»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'p6: say 3;' or rakudo:, std:, or /msg camelia p6: ... | irclog: irc.perl6.org | UTF-8 is our friend!
Set by masak on 12 May 2015.
ShimmerFairy poking at the p6u archives just because, I found www.nntp.perl.org/group/perl.perl6....g2202.html , and I wish someone could've mentioned what substr-rw actually does (and why .= wouldn't cut it) 00:19
flussence .substr returns a copy, .substr-rw returns an lvalue/proxy that mutates the original string? That's my guess anyway. 00:21
ShimmerFairy flussence: yeah, substr-rw is about letting you modify things in the original string, which is different from what .= is for. And the thread seems to be based in a misunderstanding of what the -rw methods in core do. 00:23
flussence trim-rw sounds like complete nonsense to me. 00:24
ShimmerFairy Yes. Again, the thread seems to have been started on a misunderstanding that -rw methods mutates the invocant, i.e. that they're the same as .=, and I just wish someone pointed that out. 00:25
(My understanding of the mailing list is that I couldn't just send a reply about half a month after the discussion happened, unfortunately ☺) 00:26
00:32 llfourn joined, breinbaas left
TEttinger jdv79: interestingly, I hit some... issues with one of the existing ones 00:33
github.com/atom/language-perl I believe is the right link
it highlights any ' or " as the start of a string, which uh causes issues when writing grammars that use those. thankfully... \c[APOSTROPHE] 00:34
ShimmerFairy TEttinger: does it choke on $legal'name too? 00:35
TEttinger I didn't try it. 00:36
replacing bare ' with \c[APOSTROPHE] worked everywhere I needed it
00:36 llfourn left
ShimmerFairy TEttinger: looking at that syntax highlighter, it's cute how it thinks it can define enough possible brackets for Q-lang strings :P (such repetition, jeez) 00:37
01:00 llfourn joined 01:05 llfourn left 01:18 BenGoldberg joined 01:30 cwDYKARN left
dalek ast/glr: 0de14e4 | skids++ | S04-statement-modifiers/without.t:
Better nonfatalization of test. Now we pass this file.

Doh. Using try to nonfatalize this test made it fail because
  "use fatal" is turned on inside of try blocks. The test has
since been fixed. Use a non-use-fatal defatalizer.
02:09
02:15 noganex_ joined 02:18 noganex left
dalek ast/glr: 57591f9 | skids++ | integration/precompiled.t:
unfudge now-passing test
02:23
02:26 telex left 02:28 telex joined 02:42 muraiki joined
muraiki meow 02:56
timotimo has a spurious wakelup 02:57
wakeup*
03:01 llfourn joined 03:05 CQ_ joined 03:06 llfourn left 03:07 CQ left, CQ_ is now known as CQ
muraiki night 03:18
03:18 muraiki left
ShimmerFairy nqp: my @a := nqp::list_s(); nqp::push_s(@a, "foo"); say(@a[0]); 03:18
03:18 camelia left, tinyblak joined 03:19 camelia joined 03:20 khw joined
ShimmerFairy nqp-m: my @a := nqp::list_s(); nqp::push_s(@a, "foo"); say(@a[0]); 03:20
03:20 prevost joined
camelia nqp-parrot: OUTPUT«Can't exec "./rakudo-inst/bin/nqp-p": No such file or directory at lib/EvalbotExecuter.pm line 193.␤exec (./rakudo-inst/bin/nqp-p /tmp/tmpfile) failed: No such file or directory␤Server error occurred! Closing Link: ns1.niner.name (Quit: camelia)␤Lost connect…» 03:21
..nqp-jvm: OUTPUT«(signal ABRT)#␤»
..nqp-moarvm: OUTPUT«MVMArray: atpos expected string register␤ at /tmp/tmpfile:1 (<ephemeral file>:<mainline>:26)␤ from gen/moar/stage2/NQPHLL.nqp:1289 (/home/camelia/rakudo-inst-1/share/nqp/lib/NQPHLL.moarvm:eval:190)␤ from gen/moar/stage2/NQPHLL.nqp:1492 (/home/camelia/ra…»
03:21 ChanServ sets mode: +v camelia
camelia nqp-moarvm: OUTPUT«MVMArray: atpos expected string register␤ at /tmp/7hfoIvQdk8:1 (<ephemeral file>:<mainline>:26)␤ from gen/moar/stage2/NQPHLL.nqp:1289 (/home/camelia/rakudo-inst-1/share/nqp/lib/NQPHLL.moarvm:eval:190)␤ from gen/moar/stage2/NQPHLL.nqp:1492 (/home/camelia…» 03:21
ShimmerFairy let's try this without the crashing this time :)
nqp-m: my @a := nqp::list_s(); nqp::push_s(@a, "foo"); say(@a[0]);
camelia nqp-moarvm: OUTPUT«MVMArray: atpos expected string register␤ at /tmp/Nt7dj3_3BC:1 (<ephemeral file>:<mainline>:26)␤ from gen/moar/stage2/NQPHLL.nqp:1289 (/home/camelia/rakudo-inst-1/share/nqp/lib/NQPHLL.moarvm:eval:190)␤ from gen/moar/stage2/NQPHLL.nqp:1492 (/home/camelia…»
ShimmerFairy nqp-m: my str @a; nqp::push_s(@a, "foo"); say(@a[0]);
camelia nqp-moarvm: OUTPUT«Only typed scalars are currently supported in NQP at line 2, near "; nqp::pus"␤ at gen/moar/stage2/NQPHLL.nqp:516 (/home/camelia/rakudo-inst-1/share/nqp/lib/NQPHLL.moarvm:panic:105)␤ from gen/moar/stage2/NQP.nqp:2930 (/home/camelia/rakudo-inst-1/share/nqp/…»
ShimmerFairy nqp-m: my $a := nqp::list_s(); nqp::push_s($a, "foo"); say($a[0]); 03:22
camelia nqp-moarvm: OUTPUT«MVMArray: atpos expected string register␤ at /tmp/M7VzupNiO3:1 (<ephemeral file>:<mainline>:26)␤ from gen/moar/stage2/NQPHLL.nqp:1289 (/home/camelia/rakudo-inst-1/share/nqp/lib/NQPHLL.moarvm:eval:190)␤ from gen/moar/stage2/NQPHLL.nqp:1492 (/home/camelia…»
ShimmerFairy As far as I can tell, nqp::list_s() (and presumably the other typed variants) are basically useless in NQP at the moment. I'm guessing that's not the idea? :)
03:23 tinyblak left 03:24 aborazmeh joined, aborazmeh left, aborazmeh joined
ShimmerFairy Also, that camelia crash reminded me of a serious question I had: who actually uses nqp-p anymore? I mean, I know the Parrot VM was kept for NQP because the big changes for rakudo that lead to tearing out parrot didn't need anything in NQP, so you didn't _have_ to deal with nqp-p (recalling pmichaud's post). 03:28
But really, is there any point to keeping it? I have half a mind to just go ahead and tear out any trace of parrot in nqp and see how much people don't care :P .
(camelia reminded me because, if you look at the results after it came back from the crash, it's specifically nqp-p that caused a server error) 03:29
03:43 tinyblak joined 03:45 rangerprice joined 03:51 khw left 03:53 khw joined 03:57 skids left
timotimo huh, what broke list_s in nqp?! 04:05
nqp-m: my $a := nqp::list_s(); nqp::push_s($a, "foo"); say(nqp::atpos_s($a[0])); 04:06
camelia nqp-moarvm: OUTPUT«Arg count 1 doesn't equal required operand count 3 for op 'atpos_s'␤ at gen/moar/stage2/QAST.nqp:1604 (/home/camelia/rakudo-inst-1/share/nqp/lib/QAST.moarvm:compile_mastop:209)␤ from gen/moar/stage2/QAST.nqp:1820 (/home/camelia/rakudo-inst-1/share/nqp/lib/…»
timotimo nqp-m: my $a := nqp::list_s(); nqp::push_s($a, "foo"); say(nqp::atpos_s($a, 0));
camelia nqp-moarvm: OUTPUT«foo␤»
timotimo seems like it's just the postcircumfix:<[ ]> in nqp that just compiles down to atpos rather than atpos_*
dalek kudo/glr: 86dbc33 | TimToady++ | src/core/Hash.pm:
fix .perl output of object hashes
04:08
timotimo m: my ($first, $second) = "hello/world".split: "/"; say $first; say $second 04:10
camelia rakudo-moar 5fb81f: OUTPUT«hello␤world␤»
GLRelia rakudo-moar 7591fc: OUTPUT«hello␤world␤»
timotimo m: my @exploded; my ($first, $second) = "hello/world,friends,mom".split: "/"; my @seconds; if $second.defined && $second.chars { @seconds = $second.split: ","; }; @exploded.push: { name => $first, seconds => @seconds }; say @exploded 04:13
camelia rakudo-moar 5fb81f: OUTPUT«name => hello, seconds => world friends mom␤»
GLRelia rakudo-moar 7591fc: OUTPUT«name => helloseconds => [world friends mom]␤»
timotimo m: my @exploded; my ($first, $second) = "hello/world,friends,mom".split: "/"; my @seconds; if $second.defined && $second.chars { @seconds = $second.split: ","; }; @exploded.push: { name => $first, seconds => @seconds }; say @exploded; @seconds = "frob,boz,wop".split(","); @exploded[0]<seconds>.push: |@seconds; say @exploded.perl 04:14
camelia rakudo-moar 5fb81f: OUTPUT«name => hello, seconds => world friends mom␤[{:name("hello"), :seconds(["frob", "boz", "wop", "frob", "boz", "wop"])}]<>␤»
GLRelia rakudo-moar 7591fc: OUTPUT«name => helloseconds => [world friends mom]␤Cannot modify an immutable Any␤ in block <unit> at /tmp/VsNRkO_exv:1␤␤»
timotimo m: my @exploded; my ($first, $second) = "hello/world,friends,mom".split: "/"; my @seconds; if $second.defined && $second.chars { @seconds = $second.split: ","; }; say @exploded.WHAT; @exploded.push: { name => $first, seconds => @seconds }; say @exploded; @seconds = "frob,boz,wop".split(","); @exploded[0]<seconds>.push: |@seconds; say @exploded.perl 04:15
camelia rakudo-moar 5fb81f: OUTPUT«(Array)␤name => hello, seconds => world friends mom␤[{:name("hello"), :seconds(["frob", "boz", "wop", "frob", "boz", "wop"])}]<>␤»
GLRelia rakudo-moar 7591fc: OUTPUT«(Array)␤name => helloseconds => [world friends mom]␤Cannot modify an immutable Any␤ in block <unit> at /tmp/nI_DMbumav:1␤␤»
timotimo m: my @seconds = "world,friends,mom".split(","); @seconds.push: |("a,b,c".split(","))
camelia ( no output )
GLRelia ( no output )
timotimo m: my @seconds = "world,friends,mom".split(","); @seconds.push: |("a,b,c".split(",")); say @seconds
camelia rakudo-moar 5fb81f: OUTPUT«world friends mom a b c␤»
GLRelia rakudo-moar 7591fc: OUTPUT«worldfriendsmomabc␤»
timotimo i want spaces back :| 04:16
m: my @exploded; my ($first, $second) = "hello/world,friends,mom".split: "/"; my @seconds; if $second.defined && $second.chars { @seconds = $second.split: ","; }; say @exploded.WHAT; @exploded.push: { name => $first, seconds => @seconds }; say @exploded.perl; @seconds = "frob,boz,wop".split(","); @exploded[0]<seconds>.push: |@seconds; say @exploded.perl 04:17
camelia rakudo-moar 5fb81f: OUTPUT«(Array)␤[{:name("hello"), :seconds(["world", "friends", "mom"])}]<>␤[{:name("hello"), :seconds(["frob", "boz", "wop", "frob", "boz", "wop"])}]<>␤»
GLRelia rakudo-moar 86dbc3: OUTPUT«(Array)␤[:name("hello"), :seconds($["world", "friends", "mom"])]␤Cannot modify an immutable Any␤ in block <unit> at /tmp/Y08mm8sRYq:1␤␤»
timotimo m: my @exploded; my ($first, $second) = "hello/world,friends,mom".split: "/"; my @seconds; if $second.defined && $second.chars { @seconds = $second.split: ","; }; say @exploded.WHAT; @exploded.push: { name => $first, seconds => @seconds }; say @exploded.perl; @seconds = "frob,boz,wop".split(","); say @exploded[0].WHAT 04:18
camelia rakudo-moar 5fb81f: OUTPUT«(Array)␤[{:name("hello"), :seconds(["world", "friends", "mom"])}]<>␤(Hash)␤»
GLRelia rakudo-moar 86dbc3: OUTPUT«(Array)␤[:name("hello"), :seconds($["world", "friends", "mom"])]␤(Pair)␤»
timotimo oh, whoops?
i hadn't noticed
m: my @exploded; my ($first, $second) = "hello/world,friends,mom".split: "/"; my @seconds; if $second.defined && $second.chars { @seconds = $second.split: ","; }; say @exploded.WHAT; @exploded.push: ${ name => $first, seconds => @seconds }; say @exploded.perl; @seconds = "frob,boz,wop".split(","); say @exploded[0].WHAT
04:18 aborazmeh left
camelia rakudo-moar 5fb81f: OUTPUT«5===SORRY!5=== Error while compiling /tmp/eOhhwFNMgJ␤Unsupported use of ${ name => $first, seconds => @seconds }; in Perl 6 please use $( name => $first, seconds => @seconds )␤at /tmp/eOhhwFNMgJ:1␤------> 3${ name => $first, seconds => @seconds …» 04:18
GLRelia rakudo-moar 86dbc3: OUTPUT«(Array)␤[{:name("hello"), :seconds($["world", "friends", "mom"])},]␤(Hash)␤»
timotimo m: my @exploded; my ($first, $second) = "hello/world,friends,mom".split: "/"; my @seconds; if $second.defined && $second.chars { @seconds = $second.split: ","; }; say @exploded.WHAT; @exploded.push: { name => $first, seconds => @seconds }.item; say @exploded.perl; @seconds = "frob,boz,wop".split(","); say @exploded[0].WHAT
camelia rakudo-moar 5fb81f: OUTPUT«(Array)␤[{:name("hello"), :seconds(["world", "friends", "mom"])}]<>␤(Hash)␤»
GLRelia rakudo-moar 86dbc3: OUTPUT«(Array)␤[{:name("hello"), :seconds($["world", "friends", "mom"])},]␤(Hash)␤»
timotimo this may help perl6-bench quite a bit 04:19
with a simple .item in the right place, bench list-checkouts is happy again
ShimmerFairy nqp-m: my @a := nqp::list_s(); nqp::push_s(@a, "foo"); say(nqp::atpos_s(@a, 0)); 04:21
camelia nqp-moarvm: OUTPUT«foo␤»
timotimo ohai ShimmerFairy
04:21 kaare__ joined
timotimo how are you today? 04:21
ShimmerFairy good :)
timotimo i'm glad to hear that!
dalek rl6-bench: 7c9e687 | timotimo++ | lib/Bench/Handling.pm6:
make bench readier for the post-GLR world
04:24 tinyblak left
dalek rl6-bench: b2ca0e1 | timotimo++ | / (3 files):
kebapcase some deprecated thingies
04:24
04:25 bin_005_l joined
timotimo with that and a "git pull" inside File-Find, i can now run bench without deprecation warnings 04:26
04:28 bin_005 left
colomon o/ 04:28
dalek rl6-bench: 56b7bbf | timotimo++ | microbenchmarks.pl:
fix one more constant 1 in sink context
04:29
timotimo m: my @show = " ", "X", "O"; my @grid = [ (0, 1, 2) xx 5 ] xx 5; say @show[@grid[0]] 04:36
camelia rakudo-moar 5fb81f: OUTPUT«(Any)␤»
GLRelia rakudo-moar 86dbc3: OUTPUT«(Any)␤»
timotimo m: my @show = " ", "X", "O"; my @grid = [ (0, 1, 2) xx 5 ] xx 5; say @show[@grid[0].list]
camelia rakudo-moar 5fb81f: OUTPUT« X O X O X O X O X O␤»
GLRelia rakudo-moar 86dbc3: OUTPUT«(Any)(Any)(Any)(Any)(Any)␤»
timotimo how do we do this in glr-land?
m: my @show = " ", "X", "O"; my @grid = [ (0, 1, 2) xx 5 ] xx 5; say @show[|@grid[0]]
GLRelia rakudo-moar 86dbc3: OUTPUT«(Any)(Any)(Any)(Any)(Any)␤»
camelia rakudo-moar 5fb81f: OUTPUT«5===SORRY!5=== Error while compiling /tmp/KSsnm4ChdE␤Arg-flattening | is only valid in an argument list␤at /tmp/KSsnm4ChdE:1␤------> 3d = [ (0, 1, 2) xx 5 ] xx 5; say @show[|7⏏5@grid[0]]␤»
timotimo m: my @show = " ", "X", "O"; my @grid = [ (0, 1, 2) xx 5 ] xx 5; say @grid[0].perl; say @show[@grid[0]]
camelia rakudo-moar 5fb81f: OUTPUT«[0, 1, 2, 0, 1, 2, 0, 1, 2, 0, 1, 2, 0, 1, 2]␤(Any)␤»
GLRelia rakudo-moar 86dbc3: OUTPUT«[(0, 1, 2), (0, 1, 2), (0, 1, 2), (0, 1, 2), (0, 1, 2)]␤(Any)␤»
timotimo oh, that's why
m: my @show = " ", "X", "O"; my @grid = [ |(0, 1, 2) xx 5 ] xx 5; say @grid[0].perl; say @show[@grid[0]] 04:37
camelia rakudo-moar 5fb81f: OUTPUT«5===SORRY!5=== Error while compiling /tmp/Ix1UgagV4H␤Arg-flattening | is only valid in an argument list␤at /tmp/Ix1UgagV4H:1␤------> 3my @show = " ", "X", "O"; my @grid = [ |7⏏5(0, 1, 2) xx 5 ] xx 5; say @grid[0].perl␤»
GLRelia rakudo-moar 86dbc3: OUTPUT«[0, 1, 2, 0, 1, 2, 0, 1, 2, 0, 1, 2, 0, 1, 2]␤(Any)␤»
timotimo m: my @show = " ", "X", "O"; my @grid = [ |(0, 1, 2) xx 5 ] xx 5; say @grid[0].perl; say @show[@grid[0].list]
camelia rakudo-moar 5fb81f: OUTPUT«5===SORRY!5=== Error while compiling /tmp/4YwQewXura␤Arg-flattening | is only valid in an argument list␤at /tmp/4YwQewXura:1␤------> 3my @show = " ", "X", "O"; my @grid = [ |7⏏5(0, 1, 2) xx 5 ] xx 5; say @grid[0].perl␤»
GLRelia rakudo-moar 86dbc3: OUTPUT«[0, 1, 2, 0, 1, 2, 0, 1, 2, 0, 1, 2, 0, 1, 2]␤ XO XO XO XO XO␤»
timotimo m: my @show = " ", "X", "O"; my @grid = [ flat (0, 1, 2) xx 5 ] xx 5; say @grid[0].perl; say @show[@grid[0].list]
camelia rakudo-moar 5fb81f: OUTPUT«[0, 1, 2, 0, 1, 2, 0, 1, 2, 0, 1, 2, 0, 1, 2]␤ X O X O X O X O X O␤»
GLRelia rakudo-moar 86dbc3: OUTPUT«[0, 1, 2, 0, 1, 2, 0, 1, 2, 0, 1, 2, 0, 1, 2]␤ XO XO XO XO XO␤»
dalek kudo/glr: c3f975b | TimToady++ | src/Perl6/Grammar.nqp:
deconfuse shaped declarations from contextualizers
04:38
kudo/glr: b05b863 | TimToady++ | src/core/Hash.pm:
use declarative syntax for object hash .perl
ast/glr: 3b3c42d | TimToady++ | S29-conversions/hash.t:
Z needs flat now
04:41
rl6-bench: 596ec1c | timotimo++ | perl6/rc-forest-fire (2 files):
the forest-fire benchmarks now work properly post-glr as well
04:42
rl6-bench: cb96089 | timotimo++ | microbenchmarks.pl:
give more X operators the flat treatment
04:45
04:45 tinyblak joined
timotimo re-running the benchmarks now that they are semantically correct(er) 04:46
04:50 tinyblak left
timotimo i should time if "for ^1000 X ^1000 -> $i, $j { ... }" is drastically faster than "for flat(^1000 X ^1000) -> $i, $j { ... }", otherwise the benchmarks will be unfairly impacted by this compatibility code :| 04:50
m: my $starttime = now; (^1000 X ^1000).list; say now - $starttime
GLRelia rakudo-moar b05b86: OUTPUT«0.0084918␤»
timotimo m: my $starttime = now; flat(^1000 X ^1000).list; say now - $starttime
GLRelia rakudo-moar b05b86: OUTPUT«0.00877925␤»
timotimo did someone kick out camelia? o_O
oh, glrelia just doesn't iterate anything there 04:51
camelia rakudo-moar 5fb81f: OUTPUT«16.092873␤»
rakudo-moar 5fb81f: OUTPUT«16.3938472␤»
timotimo m: my $starttime = now; my int $sum; for flat(^100 X ^100) -> $i, $j { $sum += $i + $j }; say now - $starttime
camelia rakudo-moar 5fb81f: OUTPUT«0.2192266␤»
GLRelia rakudo-moar b05b86: OUTPUT«3.3929676␤»
04:51 khw left
timotimo m: my $starttime = now; my int $sum; for (^100 X ^100) -> $i, $j { $sum += $i + $j }; say now - $starttime 04:51
camelia rakudo-moar 5fb81f: OUTPUT«0.21331947␤»
GLRelia rakudo-moar b05b86: OUTPUT«1.4659768␤»
timotimo OK, so no difference on camelia, but the base hit of using X to iterate into two vars is SUPER BAD 04:52
m: my $starttime = now; my int $sum; for (^100 X ^100) -> ($i, $j) { $sum += $i + $j }; say now - $starttime
camelia rakudo-moar 5fb81f: OUTPUT«Too few positionals passed; expected 2 arguments but got 0 in sub-signature␤ in block <unit> at /tmp/KvFmuPQlUJ:1␤␤»
GLRelia rakudo-moar b05b86: OUTPUT«2.10549970␤»
timotimo that's a tiny bit better than flattening it, at least, but still ~10x slower than pre-glr 04:53
dalek kudo/glr: 4f17b9c | TimToady++ | src/core/Hash.pm:
use declarative syntax for typed hash .perl
04:55
ast/glr: 7b4a9ac | TimToady++ | S32-hash/perl.t:
check against new .perl for weird hashes
04:55 breinbaas joined 05:01 rurban joined 05:02 llfourn joined
dalek kudo/glr: 1afdc6c | TimToady++ | src/core/Hash.pm:
weird Hash.perl should also allow self-ref
05:06
05:06 llfourn left
TimToady m: say Hash[Any,Any].new((:a, :!b)).perl 05:07
camelia rakudo-moar 5fb81f: OUTPUT«Hash[Any,Any].new("a" => Bool::True, "b" => Bool::False)␤»
GLRelia rakudo-moar 4f17b9: OUTPUT«(my Any %{Any} = :a, :!b)␤»
TimToady .tell lizmat I fixed yer Hash.perls 05:09
yoleaux TimToady: I'll pass your message to lizmat.
05:10 tinyblak joined 05:14 colomon left, colomon joined 05:22 prevost left 05:26 laouji joined
moritz m: my %h; %h<a> = %h; say %h.perl 05:28
GLRelia rakudo-moar 1afdc6: OUTPUT«(my \Hash_140079533019120 = {:a(Hash_140079533019120)})␤»
moritz m: my %h; %h<a> = %h; say EVAL(%h.perl).perl
GLRelia rakudo-moar 1afdc6: OUTPUT«{:a(Mu)}␤»
05:28 laouji left
camelia rakudo-moar 5fb81f: OUTPUT«(timeout)» 05:28
colomon m: my @array = ("o", "k"); print @array 05:29
camelia rakudo-moar 5fb81f: OUTPUT«ok»
GLRelia rakudo-moar 1afdc6: OUTPUT«ok»
colomon m: my $array-ref = ("o", "k"); print $array-ref
camelia rakudo-moar 5fb81f: OUTPUT«o k» 05:30
GLRelia rakudo-moar 1afdc6: OUTPUT«o k»
05:31 rurban left
TimToady after sleeping on it, I'm thinking the best way forward on say is to put spaces between top-level arguments, and then the one-arg rule doens't matter 05:40
colomon is t/spec/S16-io/print.t failing for everyone on GLR or just me? the $array-ref one above is failing for me, yet looks perfect here
moritz colomon: recompiling now to check
colomon moritz++ 05:41
TimToady breakfast &
moritz colomon: tests 4 and 12 fail here 05:42
colomon moritz: same here
m: my $array-ref = ("o", "k"); print $array-ref
camelia rakudo-moar 5fb81f: OUTPUT«o k»
GLRelia rakudo-moar 1afdc6: OUTPUT«o k»
colomon oh, is it also failing on nom? 05:43
it looks like it wants “ok” for that answer
JimmyZ m: my @array = ("o", "k"); print @array 05:44
GLRelia rakudo-moar 1afdc6: OUTPUT«ok»
camelia rakudo-moar 5fb81f: OUTPUT«ok»
06:03 tinyblak left, tinyblak joined
ShimmerFairy TimToady: I'm not so sure I'd agree with the space separation, it feels like a bad idea to have a default separator you can't get rid of when it's not desired. There is [~] to not have spaces, of course, but on the other hand you also have .join(" ") to get spaces in the first place :) 06:04
06:04 cognominal left 06:05 domidumont joined
ShimmerFairy Also, if we go with spaces by default instead of calling .gist on the list object, then that could cause some issues for any custom list types that have a non-standard .gist (namely, one not built around space-separated values) 06:08
06:10 domidumont left, Sqirrel left 06:11 domidumont joined
ShimmerFairy &say definitely seems like one of those functions that will want to treat lists as singular objects more often than not. I'm starting to get the feeling that function signatures need some way of choosing to go by single-arg rule (either decided by the caller or callee), since it appears to be causing quite a bit of struggle. 06:12
06:12 diana_olhovik_ joined 06:17 mr-foobar joined 06:20 cognominal joined 06:22 mr-foobar left 06:23 rangerprice left, tinyblak left 06:25 mr-foobar joined 06:28 domidumont left 06:29 domidumont joined 06:34 ab6tract joined
ab6tract m: sub ret-array { [6,7,8] }; my $f = ret-array; say $f 06:35
camelia rakudo-moar 5fb81f: OUTPUT«6 7 8␤»
GLRelia rakudo-moar 1afdc6: OUTPUT«678␤»
ab6tract m: sub ret-array { [6,7,8] }; my $f = ret-array; say $f.perl
camelia rakudo-moar 5fb81f: OUTPUT«[6, 7, 8]␤»
GLRelia rakudo-moar 1afdc6: OUTPUT«[6, 7, 8]␤»
ab6tract *phew*
that haunted my dreams last night :)
m: sub ret-array { [6,7,8] }; my ($f,$g,$h) = ret-array; say $f.perl 06:36
camelia rakudo-moar 5fb81f: OUTPUT«[6, 7, 8]␤»
GLRelia rakudo-moar 1afdc6: OUTPUT«6␤»
06:40 ab6tract left, erdic left 06:49 erdic joined
TimToady otoh, .join is 5 characters shorter than .join(" ") 06:54
TimToady --> venue 06:55
06:59 itz joined
timotimo it feels like i got a little bit of sleep in this time 07:01
07:02 rurban joined 07:03 llfourn joined, mephinet joined 07:04 BenGoldberg left 07:07 llfourn left 07:08 laben joined
laben hello p6'ers 07:09
colomon o/ 07:14
ooo, finally got a complete smoke test run! 07:17
colomon is off to the hackathon 07:18
07:18 colomon left
nine Oh I figured out why seq.t is failing. 07:18
What I'm not sure about is if it's the test or List.STORE that's wrong. I'm inclined to put the blame on the test. 07:19
JimmyZ: you here? 07:21
timotimo i hope i won't miss breakfast =o
nine timotimo: well, there's still coffee and food but I fear no company
timotimo i don't like coffee
i only care for food
t.h8.lv/p6bench/2015-08-29-nom_vs_glr.html - benchmark run updated 07:22
07:22 itz left
nine jnthn: you awake? Your session starts in 10 minutes 07:22
dalek kudo/glr: 741d14a | TimToady++ | src/core/Slip.pm:
treat slip() as undefined
timotimo nine: want to figure out why for_empty is almost 10x slower on glr than on nom? :) 07:23
nine for_empty?
timotimo one of those benchmarks 07:24
it's basically a for loop over (1 .. SCALE) { } 07:25
oh
but let me check out the ast and optimized ast first
ah, yeah
we've lost the optimization that turns for over a range into a while loop
07:25 colomon joined
TimToady m: say (if 0 { 1 }).defined 07:26
camelia rakudo-moar 5fb81f: OUTPUT«False␤»
GLRelia rakudo-moar 1afdc6: OUTPUT«True␤»
07:27 lizmat joined
moritz m: say slip().defined 07:27
GLRelia rakudo-moar 1afdc6: OUTPUT«True␤»
camelia rakudo-moar 5fb81f: OUTPUT«5===SORRY!5=== Error while compiling /tmp/BAC_prljQD␤Undeclared routine:␤ slip used at line 1. Did you mean 'flip'?␤␤»
moritz m: say slip(1).defined
GLRelia rakudo-moar 1afdc6: OUTPUT«True␤»
camelia rakudo-moar 5fb81f: OUTPUT«5===SORRY!5=== Error while compiling /tmp/Xnc5EqXozo␤Undeclared routine:␤ slip used at line 1. Did you mean 'flip'?␤␤»
TimToady m: say (if 0 { 1 }).defined 07:31
camelia rakudo-moar 5fb81f: OUTPUT«False␤»
GLRelia rakudo-moar 741d14: OUTPUT«False␤»
07:34 ab6tract joined
colomon o/ 07:35
[ptc] \o
hoelzro o/ from the salon @ SPW
timotimo breakfast-time
hoelzro it's breakfast...it's breakfast time
JimmyZ nine: I am here 07:37
nine: but I think it test is right, TimToady may know it better :) 07:38
nine JimmyZ: seq.t is testing ($x, @a) = ('xxx', (1,2,3).map(*.Int)) and expecting @a to contain the mapped values instead of a seq. Why?
[Tux] glr + panda = fail (current checkout)
==> Bootstrapping Panda
resolve stage failed for File::Find: Dependency True is not present in the module ecosystem
JimmyZ nine: because the first test works 07:39
[Tux] NOM: 07:40
test 50000 43.709 43.616
test-t 50000 45.639 45.546
nine JimmyZ: the first test is assignment of a single Iterable to a list. Under the single-arg-rule we flatten that. The other tests assign lists of multiple values to lists. We're not supposed to flatten those.
JimmyZ nine: TimToady is here :P
07:41 rurban left 07:42 FROGGS joined, Ven joined, darutoko joined
lizmat good *, #perl6 07:42
yoleaux 05:09Z <TimToady> lizmat: I fixed yer Hash.perls
lizmat TimToady++ :-)
FROGGS o/
colomon \o 07:43
07:43 virtualsue joined
FROGGS most of you sitting in the Salon? 07:43
nine Outside
colomon lizmat, nine, and I are outside
lizmat nine colomon lizmat sitting outside
JimmyZ m: my ($x, @b) = ('xxx', (1,2,3).map(*.Int)); say @b.WHAT
camelia rakudo-moar 5fb81f: OUTPUT«(Array)␤»
GLRelia rakudo-moar 741d14: OUTPUT«(Array)␤»
lizmat :-)
[Tux] m: class C{method foo{gather { take 3..Inf}}};C.new.foo[^9].perl.say 07:44
camelia rakudo-moar 5fb81f: OUTPUT«(3, 4, 5, 6, 7, 8, 9, 10, 11)␤»
GLRelia rakudo-moar 741d14: OUTPUT«(3..Inf)␤»
FROGGS we (itz, virtualsue, ven and me) are upstairs
laben |Tux|: checking out panda on glr again today, it has become my stated mission to make it work and so shall be it
JimmyZ m: my ($x, @b) = ('xxx', (1,2,3).map(*.Int)); say @b
camelia rakudo-moar 5fb81f: OUTPUT«1 2 3␤»
GLRelia rakudo-moar 741d14: OUTPUT«(1 2 3)␤»
dalek ast/glr: a9bc960 | (Stefan Seifert)++ | S32-list/seq.t:
Fix seq.t to actually adhere to the single-arg-rule
07:45
07:45 itz joined
colomon m: my $s = <a b c>.set; say $s.perl; say ($s (&) $s).perl 07:46
camelia rakudo-moar 5fb81f: OUTPUT«Method 'set' not found for invocant of class 'Parcel'␤ in block <unit> at /tmp/iS1U3Meeow:1␤␤»
GLRelia rakudo-moar 741d14: OUTPUT«Method 'set' not found for invocant of class 'List'␤ in block <unit> at /tmp/Ej1g3DzSYN:1␤␤»
colomon m: my $s = <a b c>.Set; say $s.perl; say ($s (&) $s).perl
GLRelia rakudo-moar 741d14: OUTPUT«set("a","c","b")␤set("a","c","b")␤»
camelia rakudo-moar 5fb81f: OUTPUT«set("a","c","b")␤set("a","c","b")␤»
JimmyZ m: my ($x, @b) = ('xxx', flat (1,2,3).map(*.Int)); say @b
camelia rakudo-moar 5fb81f: OUTPUT«1 2 3␤»
GLRelia rakudo-moar 741d14: OUTPUT«(1 2 3)␤»
JimmyZ m: my ($x, @b) = ('xxx', flat (1,2,3).map(*.Int).list); say @b 07:47
camelia rakudo-moar 5fb81f: OUTPUT«1 2 3␤»
GLRelia rakudo-moar 741d14: OUTPUT«(1 2 3)␤»
FROGGS nine: were there any discussions about NEXT/LAST phasers already?
ab6tract m: my ($x, @b) = 'xxx', flat (1,2,3).map(*.Int); say @b 07:48
camelia rakudo-moar 5fb81f: OUTPUT«1 2 3␤»
GLRelia rakudo-moar 741d14: OUTPUT«(1 2 3)␤»
ab6tract hmmm.. was wondering if those parens made the RHS a single-arg
nine FROGGS: not that I'm aware of. jnthn++ is teaching Perl 6 right now
FROGGS nine: yeah, seen that a few minutes ago... well, then I try to make my patch cleaner so it becomes applyable eventually 07:49
but first... spectest
JimmyZ m: my (@b) = (1,2,3).map(*.Int); say @b
camelia rakudo-moar 5fb81f: OUTPUT«1 2 3␤»
GLRelia rakudo-moar 741d14: OUTPUT«123␤»
JimmyZ m: my (@b, $x) = (1,2,3).map(*.Int), 'xxxx'; say @b 07:51
camelia rakudo-moar 5fb81f: OUTPUT«1 2 3 xxxx␤»
GLRelia rakudo-moar 741d14: OUTPUT«(1 2 3)xxxx␤»
colomon m: my $s = set < a b c >; say $s ∪ $s 07:52
camelia rakudo-moar 5fb81f: OUTPUT«set(a, c, b)␤»
GLRelia rakudo-moar 741d14: OUTPUT«(a c b) => True(a c b) => True␤»
colomon > my $s = set < a b c >; say $s ∪ $s 07:53
(a c b) => True(a c b) => True
07:54 Ven left
JimmyZ m: my (@b, $x) = (1,2,3).map(*.Int), 'xxxx'; say $x # how to assign the value to $x? 07:54
camelia rakudo-moar 5fb81f: OUTPUT«(Any)␤»
GLRelia rakudo-moar 741d14: OUTPUT«(Any)␤»
FROGGS m: my (@b, $x) = [(1,2,3).map(*.Int)], 'xxxx'; say $x # how to assign the value to $x? 07:55
camelia rakudo-moar 5fb81f: OUTPUT«(Any)␤»
GLRelia rakudo-moar 741d14: OUTPUT«(Any)␤»
FROGGS m: my (@b, $x) = [[(1,2,3).map(*.Int)], 'xxxx']; say $x # how to assign the value to $x?
camelia rakudo-moar 5fb81f: OUTPUT«(Any)␤»
GLRelia rakudo-moar 741d14: OUTPUT«(Any)␤»
FROGGS m: my (@b, $x) = [[(1,2,3).map(*.Int)], 'xxxx']; say @b
camelia rakudo-moar 5fb81f: OUTPUT«1 2 3 xxxx␤»
GLRelia rakudo-moar 741d14: OUTPUT«[1 2 3]xxxx␤»
FROGGS no idea
JimmyZ m: my (@b, $x) = $((1,2,3).map(*.Int)), 'xxxx'; say $x # how to assign the value to $x? 07:56
camelia rakudo-moar 5fb81f: OUTPUT«(Any)␤»
GLRelia rakudo-moar 741d14: OUTPUT«(Any)␤»
JimmyZ m: my (@b, $x) = $[(1,2,3).map(*.Int)], 'xxxx'; say $x # how to assign the value to $x?
camelia rakudo-moar 5fb81f: OUTPUT«(Any)␤»
GLRelia rakudo-moar 741d14: OUTPUT«(Any)␤»
07:56 espadrine joined
lizmat m: my $s = set < a b c >; say $s (|) $s 07:56
camelia rakudo-moar 5fb81f: OUTPUT«set(a, c, b)␤»
GLRelia rakudo-moar 741d14: OUTPUT«(a c b) => True(a c b) => True␤»
JimmyZ nine: ^^ 07:57
colomon m: my $s = set < a b c >; say $s.Set(:view).keys
camelia rakudo-moar 5fb81f: OUTPUT«a c b␤»
GLRelia rakudo-moar 741d14: OUTPUT«acb␤»
colomon m: my $s = set <a b c>;($s, $s).map( { .say } ) 07:58
GLRelia rakudo-moar 741d14: OUTPUT«a => Truec => Trueb => True␤a => Truec => Trueb => True␤»
camelia rakudo-moar 5fb81f: OUTPUT«set(a, c, b)␤set(a, c, b)␤»
nine JimmyZ: you can't and you never have been able to. An array on the left hand side will slurp all the remaining right hand isde
07:58 Ven joined
[ptc] FROGGS: wow, you're fast :-) Thanks for the merge! 07:59
nine JimmyZ: has always been that way, even in Perl 5
timotimo hoelzro: you know when i'm down to the lobby it's time for breakfast that's why it's called breakfast lobby UUH!
FROGGS [ptc]: thanks for the patch :o)
colomon m: my $s = set <a b c>;($s, $s).map( { .perl.say } )
camelia rakudo-moar 5fb81f: OUTPUT«set("a","c","b")␤set("a","c","b")␤»
GLRelia rakudo-moar 741d14: OUTPUT«set("a","c","b")␤set("a","c","b")␤»
07:59 xfix joined
timotimo heads out 08:00
[ptc] FROGGS: my pleasure :-) Hopefully I'll get a few more out today
FROGGS :o)
timotimo head out-ah, head on out-ah! head out-ah, head on out-ah!
colomon m: my $s = set <a b c>;($s, $s).map( { .WHAT.say } )
camelia rakudo-moar 5fb81f: OUTPUT«(Set)␤(Set)␤»
GLRelia rakudo-moar 741d14: OUTPUT«(Set)␤(Set)␤»
FROGGS timotimo: :D
timotimo get to the venue (head on out-ah!) like a perl6 dev (head on out-ah!)
dalek ast/glr: 6bb3a0a | (Stefan Seifert)++ | S04-statements/for.t:
Remove obsolete todo marker. We can iterate growing arrays now :)
08:02
colomon m: my $s = set <a b c>;($s, $s).map(*.Set(:view).keys).perl.say 08:05
camelia rakudo-moar 5fb81f: OUTPUT«("a", "c", "b", "a", "c", "b")␤»
GLRelia rakudo-moar 741d14: OUTPUT«(("a", "c", "b").Seq, ("a", "c", "b").Seq).Seq␤»
colomon m: my $s = set <a b c>;($s, $s).map(flat *.Set(:view).keys).perl.say 08:07
camelia rakudo-moar 5fb81f: OUTPUT«Cannot call map(Set: Set, WhateverCode); none of these signatures match:␤ ($: Whatever, *%_)␤ ($: &block, :$label, *%_)␤ in block <unit> at /tmp/1IBoSMnfHS:1␤␤»
GLRelia rakudo-moar 741d14: OUTPUT«Cannot call map(List: List); none of these signatures match:␤ ($: &block, :$label, *%_)␤ (HyperIterable:D $: &block, :$label, *%_)␤ in block <unit> at /tmp/demEm_WqkF:1␤␤»
colomon m: my $s = set <a b c>;($s, $s).map({flat $_.Set(:view).keys}).perl.say
camelia rakudo-moar 5fb81f: OUTPUT«("a", "c", "b", "a", "c", "b")␤»
GLRelia rakudo-moar 741d14: OUTPUT«(("a", "c", "b"), ("a", "c", "b")).Seq␤»
colomon m: my $s = set <a b c>;($s, $s).map({ $_.Set(:view).keys}).flat.perl.say 08:08
camelia rakudo-moar 5fb81f: OUTPUT«("a", "c", "b", "a", "c", "b")␤» 08:09
GLRelia rakudo-moar 741d14: OUTPUT«("a", "c", "b", "a", "c", "b").Seq␤»
colomon m: my $s = set <a b c>;($s, $s).map({ $_.Set(:view).keys}.lsip)..perl.say
camelia rakudo-moar 5fb81f: OUTPUT«5===SORRY!5=== Error while compiling /tmp/oDDny6drzM␤Undeclared routine:␤ perl used at line 1␤␤»
GLRelia rakudo-moar 741d14: OUTPUT«5===SORRY!5=== Error while compiling /tmp/FAfAOr2mNk␤Undeclared routine:␤ perl used at line 1␤␤»
colomon m: my $s = set <a b c>;($s, $s).map({ $_.Set(:view).keys}.slip)..perl.say
camelia rakudo-moar 5fb81f: OUTPUT«5===SORRY!5=== Error while compiling /tmp/PjVqEI9ICI␤Undeclared routine:␤ perl used at line 1␤␤»
GLRelia rakudo-moar 741d14: OUTPUT«5===SORRY!5=== Error while compiling /tmp/n4Pvno19Eu␤Undeclared routine:␤ perl used at line 1␤␤»
colomon m: my $s = set <a b c>;($s, $s).map({ $_.Set(:view).keys.slip}).perl.say
camelia rakudo-moar 5fb81f: OUTPUT«Method 'slip' not found for invocant of class 'List'␤ in block <unit> at /tmp/EPXAuw3lCc:1␤␤»
GLRelia rakudo-moar 741d14: OUTPUT«Method 'slip' not found for invocant of class 'Seq'␤ in block <unit> at /tmp/gX9hsAkarc:1␤␤»
colomon m: my $s = set <a b c>;($s, $s).map({ $_.Set(:view).keys.Slip}).perl.say 08:10
camelia rakudo-moar 5fb81f: OUTPUT«Method 'Slip' not found for invocant of class 'List'␤ in block <unit> at /tmp/T1SKRUHyaU:1␤␤»
GLRelia rakudo-moar 741d14: OUTPUT«("a", "c", "b", "a", "c", "b").Seq␤»
dalek kudo/glr: c344b26 | (Stefan Seifert)++ | src/core/Array.pm:
Fix STOREing junctions into arrays
08:13
08:13 Ven left
JimmyZ hmm, jnthn++ is teaching? I didn't see the scheule on the spw site 08:14
08:15 Ven joined
FROGGS JimmyZ: he's in Bernstein 08:15
08:16 itz left 08:17 domidumont left
JimmyZ oh, that room 08:17
dalek kudo-star-daily: da61f73 | coke++ | log/ (9 files):
today (automated commit)
rl6-roast-data: 50a5a30 | coke++ | / (2 files):
another glr run for 'today'
08:19 _itz_ joined
dalek kudo/glr: 1a2b583 | TimToady++ | src/core/operators.pm:
just iterate Seq value, don't test first
08:19
08:20 _itz left 08:21 itz joined
[Coke] nqp-p: we're keeping the lights on there. tests are run every day to insure it doesn't get any more bitrotted. I'd rather see you add something than rip that out. 08:22
08:22 camelia left
[Coke] (github.com/coke/perl6-roast-data/b...mmary.out) (some tests are failing, but it is mostly in good health) 08:23
08:23 RabidGravy joined, camelia joined
camelia nqp-parrot: OUTPUT«Can't exec "./rakudo-inst/bin/nqp-p": No such file or directory at lib/EvalbotExecuter.pm line 193.␤exec (./rakudo-inst/bin/nqp-p /tmp/HY3ZRsY9n7) failed: No such file or directory␤Server error occurred! Closing Link: ns1.niner.name (Quit: camelia)␤Lost conn…» 08:25
08:25 ChanServ sets mode: +v camelia
ShimmerFairy [Coke]: that's nice to hear, but does anyone actually use it? I don't see much value in maintaining something nobody uses. 08:25
[Coke] no. and we're not maintaining it. 08:26
not actively, anyway.
if camelia isn't going to build it, it shouldn't be an option for the bot, but I don't see a need to remove it from the repo. 08:27
FROGGS m: for 42, 43 { LAST { say "ohh noes" }; .say; last if $_ == 44 } 08:29
GLRelia rakudo-moar 1a2b58: OUTPUT«42␤43␤»
camelia rakudo-moar 5fb81f: OUTPUT«42␤43␤ohh noes␤»
08:29 Ven left
ShimmerFairy [Coke]: for the record, despite my personal misgivings on Parrot, if it's shown that nqp-p is still be used by people, I'm fine with keeping it. But I don't want to keep it around just because it's not actively harmful :) 08:29
08:30 rurban joined
colomon m: my $a = set < a b c >; my $b = set < b d f >; say ($a (-) $b).perl 08:30
GLRelia rakudo-moar 1a2b58: OUTPUT«set("a","c")␤»
camelia rakudo-moar 5fb81f: OUTPUT«set("a","c")␤»
[Coke] Technically, I'm using it. Thanks.
ShimmerFairy Running a daily test just to see if it functions doesn't really count for me. 08:33
moritz m: my @a = [1, 2], [3, 4]; say @a>>.shift 08:34
GLRelia rakudo-moar 1a2b58: OUTPUT«Method 'shift' not found for invocant of class 'Int'␤ in block <unit> at /tmp/mFYuOw06Y1:1␤␤»
camelia rakudo-moar 5fb81f: OUTPUT«1 3␤»
[Coke] m: / $. / 08:35
GLRelia rakudo-moar 1a2b58: OUTPUT«Method 'match' not found for invocant of class 'Any'␤ in block <unit> at /tmp/TFShklg1xA:1␤␤»
timotimo does "will first" and such work with while loops in glr at the moment?
camelia rakudo-moar 5fb81f: OUTPUT«Method 'match' not found for invocant of class 'Any'␤ in block <unit> at /tmp/d3k6Y2ajJ4:1␤␤»
moritz huh, but shift is marked as nodal
timotimo: I don't think so (not sure though)
colomon m: say (:I'm=>1,:your=>1,:isn't=>1,:day=>1,:it=>1,:afraid=>1).Bag
GLRelia rakudo-moar 1a2b58: OUTPUT«(isn't => True) => 1(afraid => True) => 1(I'm => True) => 1(day => True) => 1(it => True) => 1(your => True) => 1␤»
camelia rakudo-moar 5fb81f: OUTPUT«bag(afraid => True, day => True, isn't => True, it => True, your => True, I'm => True)␤»
moritz timotimo: there's a spectest file for it; see if it passes? 08:36
nine timotimo: sounds like a question for FROGGS++
[Coke] ShimmerFairy: I'm sorry my efforts don't count for you. I'll try not to take it personally. :)
ab6tract ShimmerFairy: with all due respect, it feels a bit awkward that you are using such an authoritative voice here
timotimo because reinstating the "turn a for loop on a range into a while loop if we can" gives me many more failures
like a for loop over a range with will first, next, last, keep and undo causing broken tests
nine moritz: the reason is probably me not really knowing what the nodal thing is about, so I may have missed something there
FROGGS timotimo: I am working on phaser handling for for loops atm 08:37
timotimo oh, turning optimize off doesn't change the failure count
ah! 08:38
ShimmerFairy [Coke]: nothing against you, I just don't think the only use case I've heard being a test run is very convincing for the "does anyone use it?" (alternatively, "does anyone still need nqp-p specifically?") argument :)
ab6tract: I don't understand, what authoritative voice? (I'm voicing a strong opinion, but I'm not claiming any authority on anything)
timotimo my 'net connection is unhappy :( 08:39
dalek kudo/glr: 0498aca | moritz++ | src/core/Supply.pm:
Partially fix Supply.zip

