry
Ulti np
timotimo don't have to do it quick, or soon
Ulti hah I do if you want it done
timotimo thing is, one of the things that currently weren't being jitted is "return", which might be hot 10:44
10:45 pecastro joined, CIAvash left, ufobat joined
timotimo actually ... 10:46
10:46 seatek left
timotimo it looks like almost all of those are either in the compiler or in emitting the profile 10:46
so maybe it's not even worth taking a look
10:48 seatek joined
Ulti well I'll set up on head anyway 10:48
ufobat good morning :-) 10:52
Ulti morning 10:53
timotimo greetings bat 10:54
moritz m: say EVAL '4+2' 10:57
camelia rakudo-moar 4e7682: OUTPUT«6␤»
moritz didn't we have some kind of requirement to use MONKEY; or so?
ah well, glad it's gone 10:58
timotimo no
it's not gone
m: my $unsafe-code = "say 'owned'"; EVAL $unsafe-code
camelia rakudo-moar 4e7682: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤EVAL is a very dangerous function!!! (use the MONKEY-SEE-NO-EVAL pragma to override this error,␤but only if you're VERY sure your data contains no injection attacks)␤at <tmp>:1␤------> 3-code = "say '…»
RabidGravy tadzik, if you get a minute I'd heartily commend github.com/tadzik/JSON-Unmarshal/pull/22 as my advent post will depend on it ;-) 11:00
11:00 kurahaupo_ joined, kurahaupo__ left
moritz timotimo: :( 11:01
timotimo RabidGravy: not bad
moritz: TT came up with it :)
11:06 cyphase left
RabidGravy though having thought about github.com/jonathanstowe/Tinky-JSON in the shower I'm going to re-design it to be more generally useful 11:06
11:07 cyphase joined
lizmat RabidGravy: I hope tadzik will find the time... he's doing some R&R in Lpn 11:08
London atm and is travelling back home tonight
11:09 hermes_ joined, hermes_ is now known as zamanf_
lizmat doing R&R& 11:09
11:11 CIAvash joined 11:12 zamanf left 11:16 complain_ joined 11:22 ugjka is now known as UgJkA 11:24 FROGGS joined
dalek c: 51a2391 | (Jan-Olof Hendig)++ | doc/Type/Cool.pod6:
Clarified that the argument to tanh is in radians and fixed incorrect example
 11:28
synopsebot6 Link: doc.perl6.org/type/Cool
11:34 firstdayonthejob left, mefzz joined 11:35 dogbert2 left 11:36 dogbert17 joined 11:44 abraxxa joined, abraxxa left
dalek c: 2e32d84 | (Jan-Olof Hendig)++ | doc/Type/Cool.pod6:
Added examples for acos
 11:45
synopsebot6 Link: doc.perl6.org/type/Cool
11:46 mefzz left 11:47 CIAvash left 11:51 rburkholder joined 12:02 cognominal left 12:09 gregf_ left
dalek c: a9f077d | (Jan-Olof Hendig)++ | doc/Type/Cool.pod6:
Replaced bogus asec examples
 12:11
synopsebot6 Link: doc.perl6.org/type/Cool
12:14 cognominal joined 12:17 cognominal left 12:20 cognominal joined 12:31 ab6tract joined, canopus left
ab6tract m: my $b = bag <blood love>; my $m = ("blood" => 1.1, "rhetoric" => 1, "love" => 1.2).Mix; dd ($m (-) $b) (+) ($b (-) $m); dd ($b (-) $m) (+) ($m (-) $b); # shouldn't these be the same? 12:32
camelia rakudo-moar 4e7682: OUTPUT«("love"=>0.2,"rhetoric"=>1,"blood"=>0.1).Mix␤("rhetoric"=>1).Bag␤»
12:33 notostraca joined
hahainternet can i define an infix operator that takes a parameter? similar to Z= 12:33
12:34 TEttinger left
timotimo you mean a meta-operator? 12:34
hahainternet oh is Z counted as a metaoperator? if so, yes :)
timotimo i actually don't think implementing your own metaops with just a simple definition is implemented :( 12:35
hahainternet not sure how in the repl i'd explore them
timotimo: that's a shame, but i appreciate that so much more seems to work and be coherent this time i'm making a good go at it 12:36
coercion / semicolon array list slips / custom metaoperators, two of those 3 are ludicrously complex
and other than minor annoyances it's been really fun so far 12:37
plus, i can just use a sub to do what i'm talking about
timotimo good to hear :)
hahainternet oh, 'Array of X' also broken, but again
12:38 canopus joined
hahainternet i would say at the moment it's quite difficult to know how to (for example) say 'type X is an Array of Integers with shape (2)' 12:38
timotimo broken in what way?
hahainternet m: class C is Array of Int {}
camelia rakudo-moar 4e7682: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤No such method 'set_of' for invocant of type 'Perl6::Metamodel::ClassHOW'␤at <tmp>:1␤»
timotimo oh
i see 12:39
well, that's because of the way Array gets parameterized
i think?
hahainternet shrugs
this is just for fun code so it's not a problem
12:39 notostraca left
hahainternet but it is a problem if i wanted to use it to compete with a statically typed language i use 12:39
timotimo perl6 is also supposed to be fun, so if you can't fun with it, that *is* a problem ;)
hahainternet nah it's still fun 12:40
the fact i can do Z↹ as an operator is superb
i just wish i could parameterise that operator, so i can really express what i intend
12:41 xinming_ is now known as xinming
timotimo right 12:42
Array gets its parameterization y having a custom method ^parameterize
hahainternet yeah and typedarray.pm
i had a good read
timotimo m: class C is Array[Int] { }
camelia rakudo-moar 4e7682: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Cannot resolve caller trait_mod:<is>(C, Array, Array); none of these signatures match:␤ (Mu:U $child, Mu:U $parent)␤ (Mu:U $child, :$DEPRECATED!)␤ (Mu:U $type, :$rw!)␤ (Mu:U $type, :$nati…»
hahainternet yeah i've tried every permutation i could think of 12:43
timotimo that also don't work :\
hahainternet providing a custom :of(Int) also doesn't
timotimo m: constant ArrayInt = Array[Int]; class C is ArrayInt { }
camelia ( no output )
timotimo there we go, that works
hahainternet which is mentioned in one of the synopses i think
that's interesting
i didn't try that obviously
timotimo not everybody can think of everything
hahainternet still, it's a little surprising and certainly a bug(?) 12:44
seatek like monkeys on typewriters ;)
hahainternet i need to get filing tbqh
just don't have very much time
i'm off to try and fit a wardrobe into a space it won't fit into now :D
timotimo that's fine. we got aware of it and maybe we'll make it work somehow
hahainternet cheers timotimo o/
timotimo cheers
i'm AFK, too
moritz m: class C does Array[Int] { } 12:54
camelia rakudo-moar 4e7682: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Array[Int] is not composable, so C cannot compose it␤at <tmp>:1␤»
skaji Hi, 13:02
I have a question.
