timotimo what i had would make formatting codes that span multiple rows work, though 00:00
tbrowder not sure it’s better ref formatting codes, at least verbatim ones, but otherwise i think they are preserved ok. 00:01
i did get Z comments deleted ok i think. 00:02
i do need to add some tests for formatting codes... 00:03
timotimo always a good idea
tbrowder but i don’t think there weren’t any before my changes, just to be fair. 00:04
timotimo there were a few, i believe 00:05
samcv this doesn't seem like a bug rt.perl.org/Ticket/Display.html?id=132217
tbrowder timotimo: can you point them out? 00:06
anyway, all current pod tests pass. 00:07
plus the tests i did add, of course.
timotimo er, no clue 00:11
agree, samcv 00:12
tbrowder there is a roast test for formatting code, but not in tab
es
*tables
samcv going to fix this LTA where using a glyph with a combiner just gives a normal error and change it to 00:18
Starter [̈ is immediately followed by a combining codepoint. Please use [ without a combining glyph ([̈ was at line 1) 00:19
i think that's much clearer
there's some other "Couldn't find terminator" $/.panic's in Grammar.nqp for rakudo but i'm not sure if i should add those to do the same thing 00:20
since i'm not really sure what triggers them
in `tribble` and `sibble` 00:21
i changed the one in quibble
can anybody tell me what tribble and sibble are :P
timotimo yup 00:22
they are nibblers for tr/// and s///
samcv ah
should i add that as a comment? 00:23
Geth rakudo: ad7c97dfb4 | (Samantha McVey)++ | src/Perl6/Grammar.nqp
Fix LTA error when starter is followed by a combining codeponit

Affects q quoting, s/// and possibly tr///. Notifies the user that they
  are using a combining codepoint after the starter and to remove it for
it to properly parse.
00:46
roast: f99444b18f | (Samantha McVey)++ | 2 files
Add test for .uc/.lc with prepend codeponints. RT132020

Add a test to close RT132020. This test makes sure .uc works properly even when a GCB=Prepend codepoint comes in front of it.
01:00
synopsebot RT#132020 [new]: rt.perl.org/Ticket/Display.html?id=132020 [UNI] Prepend character breaks case changing functions on MoarVM
timotimo why can't i quote with brackets that have combiners on them ... 01:11
samcv heh. well you *might* be able to. not sure. regardless this doesn't impose any extra restrictions 01:31
it just makes the error better 01:32
timotimo, you can
m: "q[\x[308]foo[\x[308]".EVAL.perl.say
camelia "foo"
samcv but it doesn't identify them as brackets
but as the *any* symbol type thing 01:33
timotimo ah 01:34
samcv i mean we could probably fix that so it could work 01:35
though really it's kind of "a bad idea"
i mean it would be nice if it could include the combiner as part of the string. but that just seems to be asking for problems for an edge case. and it only affects things typed into the .p6 document. so in fear of violating unicode principles it's prolly best to not let you do that 01:37
timotimo yeah, users will probably ahve to resort to \x or \c to get a dangling combiner as the first character 01:39
Geth tap-harness6: 0e2172394d | (Zoffix Znet)++ (committed using GitHub Web editor) | META6.json
Bump vesion
03:28
samcv is working on some bugs right now and going through and tagging untagged stuff- 06:18
in RT that is
is it just me or is compiling rakudo taking a lot longer than it used to? 06:21
i haven't measured it so no clue if that's actually true or not 06:22
travis-ci Rakudo build passed. Samantha McVey 'Fix LTA error when starter is followed by a combining codeponit 06:27
travis-ci.org/rakudo/rakudo/builds/298858613 github.com/rakudo/rakudo/compare/0...7c97dfb4b1
Geth roast: 9bb245211c | usev6++ | S17-supply/syntax.t
[jvm] Skip new test that uses atomicint
06:58
nqp: 2b13b1f781 | (Samantha McVey)++ | src/HLL/sprintf.nqp
Add support for '+' flag for %b in sprintf
08:01
nqp: a02ff393db | (Samantha McVey)++ | t/hll/06-sprintf.t
Correct incorrect tests with %b and + option in sprintf
rakudo: b0cda59091 | (Samantha McVey)++ | tools/build/NQP_REVISION
Bump nqp for sprintf fix

