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 cognominal-p6 left 00:08 MasterDuke joined, p6bannerbot sets mode: +v MasterDuke 00:16 MasterDuke left, MasterDuke joined, herbert.freenode.net sets mode: +v MasterDuke, p6bannerbot sets mode: +v MasterDuke 00:33 leont left, travis-ci joined, p6bannerbot sets mode: +v travis-ci
travis-ci Rakudo build failed. Tom Browder 'tweak format and spelling' 00:33
travis-ci.org/rakudo/rakudo/builds/450346147 github.com/rakudo/rakudo/compare/b...59654014fc
00:33 travis-ci left
buggable [travis build above] ☠ Did not recognize some failures. Check results manually. 00:33
02:56 cognominal-p6 joined, p6bannerbot sets mode: +v cognominal-p6 03:09 ufobat_ joined 03:10 p6bannerbot sets mode: +v ufobat_ 03:13 ufobat___ left 03:19 cognominal-p6 left 05:10 lizmat left 05:26 lizmat joined, p6bannerbot sets mode: +v lizmat 05:40 fake_space_whale left 06:16 cognominal-p6 joined, p6bannerbot sets mode: +v cognominal-p6 06:19 cognominal-p6 left, cognominal-p6 joined 06:20 p6bannerbot sets mode: +v cognominal-p6
Geth rakudo: JJ++ created pull request #2472:
Changes variable name
07:42
rakudo: 7d4b57dea4 | (Juan Julián Merelo Guervós)++ (committed by Aleks-Daniel Jakimenko-Aleksejev) | src/core/Cool.pm6
Changes variable name (#2472)

To better reflect what it does.
08:52
09:21 lizmat left
[Tux] Rakudo version 2018.10-70-g7d4b57dea - MoarVM version 2018.10-74-g2fdde4a21
csv-ip5xs0.914 - 0.989
csv-ip5xs-207.246 - 7.370
csv-parser24.693 - 24.702
csv-test-xs-200.431 - 0.446
test7.976 - 8.238
test-t1.794 - 1.888
test-t --race0.883 - 0.911
test-t-2032.401 - 33.968
test-t-20 --race10.878 - 11.399
09:29
nine Can someone explain this block to me? github.com/perl6/nqp/blob/master/s...T.nqp#L865 09:36
We allocate $coerce-reg. We never generated code to write into that register, yet we use it as condition in the if_i/unless_i/whatever op. Then we release it again. But apparently it still works? 09:37
timotimo hm? we do call the generator function with $coerce_reg though? 10:02
nine Yes, but that just generates the if_i r123, ins234 bytecode 10:03
timotimo oh?
did you check yet if the block runs at all? 10:04
nine it does
timotimo oh, you mean the register is never written to? 10:05
nine Used to look like this: github.com/perl6/nqp/blob/2a1adeec...T.nqp#L842
I don't see anything that writes into $coerce-reg
And nothing that connects it in any way to @comp_ops[0].result_reg 10:06
timotimo it could have been copy-pasted from below
almost the entire block shows up a bit further below
oh, that also doesn't use the register? 10:07
- I applied the lookup of extra types in one place in the Regex compiler that 10:08
used to use the old lookup. However, I did not add coercion as I'm unsure
what code exercises that codepath and whether the coercion is needed at all.
nine You mean in loop compilation? Looks just as broken
timotimo perhaps that's what that's about?
nine The only explanation that'd make sense to me is that $coerce-reg just happens to be a re-used register that by accident still holds a useful value. 10:09
timotimo if this is about coercing smaller values to bigger values, i.e. i8 to i64, coercion doesn't actually change the truth/falsity of a value?
nine It is about that, yes.
timotimo i don't think we have ops that ever overwrite only the lower parts of a value 10:10
so storing a 0 will always zero out the entire register
if we ever change that, this code would probably asplode violently
or quietly
nine Note that we also determine the condition op by @comp_ops[0].result_kind, not by $coerce-kind
timotimo is this about using if_s on an int for example? 10:11
and expecting it to stringify the int?
nine I'm less confused about a non-coerced value working correctly. More about a fresh register luckily holding a sensible value
timotimo i wonder, though. what happens if you allocate a random number of registers of the same kind first
nine Nah, the @Full-width-coerce-to is just about int8 -> int64
testing 10:12
10:13 travis-ci joined, p6bannerbot sets mode: +v travis-ci
travis-ci Rakudo build passed. Juan Julián Merelo Guervós 'Changes variable name (#2472) 10:13
travis-ci.org/rakudo/rakudo/builds/450425497 github.com/rakudo/rakudo/compare/2...4b57dea4f3
10:13 travis-ci left
nine Yep, NQP's build breaks just by allocating a dummy register! 10:14
It breaks late but it breaks
timotimo sounds like your assessment has been 100% correct right from the start 10:16
maybe we should make the register allocator chaotic as a "smoke out the bugs" option
always have like 5 free register slots available and pick at random
could even have a parameter in the environment that lets us bisect which exact register allocation b0rks things 10:17
Zoffix "The
compiler did not know about non-fullwidth natives, and complained about
that with a crash. Fix by teaching it about all the types and coercing
non-fullwidth natives to fullwidth when calling the op.
