MasterDuke i don't use twitter, but someone who does might like to reply to twitter.com/colomon/status/964127177229918208 saying that just got a bit faster 01:11
and it's really just creating the string representations that's slow, calculating the sequence itself is pretty fast 01:12
Zoffix Done: twitter.com/zoffix/status/967207578076934144 01:22
AlexDaniel greppable6: undefine 02:34
yoleaux 23 Feb 2018 22:47Z <tbrowder> AlexDaniel: NQP#424 has an iteresting mod to .travis.yml to help get past stallevjobs
synopsebot NQP#424 [open]: github.com/perl6/nqp/pull/424 add a new built-in sub: run-command
greppable6 AlexDaniel, 1151 lines, 104 modules: gist.github.com/19db2bf219677cd70c...3fa1663d8a
AlexDaniel greppable6: undefine\b
greppable6 AlexDaniel, 17 lines, 12 modules: gist.github.com/110496947900da1790...3406ea9c10
AlexDaniel oh wow it's used a few times 02:35
greppable6: undefine\b[$@%]
greppable6 AlexDaniel, Found nothing!
AlexDaniel greppable6: undefine \W 02:36
greppable6 AlexDaniel, 6 lines, 5 modules: gist.github.com/c4c56e292b3a6a0a0d...6504efbefa
MasterDuke Zoffix++ 02:55
Geth roast/master: 5 commits pushed by (Nick Logan)++, (Zoffix Znet)++ 03:11
rakudo/master: 4 commits pushed by (Zoffix Znet)++, (Aleks-Daniel Jakimenko-Aleksejev)++ 04:05
AlexDaniel also I'm pretty sure that release/2018.02.1 branch can be just deleted now 04:07
AlexDaniel just did that 04:09
stmuk: thanks 04:12
samcv what was the release bumped for? 05:26
AlexDaniel samcv: github.com/rakudo/rakudo/issues/1550 05:27
possibly related: irclog.perlgeek.de/perl6-dev/2018-...i_15851641 :) 05:28
lizmat Files=1235, Tests=76226, 317 wallclock secs (14.98 usr 5.18 sys + 2177.86 cusr 213.62 csys = 2411.64 CPU) 09:16
[Tux] Rakudo version 2018.02.1-29-gd3a7d7d2e - MoarVM version 2018.02-6-g1849ae6d6
csv-ip5xs0.820 - 0.930
csv-ip5xs-207.845 - 7.937
csv-parser12.505 - 12.845
csv-test-xs-200.435 - 0.444
test9.626 - 9.836
test-t2.703 - 2.706
test-t --race1.130 - 1.155
test-t-2047.504 - 47.790
test-t-20 --race16.851 - 17.712
 10:09
first double run was slower
timotimo Zoffix, regarding the point release, i'd say ... all is for given! 10:50
Geth rakudo: 4f47386762 | (Elizabeth Mattijsen)++ | src/core/Date.pm
Make sub sleep/sleep-timer/sleep-until multi's
 11:18
rakudo: bf5e335730 | (Elizabeth Mattijsen)++ | src/core/Hash.pm
Make sub hash a multi

  - also handle named only's faster		 11:51
nqp/master: 10 commits pushed by (Timo Paulssen)++
review: github.com/perl6/nqp/compare/b5da0...f77a7edb33		 11:59
nqp: cb3c1a9b00 | (Timo Paulssen)++ | tools/build/MOAR_REVISION
bump moar to get new profiler data formats

mostly changes for multithreading, also makes the profiler less crashy when running it with multi-threaded programs.		 12:01
Ā¦ nqp: version bump brought these changes: github.com/MoarVM/MoarVM/compare/2...gc6543e21f
dogbert11 bisect: say Blob[uint8].new(233).decode("utf8-c8").encode("utf8-c8").perl 12:02
bisectable6 dogbert11, Bisecting by output (old=2015.12 new=bf5e335) because on both starting points the exit code is 0
Geth rakudo: 26e34cd1f4 | (Elizabeth Mattijsen)++ | src/core/Hash.pm
Update error description on sub circumfix:<:{ }>()

