Perl 6 language and compiler development | Logs at colabti.org/irclogger/irclogger_log/perl6-dev | For toolchain/installation stuff see #perl6-toolchain | For MoarVM see #moarvm
Set by Zoffix on 27 July 2018.
00:00 ufobat left 00:22 cognominal joined, p6bannerbot sets mode: +v cognominal 00:31 lucasb left
Kaiepi not to nag but is github.com/rakudo/rakudo/pull/2475 ready to merge? i need one of the changes i needed to make it possible to implement so i can finish implementing getsockopt/setsockopt support 00:36
s/needed to/needed that/ 00:44
00:51 cognomin_ joined 00:52 p6bannerbot sets mode: +v cognomin_ 00:54 cognominal left 00:56 cognomin_ left 01:10 cognominal joined 01:11 p6bannerbot sets mode: +v cognominal 01:15 MasterDuke joined, p6bannerbot sets mode: +v MasterDuke 01:16 MasterDuke left, MasterDuke joined, herbert.freenode.net sets mode: +v MasterDuke, p6bannerbot sets mode: +v MasterDuke 01:20 MasterDuke left
timotimo i haven't looked too closely, but i'd say go ahead 01:51
02:42 cognominal left 02:51 cognominal joined, p6bannerbot sets mode: +v cognominal 03:05 cognominal left 03:09 jkramer left 03:15 jkramer joined 03:16 p6bannerbot sets mode: +v jkramer 03:48 cognominal joined, p6bannerbot sets mode: +v cognominal 04:03 llfourn joined, p6bannerbot sets mode: +v llfourn 04:07 nine joined 04:08 p6bannerbot sets mode: +v nine, camelia joined 04:09 p6bannerbot sets mode: +v camelia 04:10 cognomin_ joined 04:11 p6bannerbot sets mode: +v cognomin_ 04:13 cognominal left 04:14 cognomin_ left 04:42 cognominal joined, p6bannerbot sets mode: +v cognominal 04:50 cognominal left 04:51 cognominal joined 04:52 p6bannerbot sets mode: +v cognominal 04:55 cognominal left 05:18 llfourn left, llfourn joined 05:19 p6bannerbot sets mode: +v llfourn 05:27 cognominal joined 05:28 p6bannerbot sets mode: +v cognominal
Geth nqp: bb35cf071c | (Ben Davies)++ | src/vm/jvm/runtime/org/perl6/nqp/io/AsyncServerSocketHandle.java
Reimplement async socket introspection
 05:58
nqp: 94d35fa12d | (Aleks-Daniel Jakimenko-Aleksejev)++ (committed using GitHub Web editor) | src/vm/jvm/runtime/org/perl6/nqp/io/AsyncServerSocketHandle.java
Merge pull request #513 from Kaiepi/async-introspection

Reimplement async socket introspection
nqp: 75ecd00dc5 | (Aleks-Daniel Jakimenko-Aleksejev)++ | tools/build/MOAR_REVISION
[MoarVM Bump] Brings 5 commits

MoarVM bump brought: github.com/MoarVM/MoarVM/compare/2...g40400fdcd 40400fdcd Merge pull request #991 from Kaiepi/async-introspection 437d5b010 Fix big endian (read|write)u?int of < qword on big endian systems 35ea95818 Fix bytestream ops always converting to big endian if a switch is necessary 0d5f389c3 Don't Dump "Facts" in "Before" Part Of Spesh Log 1dc540f5f Reimplement async socket introspection		 06:00
¦ nqp: version bump brought these changes: github.com/MoarVM/MoarVM/compare/2...g40400fdcd
rakudo: 2fd90b1ecc | (Aleks-Daniel Jakimenko-Aleksejev)++ | tools/build/NQP_REVISION
[NQP Bump] Brings 12 commits

NQP bump brought: github.com/perl6/nqp/compare/2018....g75ecd00dc 75ecd00dc [MoarVM Bump] Brings 5 commits 94d35fa12 Merge pull request #513 from Kaiepi/async-introspection f8e87c58c Add constants to list of constants a50a2a6ee Fix label ... (8 more lines)
06:00 cognominal left
rakudo: version bump brought these changes: github.com/perl6/nqp/compare/2018....g75ecd00dc
a4db91394f | (Ben Davies)++ | src/core/IO/Socket/Async.pm6

