00:02 sena_kun left
AlexDaniel jnthn: speaking of operators, what we do actually need is an op that checks for Nil before calling a method 00:28
I think in Ruby that's &.
for example: 00:29
die ‘No joystick device found!’ unless $input-device&.IO.e 00:30
if you know how to write that right now without `and` please let me know :)
because I have this kind of stuff all over my code
actually, in this case .? sorta works? 00:31
m: say Nil.IO 00:32
camelia (Path)
AlexDaniel m: say Nil.?IO
camelia (Path)
AlexDaniel m: say Nil.?IO.e 00:33
camelia Invocant of method 'e' must be an object instance of type 'IO::Path', not a type object of type 'IO::Path'. Did you forget a '.new'?
in block <unit> at <tmp> line 1
AlexDaniel ok, it sorta doesn't
vrurg AlexDaniel: sounds like you need 'invoke if defined' kind of thing. 01:24
AlexDaniel: There is an op for you. It's called andthen. 01:28
m: my $foo; die "ok" unless $foo andthen .IO.e;
camelia ok
in block <unit> at <tmp> line 1
vrurg m: my $foo = "no-file"; die "ok" unless $foo andthen .IO.e; 01:29
camelia ok
in block <unit> at <tmp> line 1
AlexDaniel oh
vrurg It's the first time I see a real use for it. 01:30
AlexDaniel so 01:35
die ‘No joystick device found!’ unless $input-device andthen .IO.e;
vrurg I think it is. 01:36
AlexDaniel geez that's a weird operator :D 01:39
but I honestly didn't realize that I could use it this way
vrurg I think there was a discussion about removing it somewhere. Or was it about `notandthen`? Can't really remember now. 01:41
AlexDaniel: BTW, an example in the docs is about your situation: docs.perl6.org/language/operators#...notandthen 01:44
AlexDaniel “At first glance, notandthen might appear to be the same thing as the orelse operator. The difference is subtle” 01:47
yeah, I remember reading this
and then not using any of these :D
vrurg Me too. Though I probably spend ~60-70% or my time in NQP lately... 01:49
AlexDaniel sounds like you're not using raku in production enough! 01:55
vrurg It is how things are now. 02:01
I've some spare time on my hands and trying to use as much of it as possible for bug fixing. 02:03
*I've got
Hm, is there a dumper for moar precompiled files? 02:05
AlexDaniel vrurg++ 02:15
Geth problem-solving: 7af1dfc3f9 | (Fernando Correa de Oliveira)++ (committed by Aleks-Daniel Jakimenko-Aleksejev) | README.md
Add my name to the list
problem-solving: 5e9d288a3b | (Aleks-Daniel Jakimenko-Aleksejev)++ (committed using GitHub Web editor) | README.md
Merge pull request #84 from perl6/merge-mishap

Add FCO to the list
problem-solving/master: 4 commits pushed by (Vadim Belman)++, (Aleks-Daniel Jakimenko-Aleksejev)++ 03:01
vrurg AlexDaniel: it means less time for the bugs now. 03:05
AlexDaniel weekly: Welcome @FCO and @vrurg to the list of problem-solving reviewers! github.com/perl6/problem-solving/pull/84 github.com/perl6/problem-solving/pull/87 03:06
notable6 AlexDaniel, Noted! (weekly)
AlexDaniel SmokeMachine: ↑
vrurg: what do you mean?
vrurg I guess, roast and language release process are my direct responsibility now. 03:07
AlexDaniel vrurg: well, being on that list simply means that you need to review stuff… you can still send everyone to jnthn :D 03:09
not how it should be, but I guess things will improve as more people get involved with stuff :) 03:10
vrurg AlexDaniel: I applied for the reason to take over the language release management. Thus, I consider it my duty now.
AlexDaniel vrurg: yeah, but we haven't added a new label yet github.com/perl6/problem-solving/p...-541940054 03:11
so tickets will still be assigned to jnthn by default 03:12
I guess jnthn will reassign anything that belongs to you?
vrurg The label is one part of the overall process. I would need to think of it in general. Like, I tried to implement a project management on my fork of problem-solving. 03:13
AlexDaniel nice! 03:14
vrurg It's up to jnthn, of course.
AlexDaniel speaking of time, I won't have any this week :(
like, at all 03:15
there are release braches in nqp and rakudo, so once moarvm is released nobody should feel blocked by lack of release
vrurg Something good, I hope?
