00:14 wayland76 joined 00:19 kurahaupo left, kurahaupo joined 00:20 kurahaupo left 00:21 kurahaupo joined 00:23 [Coke] left 00:24 [Coke] joined 00:55 kurahaupo left
wayland76 Do we have an automated system for doing the packages/binaries? 00:58
01:07 kurahaupo joined
[Coke] github.com/rakudo/rakudo/blob/main...-binary.md 01:14
I haven't ever done it, patrickb & el_che are the owners there. 01:16
01:33 kylese left 01:34 kylese joined 01:41 eseyman left 01:43 eseyman joined
wayland76 Thanks! That's good info! 01:55
02:15 kylese left, kylese joined 02:28 abbe_ joined 02:31 abbe left, abbe_ is now known as abbe 02:39 wayland76 left 02:40 wayland76 joined
SmokeMachine lizmat: my playground goes run on the browser… all 4 runtimes 02:52
lizmat: it has no backend 02:53
03:06 kurahaupo left 03:07 kurahaupo joined 03:17 kurahaupo1 joined, kurahaupo1 left 03:18 kurahaupo1 joined 03:20 kurahaupo left 03:24 kurahaupo joined 03:25 kurahaupo1 left 03:36 kurahaupo left, kurahaupo joined 03:42 kurahaupo left 03:45 wayland76 left, wayland76 joined 03:56 kurahaupo joined
wayland76 In relation to the weekly, just a note that the creation of Qwiratry--Location--HTTP is something I'd like to undo, and people should instead use Qwiratry::Location::HTTP (which didn't get listed in the weekly). There will be tutorials coming, hopefully in the next month or so. 04:41
05:26 hwj joined 06:14 constxd left 06:15 perryprog left 06:16 abraxxa joined, perryprog joined, constxd joined 06:19 Sgeo left 06:25 abraxxa left, abraxxa joined 06:53 belluzj joined 07:21 belluzj left 07:22 belluzj joined 07:47 dakkar joined 07:55 _________ left, _________ joined 08:22 belluzj left 08:43 belluzj joined 08:44 nine joined 08:45 nine left
El_Che hi 09:56
building it now
tbrowder: building it now
10:00 sp1983 joined 10:04 sp1983 left 10:05 Sp1983_ joined 10:07 Sp1983_ left, Sp1983 joined
Sp1983 o/ 10:07
Have created a draft of the first post in DSCI series , 10:08
dev.to/sp1983/dsci-series-rakulang-ci-4472
El_Che mm, a problem 10:11
going to dev
patrickb rakudo 2026.07 precompiled archives are released.
10:11 Sp1983 left
El_Che patrickb: can you have a look at #raku-dev, in case you encountered the same error? 10:12
10:13 Sp1983 joined
Sp1983 I don’t know why is that , but looks like all my dev to post are not accessible recently … sigh 10:14
Ok. Let’s do it other way then ) github.com/melezhik/DSCI/blob/main...lang_ci.md 10:16
This should work ))
patrickb El_Che: I didn't hit an error. It went through smoothly.
Sp1983 Ok. Anyways . We have two identical post now )) 10:18
weekly: github.com/melezhik/DSCI/blob/main...lang_ci.md 10:19
notable6 Sp1983, Noted! (weekly)
Sp1983 weekly: dev.to/sp1983/dsci-series-rakulang-ci-4472
notable6 Sp1983, Noted! (weekly)
El_Che patrickb: mm, annoying
Sp1983 Are you logged in into dev.to patrickb: ? 10:20
10:20 wayland76 left, wayland76 joined
patrickb Sp1986 not at a PC atm 10:21
why?
El_Che patrickb: gist.github.com/nxadm/33d7033d04a5...d32e432846 10:22
Sp1983 PC?
Ahh
El_Che I had to add the block to update the submodules in the past because moarvm not getting them correctly in the past
just removing the block will fix it, I guess
Sp1983 Cause there is is issue that my recent devto posts are only sometimes accessible for logged in users
patrickb Sp1983 that link isn't working for me as a non logged in uses 10:23
user
Sp1983 Ok 10:24
So let’s forget about it and use gh post instead ))
patrickb El_Che: that submodule script needs more robustness work. it simply doesn't cover all corner cases. But it's very annoying work.
xinming Hi, today, I just found something strange behavior not clear to me, in CATCH block, if I do something like CATCH { when X::MyException and $str ~~ &named-regex { $<captured-from-named-rex> } } 10:25
With this example, the $<captured-from-named-rx> is Any, but the 'when' block is matched.