If I use $watcher = IO::Notification.watch-path("."), how can I stop the $watcher?
moritz $watcher.done maybe? 13:04
13:05 espadrine joined
jnthn m: say IO::Notification.watch-path(".").WHAT 13:05
yoleaux 00:26Z <AlexDaniel> jnthn: hey! Turns out there was a bug in our code, so I guess you fixed everything. However, I found another issue: gist.github.com/AlexDaniel/44e5cde...6fe04b9c1d
camelia rakudo-moar 4e7682: OUTPUT«(Supply)␤»
yoleaux 00:28Z <AlexDaniel> jnthn: is it a known issue? Somehow I feel that it is. Basically, a lot of output on stdout of the process started with Proc::Async clogs up something in rakudo (well, the tap itself), and Promise.in(1) does not fire up at all because of that
jnthn So $watcher is a Supply, so it doesn't watch until you tap it. .tap gives back a Tap object, and that has a .close method on it that you use to stop watching.
13:05 sena_kun joined 13:08 seatek left 13:09 flaviusb left
skaji moritz, jnthn: thank you! 13:09
So in general, supplies themselves does not watch anything, so we don't need to "stop" them.
On the other hand, once we tap it, we must close the taps. Is this correct?
moritz yes 13:10
jnthn Indeed. NOte that if you do stuff like $watcher.map(...blah...).tap then the .close on that tap cascades up to the Watcher also 13:11
And supply/react blocks do the appropriate subscription management also
13:12 pierre_ joined, pyrimidine left
skaji moritz, jnthn: I understand. Thank you agian. 13:13
I hope it is clearly documented somewhere:) maybe in docs.perl6.org/language/concurrency.html 13:14
13:14 kalkin-_ joined
moritz skaji: if not, please add it yourself 13:15
kalkin-_ I'm currently reading the todays advent article, but it doesn't explain what Z=> is
mscha m: multi sub foo(Int $x) { state $bar = 42; say $bar; }; foo(3); # OK, but ... 13:16
camelia rakudo-moar 4e7682: OUTPUT«42␤»
moritz (to the docs, I mean)
mscha m: multi sub foo(Int $x where 1..10) { state $bar = 42; say $bar; }; foo(3); # huh?
camelia rakudo-moar 4e7682: OUTPUT«(Any)␤»
kalkin-_ I know that Z is zip, and => is used to assign pairs(?), but is Z=> zip it as kv?
13:16 kalkin-_ is now known as kalkin-
moritz kalkin-: docs.perl6.org/language/operators#...perator%29 13:17
Z as a meta operator applies the operator after it (here =>) pair-wise to the items from both lists
kalkin- Ahh this makes sense, thanks 13:18
13:19 domidumont left
skaji moritz: year, you're right. But I don't know perl6 (or moarvm) internals enough to write documentation. 13:20
moritz skaji: you don't need to know internals to write documentation
skaji: documentation are for user-visible features, and you just learned about some of them :-)
skaji: I don't want to pressure you to write docs, just wanting to argue against your imposter syndrome 13:21
13:21 cog_ joined, cognominal left
kalkin- skaji: I don't know perl6 internals either, but still was able to submit few helpful patches. If you do something wrong, it just won't be accepted and you will learn something new 13:23
13:23 pyrimidine joined
mscha I'm doing adventofcode.com/2016 with Perl 6, it rulez... Today I used the following piece of code: 13:25
$name.comb.grep(/<[a..z]>/).Bag.pairs.sort(*.key).sort(-*.value)».key[^5].join;
skaji moritz, kalkin-: thanks, I hope I will in the near future. 13:26
13:27 pierre_ left
skaji meanwhile, I must write perl6 a lot. 13:28
13:28 lucasb_ joined, pyrimidine left
kalkin- Is there a way to read the documentation with p6doc without explicetly providing the pagename each time via cmdline invocation? 13:28
kalkin- has something more like `info info` in mind
Hmm actually may be Pod::To::Info would be more useful, then reproducing info features in p6doc 13:29
lucasb_ kalkin-: you want browsable experience in the terminal? 13:30
kalkin- I want a browsable expirience which is faster then my browser somewhere 13:31
lucasb_ kalkin-: if you had the docs html site locally in your machine would help? (I'm just curious) 13:32
I think a downloadable tarball of the docs site would be interesting
kalkin- lucasb_: Configuring to use uzbl for reading docs would be helpful 13:33
i even think it would be okay for p6doc to have the abillity to follow links 13:34
And by default when invoked to display a similar welcome page like docs.perl6.org
but i guess once you have that you want to have some search feature and then you are to reproducing `info` features … 13:35
lucasb_ yeah, maybe something like pod2info is a better investiment of time 13:36
sena_kun The one interested in .tar.gz version of documentation can look at github.com/perl6/doc/issues/718 13:37
lucasb_ but then, I'm not a 'info' user... I always type 'man something'. if it doesn't exist, then I search for the rendered info HTML manual in the web
kalkin- IMHO the best way would be to have Pod::To::Info, provide some simple way for OS package maintainers to generate the info files during building the (rpm|deb|...) package
MasterDuke also, docs.perl6.org/perl6.xhtml is the docs in one page
kalkin- something like p6doc --to=Info --all --path foo/ 13:38
lucasb_ sena_kun, MasterDuke: thanks for the tips 13:39
moritz kalkin-: I think you're the first "info" user I met :-) 13:40
to me, it was always frustrating
dogbert17 hmm, the example on the IO::Notification page implies that $?FILE is the current directory, that can't be right
timotimo did you know there's a all-docs-on-one-page thing?
