tbrowder hi devs, if i want to try out an old rakudo release how do i find the one named "2017.04.3-275-g84502dc"? it's not a tag on the nom branch. 01:34
[Coke] I believe this was answered elsewhere, but that's commit 84502dc 01:43
so just 'git checkout 84502dc' 01:47
tbrowder thanks, [Coke] 01:54
Zoffix You can check out the full thing too. `git checkout 2017.04.3-275-g84502dc` and that'll work even when there are more than one commits with the same shorthash 02:07
geekosaur oh the full one does work. I had wondered but not enough to test it :p 02:08
Zoffix c: 2017.04.3-275-g84502dc say $*PERL.compiler.version 02:39
committable6 Zoffix, ¦2017.04.3-275-g84502dc: «v2017.04.3.275.g.84502.dc.20»
ugexe if using rakudobrew you can do `rakudobrew build moar 84502dc` 03:01
rakudobrew is actually the right tool for comparing multiple versions against each other 03:08
[Tux] This is Rakudo version 2017.06-133-ga7b2304e7 built on MoarVM version 2017.06-30-g389e9732 05:58
csv-ip5xs 2.647
test 12.623
test-t 4.141 - 4.227
csv-parser 13.273
nine Zoffix: just read through your depressing reddit fight with chromatic. If you actually want any other outcome than what occured, you have to give the other party a chance to safe face. I loved your suggestion for chromatic to come back, but he absolutely could not in that conversation. 06:45
stmuk I've abandoned the perl reddit for perl6 stuff and wish I could unread *that* thread too 09:25
Zoffix Isn't that manipulation? It amuses me how humans made what-you're-supposed-to-say-when-into-a-science, but half the time what you're "supposed" to do is lie or say things a specific way that's most likely to manipulate the other human to act in a specific way. 10:07
ZofBot: your hair looks lovely today! wink, wink
ZofBot Zoffix, 122)
nine Zoffix: what you did was manipulation just as well. You pushed him into a corner where he could not act in a different way even if he desparately wanted to (which may or may not be the case). 10:08
And most importantly: you want to manipulate. You want the negativity to stop and you want to have positive interaction. That's your goal. I'm telling you that you won't reach the goal with your current tactics :) 10:10
Zoffix Not at all. I pointed out to the audience the depressing whining about how no one uses Perl anymore reminds me of #perl channels I no longer frequent and that demoralizing Perl 6 core devs while expecting them to get cracking and produce stellar product is contradictory. There isn't any goals, or corners or checkpoints to save faces, or feet, or whatever else people like to save. 10:30
And just as predicted complains about Perl 6 irclog.perlgeek.de/mojo/2017-06-04#i_14684821 and depressing debates about whether Perl is dead irclog.perlgeek.de/mojo/2017-07-03#i_14819507 10:32
That's Perl 5 community to me in a nutshell and I'm not deluded enough to think I'll somehow change them all, only if I manipulate people by letting them save bodyparts or whatever. 10:33
'In the English-speaking world, the expression "to save face" describes the lengths that an individual may go to in order to preserve their established position in society, taking action to ensure that one is not thought badly of by his or her peers.' 10:36
nine In a nutshell: giving him a way out of the situation without him looking bad. 10:37
Zoffix I don't even know what chromatic's "established position in society" is.
nine And honestly, if you go out of your way to _find_ complaints about Perl 6 in the Perl 5 community, it's no wonder that you actually succeed. But that's not a way to arrive at a fair characterization and it's unclear to me how you will help anyone doing that. 10:39
stmuk I don't there is any point in enaging with most of these people 10:44
Zoffix I didn't have to go out of anything. The discussion *started* at with an article complaining about Perl 6, based on something vague TimToady said. 10:45
And a week later I'm being told to save face of an individual who obsessively reads Perl 6 chat logs and social media feeds to find anything to weaponize 10:46
nine Zoffix: I meant the IRC logs you linked
Zoffix: and I did not tell you anything!