From github.com/perl6/nqp/commit/df45cb...d8397e8176
Don't know if that explains the part of code you're asking about.
timotimo yeah, that's where i pasted the quote from up above 10:18
about you not adding coercion
Zoffix Also, that code was over my head, so it's possibly it does wrong things
timotimo the coercion wasn't actually needed, but for a wrong reason :D
don't worry about it 10:19
Zoffix k :)
timotimo the internet was built upon "rough consensus and working code" after all
and we're all still alive on the www
nine Zoffix: so now I wonder which part of that commit actually fixed or at least worked around the problem. Because you clearly didn't coerce anything :)
Zoffix haha :) 10:20
nine Btw. another funny find. Of all the expr JIT templates, the one for set seems to be broken. Though where in this mass of code the bug may hide is gonna be hard to find out: (template: set (copy $1)) 10:21
timotimo if there's a bug in there, it's probably going to be about how the rest of the compiler handles copies and such 10:22
nine Causes: MoarVM panic: Register types do not match between value and node in t/spec/S02-names-vars/list_array_perl.t when ran with MVM_SPESH_NODELAY=1 MVM_SPESH_BLOCKING=1
timotimo maybe the compiler is just stumbling upon broken code from earlier
since spesh is allowed to spit out code without going through the bytecode validator 10:23
it'd probably be good to output the two register types, and maybe also the register numbers
if it's just about set-ing between a 32bit and a 64bit integer, it could be fine depending on the direction
nine But....we don't have 32bit registers. How could the expr JIT compiler even notice such a discrepancy? 10:25
timotimo oh? hum. 10:27
nine The coercion thing is also surprisingly difficult to fix. Because $qastcomp.coerce will also release the source register unconditionally. But in this case we may want to coerce it a second time to the conditional's result type 10:28
Zoffix gah. So who owns goraku.org? 10:30
It was available when I tried to buy it a week ago, but somehow it wasn't purchased and now it tells me it's owned by someone :/
timotimo hum. namecheap.org seems to have it 10:31
but there's no contacts in there
Zoffix rakulang.org it is! 10:33
timotimo raku.code?
Zoffix Wonder... goraku.party or stay with rakudo.party?
Is .code a TLD? 10:34
nine Surprisingly not
.codes is 10:35
Zoffix Looks like all raku.* domains is taken by squatters
raku.codes is available… for $70/year
nine Suggested Domains 10:37
r.xyz
$12499.99*
Zoffix yikes :)
raku.org for C$2,517.77 10:43
nine There's .website, .wiki, .tools, .team, .support, .software, .run, .guide, .foundation, .expert, .download, .cool, .computer, .community, .codes, .build, .blog, .business, .global, .co.com, .biz, .club, .help available 10:44
Zoffix $40/year for cool 10:45
Zoffix is too cheap to be cool
nine DirectNIC sells raku.codes for $7.50 when registered for a year 10:46
Surprisingly it becomes more expensive per year when registered for a longer period
Zoffix Yeah and then it's jacked to $70 :)
nine Aaah, because the first year is 7.50 while the rest will be $49.99 per year 10:47
Zoffix I found the perfect one: raku.ws :)
For <ws> regex token :) 10:48
raku.wtf for $40 :) 10:50
raku.dog :) 10:57
I got raku.one, raku.ws, raku.ninja, rakulang.org, and goraku.party 10:59
11:01 pmurias joined, p6bannerbot sets mode: +v pmurias
timotimo i hope people won't misunderstand raku.ws to mean that they can reach us at raku. 11:07
11:17 lizmat joined, p6bannerbot sets mode: +v lizmat 11:27 lizmat_ joined, p6bannerbot sets mode: +v lizmat_ 11:30 lizmat left
Zoffix lizmat: why do you pay so much attention to "edgy" randos on Twitter dissing perl? :) 11:37
lizmat_ because it's those "edgy" randos that you're trying to please with "Raku", and see how it's working out 11:38
11:38 lizmat_ is now known as lizmat
Zoffix Not at all. The alias is for our community 11:38
lizmat I'm afraid I'm not part of that community 11:39
Zoffix So use "Perl 6"
I don't see what trying to party-poop people who are happy about the alias accomplishes 11:40
lizmat so you concur that there is a "Raku" and a "Perl 6" community ?