For some reason it now thinks we're changing the Perl 6 grammar now. So either we have a problem in the Perl 6 grammar in the handling of :{ } or we have a false positive on the error message.
Not pursuing this further at this point.
bisectable6 dogbert11, bisect log: gist.github.com/0de4ac173a1cfcfec9...36d4af058e
dogbert11, (2016-04-01) github.com/rakudo/rakudo/commit/a8...b417a0d372
travis-ci NQP build failed. Timo Paulssen 'ensure threads without call graph don't mess up sql syntax' 12:18
travis-ci.org/perl6/nqp/builds/345623058 github.com/perl6/nqp/compare/b5da0...f77a7edb33
timotimo we can't install the clang we want on travis for some reason 12:19
Depends: libstdc++-4.9-dev but it is not installable
Depends: libgcc-4.9-dev but it is not installable
Depends: libobjc-4.9-dev but it is not installable
the failure got resolved by trying again ... 12:26
travis-ci.org/perl6/nqp/builds/345623764 - does this show any actual builds for anybody? all i get is "it's been running for 25 minutes" but no logs to be found
lizmat I see a log for travis-ci.org/perl6/nqp/jobs/345623767 12:29
timotimo interesting; i don't get a link to that job
lizmat perhaps some JS fun ? 12:30
timotimo no clue :<
tbrowder i note that NQP has six or so PRs passing all tests (one has a conflict). i say they should either be merged or the author told why not. 12:35
lizmat timotimo: perhaps a reboot of your browser ? 12:36
timotimo it's not that important, i reckon 12:38
Geth rakudo: 5210d7028d | (Elizabeth Mattijsen)++ | src/core/Int.pm
Make sub is-prime a multi