We now pass 8 out of 10 tests for it (used to abort after two tests)
08:39 Ven joined
ab6tract ShimmerFairy: probably my interpretation then. the way you spoke about the repo gave me the sense that it was your choice in the end. anyway, I like your opinions in general, even if not this specific case, so please don't take my criticism the wrong way :) 08:41
moritz fwiw I'd be happy to remove nqp-p, but I also respect peoples wish to resurrect rakudo-p eventually, and don't want to cause them any extra work 08:42
ShimmerFairy ab6tract: ah, sorry to accidentally give that impression :) . The most I could do is one of those "forgiveness > permission" type commits; I certainly have no authority to keep it that way though :P
dalek kudo/glr: 201a509 | colomon++ | src/core/set_operators.pm:
Fix set intersection for the normal set case.

Desperately needed a strategic .Slip.
ab6tract ShimmerFairy: :) 08:43
ShimmerFairy moritz: I understand that argument, however is there any reason why making someone dive into nqp's commit history for nqp-p is much worse than diving into rakudo's for rakudo-p ?
nine [Coke]: rt.perl.org/Public/Bug/Display.html?id=113026 is fixed by GLR, I gather you can close it?
moritz ShimmerFairy: please don't do that; deletions on that scale affect repo size negatively (especially if reverted afterwards)
08:45 rurban left
JimmyZ cloning nqp is already slow :P 08:46
ShimmerFairy And I fully admit that my personal experience with parrot makes me want to wipe out any and all traces of nqp-p with glee, whereas most others are content with letting it sit there :)
08:46 rurban joined
ShimmerFairy (though I can't imagine why anyone would want to give off the illusion that parrot is still supported anywhere in the rakudo toolchain :P) 08:47
08:49 xyf joined
JimmyZ I think we cant delete history since parrot fed nqp ;) 08:50
moritz nor the illusion that parrot is still actively developed or maintained
08:50 diana_olhovik_ left
colomon m: my $b = bag <a b c a c>; dd $b; say ($b.keys.perl) 08:52
GLRelia rakudo-moar 201a50: OUTPUT«$b = ("a"=>2,"c"=>2,"b"=>1).Bag␤("a", "c", "b").Seq␤»
camelia rakudo-moar 5fb81f: OUTPUT«$b = ("a"=>2,"c"=>2,"b"=>1).Bag␤("a", "c", "b")␤»
ShimmerFairy Well, I do appreciate the historical significance of Parrot, but I have no qualms about getting rid of it from nqp and finally letting it retire to the history books of P6 :)
colomon m: my $b = bag <a b c a c>; dd $b; say $b.keys 08:53
GLRelia rakudo-moar 201a50: OUTPUT«$b = ("a"=>2,"c"=>2,"b"=>1).Bag␤acb␤»
camelia rakudo-moar 5fb81f: OUTPUT«$b = ("a"=>2,"c"=>2,"b"=>1).Bag␤a c b␤»
JimmyZ we can have nqp-old repo.
colomon m: my $s = set < a b c >; say $s.Bag.perl
GLRelia rakudo-moar 201a50: OUTPUT«(:c=>1,:a=>1,:b=>1).Bag␤»
camelia rakudo-moar 5fb81f: OUTPUT«("a"=>1,"c"=>1,"b"=>1).Bag␤»
colomon m: my $s = set < a b c >; say $s.Bag.keys 08:54
GLRelia rakudo-moar 201a50: OUTPUT«b => Truea => Truec => True␤»
camelia rakudo-moar 5fb81f: OUTPUT«a c b␤»
JimmyZ have a
colomon m: my $s = set < a b c >; say $s.Bag.keys.perl
GLRelia rakudo-moar 201a50: OUTPUT«(:a, :b, :c).Seq␤»
camelia rakudo-moar 5fb81f: OUTPUT«("a", "c", "b")␤»
ShimmerFairy JimmyZ: yeah, if others agreed to get nqp-p out of the repo, I think a separate repo for easy access to nqp-p for interested resurrectors would be nice :) 08:55
(though I still haven't heard why purging nqp-p would be so much worse than purging rakudo-p was, for possible resurrections) 08:56
colomon m: my $s = set < a b c >; say $s.Bag.kv.perl
GLRelia rakudo-moar 201a50: OUTPUT«((:b, 1), (:a, 1), (:c, 1)).Seq␤»
camelia rakudo-moar 5fb81f: OUTPUT«(("a", 1), ("c", 1), ("b", 1))␤»
[Coke] can glr folks review rt.perl.org/Ticket/Display.html?id=112716 ?
dalek pan style="color: #395be5">perl6-examples: f4282ff | (Steve Mynott)++ | / (4 files):
add 15interactivity to HTML and tweak it
JimmyZ ShimmerFairy: becase of bootstrap
which increases the repo size
ShimmerFairy JimmyZ: oh yeah, I suppose regenerating nqp-p's bootstrap would be interesting when the last copy is hidden deep in the git log somewhere :) 08:57
moritz ShimmerFairy: well, purging it once we know for sure it's dead is fine; but purging and then later reverting the commit would be bad 08:58
and yes, it's bootstrapping that mostly contributes to NQP's repo size
JimmyZ I think it is is a pain to keep it in nqp 08:59
timotimo so ... with --optimize=off i get the very same spectest failures apparently
JimmyZ is not
08:59 Loren joined
timotimo not only apparently 08:59
ShimmerFairy moritz: and I'm guessing that rakudo-p wasn't an issue because either A) there's no bootstrap binary to worry about, and/or B) bringing back parrot would be able to use almost nothing of the old implementation on the rakudo side
JimmyZ most people cares rakudo
moritz ShimmerFairy: yes 09:00
ShimmerFairy moritz: A or B or both, ooc? :)
moritz ShimmerFairy: the idea was that if r-p was brought back, nqp-p would first need to be updated to avoid all the parrot special cases
ShimmerFairy: both
[Coke] my glr test runs keep flapping with: unhandled Failure detected in DESTROY 09:01
FROGGS now I think my NEXT/LAST patch finally gets shape
[Coke] they have come and gone multiple times.
09:02 diana_olhovik_ joined
dalek kudo/glr: 5996d83 | timotimo++ | src/Perl6/Optimizer.nqp:
re-instate for -> while opt