This gets the file descriptor of the socket. This was originally implemented before release 2018.10, but was reverted due to bugs.
rakudo: 212193cb4c | (Aleks-Daniel Jakimenko-Aleksejev)++ (committed using GitHub Web editor) | src/core/IO/Socket/Async.pm6
Merge pull request #2475 from Kaiepi/async-introspection

Reimplement .native-descriptor for async sockets
06:01 nebuchadnezzar left
AlexDaniel Kaiepi: \o/
I'm under the weather, can't tell how clearly I'm able to think, maybe that was a mistake… :) 06:02
but I guess tomorrow I'll be able to run blin and we'll see 06:03
07:44 ufobat joined 07:45 p6bannerbot sets mode: +v ufobat 07:53 nebuchadnezzar joined 07:54 p6bannerbot sets mode: +v nebuchadnezzar 07:58 cognominal joined 07:59 p6bannerbot sets mode: +v cognominal 08:01 xlat joined 08:02 p6bannerbot sets mode: +v xlat 08:12 robertle joined, ufobat left 08:13 p6bannerbot sets mode: +v robertle
lizmat Files=1256, Tests=81568, 342 wallclock secs (15.12 usr 5.50 sys + 2424.70 cusr 216.21 csys = 2661.53 CPU) 08:28
08:29 cognomin_ joined 08:30 p6bannerbot sets mode: +v cognomin_ 08:32 cognominal left 08:33 cognomin_ left
Geth roast: 731a3f211a | (Elizabeth Mattijsen)++ | S03-buf/read-int.t
Add Buf.read-int out-of-bounds tests

Resulting in MoarVM #1021, as there is a bug in the big-endian case.		 09:15
09:23 ufobat joined 09:24 p6bannerbot sets mode: +v ufobat 10:31 cognominal joined 10:32 p6bannerbot sets mode: +v cognominal 10:35 lucasb joined, p6bannerbot sets mode: +v lucasb
Geth roast: 848473a8cb | (Elizabeth Mattijsen)++ | S03-buf/read-int.t
Add tests for Blob/Buf[uint8]

And don't separate blob / buf anymore: we're not going to test 'write-int' here, so no need to separate them.		 10:59
releasable6 Next release in ≈3 days and ≈7 hours. 4 blockers. Please log your changes in the ChangeLog: github.com/rakudo/rakudo/wiki/ChangeLog-Draft 11:00
11:02 cognomin_ joined 11:03 p6bannerbot sets mode: +v cognomin_ 11:05 cognominal left 11:07 cognomin_ left
Geth roast: 736e152765 | (Elizabeth Mattijsen)++ | S03-buf/read-int.t
Fix comment
 11:24
roast: 3985f5daa0 | (Elizabeth Mattijsen)++ | S03-buf/read-num.t
Add basic read-num tests

Basic because I'm not sure how to construct a num32 / num64 manually to be able to check correctness. So for now, just check for not 0. Also
  because of MoarVM#1020, a lot of the read-num32 tests fail.
lizmat .tell AlexDaniel I think I have logged all of my commits 11:40
yoleaux lizmat: I'll pass your message to AlexDaniel.
12:27 cognominal joined, p6bannerbot sets mode: +v cognominal 12:35 cognominal left 12:36 cognominal joined 12:37 p6bannerbot sets mode: +v cognominal 12:40 cognominal left 13:10 cognominal joined 13:11 p6bannerbot sets mode: +v cognominal
Geth rakudo: f66861dc44 | (Elizabeth Mattijsen)++ | src/core/Buf.pm6
Workaround signedness issue / add support for 128 bit

Currently reading nqp::readint8/16/32 does not expand the sign bit. Work around that for now for the 2018.12 release. Also add support for read-int128/uint128, which was pretty trivial thanks to the miracle of bigints being builtin!		 13:26
jnthn releasable6: status 13:38
releasable6 jnthn, Next release in ≈3 days and ≈5 hours. 4 blockers. 13 out of 42 commits logged
jnthn, Details: gist.github.com/0c8e46b63e7b8dfb28...23ee49202a
13:41 ufobat left
Geth rakudo: 3c2cfb224f | (Jonathan Worthington)++ | src/Perl6/Metamodel/EnumHOW.nqp
Allow an enum to have private methods mixed in

Resolves #2535.		 14:11
roast: 93c620ff26 | (Jonathan Worthington)++ | S12-enums/thorough.t
Test mixing a private method into an enum value

Covers github.com/rakudo/rakudo/issues/2535
rakudo: f027931304 | (Elizabeth Mattijsen)++ | src/core/Buf.pm6
Make sure read-int128/uint128 take endianness into account
 14:13
roast: efbcecac79 | (Elizabeth Mattijsen)++ | S03-buf/read-int.t
Add tests for read-int128/uint28