Zoffix "<nine> [...] If you actually want any other outcome than what occured [...]" 10:47
nine Zoffix: I suggested that if you actually want to accomplish something, you would have to change tactics. But apparently that's not what you want. Looks like you're quite happy just bickering about those others.
So I'm gonna shut up now instead of wasting my work time on this.
Zoffix I was quite happy enjoying my morning coffee without incessant bullshit about language politics
nine There's a god damn "if" in there!
Zoffix Or what I'm supposed to do 10:48
I'm here to write code not to be told how to manipulate people.
nine Oh sorry for trying to help. I won't make that mistake again. 10:49
masak ...somebody called for hugs?
lizmat hugs nine and Zoffix 10:50
masak hugs Zoffix and nine
lizmat fwiw, I've plonked mr_chromatic mentally 10:51
masak clearly, he hasn't forgotten about us, though :P
masak waves to chromatic
lizmat "Hell hath no fury like a woman scorned" 10:52
my mental label for mr_chromatic
jnthn I've found the Perl community at large both reasonable and hospitable with regard to Perl 6, including during times when the state was much less good than it is today. There have been (and are) reasonable sceptics, who have had worthwhile input over the years. There are then a relatively very small handful of people who seemingly will be satisfied by nothing we do. 10:53
masak well, chromatic likes to talk about opportunity costs. imagine if all of chromatic's ranting were put to some good use. 10:54
the time spent on it, I mean.
lizmat maybe we should introduce a hot air tax :-)
masak .oO( well, that ballooned quickly ) 10:55
jnthn: I agree with your categorization. 10:58
nine Conflicts have been solved on a much, much larger scale. I firmly believe that with enough good will and the right approach this particular one can be, too. Apparently and surprisingly though, that's not wanted. Which in my view kinda removes any moral high ground for judging anyone.
stmuk it's a pity the original blog post was ever posted to reddit .. the title should have been enough to warn people off 10:59
it's not like any of this is anythng new 11:00
masak is wizened enough not to go find any links to waste his time on
oh, that's what "wizened" means. not quite what I meant, then :P 11:01
jnthn haha
masak but yeah, "old" as in "I've read reddit quite enough, thank you"
and I find a soupçon of chromatic doth not improve the taste of the broth 11:02
jnthn Righty, lunch, then incorporating type feedback into the new spesh model 11:03
uh, that woulda made more sense on #moarvm 11:04
Anyway, for those who aren't there, over there I'm doing a bunch of work on the specializer. :-)
bbi30
tbrowder Zoffix: can i extract or find out nqp version from the rakudo id? looks like i need to explicitly define nqp and moar branches to Configure.pl to get the proper Makefile for full installation. 12:20
timotimo tbrowder: tools/build/NQP_REVISION in rakudo and tools/build/MOAR_REVISION in nqp 12:24
tbrowder thnx timotimo 13:34
dogbert17 m: my ($a, $b) = BagHash.new(2, 2, 4), BagHash.new(2, 3, 3, 4); say $a (^) $b 13:36
camelia bag(3(2), 2)
lizmat feels correct ? 13:38
or do you expect setty semantics?
masak (^) is symmetric difference, yes? 13:41
lizmat yes 13:43
masak well, hmhm, seems to me the baggy generalization of symmetric set difference would be something like "let the lesser count of the element consume the greater" 13:45
which seems to be what's happening above, with the 2
dogbert17 commit: 2017.01 ($a, $b) = BagHash.new(2, 2, 4), BagHash.new(2, 3, 3, 4); say $a (^) $b
committable6 dogbert17, ¦2017.01: «04===SORRY!04=== Error while compiling /tmp/dLzTAvBkne␤Variable '$a' is not declared␤at /tmp/dLzTAvBkne:1␤------> 03(08⏏04$a, $b) = BagHash.new(2, 2, 4), BagHash. «exit code = 1»»
dogbert17 commit: 2017.01 my ($a, $b) = BagHash.new(2, 2, 4), BagHash.new(2, 3, 3, 4); say $a (^) $b 13:46
committable6 dogbert17, ¦2017.01: «bag(3(2), 2)»
dogbert17 commit: 2016.06 my ($a, $b) = BagHash.new(2, 2, 4), BagHash.new(2, 3, 3, 4); say $a (^) $b
committable6 dogbert17, ¦2016.06: «set(3)»
dogbert17 at least it used to be different
masak yeah, that old semantics looks like a more setty take 13:48
dogbert17 which one do you prefer?