I kept the "is pure" on the proto, although I think this is technically incorrect as an added candidate could not be guaranteeing a "pure" result necessarily (altough in this case that would probably be folly).
Additional notes:
  - is-prime(3) still constant folds (if the "is pure" is on the candidate,
   it wouldn't). Not sure what is going on there.
  - 3.is-prime does *not* constant fold, even when annotated with "is pure"		 12:39
tbrowder timotimo: if you look at my NQP#424 you will see my .travis.yml mod that got my stalled job working again 12:40
synopsebot NQP#424 [open]: github.com/perl6/nqp/pull/424 add a new built-in sub: run-command
tbrowder my money is on it working for your job 12:41
travis-ci NQP build passed. Timo Paulssen 'bump moar to get new profiler data formats 12:45
travis-ci.org/perl6/nqp/builds/345623764 github.com/perl6/nqp/compare/09f77...3c1a9b00de
tbrowder well shut my mouth!! 12:46
timotimo tbrowder: i think the problem was only on the display side 12:47
because liz did see the jobs 12:48
tbrowder it did take a long time, though. the difference between yr job and mine was in mine travis reported a failure due to a process time out. while yourā€™s didnā€™t. 12:50
Geth rakudo: b704a17513 | (Elizabeth Mattijsen)++ | src/core/io_operators.pm
Make sub gist a multi
 12:54
tbrowder timotimo: why havenā€™t you merged NQP#283 12:57
synopsebot NQP#283 [open]: github.com/perl6/nqp/pull/283 expose coerce_in and _ni for direct int <-> num
timotimo don't knw tbh 12:58
tbrowder there are several open nqp PRs with good tests just waiting to be merged 12:59
that is, unless core devs disapprove for some reason..l 13:00
lizmat tbrowder: if someone would disapprove, they would say, I'm sure 13:04
seems your PR's got warnocked somehow, which isn't too strange with the low bus number on nqp internals :-( 13:05
tbrowder ā€œwarnockedā€, as in low priority? 13:06
lizmat en.wikipedia.org/wiki/Warnock%27s_dilemma 13:07
tbrowder oh...that sounds familiar, i think AD gets tired... 13:09
lizmat www.nntp.perl.org/group/perl.perl6...15407.html # more specific to Perl 6 13:11
Geth rakudo: 4c67498f90 | (Elizabeth Mattijsen)++ | src/core/io_operators.pm
Make subs chdir/chmod/unlink/rmdir/mkdir/rename/copy/move/symlink/link a multi
 13:13
timotimo wow, i hadn't realized warnock was perl6-related at all 13:18
tbrowder timotimo: any objection to my merging yr old pr? 13:23
going onto 2yrs
MasterDuke timotimo: you said something a couple days ago about getting that prime code someone (rindolf?) had a bunch faster. was that by optimizing his code, or rakudo/moarvm? 13:24
Geth rakudo: 7d72387bc0 | (Elizabeth Mattijsen)++ | src/core/List.pm
Make sub combinations/permutations a multi
 13:33
nqp/master: 12 commits pushed by (Tom Browder)++
review: github.com/perl6/nqp/compare/cb3c1...acf2c303a9		 13:35
timotimo MasterDuke: i split a mathy expression into multiple assignments into native int typed variables 13:44
i wonder if i can golf it down significantly into a one-liner 13:48
the expression used to be @d.ASSIGN-POS($m, (@d.AT-POS($m) + @d.AT-POS($m - $p)) % $BASE) 13:51
now it's that over four lines with three variables
MasterDuke heh, sounds somewhat less readable
tbrowder .ask pmurias any debugging ideas about situation where nqp tests on jvm pass on my local host but fail on travis-ci? 14:02
yoleaux tbrowder: I'll pass your message to pmurias.
timotimo well, i can probably put all the "my"s into one line again 14:05
i don't remember what i wanted coerce_ni and coerce_in for 14:13
dogbert17 my @primes = grep { .is-prime }, 1 .. *; my @p = gather for 4000, 5, 100, 2000 -> $n { take start { @primes[$n] }; }; .say for await @p; 15:41
m: my @primes = grep { .is-prime }, 1 .. *; my @p = gather for 4000, 5, 100, 2000 -> $n { take start { @primes[$n] }; }; .say for await @p; 15:42
camelia An operation first awaited:
in block <unit> at <tmp> line 1