Zoffix I concur that there are people who understand the major branding problem with the language and those who think everything is fine.
lizmat did Microsoft change its name to Win4U ? 11:41
Zoffix Did we change the name? 11:42
Did 1and1 change to IONOS?
lizmat some edgy randos think we did 11:43
Zoffix Well, if you want to ruin your day by listening to people who actively brag about "upsetting a Perl 6 fanboi", be my guest. 11:44
I won't take part in that
11:44 Zoffix left 11:52 woolfy joined, p6bannerbot sets mode: +v woolfy
Geth rakudo: 45bbc9465c | (Zoffix Znet)++ | src/core/Proc/Async.pm6
Consistify Proc::Async.command with Proc.command

The Proc.command is part of 6.c spec but Proc::Asyc.path/.args do not appear to be tested even in 6.d propspec.
Make interface consistent by putting up .path/.args up for deprecation and adding Proc::Async.command
Fixes R#2444 github.com/rakudo/rakudo/issues/2444
11:53
synopsebot R#2444 [open]: github.com/rakudo/rakudo/issues/2444 [6.d review][consistency] Inconsistency: Proc.command vs Proc::Async.path + .args
roast: 51639a8efe | (Zoffix Znet)++ | S17-procasync/basic.t
Spec Proc::Async.command

PoV: github.com/rakudo/rakudo/commit/45bbc9465c Covers R#2444 github.com/rakudo/rakudo/issues/2444
11:54
11:58 woolfy left
Geth roast: 90ca958603 | (Zoffix Znet)++ | S32-io/chdir-process.t
Document intent of &*chdir tests

Can't figure out a way to prove from within roast that process dir changes as well (without using `nqp` and its `nqp::cwd` op).
For purposes of release, we'll just document the intent of the test in the comments, and hopefully will come up with proper proof later on
Addresses S#471 github.com/perl6/roast/issues/471 sufficiently enough for 6.d release purposes, to not block the release
12:07
synopsebot S#471 [open]: github.com/perl6/roast/issues/471 [6.d review] Bogus &*chdir tests
12:12 leont joined 12:13 p6bannerbot sets mode: +v leont 12:18 Zoffix joined, p6bannerbot sets mode: +v Zoffix
Zoffix AlexDaniel: I just realized there's no usual URL to toaster on github.com/rakudo/rakudo/issues/2451 Is that info available anywhere? 12:18
I mean, sure it bisected to 46ef0ea08cae96db25c7b5a954 but what was the error? Are all of them "version too late" or are there are other failure modes for the same commit 12:19
12:30 brrt joined 12:31 p6bannerbot sets mode: +v brrt 12:45 dct joined, p6bannerbot sets mode: +v dct 12:47 dct left 12:49 dct joined, p6bannerbot sets mode: +v dct 12:55 travis-ci joined, p6bannerbot sets mode: +v travis-ci
travis-ci Rakudo build failed. Zoffix Znet 'Consistify Proc::Async.command with Proc.command 12:55
travis-ci.org/rakudo/rakudo/builds/450458313 github.com/rakudo/rakudo/compare/7...bbc9465c9d
12:55 travis-ci left
buggable [travis build above] ☠ Did not recognize some failures. Check results manually. 12:55
13:04 dct left 13:37 MasterDuke left 13:41 Zoffix left
Geth star/2018.10: 796b7a4eb9 | (Steve Mynott)++ | docs/announce/2018.10.md
add one reference to Raku as a perl 6 alias
13:42
rakudo: 125789ae1e | (Tom Browder)++ (committed using GitHub Web editor) | docs/rakudo-nqp-and-pod-notes.md
update table heading and content
13:44
stmuk_ .tell Zoffix I'm probably opening a can of worms but I wonder if a change like 'Hi, my name is Camelia. I'm the spokesbug for Perl 6, the plucky little sister of Perl 5 and now nicked "Raku"' should be added to the home page 13:53
yoleaux stmuk_: I'll pass your message to Zoffix.
stmuk_ .tell zoffix s/nicked/nicknamed/ 13:54
yoleaux stmuk_: I'll pass your message to zoffix.