we used to generate a call to &infix:<,> that's no longer there.
09:02
timotimo now i can run a proper benchmark comparison
dalek rl6-roast-data: 168f3c6 | coke++ | / (2 files):
another glr run for 'today'
09:03
FROGGS timotimo: does your patch involve spectest pro/re-gressions?
ShimmerFairy moritz: Like I said, it's just my personal feelings that make me inclined to delete nqp-p. If there are real reasons to keep it around, I'll probably survive :P .
09:04 llfourn joined
ShimmerFairy (and I only remembered it because I was making Makefile changes to nqp the other day, remembered parrot at the last second, and when I saw the parrot makefile was different enough from JVM and MVM I couldn't care enough to fix things there too.) 09:04
timotimo FROGGS: i ran the spectests regularly and with --optimize=off and there was not a single difference
[Coke] There are reasons to keep it around, yes.
timotimo huh, actually ...
is it enough to put --optimize=off at the very end of the harness?
oh, hm. i may have done it wrong
FROGGS timotimo: cool!
I was fearing bad things ;o) 09:05
[Coke] one of the destroy errors is right at the top of github.com/coke/perl6-roast-data/c...168f3c6bef 09:06
timotimo it's not an error, it's a warning
[Coke] 179 failures left on glr branch.
timotimo that's a bit more than it was yesterday :| 09:07
[Coke] shouldn't be visible in the test output, though.
timotimo: it's been steadily going down, I thought.
(that's tests, not files)
09:07 rindolf joined 09:08 colomon left
tadzik I wonder if those failures will still be there by the time I build :) 09:08
09:09 colomon joined, llfourn left
timotimo that run doesn't contain my last "for -> while opt" commit, right? it'd take you longer than 2 minutes to run a full spectest thingie, right? 09:09
tadzik prolly
[Coke] I'm doing another rn right now 09:10
up to S03
sadly the test_summary thing is still one-test-at-a-time
dalek kudo/glr: 766e702 | (Stefan Seifert)++ | src/core/Any (2 files):
Fix for $a -> $v is rw { $v = 1 }
09:11
[Coke] crap, now I'm already behind again. :)
timotimo so, env TEST_JOBS=4 /usr/bin/perl t/harness --fudge --moar --keep-exit-code --tests-from-file=t/spectest.data -- --optimize=off 09:12
will this pass --optimize=off to perl6?
tadzik ¯\_(ツ)_/¯ 09:13
timotimo ah
of course it doesn't
why did i even think "@pass-through-options" would do that?
FROGGS \o/ 09:14
tadzik /o\
FROGGS looks like my next/last phaser implementation does work now!
tadzik noice
FROGGS and the patch does not look that bad either
tadzik: grab your own noice :o) 09:15
tadzik I read that as "next laser phaser implementation"
timotimo FROGGS: i'm having serious difficulty figuring out if my patch actually makes things worse or not :(
FROGGS :/ 09:16
timotimo now i've reverted the patch locally
colomon m: my $b = bag < a b c c >; say $b.keys.perl
GLRelia rakudo-moar 766e70: OUTPUT«("a", "c", "b").Seq␤»
camelia rakudo-moar 5fb81f: OUTPUT«("a", "c", "b")␤»
timotimo and i'll run the spec tests yet again
dalek ast/glr: a231a63 | (Stefan Seifert)++ | S04-statement-modifiers/if.t:
Fix test for Empty broken by relying on stringification
09:17
laben tadzik: i found which other problem broght upon panda by json::fast broke dem deps: it returns an $[] instead of [] for arrays and that doesn't get flattened as one would expect
dalek ast/glr: 3bcaad9 | (Stefan Seifert)++ | S04-statement-modifiers/with.t:
Fix another test for Empty broken by relying on stringification
09:18
tadzik timotimo: ^
laben tadzik: i fixed that by explicitly making them arrays with @() but i still have not fixed rebootstrap
[Coke] rt.perl.org/Ticket/Display.html?id=113026 - another GLR ticket - still todo'd in glr branch. Not sure if request makes sense in GLR, needs review. 09:19
colomon m: (:isn't=>1,:your=>1,:afraid=>1,:it=>1,:I'm=>1,:day=>1).Bag.perl.say
GLRelia rakudo-moar 766e70: OUTPUT«(:afraid=>1,:isn't=>1,:it=>1,:I'm=>1,:your=>1,:day=>1).Bag␤»
camelia rakudo-moar 5fb81f: OUTPUT«(:isn't=>1,:afraid=>1,:it=>1,:your=>1,:I'm=>1,:day=>1).Bag␤»
colomon m: (:isn't=>1,:your=>1,:afraid=>1,:it=>1,:I'm=>1,:day=>1).Bag.keys.say
GLRelia rakudo-moar 766e70: OUTPUT«your => Trueisn't => Trueafraid => Trueit => TrueI'm => Trueday => True␤»
camelia rakudo-moar 5fb81f: OUTPUT«I'm => True afraid => True isn't => True your => True it => True day => True␤»
tadzik laben: maybe we can get a test for both JSONs and make sure they behave alike?
laben: then perhaps it can be fixed on the JSON side 09:20
since J::F is to be a drop-in replacement for J::T they should work exactly the saem
laben tadzik: they most definitely dont with glr in mind, as per the difference in flattening 09:21
dalek ast/glr: ea09191 | (Stefan Seifert)++ | S04-statements/if.t:
Fix more tests trying to test for empty lists
tadzik laben: right
and I think they should :)
laben: do you have a testcase in handy?
laben tadzik: for json parsing? afraid not 09:22
tadzik: anyway, there's another strange problem which i'm trying to locate now that makes the deps from an array into an array of pairs...
tadzik laben: waiwait :)
laben: you said jsonfast returns $[] instaed of [] as jsontiny does 09:23
that's a testcase :)
let's add it to testsuite
[Coke] timotimo: can you review rt.perl.org/Ticket/Display.html?id=81974 in light of the glr?
FROGGS m: $_ = ""; / $. /
GLRelia ( no output )
camelia ( no output )
FROGGS [Coke]: ^^
xyf where can i find doc about DateTime?
laben tadzik: ofc that is, but i did not write it down as a test. btw you can add also testing of empty arrays? Tiny returned [] while Fast returns Any
tadzik laben: yeah, let's add that 09:24
it's definitely needed
timotimo damn 09:25
FROGGS: now i have a diff between spec test runs :(
moritz xyf: doc.perl6.org/type/DateTime
nine [Coke]: RT #113026 is fixed by GLR! 09:26
synbot6 Link: rt.perl.org/rt3/Public/Bug/Display...?id=113026
09:26 CIAvash joined
[Coke] m: :2 09:26
GLRelia rakudo-moar 766e70: OUTPUT«5===SORRY!5=== Error while compiling /tmp/hFXXIjJB2o␤Malformed radix number␤at /tmp/hFXXIjJB2o:1␤------> 3:27⏏5<EOL>␤ expecting any of:␤ number in radix notation␤»
camelia rakudo-moar 5fb81f: OUTPUT«5===SORRY!5=== Error while compiling /tmp/YOg3av1qn0␤Malformed radix number␤at /tmp/YOg3av1qn0:1␤------> 3:27⏏5<EOL>␤ expecting any of:␤ number in radix notation␤»
[Coke] m: :2()
GLRelia rakudo-moar 766e70: OUTPUT«This call only converts base-2 strings to numbers; value () is of type List, so cannot be converted!␤(If you really wanted to convert () to a base-2 string, use ().base(2) instead.)␤ in block <unit> at /tmp/77_BQcomON:1␤␤»
camelia rakudo-moar 5fb81f: OUTPUT«This call only converts base-2 strings to numbers; value $( ) is of type Parcel, so cannot be converted!␤(If you really wanted to convert $( ) to a base-2 string, use $( ).base(2) instead.)␤ in block <unit> at /tmp/RPW0YbHDh6:1␤␤»
[Coke] m: say :2(1)
camelia rakudo-moar 5fb81f: OUTPUT«This call only converts base-2 strings to numbers; value 1 is of type Int, so cannot be converted!␤(If you really wanted to convert 1 to a base-2 string, use 1.base(2) instead.)␤ in block <unit> at /tmp/Ry2ZL63zTB:1␤␤»
[Coke] m: say :2<1> 09:27
camelia rakudo-moar 5fb81f: OUTPUT«1␤»
GLRelia rakudo-moar 766e70: OUTPUT«This call only converts base-2 strings to numbers; value 1 is of type Int, so cannot be converted!␤(If you really wanted to convert 1 to a base-2 string, use 1.base(2) instead.)␤ in block <unit> at /tmp/qcOUhRV4ga:1␤␤»
rakudo-moar 766e70: OUTPUT«1␤»
timotimo FROGGS: gist.github.com/timo/84a1acb434ab2b5cccc5
does this look worrisome?
nine [Coke]: and I removed the todo marker in the spec tests
moritz colomon: since you've been working on sets: t/spec/integration/advent2012-day13.t has a failing test for set difference
[Coke] nine: ok, but the test that was written is still todo'd in the glr roast branch.
... ok, it was still marked TODO there. passing?
colomon moritz: we’re working on set difference right this moment.
(well, or talking about Paul Simon) 09:28
nine timotimo: yes, worrysome
[Coke] nine: Danke.
09:28 lizmat left
timotimo shit. 09:28
nine [Coke]: it's only marked todo for niecza now
laben tadzik: silly me, i did change deps calc in Ecosystem.pm but did not change them in Panda.pm (get-project-from...), testing now
09:29 lizmat joined
[Coke] :2<1> is a glr bug, looks like. 09:30
FROGGS timotimo: that might not your patch
moritz m: say (:2<1>) 09:31
GLRelia rakudo-moar 766e70: OUTPUT«1␤»
camelia rakudo-moar 5fb81f: OUTPUT«1␤»
FROGGS timotimo / nine: I also get t/spec/S17-supply/zip.t aborting locally
timotimo t.h8.lv/p6bench/2015-08-29-nom_vs_glr_2.html - with the for optimization in place, the benchmarks aren't all so terrible any more
array access (assignment, mostly) is still very slow. as is the X metaop
FROGGS ohh no, it was unhappy before too 09:32
cdc_ m: say (:16<a>)
GLRelia rakudo-moar 766e70: OUTPUT«10␤»
camelia rakudo-moar 5fb81f: OUTPUT«10␤»
FROGGS I'm going to push my test because it is making three test files pass
s/test/patch/ 09:33
tadzik while_array_set looks a lot slower
and for_ :) 09:34
JimmyZ looks like push join is slower
09:34 rurban left
timotimo our array_set_xx could become a ton faster if we implement the xx iterator's push_all or what it's called method 09:34
laben tadzik: i had enough of strange set semantics, i rewrote the deps with lists/arrays and here it is gist.github.com/ec4368a804fbadeab11a 09:35
09:35 Ven left
dalek ast/glr: 2a1e0eb | coke++ | S02-literals/radix.t:
Add tests for RT #112728
09:35
synbot6 Link: rt.perl.org/rt3/Public/Bug/Display...?id=112728
laben tadzik: makes rebootstrap work. btw if i were you, i would consider bringing the project creation code in a single place 09:36
tadzik laben: yes, that's the plan
it became a mess over time :)
not a fan
dalek kudo/glr: d65ca02 | FROGGS++ | src/core/Any-iterable-methods.pm:
make NEXT/LAST phaser firing work
[Coke] m: say 2**10000000000; say 1/(2**10000) 09:37
GLRelia rakudo-moar 766e70: OUTPUT«0␤0␤»
camelia rakudo-moar 5fb81f: OUTPUT«0␤0␤»
tadzik but honestly I quite like the idea of using sets here, it's just that they got so confusing for some reason now that glr's in
FROGGS m: $_ = "a"; say / $. /
GLRelia rakudo-moar 766e70: OUTPUT«/ $. /␤»
camelia rakudo-moar 5fb81f: OUTPUT«/ $. /␤»
tadzik I'm happy with what makes rebootstrap work though :)
colomon m: say +True
GLRelia rakudo-moar 766e70: OUTPUT«1␤»
camelia rakudo-moar 5fb81f: OUTPUT«1␤»
FROGGS m: $_ = "a"; say m/ $. /
GLRelia rakudo-moar 766e70: OUTPUT«Nil␤»
camelia rakudo-moar 5fb81f: OUTPUT«Nil␤»
colomon m: say True.Int
GLRelia rakudo-moar 766e70: OUTPUT«1␤»
camelia rakudo-moar 5fb81f: OUTPUT«1␤»
FROGGS m: $_ = "a"; say "a" ~~ / $. / 09:38
GLRelia rakudo-moar 766e70: OUTPUT«Nil␤»
camelia rakudo-moar 5fb81f: OUTPUT«Nil␤»
laben tadzik: i can understand the feeling, it's just that the transition to j::f plus new set semantics from glr made it explode badly
tadzik right
I'm onto the json differences right now
09:38 xyf left, Ven joined
dalek rl6-roast-data: 567bd23 | coke++ | / (2 files):
another glr run for 'today'
09:39
09:39 rurban joined
tadzik laben: the patch still works on nom, right? 09:39
asking just in case :) 09:40
dalek ast/glr: d8a7786 | coke++ | S02-types/set.t:
unfudge passing test
laben tadzik: i'll test it
tadzik awesome, thank you
[Coke] 163 failures, running again... 09:41
tadzik laben: actually, now that I tried it, JSON::Fast doesn't work on glr anymore
fails tests
I think I'll fallback to JSON::Tiny
laben tadzik: wait a sec, did you get the latest json::fast?
tadzik I... think
yeah, the subtree may not be up-to-date 09:42
laben tadzik: subtree not updated, does not have the latest fix. i was just going to ask you about this: is it needed to have the subtrees directly in the repo instead of git submodules? 09:43
09:43 rmgk left
tadzik yes 09:43
submodules were a much bigger pain to work with :)
ab6tract gist.github.com/ab5tract/ebda98cc14e25881adbd
tadzik updated it right now
ab6tract colomon ^^ 09:44
colomon ab6tract: on it
tadzik laben: pushed
09:44 rmgk joined
timotimo ah, so xx build a List, and it seems like then it gets passed on to push-until-lazy, which is responsible for 50% of the whole time the benchmark runs 09:44
laben tadzik: else, how do i tell (re)bootstrap to go look for the modules source in another location instead of the remote? 09:45
tadzik you don't
I mean, it can't do that
but it's an interesting idea
FROGGS m: while 42 { FIRST { say "hello" }; last }
GLRelia ( no output )
camelia ( no output )
laben :( it was possible with bootstrap before
tadzik it was?
I mean, it's not a bootstrap thing 09:46
rebootstrap does bootstrap and then tells panda to reinstall what was there before
FROGGS m: my $x; while 42 { FIRST { say "hello" }; last if $x++ > 3 }
GLRelia ( no output )
camelia ( no output )
tadzik and panda doesn't even have the copy of the files from remotes
laben yes, instead of calling "bin/panda File::Find etc" i modded it to call "bin/panda dir/to/file/find ..."
tadzik oh, you can do that alright 09:47
but it's not a built-in feature
[Coke] m: DateTime.new('2015-08-23t02:27:33-07:00');
GLRelia ( no output )
camelia ( no output )
tadzik you could put a hook for Ecosystem.update that mangles projects.json to your liking
and turns all those git:// into file:// or so 09:48
[Coke] b: DateTime.new('2015-08-23t02:27:33-07:00');
tadzik laben: btw I don't see the json-fast difference that you mentioned 09:49
laben the idea is, generally i use bootstrap to try an env or commit that may not be compatible with remote repos, which means i first modify the deps then call bootstrap on the modded deps and local (modded) panda
tadzik gist.github.com/tadzik/b10f4469b0adc2c16cf1
well, bootstrap itself shouldn't touch network anyway
rebootstrap might
dalek ast: 28e4d44 | coke++ | S32-temporal/DateTime.t:
Add tests for RT #125872
kudo/glr: 898d0c7 | colomon++ | src/core/Setty.pm:
ab6tract++'s Setty.Bag and Setty.Baghash.
synbot6 Link: rt.perl.org/rt3/Public/Bug/Display...?id=125872
FROGGS m: say ([Z] () eq []).perl
GLRelia rakudo-moar d65ca0: OUTPUT«((Bool::True)).Seq␤»
camelia rakudo-moar 5fb81f: OUTPUT«((Bool::True,),)␤»
09:49 itz left
laben tadzik: ? afaics, it uses the bundled deps only to call panda on site, the ones it installs are fetched by net? 09:50
tadzik hmmmmm 09:51
I don't think so
see projects.json.bootstrap
FROGGS m: use Test; is ([Z] ()), [], "[Z] () eq []";
GLRelia rakudo-moar d65ca0: OUTPUT«not ok 1 - [Z] () eq []␤␤# Failed test '[Z] () eq []'␤# at /tmp/HjIoyxBm50 line 1␤# expected: ''␤# got: (Nil)␤»
camelia rakudo-moar 5fb81f: OUTPUT«ok 1 - [Z] () eq []␤»
timotimo i think in order to make xx any better, i'd have to build a custom iterator
FROGGS m: use Test; is-deeply ([Z] ()), [], "[Z] () eq []";
GLRelia rakudo-moar d65ca0: OUTPUT«not ok 1 - [Z] () eq []␤␤# Failed test '[Z] () eq []'␤# at /tmp/5qF3z7yIww line 1␤# expected: []␤# got: Nil␤»
camelia rakudo-moar 5fb81f: OUTPUT«not ok 1 - [Z] () eq []␤␤# Failed test '[Z] () eq []'␤# at /tmp/RywMh5LK8y line 1␤# expected: []␤# got: $()␤»
colomon m: my $s = set <a b c> ; say ($s (^) $s).perl
GLRelia rakudo-moar d65ca0: OUTPUT«set(("a", "c", "b").Seq,("a", "c", "b").Seq)␤»
camelia rakudo-moar 5fb81f: OUTPUT«set()␤»
laben timotimo: you could take inspiration from DwimIterator by nine++ in HYPER 09:52
timotimo isn't that quite a bit too complicated? :)
laben ok maybe i grew accostumed to it, did not seem to complicated
s/to/too/
timotimo surprisingly (IMO) i'm not able to get a slip into the LHS of xx without it going for the thunked version of xx 09:53
that handles the flattening correctly
the other variant just creates a List that has the thing on the LHS bindpos'd to each slot from 0 to ^$n
laben tadzik: then, it remains a single doubt (not much of an issue). why does panda precomp itself, even if it doesnt for any modules? 09:55
tadzik laben: it doesn't 09:56
09:56 avalenn_ left
tadzik is it possible that you have a leftover blib? 09:56
dalek kudo/glr: 05ce1e8 | colomon++ | src/core/set_operators.pm:
ab6tract++'s symmetric difference fix.

Just needed a slip.
timotimo huh 09:57
actually
it's more like 75% of the whole time is spent copying the list from the xx operator to the @a 09:58
moritz that's good
because it means there's an opportunity for optimization
laben tadzik: you're right, but then it shouldn't copy the .moarvm files 09:59
timotimo yeah
96.65% STORE
3.33% infix:<xx>
tadzik laben: yeah, it just copies everything from blib :/
09:59 Loren left
tadzik in case it's put there by the custom build hook or so 09:59
laben but if there is a custom build hook, let them put a custom install filter 10:00
10:01 Alina-malina joined
tadzik hm, what kind of custom install folder do you mean? 10:01
Ven m: $_ = 'a'; if !/a/ { 1 } # does anyone know what's going on? works with "unless" :)
GLRelia rakudo-moar 898d0c: OUTPUT«Method 'match' not found for invocant of class 'Any'␤ in block <unit> at /tmp/bq8kH5dF39:1␤␤»
camelia rakudo-moar 5fb81f: OUTPUT«Method 'match' not found for invocant of class 'Any'␤ in block <unit> at /tmp/1iOm6wsrNa:1␤␤»
10:02 itz joined
moritz m: $_ = 'a'; if !.say { 1 } 10:03
GLRelia rakudo-moar 05ce1e: OUTPUT«a␤»
camelia rakudo-moar 5fb81f: OUTPUT«a␤»
laben filter, not folder. so by default you only copy .p[ml][6]? and if they want to also copy .so etc, they pass a option or make a custom hook to do that
moritz m: $_ = 'a'; if !m/a/ { 1 } 10:04
GLRelia ( no output )
camelia ( no output )
moritz Ven: I have a hunch
[Coke] colomon: where are you? I have your bag.
colomon [Coke]: thank you! in the courtyard 10:05
masak could someone with Mac OS X try something for me?
colomon masak: like what?
masak run `p6doc List.map` and see if it fails?
moritz Ven: Regex.Bool basically looks for the caller's $_
colomon masak: p6doc: command not found
how do I install it?
masak colomon: panda
moritz Ven: and somehow the ! seems to introduce a callframe that makes it search for the wrong $_
colomon masak: I’m on GLR 10:06
masak ok :/
(it works for me on my Linux laptop, but it fails for my neighbor who has Mac OS X)
ab6tract masak: it works for me
tadzik colomon: panda should work on glr
ab6tract let me panda install the latest
masak ab6tract: cool.
10:07 larion joined
colomon tadzik: ooooo 10:08
ab6tract again it works :/
let me rakudobrew the latest nom :)
dalek rl6-roast-data: 8b10476 | coke++ | / (2 files):
another glr run for 'today'
10:10
[Coke] ugh. I feel like we've closed more tickets than this. only at 1,039
dalek kudo/glr: 00a34b8 | colomon++ | src/core/set_operators.pm:
set, bag, and mix composers fixed in GLR.

  ab5tract++'s patch.
masak ab6tract++ 10:11
it's weird that it behaves differently on our two systems -- and worth checking out ;)
[Coke] down to 144 failures in glr, starting another run
r: sub s_s(*%n) { %n>>.say }; s_s(|{:assoc<list>},:assoc<left>); 10:12
GLRelia rakudo-moar 05ce1e: OUTPUT«list␤»
..rakudo-jvm 05ce1e: OUTPUT«Can't open perl script "/home/steve/rakudo-inst/bin/eval-client.pl": No such file or directory␤»
camelia rakudo-jvm 5fb81f: OUTPUT«left␤»
..rakudo-moar 5fb81f: OUTPUT«list␤»
masak 144 *test* failures?
not test files? 10:13
[Coke] masak: yes.
moritz masak: we're below 144 test files for days :-)
FROGGS should be <= 30 test files
[Coke] here's the listing: github.com/coke/perl6-roast-data/b....out#L5263
and I just saw a set/bag fix go in, so should be even better after next run completes.
colomon yes, ab6tract++ fixed a bunch of stuff in set/bag 10:14
ab6tract masak: works just fine here :(
moritz in 8 hours we'll be able to give the remaining bugs names :-)
dalek ast/glr: 787837c | lizmat++ | S02-types/sethash.t:
Make S02-types/sethash.t clean
masak ab6tract: thank you, that's a data point
[Coke] crap. killing and starting over to catch that one. :)
moritz lizmat: fwiw I've given up on the remaining Supply.zip test failures; maybe a job for you? 10:15
10:15 oetiker left
dalek ast: b19f209 | hoelzro++ | S17-procasync/many-processes-no-close-stdin.t:
Add test to make sure we release stdin on child exit
10:16
ab6tract [Coke]: lizmat++ is about working on a mixhash.t test, maybe you want to wait for that one too :D
[Coke] when flattening duplicate named args, does the one that gets picked matter?
10:16 Ven left
moritz nah, don't wait, just run another later :-) 10:17
[Coke]: iirc the last one wins
[Coke] m: (my @ = 'a', <b c>)[1].perl.say; (constant @ = 'a', <b c>)[1].perl.say;
GLRelia rakudo-moar 00a34b: OUTPUT«("b", "c")␤("b", "c")␤»
camelia rakudo-moar 5fb81f: OUTPUT«"b"␤("b", "c")␤»
dalek kudo/nom: fd1612f | hoelzro++ | t/spectest.data:
Add many-processes-no-close-stdin.t to spectest.data
moritz m: sub f(:$a) { say $a }; f :a :!a
GLRelia rakudo-moar 00a34b: OUTPUT«False␤»
camelia rakudo-moar 5fb81f: OUTPUT«False␤»
10:22 Ven joined
[Coke] if someone tells me what file tests for RT #122895 belong in, I'll write them. 10:23
synbot6 Link: rt.perl.org/rt3/Public/Bug/Display...?id=122895
[Coke] S02-types/lists.t ? 10:24
timotimo with the for -> while optimization and putting @a[ SCALE ] = SCALE in front of the @a[ $_ ] = $_ for 0 .. SCALE loop, we get as fast as nom was
btyler_ design.perl6.org/S04.html#line_1380 <-- lizmat
[Coke] ah, FROGGS++ #IRL
timotimo but how can nom be as fast as glr is now without knowing up front how big the array is going to be?
colomon [Coke]: or S04-declarations/constant.t ?
hoelzro FROGGS, jnthn: if one of you gets a chance today, could you try Linenoise on Windows? I'd like to make sure it works =) 10:25
10:26 YuviPanda is now known as PastJynus
FROGGS hoelzro: I can just copy the Linenoise.pm into the install/lib/[etc] folder, right? 10:26
10:26 PastJynus is now known as YuviPanda
FROGGS hoelzro: I'm booting my windows machine right now 10:26
hoelzro FROGGS: I've only used panda to do it, but there's a small installation process
10:27 xfix left
dalek ast/glr: d51f2cc | coke++ | S04-declarations/constant.t:
add tests for RT #122895
10:27
synbot6 Link: rt.perl.org/rt3/Public/Bug/Display...?id=122895
hoelzro FROGGS: are you somewhere we can pair on this?
that might be easier =)
laben masak: github.com/masak/xml-writer/pull/5 10:28
FROGGS I'm upstairs
10:28 oetiker joined
[Coke] colomon: yes, that's what FROGGS suggested IRL. :) 10:28
FROGGS hoelzro: I need to rebuild all the things on windows, so no hurry :o) 10:29
[Coke] tadzik: is RT #122839 panda's fault or star's fault?
synbot6 Link: rt.perl.org/rt3/Public/Bug/Display...?id=122839
tadzik synbot6++ 10:30
botsnack!
synbot6 om nom nom
[Coke] m: m: class Foo { has @.bar }; say Foo.new( bar => [1,2,3] ).bar.elems
GLRelia rakudo-moar 00a34b: OUTPUT«3␤»
camelia rakudo-moar 5fb81f: OUTPUT«3␤»
tadzik [Coke]: I feel like 2014.08 may simply be too old for Compress::Zlib
10:31 average joined
[Coke] Ok. I'll close it with a polite note to reopen if it's still an issue with recent stars. 10:31
colomon lizmat: I’m getting 7 failures in S02-types/bag.t ? Did you say it was passing now?
tadzik FROGGS FROGGS
I meant to ask
why exactly is panda's gen-meta called Bundler?
average perl6 is a monument of stubborness. you guys are really pushing in that release on christmas ?
tadzik average: yes
[Coke] tadzik: ah, it was an issue on 2015.01..
average damn.. this is really happening
tadzik indeed! 10:32
We're quite excited
average i'm surprised
moritz m: say ().perl, ().item.perl
GLRelia rakudo-moar 00a34b: OUTPUT«()()␤»
camelia rakudo-moar 5fb81f: OUTPUT«()$( )␤»
virtualsue smooches average
tadzik [Coke]: I'm guessing some of the Compress::Zlib patches in the meantime take advantage of features introduced later than that Star release
average virtualsue: yo, sue ! 10:33
colomon m: my %m = bag <a b o p>,< a p o o>; dd %m
GLRelia rakudo-moar 00a34b: OUTPUT«%m = {:a(2), :b(1), :o(3), :p(2)}␤»
camelia rakudo-moar 5fb81f: OUTPUT«Odd number of elements found where hash initializer expected␤ in block <unit> at /tmp/WJIsurcJ2a:1␤␤»
tadzik for instance NativeCall getting into core :)
which is not a feature, but you get the idea
ab6tract colomon: :) 10:35
dalek kudo/glr: ec355fc | moritz++ | src/core/List.pm:
Fix itemization of List.perl
10:36
tadzik [Coke]: so basically, the correct response should be something along "consider using Compress::Zlib from 2014.08 when using Star from 2014.08"
colomon ab6tract: yeah, poor lizmat++ wrote that test yesterday, and we just broke it. :)
average O'Reilly should really interview the top #10 contributors on Perl6
tadzik unless that's broken too :)
average I'm really really curious for an in-depth interview revolving around motivations for writing on it
moritz average: somebody else already does that
average moritz: yes, I'm really curious to read that if it's already out or if it will be published soon
moritz some of those interviews are already published
[Coke] |Tux|: can you note the expected behavior you're not seeing in rt.perl.org/Ticket/Display.html?id=123980 ? thanks. 10:37
tadzik I know Gabor Szabo did some interviews
moritz somebody please help find those URLs
tadzik Also perl.com?
[Coke] tadzik: but someone reported "same issue on 2015.01".
average I've seen Gabor's interviews, they were nice, but some topics were overlooked
tadzik [Coke]: ah, fair enough
[Coke] tadzik: is looking for easy closes, this isn't one, moving on; feel free to grab it though. :) 10:38
FROGGS tadzik: because it will create a tar.gz at the end in Bundler me thinks
tadzik: there also should be a bundle target
10:39 laben left, laben joined
moritz average: perlhist.com/perl6 10:39
tadzik [Coke]: :) fair enough
oh, those are good
[Coke] m: class Foo:D {}; say Foo.new
GLRelia rakudo-moar 00a34b: OUTPUT«Foo.new␤»
camelia rakudo-moar 5fb81f: OUTPUT«Foo.new␤»
[Coke] I. WAT?
average moritz: thanks ! 10:40
[Coke] m: class Foo:U {}; say Foo.new
GLRelia rakudo-moar 00a34b: OUTPUT«Foo.new␤»
camelia rakudo-moar 5fb81f: OUTPUT«Foo.new␤»
[Coke] m: class Foo:ABC {}; say Foo.new
GLRelia rakudo-moar 00a34b: OUTPUT«Foo.new␤»
camelia rakudo-moar 5fb81f: OUTPUT«Foo.new␤»
[Coke] m: class Foo:any-ole-adverbial {}; say Foo.new
GLRelia rakudo-moar 00a34b: OUTPUT«Foo.new␤»
camelia rakudo-moar 5fb81f: OUTPUT«Foo.new␤»
tadzik class Foo {}; class FooD is Foo {}; say FooD ~~ Foo:D
m: class Foo {}; class FooD is Foo {}; say FooD ~~ Foo:D
GLRelia rakudo-moar 00a34b: OUTPUT«True␤»
camelia rakudo-moar 5fb81f: OUTPUT«True␤»
tadzik \o/
average you guys have seen the AOSA books right ?
tadzik I don't think so 10:41
average tadzik: aosabook.org/en/index.html
moritz has 10:42
and read most of it, even
average when you feel ready, you should try to get published in volume 1,2 or 3. I really see Perl6 as the sort of thing that should be in there(of course, when the time is right)
moritz of volume one, that is
average Audrey Tang has 2 articles in the AOSA books 10:43
maybe she could pave the way
laben for those who have merge rights: github.com/cosimo/perl6-string-crc32/pull/7 10:44
10:44 user3 joined
dalek ast/glr: 81c2001 | lizmat++ | S02-types/mixhash.t:
Make S02-types/mixhash.t pass
10:45
10:46 user3 left 10:47 rurban left
[Tux] [Coke] in RT#123980, I hoped .map(*//"-") to do the same as .map({*//"-"}) 10:47
synbot6 Link: rt.perl.org/rt3/Public/Bug/Display...?id=123980
10:47 Ven left
dalek ast/glr: 01588b2 | colomon++ | S02-types/bag.t:
Change bag to Bag.new as needed.