moritz not very intuitive, and often when the man page said "for more information, read the info", and then the info page was exactly the same as the mang page 13:41
kalkin- moritz: actually I'm not, but there're some manpages which are just more awesome as info docs because they are BIG. i.e: zsh
moritz s/mang/man/
kalkin-: why dream about a Pod::To::Info when you don't use info? :-)
kalkin- generally i think info is underrated. It's really a good tool for shipping documentation (not manpages!) 13:42
timotimo docs.perl6.org/perl6.xhtml
kalkin- moritz: i use it only for things where it worse it
moritz I think it has been superseeded by browsers, and HTML
and if your browser is laggy, fix your browser problems :-)
timotimo wow, that page is BROKEN AF
kalkin- moritz: btw I completely agree that info has confusing interface 13:43
timotimo it allows scrolling to the side and the text goes below the menu
kalkin- moritz: also sometimes i don't want to have a huge html engine + browser in my OS
lucasb_ The PDF output of Texinfo is very nice and readable 13:44
about the info browser, I wish I used it more, but I've never felt comfortable with it. I guess emacs people likes it better 13:45
the standard info2html or whatever it's called is good enought. I read the GNU manuals online generated from them 13:46
13:47 cog_ left
sena_kun Hmmm, is there a reason to mention proto methods in the docs? 13:49
timotimo kalkin-: did you see what i wrote? the one-html-page version of the docs? 13:50
13:50 bjz_ left
kalkin- timotimo: no I missed it. This is awesome 13:50
I think this will do it for 80% of use cases 13:52
moritz kalkin-: if you want to have a slim system (no browser on it), simply read the docs on your desktop or laptop 13:53
timotimo like, with less? 13:55
kalkin- moritz: our definition of slim system may differ
13:55 cognominal joined
sena_kun Who is responsible for docs.perl6.org/perl6.xhtml, by the way? 13:56
lucasb_ bigpage:
pod2onepage --threads=1 -v --source-path=./doc --exclude=404.pod6,/.git,/precompiled > html/perl6.xhtml
from Makefile
kalkin- IMHO it doesn't matter how slim your system is if a human interacts with it, he/she should be able to study and change it. this means docs + development stuff
lucasb_ I was just looking at it :)
ah, pod2onepage comes from github.com/gfldex/perl6-pod-to-bigpage 14:01
gfldex++ :)
kalkin- How can i make the repl to keep the history between invocations? Is there someway to make it write to some history file? 14:03
sena_kun kalkin-, did you install Linenoise? 14:04
kalkin- sena_kun: I use Readline instead
Because readline allows me zsh style completion
Is this the issue?
sena_kun kalkin-, if Readline doesn't support history, then yes. Linenoise gives you history among other features. 14:05
kalkin- I'm talking about part of history completion, which Linenoise lacks
like if you start with say press Up and it will only complete the elements starting with say 14:06
Linenoise AFAIK provides only simple prior/next history completion
sena_kun kalkin-, hm, sorry. Ah, I understood. No, Linenoise can't do that.
On the other hand, it should be not so hard to write. 14:07
On the up arrow press you need just take input and grep history with it as a regex.
Except that linenoise is just a C library, of course. 14:08
kalkin- I think it wouldn't be hard to add history persistence to Readline either if it's missing it(?) :)
sena_kun: you probably could wrap it at the perl6 level couldn't you? 14:09
sena_kun github.com/drforr/perl6-readline/issues/4 - almost a year passed.
kalkin-, dunno, I've never tried that. 14:10
14:11 pierre_ joined
dalek c: adb12e6 | (Siavash Askari Nasr)++ | doc/Language/typesystem.pod6:
"handles" trait accepts Pairs
 14:15
synopsebot6 Link: doc.perl6.org/language/typesystem
14:17 wamba left 14:23 rurban joined 14:32 cognominal left
dalek c: 0d5619b | Altai-man++ | doc/Type/Cool.pod6:
Add missing semicolons
 14:43
synopsebot6 Link: doc.perl6.org/type/Cool
14:43 ZzZombo joined 14:47 ufobat left 14:48 pierre_ left
RabidGravy in P5 news, I got a bug report against Term::Screen in 2010, I replied that I wouldn't fix it because the approach to processing the events was wrong and they should use POE, the same guy just reported the same bug 14:54
15:01 CIAvash joined 15:04 rurban left, rurban joined, rurban left 15:05 cognominal joined 15:09 pyrimidine joined 15:14 pyrimidine left
RabidGravy gah, this is a really awful module how come I'm still supporting it after nearly twenty years 15:22
dalek c/small-example-fixes: 7d946fe | Altai-man++ | doc/Type/ (8 files):
Small bunch of changes to make examples compileable.

1)Missed semicolons; 2)Typos; 3)Endless code without :skip-test; 4)Convert flags table to actual table, not code.		 15:25
15:26 zacts left 15:27 CIAvash left 15:28 zacts joined 15:33 mscha left 15:35 lucasb_ left 15:40 sena_kun left 15:46 travis-ci joined
travis-ci Doc build passed. Altai-man 'Small bunch of changes to make examples compileable. 15:46
travis-ci.org/perl6/doc/builds/181129126 github.com/perl6/doc/commit/7d946fe8cab7
15:46 travis-ci left
dalek c: 7d946fe | Altai-man++ | doc/Type/ (8 files):
Small bunch of changes to make examples compileable.

1)Missed semicolons; 2)Typos; 3)Endless code without :skip-test; 4)Convert flags table to actual table, not code.		 15:49
c: d5b10eb | RabidGravy++ | doc/Type/ (8 files):
Merge pull request #1034 from perl6/small-example-fixes

Small bunch of changes to make examples compileable.
15:49 pierre_ joined 15:54 pierre_ left 15:56 wamba joined 15:57 AlexDaniel joined 16:02 Actualeyes left 16:03 complain_ left 16:07 rindolf left 16:12 rindolf joined 16:16 pyrimidine joined 16:20 pyrimidine left 16:21 khw joined 16:46 domidumont joined 16:49 pyrimidine joined 16:53 BenGoldberg joined 16:54 pyrimidine left 17:10 cdg joined, araujo left 17:11 rburkholder left 17:12 araujo joined, araujo left, araujo joined, skids joined 17:18 cdg left 17:28 sena_kun joined 17:33 James0r left
babydrop CIAvash, sure sounds like it. Thanks. 17:35
17:37 dogbert17 left 17:38 dogbert17 joined
AlexDaniel sena_kun: “Because we can't automatically compile and check this example otherwise?” – what is the error? 17:39
sena_kun AlexDaniel, let's see...
17:40 ufobat joined 17:41 CIAvash joined
sena_kun AlexDaniel, you're totally right here, my bad. It actually compiles ok even without semicolon. Though now I somewhat don't like that `sub f(){ 17:41
Argh.
}` without whitespace, but it's okay.
AlexDaniel so the semicolon has to go… :) 17:42
by the way, has anybody thought about changing regular single quotes in the docs to ‘’ ?
sena_kun AlexDaniel, if you're insisting on removing, just do it! ;) 17:43
moritz AlexDaniel: please don't (changing quotes)
AlexDaniel moritz: I'm not saying that we should. But if not, why not?