13:55 patrickb joined 13:56 p6bannerbot sets mode: +v patrickb 13:58 Zoffix joined, p6bannerbot sets mode: +v Zoffix
pmurias stmuk_: I wouldn't use the word nicked 14:01
stmuk_ pmurias: it was a typo. re-read above lines 14:02
I'll do a PR
Zoffix stmuk_: my suggestion would be to hold off a little on trying to stick "Raku" in every place we can. The people who weren't on board with the alias are understandably a bit freaked out right now and I don't want to make that group feel like the entire world is immediatelly turning upside down with all the changes. I'm including this blurb in 6.d release brochure i.imgur.com/YPEjb7S.png . 6.d is 14:03
getting cut. then we have a month of Advent blogs many of which will mention existence of the secondary name, and so as we enter 2019 we have a bunch of "3rd party" use of the second name so we can see how much we want to use it in core assets.
stmuk_ Zoffix: you have misunderstood me that's not my intention. 14:04
Zoffix " 14:05
"I'm the spokesbug for Raku Perl 6"
14:05 nige joined
Zoffix It doesn't have to be one or the other. The two names combine well. 14:05
stmuk_ that's not what I said! please re-read it
14:06 p6bannerbot sets mode: +v nige
stmuk_ '...the plucky little sister of Perl 5 and now nicknamed "Raku"' 14:06
nige a suggestion for the domain name is .... raku.perl.org
stmuk_ it seems to me that "nickname" fits in with alias or stage name 14:07
Zoffix decides to re-focus on 6.d release
14:07 Zoffix left
stmuk_ I'll do a PR and leave it open and hopefully it can be reconsidered once things are calmer 14:08
nige: I don't think changing domains is a Good Idea 14:10
nige ok - just saw discussion of different domains earlier 14:11
I think it's a good idea for the Perl foundation to own the domain 14:12
14:17 as joined, p6bannerbot sets mode: +v as 14:18 as is now known as Guest85603
nine Seeing github.com/perl6/mu/pull/34 I'm kinda glad that my interaction with the community is restricted to IRC and conferences. Seems like I'm missing a somewhat heated discussion. OTOH this leaves me mostly guessing what's going on... 14:30
14:34 cygx joined, p6bannerbot sets mode: +v cygx
cygx Zoffix: I would still maintain that Perl6 *is* "the next" Perl, in the same way that Modula-2 was the next Modula. 14:37
brrt nine: me too, a bit
cygx So I'd prefer slightly different language in any 'What is Raku?' announcement
Perhaps something like: 'Some who love "Perl" come to "Perl 6" and feel tricked because it is not the next release of that language, but an entirely different one.'
brrt
.oO( On the level from typo-fix to go module system, how bad is the drama? )
14:40
14:47 travis-ci joined, p6bannerbot sets mode: +v travis-ci
travis-ci Rakudo build passed. Tom Browder 'update table heading and content' 14:47
travis-ci.org/rakudo/rakudo/builds/450483172 github.com/rakudo/rakudo/compare/4...5789ae1ebe
14:47 travis-ci left
Geth rakudo: 8a26025750 | (Tom Browder)++ (committed using GitHub Web editor) | docs/rakudo-nqp-and-pod-notes.md
Update rakudo-nqp-and-pod-notes.md
14:50
rakudo: 316ad5256e | (Tom Browder)++ (committed using GitHub Web editor) | docs/rakudo-nqp-and-pod-notes.md
escape left angle brackets for visibility
14:51
14:59 Zoffix joined, p6bannerbot sets mode: +v Zoffix
Zoffix nine: a month ago, I posted my letter to Larry for alias creation on Facebook and that user went all out dissing it, saying they were never consulted and they apparently missed the year and a half of blog posts on the topic because they don't follow all of the channels. Then they created another Facebook post to mock my post about proposing to remove sigils. At that point, I left our FB group, so I don't 15:01
know what else they posted.
nine Oh darn....I finally realized how to utilize bytecode writing in improving compile_mastop. That will save us allocation of a couple 100K arrays. But that's gonna be a pretty hefty rewrite of that function :/ 15:02
Zoffix TBH, I'm not sympathetic to throwing a shit fit just because things didn't go exactly the way you wanted them to go. Nothing's stopping that user from writing about "Perl 6" 15:03
cygx: updated: i.imgur.com/kzuSv22.png 15:06
cygx Zoffix++ 15:07
brrt Zoffix++, whatever I think of the new alias, there's no denying your work on making perl6 'acceptable' 15:11
15:15 fake_space_whale joined, brrt left
Guest85603 @Zoffix you are behaving as a child 15:16
15:16 p6bannerbot sets mode: +v fake_space_whale
cygx btw, how's the 'Raku' brand supposed to be versioned? 15:16
Perl 6.d == Raku 2018?
Zoffix Guest85603: At least I don't hide behind a "Guest" 15:17
cygx: Perl 6 6.d == Rakud 6.d
*Raku
:)
cygx so Raku is stuck with 6.* for the forseeable future?