Also just test the first two and last two positions in the buffer. Assume that all others will work out ok.		 14:15
lizmat afk for a bit& 14:17
14:26 cognominal left 14:27 cognominal joined 14:28 p6bannerbot sets mode: +v cognominal 14:34 cognomin_ joined 14:35 p6bannerbot sets mode: +v cognomin_ 14:37 cognominal left 14:51 cognomin_ left, cognominal joined 14:52 p6bannerbot sets mode: +v cognominal 14:56 cognominal left, pmurias joined, p6bannerbot sets mode: +v pmurias
pmurias does it make sense to add a :global parameter to EVAL so that one could eval code with a custom global namespace? 14:57
also maybe allow overiding the $?FILES it compiles code as 14:58
lizmat pmurias: have you looked at the implementation of Cool.EVAL? :-) 14:59
or are you not talking about just context >
?
|Tux| Rakudo version 2018.11-44-gf02793130 - MoarVM version 2018.11-30-g8e078a2fe
csv-ip5xs0.914 - 1.013
csv-ip5xs-206.935 - 6.950
csv-parser21.456 - 21.485
csv-test-xs-200.422 - 0.428
test7.760 - 7.816
test-t1.714 - 1.744
test-t --race0.761 - 0.777
test-t-2030.372 - 30.924
test-t-20 --race9.294 - 9.659
 15:00
15:00 pmurias left
lizmat pmurias: ah, I see now... I guess it's up to jnthn to say if its ok to add :$global = GLOBAL and then do :$global instead of :global(GLOBAL)
15:03 SourceBaby joined, ChanServ sets mode: +v SourceBaby 15:04 p6bannerbot sets mode: +v SourceBaby, pmurias joined, p6bannerbot sets mode: +v pmurias
pmurias lizmat: I could just copy & paste EVAL from the CORE, but it seems like that might be something that's generally useful 15:05
lizmat jnthn: ^^^ 15:09
15:10 pmurias left
lizmat I recall that jnthn wanted to get rid of the :context parameter, so I have no idea how he would feel about adding another parameter like that in the same vein 15:10
15:10 pmurias joined, p6bannerbot sets mode: +v pmurias
Geth rakudo: lizmat self-assigned Pythonic native arrays github.com/rakudo/rakudo/issues/2525
b781f8ccd4 | (Elizabeth Mattijsen)++ | src/core/Buf.pm6

native-endian -> NativeEndian
   little-endian -> LittleEndian
   big-endian -> BigEndian
As suggested by jnthn.		 15:12
15:12 pmurias left 15:24 cognominal joined, p6bannerbot sets mode: +v cognominal 15:28 cognominal left 15:52 cognominal joined 15:53 p6bannerbot sets mode: +v cognominal
AlexDaniel . 16:00
yoleaux 11:40Z <lizmat> AlexDaniel: I think I have logged all of my commits
AlexDaniel lizmat: thank you!
16:11 llfourn left 16:16 llfourn joined 16:17 p6bannerbot sets mode: +v llfourn 16:21 robertle left
roast: 30d9c375a4 | (Elizabeth Mattijsen)++ | 2 files
Adjust to new naming of the Endian enum
 16:38
16:47 Tux__ joined, |Tux| left, p6bannerbot sets mode: +v Tux__ 17:38 cognominal left 17:39 cognominal joined, p6bannerbot sets mode: +v cognominal 17:45 lucasb left
jnthn m: class C { method TWEAK() { Failure.new } }; C.new 17:54
camelia Failed
in block <unit> at <tmp> line 1
Geth rakudo: dd2af90cca | (Elizabeth Mattijsen)++ | 2 files
Add check for negative indexes to native array AT-POS/ASSIGN-POS

In order to fix R#2525. This *slows down* element access to native arrays by about 60%. So perhaps we should fix this at the VM level rather than in the setting.
Also add missing return type indicators.		 17:58
18:06 lucasb joined, p6bannerbot sets mode: +v lucasb 18:10 cognominal left 18:11 cognominal joined 18:12 p6bannerbot sets mode: +v cognominal 18:15 cognominal left 18:51 cognominal joined 18:52 p6bannerbot sets mode: +v cognominal 18:55 pmurias joined, p6bannerbot sets mode: +v pmurias
pmurias jnthn: what do you think about adding a :global parameter to &EVAL? 19:05
19:25 cognominal left
Kaiepi what would the global parameter do? 19:37
lizmat Kaiepi: determine which namespace to use as "GLOBAL::" 19:40
Kaiepi ah
Geth rakudo: 2a761ca7af | (Elizabeth Mattijsen)++ | src/core/Kernel.pm6
Introducing Kernel.Endian