masak I could imagine using both under different circumstances 13:49
but the baggy semantics is perhaps slightly nicer to have there, in that if I wanted the setty semantics I could drop down to Sets and then do (^) 13:50
lizmat note that if you want Setty semantics, you can coerce to Set :-)
masak right
lizmat masak: we have a similar issue with (<) and (<=)
which I gave baggy semantcs
masak we talked a little bit about it the other week
lizmat but reverted after dogbert17's finds 13:51
dogbert17 I have to change two rows here but that poses no problem :) docs.perl6.org/type/BagHash#Operators
lizmat masak: I was thinking of deprecating (<+) in favour of baggy semantics of (<=) 13:52
two reasons: we don't have a baggy equivalent of (<) currently
and it would reduce the number of set operators 13:53
fwiw, when I added baggy semantics to (<) and (<=) there were no spectest fails if I remember correctly 13:54
masak I'd love to discuss this further, but I'm about to decommute
lizmat masak: ack
lizmat is currently being commuted
dogbert17 lizmat: curveball RT #131561, is it resolved? 13:55
synopsebot6 Link: rt.perl.org/rt3/Public/Bug/Display...?id=131561
lizmat dogbert17: I beleive so 13:56
dogbert17 the tests as well?
lizmat m: dd (a => 1, a => -1).Mix
camelia mix()
lizmat m: dd (a => 0, a => 0).Mix 13:57
camelia mix()
lizmat tests?
dogbert17 where's the dot :)
m: dd (0 => 1, 0 => -1, 0 => 0, ).MixHash
camelia ()MixHash
lizmat ah.... 13:58
hmmm... different issue, I think
dogbert17 commit: 2016.12 dd (0 => 1, 0 => -1, 0 => 0, ).MixHash
committable6 dogbert17, ¦2016.12: «().MixHash»
lizmat dogbert17: fixing now 13:59
dogbert17 while commuting ?
jnthn The wonders of 21st century technology :) 14:00
lizmat yeah... 4G tethering :-)
dogbert17 cool, I guess you're using your laptop and not a mobile phone like Zoffix 14:01
lizmat yup
that makes things a lot easier :-)
dogbert17 indeed
dunno if we have an RT for this release blocker 'dot bug' :-) 14:02
Geth rakudo/nom: 9ba5eaef45 | (Elizabeth Mattijsen)++ | src/core/Baggy.pm
Oops, seems I forgot a dot, dogbert17++
dogbert17 lizmat++ 14:03
AlexDaniel c: 2017.04.3-275-g84502dc say 42 14:19
committable6 AlexDaniel, ¦2017.04.3-275-g84502dc: «42»
AlexDaniel holy…
lizmat AlexDaniel ? 14:23
AlexDaniel lizmat: I have never thought that something like this is going to work :)
especially in committable6 :)
lizmat ah :-)
AlexDaniel I mean, sure, it asks git to resolve this stuff, but omg that was unexpected
MasterDuke m: -> UInt @ {}(Array[UInt].new) 14:27
camelia Constraint type check failed in binding to parameter '<anon>'; expected UInt but got Array[UInt] (Array[UInt].new())
in block <unit> at <tmp> line 1
MasterDuke the ^^^ problem seems to be here: github.com/rakudo/rakudo/blob/nom/...P.nqp#L497
$cons_type is 'UInt' 14:28
$oval is 'Array[UInt]'
should $cons_type be something like 'Array[UInt]', or instead of `$result := $cons_type.ACCEPTS($oval);`, should $cons_type be ACCEPTS()ing the parameterized type of $oval? 14:33
jnthn MasterDuke: I suspect it shouldn't have a cons_type at all 14:34
MasterDuke jnthn: but UInt is a subset, i thought those created constraint types? 14:36
jnthn On $ surely 14:37
But Int @ means Positional[Int]
So surely UInt @ means Positional[UInt]
I think the mistake is probably being made at signature construction time 14:38
MasterDuke m: -> UInt @ {}.signature.say 14:41
camelia (Int @ where { ... })
jnthn Yeah, that's wrong 14:42
Shoudln't be pulling the type into pieces in the parametric case
MasterDuke hm, where do signatures get constructed? 14:43
jnthn actions/world
MasterDuke thanks, i'll look around 14:44
jnthn github.com/rakudo/rakudo/blob/nom/....nqp#L5040
I suspect it's wrong a bit before then
But it'll be close
Maybe around here github.com/rakudo/rakudo/blob/nom/....nqp#L5242 14:45
MasterDuke 'dissect_type_into_parameter', sounds relevant 14:48
dogbert17 lizmat: I'm going to add tests for RT #131303. Where should I put them? 16:13
synopsebot6 Link: rt.perl.org/rt3/Public/Bug/Display...?id=131303
Zoffix m: dd 1..4, 6..7 ... 9..12 16:39
camelia No such method 'succ' for invocant of type 'Range'. Did you mean any of these?
