| travis-ci | Rakudo build passed. Jonathan Worthington 'Fix bad assumption about methods being closures. | 00:09 | |
| travis-ci.org/rakudo/rakudo/builds/264549001 github.com/rakudo/rakudo/compare/2...1cb3f5fe8f | |||
| [Coke] | 6hhh | 00:56 | |
| mst | [Coke]: yes? | ||
| timotimo | what's time to the third power useful for in physics? | 00:57 | |
| geekosaur | usualy more useful as a reciprocal | 01:00 | |
| e.g. s^-1 velocity, s^-2 acceleration, and s^-3 (change in acceleration) comes up in various contexts, notably rocketry | 01:01 | ||
| (s there being seconds) | |||
| [Coke] | mst: missend. | 01:02 | |
| geekosaur | (all of those having a distance, typically meters, as numerator) | ||
| mst | [Coke]: cat on keyboard was my first thought :D | ||
| [Coke] | I use ssh to a tmux session; if it hangs, I will type out a few characters and if nothing shows up, I'll disconnect the ssh session. sometimes it's up even though it's not responding to me. | 01:03 | |
| timotimo | i get that sometimes | 01:11 | |
| lizmat | Files=1217, Tests=67433, 271 wallclock secs (13.89 usr 5.26 sys + 1810.61 cusr 178.11 csys = 2007.87 CPU) | 07:16 | |
| yoleaux | 02:53Z <zengargoyle> lizmat: did you do that talk thing from a while back? | ||
| lizmat | .tell if you're referring to my "talk" at the TPCiA: no. This was a completely unscripted Q&A session with most of the answers provided by people in the audience | 07:17 | |
| yoleaux | lizmat: I'll pass your message to if. | ||
| lizmat | .tell zengargoyle if you're referring to my "talk" at the TPCiA: no. This was a completely unscripted Q&A session with most of the answers provided by people in the audience | 07:18 | |
| yoleaux | lizmat: I'll pass your message to zengargoyle. | ||
| zorin-os | can anyone help me statically compile perl with -Drelocatableinc and with thread support | 07:41 | |
| lizmat | zorin-os: are you talking about Rakudo Perl 6 or Perl 5 ? | 07:49 | |
| indeed: wrong channel more than once :-( | 07:50 | ||
| zorin-os | depends is perl 6 is compatible with perl5 scripts | 07:54 | |
| are you able to determine how a perl was compiled | 07:59 | ||
| nine | zorin-os: Perl 6 is a sister language to Perl 5. You will get better help over in #perl though perlbrew.pl/ might be all you need | 08:13 | |
| yoleaux | 00:22Z <AlexDaniel> nine: Can you take a look at RT #131898? It looks scary, I'd say a blocker. Please tell me what you think. | ||
| synopsebot6 | Link: rt.perl.org/rt3/Public/Bug/Display...?id=131898 | ||
| zorin-os | i cant use anything that modifies the environment as it may collide with an existing perlbrew installation | 08:14 | |
| nine | .tell AlexDaniel yes, that's definitely a blocker. Will have a look (though probably only tomorrow) | ||
| yoleaux | nine: I'll pass your message to AlexDaniel. | ||
| zorin-os | for example, chroot ./perl /bin/perl /script.pl AND the /lib/perl5 directory exists (outside of the chroot it will be ./perl/lib/perl5/) and succesfull execution of the script means the script is then fully portable (except for its shebang) | 08:21 | |
| Geth | roast: fd43387d40 | (Jonathan Worthington)++ | 2 files Add new atomics tests. |
09:22 | |
| roast: 456b87ddbb | (Jonathan Worthington)++ | S17-promise/allof.t Correct test that tried to mis-use CAS. The cheating one didn't have to care for how the world actually works. :-) |
09:23 | ||