Zoffix cygx: pretty much. Since it's not merely something we put in print. It has to also work and correctly compare against the existing `v6.d` version pragma. 15:18
Guest85603: did you have more to discuss with me? I'm about to head out. 15:19
v6.c I meant
m: say v6.c before v6.d
camelia True
cygx well, Java did make a distinction between 'product version' 5.0 and 'developer version' 1.5.0 when they pulled that particular rebranding stunt in '04 15:21
Zoffix & 15:22
Guest85603: but I read the backlog, so… let it all out, don't hold back
15:22 Zoffix left
patrickb The more bikesheddable a topic is, the more potential there is for the discussion heating up and fighting. Everyone can have an opinion, irrespective of the level of their involvement. A naming discussion is *very* bikesheddable. In that light I very much like the idea behind the "marketing group" recently discussed. 15:31
15:35 Guest85603 left
jdv79 i stiil dont get it because anyone that digs enough will find out it is still, or was, Perl6. 15:37
stmuk_ its not a replacement name, you don't have to use it. It's very Perl like "there is more than one name to call it" and like his original licence hack 15:39
I really can't see what the fuss is about
15:42 andrewshitov_ joined, p6bannerbot sets mode: +v andrewshitov_
stmuk_ and the vases look cool 15:43
jdv79 im fine with it. just based on the original reasoning it seems less than sufficuent:) 15:45
so to do raku you rakudo? 15:46
should be some fun word play in there 15:47
15:52 geekosaur left 15:56 Zoffix joined, p6bannerbot sets mode: +v Zoffix, Zoffix left 16:16 lizmat left
b2gills I was thinking that `use Raku;` could be an alias to `use v6;` (Perhaps extending to `use Raku.d` -> `use v6.d`) 16:20
16:27 lizmat joined, p6bannerbot sets mode: +v lizmat
cygx b2gills: if Raku and Perl6 version numbers remain aligned, I see no need for that... 16:29
but feel free to implement a slang that translates `.raku` to `.perl` ;)
~~ 16:44
16:44 cygx left 16:59 ExtraCrispy joined, p6bannerbot sets mode: +v ExtraCrispy 17:08 martin1969 joined 17:09 p6bannerbot sets mode: +v martin1969
nine Oh boy, \qq[] is a real life saver when generating Perl code :) 17:12
17:32 martin1969 left
Geth rakudo: 00eb7b8561 | (Elizabeth Mattijsen)++ | src/core/Rakudo/Iterator.pm6
Simplify some iterators
17:37
17:48 cognominal-p6 left 17:55 pmurias left
samcv .tell Zoffix i'm back from vacation and looking over the 6.d changelog 17:59
yoleaux 3 Nov 2018 00:21Z <Zoffix> samcv: I've logged unicode changes as "- Expanded specification coverage of Unicode routines and features - Upgraded coverage to Unicode version 11" and will log $*COLLATION and any new routines I see. So I guess it's logged now, but feel free to make it more specific. I plan to have ChangeLog in completed form by Nov 3. ~11PM EST
samcv: I'll pass your message to Zoffix.
18:02 Zoffix joined, p6bannerbot sets mode: +v Zoffix
Zoffix samcv: I'm already entering it into releas brochure BTW, but we can still make changes 18:02
yoleaux 17:59Z <samcv> Zoffix: i'm back from vacation and looking over the 6.d changelog
dogbert17 Zoffix: do you still have your 32 bit box handy? 18:03
Zoffix dogbert17: yeah 18:04
dogbert17 Zoffix: can you build the latest Rakudo on it, suddenly it hangs on my system
Zoffix OK. FWIW my latest one that successfully built is 2018.10-53-g8e858c8 18:05
dogbert17 that's a couple of days old isn't it?