(yet, at least)
moritz AlexDaniel: because people copy&paste, don't see that they are special, and try to close them with an ordinary single quote
and then get confused about the error message 17:44
AlexDaniel moritz: we can improve the error message
moritz or they have a step somewhere that doesn't handle UTF-8 right (like email)
AlexDaniel moritz: at the same time, beginners often copy&paste into perl6 -e '' and get confused too
moritz AlexDaniel: that's a valid point, I concede 17:45
AlexDaniel well that's how I got the idea
moritz AlexDaniel: though if we change the compiler now to give a better error message, we should wait about half a year before changing the docs on a large scale
AlexDaniel moritz: that's OK, I am not saying that we should rush. In fact, I'm not even saying that something should be changed, just trying to see what other folks think 17:46
moritz AlexDaniel: the traditionalist inside my chest is very much against it. The pragmatist is convinced by the -e '...' argument :-)
AlexDaniel m: say ‘hello world'; say 42 17:47
camelia rakudo-moar 6f6e6d: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Unable to parse expression in smart single quotes; couldn't find final "’" ␤at <tmp>:1␤------> 3say ‘hello world'; say 427⏏5<EOL>␤ expecting any of:␤ argument list␤ smart…»
17:48 cognominal left
AlexDaniel “smart quotes”? 17:48
moritz smart single quotes even 17:49
AlexDaniel isn't it “curly quotes”?
moritz .u ‘
yoleaux U+2018 LEFT SINGLE QUOTATION MARK [Pi] (‘)
moritz .u “
yoleaux U+201C LEFT DOUBLE QUOTATION MARK [Pi] (“)
AlexDaniel I think “smart” stands for the feature in some software that changes ' to ‘’ 17:50
and has nothing to do with this case
moritz it's how people tend to get them in first place :-)
17:51 pierre_ joined
AlexDaniel I type them myself and am pissed off by the message :P 17:51
babydrop
.oO( typographic quotes... )
AlexDaniel eval: say ‘hello world'; say 42
evalable6 AlexDaniel, rakudo-moar 6f6e6db: OUTPUT«(exit code 1) ===SORRY!=== Error while compiling /tmp/PvVi5swjgZ␤Unable to parse expression in…»
AlexDaniel, Full output: gist.github.com/d833afa6093c21559b...c85205def5
sena_kun AlexDaniel, personally, I somewhat dislike it. Not like I hate UTF, but after such change *every* person who commites to the docs will have to use exactly this quote character. It is a new work, test, explanations to the contributors for almost no benefit. I can be wrong, surely. 17:52
AlexDaniel sena_kun: what if some examples will use '' and some ‘’? 17:53
sena_kun: I mean, what do you think about this outcome?
17:53 espadrine left, rurban joined
sena_kun AlexDaniel, this is how non-determinism was born! 17:53
17:54 cognominal joined
AlexDaniel sena_kun: I am asking because those who can type ‘’ actually have to use “exactly this quote character” right now anyway, don't they? 17:54
17:54 CIAvash left
AlexDaniel I do agree that there are more people who would much rather use ', but still… 17:54
but yes, it will require a lot of work, you are right 17:55
17:55 pierre_ left
TimToady the quotes aren't actually curly in my font, just slanted 17:55
sena_kun AlexDaniel, majority/minority. Of course, we don't have a place to hide people who can type ‘’, but... You know. 17:56
babydrop TimToady: what does this line do? Git blame blames you: github.com/rakudo/rakudo/blob/nom/...re.pm#L112 17:57
17:57 CIAvash joined
babydrop m: say sub ($,$,$,$,$) {}.signature 17:57
camelia rakudo-moar 6f6e6d: OUTPUT«($,,,,)␤»
babydrop It causes this bug, but I don't understand what it's meant to do in there...
17:59 dwarring joined
moritz babydrop: probably shorten (Int $) to (Int) 17:59
babydrop m: say sub (Int $, Str $, Int @, Str %,$) {}.signature 18:00
camelia rakudo-moar 6f6e6d: OUTPUT«(Int, Str, Int @, Str %,)␤»
babydrop Makes sense. Thanks.
TimToady maybe the pattern needs a » before the space 18:01
babydrop Seems to do the trick. Thanks, TimToady++ 18:02
TimToady or exempt Any $ maybe
18:02 CIAvash left 18:03 cognominal left
AlexDaniel m: sub foo (--> -42) {} 18:05
camelia rakudo-moar 6f6e6d: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Malformed return value␤at <tmp>:1␤------> 3sub foo (-->7⏏5 -42) {}␤»
18:06 cognominal joined
rindolf Hi all. 18:09
18:09 mst left, mst joined
babydrop \o 18:11
AlexDaniel m: say #`🤖 42
camelia rakudo-moar 6f6e6d: OUTPUT«5===SORRY!5===␤Argument to "say" seems to be malformed␤at <tmp>:1␤------> 3say7⏏5 #`🤖 42␤Other potential difficulties:␤ Unsupported use of bare "say"; in Perl 6 please use .say if you meant $_, or use an explicit invocant or argum…»
AlexDaniel aww
babydrop u: 🤖 18:12
unicodable6 babydrop, U+1F916 ROBOT FACE [So] (🤖)
TimToady on Day 4, Damian is spelled with an 'a', not an 'e' 18:17
babydrop Fixed. Thanks. 18:19
geekosaur movie flashbacks? :p 18:22
18:26 CIAvash joined 18:31 stevieb left
dalek c: 6379dff | paultcochrane++ | doc/Type/IO/Path.pod6:
Fix a genitive

  ... and the placement of a full stop.		 18:31
synopsebot6 Link: doc.perl6.org/type/IO/Path
18:32 pyrimidine joined 18:33 cognominal left 18:34 firstdayonthejob joined, cognominal joined 18:37 pyrimidine left
dalek c: 189fb85 | paultcochrane++ | doc/Type/IO/Path.pod6:
Format code-like textual elements with C<>
 18:37
synopsebot6 Link: doc.perl6.org/type/IO/Path
18:37 CIAvash left
dalek c: 6401a49 | (Aleks-Daniel Jakimenko-Aleksejev)++ | doc/Type/WrapHandle.pod6:
There is no need for a ; here

Also add a space character before the block		 18:38
synopsebot6 Link: doc.perl6.org/type/WrapHandle
japhb I don't know if this is still an active discussion, but FWIW I'd be happy to see all 'returns' in the docs replaced by -->. I think the latter is a much clearer way to represent the routine's "interface" in a unified way, though perhaps that's biased by time looking at Haskell.
[ptc] japhb++ 18:41
is it possible to concatenate IO::Path objects in a platform-independent way? 18:42
geekosaur I actually feel like it's in the wrong place (it looks like part of the formal parameters; Haskell deliberately fuzzes that, but Perl 6 doesn't have its excuse for doing so)
[ptc] for instance, if one has a dir as an IO::Path and wants to add another dir to the end of that, is there a platform independent way of doing that 18:43
or can I just use Unix slashes and Perl6 works the rest out for me?
geekosaur docs.perl6.org/type/IO$COLON$COLON...thod_child 18:44
[ptc] geekosaur: brilliant! Thanks! Have been going through those docs but didn't use the right search term :-)
geekosaur yeh, "child" is a slightly weird name for that
AlexDaniel japhb: it is an active discussion, kinda. Thanks for your comment 18:45
geekosaur I mean, it sorta makes sense? but it's not where I would look for that; and "child" makes me think it's an iterator for the children of the current IO::Path
(except that's weird too since it should be plural in that case...) 18:46
AlexDaniel japhb: if you were a begginner, wouldn't you find --> notation a bit confusing? Or is it not a problem?