| rakudo/atomics: de3124522f | (Jonathan Worthington)++ | 2 files Add Scalar atomic load/store, plus full barrier. For now as `Atomic.fetch`, `Atomic.assign`, and `full-barrier`, but they may well become functions instead. |
09:24 | ||
| rakudo/atomics: 7ab6eb6ac4 | (Jonathan Worthington)++ | t/spectest.data Run S17-lowlevel/atomic.t. |
|||
| nqp/atomics: fa974bc795 | (Jonathan Worthington)++ | src/vm/moar/QAST/QASTOperationsMAST.nqp Add constant mapping atomic integer type. |
09:37 | ||
| nqp/atomics: 59fd391f29 | (Jonathan Worthington)++ | src/vm/moar/QAST/QASTOperationsMAST.nqp Add constant mapping atomic integer type. |
09:41 | ||
| pmurias | jnthn: do we want to sink for $foo.?method $foo.+method etc.? | 10:10 | |
| jnthn: in those cases we set a nosink and then wrap them in a hllize op ignoring the nosink | 10:12 | ||
| jnthn | Yes | ||
| Well, for .? for sure | 10:13 | ||
| I guess the .* and .+ return a list, which won't do anything on sink | |||
| pmurias | seems like we shouldn't be hllizing .+ and .* | 10:16 | |
| jnthn | It's probably pointless, indeed. | ||
| Geth | roast: 45b159756f | (Elizabeth Mattijsen)++ | S03-junctions/autothreading.t One should use Junction.perl instead of .Str for testing |
10:54 | |
| roast/6.c-errata: ce3e1ec466 | (Elizabeth Mattijsen)++ | S03-junctions/autothreading.t One should use Junction.perl instead of .Str for testing |
10:56 | ||
| lizmat | that should take care of the autothreading.t fails | ||
| the only fails I now see on spectest is t/spec/S32-io/IO-Socket-INET.t (12-17) | 11:06 | ||
| is that just me or just MacOS or what? | |||
| jnthn | No, I see it on Linux too. I thought they were about fallout from is default etc.? | ||
| lizmat | ah! | 11:10 | |
| ok, /me investigates | |||
| jnthn | lunch & | 11:11 | |
| dogbert17_ | lizmat: chances are that you have to revert a revert of a revert :) | 11:14 | |
| lizmat | yeah, doing that right now | ||
| yup, reverting is default fixes that | |||
| dogbert17_ | guess it was 7fdbb49ccc4b | 11:15 | |
| Geth | rakudo/nom: f1464e22f0 | (Elizabeth Mattijsen)++ | 5 files Revert "Revert "Revert potentially confusing "is default" changes"" This reverts commit 7fdbb49ccc4b7edc22dd561e3f42a960ba934f95. In light of release being near, feels safer to do without "is default" for now and investigate after release. |
||
| ilmari | revertception | ||
| lizmat | and t/spec/S32-io/IO-Socket-INET.t is also clean | ||
| dogbert17_ | yay | ||
| m: say "Revert " x 3 | 11:17 | ||
| camelia | Revert Revert Revert ? | ||
| Zoffix | FWIW, when is default went in, there weren't failures in S32-io/IO-Socket-INET.t. It likely merely exposes another bug that was introed sometime between is default going in and like last week | 12:14 | |
| AlexDaniel | .tell nine OK, thank you! | 12:28 | |
| yoleaux | 08:14Z <nine> AlexDaniel: yes, that's definitely a blocker. Will have a look (though probably only tomorrow) | ||
| AlexDaniel: I'll pass your message to nine. | |||
| Geth | rakudo/atomics: 5ea8596595 | (Jonathan Worthington)++ | 2 files Add atomicint type. Which will always be of an appropriate size for use with atomic ops on integers. What that size will be is platform specific. The integer atomic ops will work on any native int value that happens to be the current platform's atomic size, but we'll do the documentation so as to strongly encourage using atomicint instead to achieve portability. |
12:29 | |