Returns the appropriate Endian enum for the current kernel architecture, either LittleEndian or BigEndian.		 19:56
roast: 0af3a3be2d | (Elizabeth Mattijsen)++ | 2 files
Use the new Kernel.Endian feature
 19:58
jnthn pmurias: It's probably OK; would it just take a Stash? 20:12
Geth rakudo: 543219c927 | (Elizabeth Mattijsen)++ | src/core/Buf.pm6
Remove read-uint128 hack now that we have Kernel.Endian
 20:16
lizmat afk for a bit& 20:17
lucasb was Endianness considered as the name? 20:19
too long? :) 20:20
"ness"... 4 chars longer
jnthn Not sure that it adds anything 20:21
Well, other than 4 chars :)
lizmat: I'd probably make it Kernel.endian, since I think all the other methods on there are lowercase
Well, and methods are in general :)
lucasb Can you believe ruby devs prefered 'RubyVM::AbstractSyntaxTree' over 'RubyVM::AST'? :) 20:22
because longer names discourages casual usage
jnthn I couldn't be ast with typing that all the time... :P 20:23
lucasb (Just to be clear, my name sugggestion was for the enum: enum Endianness < NativeEndian LittleEndian BigEndian >) 20:28
(or... EndianType) 20:29
nine is glad he didn't pick 3 as the value for BigEndian 20:32
pmurias jnthn: Stash seems good enough 20:37
jnthn: and what about something that sets $?FILES? 20:43
20:54 pmurias left, pmurias joined, p6bannerbot sets mode: +v pmurias 21:09 cognominal joined 21:10 p6bannerbot sets mode: +v cognominal 21:19 cognominal left 21:20 cognominal joined, p6bannerbot sets mode: +v cognominal 21:24 cognominal left 21:53 xlat left 21:57 cognominal joined 21:58 p6bannerbot sets mode: +v cognominal 22:00 lucasb_ joined 22:01 p6bannerbot sets mode: +v lucasb_ 22:04 lizmat left 22:15 lucasb_ left 22:25 lizmat joined, p6bannerbot sets mode: +v lizmat, hoelzro joined 22:26 Ulti_ joined, p6bannerbot sets mode: +v hoelzro 22:27 p6bannerbot sets mode: +v Ulti_
Geth rakudo: af43b1592b | (Elizabeth Mattijsen)++ | 2 files
Kernel.Endian to be called .endian
 22:28
roast: 7e29de3229 | (Elizabeth Mattijsen)++ | 2 files
Kernel.Endian to be called .endian
 22:29
22:29 Ulti left 22:30 cognominal left
pmurias lucasb: doesn't RubyVM::AbstractSyntaxTree have such a lame name because it's something for mostly debugging internal use rather than a proper AST? 22:32
Geth nqp: 7e9594a968 | (Elizabeth Mattijsen)++ | tools/build/MOAR_REVISION
Bump MoarVM to get read-int fixes of nine++
 22:40
¦ nqp: version bump brought these changes: github.com/MoarVM/MoarVM/compare/2...2-gc32000b
nqp/master: 5 commits pushed by (Paweł Murias)++ 22:43
rakudo: fd7f8b6d94 | (Elizabeth Mattijsen)++ | tools/build/NQP_REVISION
Bump NQP to get nine++'s read-int fixes
 22:56
rakudo: version bump brought these changes: github.com/perl6/nqp/compare/2018....6-g7e9594a
7718271334 | (Elizabeth Mattijsen)++ | src/core/Buf.pm6
23:09 cognominal joined 23:10 p6bannerbot sets mode: +v cognominal
Geth rakudo: 06156a7ca8 | (Elizabeth Mattijsen)++ | t/spectest.data
Add blob8/buf8.read-intX tests

They now all pass, nine++		 23:14
23:36 pmurias left 23:41 cognominal left 23:42 cognominal joined, p6bannerbot sets mode: +v cognominal 23:46 cognominal left, lucasb left 23:48 cognominal joined 23:49 p6bannerbot sets mode: +v cognominal
AlexDaniel nine: thank you! 23:50

Powered by Cro and the Raku® Programming Language.

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

Thank you!