Geth rakudo: ugexe++ created pull request #6506:
RakuAST: extend lexical lowering to anonymous variables and BEGIN-compiled routines
01:23
[Coke] (this hopefully will also let me do some more work on blin itself since I'm not going to have worry about leaving the box turned on) 02:51
... this seems faster. 02:55
I have half as much memory, but 4 times the cores as the other box.
gist.github.com/coke/f78decdb75da5...b66731e9db - there's the first blin run on this machine (no new failures yet on main) 03:55
04:55 Pixi` joined, ShimmerFairy left 04:56 ShimmerFairy joined 04:58 Pixi left 04:59 Pixi joined 05:00 Pixi` left
ab5tract Hash.prepend doesn’t make sense to me 09:13
It’s not ordered, so how would you prepend?
Hash.append seems to be broken: 09:17
m: my %h; dd %h.append: :test(66); dd %h
camelia {}
{}
lizmat you need to specify the key *and* one or more values 09:18
also: :test(66) is a named arg to append, that will just get eaten
ab5tract I did specify a key and a value 09:19
And I still hate %_
m: my %h; dd %h.append: %( :test(66) ); dd %h
camelia {:test(66)}
{:test(66)}
ab5tract If it isn’t ordered, append should be the same as push which should be the same as unhurt 09:20
*unshift
If they aren’t the same, what on earth are they actually doing differently? 09:21
Do any of these methods actually take a named argument?
lizmat nope 09:22
m: my %h; %h.push("a",42); dd %h; %h.push("a",666); dd %h 09:23
camelia {:a(42)}
{:a($[42, 666])}
ab5tract Then surely it would be better for them to take *%_ as a list of keys and values and not as a (nonexistent) named argument
lizmat you're misunderstanding what Hash.push/append do
m: my %h; %h.push("a",42); dd %h; %h.push("a",[666,777]); dd %h
camelia {:a(42)}
{:a($[42, [666, 777]])}
lizmat m: my %h; %h.push("a",42); dd %h; %h.append("a",[666,777]); dd %h 09:24
camelia {:a(42)}
{:a($[42, 666, 777])}
ab5tract ah, right your are
ab5tract scrambles for a hole in the rocks
09:28 finanalyst joined
ab5tract lizmat: after that embarrassing demonstration of my lack of recall, I think it makes sense that we would have prepend and unshift as well 09:42
lizmat well, in the past 14 years of doing raku, yesterday was the first day I could have used it 09:43
so it's definitely not a big thing
but for consistency's sake, it feels like an omission not to have Hash.unshift/prepnd 09:44
*prepend
ab5tract Indeed, it feels like an oversight 09:54
Geth rakudo/lizmat-30: 6f266b9cdd | (Elizabeth Mattijsen)++ | src/Raku/Actions.nqp
RakuAST: worry about unknown use adverbs in 6.e+

As a result of discussion about 'use' adverbs in
   github.com/Raku/problem-solving/issues/525
   use Test:foo:ver:bar; ok 1; # no worries
   use v6.e.PREVIEW; use Test:foo:ver:bar; ok 1;
   Potential difficulties:
   Detected unsupported adverbs in 'use' statement: :foo:bar
10:32
rakudo: lizmat++ created pull request #6508:
RakuAST: worry about unknown use adverbs in 6.e+
finanalyst .tell coleman The docs.raku.org actions seem to have stalled on June 3. The docs site 'about' shows last rendering was June 3 12:46
tellable6 finanalyst, I'll pass your message to coleman
lizmat m: my @fail; for <One Two> { require ::($_); CATCH { @fail.push($_); .resume } }; dd @fail 15:28
camelia Method BOOTException.raku not found
in block <unit> at <tmp> line 1
lizmat m: my @fail; for <One Two> { require ::($_); CATCH { @fail.push($_); .resume } }; say @fail
camelia [Could not find One in:
/home/camelia/.raku
/home/camelia/rakudo-m-inst-1/share/perl6/site
/home/camelia/rakudo-m-inst-1/share/perl6/vendor
/home/camelia/rakudo-m-inst-1/share/perl6/core
CompUnit::Repository::AbsolutePath
lizmat m: my @fail; for <One Two> -> $m { require ::($m); CATCH { @fail.push($m); .resume } }; dd @fail 15:29
camelia ["One", "One", "One", "One", "One", "One", "One", "One", "One", "Two", "Two", "Two", "Two", "Two", "Two", "Two", "Two", "Two"]
lizmat turns out backtracking through the failed require throws for not meeting the return constraint 15:31
for each of the repo objects in the "stack"
Type check failed for return value; expected CompUnit:D but got BOOTException (BOOTException) 15:32
m: my $seen; require ::("Foo"); CATCH { ++$seen; .resume }; say $seen # golf 16:34
camelia 9
Geth rakudo/lizmat-31: ccae2d0526 | (Elizabeth Mattijsen)++ | 6 files
Remove return constraints from CompUnit::Repos

If a "require ::($foo)" failed, then the initial execution error would trigger additional execution errors if .resume was called on it. One for each recursion level in $*REPO.
   my $s; require ::("Foo"); CATCH { ++$s; .resume }; say $s; # 9
This commit reduces that to 2. Not sure yet where the last additional execution error is coming from.
16:36
rakudo: lizmat++ created pull request #6509:
Remove return constraints from CompUnit::Repos
lizmat m: my $seen; for <foo bar> -> $m { require ::($m); CATCH { ++$seen; next }; say $seen # golf 16:41
camelia ===SORRY!=== Error while compiling <tmp>
Missing block
at <tmp>:1
------> TCH { ++$seen; next }; say $seen # golf<HERE><EOL>
expecting any of:
postfix
statement end
statement modifier
statement modi…
lizmat m: my $seen; for <foo bar> -> $m { require ::($m); CATCH { ++$seen; next } }; say $seen # golf
camelia 2
lizmat hmmm.... 16:42
ugexe i thought most exceptions weren't actually resumable 17:04
removing those return constraints doesn't seem like the right thing to do regardless 17:05
lizmat yeah, you're right, I already closed the PR it was adding bandaids instead of fixing the root cause
ugexe: re github.com/Raku/nqp/pull/854 that doesn't depend on anything in MoarVM, right ? 17:06
ugexe heh, i don't think so but i've totally forgot about that PR 17:07
17:08 sp1983 joined
ugexe anything i have created a PR for wouldn't depend on moarvm changes at least 17:08
lizmat ok, well, can't see why the CI turned red, so asked for a re-run 17:09
ugexe i think all nqp ci runs are red
17:10 sp1983 left
lizmat ah, that would explain :-) 17:10
17:10 sp1983 joined
ugexe yeah, i would have run a make spectest and RAKUDO_RAKUAST=1 make spectest i presume 17:11
lizmat ok, will look into it after the weekly 17:12
[Coke] nqp is red for jvm 17:13
if its red on something Moar, thats an issue 17:14
ugexe moar windows is also red 17:16
on main
lizmat notable6: weekly 17:47
notable6 lizmat, 7 notes: gist.github.com/8dfbbd296e2fda65d5...c70128e220
lizmat notable6: reset 18:24
notable6 lizmat, No notes for “reset”
lizmat notable6: weekly reset
notable6 lizmat, Moved existing notes to “weekly_2026-08-03T18:24:09Z”
18:29 sp1983 left
lizmat and yet another Rakudo Weekly News hits the Net: rakudoweekly.blog/2026/08/03/2026-...-language/ 18:58
Geth rakudo: ugexe++ created pull request #6510:
Remove bin wrapper scripts on uninstall
22:33
rakudo: ugexe++ created pull request #6511:
Make explicitly-manage return the managed string
22:50
23:37 finanalyst left