This is because we switched back the old "bag flattens" approach.
10:48
10:51 virtualsue left 10:52 rurban joined 10:54 ab6tract left
moritz [Tux]: looks distinctly wrong 10:54
oh, you used $_ // '-' inside the block 10:55
then it doesn't look quite as wrong anymore 10:56
but I don't think * can auto-curry short-circuiting operators
ShimmerFairy m: say (* + 5).perl 11:00
GLRelia rakudo-moar ec355f: OUTPUT«WhateverCode.new␤»
camelia rakudo-moar fd1612: OUTPUT«WhateverCode.new␤»
ShimmerFairy m: say (* && 5).perl
GLRelia rakudo-moar ec355f: OUTPUT«5␤»
camelia rakudo-moar fd1612: OUTPUT«5␤»
ShimmerFairy m: say (* || 5).perl
GLRelia rakudo-moar ec355f: OUTPUT«*␤»
camelia rakudo-moar fd1612: OUTPUT«*␤»
[Tux] moritz, yes, sorry. .map(*//"-") =?=?= .map({$_//"-"}) 11:02
but for now I am more interested in the laze Range change 11:03
m: class C{method foo{gather { take 3..Inf}}};C.new.foo[^9].perl.say
GLRelia rakudo-moar ec355f: OUTPUT«(3..Inf)␤»
camelia rakudo-moar fd1612: OUTPUT«(3, 4, 5, 6, 7, 8, 9, 10, 11)␤»
[Tux] that is biting me
timotimo it's kinda hard to make sense of that gather/take 11:04
ShimmerFairy m: my @a = gather { take 3..Inf }; say @a[^9].perl
GLRelia rakudo-moar ec355f: OUTPUT«(3..Inf)␤»
camelia rakudo-moar fd1612: OUTPUT«(3, 4, 5, 6, 7, 8, 9, 10, 11)␤»
timotimo since after taking a range that goes up to Inf, you'll hardl be able to take anything else
11:05 llfourn joined
ShimmerFairy m: class C { method foo { 3..Inf } }; say C.new.foo[^9].perl 11:05
GLRelia rakudo-moar ec355f: OUTPUT«(3, 4, 5, 6, 7, 8, 9, 10, 11)␤»
camelia rakudo-moar fd1612: OUTPUT«(3, 4, 5, 6, 7, 8, 9, 10, 11)␤»
ShimmerFairy ^ that should be sufficient, since that gather/take was useless anyway 11:06
11:06 average left
timotimo but it's obviously golfed from something 11:06
11:06 RabidGravy left
[Tux] ShimmerFairy, that is simplified code. the take is a loop where a sorted set of ranged adds to the final lazy list 11:06
timotimo tux's probably being bitten by take no longer flattening its arguments?
[Tux] if that detects Inf, it stops adding more 11:07
if I flat it myself, it will hang
timotimo yeah
[Tux] or I did something wrong
ShimmerFairy It's hard to see what the intent is from the code gold
*golf
timotimo .o( Jungle Sound Golf ) 11:08
[Tux] github.com/Tux/CSV/blob/master/lib...SV.pm#L118
that is the full code (for nom)
github.com/Tux/CSV/blob/glr/lib/Te...SV.pm#L118 for GLR 11:09
dalek ast/glr: 960da6d | colomon++ | S02-types/mix.t:
Change mix to Mix.new as needed.

This is because we switched back the old "mix flattens" approach.
11:09 llfourn left
[Tux] is preparing for Granada. will leave in 3 hours 11:10
timotimo have a good travel! 11:11
11:12 rurban1 joined
dalek rl6-roast-data: 6c3eb2d | coke++ | / (2 files):
another glr run for 'today'
11:12
[Coke] 122 failures. rerunning.
11:13 rurban left
[Coke] m: class A { multi method foo($a) { "general" }; multi submethod foo(Str $a) { "specific" } }; class B is A {}; say .new.foo("OH HAI") for A, B 11:15
dalek ast/glr: e13c1af | colomon++ | S02-types/set.t:
Change set to Set.new as needed.

This is because we switched back the old "set flattens" approach.
GLRelia rakudo-moar ec355f: OUTPUT«specific␤specific␤»
camelia rakudo-moar fd1612: OUTPUT«specific␤specific␤»
moritz btw Str.trans doesn't flatten Ranges 11:17
I guess it should, but I haven't found where to add it in the code
11:17 cognominal left
moritz m: List.sink 11:19
GLRelia rakudo-moar ec355f: OUTPUT«Method 'sink' not found for invocant of class 'List'␤ in block <unit> at /tmp/yDVgWcbQWZ:1␤␤»
timotimo don't we have Range => Range for the arguments to trans?
camelia ( no output )
moritz timotimo: I'm talking about things like (7, 1..6) => (1..7)
timotimo oh
right 11:20
[Coke] Is it worth doing anything about RT #77170 or should we close it as a DIHWIDT.
synbot6 Link: rt.perl.org/rt3/Public/Bug/Display...l?id=77170
11:20 brrt joined
moritz m: for List.^mro { say .^name if .^method_table<sink> } 11:21
GLRelia ( no output )
camelia rakudo-moar fd1612: OUTPUT«List␤»
moritz m: for List.^roles { say .^name if .^method_table<sink> }
GLRelia rakudo-moar ec355f: OUTPUT«Method 'method_table' not found for invocant of class 'Perl6::Metamodel::ParametricRoleGroupHOW'␤ in block <unit> at /tmp/QbgRK_RGFa:1␤␤»
camelia rakudo-moar fd1612: OUTPUT«Method 'method_table' not found for invocant of class 'Perl6::Metamodel::ParametricRoleGroupHOW'␤ in block <unit> at /tmp/Cfwn4i0Q_s:1␤␤»
moritz where does List.sink come from? 11:22
m: say List.^can('sink')[0].package 11:23
GLRelia rakudo-moar ec355f: OUTPUT«Nil␤»
camelia rakudo-moar fd1612: OUTPUT«(List)␤»
moritz wtf? 11:24
why does Nil.sink end up in List?
11:26 kid51 joined
[Coke] m: .WHY, not Nil 11:27
GLRelia ( no output )
camelia rakudo-moar fd1612: OUTPUT«WARNINGS:␤Useless use of "," in expression ".WHY, not " in sink context (line 1)␤»
11:27 kid51 left
[Coke] see I was not expecting that to do something in glr vs. nom 11:27
11:29 virtualsue joined
[Coke] closes RT #77170 - feel free to re-open if you feel it's something we should fix. 11:29
[Coke] has a change of heart and doesn't close it. bah. 11:30
11:30 ab6tract joined
ab6tract [Coke]: how are we doing? :) 11:30
[Coke] ab6tract: latest run for glr is at S15 and pending. last run had 122 failures. down to 1038 RTs. 11:32
m: say "foo" & "a nice old foo" ~~ /foo/
GLRelia rakudo-moar ec355f: OUTPUT«␤»
camelia rakudo-moar fd1612: OUTPUT«「all(Failure.new(exception => X::OutOfRange.new(what => "Start argument to substr", got => "5", range => "0..3", comment => "use *5 if you want to index relative to the end"), backtrace => Backtrace.new), "e o")」␤»
11:33 colomon_ joined
lizmat m: say gist(1) 11:33
GLRelia rakudo-moar ec355f: OUTPUT«(1)␤»
camelia rakudo-moar fd1612: OUTPUT«1␤»
colomon_ m: my @a := [1,2],{ [.[0]+2,.[1]+2] } ... *; dd @a
GLRelia rakudo-moar ec355f: OUTPUT«Index out of range. Is: 1, should be in 0..0␤ in block <unit> at /tmp/UF0YNPCbcg:1␤␤Actually thrown at:␤ in block at /tmp/UF0YNPCbcg:1␤␤»
camelia rakudo-moar fd1612: OUTPUT«([1, 2], [3, 4], [5, 6], [7, 8], [9, 10], [11, 12], [13, 14], [15, 16], [17, 18], [19, 20]...Inf)␤»
colomon_ can anyone explain that one?
lizmat m: say gist(my $a, my $b)
GLRelia rakudo-moar ec355f: OUTPUT«((Any) (Any))␤»
camelia rakudo-moar fd1612: OUTPUT«(Any) (Any)␤»
lizmat m: my @a := (1,2),{ [.[0]+2,.[1]+2] } ... *; dd @a 11:34
GLRelia rakudo-moar ec355f: OUTPUT«Index out of range. Is: 1, should be in 0..0␤ in block <unit> at /tmp/mR5zAwbFJt:1␤␤Actually thrown at:␤ in block at /tmp/mR5zAwbFJt:1␤␤»
camelia rakudo-moar fd1612: OUTPUT«Index out of range. Is: 1, should be in 0..0␤ in block <unit> at /tmp/igNADeBgbW:1␤␤Actually thrown at:␤ in block at /tmp/igNADeBgbW:1␤␤»
ab6tract \o/ 11:35
[Coke] m: nom: class A { has str $.x; method BUILD(:$!x) { } };say A.new(:x<foo>).x
GLRelia rakudo-moar ec355f: OUTPUT«Cannot modify an immutable str␤ in method BUILD at /tmp/uE6p2E7mQO:1␤ in block <unit> at /tmp/uE6p2E7mQO:1␤␤»
camelia rakudo-moar fd1612: OUTPUT«Cannot modify an immutable str␤ in method BUILD at /tmp/3joEvTJ3zW:1␤ in block <unit> at /tmp/3joEvTJ3zW:1␤␤»
[Coke] m: if 42, 44, 22 -> *@a { say @a.perl } 11:36
GLRelia rakudo-moar ec355f: OUTPUT«[]␤»
camelia rakudo-moar fd1612: OUTPUT«[]<>␤»
[Coke] m: while 42, 44, 22 -> *@a { say @a.perl; last }
GLRelia rakudo-moar ec355f: OUTPUT«[]␤»
camelia rakudo-moar fd1612: OUTPUT«[]<>␤»
11:36 xfix joined
ab6tract m: while [42, 44, 22] -> *@a { say @a.perl; last } 11:37
GLRelia rakudo-moar ec355f: OUTPUT«[]␤»
camelia rakudo-moar fd1612: OUTPUT«[]<>␤»
[Coke] m: constant k = k; say k
GLRelia rakudo-moar ec355f: OUTPUT«5===SORRY!5=== Error while compiling /tmp/p3Cx73dLWP␤An exception occurred while evaluating a constant␤at /tmp/p3Cx73dLWP:1␤Exception details:␤ 5===SORRY!5=== Error while compiling ␤ Cannot invoke this object (REPR: Null, cs = 0)␤ at…»
camelia rakudo-moar fd1612: OUTPUT«5===SORRY!5=== Error while compiling /tmp/g9ROvqM1xF␤An exception occurred while evaluating a constant␤at /tmp/g9ROvqM1xF:1␤Exception details:␤ 5===SORRY!5=== Error while compiling ␤ Cannot invoke this object (REPR: Null, cs = 0)␤ at…»
[Coke] we have five tickets marked as [GLR] in RT. (more may be there unmarked)
10 tickets marked testneeded. 11:39
The role of build bot today will be played by [Coke]
colomon m: my @a; say @a[1] 11:40
GLRelia rakudo-moar ec355f: OUTPUT«(Any)␤»
camelia rakudo-moar fd1612: OUTPUT«(Any)␤»
colomon m: my $a; say $a[1] 11:41
GLRelia rakudo-moar ec355f: OUTPUT«(Any)␤»
camelia rakudo-moar fd1612: OUTPUT«(Any)␤»
dalek kudo/glr: 093a4fe | (Stefan Seifert)++ | src/core/io_operators.pm:
Fix sub gist enclosing in parentheses even when passed a single argument

gist(1) returned "(1)" instead of just "1" since commit 2023f0218f19cd1a625f5a5b6ffb69a532db1053
11:42
11:46 brrt left 11:48 CIAvash left 11:51 TEttinger left
dalek ast/glr: 9d89274 | (Stefan Seifert)++ | S16-io/print.t:
Update S16-io/print.t to GLRR rules

Under GLRR itemization prevents even a single arg from beeing flattened. So print should treat these as single items and .Str them
11:52
jdv79 is there an easy way to resolve a pause id from the p6 git repos? 11:53
dalek ast/glr: e72a073 | FROGGS++ | S04-statements/for_with_only_one_item.t:
adjust test to One Arg Rule™
11:54
lizmat jdv79: perhaps from CREDITS ? 11:56
jdv79 i guess it'll have to be written. was more asking if something already existed. 11:57
FROGGS @all: this is the current spectest fallout: github.com/perl6/glr-test-failures...ilures.txt 12:02
m: say (1 ... 5 ... 10).join(' ') 12:04
GLRelia rakudo-moar 093a4f: OUTPUT«Method 'splice' not found for invocant of class 'List'␤ in block <unit> at /tmp/Xk5VO29zk_:1␤␤»
camelia rakudo-moar fd1612: OUTPUT«1 2 3 4 5 6 7 8 9 10␤»
nine FROGGS: that change to for_with_only_one_item.t is actually wrong because the item is itemized which should prevent it from being flattened even under single-arg-rule 12:05
FROGGS: the difficulty there is that we want my @a = 1,2; for @a to iterate twice, for $a to iterate once, @a.map to iterate twice and $a.map also to iterate twice. I don't really know how to do that since for compiles to a .map 12:06
FROGGS nine: I guess 'for X' should translate more to 'X.list.map' 12:07
(metaphorically)
nine But $@a.list will just return @a
FROGGS and ($@a).list? 12:08
nine Maybe we need an :item option on method map
FROGGS: will return @a as well
FROGGS: List.list is just return self
FROGGS that seems correct
12:09 domidumont joined
dalek pan style="color: #395be5">perl6-examples: 0b2f96d | (Steve Mynott)++ | categories/cookbook/15interactivity/15-01-parse-program-args.pl:
this comment breaks the whole make html since the pod eval exits
12:09
12:10 Ven joined 12:13 virtualsue left
nine sub fsort-only([$p?,*@r]) { 12:15
12:15 virtualsue joined
dalek ast/glr: 6b8d709 | FROGGS++ | S04-statements/for_with_only_one_item.t:
Revert "adjust test to One Arg Rule™"

This reverts commit e72a073e12f5dafaff35c1fb0e124562858ff40c.
12:16
12:16 Ven left 12:18 cognominal joined 12:21 sunnavy left 12:22 sunnavy joined
dalek rl6-roast-data: 2f1df5b | coke++ | / (2 files):
another glr run for 'today'
12:23
12:23 rarara joined
[Coke] down to 95 fails. 12:24
nine epic!
12:26 sunnavy left
timotimo very nice 12:26
being below 128 is good
[Coke] do we have a builtin somewhere that escapes problematic shell characters?
(The ticket is mainly about " ", but figured I'd do it more generically if possible.) 12:27
12:27 sunnavy joined 12:28 bin_005_l left
laben [Coke]: no builtin i know of, but a module is present github.com/xfix/perl6-Acme-Addslashes/ XD 12:28
xfix Addslashes is a joke, it's not for escaping shell characters.
[Coke] needed for core. I'll see what I can steal there.
... or not. :)
laben i know that, it's the reason why i added "XD"
xfix I don't like the idea of "escaping" myself.
It's quite... error-prone. 12:29
timotimo we should have :XD, too
xfix For bash however, I guess you can put everything between single quotes, and replace single quotes with '\''
But I don't like the idea myself, it's probably better to just directly call a program using some exec API I don't remember. 12:30
Ignoring shell.
huf you better have some kind of shell escape functionality in place, because if you dont, people will *still* interpolate strings into shell calls
[Coke] xfix - this ideally would be cross shell, so there's probably not much to do. 12:31
huf and you'll have a harder time telling them do escape it
laben meanwhile Text::CSV is getting in better health, i hope to be able to use it as benchmark today or tomorrow at most
[Coke] this is for RT #113954
synbot6 Link: rt.perl.org/rt3/Public/Bug/Display...?id=113954
xfix To support fish shell, also add conversion of \ to '\\'. This will work in both bash and fish, but won't be compatible with cmd.exe. 12:32
cmd.exe has a really strange syntax, I don't quite remember how to escape it.
ShimmerFairy [Coke]: .perl 'ing the string should be good enough to show that the string needs to be quoted
FROGGS hoelzro++
ShimmerFairy (and if the MAIN program there asks for a literal string that needs special escaping for the given shell, that's probably a design error on the programmer's part :P) 12:33
12:33 domidumont left
xfix But looking at RT, I wouldn't consider this to be a huge problem. It feels strange if someone would actually do this. 12:33
I guess special characters could be simply escaped with \, without quotes. 12:34
12:34 mprelude joined
xfix For example, -e '...' foo\ bar 12:34
[Coke] I am always happy to close things as DIHWIDT.
xfix This is strange, but if someone does something silly like that, oh well.
[Coke] xfix: not sure if that works on CMD.exe
xfix True, CMD requires a different syntax.
[Coke] .ask masak do we really need to fix 113954, or is it: DIHWIDT ?
yoleaux [Coke]: I'll pass your message to masak.
[Coke] m: .say for gather take do given 1 { { a => 1, b => 2 } } 12:36
GLRelia rakudo-moar 093a4f: OUTPUT«a => 1b => 2␤»
xfix Escape character for cmd.exe was ^, I think, but the user may not necessarily use cmd.exe on Windows making things tricker.
camelia rakudo-moar fd1612: OUTPUT«a => 1, b => 2␤»
nine m: lazy { 42 }
GLRelia rakudo-moar 093a4f: OUTPUT«Method 'lazy' not found for invocant of class 'Int'␤ in block <unit> at /tmp/6TMJeS9A3n:1␤␤»
camelia ( no output )
ShimmerFairy [Coke]: I can't imagine that problem coming up except in that situation. And if you're going to run against shell character parsing while requesting a literal in a MAIN multi, my suspicion is that you should really have to write your USAGE too. 12:37
nine m: say lazy { 42 }
GLRelia rakudo-moar 093a4f: OUTPUT«Method 'lazy' not found for invocant of class 'Int'␤ in block <unit> at /tmp/Onsm60FEMA:1␤␤»
camelia rakudo-moar fd1612: OUTPUT«42␤»
dalek ast/glr: ba06e5e | lizmat++ | S17-supply/on.t:
Make S17-supply/on.t pass

Actually for the wrong reason: we seem to have an issue with on-demand supplies of different lengths in here. So I fixed the test by making them the same length. At one point, we probably want to add a test for different length on-demand supplies.
12:38
12:40 virtualsue left, virtualsue joined 12:41 laouji joined 12:46 tinyblak joined
nine lizmat: d04396141 12:46
[Coke] gist.github.com/coke/3750096a7b0996a9cf87 - give me updates.
laben [Coke]: you could mention under the first item (glr) that panda works (for now) 12:47
12:48 Ven joined
[Coke] laben++ 12:48
laben i hope that brings more people to try out their modules or others
[Coke] updated. 12:49
dalek pan style="color: #395be5">perl6-examples: e71314c | (Steve Mynott)++ | lib/Pod/Htmlify.pm6:
sort sub categories by filename
osystem: 1bf4f1a | ab5tract++ | META.list:
Release Terminal::Print
12:50
12:50 virtualsue left, tinyblak left
dalek kudo/glr: 966096d | (Stefan Seifert)++ | t/spectest.data:
Don't run lazy.t at all since lazy {} has never been implemented

lazy.t is failing now for good reasons, so do not run it until we are actually trying to implement the "lazy" statement prefix.
12:54
[Tux] laben, do you want co-maint, or do you plan PR's? 12:55
I'll be off-line in 45 minutes
laben [Tux]: i'll send you some PR, no hurry though, much work needs to be done still
[Coke] I have spicy cheese if anyone wants a hunk. 12:56
I WILL eat it all if you don't stop me.
FROGGS :D
nine m: ([Z] ()).perl.say 12:57
GLRelia rakudo-moar 093a4f: OUTPUT«Nil␤»
camelia rakudo-moar fd1612: OUTPUT«()␤»
12:57 itz left
nine FROGGS: we get a Nil now and I'm unsure if that's correct or where it's actually coming from. Could you have a look? 12:57
FROGGS nine: what test/code exactly? 12:58
dalek osystem: ca9075a | (Tadeusz “tadzik” Sośnierz)++ | META.list:
Add Terminal::ANSIColor
12:59
tadzik where do I teach dalek who I am?
FROGGS nine: nvm
tadzik moritz?
nine FROGGS: t/spec/S03-operators/reduce-le1arg.rakudo.moar
FROGGS nine: yeah, I already looked at it but gave up earlier
nine :( 13:00
FROGGS let's try again
TimToady m: say infix:<Z>()
camelia rakudo-moar fd1612: OUTPUT«Nil␤»
ShimmerFairy tadzik: it involves your git settings
GLRelia rakudo-moar 966096: OUTPUT«Nil␤» 13:01
FROGGS m: say &infix:<,>()
GLRelia rakudo-moar 966096: OUTPUT«␤»
camelia rakudo-moar fd1612: OUTPUT«␤»
tadzik ShimmerFairy: yes, I know
13:01 smls joined
tadzik ShimmerFairy: but dalek knows how to translate my git name to my irc name 13:01
FROGGS m: say METAOP_REDUCE_LISTINFIX()
GLRelia rakudo-moar 966096: OUTPUT«5===SORRY!5=== Error while compiling /tmp/Z47syPYlBt␤Calling METAOP_REDUCE_LISTINFIX() will never work with any of these multi signatures:␤ (\op, \triangle) ␤ (\op)␤at /tmp/Z47syPYlBt:1␤------> 3say 7⏏5METAOP_REDUCE_LISTINFIX()␤»
camelia rakudo-moar fd1612: OUTPUT«5===SORRY!5=== Error while compiling /tmp/L9kWQJZq73␤Calling METAOP_REDUCE_LISTINFIX() will never work with any of these multi signatures:␤ (\op, \triangle) ␤ (\op)␤at /tmp/L9kWQJZq73:1␤------> 3say 7⏏5METAOP_REDUCE_LISTINFIX()␤»
tadzik and now that I have a new one it's confused :)
FROGGS m: say METAOP_REDUCE_LISTINFIX(())
GLRelia rakudo-moar 966096: OUTPUT«sub (|values) { #`(Sub|47703536) ... }␤»
camelia rakudo-moar fd1612: OUTPUT«sub (|values) { #`(Sub|64323176) ... }␤»
FROGGS m: say METAOP_REDUCE_LISTINFIX(())()
GLRelia rakudo-moar 966096: OUTPUT«Invocant requires a type object of type List, but an object instance was passed. Did you forget a 'multi'?␤ in block <unit> at /tmp/eLvE7I8hF3:1␤␤»
camelia rakudo-moar fd1612: OUTPUT«Cannot find method 'postcircumfix:<( )>'␤ in block <unit> at /tmp/Mo3lNMzlo7:1␤␤»
laben nine: do you know why "sub foo(*@m) {@m.say}; foo(Array)" hangs?
FROGGS ahh w/e
nine laben: looking at it 13:03
laben nine: golfed it to "(flat Array).say"
[Coke] m: my @a = ^3;say (state@ = @a).pop xx 2;say @a; 13:04
GLRelia rakudo-moar 966096: OUTPUT«Cannot pop from an empty Array␤ in block <unit> at /tmp/xbNrteOudj:1␤␤Actually thrown at:␤ in block <unit> at /tmp/xbNrteOudj:1␤␤»
camelia rakudo-moar fd1612: OUTPUT«Cannot pop from an empty Array␤ in block <unit> at /tmp/QnwaeTMaAk:1␤␤Actually thrown at:␤ in block <unit> at /tmp/QnwaeTMaAk:1␤␤»
laben nine: it's surely flat, as even "Array.flat" hangs
[Coke] m: sub p(*@l) { gather { if @l == 1 { take @l } else { @l.shift; for p(@l) -> @p {} } } }; say p 1..4 13:05
GLRelia rakudo-moar 966096: OUTPUT«␤»
camelia rakudo-moar fd1612: OUTPUT«Type check failed in binding @p; expected 'Positional' but got 'Int'␤ in block at /tmp/SqXtcgfOAz:1␤ in block <unit> at /tmp/SqXtcgfOAz:1␤␤»
13:05 tinyblak joined 13:06 llfourn joined, tinyblak left, itz joined
tadzik github.com/rakudo/rakudo/blob/nom/...ons.pm#L13 :| 13:06
can we not
liek, have an env var that turns them off or so 13:07
13:07 tinyblak joined
JimmyZ you can ! 13:07
tadzik then can we teach camelia to do it and not have this in core? :) 13:08
lizmat m: my @a; say all(@a)
camelia rakudo-moar fd1612: OUTPUT«all()␤»
GLRelia rakudo-moar 966096: OUTPUT«(timeout)» 13:09
lizmat m: all(my @)
GLRelia ( no output )
camelia ( no output )
ShimmerFairy tadzik: well, # until something better
:P
lizmat m: say all(my @)
camelia rakudo-moar fd1612: OUTPUT«all()␤»
lizmat m: say all([])
camelia rakudo-moar fd1612: OUTPUT«all()␤»
ab6tract whoever knows how to fiddle camelia, all we need is RAKUDO_NO_DEPRECATIONS 13:10
tadzik ya
GLRelia rakudo-moar 966096: OUTPUT«(timeout)»
lizmat ab6tract: maybe we should teach the deprecation mechanism to not be as verbose on camelia?
13:10 tinyblak left 13:11 llfourn left, diana_olhovik_ left
ab6tract lizmat: maybe having all deprecations off for camelia would be a good canary cage for #perl6 golfing? 13:11
13:12 itz left 13:13 larion left
[Coke] every change we make to how cam processes things is going to involve us having to explain it to someone later about why their output is different. 13:13
ShimmerFairy
.oO( use warnings :verbose/:brief/:superbrief )
13:14
tadzik lizmat: I think it's silly for Rakudo core to know the hostname camelia runs on :)
laben we could have a modded RESTRICTED core called CAMELIA or EVALBOT with all these useful mods 13:15
ab6tract [Coke]: so does that count you against having her use RAKUDO_NO_DEPRECATIONS?
[Coke] twitter.com/flipzagging/status/634...2040162304 Which may or may not sum up xmas this year.
I'm about -.1 on that. 13:16
13:16 itz joined
[Coke] no big deal. 13:16
tadzik or we could use the env var on caemlia
[Coke] (tweet comes via bacek)
tadzik oh, bacek :)
[Coke] tadzik: I think that's what's being suggested, aye.
tadzik long time no see
13:21 skids joined 13:23 araujo left, domm_holiday left, luiz_lha left, huf left, bartolin left, esh left, mattp_ left, mtj_ left, frew left, bartolin joined, domm_holiday joined 13:24 huf joined, luiz_lha joined, frew joined, mattp_ joined, luiz_lha is now known as Guest12682, araujo joined
FROGGS Ven: are you still here? 13:24
Ven FROGGS: yes
I'm afraid masak isn't anymre
13:24 mtj_ joined, _itz_ left, mr-foobar left, preyalone left, nebuchadnezzar left, lea left, krunen left, geekosaur left, tadzik left, ShimmerFairy left, robinsmidsrod left 13:25 esh joined, nebuchadnezzar joined
FROGGS Ven: I thought you also had to leave already 13:25
13:25 lea joined, _itz joined, krunen joined
Ven FROGGS: in a bit less than 2 hrs! 13:25
FROGGS k
13:25 ShimmerFairy joined, tadzik joined, mr-foobar joined, robinsmidsrod joined, geekosaur joined
FROGGS I think your charger and your glasses etui is up here 13:25
Ven ..and my bag :) 13:26
13:26 camelia left, araujo left, camelia joined 13:27 mrf joined, virtualsue joined
nine m: say nqp::iscont(my $a); 13:27
13:27 Loren_ joined
GLRelia rakudo-moar 966096: OUTPUT«===============================================================================␤The use of nqp::operations has been deprecated for non-CORE code. Please␤change your code to not use these non-portable functions. If you really want␤to keep using nqp:…» 13:27
13:27 araujo joined
FROGGS nine: I think I've got a fix for the [Z] () test 13:27
13:27 Ven left
ShimmerFairy github.com/perl6/roast/commit/5e7d...234c2a7e57 13:27
nine \o/
ShimmerFairy was this commit meant for glr?
Because the changes made break the test file for me on nom 13:28
13:28 ChanServ sets mode: +v camelia
camelia rakudo-moar fd1612: OUTPUT«1␤» 13:28
dalek kudo/glr: 71d4f4c | (Stefan Seifert)++ | src/core/Array.pm:
Avoid endless loop on flat(Array)
kudo/nom: 6a1879d | lizmat++ | src/core/Deprecations.pm:
Eradicate special casing for camelia
kudo/glr: b33f31c | (Stefan Seifert)++ | src/core/Iterable.pm:
Move "undefined" error up the stack
13:29
ShimmerFairy m: for @(1..3 X 'a'..'b') -> $a, $b { say $a, $b }
camelia rakudo-moar fd1612: OUTPUT«1a␤1b␤2a␤2b␤3a␤3b␤»
GLRelia rakudo-moar 966096: OUTPUT«(1 a)(1 b)␤(2 a)(2 b)␤(3 a)(3 b)␤»
ShimmerFairy m: for 1..3 X 'a'..'b' -> ($a, $b) { say $a, $b }
camelia rakudo-moar fd1612: OUTPUT«Too few positionals passed; expected 2 arguments but got 0 in sub-signature␤ in block <unit> at /tmp/XBI3orPjSi:1␤␤»
GLRelia rakudo-moar 966096: OUTPUT«1a␤1b␤2a␤2b␤3a␤3b␤»
ShimmerFairy seems so :)
laben nine++ great work as always 13:30
timotimo github.com/perl6/gtk-simple/blob/m...Simple.pm6nine
dalek Heuristic branch merge: pushed 17 commits to roast/glr by lizmat 13:31
13:32 preyalone joined
dalek ast: e35f48e | ShimmerFairy++ | S03-metaops/cross.t:
Revert "fix cross metaop with nested/itemized lists"

