MasterDuke re RT #127395 and RT #123078, should the exception say the privacy of the suggested method if it's opposite the given one? e.g., Int!sun would suggest 'public method sin', and Int.sun might suggest 'private method sub' 00:45
or maybe just put a '!' before any private methods regardless? 00:46
Geth rakudo: MasterDuke17++ created pull request #1091:
Suggest misspelled method names
01:47
roast: c4dc7ba410 | usev6++ | 5 files
[JVM] Fudge newly failing tests
06:27
[Tux] This is Rakudo version 2017.05-297-g608e886a9 built on MoarVM version 2017.05-25-g62bc54e9 06:49
csv-ip5xs 2.741
test 13.498
test-t 4.438 - 4.607
csv-parser 13.494
Error encoding UTF-8 string: could not encode Unicode Surrogate codepoint 55600 (0xD930)
should I tell those 88.t fails on a daily basis now that jnthn is/was working on it?
jnthn [Tux]: Is it a regression? 08:49
|Tux| nope 08:50
that is why 88.t has never been in the default test suite
github.com/Tux/CSV/blob/master/88.t ā† run till fail 08:51
once that passes always, I can finally try to support real-binary data in CSV, like jpg/png images and serialized objects 08:52
Zoffix m: use nqp; my $x = nqp::null_s; dd nqp::ifnull($x, 42) 10:38
camelia chars requires a concrete string, but got null
in block <unit> at <tmp> line 1
Zoffix Could probably get a bit of boost by adding nqp::ifnull_s() op and getting rid of these variable assignments: github.com/rakudo/rakudo/blob/nom/...pm#L22-L34 10:39
switching those to nqp::ifnull(blah, Str) shows 16%-35% speed boost; tho unsure how accurate that is, given above it gives wrong result.
Benched with this concoction: gist.github.com/zoffixznet/2e7bfbc...fc844ffc47 10:40
Geth rakudo/js: 2e0d8b362d | pmurias++ | 3 files
[js] Add a 'make js-spectest' which run a bunch of passing tests
jnthn Yes it's totally wrong
You're boxing a null string 10:41
And then null-testing the box 10:48
Zoffix ah
jnthn I'm not sure why the assignment would be so costly though
Will have to look at what code it's generating
There ain't even a Scalar container being created/used 10:49
And it shouldn't need a nativeref either
Geth nqp/nqp-io: 5 commits pushed by (Jonathan Worthington)++ 12:22
jnthn Just a rebase
Geth nqp/nqp-io: 305c8fed7d | (Jonathan Worthington)++ | src/core/IO.nqp
Add a way to get NQP wrapped standard handles.
12:33
nqp/nqp-io: dff957e5dd | (Jonathan Worthington)++ | src/HLL/Compiler.nqp
Switch HLL compiler code reading to high level I/O
12:36
Zoffix NeuralAnomaly: status 12:38
NeuralAnomaly Zoffix, [āœ˜] Next release will be in 2 weeks and 4 days. Since last release, there are 27 new still-open tickets (24 unreviewed and 0 blockers) and 295 unreviewed commits. See perl6.fail/release/stats for details
Zoffix starts working on bot-ified ecosystem smoker to watch for regressions
... eventually... it was a mistake to run update-perl6 on this network :| cloning at 23KiB/s 12:48
jnthn Zoffix++
Geth nqp/nqp-io: aa9a813a02 | (Jonathan Worthington)++ | src/core/IO.nqp
Add further methods to NQP IO handle.
12:53
nqp/nqp-io: 1c4350e5dd | (Jonathan Worthington)++ | t/docs/opcodes.t
Switch opcodes test to high level I/O.
nqp/nqp-io: 72e6b4d59a | (Jonathan Worthington)++ | t/docs/tests.t
Update t/docs/tests.t to use high level I/O.
nqp/nqp-io: 6261004f93 | (Jonathan Worthington)++ | t/nqp/019-file-ops.t
Start to rework I/O tests to use handle.