sec
sech
sum
uc

in block <unit> at <tmp> line 1
Zoffix m: dd infix:<...>( 1..4, 6..7, 9..12 ) 16:40
camelia (1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12).Seq
Zoffix m: dd infix:<...^>( 1..4, 6..7, 9..12 )
camelia Cannot resolve caller infix:<...^>(Range, Range, Range); none of these signatures match:
(\a, Mu \b)
in block <unit> at <tmp> line 1
Zoffix Where is that middle candidate used? Seems it's not reachable via normal operator use and looks like its extra features aren't available in the ...^ version.
m: dd [...] 1..4, 6..7, 9..12 16:41
camelia (1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12).Seq
Zoffix I guess this is it
And there's no good meaning for ...^ version, which is why it's lacking 16:44
m: dd 'a' … 'b'; dd 'a' …^ 'b'; 17:06
camelia ("a", "b").Seq
("a",).Seq
Zoffix m: dd 'aa' … 'bb'; dd 'aa' …^ 'bb';
camelia ("aa", "ab", "ba", "bb").Seq
("aa", "ab", "ba", "bb").Seq
Zoffix Wonder how good a coverage we got for &SEQUENCE 17:07
Undercover: &SEQUENCE
Undercover: coverage &SEQUENCE
coverage: &SEQUENCE
cover: &SEQUENCE 17:08
Undercover Zoffix, Failed to figure out coverage status on perl6.WTF/SETTING__src_core_operato....html#L112
Zoffix Undercover: you're useless 17:09
MasterDuke Zoffix: gist.github.com/Whateverable/72021...9d01acc6c5 17:12
different kind of coverage report though
Zoffix MasterDuke: how does it figure out the lines? 17:13
MasterDuke runs with MVM_COVER_LOG=1 and then parses the log file 17:15
Zoffix Ah. Then it has same limitations as perl6.wtf
MasterDuke yep
Geth rakudo/nom: 31eefb0d00 | TimToady++ | src/core/CompUnit/Repository/FileSystem.pm
Don't use .<<$key>> where .{$key} is intended
17:16
rakudo/nom: d5d3bd2763 | TimToady++ | src/Perl6/Grammar.nqp
Warn on ambiguous hyper inside quote words

Fixes #131695
synopsebot6 Link: rt.perl.org/rt3/Public/Bug/Display...?id=131695
MasterDuke but you get to choose the code to run and get coverage of
i.e., doesn't rely on a roast test existing 17:17
Zoffix Looks like there's some amount of tests for it in S03-sequence/* dir 17:21
781 lines in total
Zoffix is chaffed to use IO::CatHandle in real code for the first time :) perl6 -e 'IO::CatHandle.new(dir).lines.elems.say'
Guess `perl6 -e 'say +lines' *` is shorter, but still :) 17:22
spec (master)$ perl6 -MFile::Find -e 'say +lines IO::CatHandle.new: :encoding<latin1>, find :dir<.> :type<file>' 17:27
426222
:)
TimToady m: my $x = 42; say <<$x>> 17:33
camelia Potential difficulties:
Ambiguous use of >>; use » instead to mean hyper, or insert whitespace before >> to mean a quote terminator (or use different delimiters?)