Died with the exception:
No such method 'is-prime' for invocant of type 'Mu'
in block at <tmp> line 1
in block at <tmp> line 1
dogbert17 what's wrong here?
if anything :-) RT #127208 15:43
synopsebot RT#127208 [open]: rt.perl.org/Ticket/Display.html?id=127208 [SEGV][CONC] Non-deterministic segfaults in parallel code 15:44
MasterDuke *.is-prime
dogbert17 m: my @primes = grep { *.is-prime }, 1 .. *; my @p = gather for 4000, 5, 100, 2000 -> $n { take start { @primes[$n] }; }; .say for await @p; # like this ? 15:46
camelia 5===SORRY!5=== Error while compiling <tmp>
Malformed double closure; WhateverCode is already a closure without curlies, so either remove the curlies or use valid parameter syntax instead of *
at <tmp>:1
------> 3my @primes = grep { *.is-priā€¦
MasterDuke whoops, my bad
dogbert17 it's a bit strange, I'm probably missing something obvious 15:48
MasterDuke m: my @primes = grep { .is-prime }, 1 .. *; my @p = gather for 40, 5, 1, 20 -> $n { take { @primes[$n] }; }; .().say for @p; 15:52
camelia 179
13
3
73
MasterDuke m: my @primes = (1 .. *).grep(*.is-prime); my @p = gather for 4000, 5, 100, 2000 -> $n { take start { @primes[$n] }; }; .say for await @p; 15:58
camelia 37649
13
547
17333
MasterDuke m: my @primes = (1 .. *).grep({ .is-prime }); my @p = gather for 4000, 5, 100, 2000 -> $n { take start { @primes[$n] }; }; .say for await @p; 15:59
camelia 37699
13
547
17359
MasterDuke dogbert17: ^^^, works with the method form of grep
mst how come the block form and the whatever star form return different results? did I miss something? 16:03
MasterDuke i had different indices in the block form 16:06
mst both seem to be 4000, 5, 100, 2000 ?
I mean .grep(*.is-prime) versus grep({ .is-prime }) 16:07
MasterDuke oh, the block form of the method call, yeah, that's weird
mst the 5 and 100 results are the same, the 2000 and 4000 results are different
that ... doesn't strike me as something that should happen
MasterDuke i get a bunch of different values if i run it a bunch more times 16:08
i'm guessing some sort of race condition with the iterator for @primes? 16:09
m: my @primes = (1 .. *).grep({.is-prime}); my @p = gather for 4000, 5, 100, 2000 -> $n { take { @primes[$n] }; }; .().say for @p; 16:10
camelia 37831
13
547
17393
MasterDuke because this always returns the same values
mst: btw, don't know if you saw it, but Zoffix was asking about a Perl 6 dev cloak 16:12
mst I say alexdaniel ask
*saw
that one's done
AlexDaniel oh 16:13
yes :)
mst of course you'll have to log in :P
MasterDuke whoops, you're right, my bad
dogbert17 MasterDuke: thx 16:14
AlexDaniel mst: thanks!
MasterDuke mst: does one have to have signed the CLA to get one? 16:16
mst I can give out @perl6/name to anybody who wants one
@perl6/developer/name I was *guessing* should be people with commit bits 16:17
but
I'm mostly "the dude doing the paperwork" here, so if people with commit bits tell me to give 'em to other people, I'm not going to argue
like, the perl community in general has never exactly been fussy about formal status, so I'm not going to go all bureaucratic about it ;) 16:18
AlexDaniel MasterDuke: soooā€¦ with 178 commits to rakudo, isn't it time to submit a CLA? ;)
pmurias what are the Perl 6 dev cloaks useful for? 16:20
mst beiny shiny?
timotimo hides your ip from evil haxors! 16:21
mst cloaks don't hide your ip particularly reliably
MasterDuke i'll get around to it eventually
mst basically, cloaks are just there to show project affiliation 16:22
AlexDaniel pmurias: I was thinking this: irclog.perlgeek.de/perl6-dev/2018-...i_15850058
mst oh, yeah, they do make hostmask-based matching easier 16:23
AlexDaniel .tell Zoffix now that I have a dev cloak, can you add me to the list for p6lert? 16:24
yoleaux AlexDaniel: I'll pass your message to Zoffix.
pmurias tbrowder: haven't dealt with travis jvm behaving differently, so I don't have any good hints 16:25
Zoffix AlexDaniel: done. Docs for interface are: github.com/perl6/alerts#posting-alerts 16:29
yoleaux 16:24Z <AlexDaniel> Zoffix: now that I have a dev cloak, can you add me to the list for p6lert?
Zoffix FWIW, anyone with sudo access to hack can add users. You'd edit /home/p6lert/alerts/secret.json config file. 16:30
mst Zoffix: tempted to suggest that your cloak should be @perl6/botguru/zoffix :)
Zoffix heh 16:31
MasterDuke AlexDaniel could be botherder, or maybe botfather, botdad 16:34
timotimo botherer?
AlexDaniel ā€¦ botherable
MasterDuke: btw there are two dads :P github.com/perl6/whateverable/blob...le.pm6#L47 16:43
MasterDuke heh 16:44
mst # TODO roles should not have TWEAK method 16:58
why not?
having to do 'sub BUILD {} after BUILD => sub { ... };' in Moo(se) roles has always been a bit annoying
timotimo a role's TWEAK method will be merged into the class it's applied to by overwriting what's there already 17:00
AlexDaniel yeah, something like that. IMO it should not be that way 17:01
TimToady in general, it seems like submethods should be glommed together, not overridden 17:02
japhb
.oO( Glomming Submethods for Fun and Profit )
... and FWIW I agree 17:03
mst I thought TWEAK and BUILD being submethods it wouldn't overwrite 17:04
TimToady perhaps it's currently "shouldn't" :) 17:06
mst oh, right, I misread your glomming comment 17:08
TimToady I don't actually know whether there is an attempt to glom currently; haven't been in that code much...
mst timotimo: also, surely if the class already had one, it wouldn't be overwritten - and if the superclass has one, that'll still get called via whatever equivalent mechanism to 'call all BUILDs' 17:09
TimToady certainly we glom at the class level, that's what the whole buildplan is about
but glomming should likely be guaranteed for role submethods as well 17:10
mst I mean, still not really helpful, but that's why 'sub BUILD {} after BUILD => sub { ... };' works in perl5 M* OO to emulate glomming
Geth nqp: 1f96304a73 | usev6++ | src/vm/jvm/runtime/org/perl6/nqp/sixmodel/reprs/DecoderInstance.java
[JVM] Don't add to decoder if there is no byte