Since readallfh, readlinefh, and friends are going away.
13:16
nqp: 1c6dac67ed | pmurias++ | 2 files
[js] Make nqp::mkdir create the missing directories in the path it gets
13:26
nqp: fda389d156 | pmurias++ | t/nqp/019-file-ops.t
Test nqp::mkdir with a nested path of missing directories
nqp/nqp-io: c0ec9e725f | (Jonathan Worthington)++ | src/core/IO.nqp
Make print/say/note use high level handle.
13:30
nqp/nqp-io: ea0b05884e | (Jonathan Worthington)++ | 3 files
Eliminate final uses of nqp::sayfh.

Getting cleaner code in the process.
nqp/nqp-io: 71982bcac1 | (Jonathan Worthington)++ | src/core/IO.nqp
Implement seek in IO handle class.
13:42
nqp/nqp-io: 46f5261b43 | (Jonathan Worthington)++ | t/nqp/019-file-ops.t
Switch seek/tell tests to use new IO handle.
Zoffix more inconsistency spotting: IO::Path.new takes CWD arg :CWD, Proc.spawn takes it as :cwd, and IO::Path.relative takes it as a positional :) 13:43
Geth rakudo/io-adaptation: 31c433edc4 | (Jonathan Worthington)++ | tools/build/gen-cat.nqp
Update gen-cat to use NQP high level I/O handles.
13:49
rakudo/io-adaptation: c9347d483d | (Jonathan Worthington)++ | 4 files
Eliminate sayfh/printfh use in the compiler.
travis-ci NQP build failed. pmurias 'Test nqp::mkdir with a nested path of missing directories' 13:52
travis-ci.org/perl6/nqp/builds/237516616 github.com/perl6/nqp/compare/08bfc...a389d1568e
Geth nqp/nqp-io: 3a591e5d16 | (Jonathan Worthington)++ | src/core/IO.nqp
Implement readchars in NQP IO handle.
nqp/nqp-io: 78ca850114 | (Jonathan Worthington)++ | t/nqp/019-file-ops.t
Update tests that used nqp::readcharsfh.
Zoffix ZOFVM6: Files=1252, Tests=136835, 211 wallclock secs 14:04
ZOFVM: Files=1252, Tests=136835, 118 wallclock secs (22.28 usr 3.00 sys + 2373.18 cusr 140.49 csys = 2538.95 CPU) 14:09
m: say 211/118
camelia 1.788136
Geth nqp/nqp-io: 919ba201a6 | (Jonathan Worthington)++ | src/core/IO.nqp
Allow setting line seps on IO handle.
14:14
nqp/nqp-io: 06d1a6ea69 | (Jonathan Worthington)++ | t/nqp/019-file-ops.t
Further re-working of file ops tests.

Also unskip some that are passing on MoarVM and should also pass on JVM too, given the multi-char sep logic is covered in decoder tests.
Zoffix update finally done :) took almost two hours. 14:21
Geth nqp/nqp-io: 9d6643c693 | (Jonathan Worthington)++ | t/nqp/019-file-ops.t
Fix outdated comment.
14:22
nqp/nqp-io: 913012ca81 | (Jonathan Worthington)++ | t/moar/03-line-seps.t
Re-work line seps tests to use high-level I/O.
jnthn Zoffix: Wow...
What a connection...
Zoffix :)
Geth rakudo/io-adaptation: c0b2b020d5 | (Jonathan Worthington)++ | src/core/Proc.pm
Eliminate nqp::readallfh in Proc.pm.

It should use the IO::Pipe handle instead.
14:25
nqp/nqp-io: 83abc6ffde | (Jonathan Worthington)++ | docs/ops.markdown
Undocument char-level I/O ops, which will go away.
14:28
rakudo/io-adaptation: 81e3f641d3 | (Jonathan Worthington)++ | src/core/CompUnit/Repository/FileSystem.pm
Use slurp rather than re-implementing it.

Especially now re-implementing it has become a good bit trickier.
14:29
rakudo/nom: ae891f9367 | (Zoffix Znet)++ | 4 files
Remove TAP.pm6 from core