Zoffix m: say $*PERL.compiler.version.say
camelia v2018.10.75.g.00.eb.7.b.856
True
Zoffix yeah, like 12 commits behind 18:06
22
dogbert17 I wonder if something has happened today or yesterday unless my Linux install has somehow gotten busted
18:08 Zoffix left
dogbert17 it hangs on either 'Creating tools/build/install-jvm-runner.pl ...' or '/usr/bin/perl -MExtUtils::Command -e cp 3rdparty/dyncall/dyncallback/*.h /home/dogbert/repos/rakudo/install/include/dyncall' 18:08
18:08 Zoffix joined, p6bannerbot sets mode: +v Zoffix, Zoffix left, Zoffix joined, p6bannerbot sets mode: +v Zoffix
Zoffix dogbert17: yeah, hangs now: R#2474 18:11
synopsebot R#2474 [open]: github.com/rakudo/rakudo/issues/2474 [build][⚠ blocker ⚠] Build hangs on 32-bit boxes
dogbert17 Zoffix: thanks for checking 18:12
18:15 ExtraCrispy left
Geth 6.d-prep: 021095fb28 | (Zoffix Znet)++ (committed using GitHub Web editor) | ChangeLog.md
Remove "Removals"

It's a minor housekeeping detail that doesn't need to be relayed to users
18:18
6.d-prep: 9d2ac2c277 | (Zoffix Znet)++ (committed using GitHub Web editor) | ChangeLog.md
Move deprecations closer to the start

Since the first two sections are the only ones users might need to modify their code for
18:19
6.d-prep: e8f63d23f5 | (Zoffix Znet)++ (committed using GitHub Web editor) | ChangeLog.md
Formatting fixups
18:21
18:23 cognominal-p6 joined 18:24 p6bannerbot sets mode: +v cognominal-p6
Geth 6.d-prep: 7655708989 | (Zoffix Znet)++ (committed using GitHub Web editor) | ChangeLog.md
Remove Proc::A TODO

It's been TODONE
18:28
Zoffix Hm... A thought occurs: release 6.d a few dozen hours earlier. To let the 6.d release and alias thing ride the same wave instead of 6.d being muted by it 18:30
m: DateTime.new('2018-11-06TOO:OO:OOT+14:00').in-timezone(+5*60).say 18:31
camelia Invalid DateTime string '2018-11-06TOO:OO:OOT+14:00'; use an ISO 8601 timestamp (yyyy-mm-ddThh:mm:ssZ or yyyy-mm-ddThh:mm:ss+01:00) instead
in block <unit> at <tmp> line 1
Zoffix m: DateTime.new('2018-11-06TOO:OO:OO+14:00').in-timezone(+5*60).say
camelia Invalid DateTime string '2018-11-06TOO:OO:OO+14:00'; use an ISO 8601 timestamp (yyyy-mm-ddThh:mm:ssZ or yyyy-mm-ddThh:mm:ss+01:00) instead
in block <unit> at <tmp> line 1
Zoffix gah
what is its problem :/ 18:32
oh, I somehow used "O" instaed of "0" :O
japhb I was just about to say. :-)
Zoffix m: DateTime.new('2018-11-06T00:00:00+14:00').in-timezone(+5*60).say 18:33
camelia 2018-11-05T10:05:00+00:05
Zoffix So I guess if I cut the release tonight, it's more or less Diwali time.
stmuk_ yay!
japhb I assume you meant +5*3600 ...
Zoffix m: DateTime.new('2018-11-06T00:00:00+14:00').in-timezone(+5*3600).say 18:34
camelia 2018-11-05T15:00:00+05:00
Zoffix oh
OK, Monday, 5pm EST. That will be the time 18:35
Hmmm 18:36
japhb EST would be -5, yes?
Zoffix Ah :)
Then tonight. And it'll make to Weekly too then
"tonight" == in ~10hr or so 18:37
Zoffix leaves for the Final Countdown finish ups
japhb Triple shot seems good indeed. (Diwali + Raku + Weekly)
Zoffix \o/ www.youtube.com/watch?v=9jK-NcRmVcw \o/
18:37 Zoffix left
stmuk_ holds up lighter flame phone app 18:39
18:50 travis-ci joined, p6bannerbot sets mode: +v travis-ci
travis-ci Rakudo build failed. Elizabeth Mattijsen 'Simplify some iterators' 18:50
travis-ci.org/rakudo/rakudo/builds/450541089 github.com/rakudo/rakudo/compare/3...eb7b856129
18:50 travis-ci left
buggable [travis build above] ☠ Did not recognize some failures. Check results manually. 18:50
stmuk_ actually this is more rawk
youtu.be/zC7uPC6KeA4?t=206
Geth 6.d-prep: 68325977a1 | (Zoffix Znet)++ (committed using GitHub Web editor) | ChangeLog.md
Infer more strongly that `is-approx` has new behaviour along with new name
18:58
6.d-prep: 8aef89192f | (Zoffix Znet)++ (committed using GitHub Web editor) | ChangeLog.md
s/Rat/Rational/
19:10
6.d-prep: 4317e5ee5a | (Zoffix Znet)++ (committed using GitHub Web editor) | ChangeLog.md
Reword for clarity
19:12
6.d-prep: 0d3adcf365 | (Zoffix Znet)++ (committed using GitHub Web editor) | ChangeLog.md
Use better wording for TWEAK
19:25
6.d-prep: f455352bdf | (Zoffix Znet)++ (committed using GitHub Web editor) | ChangeLog.md
Consistify notation for sub-only routines
19:29
6.d-prep: c7ca5c182a | (Zoffix Znet)++ (committed using GitHub Web editor) | ChangeLog.md
Add missing comma
19:33
19:34 cognominal-p6 left
Geth 6.d-prep: 7cb07fee32 | (Zoffix Znet)++ (committed using GitHub Web editor) | ChangeLog.md
Remove stray words
19:39
roast/reorganize-modules: 0a737a5aba | (Nick Logan)++ | 557 files
Move all module code to segregated namespace under packages/

Move all module code that will be loaded under packages/ -- while also structuring their filesystem layout as distributions. This should make it easier to keep track of what namespaces a given
  `use lib ...` is bringing in.