AlexDaniel vrurg: no, just wörk (ish), but generally life is good :) 03:16
btw if anyone is interested, previously I said that I'll have to do alternative service in my country starting Jan 2020 03:17
but that's no longer the case!
I'm now free till at least Nov 2020, and a bit more after that, and possibly forever
vrurg Oh, unblocking merges! Eventually! BTW, will you have time for commenting on issues/PRs? I hope to get jnthn a bit more involved into the operators discussion. Your opposition could be useful.
AlexDaniel hah, useful opposition :) 03:18
vrurg: I'll try. Thing is, we already have .+ and .?, which I didn't really know about
vrurg It is always useful. Total consensus is the direct path to swamping death. Prooved by USSR.
AlexDaniel like I knew about .? but kinda forgot about .+
vrurg: some time ago Zoffix proposed this idea that whenever you add a new feature you have to remove one too :) 03:19
which is a bit extreme, but it's true that currently we're only adding and not removing anything
vrurg I was saying it once, .+ is pretty much useless without fully implemented submethods. It would have more reasons to be used with my PR.
Oh, the new ops make .* obsolete. I would gladly remove it. 03:20
Can I do it tomorrow??? Pleeeeeaseeee?! :D
AlexDaniel and somehow lately I really want to just delete and simplify stuff… 03:21
“Raku” change is very refreshing :) 03:22
no need for no-break space nonsense, no need to explain the name to others 03:23
timotimo well rak u too
AlexDaniel :)
vrurg And here we go, people started kidding on the new name. OMG... :D 03:25
AlexDaniel vrurg: btw notandthen is the first item here: github.com/perl6/problem-solving/issues/64 03:26
vrurg AlexDaniel: aha, and there too. 03:27
Oh, and roast rename... Though I would leave it as is. rakus-spec sounds cumbersome. 03:29
AlexDaniel vrurg: once I have a bit more time in a week I'll start figuring out what to do with the github org
vrurg: but if we're moving perl6/roast to rakulang or, then I'd much rather move it to something better like rakulang/spec 03:30
vrurg: even if it remains being “roast”
… or should it be /raku-spec ? ok now I don't know :) 03:31
vrurg AlexDaniel: it'd be confusing to people then. Project name better be consistent to the repo
Just 'spec' is no good.
AlexDaniel vrurg: it doesn't even mention “roast” anywhere in the readme 03:32
so I don't think it'll be more confusing than it is now :) 03:33
vrurg It's in announces and release-guide. Not to mention unicode-generated-tests and New-Release-Policy
AlexDaniel mmm that's true
anyway, I'm not sure, but point is that if repos are moving we can as well consistify and make them better 03:34
for example, that is *definitely* moving to something with a different name: github.com/perl6/specs/
this, I mean 03:35
“and the "specs" in the repository name of the design documents tends to mean "speculations" now”
mental gymnastics instead of fixing the issue :) 03:36
vrurg That repo is likely to become something like 'historical-docs'
They still serve a good source of design resoning. I use them to find out why something was implemented this or that way. 03:37
AlexDaniel old-design-docs ? :)
vrurg Anyway, we could expand ROAST abbreviation to something more meaningful. 'R' will stand for Raku, apparently. 03:38
AlexDaniel: either 'original-design-docs' or 'historical-design-docs'
AlexDaniel we'll figure it out 03:39
vrurg ROAST: Raku Official ... Set of Tests? :) 03:40
Anyway, it was once proposed to reconsider roast structure because basing it on SNN is really confusing for people not historically involved (it took me half a year to realize the link between the old specs and directories). 03:42
AlexDaniel yeah, it is confusing
vrurg added an item on my Roast TODO list 03:43
AlexDaniel you mean RakuSpec TODO list 03:46
ok just kidding :) 03:47
vrurg Back to the label thing... 6e-release label would be needed, as well as labels for any future language versions.