For easier maintenance, the module will live solely in the ecosystem[^1] from now on.
Since HARNESS_TYPE=6 make spectest/stresstest require this module, t/harness6 script now clones the module from the repo[^1], or ... (5 more lines)
14:30
Zoffix .tell ugexe TAP.pm6 is gone from core now: github.com/rakudo/rakudo/commit/ae891f9367 14:31
yoleaux Zoffix: I'll pass your message to ugexe.
Geth nqp/nqp-io: 1d6ab48abd | (Jonathan Worthington)++ | src/core/IO.nqp
Add flush method to IO handle.
14:43
nqp/nqp-io: 3ef595d8d1 | (Jonathan Worthington)++ | t/nqp/019-file-ops.t
Update outdated comment.
nqp/nqp-io: f7b19f81bb | (Jonathan Worthington)++ | src/vm/jvm/HLL/Backend.nqp
Use high-level I/O in JVM HLL::Backend.
Zoffix huggable: swear jar 14:47
huggable Zoffix, Help Perl 6. Donate (specify "Perl 6 fund" in purpose field): donate.perlfoundation.org/
Zoffix huggable: swear jar :is: Help Perl 6. Donate: donate.perlfoundation.org/
huggable Zoffix, Added swear jar as Help Perl 6. Donate: donate.perlfoundation.org/
Zoffix hehe, there are encoding issues on donate.perlfoundation.org/ 14:48
"attempting to claim Ć¢ā‚¬Å“PerlĆ¢ā‚¬ as their own."
[Coke] Zoffix: Reported on the tpf issue site. 14:49
(just now.) thanks.
Zoffix cool
Geth nqp/nqp-io: 93aba4941b | (Jonathan Worthington)++ | src/core/IO.nqp
Implement set-encoding in NQP file handle.
14:58
nqp/nqp-io: fcbda508e8 | (Jonathan Worthington)++ | src/core/IO.nqp
There is no die in NQP; it's nqp::die.
[Coke] jnthn: github.com/perl6/nqp/blob/nqp-io/s...IO.nqp#L75 15:01
Geth nqp/nqp-io: 6ed77ec37a | (Jonathan Worthington)++ | src/core/IO.nqp
One more die -> nqp::die; [Coke]++.
15:10
jnthn Finally getting there clearing up the string I/O op use 15:11
I hadn't really considered just how much there'd be to do in NQP 15:12
Geth nqp/nqp-io: f6b79387ab | (Jonathan Worthington)++ | src/HLL/Compiler.nqp
Use new IO handle in HLL::Compiler.
15:13
rakudo/io-adaptation: a6cd73af77 | (Jonathan Worthington)++ | src/Perl6/Compiler.nqp
Update Perl6::Compiler for NQP I/O changes.
nqp/nqp-io: 5d3e388264 | (Jonathan Worthington)++ | 6 files
Eliminate remaining nqp::printfh uses.
15:32
rakudo/io-adaptation: 837c73752b | (Jonathan Worthington)++ | src/core/Exception.pm
Eliminate final nqp::printfh use in Rakudo.
15:44
jnthn This all still passes NQP test and spec test :) 15:49
Zoffix :} 15:50
[Coke] ooh, I wonder if this last one will eliminate any of the exceptions-going-to-wrong-stderr issue. 15:52
jnthn [Coke]: Yes, I think it might :) 15:54
[Coke] sweet!
Geth rakudo: tbrowder++ created pull request #1092:
add fix for RT #131400
15:57
nqp/nqp-io: 71e653a1f9 | (Jonathan Worthington)++ | t/nqp/086-pipes.t
Update final test using char-level I/O.
15:58
nqp/nqp-io: 56178e2f8e | (Jonathan Worthington)++ | src/vm/js/HLL/Backend.nqp
Eliminate final use of a char-level I/O op.

Except nqp::say and nqp::print, which we'll leave be as they can be very handy for debugging purposes.
16:00
nqp/nqp-io: 0adb0cf78c | (Jonathan Worthington)++ | src/vm/moar/QAST/QASTOperationsMAST.nqp
Un-map the char-level I/O ops for MoarVM.
16:02
nqp/nqp-io: e8ef675e5b | (Jonathan Worthington)++ | 11 files
Update the MoarVM stage0.