This reverts commit 5e7d42fb821b461fd491173228430d234c2a7e57, which is a change seemingly meant for the glr branch. Breaks on nom.
13:32
kudo/glr: 30ba99b | FROGGS++ | src/core/List.pm:
fix "[Z] ()" to return (), returned Nil before
13:33
13:33 mvuets joined
mvuets o hai! 13:34
hoelzro o/ mvuets
timotimo o/
13:34 itz left
nine timotimo: why not just $code($task, Any);? 13:35
13:38 colomon left, virtualsue left
mvuets masak: are you coming back to the venue tonight by chance? 13:38
13:39 itz joined
hoelzro mvuets: I think he's heading home =( 13:40
13:40 virtualsue joined
[ptc] o/ mvuets 13:41
13:41 cognominal left
mvuets while i have a chance to talk to people in person, i was wondering who can help me and answer some question? earlier this year i wanted to hack some P6 specs and testing modules. but not sure if that's a good idea yara yara 13:42
i am too dumb to hack on the core or whatnot, but could do some nitty-gritty fixes or enhancements. i tend to micro manage and focus on tiny details 13:44
[ptc] mvuets: sounds like the kind of thing I do ;-)
13:44 Ven joined, colomon joined
colomon takes advantauge of his restored connection to post something pointless: appcordions.com/ 13:45
[ptc] \o/ bagpipes on Android!
mvuets: what do you want to work on? 13:46
hoelzro mvuets: we've migrated to the salon, if you'd like to join us
dalek rl6-roast-data: db103c4 | coke++ | / (2 files):
another glr run for 'today'
[ptc] mvuets: btw: we're all in the Salon downstairs now, so come on down!
colomon [ptc]: only the keyboard interface. :(
laben m: class C {method foo (*@a is copy) {}}; my $c = C.new; $c.foo(Hash)
camelia ( no output )
GLRelia rakudo-moar 30ba99: OUTPUT«Invocant requires an instance of type Hash, but a type object was passed. Did you forget a .new?␤ in method foo at /tmp/ZIBCgltryv:1␤ in block <unit> at /tmp/ZIBCgltryv:1␤␤»
[Coke] 82 failures.
hoelzro .oO( not to me confused with the saloon )
laben can someone explain this ^^^ msg to me?
and if it's intended that it is different from nom 13:47
nine laben: it's an error 13:48
[ptc] itz: you have a commit bit on perl6-examples, don't you?
moritz laben: looks like a bug to me
nine laben: we're probably trying to call .list on Hash
ab6tract m: Hash.copy.say
camelia rakudo-moar fd1612: OUTPUT«Method 'copy' not found for invocant of class 'Hash'␤ in block <unit> at /tmp/Pwuv47e98g:1␤␤»
GLRelia rakudo-moar 30ba99: OUTPUT«Method 'copy' not found for invocant of class 'Hash'␤ in block <unit> at /tmp/n_4jGdicSi:1␤␤»
ab6tract oops, nm me :)
ShimmerFairy nine: it looks to me like perhaps it's thinking 'is copy' needs the same things 'is rw' does, at first glance 13:49
timotimo nine: p6parcel would turn $task into a list that gets filled up with Any for any nulled entries
nine timotimo: what is $task exactly? 13:50
13:51 cognominal joined
timotimo good question 13:53
i thought you were going to figure all of this out for me :P
13:55 colomon_ left, brrt joined, itz left 13:56 virtualsue left, ddeimeke joined
dalek ast/glr: 38af820 | (Stefan Seifert)++ | S04-statements/sink.t:
Remove obsolete todo marker from test
13:56
lizmat m: sub foo { say &?ROUTINE.name }
camelia ( no output )
GLRelia ( no output )
nine timotimo: sorry, if it were something to do with Qt, I'd be interested ;) Couldn't care less about GTK though. 13:57
timotimo heh
lizmat m: sub foo { if 1 { say &?ROUTINE.name } }()
camelia rakudo-moar 6a1879: OUTPUT«<unit>␤»
GLRelia rakudo-moar 30ba99: OUTPUT«<unit>␤»
13:57 colomon left
timotimo understandably 13:57
13:58 colomon joined
lizmat m: sub foo { if 1 { say &?ROUTINE.name }; &?ROUTINE }() 13:58
camelia rakudo-moar 6a1879: OUTPUT«<unit>␤»
GLRelia rakudo-moar 30ba99: OUTPUT«<unit>␤»
13:59 khw joined
ddeimeke Hi guys @spw2015, we like to do a "before / after" retrospective. Can you please tell me the current status of tests failing? It was 102 on Thursday morning, correct? 13:59
[ptc] ddeimeke: afaik, we're under 20 files failing now. People have been working hard :-) 14:00
ddeimeke Applause \O/
dalek ast/glr: a645d40 | lizmat++ | S17-promise/start.t:
Since #122715 is not closed yet, skip this
14:01
synbot6 Link: rt.perl.org/rt3/Public/Bug/Display...?id=122715
14:01 average joined
average jnthn is pushing those presentations hard 14:02
laben mh, i remember reading about possible native c++ bindings, are they planned? 14:03
dalek ast/glr: 0aedac0 | (Stefan Seifert)++ | S03-metaops/cross.t:
Fix failure caused by faulty merge conflict resolution
14:04
tadzik laben: hoelzro was talking about it today
ab6tract average: what do you mean?
tadzik laben: there's a branch in moarvm that makes them possible iirc
ab6tract laben: supposedly it just needs a small amount of love 14:05
14:05 fame joined
laben tadzik: interesting, how much would be possible? it would be amazing to be able to map c++ overload to p6 mmd 14:07
dalek kudo/glr: 33b9374 | (Jimmy Zhuo)++ | src/core/Int.pm:
add native int version of pow
rakudo/glr: 905eea0 | coke++ | src/core/ (2 files):
rakudo/glr: RT #123760 - add Str matchers for comb
14:07 dalek left
synbot6 Link: rt.perl.org/rt3/Public/Bug/Display...?id=123760 14:07
mvuets [ptc]: last time i checked, i wanted to work on S24 (testing) related affairs 14:08
tadzik laben: you'd have to pester hoelzro :)
I dunno
14:08 dalek joined, ChanServ sets mode: +v dalek
hoelzro hides 14:08
laben hoelzro: can you tell me moar? 14:09
14:09 tinyblak joined
hoelzro oh noes, he found me! 14:10
14:10 xyf joined
tadzik . o O ( talk moar to me ) 14:10
hoelzro so FROGGS authored it, and you need the moarvm and rakudo cpp branches
I think one of them is called cpp2?
colomon am I still on here? 14:11
hoelzro looking at the tests in rakudo under t for the C++ native call is a good guide
colomon: yes
colomon \o/
github.com massively slow
tadzik internet bad
lizmat too much perl6 hacking
laben hoelzro: no need for special nqp?
hoelzro laben: I don't think so, but you can check for a C++ branch 14:12
ShimmerFairy hoelzro: how many FFIs out there support C++ libraries, ooc? My limited knowledge tells me it's kind of significant that rakudo can possibly do it :)
hoelzro ShimmerFairy: very few; I think Julia does?
colomon tadzik: what’s the diff between perl6-Term-ANSIColor and Terminal::ANSIColor? 14:13
hoelzro I had notes on this...
tadzik colomon: the latter is newer :)
nine Maybe merging nom wasn't the best idea today...
hoelzro ShimmerFairy: it's definitely cool =)
laben python does, iirc
tadzik colomon: ab6tract had a good idea to stop using Term:: as a namespace for Terminal stuff, since Term is a thing in perl 6
hoelzro laben: really? I didn't konw that!
ShimmerFairy hoelzro: it definitely gives the impression that Perl 6 is one of the few languages that can meet C++ on its own level :P
tadzik: what Term?
hoelzro ShimmerFairy: on more than one level ;) 14:14
laben hoelzro: not sure, but i remember qt and kde have python bindings
ShimmerFairy that might be via C bindings, at first guess
hoelzro ahhh, that was one of my links
about how they did Qt <> Python bindings
14:14 itz joined
hoelzro I think it may be because Qt is special 14:14
laben ShimmerFairy: actually it seems they have a special way, Smoke they called it
colomon tadzik: I’m going to quickly design a new smoker-thingy for emmentaler. Trying to smoke everything in one run is Simply Not Working.
hoelzro and has introspectability 14:15
tadzik colomon: sounds awesome
ShimmerFairy hoelzro: yeah, Qt is (as far as I understand it) very much a product of pre-standard-library days, so it tends to be visibly different from other C++ code :)
colomon tadzik: my design goal is to have it be dead simple and work. :)
tadzik colomon: that used to be true for smoker :P
colomon tadzik: yeah, if these people hadn’t written all these darned modules… 14:16
;)
14:16 fame is now known as cwDYKARN
tadzik hehe 14:16
14:18 RabidGravy joined
nine Could people PLEASE learn how git works? 14:18
lizmat m: sub a(@a) { say @a.perl }; my &w = &[,]; a [[&w]] <a b c>.map: {$_}
camelia rakudo-moar 6a1879: OUTPUT«(("a", "b"), "c")␤»
GLRelia rakudo-moar e6c4b9: OUTPUT«(("a", "b"), "c")␤»
14:18 Ven left
ShimmerFairy hoelzro: the first time I saw dalek report on C++ bindings here, my reaction was "but that's impossible o_o" :P 14:18
hoelzro ShimmerFairy: I just gave a talk about how hard is gets =) 14:19
ShimmerFairy nine: what's the issue?
hoelzro I wasn't using the C++ native call stuff, though =/
xyf join C++
laben hoelzro: dont you "just" need a demangler? XD
xyf exit
14:19 xyf left
hoelzro laben: hehehhe 14:19
14:20 rurban1 left
ab6tract laben: as i understand it, C++11 finally includes a stable ABI 14:20
which means we would not need any C binding jibber jabber
ShimmerFairy The point of mangling, AIUI, is to keep code from different compilers mixing together, so I'd be surprised if C++11 standardized it in some way
nine ShimmerFairy: you reverted a commit on master that was meant for glr. So when we merge master into glr now, we get the revert on glr as well. When I revert the revert there, I get conflicts.
14:21 tinyblak left
[Coke] regression, now have 163 tests failing. :P 14:21
ShimmerFairy nine: well, I'm not aware of a better solution to handling that specific issue, I simply wanted to fix the nom branch :)
14:22 tinyblak joined
colomon way to get a unique filename in p6? 14:22
laben hoelzro: i promise you, if we have proper c++ bindings, i'll work my ass off to have proper Qt bindings!
[Coke] File::Temp ?
tadzik yes
[Coke] m: use File::Temp;
camelia rakudo-moar 6a1879: OUTPUT«===SORRY!===␤Could not find File::Temp in any of:␤ file#/home/camelia/.perl6/2015.07.1-179-g6a1879d/lib␤ inst#/home/camelia/.perl6/2015.07.1-179-g6a1879d␤ file#/home/camelia/rakudo-inst-1/share/perl6/lib␤ file#/home/camelia/rakudo-inst-1/shar…»
GLRelia rakudo-moar e6c4b9: OUTPUT«===SORRY!===␤Could not find File::Temp in any of:␤ file#/home/steve/.perl6/2015.07.1-634-ge6c4b9d/lib␤ inst#/home/steve/.perl6/2015.07.1-634-ge6c4b9d␤ file#/home/steve/sandbox/perl6/rakudo/install/share/perl6/lib␤ file#/home/steve/sandbox/per…»
nine Would have been nice to clean up the fallout in glr afterwards, so I wouldn't have had to waste time trying to figure out what happened.
hoelzro laben++ 14:23
laben oh yeah, who takes PR for File::Temp? i got a patch around since some time to make it work for glr
hoelzro laben: feel free to ping me or FROGGS if you have questions
[Coke] +S32-str/comb.rakudo.moar aborted 48 test(s)
tadzik laben: I'm pretty sure I fixed it for glr
laben: but it may need another one :D 14:24
glr crazy
ab6tract ShimmerFairy: standardized across compilers? no. but stable enough so that you don't need to extern a bunch of crap: suppsoedly. I am just repeating what I came to understand through a previous discussion with FROGGS
nine [Coke]: yeah, that's from merging master
dalek rl6-roast-data: ea7ea59 | coke++ | / (2 files):
another glr run for 'today'
lizmat was I premature in merging nom (again) ? :-(
14:25 tinyblak left
colomon [Coke]++ tadzik++ FROGGS++ 14:25
laben tadzik: didnt see, it's the same fix just without parens. good!
[Coke] m: print 163-48
camelia rakudo-moar 6a1879: OUTPUT«115»
GLRelia rakudo-moar e6c4b9: OUTPUT«115»
[Coke] still not enough to account for the new failures.
14:25 rurban joined
ShimmerFairy ab6tract: I'm not familiar with binding stuff for C++, just that one tidbit about how mangling is not defined by the standard for that reason of avoiding inter-compiler messing :) 14:25
[Coke] rebuilding again...
14:27 tinyblak joined
FROGGS lizmat: I'm starting now to work on CUR 14:27
lizmat FROGGS: cool, shall I join you ?
nine Oh and now I lost a commit in rakudo/glr because I tried to clean up the merge fallout
14:27 itz left
lizmat :-( 14:28
FROGGS lizmat: not yet I think, I need to get into it first
lizmat :-( :-(
laben hoelzro: unfortunately, i dont know my way around moarvm, the changes done for c++ support are totally undecipherable to me
lizmat FROGGS: ok
m: sub a(@a) { say @a.perl }; my &w = &[,]; a [[&w]] <a b c d>; a [,] <a b c d> # I don't think these should be different
camelia rakudo-moar 6a1879: OUTPUT«((("a", "b"), "c"), "d")␤("a", "b", "c", "d")␤»
GLRelia rakudo-moar e6c4b9: OUTPUT«((("a", "b"), "c"), "d")␤("a", "b", "c", "d")␤»
tadzik does anyone know if Jonathan Stowe is on IRC?
lizmat RabidGravy: are you online? 14:29
tadzik: ^^^
tadzik lizmat: thanks :)
RabidGravy I might be
tadzik awesome
hoelzro laben: have you seen the tests for it in Rakudo?
tadzik I tracked you down as an author of cookie tests in HTTP::UA
hoelzro I'm hoping you don't need to look at MoarVM =)
tadzik they break on glr and I have no clue wtf they mean :)
ab6tract ShimmerFairy: I'm only familiar with the fact that it has been a huge PITA in the past but that there is a NativeCall branch that requires no externing, and that this is somehow connext to c++11 14:30
RabidGravy tadzik, it was an ordering thing
lizmat m: sub a(@a) { say @a.perl }; my &w = &[,]; a [[&w]] <a b c d>; a [,] <a b c d> # fixing this will fix Supply.zip/zip-latest
camelia rakudo-moar 6a1879: OUTPUT«((("a", "b"), "c"), "d")␤("a", "b", "c", "d")␤»
GLRelia rakudo-moar e6c4b9: OUTPUT«((("a", "b"), "c"), "d")␤("a", "b", "c", "d")␤»
tadzik RabidGravy: okay, but let's say 03-cookies.t:62
what does 1 mean there? 14:31
14:31 cwDYKARN left
ab6tract m: my $raw = 'd22;e55'; $raw ~~ s:nth(*-2)[\d+] = 675 14:31
camelia ( no output )
GLRelia rakudo-moar e6c4b9: OUTPUT«Attempt to retrieve before :1st match -- :nth(0)␤ in block <unit> at /tmp/tf71xSDyGR:1␤␤»
colomon Method 'build-stderr' not found for invocant of class 'Str' ??? 14:32
tadzik ab6tract: looks like there's 0 matches :) 14:33
colomon finds the string in emmentaler. odd
tadzik m: my $raw = 'd22;e55'; say($raw ~~ m[\d+]).perl
camelia rakudo-moar 6a1879: OUTPUT«「22」␤»
GLRelia rakudo-moar e6c4b9: OUTPUT«␤»
tadzik ha 14:34
dalek ast/glr: e35f48e | ShimmerFairy++ | S03-metaops/cross.t:
Revert "fix cross metaop with nested/itemized lists"

This reverts commit 5e7d42fb821b461fd491173228430d234c2a7e57, which is a change seemingly meant for the glr branch. Breaks on nom.
ast/glr: 624f5b7 | (Stefan Seifert)++ | S03-metaops/cross.t:
Merge branch 'master' into glr

Conflicts: S03-metaops/cross.t
lizmat m: sub a(@a) { say @a.perl }; a [~] <a b c d>
camelia rakudo-moar 6a1879: OUTPUT«Type check failed in binding @a; expected 'Positional' but got 'Str'␤ in sub a at /tmp/gF116ipQ8d:1␤ in block <unit> at /tmp/gF116ipQ8d:1␤␤»
GLRelia rakudo-moar e6c4b9: OUTPUT«Type check failed in binding @a; expected 'Positional' but got 'Str'␤ in sub a at /tmp/WczcASLFPE:1␤ in block <unit> at /tmp/WczcASLFPE:1␤␤»
ast/glr: 60a9ada | (Stefan Seifert)++ | S03-metaops/cross.t:
Fix fallout from reverting a glr commit on master
Heuristic branch merge: pushed 17 commits to rakudo/glr by niner
tadzik regex brokens
RabidGravy tadzik, the test in general means "exactly one each of those" but I can't remember why I put the 14:35
ab6tract tadzik: crap... that's way worse
RabidGravy put the 1 ==
nine t/spec/S02-literals/numeric.t, t/spec/S22-package-format/local.t and t/spec/S32-exceptions/misc.rakudo.moar are failures we imported from nom 14:36
ab6tract m: my $raw = 'd22;e55'; $raw ~~ s:nth(*-2)[\d+] = 675; $raw ~~ s:nth(1) = 555; say $raw
camelia rakudo-moar 6a1879: OUTPUT«5===SORRY!5===␤Unrecognized regex metacharacter ; (must be quoted to match literally)␤at /tmp/dVjQQ3FqxQ:1␤------> 3(*-2)[\d+] = 675; $raw ~~ s:nth(1) = 5557⏏5; say $raw␤Couldn't find terminator =␤at /tmp/dVjQQ3FqxQ:1␤------> 3*-2)[\…»
GLRelia rakudo-moar e6c4b9: OUTPUT«5===SORRY!5===␤Unrecognized regex metacharacter ; (must be quoted to match literally)␤at /tmp/r0xGqrnOLc:1␤------> 3(*-2)[\d+] = 675; $raw ~~ s:nth(1) = 5557⏏5; say $raw␤Couldn't find terminator =␤at /tmp/r0xGqrnOLc:1␤------> 3*-2)[\…»
ab6tract m: my $raw = 'd22;e55'; $raw ~~ s:nth(*-2)[\d+] = 675; $raw ~~ s:nth(1)[\d+] = 555; say $raw 14:37
camelia rakudo-moar 6a1879: OUTPUT«d555;e55␤»
GLRelia rakudo-moar e6c4b9: OUTPUT«Attempt to retrieve before :1st match -- :nth(0)␤ in block <unit> at /tmp/n6CKIZfkyb:1␤␤»
ab6tract that works in glr locally
dalek c: 5c0d9f9 | (Will Coleda)++ | WANTED:
Update WANTED
14:38
laben hoelzro: what is missing for c++ support? fixing bugs or actual features? 14:39
[Coke] 1036 tickets. 14:40
smls [Coke]: Isn't that the wrong repository?
[Coke] smls: ... for what? 14:41
ab6tract laben: FROGGS++ knows more
hoelzro laben: FROGGS is more in the know on that, but I think it's pretty much ready
[Coke] I thought perl6/doc was for docs?
smls [Coke]: For advent posts stuff
[Coke] there is no "repo" for advent posts.
tadzik there is
14:41 telex left
[Coke] but the request is for a tutorial. 14:41
tadzik github.com/perl6/mu/tree/master/mi...dvent-2014
[Coke] for docs. not specifically for an advent post. and we don't record requests for advent post.
FROGGS laben: using it and fixing bugs, aye 14:42
I mean, discovering bugs...
smls [Coke]: Ok.
14:42 telex joined
[Coke] anyway, RT was not the right spot, this is slightly better. 14:42
14:42 spider-mario joined
jdv79 is it right that perl6 --doc actually runs the file? 14:43
tadzik yes
jdv79 so to get the pod all deps have to be available?
tadzik yes
ab6tract m: my $raw = 'd22;e55'; $raw ~~ s:nth(*-2)[\d+] = 675; print "\n"; $raw ~~ s:nth(1) = 555; say $raw
camelia rakudo-moar 6a1879: OUTPUT«5===SORRY!5===␤Unrecognized regex metacharacter ; (must be quoted to match literally)␤at /tmp/dE2JBpjNTk:1␤------> 3 675; print "\n"; $raw ~~ s:nth(1) = 5557⏏5; say $raw␤Couldn't find terminator =␤at /tmp/dE2JBpjNTk:1␤------> 03675; …»
GLRelia rakudo-moar 6c97f3: OUTPUT«5===SORRY!5===␤Unrecognized regex metacharacter ; (must be quoted to match literally)␤at /tmp/LwZBhkwruS:1␤------> 3 675; print "\n"; $raw ~~ s:nth(1) = 5557⏏5; say $raw␤Couldn't find terminator =␤at /tmp/LwZBhkwruS:1␤------> 03675; …»
skids tadzik, ab6tract: :nth is 1-based. use *-1.
laben FROGGS: good! but does it do only simple C-style bindings or does it also do overloads?
ab6tract m: my $raw = 'd22;e55'; $raw ~~ s:nth(*-2)[\d+] = 675; print "\n"; $raw ~~ s:nth(1)[\d+] = 555; say $raw
camelia rakudo-moar 6a1879: OUTPUT«␤d555;e55␤»
GLRelia rakudo-moar 6c97f3: OUTPUT«Attempt to retrieve before :1st match -- :nth(0)␤ in block <unit> at /tmp/Q1J8u71LCB:1␤␤»
ab6tract skids: why would that matter? as demonstrated, this behavior works 14:44
tadzik skids: it's worse than that
ab6tract *on nom
tadzik m: my $raw = 'd22;e55'; say($raw ~~ m[\d+]).perl
camelia rakudo-moar 6a1879: OUTPUT«「22」␤»
GLRelia rakudo-moar 6c97f3: OUTPUT«␤»
ab6tract the whole point is that I need to change the last two elements without toughing anything earlier
tadzik skids: ^
ab6tract *touching 14:45
skids Yeah I changed it when glrizing because nom is wrong.
ab6tract skids: then explain how I can get the behavior I want
skids Ten you need :nth(*,*-1) or vice versa 14:46
FROGGS laben: atm it does mangled symbol lookup but no overloads
ab6tract why change the semantics from "hey it's the same as array access" ?
skids Because it never was.
:nth(1) is the first element, as *opposed* to array access. 14:47
ab6tract m: my $raw = 'd22;e55'; $raw ~~ s:nth(*-1)[\d+] = 675;
camelia ( no output )
GLRelia ( no output )
ab6tract skids: but you have changed the "end of list" semantics
ShimmerFairy Think of the * as the 'n' in 'nth' for WhateverCodes passed to :nth :)
ab6tract but I see your meaning now 14:48
skids Those semantics are meant for 0-based things.
It screws with divisions.
ShimmerFairy *-1 doesn't technically mean "end of list", it's just that array subscripting chooses to give WhateverCode objects the number of elements in a list. 14:49
ab6tract ShimmerFairy: yeah, I get it. thanks :) 14:50
ShimmerFairy you're welcome :)
14:51 fame joined
ab6tract and skids++ for fixing it 14:51
skids np
14:51 pflanze joined
[Coke] is there a way to get git diff to NOT show long line change as multi lines? I normally want that, but I have this one repo where one of the lines is a few 1000 chars. 14:55
14:56 Ven joined
tadzik [Coke]: I think your $PAGER is to blame 14:57
dalek rl6-roast-data: dc546fc | coke++ | / (2 files):
another glr run for 'today'
14:58
pflanze Why do I not get the path that I -I printed when perl6 doesn't find a module?: perl6 -I `pwd`/lib -M Tree -e '' 2>&1|grep `pwd` # returns nothing
[Coke] down to 126 test failures; this is still higher than our min of 80-something.
ShimmerFairy [Coke]: in `less` , I know you can hit arrow right and it'll suddenly convert over to unwrapped lines (since by pressing right you're saying you want to see more of a line). Not sure of any settings beyond that
tadzik pflanze: oh, p5 does that, right?
pflanze: I guess no one ever thought of implementing it :) 14:59
[Coke] tadzik: PAGER= git diff, same thing.
skids Need to add an -S to less invoke somehow.
Different users of less sometimes have their own options env var. 15:00
[Coke] ShimmerFairy: yes, but that actuallys moves the "window" to the right. when you move back to the left again, it re-warps.
*wraps
mvuets export LESS=FRXi
tadzik [Coke]: I just did git diff | less -S and it worked
mvuets or press "-S" while in `less`
tadzik pflanze: if you want to I can probably try to help you implement it in rakudo :) 15:01
or at least discover why it's actually hard together
[Coke] mvuets++ wins, as I can do that on the fly and not need to have 2 different ways to git diff; thanks everyone! 15:02
pflanze tadzik, ok so well it loads the modules lazily, fun. Anyway, it prints a message including the search path when reading the program from stdin.
And typing some code there.
And it doesn't include my path there.
tadzik, a better question would be, how do I get a repl from within a file? (i.e. perl6 myfile.p6, where that file contains a repl call) 15:03
tadzik oh, I'm not sure if you can
but it would be nice to have
you can always use a debugger
dalek p/proc-async-ready: a9f1c5d | hoelzro++ | tools/build/MOAR_REVISION:
Bump MOAR_REVISION for async process ready callback
15:04
skids Funny thing when I was looking at commandline flag emulation I said, heck, why not choose "less" as an example, to find it had completely exhausted the [a-zA-Z] flagspace.
15:04 Ven left
dalek kudo/proc-async-ready: b89f786 | hoelzro++ | src/core/Proc/Async.pm:
Represent Proc::Async state with status