Oh, let's bikeshed! Such a fun! :D :D :D
AlexDaniel vrurg: speaking of bikeshedding, my proposal is to have 6.e mean just e 03:49
vrurg: last time there was a very long discussion on what D is going to be
vrurg AlexDaniel: I basically like it. But technically... We have too many changes related to the renaming to introduce another one without extra need. 03:50
I'd use `Raku E` in papers and media. But stick to `use v6.e` until the next release, perhaps. 03:51
We could even consider switching to all-numeric versioning, but that would likely mean jumping over to 7.x
But, again, not before 6.e. 03:53
AlexDaniel yeah, that's ok 03:57
I think Raku 6.e is more correct in media, even if it does contain a little bit of Perl 6 in it 03:58
vrurg A problem-solving would be needed on this around this Jan/Feb, after we sort out most of the renaming problems. 04:04
And people need to cool down because it could turn into another round of hot discussions. 04:05
vrurg is afk& 04:09
06:10 jmerelo joined 06:28 sena_kun joined 07:18 sena_kun left 07:33 jmerelo left 07:57 ufobat joined 08:40 sena_kun joined 09:08 sena_kun left 09:28 maettu joined 10:20 sena_kun joined
|Tux| Rakudo version 2019.07.1-454-g7e7aa2626 - MoarVM version 2019.07.1-313-g836396c74
csv-ip5xs0.707 - 0.765
csv-ip5xs-206.329 - 6.366
csv-parser21.450 - 21.638
csv-test-xs-200.429 - 0.439
test5.954 - 6.211
test-t1.747 - 1.767
test-t --race0.850 - 0.908
test-t-2027.157 - 29.223
test-t-20 --race9.047 - 9.371
10:50 Guest13443 joined 11:43 Altai-man_ joined 11:45 sena_kun left 11:55 ggoebel joined 12:03 |Tux| joined
Geth Blin: nxadm++ created pull request #14:
Merge nxadm/raku-blin-docker
tbrowder Speaking of hot discussions: I really don't want to rename roast stuff away from the current SNN. Having trouble knowing the meaning is rectified with more questioning and docs. The synopses are too valuable to hide away under yet another layer of indirection. we don't have enough documentation for new developers as it is, renaming to confuse current folks doesn't help the cause IMHO. 13:31
13:44 sena_kun joined 13:46 Altai-man_ left
Geth Blin: nxadm++ created pull request #15:
migration from nxadm/raku-blin-docker
14:34 lucasb joined 14:36 ggoebel left
tbrowder One more comment ref roast and SNN "speculations": there are still NYI design features listed in them, at least in S26 (Documentation), so there is still Perl-covered gold there that hasn't been fully mined. 15:27
vrurg tbrowder: making them 'historical' doesn't mean they won't be a good source of information. 15:34
tbrowder i know, but, to me, using the term "speculations" for the original design docs has always had a bad connotation that might give new folks the wrong idea of them. 15:38
15:44 Altai-man_ joined 15:45 sena_kun left 15:46 ggoebel joined 15:49 Altai-man_ is now known as sena_kun
tbrowder AlexDaniel: i got some info from git folks on upcoming integrated CI testing with git actions that you've probably seen <github.com/features/actions>. they are taking sign-ups for beta testing. 2000 free minutes/month for open source, excess minutes over 2000 per month for linux, win, mac, respectively: $0.008, $0.016, $0.08. 15:52
the beta requires agreeing to an NDA (non-disclosure agreement). the link also mentions free overtime for self hosting (coming soon, no mention of cost). 15:56
checking travis... 16:06
Geth nqp/truffle: fb8237944d | (Paweł Murias)++ | src/vm/jvm/Truffle.nqp
[truffle] Remove the unusued :side_effects argument when adding nodes

We are not using this when emitting AST for the truffle backend
nqp/truffle: 0dc156cc02 | (Paweł Murias)++ | 2 files
[truffle] Implement nqp::setmethcache
16:24 moon_child left 16:30 moon-child joined 16:32 moon-child left, moon-child joined 16:34 Kaiepi joined 16:35 Kaiepi left, Kaeipi joined 16:39 Merfont joined 16:40 Kaeipi left
El_Che tbrowder: I am in the beta, but I haven't tested yet. Spent quite some time getting to learn Travis. However, travis has been quite flakey lately 17:06
AlexDaniel: I sent a PR with the Docker changes to Blin 17:09
17:32 cognominal joined 17:43 Altai-man_ joined 17:45 sena_kun left 18:16 Merfont is now known as Kaiepi
Kaiepi when i need to use an array with REPR VMArray, is there a class that already exists that i can use or do i need to make my own? 18:17