[ptc] geekosaur: yeah, I was looking for "join" and "concat" and stuff. I guess "child" makes sense in a broader sense, but it's not where I'd instinctively look... 18:47
18:48 cibs left
AlexDaniel japhb: I mean, we are going to replace ‘returns’ with --> anyway, but I'm just trying to understand what would be the best way to explain this to newcomers 18:49
japhb AlexDaniel: I read --> as 'yields', and :( as 'takes', so I mentally read a signature like :(Int x, Int y --> Int) as 'takes an Int x and Int y and yields another Int'
I don't know if I would have read it the same way as a beginner, but I think it would be really easy to teach 18:50
18:50 cibs joined
japhb But that also makes me think: The error messages use -->, so people will be seeing it a lot early on anyway. 18:50
ufobat is it possible to have a junction of certain type? like it should only contain Integers > 10 or something? what would be its signature? 18:51
[ptc] the --> is fairly clear; I'd guess that the "returns" could very easily get confused with "return" for newcomers 18:52
AlexDaniel japhb: all arguments are in favor of -->, the only exception is ‘returns’ being “more readable” to those who are unfamiliar with the syntax
and some will probably argue with that one too :)
babydrop ufobat: sounds like you're abusing junctions :) 18:53
AlexDaniel ufobat: why?
babydrop ufobat: there's a way by poking in the guts, but you're not meant to do that 18:54
18:54 darutoko left
babydrop m: sub (Junction $ where $_ > 10) { say "feels good" }(9&11) 18:55
camelia rakudo-moar 219f52: OUTPUT«Constraint type check failed for parameter '<anon>'␤ in sub at <tmp> line 1␤ in block <unit> at <tmp> line 1␤␤»
babydrop m: sub (Junction $ where $_ > 10) { say "feels good" }(9|11)
camelia rakudo-moar 219f52: OUTPUT«feels good␤»
babydrop ufobat: ^ junctions are different, so "contains" is kinda a tricky question
ufobat maybe i am not wanting junctions, i dont know. i am thinking of routes in bailador.. i'd like to have a route that could be <GET PUT>.any but it should not allow http method HMPF
AlexDaniel hahaha
u: sub (Junction[Int] $) { say "feels good~" }(9|11) 18:56
unicodable6 AlexDaniel, U+0073 LATIN SMALL LETTER S [Ll] (s)
AlexDaniel, U+0075 LATIN SMALL LETTER U [Ll] (u)
AlexDaniel, U+0062 LATIN SMALL LETTER B [Ll] (b)
AlexDaniel, gist.github.com/a7a5f6d20761210aae...8ab4664876
AlexDaniel oops
eval: sub (Junction[Int] $) { say "feels good~" }(9|11)
evalable6 AlexDaniel, rakudo-moar 219f527: OUTPUT«(exit code 1) ===SORRY!=== Error while compiling /tmp/2CfofqOAF9␤An exception occurred while p…»
AlexDaniel, Full output: gist.github.com/07f6953112d6b8fcd9...7c1c28ba59
AlexDaniel SORRY! SORRY!
babydrop ufobat: I don't get where the 10> ints come in 18:57
18:57 shadowpaste joined
ufobat it whats just an example 18:58
babydrop m: multi dispatch ('GET') { say "getting" }; multi dispatch ('PUT') { say "putting" }; dispatch 'GET'|'PUT'
camelia rakudo-moar 219f52: OUTPUT«getting␤putting␤»
18:59 zamanf_ left
ufobat my real world problem would be nopaste.me/view/4499b757 instead of a single HTTP method a junction of HTTP methods 18:59
babydrop m: sub route ($where) { say so $where eq 'PUT' }; route 'GET'|'PUT'
camelia rakudo-moar 219f52: OUTPUT«False␤True␤»
babydrop m: sub route (Mu $where) { say so $where eq 'PUT' }; route 'GET'|'PUT' 19:00
camelia rakudo-moar 219f52: OUTPUT«True␤»
babydrop m: sub route (Mu $where) { say so $where eq 'GET' }; route 'GET'|'PUT'
camelia rakudo-moar 219f52: OUTPUT«True␤»
ufobat brb, sorry - children need attention
babydrop m: sub route (Mu $where) { say so $where eq 'HMPF' }; route 'GET'|'PUT'
camelia rakudo-moar 219f52: OUTPUT«False␤»
babydrop doesn't see the issue
Ah 19:01
Actually no Ah 19:02
ufobat no?
a few days ago i was told i am rather bad in explaining my problems *g* 19:03
babydrop ufobat: just put `of Mu` rather than `of Str` on the subset and it'll take junctions 19:04
m: class { subset HttpMethod of Mu where * eq any <GET PUT POST HEAD PUT DELETE TRACE OPTIONS CONNECT PATCH>; has HttpMethod $.method; }.new: :method('GET'|'PUT')
camelia ( no output )
babydrop and you can add the type check inside the `where` if you really cared :)
ufobat: it's just a classical XY Problem 19:05
huggable: xy
huggable babydrop, nothing found
babydrop huggable: xy problem
huggable babydrop, You want to do X, but don't know how. You think you can solve it using Y, but don't know how to do that, either. So, you ask about Y in order to solve X, which doesn't make sense. You should ask about X.
ufobat yeah, that is true
babydrop Like, I still have no idea what issue you're having. I can only assume your subset doesn't accept a Junction based on the scanty 2 lines of code you pasted
ufobat you're guessing it right 19:06
babydrop \o/
ufobat thanks for the help :-)
babydrop Any time.