... (6 more lines)
6.d-prep: 9e751589ad | (Elizabeth Mattijsen)++ (committed using GitHub Web editor) | ChangeLog.md
Remove Raku from header

The Perl 6.d **specification** has **nothing** to do with `raku`.
19:42
Kaiepi i have an idea on how to reimplement .native-descriptor for async sockets 19:47
my first attempt was waaay overcomplicating things
Geth roast/reorganize-modules: 3e9291f400 | (Nick Logan)++ | 557 files
Move all module code to segregated namespace under packages/

Move all module code that will be loaded under packages/ -- while also structuring their filesystem layout as distributions. This should make it easier to keep track of what namespaces a given
  `use lib ...` is bringing in.
... (6 more lines)
19:49
Kaiepi there's no need for adding the asyncsocket op like i first thought 19:50
Geth 6.d-prep: c8011e2e82 | (Zoffix Znet)++ (committed using GitHub Web editor) | ChangeLog.md
Add missing article
19:51
6.d-prep: 4b671be627 | (Zoffix Znet)++ (committed using GitHub Web editor) | ChangeLog.md
Remove duplicate entry

Already mentioned in Versioned-Changes section
20:04
20:07 Zoffix joined, p6bannerbot sets mode: +v Zoffix
Zoffix lizmat: what do you mean by the 6.d-prep commit? 20:08
lizmat: Raku is the second name for the Perl 6 language. It has everything to do with the Specification
lizmat raku is an alias of Perl 6 . roast (aka the Perl 6 specification) in my view has nothing to do with marketing and hence not part of the spec 20:09
Zoffix lizmat: also, this is the release brochure's current look: i.imgur.com/VEJkprX.png 20:10
I'm surprised at your new pushback at this considering "Rakudo Perl 6" that started all this was birthed by your mind :) 20:12
lizmat huh???
also: "Rakudo Perl 6" is something else than "Raku Perl 6" 20:13
Zoffix to which part? The brochure, pushback, or the Rakudo Perl 6?
lizmat birthed by my mind ? huh? 20:14
Zoffix Yeah, rakudo.party/post/The-Hot-New-Lang...med-Rakudo says "The "Rakudo Perl 6" name for the language was suggested by lizmat++, so I assume there's at least one other core team member who's open to the language name tweak".
It doesn't link to logs tho
But if you find the reddit posts for that article, you'll find your comment clarifying what your suggestion meant 20:15
Or maybe clarification was made in-channel 20:16
lizmat www.reddit.com/r/perl6/comments/9m...e/e7cdgih/
is what I remember
Zoffix Ah now, my memory is correct: www.reddit.com/r/perl6/comments/6l...o/djy3yz0/
s/now/no/; 20:17
20:17 cognominal-p6 joined
lizmat that's about "Rakudo Perl 6" 20:17
not about "raku" on its own 20:18
20:18 p6bannerbot sets mode: +v cognominal-p6
lizmat BTW: a new low :-( www.reddit.com/r/perl/comments/9tw...u/e91mu3g/ 20:19
and that's by one of the core p5p people :-(
20:19 cognominal-p6 left 20:20 cognominal-p6 joined 20:21 p6bannerbot sets mode: +v cognominal-p6
Zoffix Well, both the 6.d release blog post announcement and the release brochure refer to the language by both names, "Raku Perl 6". The alias was scheduled to be decided on by 6.d release and I don't see the point of omiting "Raku" on the title on a release whose one of the main deliverables was creation of "Raku" alias. 20:21
Geth 6.d-prep: 9a56c333e1 | (Elizabeth Mattijsen)++ (committed using GitHub Web editor) | ChangeLog.md
Revert
20:22
Zoffix In fact, there's even a positive point to include it: it servers as a clarificatory statement that "Raku" isn't a full rename. 20:24
lizmat who are we kidding ? 20:25
a full rename is what *you* want to achieve without rocking the boat too much now
a full rename is what a very vocal part of the perl 5 community wants 20:26
20:26 fake_space_whale left
Zoffix If someone wants to argue against the alias because it's unclear if it's a full rename, I invite them to first argue that "Perl 6" is not confusing because it's unclear that it's not the next version of "Perl" 20:28
lizmat and there we disagree :-)
so that's basically the problem
I guess the sister language meme and I are incompatible 20:29
20:29 lizmat left
Zoffix Raku's success doesn't have to include Perl's demise. You argued yourself in one of your posts that Perl must not die. 20:30
cognominal-p6 "outvote Larry". So much for benevolent dictator, welcome to the tyranny of the majority :) en.wikipedia.org/wiki/Tyranny_of_the_majority
Zoffix cognominal-p6: what's that about? 20:31
cognominal-p6: Raku's is TimToady's ruling.
cognominal-p6 That was in the reddit thread about renaming Perl 6
Zoffix Not a majority vote of anything.