So that we can remove the char-level I/O from MoarVM.
16:20
Zoffix *sigh* 16:42
ZofBot: ignore travis-ci
ZofBot Zoffix, because you can do TWEAK now
Zoffix Doesn't look like our docs build suitable for travis anymore. Like half the builds fail due to job taking too longs 16:43
[Coke] doc build so slow. :| 16:44
Zoffix Situation can probably ameliorated by making travis install rakudo from debs instead of building it from scratch 16:45
jnthn Alright, enough for me today. Time for rest and dinner :) 16:51
dogbert17 m: my @letters = <a b c d e f>; dd @letters[3 .. 7]; # should this be a truncating slice? 16:56
camelia ("d", "e", "f", Any, Any)
perlpilot dogbert17: why? 16:57
Zoffix dogbert17: no, it does that only when the slice is lazy
m: my @letters = <a b c d e f>; dd @letters[lazy 3 .. 7];
camelia ("d", "e", "f")
dogbert17 " However if the outer object passed to (one dimension of) of a positional subscript is a Range, it will be automatically truncated to the actual size of the collection:" 16:58
docs.perl6.org/language/subscripts
Zoffix lies 16:59
dogbert17 but no statistics :)
Zoffix The person writing it probably just tested with 1..* which is lazy and hence truncates
dogbert17 the example I tried above is the example used in the docs 17:00
commit: 2016.01 my @letters = <a b c d e f>; dd @letters[3 .. 7]; 17:01
committable6 dogbert17, Ā¦2016.01: Ā«("d", "e", "f", Any, Any)Ā»
dogbert17 commit: 2015.06 my @letters = <a b c d e f>; dd @letters[3 .. 7];
committable6 dogbert17, Ā¦2015.06: Ā«("d", "e", "f")Ā»
dogbert17 commit: 2015.09 my @letters = <a b c d e f>; dd @letters[3 .. 7]; 17:02
committable6 dogbert17, Ā¦2015.09: Ā«("d", "e", "f")Ā»
dogbert17 commit: 2015.11 my @letters = <a b c d e f>; dd @letters[3 .. 7];
committable6 dogbert17, Ā¦2015.11: Ā«List $var = $("d", "e", "f")Ā»
dogbert17 commit: 2015.12 my @letters = <a b c d e f>; dd @letters[3 .. 7];
committable6 dogbert17, Ā¦2015.12: Ā«("d", "e", "f", Any, Any)Ā»
dogbert17 hmm, seems as if some changes will have to be made in that document then 17:03
perlpilot dogbert17: looks like that statement about Range needs to be modified to talk about laziness instead.
(and the example updated, of course) 17:04
dogbert17 agreed
I can try to whip something together unless someone else wants to do it 17:05
hmm, there is some text about laziness farther down: github.com/perl6/doc/blob/master/d....pod6#L261 17:09
Zoffix hm 17:10
t/get-w3-redirect.t ............. ok
Malformed UTF-8 at line 8 col 18494
in block at /home/zoffix/.zef/store/perl6-lwp-simple.git/243de4fdcc38e1d3049d39aeea7320e19cc8125b/lib/LWP/Simple.pm (LWP::Simple) line 126
False alarm. Installs fine on my other box 17:12
dogbert17 m: my @letters = <a b c d e f>; dd @letters[3 .. 7]:v; 17:15
camelia ("d", "e", "f")
Zoffix That's normal 17:17
m: dd { :42a }<foo bar ber> 17:18
camelia (Any, Any, Any)
Zoffix m: dd { :42a }<foo bar ber>:v
camelia ()
dogbert17 interesting 17:19
interesting 17:21
dogbert17 wrong window
Zoffix m: my $h; $h<booo>.elems.say
camelia 1
Zoffix m: my $h; $h<booo>:v.elems.say
camelia 0
Zoffix ^ handy when you don't know whether the key is inited or not, when looking up elems 17:22
dogbert17 nifty trick 17:25
Zoffix How dissapointing. I wanted to go fix doc Issue 1337, but it's a PR... and already merged :( 17:40
AlexDaniel that's pretty cool 17:41
I wist it was more common
tbrowder Zoffix: check PR 1092 now, please, when you get a chance 17:51
Geth rakudo/nom: 807d30c290 | (Tom Browder)++ | lib/Pod/To/Text.pm6
add fix for RT #131400
18:00
rakudo/nom: 654bb30148 | (Tom Browder)++ | lib/Pod/To/Text.pm6
changes:

  # original code on line 23 was unwrapped for debugging:
   when Positional { $pod.map({pod2text($_)}).join("\n\n") }
   # fix based on irc #perl6, 2017-05-30, word descrip from
   # kurahaupo at 23:43 and on 2017-05-30 MasterDuke's hint
   # at 00:24
... (11 more lines)
rakudo/nom: 3721e68d24 | (Zoffix Znet)++ (committed using GitHub Web editor) | lib/Pod/To/Text.pm6
Merge pull request #1092 from tbrowder/pod2text-fix

add fix for RT #131400
Zoffix buggable: draw 18:08
buggable Zoffix, The next Accidental /win Lottery draw will happen in 1 week, 2 days, 5 hours, 51 minutes, and 5 seconds. Currently have 14 ballots submitted by 10 users!
Zoffix huh 1 week? I thought it did it on the first of the month 18:09
forgot a `:1day` in .clone :) 18:12
buggable: draw
buggable Zoffix, The next Accidental /win Lottery draw will happen in 1 day, 5 hours, 47 minutes, and 17 seconds. Currently have 14 ballots submitted by 10 users!
Zoffix yey
ZOFVM: Files=1239, Tests=136749, 118 wallclock secs (23.39 usr 3.30 sys + 2415.09 cusr 125.67 csys = 2567.45 CPU) 18:21
Geth rakudo/nom: 22ef682537 | (Zoffix Znet)++ | lib/Pod/To/Text.pm6
Refactor code to be shorter
18:22
Zoffix tbrowder: seems it don't fix the actual issue, but merely avoids the crash by stringifying the things that were making it crash: gist.github.com/zoffixznet/3e9eba8...f14a9f410a 18:38
m: say Pod::FormattingCode ~~ Pod 18:39
camelia False
Zoffix m: say Pod::FormattingCode.^mro
camelia ((FormattingCode) (Block) (Any) (Mu))
Zoffix m: say Pod::FormattingCode.^mroĀ».^name
camelia List
Zoffix m: say Pod::FormattingCode.^mro.map: *.^name
camelia (Pod::FormattingCode Pod::Block Any Mu)
Zoffix m: say Pod.WHO.^name 18:41
camelia Stash
Zoffix pod don't look to be a type at all
tbrowder well, i tried all i know how to do to find the actual problem (i did make some painful debug print progress but stopped when the errors stopped, but never found the actual chunk that caused it. i was planning to send the problem to Damian to use in the grammar parsing tutorial he's having at TPC NA DC in a short while.
Zoffix oh wait
m: say Pod.HOW.^name 18:42
camelia Perl6::Metamodel::PackageHOW
tbrowder right, Pod is *not* a type, all else is Pod::Block except for Pod::Config as near as I can tell. 18:43
Zoffix Well, it's hard to fix a problem if you don't know what the problem is. I found the block by taking S26 and deleting chunks from the bottom until I found a chunk that caused the crash 18:44
what's "twine" anyway? 18:45
tbrowder well, i was zeroed in on it another misguided direction, glad you found it.
don't know--looks like a desription of how a Pod::Block can get twisted around with Declarations, leading and trailing text and such--Pod is not well documented. 18:47
twisted => twine 18:48
Geth rakudo/nom: 5e3393454c | (Zoffix Znet)++ | lib/Pod/To/Text.pm6
Fix regresion in from earlier today's commit

Most of the Pod stuff is not `Pod` type but `Pod::Block`. Stuff the rest through pod2text; that catches Pod::Config and enything else that might get added in teh future.
  [1] github.com/rakudo/rakudo/pull/1092
18:59
roast: 113dab63da | (Zoffix Znet)++ | S26-documentation/02-paragraph.t
Test nested paras don't crash