19:15 Cabanossi left, Cabanossi joined 19:18 cognominal left 19:25 cognominal joined
babydrop u: arabic digit 1 19:32
unicodable6 babydrop, U+0061 LATIN SMALL LETTER A [Ll] (a)
babydrop, U+0072 LATIN SMALL LETTER R [Ll] (r)
babydrop, gist.github.com/598f30cc07fe79d215...12c3e0ba62
babydrop u: indian digit 1
unicodable6 babydrop, U+0069 LATIN SMALL LETTER I [Ll] (i)
babydrop, U+0064 LATIN SMALL LETTER D [Ll] (d)
babydrop, U+006E LATIN SMALL LETTER N [Ll] (n)
babydrop, gist.github.com/778c59d3031ab0d6f1...45622ec2ae
babydrop AlexDaniel: why did you break your bot!
ohhh
AlexDaniel one
babydrop u: arabic digit one
unicodable6 babydrop, U+0661 ARABIC-INDIC DIGIT ONE [Nd] (١)
babydrop, U+06F1 EXTENDED ARABIC-INDIC DIGIT ONE [Nd] (۱)
19:33 cognominal left
AlexDaniel I wonder what yole has to say to this 19:33
yoleaux: arabic digit 1
.u arabic digit 1
yoleaux No characters found
AlexDaniel ?
what did it try? 19:34
.u 80
yoleaux U+E013F VARIATION SELECTOR-80 [Mn] (◌󠄿)
19:34 cognominal joined
AlexDaniel oooooh 19:34
babydrop m: :35<Z>
camelia rakudo-moar 219f52: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Invalid base-35 character 'Z': 7⏏5Z. Please use one of 0..9, A..Y␤at <tmp>:1␤------> 3:35<Z>7⏏5<EOL>␤»
babydrop I guess there's no way to trigger that error for radix 36? 19:35
AlexDaniel u: 80
unicodable6 AlexDaniel, U+0038 DIGIT EIGHT [Nd] (8)
AlexDaniel, U+0030 DIGIT ZERO [Nd] (0)
AlexDaniel m: say ‘Z’.succ
camelia rakudo-moar 219f52: OUTPUT«AA␤»
AlexDaniel m: say ‘Z’.ord.succ.chr
camelia rakudo-moar 219f52: OUTPUT«[␤»
babydrop m: try ':35<♥>'.EVAL; say $!.^name
camelia rakudo-moar 219f52: OUTPUT«X::Syntax::Malformed␤»
19:36 rburkholder joined
babydrop kinda annoying we have like 3 different error messages for what is basically one thing 19:36
AlexDaniel m: say :35<[>
camelia rakudo-moar 219f52: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Malformed radix number␤at <tmp>:1␤------> 3say :35<7⏏5[>␤ expecting any of:␤ number in radix notation␤»
babydrop m: try q|:35('♥')|.EVAL; say $!.^name
camelia rakudo-moar 219f52: OUTPUT«X::Str::Numeric␤»
babydrop m: try q|:35<Z>|.EVAL; say $!.^name
camelia rakudo-moar 219f52: OUTPUT«X::Syntax::Number::InvalidCharacter␤»
babydrop weird eh
19:37 girafe joined 19:38 bjz joined
AlexDaniel 39 open whateverable issues… should I slap “Decemberfest” label on all of them? :) 19:39
some free labour won't hurt for sure
babydrop wtf is Decemberfest
AlexDaniel babydrop: it is a monthly even organized by AlexDaniel. Submit 3 PRs in whateverable repo and receive a free “thank you” 19:42
event*
babydrop heh 19:43
19:47 sena_kun left 19:48 pyrimidine joined 19:49 girafe2 joined 19:50 pyrimidine left, pyrimidine joined
pmurias AlexDaniel: what's a whateverable issue? 19:50
geekosaur "whateverable" is the bot framework for evalable/unicodable/committable/bisectable/etc. 19:51
AlexDaniel pmurias: nothing in particular, just a lot of stuff to be done: github.com/perl6/whateverable/issues
19:52 girafe left 19:53 pyrimidine left 19:56 espadrine joined
pmurias AlexDaniel: shouldn't only the LHF be marked Decemberfest? 20:03
20:04 sivoais left
AlexDaniel that's what I thought too. Turns out not necessarily 20:04
babydrop :/
pmurias AlexDaniel: are you making fun of Hacktoberfest? ;) 20:08
20:08 lizmat left
AlexDaniel pmurias: just looking for free labour :) But yes, a bit 20:09
20:10 lizmat joined, firstdayonthejob left
pmurias likes the Hacktoberfest concept 20:11
20:11 knut_ joined 20:13 sivoais joined 20:17 cxreg left 20:18 cognominal left 20:19 rurban left
knut_ Hi all. I'm not much of a programmer, but I've used perl5 in the past. In the last two years I tried perl6 on some occasions. Everytime I stumbled about map and $_. When used as a method one cannot use $_, the same is true for map as function+expression. But inside a block it works. It took me very long to figure out that this is actually a feature of the block 20:20
20:20 cognominal joined
knut_ so maybe this could be documentet in some perl5to6 guide. I found it very confusing 20:20
moritz knut_: maybe send a pull request? the docs are on github 20:21
AlexDaniel knut_: by the way, do you know about * ?
moritz knut_: I'll be happy to review it if you make one
20:21 cxreg joined
AlexDaniel m: say <10 20 30>.map(*.succ) 20:21
camelia rakudo-moar fa82a1: OUTPUT«(11 21 31)␤»
AlexDaniel ↑ no block required
babydrop knut_: there are a number of 5-to-6 docs: github.com/perl6/doc/tree/master/doc/Language
20:22 telex left
moritz raw.githubusercontent.com/perl6/do...lfunc.pod6 20:22
this one talks about map, specifically
geekosaur but I note the doc for $_ in 5to6-variables doesn't mention its scope changed from "global" to lexical/block
babydrop knut_: pick the best one to add it to. Perhaps a general note that unlike Perl 5, we don't make $_ available in map/grep/etc without a block
20:22 ab6tract left
knut_ AlexDaniel: I know about * and found it even more confusing in the beginning ;) 20:22
geekosaur which is a pretty big change for someone used to perl 5's $_ 20:23
(granted, p5 did experimentally try to do the same thing... and broke pretty much everything and reverted it)
AlexDaniel knut_: good :) I think you are a perfect person to improve 5to6 docs, your contributions will be appreciated :)
babydrop knut_: to add to confusion, some things "thunk", so $_ *is* available without a block, like in a where clause: subset of Int where $_ %% 2; works
20:24 domidumont left, dataangel left
moritz geekosaur: I think the real difference is that in p5, you can write map $_ + 1, @array and it automatically captures the $_ + 1 as a callabl 20:24
(which is kinda spooky)
20:25 firstdayonthejob joined
AlexDaniel babydrop: so all ｢$_ %% 2｣, ｢{ $_ %% 2 }｣ and ｢* %% 2｣ are allowed… wow 20:25
moritz right, but not for a map 20:26
masak moritz: well, Perl 6 sort of trains one not to accept blocks that are "too implicit" 20:29
mst perl6 achieves the same goal with whatever star expressions, no?
map *+1, @foo; etc. 20:30
20:30 SoylentGreenSqui joined
mst which is way way neater than a special case in the parser (that I can't access via prototypes *grump*) 20:30
moritz mst: right, but there a limits to them; for example you can't use them in argument lists
mst moritz: hrm?
