🦋 Welcome to the IRC channel of the core developers of the Raku Programming Language (raku.org #rakulang). This channel is logged for the purpose of history keeping about its development | evalbot usage: 'm: say 3;' or /msg camelia m: ... | Logs available at irclogs.raku.org/raku-dev/live.html | For MoarVM see #moarvm
Set by lizmat on 8 June 2022.
jdv lizmat: why isn't your last change on nqp bumped into rakudo? 02:01
ab5tract: why isn't your last change on nqp bumped into rakudo?
Geth rakudo/main: 11e81cd83b | (Stefan Seifert)++ | t/02-rakudo/13-exceptions.t
Fix syntax errors in tests obscuring the errors we're interested in

For unknown reason the EVALed strings had a trailing } so if the expected compile time error would not happen we'd get a syntax error. The test triggers just fine without this (as it's checking for the right compile time error), so fix the syntax error to reduce confusion.
07:48
rakudo/main: 4 commits pushed by (Stefan Seifert)++ 08:24
nine With these t/02-rakudo/13-exceptions.t is finally passing
ab5tract massive 09:01
jdv: It didn't seem urgent. do we bump every commit, every time? #new-to-it 09:02
lizmat nine: confirm t/02-rakudo/13-exceptions.t passing, but seems we lost t/02-rakudo/10-nqp-ops.t :-( 10:08
seems to be caused by: 10:15
m: use nqp; nqp::p6bindattrinvres(($ := 42), Int, q/$!value/, nqp::getattr(42, Int, q|$!value|))
camelia ( no output )
lizmat the $ := 42 part
my $a := 42 is fine 10:16
so it looks like the handling of the anomyous state var is to blame
finanalyst_ [Coke] please take a look at the doc-website PR. The issue - after merging the PR - is how to inform previous users to change their settings. Simplest is to delete the local storage for doc-website through the browser. 10:48
lizmat finanalyst_: isn't there some JS code that can do that automagically? 10:49
when finding incompatible data?
ab5tract > r 'say "Xabddcabaacab" ~~ m/abaacab/' 12:24
「abaacab」
ab5tract somehow this causes a segv on when compiled with optimization/without debug (but only after printing '「abaacab」') 14:05
[Coke] lizmat: it would be nice if we could as part of this deployment, wipe the setting the first time we look at it. 14:28
ab5tract also loving the C stdlib which interacts so unwholesomely with true == 1 and false == 0 14:29
Geth rakudo: AntonOks++ created pull request #5553:
Compiler.nqp: Add $rakudo-flavor
14:33
releasable6 Next release in ≈4 days and ≈3 hours. There are no known blockers. Please log your changes in the ChangeLog: github.com/rakudo/rakudo/wiki/ChangeLog-Draft 15:00
releasable6_ Next release in ≈4 days and ≈3 hours. There are no known blockers. Please log your changes in the ChangeLog: github.com/rakudo/rakudo/wiki/ChangeLog-Draft
ab5tract could definitely use some insight from an experience C hacker here: gist.github.com/ab5tract/9030eb229...ow-exactly 15:55
nine lizmat: ah, well that's exactly why I work on make test right now. It's so easy to miss a regression there 16:08
nqp-ops.t just sounds like someone where you'd expect failures with RakuaST
lizmat_ yet it only has 2 tests ? 16:09
nine I meant the name. I did look through the list of failing files and thought I knew all of them. But then I also was in quite the hurry to catch a train 16:10
ab5tract lol, the only problem with my fix was with this un-related performance optimization I decided to include with my fix :) 16:19
lizmat *phew* 16:20
ab5tract yeah, that's a big sigh of relief from me as well
theoretically it *should* work for us our variation as well, but it clearly isn't a 1:1 port 16:24
`Stage parse : 31.338` <---- haven't seen it go this fast in a week or two (it's been above 40)
lizmat well, I guess we should start with merging the memmem fix? 16:25
ab5tract my double-check run of spectest is almost finished 16:28
yeah, we're good
lizmat ok, so I can merge the PR ? 16:30
ab5tract please go ahead and merge github.com/MoarVM/MoarVM/pull/1798 16:31
it's honestly quite wild that no issues with this have emerged until now 16:32
lizmat yeah :-)
ab5tract I guess that says something about the robustness of the overall algorithm
now I may need to play around with the broken version and EVAL in lldb and come to terms with whatever the hell was happening there 16:35
Geth nqp/main: 84c3a9bd5c | (Elizabeth Mattijsen)++ | tools/templates/MOAR_REVISION
Bump MoarVM to get weird pattern bug fix

  ab5tract++ for chasing this one down!
16:37
lizmat so far, core compilation parse time appears unaffected to me 16:42
ab5tract Curious, but not worrying 16:46
Though obviously I would have been happy to deliver a 25% speed up ;) 16:47
Geth rakudo/main: f56a39731c | (Elizabeth Mattijsen)++ | tools/templates/NQP_REVISION
Bump NQP to get weird pattern bug fix

  ab5tract++ for chasing this one down!
nine ab5tract++ # epic work! 17:46
ab5tract :D 17:55
Geth rakudo/main: 7d855c4be0 | (Stefan Seifert)++ | src/Raku/ast/variable-declaration.rakumod
RakuAST: fix anonymous vars losing their bindability

Forgot to initialize the new VarDeclaration::Simple attributes in this subclass that overrides the constructor.
Yet again lizmat++ for the heads-up
18:11
Geth rakudo/main: 4 commits pushed by (Stefan Seifert)++ 20:17
jdv ab5tract: its nice if those are always kept in sync for at least release reasons
i'll run a blin sometime today or tomorrow 20:18
ab5tract jdv: ack, will keep that in mind in the future 20:19
jdv thanks 20:44