at <tmp>:1
------> 3my $x = 42; say <<$x7⏏5>>
(42)
TimToady m: my $x = 42; say << $x >> 17:34
camelia (42)
TimToady m: my $x = 42; say <<$x>>[0]>> 17:35
camelia Potential difficulties:
Ambiguous use of >>; use » instead to mean hyper, or insert whitespace before >> to mean a quote terminator (or use different delimiters?)
at <tmp>:1
------> 3my $x = 42; say <<$x7⏏5>>[0]>>
Ambiguous…
TimToady it already found one problem in the core setting :) 17:36
Zoffix ©ol. 17:43
*cool
MasterDuke heh, never seen that typo before 17:44
Zoffix Accidental compose key press :) Compose, C+o = ©
MasterDuke timotimo: what does it mean for a type to be nominalizable? 18:07
anybody feel free to answer that 18:08
nine Zoffix: It was not my intention to order you to do anything. I was merely trying to be helpful. I obviously failed. Sorry for that. 18:20
timotimo MasterDuke: have you looked at Archetypes.nqp in rakudo's src/Perl6/Metamodel? 18:24
gotta go AFK now, but maybe that helps a little bit?
MasterDuke i haven't, but will take a look, thanks
jnthn MasterDuke: It means that it knows how to turn itself into something that is a nominal type (that is, only the name matters, not any kind of constrainty check) 18:42
MasterDuke jnthn: ah, thanks 18:45
fwiw, in the `elsif $type.HOW.archetypes.nominalizable {` branch of `dissect_type_into_parameter` i wrapped the whole post_contstraint stuff in an `if $type.HOW.name($type) ne 'UInt' {`
jnthn A role can do it by punning for example
MasterDuke and the example from the ticket doesn't die
jnthn Alright 18:46
So now it just needs a proper solution :)
MasterDuke but i don't know what the real solution is
heh. exactly
jnthn (Namely, not doing the dissection if the sigil is @ or % or &)
MasterDuke ah
Zoffix Speaking of & ... YIL: 18:47
m: -> & {}(/^/)
camelia ( no output )
Zoffix I've no idea what sort of arg it's expecting tho
m: /^/('x')
camelia No such method '!cursor_start' for invocant of type 'Str'
in block <unit> at <tmp> line 1
jnthn A Callable
m: say Regex ~~ Callable 18:48
camelia True
jnthn m: say Regex ~~ Method
camelia True
Zoffix No, I mean what has !cursor_start method.. 18:52
m: /^/(/^/)
camelia No such method '!cursor_start' for invocant of type 'Regex'
in block <unit> at <tmp> line 1
Zoffix m: /^/(grammar {})
camelia Cannot look up attributes in a <anon|63636064> type object
in block <unit> at <tmp> line 1
Zoffix m: /^/(grammar {}.new)
camelia ( no output )
Zoffix Ah. i see
jnthn A Cursor :) 18:55
I wonder a bit why it didn't stick a Curosr type constraint on the invocant there though
MasterDuke jnthn: it seems like the type constraint stuff doesn't know the containerness (i.e., sigil) of the parameter it's contraining? 18:57
Zoffix m: say /.../(/.../(/.../('foobar' ~~ /.../))) 18:58
camelia 「bar」
jnthn MasterDuke: No, so it'll probably need some refactoring to decide whether to dissect the type later 19:00
MasterDuke ok, thought that might be the case, wanted to make sure i wasn't missing something obvious 19:01
AlexDaniel I would appreciate any feedback on this: github.com/perl6/whateverable/issu...-312935676 19:03
especially from those who are smarter than me… so everyone is welcome to comment!