| rakudo/atomics: 79c62c0b61 | (Jonathan Worthington)++ | src/core/Atomic.pm Add various atomic integer operations. Again, these may become functions rather than methods on `Atomic` in the future. |
|||
| roast: 16e93354f3 | (Jonathan Worthington)++ | S17-lowlevel/atomic.t Basic tests for atomic integer operations. |
|||
| AlexDaniel | does anybody know why design docs are down? design.perl6.org/S01.html | 12:56 | |
| Zoffix | hack upgrade didn't start something it should have?> | ||
| AlexDaniel | “www.p6c.org hosts many Perl 6-related websites, including the perl6.org main site, doc.perl6.org, modules.perl6.org and the Perl 6 design documents.” | 12:59 | |
| so it's not hack, right? | |||
| Zoffix: “The following users have root/sudo access, and can potentially help you:” lists your name :) ping | 13:00 | ||
| Geth | rakudo/nom: 2b8115f0e9 | pmurias++ | src/Perl6/Actions.nqp Stop hllizing and sinking the result of .+, .* and .= method calls |
||
| roast: da9a389b28 | pmurias++ | S04-statements/sink.t Test that .= in both forms doesn't sink |
13:01 | ||
| perlpilot | AlexDaniel: read the bottom of design.perl6.org/ for more clues. | ||
| Geth | rakudo/atomics: 6eb8f7e77c | (Jonathan Worthington)++ | src/core/Atomic.pm Implement loop form of cas for integers. |
||
| AlexDaniel | ah, I see | ||
| perlpilot | maybe the cron job is broken? | ||
| pmurias | jnthn: why are the Atomic operations methods on the type object? | 13:02 | |
| Geth | roast: 06bdc9b1ea | (Jonathan Worthington)++ | S17-lowlevel/cas-loop-int.t Add cas loop form tests on lexical integers. |
13:03 | |
| AlexDaniel | I don't have access to any of this stuff, so can't really do anything | ||
| perlpilot | I have sudo, but I can't easily login from work | 13:05 | |
| jnthn | pmurias: What else would they be on? :) | 13:11 | |
| It's just a way to namespace/group them in the interim so I don't have to figure out function names for them :) | 13:12 | ||
| pmurias | ok | ||
| Zoffix | AlexDaniel: maybe in an hour. I'm occupied at work ATM | 13:16 | |
| Ok. Back. | 13:25 | ||
| upgrade Perl is missing a module | 13:27 | ||
| Well, more like "missing" :) CGI.pm | |||
| Geth | rakudo/atomics: 413b5331b0 | (Jonathan Worthington)++ | src/core/Atomic.pm Add cas operation on integers. |
13:28 | |
| Zoffix | AlexDaniel: There. Fixed forever: design.perl6.org/S01.html | 14:03 | |
| AlexDaniel | Zoffix: thanks! | 14:04 | |
| ugexe | m: my $x but Buf; | 15:17 | |
| camelia | P6opaque: missing attribute protocol in compose of Any+{Buf}? in block <unit> at <tmp> line 1?? | ||
| ugexe | m: my $x = 1 but Buf; | ||
| camelia | Cannot have attributes on an array representation? in block <unit> at <tmp> line 1?? | ||
| ugexe | two LTAs | ||
| jnthn | Probably should detect roles that demand a specific representation and just complain earlier | 15:18 | |
| Geth | rakudo/js: cbde083fa2 | pmurias++ | src/vm/js/perl6-runtime/runtime.js [js] When the scalar has a Mu type don't type check when putting stuff into it |
15:41 | |
| rakudo/js: 8c888adf92 | pmurias++ | 2 files [js] Stop nqp::p6sink from decontainerizing |
|||
| stmuk | anyone here use rakudobrew on windows? | 15:43 | |
| TimToady slept an amazingly large subset of the last twelve hourse... | 15:48 | ||
| not to mention hours | |||
| Zoffix | No, I just build manually on windows. | 15:52 | |
| Speaking of which... Just tried to install rakudo on co-workers Win32-bit box... Crashed and burned on takeoff :( | |||