This should fix the last part of github.com/rakudo/rakudo/issues/1541		 17:30
nqp: ad9762520b | usev6++ | 2 files
Test for empty decoder after adding empty buffer
nqp: 17fd55f9de | (Christian BartolomƤus)++ (committed using GitHub Web editor) | 3 files
Merge pull request #423 from usev6/jvm_decoder_empty

  [JVM] Fix eof detection after read from empty file
roast: c227876274 | skids++ | S02-magicals/dollar-underscore.t
Add tests for aliasing, or not, of $_ in various block syntaxes
 18:52
nqp: fe7cae2142 | (Nick Logan)++ (committed using GitHub Web editor) | 2 files
Add getppid op (#422)

Allows a child process to get the parent process id, and thus guess if it has become orphaned.
  See: #381		 18:58
synopsebot NQP#381 [closed]: github.com/perl6/nqp/pull/381 Add getppid op for moarvm
MasterDuke timotimo: did you see the possible race condition we were experimenting with earlier? 19:02
timotimo multiple threads reifying the same array?
MasterDuke could be 19:03
m: my @primes = (1 .. *).grep({ .is-prime }); my @p = gather for 4000, 5, 100, 2000 -> $n { take start { @primes[$n] }; }; .say for await @p; 19:04
camelia 37657
13
547
17299
MasterDuke m: my @primes = (1 .. *).grep({ .is-prime }); my @p = gather for 4000, 5, 100, 2000 -> $n { take start { @primes[$n] }; }; .say for await @p;
camelia 37589
13
547
17293
MasterDuke notice the different results
jnthn Yeah, that ain't going to work out too well, at least, not unless we decide we're going to insert locking into lazy arrays 19:05
timotimo yeah, no, we probably don't want that 19:06
MasterDuke m: my @primes = .grep { .is-prime }, 1 .. *; my @p = gather for 4000, 5, 100, 2000 -> $n { take start { @primes[$n] }; }; .say for await @p; 19:08
camelia 5===SORRY!5=== Error while compiling <tmp>
Unexpected block in infix position (missing statement control word before the expression?)
at <tmp>:1
------> 3my @primes = .grep7ā5 { .is-prime }, 1 .. *; my @p = gather f
expecting anyā€¦
MasterDuke m: my @primes = grep { .is-prime }, 1 .. *; my @p = gather for 4000, 5, 100, 2000 -> $n { take start { @primes[$n] }; }; .say for await @p;
camelia An operation first awaited:
in block <unit> at <tmp> line 1