Changes: 2017.10-46-gf263b9c..2017.10-54-ga02ff393d a02ff393d Correct incorrect tests with %b and + option in sprintf 2b13b1f78 Add support for '+' flag for %b in sprintf 4e0dfc54d [jvm] Implement close for ServerSocketHandle 9d28fccb2 [js] Add missing file 00e8bd16c Test accepts_type a05185458 [js] Take a named argument that might be passed 17ce8f83d [jvm] Implement 'getport' for ServerSocketHandle b21e0e25f [jvm] Skip not yet fixed test
08:18
roast: 77270a1df8 | (Samantha McVey)++ | S32-str/sprintf-b.t
Untodo many sprintf %b tests should close RT124696
08:20
synopsebot RT#124696 [new]: rt.perl.org/Ticket/Display.html?id=124696 Roast rakudo skip/todo test:./S32-str/sprintf-b.t line:23 reason: '%b with flag plus not correct'
samcv m: use nqp; say(nqp::sprintf('%010d', [-9])) 09:31
camelia Directive d not applicable for type VMNull
samcv m: use nqp; say(sprintf('%010d', -9))
camelia -000000009
samcv m: use nqp; say(sprintf('%010b', -9))
camelia 00000-1001
samcv well i figured how %b and %d differ in sprintf and how it adds the extra 0's in a different place for %b. which seems very odd… and will need to fix that for more things than ust %b possibly. not sure if any other directives have that issue aside from %b 09:32
i think it has something to do with %g directive though. based on how sprintf works in nqp. and the different order of things is what causes it 09:33
Geth rakudo/js: 95991afd3d | pmurias++ | src/core/Lock/Async.pm
[js] Disable locking code that serves no purpose

It needs to be fixed on the js backend as it's causing exceptions
09:34
samcv i would like the %b directive to return from the method *with* the prepended 0's and - sign instead of adding it later but will have to work on that
so it doesn't affect any other directives and break %g and maybe others
anybody know if i can differentiate between token directive:sym<b>($/) and token method directive:sym<foo>($/) by using what's in $/? 09:39
$<directive> is availible but is there a way to see if $<directive> in $/ is from sym<b> or sym<foo> 09:40
[Tux] Rakudo version 2017.10-145-gb0cda5909 - MoarVM version 2017.10-58-gad8618468
csv-ip5xs1.121 - 1.203
csv-ip5xs-2012.981 - 13.450
csv-parser12.165 - 12.232
csv-test-xs-200.444 - 0.451
test12.186 - 12.264
test-t3.248 - 3.249
test-t --race1.280 - 1.360
test-t-2059.165 - 59.543
test-t-20 --race20.542 - 21.141
09:56
(on lizmat' request "test-t --race" added
Geth nqp: 63093556e0 | usev6++ | src/HLL/sprintf.nqp
Improve handling of flags for '%b' directive
10:45
nqp: aa86255e75 | usev6++ | t/hll/06-sprintf.t
Adjust tests for '%b'
nqp: a09c78780f | (Samantha McVey)++ | 2 files
Merge branch 'sprintf' into master

Merge the sprintf branch from 2 years ago into master for better handling of %b spacing.
samcv bartolin: got the sprintf merged after doing some conflict resolution
\ol
\o/
lizmat Files=1229, Tests=75812, 318 wallclock secs (14.80 usr 5.37 sys + 2191.30 cusr 216.48 csys = 2427.95 CPU) 10:55
Zoffix is considering taking out the "community" spread (or at least the photo): twitter.com/ribasushi/status/928182573725085696 11:26
Usually, we get the people in pics to sign the photo model release document, to avoid people suing and saying you made $x benefit from my pic, you owe me. 11:27
pmurias Zoffix: we don't have a more Perl 6 focused picture? 11:29
Zoffix pmurias: the only ones I know are just 4-9 people (e.g. www.facebook.com/photo.php?fbid=90...mp;theater ). It doesn't sell "community" well. And still have the issue of someone going rogue and spoiling the broth for everyone. 11:31
And on unrelated note... didn't riba say if his fundraiser failed he'd leave Perl community forever or did I miss the "I'll stay and troll everyone" clause?
lizmat Zoffix: FWIW, I like the spread 11:48
but I guess we will have to ask every individual in that photo to actually give permission 11:49
before we can actually use it 11:50
shall I take that task ?
wrt riba: he just likes to stoke up things, I think he's done enough for Perl that he can't be labelled "troll" just like that 11:51
for that, other people come to mind (regardless of how much they've done) 11:52
but in both cases, it's the scorned lover and hell has no fury kind of thing, I guess :-( 11:53
Zoffix lizmat: that sounds like a hell of a task, so I'll leave it up to you. I like the photo too. 11:58
lizmat: and if anyone says no, ask them if they're OK with just their face covered with a camelia logo. 11:59
lizmat no, if anyone says no, it's a no go for the picture I think: blocking out people in that picture will just give the wrong signal, I think 12:00
Zoffix OK. 12:01
jnthn Also, should perhaps try to take a photo explicilty for these purposes the next time we can get sufficiently many Perl 6 developing/using folks in the same place :) 12:03
lizmat I guess LPW would be the next opportunity 12:04
Geth rakudo: f8e1a5faab | (Elizabeth Mattijsen)++ | lib/Telemetry.pm6
Separate the Instrument from the Snap