this is for something in src/core.c, so i'm not sure if NQPArray can be used or not 18:19
vrurg Kaiepi: grep comes back with class array from native_array.pm6 18:25
Kaiepi idt that can be used without being parameterized, and it doesn't let me parameterize it with Mu 18:27
there's IterationBuffer though, is that meant to be used in cases like this? 18:28
vrurg I think the starting comment in IterationBuffer could provide you with more information than I can. 18:29
Kaiepi looks like it'll work for this 18:32
AlexDaniel El_Che: I did approve your request to use travis 18:49
El_Che: btw it could've been a merge of two repos (with two separate starting commits) 18:52
El_Che: though there's probably no good reason for doing it that way besides “because we can” :D 18:53
Geth Blin: 0d09f2eb11 | (Claudio Ramirez)++ | 9 files
migration from nxadm/raku-blin-docker
Blin: 0b12bc0b33 | (Aleks-Daniel Jakimenko-Aleksejev)++ (committed using GitHub Web editor) | 9 files
Merge pull request #15 from nxadm/master

migration from nxadm/raku-blin-docker
El_Che AlexDaniel: no, the travis thing was intentional :) 19:03
AlexDaniel right? That's why I approved it? :) 19:04
19:09 jmerelo joined
El_Che AlexDaniel: master pushes will end in docker hub with :latest, version tags (eg, v1.0.1) will end up as :version (e.g. :1.0.1) 19:12
Sometimes the travis build fails because the Digest module is more than 10 minutes without output 19:13
Not always, though
I added --verbose to zef, and it helped somewhat, but maybe we need to go for --debug 19:14
Geth rakudo/ugexe-patch-1: 2c5c013558 | (Nick Logan)++ (committed using GitHub Web editor) | .github/FUNDING.yml
Create FUNDING.yml
jmerelo If you've checked enough docs-stage.perl6.org, I'll replicate it in docs.perl6.org.
tellable6 2019-11-03T17:47:13Z #raku <cfa> jmerelo: it looks like docs.raku.org isn't building, picking up updates
hey jmerelo, you have a message: gist.github.com/18ab103689fdd01aba...5769bd7625
jmerelo .tell cfa it's done by hand for the time being. Automatic deployment is yet to come... 19:24
tellable6 jmerelo, I'll pass your message to cfa
Geth rakudo: ugexe++ created pull request #3277:
Create FUNDING.yml
jmerelo .tell rba whatever is the best for you. The docs*.perl6.org are only going to be temporary. 19:25
tellable6 jmerelo, I'll pass your message to rba
Geth rakudo: 2c5c013558 | (Nick Logan)++ (committed using GitHub Web editor) | .github/FUNDING.yml
Create FUNDING.yml
rakudo: 3802cdfcad | (Aleks-Daniel Jakimenko-Aleksejev)++ (committed using GitHub Web editor) | .github/FUNDING.yml
Merge pull request #3277 from rakudo/ugexe-patch-1

Create FUNDING.yml
El_Che AlexDaniel: Can you enable this in Travis? travis-ci.org/perl6/Blin 19:33
(or give me access?)
AlexDaniel “This is not an active repository You don't have sufficient rights to enable this repo on Travis. Please contact the admin to enable it or to receive admin rights yourself”
what the? :) 19:34
Geth rakudo: 1ec69e1f6a | tony-o++ (committed using GitHub Web editor) | .github/FUNDING.yml
Add to yaml as GH approves more folks.
El_Che AlexDaniel: maybe add travis here: github.com/perl6/Blin/settings
AlexDaniel El_Che: yeah 19:36
El_Che: is it better now? 19:37
I hope I didn't break anything…
El_Che Now go here: travis-ci.org/account/repositories and enable Blin 19:38
AlexDaniel El_Che: there's no such repo O_o 19:42
jmerelo El_Che: I think I can do that. 19:43
El_Che AlexDaniel: whut?? I see it
AlexDaniel ah, wrong tab
El_Che and I have the switch to enable it
AlexDaniel enabled!
El_Che (but no rights to do it :) )
great job :)
just triggered a build 19:44
AlexDaniel El_Che: I just gave you admin rights to blin repo, if that helps
El_Che AlexDaniel: should not be needed now, travis seems ok
AlexDaniel: it's handy though to restart build
jmerelo El_Che: I just gave requested permissions, I don't know if it was needed when I did it. 19:45
El_Che see above above Digest
anyway, what's that "perl6" org? Sounds old :)
19:45 Altai-man_ left
El_Che AlexDaniel: the renaming will be repo by repo, I understand 19:46
lizmat notable6: weekly 19:47
notable6 lizmat, 8 notes: gist.github.com/d40f2774d88e4999f7...7ecb39b2c0
19:48 jmerelo left
rba . 19:53
tellable6 2019-11-04T19:25:11Z #raku-dev <jmerelo> rba whatever is the best for you. The docs*.perl6.org are only going to be temporary.