SoylentGreenSqui starts a heated diskussion by asking, "what editor to use for perl6?" 20:31
moritz mst: foo(*) doesn't become { foo($_) }
geekosaur and the * kinda alerts you to magic going on, unlike $_ which is everywhere in p5 and you can't easily tell which ones are "magical"
moritz SoylentGreenSqui: vim. No heat necessary :-)
AlexDaniel SoylentGreenSqui: emacs works OK with perl6 if you use install perl6-mode
mst moritz: right, the subroutine needs to know it's getting WhateverCode
AlexDaniel SoylentGreenSqui: atom shuold have good support for perl6 now
mst moritz: foo(*) is similar to foo({ $_ }) no?
moritz mst: no, it's a straight call to &foo, passing in a Whatever instance 20:32
it doesn't autocurry
or whatever you call it :-)
20:32 firstdayonthejob left
AlexDaniel SoylentGreenSqui: vim works too, but if I recall correctly it does not support fancy unicode quotes and a bunch of other stuff, but it should be quite usable 20:32
SoylentGreenSqui: but why do you ask?
mst moritz: ah, you have to special case Whatever? 20:33
AlexDaniel m: say <10 20 30>».succ
camelia rakudo-moar fa82a1: OUTPUT«(11 21 31)␤»
mst moritz: foo(*+1) is similar to foo({ $_ + 1 }) still though, right?
AlexDaniel knut_: there is also this ↑
SoylentGreenSqui to find out which one to use ^^ 20:34
AlexDaniel SoylentGreenSqui: what editor do you normally use?
SoylentGreenSqui Notepadd++
AlexDaniel :|
moritz mst: yes
SoylentGreenSqui But I just set up a linux system
AlexDaniel SoylentGreenSqui: perhaps try atom? 20:35
mst moritz: is there an idiom for accepting that?
moritz mst: *.&foo
mst moritz: oh, and then foo gets a whatevercode? 20:36
timotimo not gets, becomes
SoylentGreenSqui AlexDaniel: Looks nice.
mspo m: say blob16.new([1, 2, 3]).bytes; 20:37
camelia rakudo-moar fa82a1: OUTPUT«6␤»
20:37 telex joined
mspo where is that blob16 documented as a thing? 20:37
also why are pack/unpack experimental?
AlexDaniel SoylentGreenSqui: I am suggesting it because emacs/vim would probably be a bit too huge of a leap from Notepad++. Feel free to go for it though, if you feel ready :)
mspo: perhaps it is related to this issue: github.com/perl6/doc/issues/372 20:38
SoylentGreenSqui I also found "Visual Studio Code" which looks promising. 20:39
AlexDaniel mspo: github.com/rakudo/rakudo/pull/649 20:40
20:42 Guest42590 joined 20:46 firstdayonthejob joined 20:48 cognominal left, ufobat left
babydrop laughs 20:49
AlexDaniel: you're pissed off at the message? :) 20:50
AlexDaniel: why not submit a PR? :) git clone github.com/rakudo/rakudo .; grep -ER 'dba.+smart' src/
pmurias SoylentGreenSqui: a bunch of perl 6 guys seem to use atom (I don't use it because it's not a vi)
20:51 cognominal joined
babydrop wouldn't mind s/[low|high] s+ smart// 20:53
20:54 pyrimidine joined, Guest42590 left
dalek c: f19e6fc | Altai-man++ | doc/Type/Int.pod6:
Fix of example

tommystanton++		 20:58
synopsebot6 Link: doc.perl6.org/type/Int
21:00 FROGGS left
AlexDaniel babydrop: I like that you enjoy my tickets xD 21:05
babydrop: I believe it would also require changing some tests
babydrop: feel free to mark it as LHF ;)
or… maybe I'll do it 21:06
21:07 bjz left
babydrop If these are 6.c-errata tests, then we can't change them (if they specifically test for these names). It'd have to wait until 6.d 21:09
yoleaux AlexDaniel: harass mst 21:12
AlexDaniel what's the idea behind 6.c-errata by the way? 21:16
timotimo if we discover tests to have been wrong, we fix them there 21:18
21:18 bjz joined
babydrop AlexDaniel: 6.c is our promise to the users that the tested behaviour will remain unchanged in 6.c language. In a perfect world, those tests would remain frozen forever. In reality, some test are later found to be broken in one way or another or they rely on some untested feature to perform their test (like the test checking the stacktrace is printed twice by checking for a specific core filename that later 21:24
gets renamed) and they need to fixed, that's where 6.c-errata comes in.
AlexDaniel but new tests are added to 6.c directly? 21:25
oops, nope, it's just a master branch
ok
so there is 6.c tag and 6.c-errata branch, okay, it makes sense 21:26
or maybe not. Not sure what's the point of this tag.
pmurias AlexDaniel: what would be the benefit of changing the docs to use smart quotes?
AlexDaniel pmurias: so that people can copy-paste stuff into perl6 -e '' 21:27
21:28 bjz left
AlexDaniel babydrop: yea, I'm still confused, I don't understand for what purpose 6.c tag was created 21:28
pmurias: and these are curly quotes, by the way… 21:29
AlexDaniel works on a PR 21:30
pmurias AlexDaniel: for the docs?
babydrop AlexDaniel: ask the person who created it
geekosaur presumably to represent the tests as released with 6.c, to provide a reference for 6.c-errata
babydrop geekosaur: it was created in Feb 29 tho
*19
geekosaur yes. I think -errata was created at the same time
babydrop Ah 21:31
geekosaur before that nobody had considered that an errata branch might be needed
babydrop Makes sens.
AlexDaniel so 6.c tag is “here is what we promised, but nevermind, just use 6.c-errata branch”
babydrop 6.c is the release 6.c-errata is the errata to the release. 21:32
babydrop sees no issue
geekosaur neither
specs can have bugs, and in fact often do
geekosaur was marginally involved in finding a bug in the haskell '98 spec...
pmurias AlexDaniel: I don't think making copy-pasting stuff into perl6 -e easier is worth the potential confusion to a newbie who isn't aware that curly quotes work in Perl6/doesn't know how to type them etc. 21:33
dalek c: bbcafde | gfldex++ | CREDITS:
add username to credits
babydrop huh... so we have two CREDITS :/ 21:34
gfldex yes :)
we don't like it easy :->
also, there is no shame in adding yourself to the CREDITS of the docs
MasterDuke maybe this is a bad idea, but what about a JS button to swap between regular (e.g., ' and ") and fancy (e.g., <i can't type them>)?
babydrop gfldex: may be worth teaching this script to parse it then: github.com/rakudo/rakudo/blob/nom/...butors.pl6 21:35
MasterDuke: that's featuritis.
babydrop cues in www.wechall.net/ as experts in featuritis
geekosaur (mind, re h'98, about half of the then active readership of haskell-cafe could make the same claim, since building the example code triggering it was a collaborative effort on the mailing list :)
gfldex babydrop: if I read the code right it does so already, if the docs are in '../doc' when it is executed 21:36
babydrop gfldex: just Rakudo's. It does expect the doc repo clone to exists, but only so it could run `git log --since` in it 21:37
gfldex are you sure? 21:38
it's seems to .map over @repos and complain if the path doesn't exist
babydrop Yes, that checks that the repo clone exists.