  - before, the Instrument object would contain the snap
   - this seemed like a good idea at the time
   - but isn't, because we cannot instantiate an Instrument
  - introduces the Telemetry::Instrument::Snap role
   - which is mostly what Telemetry::Instrument was
   - minus "formats" and "columns"
  - should make it easy to make a Telemetry::Instrument::AdHoc
   - which could e.g. just take a number of variables to snap
12:42
lizmat do we have a (un)official way to attach a WHY to a scalar container ? 13:06
#| Because(nl)my $a = 42; dd $a.VAR.WHY
doesn't work :( 13:07
would there be anything against adding a "has str $!why" to Perl6/Metamodel/ContainerDescriptor.nqp 13:31
and trying to make at least #| work on variable declarations ?
jnthn Aren't variables more implementation detail than documentable API? 13:32
lizmat well, I guess that's true,
hmmm...
timotimo the rakudo ops (the C code) have a struct that mirrors the structure of ContainerDescriptor, so if you change one, you should also change the other, or explosions
lizmat timotimo: odd, I just added it without any explosions during build *or* spectest 13:33
jnthn But if we did do it then yeah, ContainerDescriptor would be the cheapest place. Also what timotimo said is true, though adding it as the last attribute wouldn't trip that up.
timotimo yeah, put it last and nothing explodes
try putting it in the middle for some *fun*
jnthn Putting it on Scalar would add 8 bytes to every array/hash element since they are also Scalars :)
lizmat ah, it *was* the last attribute
yeah, so maybe not such a good idea :-)
my reasoning was to have a Telemetry::Instrument::Adhoc 13:34
jnthn Well, ContainerDescriptor is shared among Scalar, so it's not quite so bad, though still makes every static var cost a bit more
lizmat that would allow you to : $*SAMPLES = T::I::Adhoc( $var1, $var2, $var3) and have the legend auto-generate from the WHY of the containers 13:35
jnthn ah
timotimo we can totally have a pass in the optimizer that replaces all lexical variables in internal functions with the same set of names :P
jnthn Alternative API would be LegendLabel => $var
And if you just want the name of the var as the legend label use :$var1 13:36
lizmat well, the named approach is not an option, as I want to snap the container (and changes in it)
I lose the container with the :$name syntax 13:37
jnthn *%maybe-slurp-them-with is raw ? :)
lizmat jnthn: no, doesn't work either :-)
jnthn Really?
lizmat pretty sure, lemme double check
jnthn m: my $a = 42; sub f(*%h is raw) { say %h<a>.VAR.name }; foo(:$a) 13:38
camelia 5===SORRY!5=== Error while compiling <tmp>
Undeclared routine:
foo used at line 1
jnthn m: my $a = 42; sub f(*%h is raw) { say %h<a>.VAR.name }; f(:$a)
camelia $a
lizmat hmmm... cool
jnthn Works on somebody's machine :-)
lizmat ok, so that solves that part 13:39
alexk6_ m: package A { package B { package C::B {}}}; say A::B::C::B.WHAT; 13:54
camelia (B)
alexk6_ m: package A { package B { package A::B {}}}; say A::B::A::B.WHAT;
camelia Could not find symbol '&B'
in block <unit> at <tmp> line 1
alexk6_ is that a bug?
lizmat alexk6_: not 100% sure, but definitely an issue that needs clarification :-) 13:56
afk for a few hours&
alexk6_ m: package A::B::A::B {}; say A::B::A::B.WHAT 13:57
camelia (B)
Zoffix lizmat: is David Golden one of the people in the pic? 14:00
lizmat Zoffix: no 14:05
David Golden was not at the PTC 2017 14:06
why ?
Zoffix lizmat: he commented on riba's tweet about not wanting to be included next year. If he were on this pic, that would've saved the trouble of asking everyone else :) 14:08
AlexDaniel releasable6: status 14:13
releasable6 AlexDaniel, Next release in 10 days and ≈4 hours. No blockers. 0 out of 146 commits logged
AlexDaniel, Details: gist.github.com/40cc059e0f703e9bd6...cfb1b4dfcf
ggoebel ./perl6-js -e 'say(123)' 14:31
ThreadPoolScheduler is not implemented on the JS backend
make js-spectest 14:34
cd t/spec && git pull --ff-only
Already up-to-date.
Use of bare << to mean <<"" is deprecated at t/harness5 line 43.
syntax error at t/harness5 line 52, near "HEAD
my "
(Might be a runaway multi-line '< string starting on line 49)
ilmari looks like you've got a merge conflict in t/harness5 14:35
pmurias ggoebel: hi 15:15
ggoebel: do you have the latest commit in the js branch, this looks like something I have fixed 15:17
Geth rakudo: ggoebel++ created pull request #1241:
fix merge conflict
15:32
synopsebot RAKUDO#1241 [open]: github.com/rakudo/rakudo/pull/1241 fix merge conflict
ggoebel pmurias: github.com/rakudo/rakudo/blob/js/t/harness5#L43 15:35
ilmari ggoebel: with perl 5.26 you would have got a more awesome error message: Version control conflict marker at t/harness5 line 43, near "<<<<<<<" 15:38
pmurias ggoebel: thanks for fixing 15:40
ggoebel ilmari: Nice. I'm only running 5.22.1 perhaps if I were still using gentoo... switched to ubuntu LTS a while ago... And I've mostly stopped installing/tracking personal current perl5 distributions.
pmurias: thanks for your continued efforts on perl6-js :-) 15:41
pmurias ggoebel: does the ThreadPoolScheduler bug occur in the latest commit
ggoebel yes
Geth rakudo/js: a628e1c608 | (Garrett Goebel)++ | t/harness5
fix merge conflict
rakudo/js: e34f66e3a2 | (Paweł Murias)++ (committed using GitHub Web editor) | t/harness5
Merge pull request #1241 from ggoebel/js