rba Hmm. Just missed jmerelo. 19:54
lizmat El_Che: re "new rakudo-pkg support (and packages) for new releases: ubuntu 19.10, Fedora 31, centos 8", should I use a specific link for that? 19:55
El_Che lizmat: maybe github.com/nxadm/rakudo-pkg#os-repositories ?
lizmat ok, thanks 19:56
notable6: weekly reset
notable6 lizmat, Moved existing notes to “weekly_2019-11-04T19:56:09Z”
El_Che I have specific link for each pkg, but is less desirable than a repo
eg nxadm.github.io/rakudo-pkg/latest-...rch=x86_64
rba Regarding the downloads page on rakudo.org. Should we promote as well the rakudo releases? At the moment you will find only the rakudo star releases. 19:59
May we link to the pkgs from El_Che too? 20:00
MasterDuke i thought we already did 20:05
20:13 ufobat_ joined 20:17 ufobat left
El_Che MasterDuke: I am not a fan of the download page: 20:18
- Rakudo Star
- Installing from binaries
- Installing from source
Rakudo Star 2019.03 Errata
lizmat and another Rakudo Weekly hits the Net: rakudoweekly.blog/2019/11/04/2019-44-comma-cro/ 20:19
El_Che Star on top and at the bottom, and references to old releases while rakudo is more recent 20:20
[Coke] Is the problem that the rakudo star release is behind?
El_Che no, the problem is that unrelated things are mashed together 20:21
1. Star 2. Binaries 3. Source 4. Star errata
2nd problem is that the impression is created that the latest star release corresponds with the latest rakudo release 20:22
AlexDaniel ha, github/stripe requires a utility bill or bank statement with your address on it 20:25
which I can't provide, I don't have that, I'm not even sure what they are asking for :D 20:26
like, maybe in some countries you do receive papers where your address is specified?
MasterDuke yeah, but even if we don't actually get it in the mail you usually can see a statement online that you could print out 20:30
AlexDaniel it doesn't have my address in it 20:37
I'm not even sure if you're required to have an address at all 20:39
given e-residency and stuff like that
[Coke] hasn't had to provide proof of address except for getting other forms of ID, like a new york state driver's license. 20:42
(and then that could be used to get a passport, etc.)
AlexDaniel I did submit a pdf of a page from the government website, which does have my address listed, but doesn't have a date or signature/stamp of any sort 20:43
[Coke] mm, doesn't sound super verifiable. :)
MasterDuke has had to a couple times recently (e.g., to get driver's license realid certified, to sign up for nhs)
AlexDaniel I know one way to do it, which is to request a document from the population registry 20:44
but it's an annoying process 20:45
Kaiepi i'm trying to pass a BOOTHash for named type parameters as the last element in the list of parameters for nqp::parameterizetype for a role, but fsr it's turning into a normal Hash when it gets assigned to %named in Metamodel::ParametricRoleGroupHOW!produce_parameterization, which makes it throw and doesn't happen when you parameterize by other means 20:57
what do i need to do to make it stay a BOOTHash?
tyil I haven't ran the tests locally yet, but gitlab-ci says these should be OK (apart from kernel signals): dist.tyil.nl/raku/rakudo-star/ 21:04
if anyone wants to do a test run on them or check other things, please do
El_Che tyil++ 21:05
tyil I would like to get stuff fixed where possible, so I can do a proper release for 2019.10 later this month
I'll contact the guy that builds the .msi package as well, to see if he can get the Windows installers to work as well
vrurg Kaiepi: I would guess it became a Hash before binding to %named.
AlexDaniel I have enabled sponsorships in doc, roast, nqp, moarvm 21:06
so feel free to, I guess
vrurg Kaiepi: most likely it happens when you pass it to nqp::parameterizetype.
AlexDaniel the system is kinda weird because you have to do it in every repo separately 21:07
21:08 travis-ci joined
travis-ci Rakudo build failed. tony-o 'Add to yaml as GH approves more folks.' 21:08
travis-ci.org/rakudo/rakudo/builds/607303410 github.com/rakudo/rakudo/compare/3...c69e1f6a81
21:08 travis-ci left
lizmat restarted the one failing job 21:10
tyil hmm, Linenoise fails to install for me with rakudo-star-2019.07.1 21:20
[Linenoise] Failed to open file /home/tyil/projects/personal/rakudo-star/work/build/rakudo-star-2019.07.1/modules/Linenoise/lib/Linenoise.pm6: No such file or directory 21:21
[Coke] AlexDaniel: any idea how to become a sponsor?