Died with the exception:
No such method 'is-prime' for invocant of type 'Mu'
in block at <tmp> line 1
in block at <tmp> line 1
MasterDuke that's the other odd thing ^^^
jnthn Same underlying cause, surely 19:09
timotimo: Well, given it only impacts upon things that are marked lazy, when you already asked to trade time for less memory etc...
timotimo: Provided we can do it without making the non-lazy case more costly I'm open to looking into it at least
timotimo fair enough 19:10
surely we can do something clever with regards to signaling 19:12
i.e. if one thread is already reifying 0 to 10_000 and we are interested in value 100, but it's not yet ready
travis-ci NQP build failed. Nick Logan 'Add getppid op (#422) 19:14
travis-ci.org/perl6/nqp/builds/345715828 github.com/perl6/nqp/compare/17fd5...7cae21420f
MasterDuke synopsebot: source 19:21
synopsebot MasterDuke, See: github.com/perl6/synopsebot
MasterDuke does anyone else think it would make sense for synopsebot to wait until it sees `review: github.com/*` before it prints its stuff? 19:27
geekosaur with a timeout in case what it thought was a commit announcement wasn't? 19:56
timotimo yeah, i wanted that feature in the past 20:00
Geth nqp: b6d98eac3f | (Samantha McVey)++ | 2 files
[moar] Add encodeconf and decodeconf ops

These ops by default decode strictly with windows-1252 and windows-1251 codetable specification, throwing on invalid input or output (as is done now for utf8 encoding). They can be set to decode/encode permissively as well, which will decode/encode codepoints that are unmapped yet still fit within one byte.		 20:04
travis-ci NQP build failed. Samantha McVey '[moar] Add encodeconf and decodeconf ops 20:20
travis-ci.org/perl6/nqp/builds/345731140 github.com/perl6/nqp/compare/fe7ca...d98eac3f0d
pmurias samcv: nqp::encode(...) is now just an alias for nqp::encodeconf(..., 1) 20:22
?
samcv no
well it maybe will be 20:23
but yeah nqp::encode is like doing nqp::encodeconf(.., 1) sorry if i wasn't clear 20:24
github.com/MoarVM/MoarVM/commit/31...07d4c2068b
Kaiepi i finally figured out what was causing me to require gmake to install rakudo from a while back, when normally bsd make works fine 20:27
passing --compiler=gcc to moarvm makes it generate a gnu makefile, but doesn't check whether make actually is gmake 20:29
so trying to build it on *bsd (apart from osx) will fail 20:30
but i'm wondering what the distinction between --compiler and --cc is meant to be, since --cc=gcc doesn't screw with the generated makefile 20:32
is --cc just meant to be used for the path to the compiler if it's different from where where the one the configure script chooses is? 20:34
pmurias samcv: any need to keep the old ops when rakudo (and other backends) updated to the new ones? 20:38
samcv pmurias: no not really
yoleaux 20:36Z <Zoffix> samcv: does Unicode have some sort of "non-break" property? In .words and with \S/\s we split on \x00A0\x2007\x202F chars because they're Zs and don't split on \xFEFF because it ain't. But when parsing quotewords we have code explitly not splitting on "\x00A0\x2007\x202F\xFEFF" and I don't know why these 4 chars were chosen to be special exactly (other than having "non-break" in the name)
tbrowder samcv: try modifying yr .travis.yml as it is now in the master branc. 21:18
*branch. that should get yr jvm job farther along. 21:19
Geth nqp: 24f90977a1 | (Samantha McVey)++ | 2 files
Only run the new decode/encode conf tests on moarvm for now
tbrowder for 21:20
.tell samcv try modifying yr .travis.yml as in the current nqp master branch. it will help push yr jvm job along in NQP#425 22:12
yoleaux tbrowder: I'll pass your message to samcv.
synopsebot NQP#425 [open]: github.com/perl6/nqp/pull/425 [moar] Add encodeconf and decodeconf ops
Geth rakudo: MasterDuke17++ created pull request #1563:
Whitespace fixes for Rational.pm		 22:37
[Coke] 23:13

Powered by Cro and the RakuĀ® Programming Language.

Please report any issues / comments / feature requests as an issue on App::Raku::Log.

Thank you!