The state of a Proc::Async is more complicated than started vs not started; this is to help some more complicated pending changes
15:04
15:04 dalek left 15:05 ddeimeke left
tadzik BAM 15:05
hoelzro ^ could I get some eyes on github.com/rakudo/rakudo/commit/b89f786cd7?
thoughts on the new Proc::Async API?
15:05 dalek joined, ChanServ sets mode: +v dalek 15:07 llfourn joined
dalek kudo/proc-async-ready: 54cb9f3 | hoelzro++ | src/core/Proc/Async.pm:
Deprecate Proc::Async.start() in favor of ready()/finish()

Proc::Async has a start() method, which prepares a process to be started. After it returns, there is no guarantee that a process is
  *actually* running, and some methods (for example: close-stdin, kill)
will silently fail, and more importantly, sometimes put the child process and the VM's view of it in a state that locks up the program.
So start() is now deprecated; and it has changed so that after it returns, the process is either running or has already exited. If you want to execute actions once a process has been started, you should use ready:
   await $p.ready();
   # do stuff
   await $p.finish();
15:08
hoelzro er, review that one =)
kudo/proc-async-ready: 3feca35 | hoelzro++ | tools/build/NQP_REVISION:
Bump NQP_REVISION for proc async ready callback from MVM
hoelzro sorry for commit spam
moritz hoelzro: .ready doesn't sound like spawning a process to me 15:10
hoelzro moritz: I was thinking .launch before, but I didn't like it 15:11
maybe I should just do .spawn?
moritz hoelzro: are there any operations that work on a Started Proc::Async object, but not on a Starting one?
hoelzro moritz: .kill, .close-stdin
15:11 llfourn left
hoelzro .print, .say, .write 15:12
moritz hoelzro: can't those operations be queued until it has actually started?
hoelzro: that strikes me as more user-friendly than forcing the user to wait
15:13 fame left
dalek kudo/glr: b7bee21 | lizmat++ | src/core/Supply.pm:
An attempt at fixing Supply.zip/zip-latest

Problem is that [,] seems to have changed meaning in the GLR. So, instead, we're giving the no :with case a separate code path. Now we only need to convince the tests that we're doing the right thing.
15:13
kudo/glr: 07a9a0a | lizmat++ | src/core/List.pm:
Merge branch 'glr' of github.com/rakudo/rakudo into glr
hoelzro moritz: part of it was the usage of the idiom "await $p.start" 15:14
15:14 mvuets left
hoelzro jnthn and I thought it read as if you're waiting for the process to start 15:14
15:16 CIAvash joined
dalek ast/glr: 27b20bf | TimToady++ | S03-sequence/misc.t:
remove two wonky sequence tests

We shouldn't use the last value generated from a closure match as a seed for intuiting the next sequence.
15:20
15:21 brrt left 15:22 laouji left
skids m: my @b = 1,2,3; @b[0,1,2] = 4,5; @b.say; # is leaving @b[2] untouched here correct behavior? 15:22
camelia rakudo-moar 6a1879: OUTPUT«4 5 (Any)␤»
GLRelia rakudo-moar 07a9a0: OUTPUT«453␤»
skids Apparently no.
dalek kudo/glr: 2a58180 | TimToady++ | src/core/operators.pm:
implement non-caching 3+ arity sequence ops
15:23
ab6tract skids: I'm still getting some quite strange behavior with Terminal::Print on the GLR 15:28
timotimo what is Terminal::Print? o_O 15:29
ab6tract it seems to still be related nths
TimToady m: (1, 2, 6 ... *)[0]
camelia rakudo-moar 6a1879: OUTPUT«Unable to deduce arithmetic or geometric sequence from 1,2,6 (or did you really mean '..'?)␤ in block <unit> at /tmp/4upm1Sw_cw:1␤␤Actually thrown at:␤ in block <unit> at /tmp/4upm1Sw_cw:1␤␤»
GLRelia rakudo-moar 07a9a0: OUTPUT«Unable to deduce arithmetic or geometric sequence from 1,2,6 (or did you really mean '..'?)␤ in block <unit> at /tmp/qRm7fQhb7L:1␤␤Actually thrown at:␤ in block <unit> at /tmp/qRm7fQhb7L:1␤␤»
TimToady m: (1, 2, 6 ... *)[1]
camelia rakudo-moar 6a1879: OUTPUT«Unable to deduce arithmetic or geometric sequence from 1,2,6 (or did you really mean '..'?)␤ in block <unit> at /tmp/IKGutkSe4a:1␤␤Actually thrown at:␤ in block <unit> at /tmp/IKGutkSe4a:1␤␤»
GLRelia ( no output )
ab6tract timotimo: github.com/ab5tract/Terminal-Print
timotimo: tl;dr -- the foundations of a PP ncurses replacement :) 15:30
TimToady m: (1, 2, 6 ... *)[^1]
camelia rakudo-moar 6a1879: OUTPUT«Unable to deduce arithmetic or geometric sequence from 1,2,6 (or did you really mean '..'?)␤ in block <unit> at /tmp/DIpQdDVCxl:1␤␤Actually thrown at:␤ in block <unit> at /tmp/DIpQdDVCxl:1␤␤»
GLRelia ( no output )
timotimo ooooh
TimToady m: (1, 2, 6 ... *).eager
camelia ( no output )
GLRelia ( no output )
timotimo go damn, the heat is making me feel very uncomfortable
god damn* 15:31
TimToady m: (1, 2, 6 ... *).sink
camelia ( no output )
GLRelia ( no output )
colomon m: say [1..10].shuffle
camelia rakudo-moar 6a1879: OUTPUT«Method 'shuffle' not found for invocant of class 'Array'␤ in block <unit> at /tmp/NSdoF2pGXU:1␤␤»
GLRelia rakudo-moar 2a5818: OUTPUT«Method 'shuffle' not found for invocant of class 'Array'␤ in block <unit> at /tmp/Z_p_YcJ5AI:1␤␤»
colomon pick!
moritz pick( 15:32
TimToady m: (1, 2, 6 ... *).say
moritz *)
camelia rakudo-moar 6a1879: OUTPUT«Unable to deduce arithmetic or geometric sequence from 1,2,6 (or did you really mean '..'?)␤ in block <unit> at /tmp/twCe8D_t_v:1␤␤Actually thrown at:␤ in block <unit> at /tmp/twCe8D_t_v:1␤␤»
GLRelia rakudo-moar 2a5818: OUTPUT«Unable to deduce arithmetic or geometric sequence from 1,2,6 (or did you really mean '..'?)␤ in block <unit> at /tmp/PAqGZa7ka9:1␤␤Actually thrown at:␤ in block <unit> at /tmp/PAqGZa7ka9:1␤␤»
ab6tract timotimo: one day declaring terminal ui will be as easy as passing some Rats to a ui constructors :)
TimToady m: (1, 2, 6 ... *).pick(*)
camelia rakudo-moar 6a1879: OUTPUT«Cannot .pick from an infinite list␤ in block <unit> at /tmp/PdZE0WcKgy:1␤␤Actually thrown at:␤ in block <unit> at /tmp/PdZE0WcKgy:1␤␤»
GLRelia rakudo-moar 2a5818: OUTPUT«Cannot .pick from a lazy list␤ in block <unit> at /tmp/fVzmbWt70R:1␤␤Actually thrown at:␤ in block <unit> at /tmp/fVzmbWt70R:1␤␤»
timotimo ab6tract: i wante dto build something like that a long time ago. i wanted to call it "libcatui"
"kiCk-Ass Terminal UI"
TimToady m: (1, 2, 6 ... *)[1].say 15:33
camelia rakudo-moar 6a1879: OUTPUT«Unable to deduce arithmetic or geometric sequence from 1,2,6 (or did you really mean '..'?)␤ in block <unit> at /tmp/Y0YmBf__vb:1␤␤Actually thrown at:␤ in block <unit> at /tmp/Y0YmBf__vb:1␤␤»
GLRelia rakudo-moar 2a5818: OUTPUT«Nil␤»
ab6tract Terminal::Print is the infrastructure, it's "only" missing async
timotimo: so, plenty of room for catui on top :) 15:34
timotimo cats do like to be on top of things
skids ab6tract: can you find an input to one of those *-template subs that produces different results on GLR vs nom?
TimToady sequences seem to work again, except for the non-failure ^^^ 15:35
timotimo ab6tract: your last-name rocks, btw
ab6tract skids: examples/zig-zag.p6 shows some really terrible behavior
on GLR 15:36
unfortunately GLR requires b937b5c8, and nom requires 2d3557f2
skids Well... there's a failed test in there but I can't see it due to cleared screens. 15:38
skids pipes through a pager
ab6tract skids: and t/basics.t gives a "cannot modify immutable Str" error
ShimmerFairy I never heard why Term had to be changed to Terminal.
ab6tract ShimmerFairy: because Term is an actual thing in Perl 6
so leave that open to things about Terms
that's the idea anyway
also, even if Term:: stayed about terminals, I don't like abbreviations in almost all cases :P 15:39
ShimmerFairy I've never heard of Term though, what is it?
ab6tract everything that is not operator
*an
ShimmerFairy that's term though
hoelzro does anyone have any opposition to a #line directive, as in Perl 5? 15:40
or an idea for something sixier?
ab6tract ShimmerFairy: so what? we are talking about module namespaces. lower case modules are usually pragmas
colomon +1 to #line
ab6tract why does it matter?
[Coke] I get 19 test files failing on the last run. 126 test fails. next run almost done.
ShimmerFairy There are no pragma modules in P6 though o_-
hoelzro: what do you mean by a #line directive? 15:41
ab6tract skids: really tough to build a programmatic test harness for this.. so it requires a human to interpret
15:41 llfourn joined
ab6tract in fact 4..7 are all broken -- should be screens full of hearts, not just one 15:42
TimToady hoelzro: gee, we could allow 'constant $?line =" to work instead of #line
hoelzro ShimmerFairy: the main reason is so that --ll-exception points to src/core/ instead of m-CORE.setting 15:43
ab6tract # at t/basics.t line 17\n # Cannot modify an immutable Str
hoelzro TimToady: so have a $?filename as well?
ShimmerFairy hoelzro: that's what $?LINE is for :)
hoelzro I like that.
ShimmerFairy and $?FILE (?)
hoelzro ah ha!
ShimmerFairy++ TimToady++ # enlightenment
ShimmerFairy but you can't modify them yet :(
15:45 prevost joined
ab6tract skids: 4..6, sorry. test 7 is not "screen full of hearts" related :) 15:46
skids: WARNING: unhandled Failure detected in DESTROY: \n Cannot shift from an empty Array 15:48
that's the one that comes from examples/zig-zag.p6
which are triggered by the nths in the cursor-template 15:49
skids m: gist.github.com/skids/d804af84ed6122ba4a53 # that part seems to work 15:50
camelia rakudo-moar 6a1879: OUTPUT«5===SORRY!5=== Error while compiling /tmp/8W3p6loMV7␤Confused␤at /tmp/8W3p6loMV7:1␤------> 3https:7⏏5//gist.github.com/skids/d804af84ed6122ba␤ expecting any of:␤ colon pair␤»
GLRelia rakudo-moar 2a5818: OUTPUT«5===SORRY!5=== Error while compiling /tmp/HTCoTXv3Ao␤Confused␤at /tmp/HTCoTXv3Ao:1␤------> 3https:7⏏5//gist.github.com/skids/d804af84ed6122ba␤ expecting any of:␤ colon pair␤»
skids hrm something pasted wrong. 15:51
Oh or gists aren't working.
Maybe the :x, :y inputs are not hygenic at some point.
Some extra subunit tests would not hurt this package IMO. 15:52
moritz m: gist.github.com/skids/d804af84ed6122ba4a53
camelia rakudo-moar 6a1879: OUTPUT«qx, qqx is disallowed in restricted setting␤ in sub restricted at src/RESTRICTED.setting:1␤ in sub QX at src/RESTRICTED.setting:11␤ in sub build-cursor-to-template at /tmp/jbGEQi2Ruz:3␤ in block <unit> at /tmp/jbGEQi2Ruz:14␤␤»
GLRelia rakudo-moar 2a5818: OUTPUT«qx, qqx is disallowed in restricted setting␤ in sub restricted at src/RESTRICTED.setting:1␤ in sub QX at src/RESTRICTED.setting:11␤ in sub build-cursor-to-template at /tmp/vf50yATd7W:3␤ in block <unit> at /tmp/vf50yATd7W:14␤␤»
moritz skids: gist URLs don't work with comments 15:53
ab6tract the "cannot shift from empty array" seems to me to indicate something in nth behavior, to my old eyes
hmm... what do you mean by not hygeinic?
moritz ab6tract: have you tried "use fatal;"? 15:54
ab6tract skids: to be sure. I was only beginning to blow some dust off. probably shouldn't have bothered with the GLR at the moment, but it was too tempting
skids Like maybe something is dumping an iterator in :x or something.
ab6tract skids: would that work even when the signature is type constrained to Int ? 15:55
skids Oh, I guess not, type safety++
ab6tract moritz: never needed to. the thing works swimmingly on nom
moritz: until I try to make it use supplies, that is ;) .. which happens periodically 15:57
hoelzro [Coke]: gist.github.com/hoelzro/65d3f16fe294d7d1bb0b 15:59
15:59 lizmat left
hoelzro (summary of what I did) 15:59
timotimo moritz: i ended up not working on a release :|
nine How do I fix Object of type QAST::Op in QAST::WVal, but not in SC?
moritz nine: you need to call $*W.add_object($obj) before sticking an object into a QAST::WVal 16:03
16:04 ab6tract left
nine I think I actually don't want a WVal 16:04
16:04 CIAvash left, smls_ joined
nine I'm really just trying to add a named parameter to a methodcall where the value is computed at runtime 16:04
16:05 smls left
arnsholt nine: What are you working on? Some context might make it easier to figure out 16:05
nine arnsholt: The for statement compiles to a simpo09 16:06
arnsholt: The for statement compiles to a simple .map now. I need to change it to my \tmp := for-list; for-list.map($block, :item(nqp::iscont(for-list));
moritz nine: you can just call .name($name-of-the-named-param) on a QAST::Node 16:07
nine moritz: so like my $iscont := QAST::Op.new(:op('iscont'), QAST::Var.new( :name($for-list-name), :scope('local') ); $iscont.name('item'); $call.push($iscont);? 16:08
16:08 smls_ is now known as smls
moritz nine: yes 16:08
16:09 cognominal left
moritz nine: and you can use $qast.unique('for-list') to generate a unique name for a temporary 16:09
arnsholt Oh, that's a neat feature. Where $qast is any QAST object? 16:10
16:10 colomon left 16:11 colomon joined
nine I've seen QAST::Node.unique('foo') 16:11
moritz: is it .name or .named? 16:12
arnsholt Right, so probably a static method on QAST::Node then, so should work on most anything
.named, IIRC
16:13 prevost left
nine \o/ It works! 16:15
16:18 ddeimeke joined, ddeimeke left
ugexe whats the point of deprecating proc::async.start if the functionality of the deprecated function completely changes? 16:18
why not leave it the same or just remove it 16:19
16:21 itz joined, yqt joined 16:22 domidumont joined, virtualsue joined
ugexe oh i am mistaken. .start awaits on the promise that its ready, not finished 16:24
16:28 colomon left 16:29 FROGGS left 16:34 lolisa joined
_itz smls: I get an rss parsing error with your feed. see PR for details 16:34
smls Well, it's an atom feed :P 16:35
smls will take a look
dalek kudo/glr: 7aee0ac | (Stefan Seifert)++ | src/ (2 files):
Make for $@arr { } iterate only once

my @arr = 1, 2; for @arr {} will iterate twice for $@arr {} will iterate once
  @arr.map: {} will iterate twice
  $@arr.map: {} will iterate twice
16:37
nine moritz++
arnsholt++
Thanks guys!
16:40 khw left, colomon joined
smls _itz: I had extraneous whitespace at the start of my feed; Should be fixed now. 16:40
dalek ast/glr: 95e773b | (Stefan Seifert)++ | integration/advent2011-day16.t:
Change test to test for the actually desired for semantics

Make for $@arr { } iterate only once
my @arr = 1, 2; for @arr {} will iterate twice for $@arr {} will iterate once
  @arr.map: {} will iterate twice
  $@arr.map: {} will iterate twice
16:44
rl6-roast-data: 854ad44 | coke++ | / (2 files):
another glr run for 'today'
16:44 colomon left 16:46 larion joined 16:47 lizmat joined
[Coke] who was it owned twitter/perlhex? 16:47
laben m: sub foo(*@m) {}; foo(Hash) 16:48
camelia ( no output )
GLRelia rakudo-moar 7aee0a: OUTPUT«Invocant requires an instance of type Hash, but a type object was passed. Did you forget a .new?␤ in sub foo at /tmp/D3LAGukejc:1␤ in block <unit> at /tmp/D3LAGukejc:1␤␤»
laben m: sub foo(*@m) {}; foo(List)
camelia ( no output )
GLRelia rakudo-moar 7aee0a: OUTPUT«Invocant requires an instance of type List, but a type object was passed. Did you forget a .new?␤ in sub foo at /tmp/XpbHDeZGxW:1␤ in block <unit> at /tmp/XpbHDeZGxW:1␤␤»
16:48 FROGGS joined
laben anyone can give me a hint on ^^^ to know where to begin to investigate? 16:48
dalek kudo/glr: 74cc8c8 | lizmat++ | src/core/Supply.pm:
One step closer to fixing Supply.zip-latest
16:49
kudo/glr: f35d9de | lizmat++ | src/ (3 files):
Merge branch 'glr' of github.com/rakudo/rakudo into glr
16:49 domm joined
JimmyZ nine: why cant just nqp:iscont(SELF) ?? (SELF,) !! SELF 16:52
16:54 MilkmanDan left
skids ab6stract: I found the source of thise failures but will have to spectest (twice) and then file a PR and wait for it to be pulled. 16:54
ab6tract: in the meantime, I left some unrelated GLR patches in that gist, and running everything 2> /dev/null should "work" 16:55
oh, he left. 16:56
laben it sounds like it mistakenly believes the Signature has a definedness check, but it actually has not
lizmat dinner at SPW wrapup&
the network is down at the SPW, so only connectivity is on 3G
and not many people have that 16:57
later!
afk&
16:57 lizmat left
JimmyZ nine: oh, I know now, for is abit different from map 16:58
17:06 MilkmanDan joined 17:10 kaare__ left 17:18 itz left
skids thought "for $foo" was going to stay exactly $foo.map as a sugar. 17:21
17:21 virtualsue left 17:22 yqt left
skids Though I really don't have a preference. 17:22
17:25 rurban left 17:29 kaare__ joined 17:30 espadrine left 17:31 tinyblak left 17:33 kaare__ left 17:44 ShimmerFairy left
nine skids: people seem to expect something else. And this way we stick to the GLRR rules regarldless of how for is implemented 17:46
17:48 kaare__ joined, skids left, cognominal joined 17:56 ShimmerFairy joined
dalek kudo/glr: 918cf63 | skids++ | src/core/Str.pm:
Prevent generation of X::Cannot::Empty Failures in :nth processing.

There was a case where the result of shift() could escape without being handled. Probably more efficient not to generate any Failure objects in the first place, so do that.
ab[56]tract++ for unearthing with a long-running script.
18:15
kudo/glr: a977dc3 | FROGGS++ | src/core/Str.pm:
Merge pull request #520 from skids/glr3

Prevent generation of X::Cannot::Empty Failures in :nth processing.
18:21 lolisa left 18:22 ChoHag joined, lolisa joined
ChoHag If I have a class which does Positional, how do I write a for loop for each of its elements? 18:22
18:22 perl6_newbee joined
ChoHag Pre-GLR. 18:22
perl6_newbee Hi all
Can someone help me and explain how I can store hashes in an array? 18:23
ChoHag The naïve "my $foo = Class::Which::Does::Positional.new(...); for $foo { ... }" didn't work. 18:24
m: my %foo = a => 1, b => 2; my @bar = 1, %foo, 2; @bar.perl.say
camelia rakudo-moar 6a1879: OUTPUT«[1, :a(1), :b(2), 2]<>␤»
GLRelia rakudo-moar a977dc: OUTPUT«[1, {:a(1), :b(2)}, 2]␤»
ChoHag Like that.
Hmm GLR changes the auto-flattening behaviour. 18:25
m: my %foo = a => 1, b => 2; my @bar = 1, {%foo}, 2; @bar.perl.say
camelia rakudo-moar 6a1879: OUTPUT«[1, {:a(1), :b(2)}, 2]<>␤»
GLRelia rakudo-moar a977dc: OUTPUT«[1, {:a(1), :b(2)}, 2]␤»
perl6_newbee hmm
I always get either a block or LoL.
ChoHag m: my %foo = a => 1, b => 2; my @bar = 1, %{%foo}, 2; @bar.perl.say 18:26
camelia rakudo-moar 6a1879: OUTPUT«5===SORRY!5=== Error while compiling /tmp/Wn3zMAqLlq␤Unsupported use of %{%foo}; in Perl 6 please use %(%foo)␤at /tmp/Wn3zMAqLlq:1␤------> 3o = a => 1, b => 2; my @bar = 1, %{%foo}7⏏5, 2; @bar.perl.say␤»
GLRelia rakudo-moar a977dc: OUTPUT«5===SORRY!5=== Error while compiling /tmp/JeUb06h4dQ␤Unsupported use of %{%foo}; in Perl 6 please use %(%foo)␤at /tmp/JeUb06h4dQ:1␤------> 3o = a => 1, b => 2; my @bar = 1, %{%foo}7⏏5, 2; @bar.perl.say␤»
ChoHag m: my %foo = a => 1, b => 2; my @bar = 1, %(%foo), 2; @bar.perl.say
camelia rakudo-moar 6a1879: OUTPUT«[1, :a(1), :b(2), 2]<>␤»
GLRelia rakudo-moar a977dc: OUTPUT«[1, {:a(1), :b(2)}, 2]␤»
ChoHag heh
18:28 kaare_ joined
timotimo [Coke]: soundcloud.com/mcfreshdawg3000/cra...me-re-edit 18:28
perl6_newbee @array.push(%( key1 => 1 )) Is that the right syntax? 18:29
18:29 breinbaas left
[ptc] m: @array.push(%( key1 => 1 )) 18:30
camelia rakudo-moar 6a1879: OUTPUT«5===SORRY!5=== Error while compiling /tmp/CFvSVYzUVE␤Variable '@array' is not declared. Did you mean any of these?␤ array␤ Array␤␤at /tmp/CFvSVYzUVE:1␤------> 3<BOL>7⏏5@array.push(%( key1 => 1 ))␤»
GLRelia rakudo-moar a977dc: OUTPUT«5===SORRY!5=== Error while compiling /tmp/wSTokQ6hBg␤Variable '@array' is not declared. Did you mean any of these?␤ array␤ Array␤␤at /tmp/wSTokQ6hBg:1␤------> 3<BOL>7⏏5@array.push(%( key1 => 1 ))␤»
18:30 kaare__ left
[ptc] m: my @array; @array.push(%( key1 => 1 )) 18:31
camelia ( no output )
GLRelia ( no output )
[ptc] m: my @array; @array.push(%( key1 => 1 )).perl.say
camelia rakudo-moar 6a1879: OUTPUT«[:key1(1)]<>␤»
GLRelia rakudo-moar a977dc: OUTPUT«[:key1(1)]␤»
[ptc] perl6_newbee: if you wanted to push a hash onto an array, that's probably ok 18:32
perl6_newbee @ptc: thx
18:32 tinyblak joined
perl6_newbee %( ... ) is the way to do it thx 18:36
18:36 lolisa left 18:37 tinyblak left 18:38 rurban joined 18:39 prevost joined 18:42 colomon joined
dalek rl6-roast-data: 7c0b649 | coke++ | / (2 files):
another glr run for 'today'
18:43
18:44 darutoko- joined
[Coke] back down to 18 files failing, 106 failures. 18:45
18:46 darutoko left
colomon nooo… where’d we backslide? 18:47
18:47 lizmat joined, rangerprice joined 18:48 prevost left
[Coke] post nom merge. 18:48
er, -with-, not post. 18:49
Who was the perlhex fellow, anyone remember his name?
tadzik nine's gonna be pissed :o 18:50
[Coke]: muvets
I think
18:50 eternaleye joined, lizmat_ joined
tadzik mvuets! 18:51
act.perl-workshop.ch/spw2015/user/2625
18:53 lizmat left 18:55 dha joined
[Coke] danke 18:55
dha wonders if he's having a perl6 problem or a un*x problem...
[Coke] use grammars, and then you'll hav.... sorry, what's the problem? 18:56
18:56 Guest12682 left 18:57 Loren_ left
dha trying to build panda. it's bailing out and asking me if /Users/dha/rakudo/install/share/perl6/site/bin is in your PATH. It is, indeed, the first thing in my PATH 18:58
ChoHag How can you have a un*x problem on a mac? 18:59
18:59 Loren joined
dha OS X is, technically, a bsd derivative. :-) 19:00
19:00 perl6_newbee left
dha If you prefer, though, I could characterize it as a PATH problem, specifically. 19:00
19:01 Loren left
ChoHag Well having not built panda, all I can do is disparage your choice of quasi-operating-system. 19:01
19:01 Loren joined
dha I suppose you could do that, but it would not make me think much of your advice. :-) 19:02
19:03 Loren left
dha Maybe I'll just try rebuilding rakudo again from scratch. *shrug* 19:03
ChoHag All I can suggest that's con-, rather than de-, structive is that somebody has taken an automake system and wrapped "magic" around it. 19:04
You need to debug the magic.
As per fucking usual.
dha i'm... unclear how that would have happened since I last built things, but, I suppose that's possible. 19:05
ChoHag Because fuck magic.
[Coke] dha: make sure you're using a recent panda? 19:07
19:07 rindolf left
dha did a git pull just before rebuilding. 19:07
ChoHag There you go. 19:08
[Coke] m: .say for OUTER::OUTER::.values;
camelia rakudo-moar 6a1879: OUTPUT«(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)…»
GLRelia rakudo-moar a977dc: OUTPUT«(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)␤(Mu)…»
ChoHag You probably didn't clean properly and it broke the magic.
[Coke] m: my $stash = OUTER::OUTER::; .say for %stash.map({ %stash{$_} })
camelia rakudo-moar 6a1879: OUTPUT«5===SORRY!5=== Error while compiling /tmp/lkxLpIIKDU␤Variable '%stash' is not declared. Did you mean any of these?␤ $stash␤ Stash␤␤at /tmp/lkxLpIIKDU:1␤------> 3= OUTER::OUTER::; .say for %stash.map({ 7⏏5%stash{$_} })␤»
GLRelia rakudo-moar a977dc: OUTPUT«5===SORRY!5=== Error while compiling /tmp/UCJEFuD5sI␤Variable '%stash' is not declared. Did you mean any of these?␤ $stash␤ Stash␤␤at /tmp/UCJEFuD5sI:1␤------> 3= OUTER::OUTER::; .say for %stash.map({ 7⏏5%stash{$_} })␤»
dha That is possible.
19:08 rindolf joined
ChoHag $ grep -i magic .profile echo "Fuck magic" 19:08
[Coke] m: my %stash = OUTER::OUTER::; .say for %stash.map({ %stash{$_} })
camelia rakudo-moar 6a1879: OUTPUT«Use of uninitialized value <element> of type Mu in string context in block at /tmp/bInF751r1z:1␤(Any)␤Use of uninitialized value <element> of type Mu in string context in block at /tmp/bInF751r1z:1␤(Any)␤Use of uninitialized value <element> of …»
GLRelia rakudo-moar a977dc: OUTPUT«Use of uninitialized value <element> of type Mu in string context in block at /tmp/7yOxAcFrNG:1␤(Any)␤Use of uninitialized value <element> of type Mu in string context in block at /tmp/7yOxAcFrNG:1␤(Any)␤Use of uninitialized value <element> of …»
ChoHag It's not like make was specifically designed for exactly that eventuality or anything. 19:09
colomon m: (1, 2, 5 … 10).perl.say
camelia rakudo-moar 6a1879: OUTPUT«Unable to deduce arithmetic or geometric sequence from 1,2,5 (or did you really mean '..'?)␤ in block <unit> at /tmp/yTv_iyXqXv:1␤␤Actually thrown at:␤ in block <unit> at /tmp/yTv_iyXqXv:1␤␤»
GLRelia rakudo-moar a977dc: OUTPUT«(Failure.new(exception => X::Sequence::Deduction.new(from => "1,2,5"), backtrace => Backtrace.new))␤»
rangerprice It would be great for perl to be a statically typed language
moritz m: signal(SIGINT).tap( {die} );
camelia ( no output )
GLRelia rakudo-moar a977dc: OUTPUT«Can only use merge to combine defined Supply objects␤ in any at src/gen/m-Metamodel.nqp:1679␤ in block <unit> at /tmp/q3MKHrjqsk:1␤␤»
moritz rangerprice: it wouldn't be perl
^^ this is a new failure, I think
[Coke] .ask hoelzro can you review RT #123154 ? 19:10
yoleaux [Coke]: I'll pass your message to hoelzro.
synbot6 Link: rt.perl.org/rt3/Public/Bug/Display...?id=123154
19:10 tybalt89 joined
ChoHag Is there a counter variable in a for loop (for $thing[0..*] { ... }) which will tell me which element of the list is being processed? 19:11
[Coke] m: class Sinker { method sink() { say "Blub" } }; my $i = 0; while $i < 5 { $i++; Sinker.new(); }; 1
camelia rakudo-moar 6a1879: OUTPUT«WARNINGS:␤Useless use of constant integer 1 in sink context (line 1)␤Blub␤Blub␤Blub␤Blub␤Blub␤»
GLRelia rakudo-moar a977dc: OUTPUT«WARNINGS:␤Useless use of constant integer 1 in sink context (line 1)␤Blub␤Blub␤Blub␤Blub␤Blub␤»
19:11 tybalt89 left
[Coke] m: class Sinker { method sink() { say "Blub" } }; for ^5 { Sinker.new(); }; 1 19:11
camelia rakudo-moar 6a1879: OUTPUT«WARNINGS:␤Useless use of constant integer 1 in sink context (line 1)␤»
GLRelia rakudo-moar a977dc: OUTPUT«WARNINGS:␤Useless use of constant integer 1 in sink context (line 1)␤»
colomon ChoHag: no, but you can $thing[0..*].kv to get pairs of index and item at that index
moritz ChoHag: you can use for @mylist.kv -> $index, $value { ... } 19:12
timotimo "use rammars
sorry
why can i start typing into ssh, notice the connection is dead, kill the connection and reconnect and in the mean time the characters have actually been sent ?!?
tadzik ssh funni 19:13
ChoHag I think that is exactly the right way to do it.
tadzik ¯\(°_o)/¯
colomon m: (1, 2, 6 ... *).perl.say
camelia rakudo-moar 6a1879: OUTPUT«(ListIter.new,)␤»
GLRelia rakudo-moar a977dc: OUTPUT«Cannot .elems a lazy list␤ in block <unit> at /tmp/Qz3BDcK8Yq:1␤␤Actually thrown at:␤ in block <unit> at /tmp/Qz3BDcK8Yq:1␤␤»
ChoHag Because your assumption was incorrect.
Tho connection wasn't dead.
moritz timotimo: because you're not using mosh :-)
ChoHag You were just impatient. 19:14
colomon thought about trying mosh on the plane flight.
timotimo weird.
colomon m: (1, 2, 6 ... *).say
camelia rakudo-moar 6a1879: OUTPUT«Unable to deduce arithmetic or geometric sequence from 1,2,6 (or did you really mean '..'?)␤ in block <unit> at /tmp/ukb8YlCqJ1:1␤␤Actually thrown at:␤ in block <unit> at /tmp/ukb8YlCqJ1:1␤␤»
GLRelia rakudo-moar a977dc: OUTPUT«Unable to deduce arithmetic or geometric sequence from 1,2,6 (or did you really mean '..'?)␤ in block <unit> at /tmp/uzItt6IMrl:1␤␤Actually thrown at:␤ in block <unit> at /tmp/uzItt6IMrl:1␤␤»
ChoHag Once I was working over a VPN in the morning, closed my laptop, commuted into the office (~ 30 minutes) and when I sat down my connections were still open.
TCP is even better than what we take for granted. 19:15
colomon I just had an ssh connection to hack survive my laptop being shutdown for a couple of hours. Was completely astounded
ChoHag shut down or put to sleep? 19:16
moritz probably hybernated
with ssh connections, I get the impression that they survive a network outage more often if I don't press a key while the network is down 19:17
colomon laptop closed, slept I would imagine. Certianly didn’t burn any noticable amount of battery
ChoHag That has been my untested experience.
tadzik yeah, if you don't do anything they don't have a reason to think anything is wrong :) 19:18
moritz
.oO( what do SSH and oppressive regimes have in common? :-)
tadzik hehe
colomon lizmat_, nine: looking at S03-sequence/misc.t, it seems like rakudo is working properly but throws-like is not verifying it properly? 19:19
tadzik . o O ( living in a shel )
ChoHag Oppressive regimes assume something is wrong whether you do anything or not. 19:20
lizmat_ colomon: the throws-like could well be faulty 19:21
ChoHag It's the self-proclaimed non-oppressive regimes which make that assumption.
dha Bleh. Pod::Parser is failing tests. And possibly because a pull request from may hasn't been merged. God, software is awful. 19:22
colomon m: use Test; throws-like { 1, 2, 5 ... 10 }, X::Sequence::Deduction, from => '1,2,5'
camelia rakudo-moar 6a1879: OUTPUT« 1..3␤ ok 1 - code dies␤ ok 2 - right exception type (X::Sequence::Deduction)␤ ok 3 - .from matches 1,2,5␤ok 1 - did we throws-like X::Sequence::Deduction?␤»
GLRelia rakudo-moar a977dc: OUTPUT« 1..3␤ not ok 1 - code dies␤ ␤# Failed test 'code dies'␤# at /tmp/1NKM6uxdXN line 1␤ ok 2 - # SKIP Code did not die, can not check exception␤ ok 3 - # SKIP Code did not die, can not check exception␤ # Looks like you failed 1…»
ChoHag Anyway on that note...
tadzik ...there's Pod::Parser
?
19:22 ChoHag left
tadzik oh, that parses POD, not Pod 19:23
TimToady colomon, lizmat_: no, the problem is sequences don't seem to be responding to .sink, .eager, or .[5]
tadzik fairy nuff
TimToady they do respond to .[0], go figure
colomon hmmm?
dha ...what's the difference between POD and Pod? 19:24
tadzik POD is Perl 5
Pod is Perl 6
TimToady m: (1, 2, 6 ... *)[0]
camelia rakudo-moar 6a1879: OUTPUT«Unable to deduce arithmetic or geometric sequence from 1,2,6 (or did you really mean '..'?)␤ in block <unit> at /tmp/wtaPd6pZSX:1␤␤Actually thrown at:␤ in block <unit> at /tmp/wtaPd6pZSX:1␤␤»
GLRelia rakudo-moar a977dc: OUTPUT«Unable to deduce arithmetic or geometric sequence from 1,2,6 (or did you really mean '..'?)␤ in block <unit> at /tmp/jaA3xC4bdy:1␤␤Actually thrown at:␤ in block <unit> at /tmp/jaA3xC4bdy:1␤␤»
19:24 bin_005_l joined
dha Ah. ok. 19:24
TimToady m: (1, 2, 6 ... *)[1]
camelia rakudo-moar 6a1879: OUTPUT«Unable to deduce arithmetic or geometric sequence from 1,2,6 (or did you really mean '..'?)␤ in block <unit> at /tmp/JrSDXMM4qX:1␤␤Actually thrown at:␤ in block <unit> at /tmp/JrSDXMM4qX:1␤␤»
GLRelia ( no output )
lizmat_ TimToady: that feels like it's being considered a scalar, not a list/array
dha Yes.
tadzik Yeah, to avoid confusion
dha Ha!
tadzik :)
like PERL and Perl
PERL is the old one
dha In that case we should call it Bruce.
TimToady it's return a List, currently experimenting with forcing return of Seq
dha Well, PERL is generally considered... wrong. :-) 19:25
tadzik it used to be right though, wasn't it?
colomon TimToady: that’s a different issue, isn’t it?
19:25 domidumont left
[Coke] PERL has never been right, IME, but I'm a n00b. 19:25
dha tadzik - I don't think so. 19:26
[Coke] I take -1 off a resume whenever I see it.
tadzik I may be wrong :)
dalek kudo/glr: 9ecd136 | lizmat++ | src/core/Supply.pm:
Fix Supply.zip
19:27
dha from the faqs: But never write "PERL",
because perl is not an acronym.
colomon …everything respond to .[0], right?
dalek kudo/glr: 817a11a | (Stefan Seifert)++ | src/core/Array.pm:
Have Array.List actually return a plain List not an Array