| (I didn't really look at the error, something about failing to make paths) | 15:53 | ||
| stmuk | different drive to C:? | ||
| </guess> | |||
| Zoffix | I was installing to C:\rakudo. | ||
| cloning the repo there and running... | 15:54 | ||
| huggable: rakudo win | |||
| huggable | Zoffix, perl Configure.pl --gen-moar --gen-nqp --backends=moar & gmake & gmake test & gmake install | ||
| Zoffix | ^ that | ||
| ugexe | well you might need to use the Visual Studio Command Tools cmd | ||
| and use nmake | |||
| thats how I build on windows anyway | |||
| Zoffix | The above works fine on my home box. | ||
| ugexe | your home box likely has its PATH set differently than your friend | 15:55 | |
| Zoffix | Ah | ||
| lizmat is glad TimToady was able to get such a large subset | 15:56 | ||
| ugexe | some cmd prompts carry different PATHs with them | ||
| Zoffix | Well, it was mostly a "maybe we'll need it later" install. Gonna debug it when/if we'll actually need it :) | ||
| ugexe | e.g. Visual Studio Command Tools cmd has the visual studio toolchain setup in PATH | ||
| Geth | rakudo/nom: c68762f46a | (Aleks-Daniel Jakimenko-Aleksejev)++ | README.md “doc/ChangeLog” ? “docs/ChangeLog” |
15:57 | |
| stmuk | C:\Program Files\Microsoft SDKs\Windows\v7.1\Bin\SetEnv.cmd" /x64 | 15:59 | |
| - " | |||
| Zoffix | AlexDaniel++ # I'm so glad I don't need to release anything this weekend. Burnt out to ashes and they're doing maintenance in my apartment this week | 16:00 | |
| Geth | rakudo/atomics: 61e1f4d5f5 | (Jonathan Worthington)++ | 2 files Fix native shaped array index to return l-value. Fixes `my @a[2;2]; @a[1;1]++`. |
16:09 | |
| rakudo/atomics: 9d5e929504 | (Jonathan Worthington)++ | t/spectest.data Run spectests for CAS on atomic ints. |
|||
| roast: 06230a444d | (Jonathan Worthington)++ | S17-lowlevel/cas-int.t Add tests for CAS on native integers. Covering all the places a native integer might live. |
|||
| Zoffix | ZofBot: . | 16:48 | |
| ZofBot | Zoffix, html?id=130900 ah, very well | ||
| pmurias | is it expected for an installed perl6-m to try to load stuff from a subdirectory of where it's running from? | 16:49 | |
| my perl6-m doesn't seem to work when run from a different rakudo checkout | |||
| if perl6-m is loading random stuff from disk rather it seems like a fairly serious security issue | 16:52 | ||
| tadzik | did you 'make install'? | 16:53 | |
| pmurias | my perl6-m is installed by 'make install' | 16:57 | |
| samcv | AlexDaniel, found another ignoremark issue as i was adding more tests into nqp before release :) | 17:21 | |
| AlexDaniel | samcv++ | ||
| samcv | but we will be good for release | ||
| all our ignoremark tests test for lowercase values | 17:22 | ||
| heh | |||
| Geth | nqp: 72be3bbad4 | (Samantha McVey)++ | tools/build/MOAR_REVISION Bump MoarVM for ignoremark fixes Fix uppercase needles not always matching when using ignoremark |
17:36 | |
| nqp: version bump brought these changes: github.com/MoarVM/MoarVM/compare/2...-g633cb35a d03c4023ab | (Samantha McVey)++ | src/vm/moar/QAST/QASTOperationsMAST.nqp Op already existed, but make it accessible for development and testing. |
|||
| nqp: 6fac0036eb | (Samantha McVey)++ | t/moar/10-eqatim.t Add tests for nqp::eqatim op on MoarVM |
17:39 | ||
| samcv | ok it's fixed in nqp master and moar bleed now. will finish when i get back from dentist | 17:42 | |
| Zoffix | NeuralAnomaly: status | 19:27 | |