RT#131400: rt.perl.org/Public/Bug/Display.html?id=131400 Rakudo fix: github.com/rakudo/rakudo/commit/3721e68d24
   github.com/rakudo/rakudo/commit/22ef682537
   github.com/rakudo/rakudo/commit/5e3393454c
19:15
dogbert17 Zoffix: I suspect that your recent IO work has fixed github.com/perl6/doc/issues/335 20:16
Zoffix closed thanks 20:35
Zoffix tries swapping buggable back to using JSON::Fast 21:27
to see if it eats a ton of ram still :)
AlexDaniel Zoffix: I don't think it does 21:30
timotimo i certainly wouldn't know!~
AlexDaniel all whateverables switched to JSON::Fast and I think it should be workingā€¦
Zoffix cool 21:32
AlexDaniel Zoffix: good question. I honestly don't know 21:41
Zoffix neither
AlexDaniel ok I'll try to reproduce it and see if it helpsā€¦
Zoffix w00t. perl 5.26 released 21:45
AlexDaniel \o/ 21:46
MasterDuke while people seem to be around, any thoughts/comments on github.com/rakudo/rakudo/pull/1091 ? 21:59
jnthn Nice touch on the private part 22:02
uh...wow, that came out bad
I meant github.com/rakudo/rakudo/pull/1091...0cc00aR175
MasterDuke ha 22:03
jnthn StrDistance?
oh wow
I didn't realize that has a levenstein impl in it :D 22:04
timotimo it became magical last year or the year before that
jnthn Cute
MasterDuke what was it before?
jnthn MasterDuke: The only think I'm not sure about is :local
Zoffix 5.24.0
jnthn *thing
Zoffix Oh, the conversation has moved lol :} 22:05
jnthn I'm torn between the default or :all 22:06
:all would auto-correct methods from Mu/Any/Cool
MasterDuke yeah, wasn't sure
jnthn Well, suggest them
.gerp .uniqeu and so on 22:07
Which feels useful
AlexDaniel MasterDuke: you're not sorting them by distance, don't you?
jnthn Hm, it seems that it's only taking those with distance 1 22:08
AlexDaniel hm
jnthn Which is a slight pitty because .uniq won't suggest unique as that's (I guess?) a distance of 2?
AlexDaniel m: StrDistance.new(ā€˜uniqā€™, ā€˜uniqueā€™).say 22:09
camelia Default constructor for 'StrDistance' only takes named arguments
in block <unit> at <tmp> line 1
AlexDaniel m: StrDistance.new(:before<uniq>, :after<unique>).say
camelia StrDistance.new(before => "uniq", after => "unique")
MasterDuke the problem with allowing a too high distance is a distance of 2 for a 3 letter method is going to suggest a lot
AlexDaniel m: say +StrDistance.new(:before<uniq>, :after<unique>)
camelia 2
jnthn m: say +StrDistance.new(:before<gerp>, :after<grep>)
camelia 2
MasterDuke but i could always scale it
jnthn Typos like that are 2 also, so I think 1 is too low 22:10
Yeah, scaling on name length also works
The only other thing
Is that if the they put length or bytes, then they might get two sets of "Did you mean" 22:11
So I might declare @suggestions a bit earlier and push those into it also
MasterDuke easy enough 22:12
Zoffix
.oO( use large distance and pick first N from the returned list... )
22:13
jnthn As to :local my feeling is :all would be more useful
Zoffix: Trouble is that with short method names, a large distance is enough to delete all the chars and type new ones
Zoffix ah
MasterDuke right now i suggest no more than three things. 'elems', 'chars', 'graphs' or 'codes' are 4 by themselves
or maybe push them all together as one? 22:14
jnthn maybe but that'd hash up the plural-or-not thing? :)
MasterDuke ugh, yeah 22:15
could just say the limit is four, and if the method is length those are the only four they get... 22:16
jnthn Guess so 22:18
MasterDuke ok, i'll update the PR later this evening. now dinner & 22:21
jnthn Enjoy :)
Zoffix .tell tony-o you got unmerged PR: github.com/tony-o/perl6-data-dump/pull/28 22:23
yoleaux Zoffix: I'll pass your message to tony-o.
Zoffix heh, all the modules that conditionally-load Terminal::ANSIColor seem to be silently broken :) github.com/tony-o/perl6-data-dump/issues/27 22:41
Same problem's in my IRC::Client
.tell tony-o and one more: github.com/tony-o/perl6-data-dump/pull/29
yoleaux Zoffix: I'll pass your message to tony-o.
Zoffix Well, if it's isn't the job for... 22:42
greppable6: Terminal\:\:ANSIColor
greppable6 Zoffix, gist.github.com/22c5cd6f93886b9841...54a3a39869
Zoffix hehe, my Test::Notice also broken :P 22:43
hm, prove6 from TAP::Harness promises colour support if Terminal::ANSIColor is installed, but doesn't use it anywhere 22:44
Zoffix slaps whoever decided it was a good idea to rename Term::ANSIColor to Terminal::ANSIColor 22:45
Thanks for all the extra typing \o
It should've been Term::ANSI
You've misspelt colour :P