For when we use an Array to collect some results but really want to return a plain List to not leak implementation details.
The generic .List coercer cannot just re-bind $!reified however, because we don't know if the Array is being changed afterwards, so we have to clone first.
19:28
rl6-roast-data: a32cbf6 | coke++ | / (2 files):
another glr run for 'today'
19:29
colomon m: my $code = { 1, 2, 5 …. 10 }; $code()
camelia rakudo-moar 6a1879: OUTPUT«5===SORRY!5=== Error while compiling /tmp/aRYHd44cQf␤Unsupported use of . to concatenate strings; in Perl 6 please use ~␤at /tmp/aRYHd44cQf:1␤------> 3my $code = { 1, 2, 5 …. 7⏏0510 }; $code()␤»
GLRelia rakudo-moar a977dc: OUTPUT«5===SORRY!5=== Error while compiling /tmp/ucmCj89ktb␤Unsupported use of . to concatenate strings; in Perl 6 please use ~␤at /tmp/ucmCj89ktb:1␤------> 3my $code = { 1, 2, 5 …. 7⏏0510 }; $code()␤»
19:30 colomon_ joined
[Coke] still failing 19 files; number of test failures INCREASED to 107 from 106 19:30
timotimo :(
colomon_ m: my $code = { 1, 2, 5 ... 10 }; $code()
camelia rakudo-moar 6a1879: OUTPUT«Unable to deduce arithmetic or geometric sequence from 1,2,5 (or did you really mean '..'?)␤ in block <unit> at /tmp/cPBet9HRhW:1␤␤Actually thrown at:␤ in block <unit> at /tmp/cPBet9HRhW:1␤␤»
GLRelia ( no output )
[Coke] looks like roast changed since the last run, but not rakudo. 19:31
dalek kudo/glr: a057dd4 | moritz++ | src/core/Supply.pm:
Supply.merge should flatten its args

otherwise signal(SIGINT) dies.
[Coke] ... kicking off another one. 19:32
moritz 14 test files failed with my last run after this commit
[Coke] here's hoping.
was it 19 before? ;)
moritz dunno
colomon_ m: my $code = { 1, 2, 5 ... 10 }; say $code().perl 19:33
camelia rakudo-moar 6a1879: OUTPUT«Unable to deduce arithmetic or geometric sequence from 1,2,5 (or did you really mean '..'?)␤ in block <unit> at /tmp/S4ZUlS1mWC:1␤␤Actually thrown at:␤ in block <unit> at /tmp/S4ZUlS1mWC:1␤␤»
GLRelia rakudo-moar 9ecd13: OUTPUT«(Failure.new(exception => X::Sequence::Deduction.new(from => "1,2,5"), backtrace => Backtrace.new))␤»
colomon_ m: my $code = { 1, 2, 5 ... 10 }; say $code()
camelia rakudo-moar 6a1879: OUTPUT«Unable to deduce arithmetic or geometric sequence from 1,2,5 (or did you really mean '..'?)␤ in block <unit> at /tmp/VeIzxvrYQh:1␤␤Actually thrown at:␤ in block <unit> at /tmp/VeIzxvrYQh:1␤␤»
GLRelia rakudo-moar 9ecd13: OUTPUT«Unable to deduce arithmetic or geometric sequence from 1,2,5 (or did you really mean '..'?)␤ in block <unit> at /tmp/7v24bO_Ubt:1␤␤Actually thrown at:␤ in block <unit> at /tmp/7v24bO_Ubt:1␤␤»
19:33 FROGGS left
TimToady .[0], .perl, and .gist are sufficient to trigger it; .eager, .sink, and .[5] are not 19:34
and making it return a Seq doesn't help
19:35 colomon left
dalek kudo/glr: 81229f7 | TimToady++ | src/core/operators.pm:
allow more smartmatches than just Callable on ...
19:35
kudo/glr: 8f3e7b3 | TimToady++ | src/core/operators.pm:
make sequence op return a Seq, duh
19:36 colomon joined
dalek ast/glr: 0fcc9eb | TimToady++ | S03-sequence/misc.t:
seq op returns a Seq now
19:36
[Coke] m: Int.Range.say; Num.Range.say; 19:37
camelia rakudo-moar 6a1879: OUTPUT«-Inf..Inf␤-Inf..Inf␤»
GLRelia rakudo-moar a057dd: OUTPUT«(timeout)»
[Coke] ^^ There's another GLR bug. 19:38
colomon TimToady: Type check failed in binding; expected 'Positional' but got 'Seq' 19:39
that’s my @rt80574 := -> { 'zero', 'one' } ... *;
[Coke] up to 9 [GLR] tickets in rt. 19:40
smls I wrote my first Perl 6 blog post: postcircumfix.com/2015-08-29-untan...-hierarchy \o/
19:41 Sgeo_ joined
TimToady colomon: see my last checkin above 19:41
dha smls - congratulations!
colomon TimToady: right, using that commit is how I got the new test failure
TimToady though why that doesn't coerce to list automatically, I'm not sure
colomon TimToady: should binding coerce?
dha smls - ...except I'm getting a 404 from that. :-/ 19:42
TimToady colomon: no, the checking to the misc.t
smls dha: huh?
tadzik smls: ooh, nice :)
smls dha: Did something add a slash to the end of the URL maybe? 19:43
my current nginx config doesn;t support that
dha oh, maybe. Let me look.
ah, there we go.
19:44 Sgeo left
dalek kudo/glr: 287360b | lizmat++ | src/core/Array.pm:
Fix typo of nine++
19:44
TimToady m: constant @foo = 1,2,4 ... *; say @foo.WHAT 19:45
camelia rakudo-moar 6a1879: OUTPUT«(List)␤»
GLRelia rakudo-moar 8f3e7b: OUTPUT«(Seq)␤»
TimToady m: constant @foo = 1,2,4 ... *; say @foo[10]
camelia rakudo-moar 6a1879: OUTPUT«1024␤»
GLRelia rakudo-moar 8f3e7b: OUTPUT«1024␤»
dha Hm. I'm trying to figure out a one-liner but I'm obviously missing some kind of magic needed.
TimToady m: my @foo = 1,2,4 ... *; say @foo[10]
camelia rakudo-moar 6a1879: OUTPUT«1024␤»
GLRelia rakudo-moar 8f3e7b: OUTPUT«1024␤»
19:46 colomon left
dha To get the lines in a file that aren't blank, I can use perl6 -e 'lines.grep(/\S/).elems.say' So, if I wanted to find the *blank* lines... how would I negate that grep condition? 19:47
(I could use, for instance, perl6 -e 'lines.grep(/^\s$/).elems.say' but this has me curious.
)
TimToady m: constant @foo = -> { 'a', 'b' } ... *; say @foo.WHAT
camelia rakudo-moar 6a1879: OUTPUT«(List)␤»
GLRelia rakudo-moar 8f3e7b: OUTPUT«(Seq)␤»
TimToady m: my @foo := -> { 'a', 'b' } ... *; say @foo.WHAT 19:48
camelia rakudo-moar 6a1879: OUTPUT«(List)␤»
GLRelia rakudo-moar 8f3e7b: OUTPUT«Type check failed in binding; expected 'Positional' but got 'Seq'␤ in block <unit> at /tmp/JX0x8oXuWM:1␤␤»
TimToady m: my (*@foo) ::= -> { 'a', 'b' } ... *; say @foo.WHAT
camelia rakudo-moar 6a1879: OUTPUT«(Array)␤»
GLRelia rakudo-moar 8f3e7b: OUTPUT«(Array)␤»
19:49 llfourn left
laben m: lines.grep(/\S/).elems.say 19:49
camelia rakudo-moar 6a1879: OUTPUT«17␤»
GLRelia rakudo-moar 8f3e7b: OUTPUT«(timeout)»
dha The first thing that occurred to me was perl6 -e 'lines.grep(! /\S/).elems.say' but that doesn't seem to work.
TimToady m: my (@foo) ::= (-> { 'a', 'b' } ... *); say @foo.WHAT
camelia rakudo-moar 6a1879: OUTPUT«Too few positionals passed; expected 1 argument but got 0␤ in block <unit> at /tmp/ZmQVMMMn9v:1␤␤»
GLRelia rakudo-moar 8f3e7b: OUTPUT«Too few positionals passed; expected 1 argument but got 0␤ in block <unit> at /tmp/evmb7wro3c:1␤␤»
laben m: lines.grep(not /\S/).elems.say
camelia rakudo-moar 6a1879: OUTPUT«Method 'match' not found for invocant of class 'Any'␤ in block <unit> at /tmp/cyRP1gkE7J:1␤␤»
GLRelia rakudo-moar 8f3e7b: OUTPUT«Method 'match' not found for invocant of class 'Any'␤ in block <unit> at /tmp/hVCjPiwl1G:1␤␤»
laben m: lines.grep(* !~~ /\S/).elems.say 19:50
camelia rakudo-moar 6a1879: OUTPUT«5===SORRY!5=== Error while compiling /tmp/7AKQ2aIW7u␤Variable '&infix:<!~~>' is not declared␤at /tmp/7AKQ2aIW7u:1␤------> 3lines.grep(* 7⏏5!~~ /\S/).elems.say␤»
GLRelia rakudo-moar 8f3e7b: OUTPUT«5===SORRY!5=== Error while compiling /tmp/TU99BRiesM␤Variable '&infix:<!~~>' is not declared␤at /tmp/TU99BRiesM:1␤------> 3lines.grep(* 7⏏5!~~ /\S/).elems.say␤»
laben m: lines.grep(* !~ /\S/).elems.say
camelia rakudo-moar 6a1879: OUTPUT«5===SORRY!5=== Error while compiling /tmp/9bFUcHkCx9␤Unsupported use of !~ to do negated pattern matching; in Perl 6 please use !~~␤at /tmp/9bFUcHkCx9:1␤------> 3lines.grep(* !~ 7⏏5/\S/).elems.say␤»
GLRelia rakudo-moar 8f3e7b: OUTPUT«5===SORRY!5=== Error while compiling /tmp/4dYE5WUxvK␤Unsupported use of !~ to do negated pattern matching; in Perl 6 please use !~~␤at /tmp/4dYE5WUxvK:1␤------> 3lines.grep(* !~ 7⏏5/\S/).elems.say␤»
laben wait why does it say infix !~~ not declared?
TimToady colomon_: anyway, it's probably a bug, just not a sequence op bug
laben m: lines.grep({ $_ !~~ /\S/}).elems.say 19:51
camelia rakudo-moar 6a1879: OUTPUT«3␤»
GLRelia rakudo-moar 287360: OUTPUT«(timeout)» 19:52
19:52 cognominal left
dha oh. the star thing. I tried $_ !~~ /\S/ ... 19:53
laben it works but dunno why it hangs on glr 19:54
TimToady m: lines.grep(none(/\S/)).elems.say 19:55
camelia rakudo-moar 6a1879: OUTPUT«3␤»
dha Huh. It wasn't working for me. let me try that again...
GLRelia rakudo-moar 287360: OUTPUT«(timeout)»
TimToady m: lines.grep(*).elems.say
camelia rakudo-moar 6a1879: OUTPUT«20␤»
[Coke] r: for our $:: ();
camelia rakudo-{moar,jvm} 6a1879: OUTPUT«5===SORRY!5=== Error while compiling /tmp/tmpfile␤Signatures as constraints on variables not yet implemented. Sorry. ␤at /tmp/tmpfile:1␤------> 3for our $:: ()7⏏5;␤»
GLRelia rakudo-moar 287360: OUTPUT«(timeout)»
rakudo-jvm 287360: OUTPUT«Can't open perl script "/home/steve/rakudo-inst/bin/eval-client.pl": No such file or directory␤»
..rakudo-moar 287360: OUTPUT«5===SORRY!5=== Error while compiling /tmp/tmpfile␤Signatures as constraints on variables not yet implemented. Sorry. ␤at /tmp/tmpfile:1␤------> 3for our $:: ()7⏏5;␤»
dha oh. I forgot to use braces. duh.
[Coke] m: $:: 19:56
camelia rakudo-moar 6a1879: OUTPUT«5===SORRY!5=== Error while compiling /tmp/qyCyFL1Z6t␤Variable '$' is not declared␤at /tmp/qyCyFL1Z6t:1␤------> 3<BOL>7⏏5$::␤»
GLRelia rakudo-moar 287360: OUTPUT«5===SORRY!5=== Error while compiling /tmp/chCtyvb9XT␤Variable '$' is not declared␤at /tmp/chCtyvb9XT:1␤------> 3<BOL>7⏏5$::␤»
TimToady m: lines.elems.say
camelia rakudo-moar 6a1879: OUTPUT«20␤»
GLRelia rakudo-moar 287360: OUTPUT«(timeout)»
TimToady m: lines.list.elems.say
camelia rakudo-moar 6a1879: OUTPUT«20␤»
GLRelia rakudo-moar 287360: OUTPUT«(timeout)» 19:57
19:57 colomon joined
colomon TimToady++ # dunno how I missed that 19:58
19:58 colomon left 19:59 kaare_ left
dha Other question that is probably indicating foolishness or senility on my part... 20:00
perl6 -ne 'say "{++$} $_"' example.txt would output the contents of example.txt with line numbers prepended. 20:01
20:01 lizmat_ left
dha But "++$" is not ringing a bell with me, and I have no idea how I would even begin to find it in the docs. 20:01
20:02 colomon_ left
TimToady a bare $ is short for state $ which is an anonymous state variable 20:03
20:04 colomon joined, colomon left
dha You don't say. :-) That's not actually in the documentation somewhere by any chance, is it? 20:05
(and by documentation I mean doc, not spec)
TimToady m: my (@foo) ::= (-> { 'a', 'b' } ... *).item; say @foo.WHAT
camelia rakudo-moar 6a1879: OUTPUT«Too few positionals passed; expected 1 argument but got 0␤ in block <unit> at /tmp/88dUmKwnDg:1␤␤»
GLRelia rakudo-moar 287360: OUTPUT«Too few positionals passed; expected 1 argument but got 0␤ in block <unit> at /tmp/Bq9kY3gpf6:1␤␤»
TimToady m: my (@foo) ::= \((-> { 'a', 'b' } ... *)); say @foo.WHAT 20:06
camelia rakudo-moar 6a1879: OUTPUT«(List)␤»
GLRelia rakudo-moar 287360: OUTPUT«(List)␤»
dha Huh. 'anonymous state' only appears to be in specs once, and isn't clearly documented. *sad face*
TimToady yeah, we put it in as an experiment, and then started using it without much speccing, alas 20:11
20:11 itz joined 20:13 dha left 20:14 xfix left 20:18 darutoko- left 20:19 colomon joined
tadzik hoelzro: ping 20:19
dalek rl6-roast-data: d60396c | coke++ | / (2 files):
another glr run for 'today'
20:21
ast/glr: f5f1ba7 | (Stefan Seifert)++ | integration/advent2010-day04.t:
Fix regressions from sequence rewrite in advent2010-day04.t
20:22
20:23 FROGGS joined
dalek ast/glr: 0e074f5 | (Stefan Seifert)++ | integration/advent2010-day11.t:
Fix sequence rework regression in advent2010-day11.t

No need to use binding for infinite sequences anymore
20:24
[Coke] Down to 53 failing tests, in 15 files 20:25
20:25 lizmat joined
tadzik I'm afraid to blink 20:25
[Coke] (so I finally caught up with liz's pronouncement from an hour ago or so. :)
dalek ast/glr: 5de4787 | TimToady++ | S03-sequence/misc.t:
previous patch better solved with assignment
ast/glr: 96058c0 | (Stefan Seifert)++ | integration/advent2012-day14.t:
Fix sequence rework regrssion in advent2012-day14.t

No need to use binding for infinite sequences anymore. Just assign them to an array.
20:26
20:27 pmurias joined
lizmat m: int8.Range.say 20:28
camelia rakudo-moar 6a1879: OUTPUT«-128..127␤»
GLRelia rakudo-moar 287360: OUTPUT«-128-127-126-125-124-123-122-121-120-119-118-117-116-115-114-113-112-111-110-109-108-107-106-105-104-103-102-101-100-99-98-97-96-95-94-93-92-91-90-89-88-87-86-85-84-83-82-81-80-79-78-77-76-75-74-73-72-71-70-69-68-67-66-65-64-63-62-61-60-59-58-57-56-55-54-5…»
TimToady huh
lizmat yeah 20:29
dalek ast/glr: aad89b6 | (Stefan Seifert)++ | integration/lazy-bentley-generator.t:
Another sequence rework regression fix
lizmat m: say Int.Range # 1st arg rule in action 20:30
camelia rakudo-moar 6a1879: OUTPUT«-Inf..Inf␤»
20:30 rurban left
lizmat m: say int8.Range # 1st arg rule in action 20:30
camelia rakudo-moar 6a1879: OUTPUT«-128..127␤»
GLRelia rakudo-moar 287360: OUTPUT«(timeout)»
rakudo-moar 287360: OUTPUT«-128-127-126-125-124-123-122-121-120-119-118-117-116-115-114-113-112-111-110-109-108-107-106-105-104-103-102-101-100-99-98-97-96-95-94-93-92-91-90-89-88-87-86-85-84-83-82-81-80-79-78-77-76-75-74-73-72-71-70-69-68-67-66-65-64-63-62-61-60-59-58-57-56-55-54-5…»
[Coke] r: role Foo { method bar {...} }; say Foo.perl 20:32
GLRelia rakudo-jvm 287360: OUTPUT«Can't open perl script "/home/steve/rakudo-inst/bin/eval-client.pl": No such file or directory␤»
..rakudo-moar 287360: OUTPUT«Method 'bar' must be implemented by Foo because it is required by a role␤ in any compose_method_table at src/gen/m-Metamodel.nqp:2666␤ in any apply at src/gen/m-Metamodel.nqp:2676␤ in any compose at src/gen/m-Metamodel.nqp:2812␤ in any make_pun…»
lizmat [Coke]: ^^
camelia rakudo-jvm 6a1879: OUTPUT«Method 'bar' must be implemented by Foo because it is required by a role␤ in any compose_method_table at gen/jvm/Metamodel.nqp:2653␤ in any apply at gen/jvm/Metamodel.nqp:2663␤ in any compose at gen/jvm/Metamodel.nqp:2799␤ in any make_pun at gen…»
..rakudo-moar 6a1879: OUTPUT«Method 'bar' must be implemented by Foo because it is required by a role␤ in any compose_method_table at src/gen/m-Metamodel.nqp:2664␤ in any apply at src/gen/m-Metamodel.nqp:2674␤ in any compose at src/gen/m-Metamodel.nqp:2810␤ in any make_pun…»
[Coke] I just touched a ticket about int ranges..
see RT #124082 20:33
synbot6 Link: rt.perl.org/rt3/Public/Bug/Display...?id=124082
[Coke] m: say Num.Range;
camelia rakudo-moar 6a1879: OUTPUT«-Inf..Inf␤»
GLRelia rakudo-moar 287360: OUTPUT«(timeout)»
lizmat [Coke]: and there it goes of making an infinite list
*off
[Coke] one of the 9 or so [GLR] RTs I found digging today 20:34
lizmat according to the 1st arg rule, it is doing the right thing
m: say $(Int.Range)
camelia rakudo-moar 6a1879: OUTPUT«-Inf..Inf␤»
[Coke] m: class T{}; class S is T{}
camelia rakudo-moar 6a1879: OUTPUT«===SORRY!===␤Cannot find method 'has_compile_time_value'␤»
lizmat que? 20:35
GLRelia rakudo-moar 287360: OUTPUT«(timeout)»
rakudo-moar 287360: OUTPUT«===SORRY!===␤Cannot find method 'has_compile_time_value'␤»
[Coke] (that last one wasn't GLR related)
[ptc] thanks everyone for the workshop and hackathon. That was lots of fun :-) 20:37
nine Still is :) 20:39
colomon It was / is great!
but I need to sleep very soon….
hahainternet so this is neat 20:41
someone posted this on reddit: rosettacode.org/wiki/Lychrel_numbers
and it looks like it can be done with p6's infinite list + sequence op
with initiail, {$^a+flip($^a)} ... *
also with better spelling, one would hope
timotimo achieving a cold shower felt much more rewarding than achieving cold fusion tonight 20:43
[ptc] yeah, it's mega hot...
and opening the windows in the hotel room isn't helping much. Lots of traffic noise...
colomon m: sub Lychrel($n) { $n, -> $a { $a + $a.flip } … -> $a { $a == $a.flip }; } sat Lychrel(4) 20:44
camelia rakudo-moar 6a1879: OUTPUT«5===SORRY!5=== Error while compiling /tmp/6JNkiPAbfC␤Strange text after block (missing semicolon or comma?)␤at /tmp/6JNkiPAbfC:1␤------> 3+ $a.flip } … -> $a { $a == $a.flip }; }7⏏5 sat Lychrel(4)␤ expecting any of:␤ infix…»
GLRelia rakudo-moar 287360: OUTPUT«5===SORRY!5=== Error while compiling /tmp/3tY5DDf4rD␤Strange text after block (missing semicolon or comma?)␤at /tmp/3tY5DDf4rD:1␤------> 3+ $a.flip } … -> $a { $a == $a.flip }; }7⏏5 sat Lychrel(4)␤ expecting any of:␤ infix…»
20:44 orbus joined
nine Current official count: 14 spec test files failing :) 20:45
And one of them is just because the todo markers don't seem to work
hahainternet colomon: interesting, where can i read about using pointy blocks like that?
i guess later in s03 maybe
timotimo nine: point me at the file in question with the todo markers and i'll see what i can do
the fudger has some known limitations
_itz glr singularity approaching?
timotimo we have #?does for example
nine timotimo: t/spec/S06-other/pairs-as-lvalues.rakudo.moar 20:46
colomon _itz: probably not tonight, at least if TimToady++ and nine++ finally go to sleep
hahainternet: …. err, not sure. probably an advent post somewhere?
nine lizmat++ could still fix a couple of tests then :) 20:47
colomon hahainternet: it sounded like you had an idea how the sequence operator works?
20:47 colomon_ joined
hahainternet colomon: yeah, i hadn't thought to terminate it on the desired one 20:47
colomon_ m: sub Lychrel($n) { $n, -> $a { $a + $a.flip } ... -> $a { $a == $a.flip }; } sat Lychrel(4)
camelia rakudo-moar 6a1879: OUTPUT«5===SORRY!5=== Error while compiling /tmp/90dBYAmlB6␤Strange text after block (missing semicolon or comma?)␤at /tmp/90dBYAmlB6:1␤------> 3$a.flip } ... -> $a { $a == $a.flip }; }7⏏5 sat Lychrel(4)␤ expecting any of:␤ infix␤ …»
GLRelia rakudo-moar 287360: OUTPUT«5===SORRY!5=== Error while compiling /tmp/W5L7oVl7ZA␤Strange text after block (missing semicolon or comma?)␤at /tmp/W5L7oVl7ZA:1␤------> 3$a.flip } ... -> $a { $a == $a.flip }; }7⏏5 sat Lychrel(4)␤ expecting any of:␤ infix␤ …»
nine [ptc]: no airco?
[Coke] no airco for me, either.
hahainternet colomon: 'sat' :) 20:48
colomon_ m: sub Lychrel($n) { $n, -> $a { $a + $a.flip } ... -> $a { $a == $a.flip }; } say Lychrel(4)
GLRelia rakudo-moar 287360: OUTPUT«5===SORRY!5=== Error while compiling /tmp/LJn_dXU2Hr␤Strange text after block (missing semicolon or comma?)␤at /tmp/LJn_dXU2Hr:1␤------> 3$a.flip } ... -> $a { $a == $a.flip }; }7⏏5 say Lychrel(4)␤ expecting any of:␤ infix␤ …»
camelia rakudo-moar 6a1879: OUTPUT«5===SORRY!5=== Error while compiling /tmp/BiehXGtQMY␤Strange text after block (missing semicolon or comma?)␤at /tmp/BiehXGtQMY:1␤------> 3$a.flip } ... -> $a { $a == $a.flip }; }7⏏5 say Lychrel(4)␤ expecting any of:␤ infix␤ …»
colomon_ m: sub Lychrel($n) { $n, -> $a { $a + $a.flip } ... -> $a { $a == $a.flip }; }; say Lychrel(4)
camelia rakudo-moar 6a1879: OUTPUT«4␤»
GLRelia rakudo-moar 287360: OUTPUT«4␤»
colomon_ doh
m: sub Lychrel($n) { $n, -> $a { $a + $a.flip } ... -> $a { $a == $a.flip }; }; say Lychrel(104)
camelia rakudo-moar 6a1879: OUTPUT«104 505␤»
GLRelia rakudo-moar 287360: OUTPUT«104505␤»
colomon_ there you go
hahainternet yeah, i'm going to steal your termination and stick on reddit if you don't mind 20:49
as it's a particularly attractive implementation
colomon hahainternet: be my guest
timotimo nine: damn, instead of havign a look at the test file i read more twitter feeds >_> 20:51
dalek kudo/glr: eac8b76 | TimToady++ | src/core/io_operators.pm:
one-arg say should just gist it