| NeuralAnomaly | Zoffix, [?] Next release will be in 3 days and 9 hours. Since last release, there are 50 new still-open tickets (3 unreviewed and 2 blockers) and 162 unreviewed commits. See perl6.fail/release/stats for details | ||
| Zoffix | .in 3d fix xtest on doc | ||
| yoleaux | Zoffix: I'll remind you on 18 Aug 2017 19:27Z | ||
| [Coke] | Zoffix++ taking onthe releases for so long | 20:53 | |
| samcv | i'm encountering an issue with Test | 21:00 | |
| here's the file: gist.github.com/samcv/e80e88ff1ff8...ae4d866337 | 21:02 | ||
| when i run it i get a failure, but when i run it manually i get the correct result | |||
| all the things expecting True get the correct result but the one wanting False always gets True | 21:03 | ||
| not ok 7 - so('?' ~~ /:m d /), False — Ignoremark: ? doesn't match d — TOPBAR is not a mark | |||
| so running that manually shows the correct result. but maybe i made a mistake on line 30? | |||
| though when i try using is-deeply, that returns Match.new(list => (), made => Any, pos => 0, hash => Map.new(()), orig => "?", from => 0) | 21:04 | ||
| which evaluates as True even through a stringified version is Bool false | |||
| so maybe that's what's causing it to fail? but it gives the correct result when i run it with so manually | 21:05 | ||
| yeah it seems to always show True... it works fine if the code is typed out and compiled but if it's programmic it acts differently | 21:08 | ||
| oh maybe it's how i'm accessing it? when i assign it to new variables instead of accessig it from the array using [0] and [1] it works | 21:10 | ||
| Geth | rakudo/nom: a3b95749f3 | (Samantha McVey)++ | tools/build/NQP_REVISION Bump NQP to get ignoremark fixes |
21:32 | |
| ¦ rakudo/nom: version bump brought these changes: github.com/perl6/nqp/compare/2017....g6fac0036e | |||
| roast: 810e218c83 | (Samantha McVey)++ | S05-modifier/ignoremark.t Improve ignoremark tests to check different cases Make sure different cases do *not* match when just ignorecase is supplied. Make a function to automate testing of this. |
21:33 | ||
| MasterDuke | samcv: should that comment say "just ignore*mark*"? | 21:34 | |
| samcv | yes it should X| | 21:35 | |
| added a comment on github | |||
| Geth | nqp: 99abffd7f7 | MasterDuke17++ | 2 files Bump Moar to get coveragecontrol op and map it nqp::coveragecontrol allows for turning line-coverage logging on and off. It is enabled with MVM_COVERAGE_CONTROL=1, which also turns off de-duping of lines logged. MVM_COVERAGE_CONTROL=2 turns off de-duping, but doesn't require an nqp::coveragecontrol(1) to start the logging. |
22:11 | |
| travis-ci | NQP build failed. MasterDuke17 'Bump Moar to get coveragecontrol op and map it | 22:40 | |
| travis-ci.org/perl6/nqp/builds/264930802 github.com/perl6/nqp/compare/6fac0...abffd7f74f | |||
| [Coke] | I have narrowed down my os x malformed-utf8 issue to: 51b63bf9f0d0594687908f0756774f29cde39c1f in rakudo, which is a bump to nqp, which brought in these commits: github.com/MoarVM/MoarVM/compare/2...5-g369c0c5 | 23:12 | |
| TimToady wonders what to do about tests that assume .parse returns Nil on failure rather than Failure... | 23:30 | ||
| TimToady has a patch to do the latter, y'see... | |||
| oh gee, one of the tests assumes it can stringify a Nil... | 23:34 | ||
| and then there's: is rt81136.parse('abc'), Nil | 23:40 | ||
| which seems rather overspecific when it's just testing for failure |