And this line loads Rakudo's CREDITS file: github.com/rakudo/rakudo/blob/nom/...rs.pl6#L36
I don't see it being changed to anything else later on. 21:39
gfldex it's using CREDITS_FILE for mapping nick to name and only for that 21:40
21:41 notostraca joined, telex left
babydrop Yes, and if your contributor is in just the doc repo... it won't get the nick/name for them. 21:41
Never mind. Ignore me. I'll fix it on next release.
gfldex ok 21:42
21:47 bjz joined
gfldex babydrop: the script will choke on any commit message that contains a ‚|‘. maybe the following is a fix (braincompiled) $msg = $line.split(|)[4..*].join('|') 21:47
21:52 pierre_ joined 21:53 bjz left 21:56 bjz joined, bjz left, pierre_ left 21:57 telex joined 21:58 Tonik left
babydrop gfldex: I don't see the breakage. There's a 4 in the .split's limit 21:59
21:59 RabidGravy left
babydrop m: dd "Zoffix Znet|Zoffix Znet|fa82a1f85ad7064614d038a993c7fb15fa133a6f|Fix crash with| X::Syntax::Number::InvalidCharacter w|hen r|adix is 11".split: '|'4 22:00
camelia rakudo-moar fa82a1: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Two terms in a row␤at <tmp>:1␤------> 3Character w|hen r|adix is 11".split: '|'7⏏054␤ expecting any of:␤ infix␤ infix stopper␤ postfix␤ statement end␤ …»
babydrop m: dd "Zoffix Znet|Zoffix Znet|fa82a1f85ad7064614d038a993c7fb15fa133a6f|Fix crash with| X::Syntax::Number::InvalidCharacter w|hen r|adix is 11".split: '|', 4
camelia rakudo-moar fa82a1: OUTPUT«("Zoffix Znet", "Zoffix Znet", "fa82a1f85ad7064614d038a993c7fb15fa133a6f", "Fix crash with| X::Syntax::Number::InvalidCharacter w|hen r|adix is 11").Seq␤»
babydrop The courtesy ++s should be expanded to be searched in bodies too :/ at least that's where I always put them :\ 22:01
22:07 bjz joined
gfldex i c 22:10
i need a better braincompiler
22:12 bjz left 22:20 _ramix_ joined
_ramix_ Hi! I have just successfully added travis build tests to the modules.perl6.org/dist/FileSystem:...olumesInfo and it appears as not configured. How much waiting time is there more or less until the travis status is updated? Thanks! 22:24
Now the status is unknow! 22:25
22:28 gfldex left, gfldex joined, robertle left 22:29 Vynce joined
babydrop checks it out 22:29
gfldex _ramix_: your build is shown as passing on my side
_ramix_ test here: modules.perl6.org/#q=filesystem%3A%3Acapacity 22:31
gfldex the modules page is updated 3x per hour (IIRC)
_ramix_ ok, thanks! 22:32
22:32 telex left
babydrop _ramix_: it works fine on a local build, so I'd say just wait for next rebuild. It's trigged each our at 20m and 40m and takes about 1 hour 22:32
*each hour 22:33
_ramix_ Perfect! Thanks a lot!
22:33 bjz joined 22:34 cognominal left
timotimo wow, an hour? 22:37
we might want to improve that somehow :)
22:38 bjz left 22:39 cognominal joined
babydrop timotimo: it'd done so on purpose, to comply with GitHub's API request limits. 22:43
s/it'd/it's/; # typing isn't my thing today it seems 22:44
22:44 _ramix_ left, synopsebot6 left, synopsebot6 joined
geekosaur github has rather draconian limits, yeh 22:44
timotimo oh, of course 22:45
we're accessing github api?
or is that also for raw.githubusercontent.com?
babydrop API
no idea about raw.blah
The API is for stars, last commit time... 22:47
The code's here: github.com/perl6/modules.perl6.org.../GitHub.pm
22:51 Herby_ joined
Herby_ o/ 22:51
babydrop \o 22:52
22:57 hipek8 joined
timotimo o/ 22:57
BenGoldberg m: print Bool.pick ?? "\r\\o " !! "\r o/" for ^40; 23:00
camelia rakudo-moar fa82a1: OUTPUT«
timotimo why isn't it animated? :< 23:04
AlexDaniel eval: print Bool.pick ?? "\r\\o " !! "\r o/" for ^40;
evalable6 AlexDaniel, rakudo-moar fa82a1f: OUTPUT«␍ o/␍\o ␍\o ␍ o/␍ o/␍ o/␍ o/␍ o/␍\o ␍\o ␍ o/␍\o ␍\o ␍ o/␍ o/␍ o/␍\o ␍\o ␍\o ␍ o/␍ o/␍\o ␍ o/␍\o ␍\o ␍\o ␍ o/␍\o ␍ o/␍ o/␍\o ␍ o/␍ o/␍ o/␍ o/␍\o ␍ o/␍ o/␍\o ␍ o/»
23:08 hipek8 left 23:09 telex joined
AlexDaniel m: say “test\b\btest” 23:10
camelia rakudo-moar fa82a1: OUTPUT«testtest␤»
AlexDaniel eval: say “test\b\btest”
evalable6 AlexDaniel, rakudo-moar fa82a1f: OUTPUT«test␈␈test»
timotimo AlexDaniel: you do know about ("\\o ", "o/ ").roll(40), right? 23:12
m: say ("\\o ", "o/ ").roll(40) 23:13
camelia rakudo-moar fa82a1: OUTPUT«(o/ \o \o o/ o/ o/ o/ \o \o o/ \o \o o/ o/ o/ \o \o o/ \o o/ \o o/ \o \o \o o/ \o \o \o \o \o o/ \o \o \o \o o/ \o o/ \o )␤»
AlexDaniel timotimo: what's the point?
timotimo it's a lot shorter? :P
AlexDaniel timotimo: ah, BenGoldberg ↑
I was just wondering whether evalable will handle it better 23:14
23:15 Vynce left 23:18 hipek8 joined 23:19 cognominal left 23:20 Vynce joined 23:23 cognominal joined 23:30 pmurias left 23:33 rindolf left 23:35 lukaramu left 23:39 Vynce left 23:40 firstdayonthejob left 23:48 knut_ left 23:53 Actualeyes joined 23:54 girafe2 left 23:57 djbkd joined 23:59 grumble 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!