Ah 20:32
cognominal-p6 In practice, this is more the tyranny of an intolerant and uninformed minority. :) 20:33
Zoffix heh
Geth 6.d-prep: 052f61c6b1 | (Zoffix Znet)++ (committed using GitHub Web editor) | ChangeLog.md
Fix typo
20:35
cognominal-p6 whatever the name, I want to thanks the core team. I love Perl 6 and its implementation 20:37
Zoffix :) 20:38
20:40 Zoffix left 20:48 andrewshitov_ left
Geth roast: cc2bc13678 | (Zoffix Znet)++ (committed using GitHub Web editor) | docs/New-Features-Policy.md
Move "Policy For Implementation of New Features" from 6.d-prep

It's still a fairly rough draft, but a starting point we can refine as we move forward
20:52
Kaiepi i really need to buy more ram 20:53
i can't run make spectest with rakudo on the jvm
just running make alone grinds my computer to a halt while it's trying to build the QAST 20:54
AlexDaniel Zoffix: yes, currently the output is not there. I'll fix that for the next run 21:06
Zoffix: most of them had version too late issue 21:07
21:14 Kaiepi left 21:15 Kaiepi joined, p6bannerbot sets mode: +v Kaiepi
dogbert17 Zoffix: 'Version-Con*s*trolled Changes" 21:17
21:19 nige left 21:26 Zoffix joined, p6bannerbot sets mode: +v Zoffix
Geth 6.d-prep: b6dda45751 | (Zoffix Znet)++ (committed using GitHub Web editor) | ChangeLog.md
Fix typo dogbert17++
21:27
21:37 cognominal-p6 left
Geth nqp: Kaiepi++ created pull request #513:
Reimplement async socket introspection
21:48
rakudo: Kaiepi++ created pull request #2475:
Reimplement .native-descriptor for async sockets
Kaiepi \o/ 21:50
21:54 lizmat joined, p6bannerbot sets mode: +v lizmat 21:55 lizmat left
Geth rakudo: d3f0286c3d | (Zoffix Znet)++ (committed using GitHub Web editor) | CREDITS
Add second name for pmurias (as it shows up in git log)
21:56
nqp: Kaiepi++ created pull request #514:
Pass correct connection port in nqp::asynclisten
22:07
roast: Kaiepi++ created pull request #495:
Implement spec tests for .native-descriptor on async sockets
22:42
23:01 lizmat joined, p6bannerbot sets mode: +v lizmat 23:16 MasterDuke joined, p6bannerbot sets mode: +v MasterDuke, MasterDuke left, MasterDuke joined, herbert.freenode.net sets mode: +v MasterDuke, p6bannerbot sets mode: +v MasterDuke
AlexDaniel Kaiepi++ 23:28
Kaiepi :)
AlexDaniel Kaiepi: I'm very satisfied with your persistence :)
Kaiepi i'm determined to get this right since i need it for my telnet library 23:29
next on my list is getsockopt/setsockopt support, or just ops for some of the commonly used options 23:30
timotimo are we okay with not implementing tcp urgent notification stuff?
as you could call it "basically broken" 23:31
AlexDaniel Kaiepi: maybe you can also take a look at github.com/MoarVM/MoarVM/pull/659 ? 23:32
Kaiepi i will AlexDaniel 23:33
AlexDaniel :O
Kaiepi: ♥
23:42 patrickb left
Kaiepi AlexDaniel, looks ok apart from a nitpick on a #define 23:43
AlexDaniel Kaiepi: yeah, I don't remember why I didn't merge it 23:44
23:47 leont left 23:52 travis-ci joined, p6bannerbot sets mode: +v travis-ci
travis-ci Rakudo build passed. Zoffix Znet 'Add second name for pmurias (as it shows up in git log)' 23:52
travis-ci.org/rakudo/rakudo/builds/450613195 github.com/rakudo/rakudo/compare/0...f0286c3d47
23:52 travis-ci left
Zoffix t/04-nativecall/12-sizeof.t (rakudo's test suite) segfaulted just now 23:53
What sucks is 6.d roast doesn't pass on Windows 23:54
Which has been the case for at least more than a year S#320 23:55
synopsebot S#320 [open]: github.com/perl6/roast/issues/320 Windows roast failures just before 2017.09 release
Zoffix There's always 6.d-errata, I guess if it comes out that some tests were badly designed for Windows rather than bugs in Rakudo… and someone badly needs a Windows-functional 6.d roast by that time. 23:58