fix merge conflict
pmurias ggoebel: I'll build a fresh rakudo.js and see why the problem persists 15:45
ggoebel pmurias: excellent. I'll run some errands and give it a go when I get back 15:46
Geth rakudo/js: b998bfe576 | pmurias++ | tools/build/NQP_REVISION
[js] Bump NQP
16:20
rakudo/js: 640491d60c | pmurias++ | t/harness5
Merge branch 'js' of github.com/rakudo/rakudo into js
nqp: af7f72c137 | pmurias++ | src/vm/js/Operations.nqp
[js] Fix if bug
16:21
nqp: 58083fe626 | pmurias++ | 2 files
[js] Implement nqp::cpucores
nqp: 5dba8525b3 | pmurias++ | src/vm/js/nqp-runtime/reprs.js
[js] Move method into correct repr
rakudo/js: 9463906b98 | pmurias++ | tools/build/NQP_REVISION
[js] Bump NQP more correctly
16:22
pmurias ggoebel: an un-updated nqp version was causing the problem 16:59
Zoffix__ lizmat, FYI: I'm cutting the pic. The Twitter thread is getting out of hand and Aristotle (who is in the picture) said they're weirded out. 17:39
Cutting the whole community spread actually. People suck. 17:44
Zoffix__ de-tethers and returns to the land of network outage :( 17:45
lizmat Zoffix: noted 18:08
yoleaux 17:21Z <wander> lizmat: "tbrowder" is working on Pod::To::Table parsing at rakudo PR#1240.
ggoebel pmurias: thanks! js-spectest is much happier :-) 18:23
All tests successful. 18:27
Files=36, Tests=1180, 199 wallclock secs ( 0.51 usr 0.09 sys + 1058.57 cusr 31.98 csys = 1091.15 CPU)
lizmat .ask samcv do we have a way to assemble a unicode name at runtime and look that up ? 20:12
yoleaux lizmat: I'll pass your message to samcv.
samcv assemble? 20:13
yoleaux 20:12Z <lizmat> samcv: do we have a way to assemble a unicode name at runtime and look that up ?
lizmat m: my $a = "LATIN CAPITAL LETTER A"; say "\c[$a]"
camelia 5===SORRY!5=== Error while compiling <tmp>
Unrecognized \c character
at <tmp>:1
------> 3 $a = "LATIN CAPITAL LETTER A"; say "\c[7⏏5$a]"
expecting any of:
argument list
double quotes
term
samcv m: my $a = "LATIN CAPITAL LETTER A"; say $a.parse-names 20:16
camelia A
samcv or you could use the nqp thing
nqp::getstrfromname()
lizmat samcv++ :-) thanks! 20:17
Geth rakudo: ufobat++ created pull request #1242:
RFC: Use IO::Spec in the method resolve of IO::Path
20:51
synopsebot RAKUDO#1242 [open]: github.com/rakudo/rakudo/pull/1242 RFC: Use IO::Spec in the method resolve of IO::Path
samcv lizmat: np :-) 21:18
Geth roast: 3560c6bf43 | usev6++ | S06-traits/misc.t
Fix typo in comment
21:33
pmurias ggoebel: rakudo.js passes a loads more spec tests but I haven't but them in the spec test list 21:56
Geth nqp: ugexe++ created pull request #381:
Add getppid op for moarvm
23:26
synopsebot NQP#381 [open]: github.com/perl6/nqp/pull/381 Add getppid op for moarvm