TimToady: that's a pretty cool fix, by the way 19:11
TimToady: I did not imagine that that would be possible
dogbert17 Zoffix: didn't you solve RT #131137 19:17
synopsebot6 Link: rt.perl.org/rt3/Public/Bug/Display...?id=131137
Zoffix dogbert17++ Thanks. Yeah. Closed it now. 19:18
dogbert17 yay, one RT down :) 19:19
Zoffix This is hilarious: pippinbarr.github.io/itisasifyouweredoingwork/ 19:20
dogbert17 Zoffix: do you remember if RT #130327 was only a case of a forgotten debug statement 19:41
synopsebot6 Link: rt.perl.org/rt3/Public/Bug/Display...?id=130327
Zoffix Don't know. 19:42
dogbert17 m: my $b = <a b b c d e f>.BagHash; $_ = 0 for $b.values; dd $b 21:37
camelia BagHash $b = ().BagHash
dogbert17 hmm, looks as if RT #130366 has been fixed 21:41
synopsebot6 Link: rt.perl.org/rt3/Public/Bug/Display...?id=130366
lizmat m: my $b = <a b b c d e f>.BagHash; $_ = 0, $_ = 2 for $b.values; dd $b
camelia BagHash $b = ("d"=>2,"e"=>2,"b"=>2,"a"=>2,"c"=>2,"f"=>2).BagHash
lizmat whee! :-)
decommute& 21:42
AlexDaniel buggable: github.com/rakudo/rakudo/commit/1c...24e4d0d5d3 21:44
blargh
dogbert17: interestingly, bisectable points to github.com/rakudo/rakudo/commit/1c...24e4d0d5d3
commit: 1cafc67b1aad14eafa^,1cafc67b1aad14eafa my $b = <a b b c d e f>.BagHash; $_ = 0 for $b.values; dd $b 21:45
committable6 AlexDaniel, ¦1cafc67b1aad14eafa^: «BagHash $b = ("a"=>0,"c"=>0,"b"=>0,"e"=>0,"f"=>0,"d"=>0).BagHash» ¦1cafc67: «BagHash $b = ("e"=>0,"d"=>0,"b"=>0,"a"=>0,"f"=>0,"c"=>0).BagHash»
AlexDaniel … and does so wrongly?
ah right, it's not sorted
dogbert17 there are quite a few RT's which have been fixed but lack tests and are therefore not closed. With this one I don't know, perhaps tests have been written. 21:47
AlexDaniel commit: c1bd844e2^,c1bd844e2 my $b = <a b b c d e f>.BagHash; $_ = 0 for $b.values; dd $b,
committable6 AlexDaniel, ¦c1bd844e2^: «BagHash $b = ("e"=>0,"d"=>0,"b"=>0,"a"=>0,"f"=>0,"c"=>0).BagHash» ¦c1bd844: «BagHash $b = ().BagHash»
AlexDaniel here it is
github.com/rakudo/rakudo/commit/c1...0d923ee44a
dogbert17: well, look for tests for RT #131241?
synopsebot6 Link: rt.perl.org/rt3/Public/Bug/Display...?id=131241
dogbert17 ok 21:48
AlexDaniel I don't see any 21:49
makes me wonder why is it resolved
dogbert17 at least none are marked as testing the RT in question
guess we'll have to ask lizmat 21:52
AlexDaniel dogbert17: I have marked both tickets as testneeded (and reopened the one that is closed) 21:54
dogbert17: and wrote comments on both
dogbert17 so instead of getting rid of one RT a closed one was reopened instead :)
AlexDaniel XD 21:55
I mean… this seemed like a right thing to do…
we can merge two tickets though 21:56
dogbert17++ for reviewing the ticket queue 21:57
dogbert17 adding tests should enable us to close quite a few 22:03
AlexDaniel mst: if I reached the limit for the number of user connections on freenode, should I just trick it by using a different ip for some bots, or is there a righter way? 22:25
committable6: uptime 22:42
committable6 AlexDaniel, 5 weeks, 1 day, 18 hours, 5 minutes, and 4 seconds
AlexDaniel does it mean that the bug was fixed?
but I didn't do anything…