AlexDaniel [Coke]: try starting from here? github.com/sponsors
[Coke]: I think once you have it set up, you go here: github.com/sponsors 21:22
[Coke]: and click the sponsor button, then choose the dev you want to sponsor
[Coke] That's to become a sponsored person, isn't it? 21:23
MasterDuke [Coke]: the rakudo repo now has a "Sponsor" button at the top 21:24
[Coke] the "join the waitlist" buttons seem to be for if I want to get money, not give it.
AlexDaniel [Coke]: I think it's both, I just clicked “Select” on one of the tiers and it worked
e.g. here github.com/sponsors/ugexe
it could be that you don't even need to set anything up 21:25
[Coke] ok. You said you had set it up on doc, but didn't see a sponsor button there; I *do* see it on rakudo, though. Thank you!
AlexDaniel right. So which is it? You just go straight to selecting a tier and that's it? 21:28
[Coke] I clicked on the sponsor link on the project, and from there can pick a person to sponsor. 21:35
AlexDaniel ok! 21:38
Kaiepi m: my IterationBuffer:D $list := nqp::create(IterationBuffer); $list.push: nqp::hash(); say $list[0].^name 21:43
camelia 5===SORRY!5=== Error while compiling <tmp>
Could not find nqp::create, did you forget 'use nqp;' ?
at <tmp>:1
------> 3:D $list := nqp::create(IterationBuffer)7⏏5; $list.push: nqp::hash(); say $list[0].
Kaiepi m: use nqp; my IterationBuffer:D $list := nqp::create(IterationBuffer); $list.push: nqp::hash(); say $list[0].^name
camelia Hash
rba Can the sponsors beeing setup for an org too? Do I alway have to choose individual developers?
21:44 lucasb left
Kaiepi so that's how that was happening vrurg 21:44
AlexDaniel rba: yeah, it's weird. I don't see a way to sponsor an org or a repo
rba: so it's “sponsor a dev”, you just choose a person
Kaiepi m: use nqp; class NativeArray is repr<VMArray> { }; my NativeArray:D $list := nqp::create(NativeArray); nqp::push($list, nqp::hash()); say nqp::atpos($list, 0).^name 21:45
rba How do you get the money then? Do I have to tell them my bank account?
camelia BOOTHash
Kaiepi so ig i can't use IterationBuffer for this
AlexDaniel rba: yep, each dev can define tiers (like 1$ a month, 5$ a month, etc.), and you pay for it
also, “GitHub will match your sponsorship”, at least in some cases 21:47
ah, I think it's possible to configure project sponsoring 21:58
we'd just need a single rakudo account somewhere, like CommunityBridge maybe? 21:59
rba: check this out: help.github.com/en/github/building...repository 22:00
[Coke] annoying you can only have 4 devs. 22:04
Geth rakudo: 654d6a1693 | (Tom Browder)++ (committed using GitHub Web editor) | docs/rakudo-nqp-and-pod-notes.md
Correct formatting error
rakudo: aa06d32abc | (Tom Browder)++ (committed using GitHub Web editor) | docs/rakudo-nqp-and-pod-notes.md
Change IRC name to #raku-dev
rakudo: 044b339024 | (Tom Browder)++ (committed using GitHub Web editor) | docs/rakudo-nqp-and-pod-notes.md
More Raku entries
23:00 AlexDaniel left
vrurg Kaiepi: what if you add 'is raw' to IterationBuffer push var parameter? Basically, this is where you get nqp::hash() hllized itnto Hash, as it looks to me. 23:22
timotimo i think a mirrored R and another R together could be stylized into a butterfly 23:25
vrurg m: use nqp; my IterationBuffer:D $list := nqp::create(IterationBuffer); nqp::push($list, nqp::hash()); say nqp::atpos($list,0).^name 23:31
camelia BOOTHash
vrurg Kaiepi: ^^^
Kaiepi hold on, i'm waiting on rakudo to build 23:32
vrurg Kaiepi: Anyway, the biggest problem is [] because it falls back to AT-POS method. But methods don't support `is raw` in return values. This is where you'd always have your BOOTHash hllized. 23:34
Use of nqp::atpos is inevitable anyway.
Kaiepi ah, ok
vrurg And I have recompiled already. `is raw` doesn't help with the parameter. You're bound to nqp::push / nqp::atpos. 23:35
Wether it makes sense using IterationBuffer is very much doubtful then.