A function that distributes .gist over all its arguments should not make an exception when there's only one argument.
20:52
20:53 colomon left
timotimo nine: it's because the extra # confuses the harness, i think 20:53
FROGGS m: my @a = 1, 2, 3; say @a
camelia rakudo-moar 6a1879: OUTPUT«1 2 3␤»
GLRelia rakudo-moar 287360: OUTPUT«123␤»
timotimo no airco in my hotel room, either
what fantastic hotel are you in, nine?
FROGGS timotimo: I've got airco too 20:54
timotimo all i got is a thing that pulls moist air out of the room
[ptc] nine: no, unfortunately no airco :-( 20:55
nine timotimo: I think I already tried that
timotimo: Oltnerhof
20:56 colomon joined
[ptc] hopes for airco in Granada... 20:56
timotimo yeah, the joke was that i'm in the same hotel as you
dalek kudo/glr: b1db8a6 | lizmat++ | src/core/io_operators.pm:
note === say except for the handle
timotimo also, the harness doesn't get confused by the extra # because it doesn't parse the output; instead it counts values, test numbers etc ...
the plot thickens
jdv79 also no airco. wah wah 20:57
timotimo :|
FROGGS m: use nqp; nqp::loadbytecode("foo/bar/baz"); 20:58
camelia rakudo-moar 6a1879: OUTPUT«While looking for 'foo/bar/baz': no such file or directory␤ in block <unit> at /tmp/eMglQlgKBV:1␤␤»
GLRelia rakudo-moar eac8b7: OUTPUT«While looking for 'foo/bar/baz': no such file or directory␤ in block <unit> at /tmp/wzPDkyqiJx:1␤␤»
jdv79 it was hotter in hawaii, thailand, and turkey "recently". i barely went to sleep as i swimming in bed in all 3.
dalek rl6-roast-data: 9ba9de8 | coke++ | / (2 files):
another glr run for 'today'
20:59
FROGGS m: my @a = 1, 2, 3; say @a 21:02
camelia rakudo-moar 6a1879: OUTPUT«1 2 3␤»
GLRelia rakudo-moar b1db8a: OUTPUT«[1 2 3]␤»
FROGGS m: my @a = 1, 2, 3; say @a; say 1, 2, 3
camelia rakudo-moar 6a1879: OUTPUT«1 2 3␤123␤»
GLRelia rakudo-moar b1db8a: OUTPUT«[1 2 3]␤123␤»
FROGGS *g* 21:03
21:03 vendethiel joined
vendethiel waves from home 21:04
nine TimToady: so say and note are the only exceptions to the single arg rule in the whole language?
21:04 woolfy joined
lizmat vendethiel o/ 21:04
vendethiel hi again to all the swiss hackers :)
woolfy I LOVE YOU ALLWITH A GREAT HUGE LOVE!
colomon woolfy!!!!! \o/ 21:05
21:05 virtualsue joined
vendethiel hugs woolfy 21:05
21:06 luiz_lha joined, bin_005_l left
laben m: say \(List) ~~ :(*@a) 21:06
camelia rakudo-moar 6a1879: OUTPUT«True␤»
GLRelia rakudo-moar b1db8a: OUTPUT«Invocant requires an instance of type List, but a type object was passed. Did you forget a .new?␤ in block <unit> at /tmp/89fEOd_QFd:1␤␤»
dalek ast/glr: 6e71b03 | TimToady++ | S03-sequence/basic.t:
more sequence binding -> assignment
21:06 luiz_lha is now known as Guest4404
woolfy colomon? You just went to bed. Go to sleep, man! 21:06
laben m: say \(Hash) ~~ :(*@a)
camelia rakudo-moar 6a1879: OUTPUT«True␤»
GLRelia rakudo-moar b1db8a: OUTPUT«Invocant requires an instance of type Hash, but a type object was passed. Did you forget a .new?␤ in block <unit> at /tmp/eCsdQ1uHPE:1␤␤»
21:06 rindolf left
moritz m: my @a = Hash; 21:07
camelia ( no output )
GLRelia ( no output )
TimToady nine: I dunno if those are the only ones, but those are certainly necessary, due to the distribution of .gist semantics
woolfy hugz vendethiel
laben why does it take the Hash or List type objects as invocants and not as params?
timotimo vendethiel: you made it! 21:08
vendethiel timotimo: somehow!
colomon woolfy: I’m packing!
moritz laben: it's an exception from somwhere in the setting code
21:09 bin_005 joined
dalek ast/glr: ccef992 | lizmat++ | S09-typed-arrays/arrays.t:
Make S09-typed-arrays/arrays.t pass
21:09
ast/glr: f729a5d | lizmat++ | S32-io/note.t:
Adapt S32-io/note.t because of b1db8a6576d23fbc7
ast/glr: aab8bf2 | lizmat++ | S03-sequence/basic.t:
Merge branch 'glr' of github.com/perl6/roast into glr
kudo-star-daily: dde104e | coke++ | log/ (9 files):
today (automated commit)
[Coke] colomon: verified. no ac.
nine [Coke]: sorry to read that 21:10
[Coke] ugh, do I pack now or in the morning?
laben moritz: i get it from src/Perl6/Metamodel/BOOTSTRAP.nqp line 325. the question is, why does it think it is an invocant and not a normal param?
21:10 JasonBaum joined
[Coke] ... now. else I'll be stealing a pillowcase and throwing it all in. 21:10
JasonBaum p6: say 5 21:11
camelia rakudo-moar 6a1879: OUTPUT«5␤»
GLRelia rakudo-moar b1db8a: OUTPUT«5␤»
dalek ast/glr: bee5dcc | TimToady++ | S16-io/bare-say.t:
say () should say ()
JasonBaum hello
FROGGS hi JasonBaum 21:12
JasonBaum quick question
I am using Ubuntu 14.04 and would like to try out Perl6. Can an install live peacefully alongside Perl 5.x?
dalek ast/glr: 0ad1a67 | TimToady++ | S16-io/say-and-ref.t:
say [1,2,3] makes [1 2 3]
ast/glr: 847b9aa | lizmat++ | S16-io/bare-say.t:
Adapt S16-io/bare-say.t because of b1db8a6576d23fbc7
21:13
ast/glr: 36abcb0 | lizmat++ | S16-io/say-and-ref.t:
Merge branch 'glr' of github.com/perl6/roast into glr
hoelzro JasonBaum: yes, they can!
yoleaux 19:10Z <[Coke]> hoelzro: can you review RT #123154 ?
synbot6 Link: rt.perl.org/rt3/Public/Bug/Display...?id=123154
FROGGS JasonBaum: yes, because it won't install a perl binary, and it wont install in the system paths 21:14
hoelzro (also, greetings)
woolfy Coke: pack now, sleep with a ease-of-mind, wake up, have a quiet breakfast, and just grab yout stuff and go. Now you are still awake.
laben JasonBaum: ofc it can, why shouldnt it be so? btw i recommend trying github.com/tadzik/rakudobrew for an easy to use installation
hoelzro [Coke]: on it
JasonBaum great thank you
hoelzro tadzik: are you around? you can haz convertors
FROGGS JasonBaum: just follow these instructions rakudo.org/how-to-get-rakudo/
JasonBaum FROGGS: Thanks for the link!
dalek ast/glr: 8613bf8 | TimToady++ | S03-sequence/nonnumeric.t:
list gist now parenthesizes
21:15
[ptc] hoelzro: are there plans for brekkie tomorrow morning?
hoelzro [ptc]++ # leaks
moritz m: class A { method foo(A:D:) { } }; A.foo
camelia rakudo-moar 6a1879: OUTPUT«Invocant requires an instance of type A, but a type object was passed. Did you forget a .new?␤ in method foo at /tmp/_tlL1druew:1␤ in block <unit> at /tmp/_tlL1druew:1␤␤»
GLRelia rakudo-moar b1db8a: OUTPUT«Invocant requires an instance of type A, but a type object was passed. Did you forget a .new?␤ in method foo at /tmp/HadfVYtgbs:1␤ in block <unit> at /tmp/HadfVYtgbs:1␤␤»
hoelzro er, leaks-- =)
[ptc]: yes, I believe so
[ptc] hoelzro: heh
moritz laben: either it's really a mis-reported error, or the binder tries to call a method on it, which produces this error 21:16
hoelzro everyone @ olterhof: who wants to meet up for breakfast, and when?
colomon oh, hey folks, while it still needs more love, I’ve got the module smoke tester in much better shape now. When GLR is ready to switch to master (or whatever), can we have a day or two to test all the modules and see what needs to be changed? 21:17
FROGGS hoelzro: I like to, and I have a lot of time 21:18
lizmat colomon: I think that can be aranged
hoelzro [Coke]: still Mus =(
FROGGS: great!
colomon \o/ 21:19
hoelzro colomon++
21:19 sufrostico joined
[ptc] colomon: sounds good 21:20
JasonBaum I don't know if any of you will find this helpful, but I've created a search function for reddit comments (and submissions). So you can quickly see comments where Perl6 was mentioned by making a call like this: api.pushshift.io/reddit/search?q=perl6 21:21
woolfy hoelzro: lizmat and me will have breakfast at 8:30
JasonBaum (you can use a limit parameter to get more comments if needed)
timotimo ah, that spits out json
JasonBaum Yes sorry. One second.
laben moritz: i really dont grep that code, afraid i cant say much, just that it checks "$flags +& $SIG_ELEM_INVOCANT" and that somehow is true, which doesnt make sense to me. im not able to go deeper
moritz m: say :(*@a).params[0].invocant 21:22
camelia rakudo-moar 6a1879: OUTPUT«False␤»
GLRelia rakudo-moar b1db8a: OUTPUT«False␤»
laben yea, but that exception is only in that codepath 21:23
JasonBaum Here is a jsfiddle for it: jsfiddle.net/n8tjvsex/
it will show comments and also submissions where perl6 is mentioned heavily
laben i guess i will ask jnthn if i see him
JasonBaum might be interesting for some of you to play with
it's very beta but usually works well 21:24
21:25 ab6tract joined
timotimo you haev "sex" in your url *giggle* 21:25
JasonBaum lol
timotimo that's the name of the dog of the aunt(?) that mr scalzi was talking abou
ab6tract horrible flashbacks of a lightning talk
FROGGS JasonBaum++ 21:26
JasonBaum but in any event, it may give you guys something to look at
timotimo what are all the "undefined"s for? :)
under the name of the subreddit 21:27
JasonBaum that should be the title of the submission
hahainternet so with the Lycherel numbers being referred to before, the rosetta code page has a requirement to limit it to 500 elements, i understand perl6 has lazy lists and the first thing that comes to mind is simply to test if @a[501] is not Nil to check if the list generates more than 500 elems
timotimo ah, and there seems to not be a link to the submission at all yet :)
hahainternet is there a nicer way, or something more explicit or idiomatic etc
JasonBaum i made some changes to the back-end but will fix that shortly
dalek ast/glr: e879e6c | lizmat++ | S32-exceptions/misc.t:
Fix some slightly changed messaging
laben .tell jnthn can you check why "say \(Hash) ~~ :(*@a)" and "say \(List) ~~ :(*@a)" die with exception "Invocant requires an instance of type Hash|List, but a type object was passed." in the binder? 21:28
yoleaux laben: I'll pass your message to jnthn.
laben have a good rest, #perl6
moritz laben: the binder calls method Capture on some value, if it's present; and Hash has a method Capture that requires a defined invocant
timotimo ah, oh. 21:29
[ptc] timotimo: I wonder what n8tjv-sex is though... And jsfiddle is also a bit dodgy
[ptc] wonders if it's the beer talking...
laben moritz: i'm sorry to say i understand barely half of what you're saying. i guess there's no hurry... 21:30
dalek ast/glr: 8403fcd | TimToady++ | S04-statements/gather.t:
testing neighbor identity calls for unique cells
laben goes to bed
moritz laben: what I'm saying is that the binder itself trips up, instead of intentionally generating the error message you see
timotimo [ptc]: well, n8 in german is for "nacht" aka "night" and tjv is the french high speed train 21:31
21:31 laben left
dalek ast/glr: 620317d | lizmat++ | S19-command-line/dash-e.t:
Adapt to changed gist output for @*ARGS
21:31
JasonBaum are you analyzing the jsfiddle link name? 21:32
night train sex?
timotimo totally
JasonBaum I need to party more with you fine people
timotimo the mood at the SPW2015 over the last few days was pretty great
even though it was rather a bit warmer than we'd all have liked
ab6tract .tell skids thanks for looking into it! I really appreciate it :) 21:33
yoleaux ab6tract: I'll pass your message to skids.
hoelzro [Coke]: oops, I misspoke, updating now
JasonBaum one more thing before I head off -- if you like huge data sets, I've made every available public reddit comment available. Maybe you can find something useful. Location is here - pan.whatbox.ca:36975/reddit/comments/monthly/ 21:34
it's around 1.75 billion comments in JSON
ab6tract JasonBaum: that sounds like a really fun data source :) 21:35
timotimo oh wow
JasonBaum yeah definitely. I tried to load it all into Sphinx Search but I just don't have enough RAM for it
21:35 cognominal joined
timotimo i'll feed that into json::fast and see how fast it exhausts my ram + swap 21:36
ab6tract sounds like a useful candidate for some streaming pipelines
JasonBaum I don't know what the latest and greatest is for Perl6 for JSON parsing, but I've been using Cpanel::JSON::XS for Perl 5
timotimo yup, we don't have a streaming json parser yet, right?
hoelzro tadzik: pong
ugexe is it possible that, with a known json structure, it could be chunked out into pieces that do need to be parsed vs pieces that dont? 21:37
ab6tract timotimo: not that I'm aware of
JasonBaum I also have a real-time SSE stream if you want to play with it. Details are here: www.reddit.com/r/redditdev/comment...rts_a_ton/
dalek ast/glr: 7995d66 | TimToady++ | S03-sequence/basic.t:
containers test of [] now needs $[]
ab6tract enjoyed reading that post from tony-o (ugexe?) ugexe.com/speed-quest-1-0-or-0-1-or-something/
timotimo we have a json parser that's a nativecall'd thingie 21:38
ugexe for instance, to shim in the default json parser to work faster for parsing the manifest
JasonBaum essentially, if you want to see Reddit comments and submissions in real time, do this "wget -qO- stream.pushshift.io"
that will connect you to a real-time stream
(details on usage are in that link)
21:38 geekosaur left
JasonBaum I also provide link info in that stream for all links posted to Reddit 21:39
21:39 geekosaur joined
JasonBaum (a head request and a call to Google API if it is relevant -- like a youtube video link) 21:39
woolfy SPW is "a little warmer" as in that I am soaking with sweat every day... 21:40
JasonBaum one more question -- are any of you going to the Amazon Re:Invent in Vegas in Oct?
I was planning to head out there. If any of you do end up going, please let me know and we can get a beer or something. If you are interested in Big Data. 21:41
(or if you like beer) 21:42
21:43 virtualsue left
timotimo perl6 is not yet ready for "big data", with the overhead we sometimes have :| 21:43
jdv79 ugexe: maybe gen a line oriented json file with a fastly searchable prefix?
21:43 itz left
JasonBaum does Perl6 support regex on a 920 gigabyte string? 21:44
jdv79 where is this massive json file, btw? 21:45
ugexe jdv79: i thought about that too. not enough to consider the pitfalls
JasonBaum the files are ordered by month
21:45 llfourn joined
timotimo split by month, even 21:45
JasonBaum there is a big file on torrent
jdv79 ugexe: i was actually thinking about testing that out the other day but got distracted
hoelzro there's no way to overwrite $?FILE, is there? 21:46
ugexe i think it could mostly work until %*RESOURCES is implemented, which allows custom pairs (so the structure of these are not known)
JasonBaum here is the posting for the big dataset
www.reddit.com/r/datasets/comments...t_comment/
timotimo smls: that can't be right: i.imgur.com/fTEhZAF.png 21:47
JasonBaum Amazon was nice enough to give a $500 grant to host it
but bandwidth fees ate that up quick
smls timotimo: What browser is that?
ab6tract JasonBaum: "I also provide link info in that stream for all links posted to Reddit" <-- do you mean said link is n the reddit comments in the post about SSE ?
timotimo gurgel chrome
JasonBaum the link was either posted in a comment or the link was a submission 21:48
yes
ab6tract smls: looks the same in safari
smls Looks like that in my Chrome too :( 21:49
ab6tract JasonBaum: just checking because "that stream" was slightly ambiguous to me
smls in Firefox, it renders right
smls will investigate
It's an SVG image, in case you're wondering.
JasonBaum if you go to reddit.com/r/test and post a message with a link (like a youtube video) and have another window open with "wget -qO- 'stream.pushshift.io/?event=link&...THOR_NAME' you will see it pop up in 2-5 seconds
21:50 llfourn left
timotimo the other svg is not problematic 21:50
but yeah, i've opened the svg separately already
ab6tract smls: also, nice blog!
smls thanks :)
timotimo: you can also replace the .svg in the URL with .png to get prerendered ones
JasonBaum ab6tract, when you connect to the stream, you can filter by a lot of methods -- one of which is event=(t1,t3,link) ... t1 are reddit comments, t3 are submissions and link just shows links 21:51
I'm assuming Perl6 allows for inline C like Perl 5? 21:53
nine JasonBaum: doc.perl6.org/language/nativecall
JasonBaum thanks
ab6tract JasonBaum: ahhh.. okay, it slowly dawns on me. I parsed your original comment as "somewhere in that reddit thread I post a link to an archive of all the links on reddit" 21:54
JasonBaum yeah sorry that wasn't clear
FROGGS JasonBaum: there is also a module Inline::C but NativeCall is the better way of doing things 21:55
ab6tract now that I've interacted with your streaming thingy, it's a lot clearer
dalek ast/glr: 61c6af6 | lizmat++ | S17-supply/classify.t:
Make S17-supply/classify.t pass
lizmat that should get us at 10 now
[ptc] 10?? that's awesome! Well done!
ab6tract lizmat!!!!!!!! 21:56
i'm too sweaty to hug you right now :D
vendethiel YAY
JasonBaum For Perl 5, to interact with the stream I use something like $ua->add_handler( response_data => sub { ... } ) ... I'm not sure what the equivalent would be for Perl6
dalek ast/glr: 6d66f99 | TimToady++ | S06-signature/unpack-array.t:
Seq interpolation needs .list
21:57
TimToady 9
JasonBaum There may even be an SSE module already for Perl6 but I haven't checked
TimToady --> bed 21:59
ab6tract JasonBaum: HTTP::UserAgent has been recently updated.
nine Good night! 22:00
ab6tract o/ TimToady :D
JasonBaum: I'm getting old.. When I see SEE, I think of Pentium II
*SSE
22:00 bin_005 left
JasonBaum I remember that! 22:00
dalek ast/glr: 0d71222 | lizmat++ | S17-supply/categorize.t:
Make S17-supply/categorize.t pass
22:01
FROGGS sleep well TimToady
JasonBaum The stream could have just been a series of JSON objects like Twitter's stream, but you can just as easily interact with an SSE stream and pull out the data fields
ab6tract proposed backronym for GLR -- "God Loves Refactoring"
lizmat TimToady: nighty night! 22:02
ab6tract if * didn't, * wouldn't have made it so easy on us!
JasonBaum the benefit with SSE is the ability to specify a beginning id number when you connect to the stream in case the script loses its connection
ab6tract immediately knocks on wood
JasonBaum most browsers will do this natively with a header string LAST-EVENT-ID
Eventually I can load the entire dataset into sphinx search but I would need around 256 GB of RAM to make it worthwhile 22:04
FROGGS :P 22:05
JasonBaum I was going to get a new workstation with a Skylake processor but Intel is busy doing more paper launches
Skylake at least allows for up to 64 GB if using DDR4 22:06
timotimo i don't know what SSE means in this context :|
JasonBaum (Server Side Events)
timotimo ah
JasonBaum it's basically one way websockets
timotimo that's the thing that separates chunks with two newlines, right?
JasonBaum yep!
timotimo and you get EventSource in JavaScript to handle them?
JasonBaum yes 22:07
timotimo i've worked with them, they're neat
JasonBaum you can create multiple EventSource objects to deal with each type of event (t1, t3, etc.)
smls timotimo: Should be fixed now. 22:08
timotimo and the browser will only have a single connection?
JasonBaum yes. It makes one connection
dalek rl6-roast-data: d4e0659 | coke++ | / (2 files):
another glr run for 'today'
lizmat is down to 7 fails
JasonBaum but if you want to conserve bandwidth, the event parameter is there to filter it down
smls timotimo: Apparently, Chrome really doesn't like it when multiple font sizes are used in the same SVG text element... 22:09
timotimo right
oh, ugh
JasonBaum you could also use match=perl and only get comments and submissions with perl mentioned in them
Eventually my goal would be to create a sample Perl6 script to interact with the stream to show new programmers how to work with big data using Perl6 22:10
reddit comment volume is usually always under 50 comments a second which is easily handled by Perl 22:11
jdv79 .tell [Coke] email
yoleaux jdv79: I'll pass your message to [Coke].
[Coke] realizes he's going to have trouble getting a can of soda home. Maybe zurich is a smidge more lenient. 22:12
yoleaux 22:11Z <jdv79> [Coke]: email
22:13 cognominal left
timotimo realizes Coke hes Coke going Coke to Coke have Coke trouble Coke getting Coke a Coke can Coke of Coke soda Coke home. Coke Maybe Coke zurich Coke is Coke a Coke smidge Coke more Coke lenient. 22:13
oetiker coke they are not ... 22:14
for drinks anyway ...
JasonBaum brb
oetiker for swiss army knifes they are
22:14 JasonBaum left
[ptc] but if it's in your main luggage that should be ok 22:15
22:15 jbaumgartner joined
jbaumgartner back 22:15
(had to switch to my main account)
22:15 lolisa joined
[Coke] I have no main luggage. 22:15
I guess I could check my backpack and use the booking.com bags for my plane needs. 22:16
I hesitate to do that, but May have to.
timotimo sub MAIN('luggage') { die "no." } 22:17
jdv79 at least its not an extra charge like us domestic
i think i paid 35 to check for yapc na:(
ab6tract jdv79: the weird flipside of the checked-baggage fees is that enormous bags (beyond the apparently toothless limits they declare) are now put into overhead, taking up all the space 22:24
jbaumgartner Compress::Zlib::Luggage 22:25
hoelzro Olterhof folks who are around tomorrow: we're meeting for breakfast at 8:30 22:26
timotimo we'll be kicked out of our rooms at 10am 22:27
[ptc] hoelzro: ok, good to know 22:30
[Coke] I'll leave the novelty soda (hello kitty) downstairs in the morning. If you see it, feel free to grab it. 22:31
[ptc] [Coke]: maybe we can send it to you in the post? 22:32
22:32 ggoebel joined
oetiker pass it on to liz and wendy 22:32
[Coke] I can't imagine it'll be cheap to do that. No worries.
they can give it away as a prize at the next thing or just drink it. :)
22:32 _itz_ joined 22:34 _itz left 22:36 jbaumgartner left 22:38 pmurias left, bin_005 joined 22:40 _itz joined
woolfy I refuse to drink hello kitty novelty soda 22:40
22:41 _itz_ left
woolfy Is it very expensive? Related to "sending my mail"... 22:41
lizmat goes to get some sleep and dream of lucky 7 22:43
22:43 lizmat left 22:45 llfourn joined
woolfy nightnight... 22:46
22:46 woolfy left 22:47 _itz left 22:48 _itz joined
[ptc] 'night * 22:49
ab6tract pleasant dreams my #perl6 heroes
22:49 ab6tract left 22:50 llfourn left
hoelzro night everyone 22:55
22:56 Sgeo_ left 23:03 _itz_ joined 23:05 _itz left 23:17 mprelude left 23:18 yqt joined 23:23 yqt left
[Coke] not expensive at all. arg, guess i will just stay awake until my flight 23:30
dalek rl6-roast-data: 023e9af | coke++ | / (2 files):
another glr run for 'today'
23:31
23:38 telex left 23:40 spider-mario left, telex joined 23:46 khw joined 23:50 RabidGravy left 23:58 [TuxCM] joined