If I change code to CATCH { when X::MyException { if $str ~~ $named-regex { $<captured-from-named-rx>.say } } } 10:26
The latter example, the $<captured-from-named-rx> worked as expected.
El_Che patrickb: mm, I clone the repo with --recurse-submodules and if not head, I do a checkout to the specific release tag wit --recurse-submodules
remove the --recurse submodules from the clone?
xinming So I got curious about why 'when' block and if block work differently in this case.
I'm using 2026.01 10:27
10:29 Sp1983 left 10:38 Sp1983 joined, Sp1983 left 10:48 nine joined 11:00 Sp1983 joined
Sp1983 . 11:00
11:00 Sp1983 left
lizmat SmokeMachine: remove the line needing a backend 11:07
wayland76: removed mention of that module
11:09 belluzj left 11:10 belluzj joined 11:17 nine left
xinming hmm, I figured out, that 11:28
This version worked. CATCH { when $_ ~~ X::MyException and $str ~~ &named-regex { $<captured-from-named-rx> } }
But still curious, why CATCH { when X::MyException and $str ~~ &named-regex { $<captured-from-named-rx> } } doesn't work as expected. 11:29
11:29 belluzj left, belluzj joined
xinming Any hint would be appreciated. 11:30
11:43 Sp1983 joined, Sp1983 left 11:45 Sp1983 joined, Sp1983 left
lizmat s/and/&&/ ? 11:48
xinming m: my regex R { $<v> = ["a"] }; my $s = "a"; class MyE is Exception { }; given MyE { when MyE and $s ~~ &R { "e1".say; $<v>.say; } }; given MyE { when $_ ~~ MyE and $s ~~ &R { "e2".say; $<v>.raku.say; } }; 11:50
camelia e1
Nil
e2
Match.new(:orig("a"), :from(0), :pos(1))
xinming This is the minimum example I get confused.
lizmat: let me try
m: my regex R { $<v> = ["a"] }; my $s = "a"; class MyE is Exception { }; given MyE { when MyE && $s ~~ &R { "e1".say; $<v>.say; } }; given MyE { when $_ ~~ MyE && $s ~~ &R { "e2".say; $<v>.raku.say; } }; 11:51
camelia e1
Nil
e2
Match.new(:orig("a"), :from(0), :pos(1))
xinming the same result
lizmat hmmm
xinming Actually, What makes me confuse is, i think both example eval'ed to true.
But what makes the latter version will correctly initialize the $/ object. 11:52
lizmat m: $_ = 42; when 42 && 666 { .say }
camelia ( no output )
lizmat m: $_ = 666; when 42 && 666 { .say } 11:53
camelia 666
xinming Or, probably, the first example is chort-circuited
lizmat so the when smartmatches with the final result of the expression
maybe that's your confusion ?
m: say 42 && 666
camelia 666
11:58 nine joined
lizmat afk& 12:11
xinming I'm still thinking about my example, Will trouble you if I'm still in confusion :-) 12:12
El_Che patrickb: I worked around it with git checkout $MOARVM_VERSION; git submodule sync --recursive 12:15
git submodule update --init --recursive
12:23 Sp1983 joined
Sp1983 Ok, dev.to/sp1983/dsci-series-rakulang-ci-4472 appears to work again . Looks like weird bug when freshly published post is not visible for awhile and then visible again 12:24
12:26 acidsys left, Sp1983 left 12:28 acidsys joined 12:49 Sp1983 joined
Sp1983 . 12:49
Small promotion on hn, hope never dies )) news.ycombinator.com/item?id=49082908 12:50
12:50 wayland76 left 12:51 wayland76 joined 12:55 Sp1983 left 13:18 Guest54 joined 13:22 nine left 13:25 Guest54 left
tbrowder i wasn't complaining, just wanted to make sure my reinstall 13:27
of rakudo-pkg is working ok
13:33 nine joined 13:37 nine left, nine joined 13:46 thundergnat joined
thundergnat Hmm. I was idly playing around with the code xinming above to see if it was maybe a precedence problem. 13:48
m: my regex R { $<v> = ["a"] }; my $s = "a"; class MyE is Exception { }; given MyE { when (MyE && ($s ~~ &R)) { "e1".say; $<v>.raku.say; } }; given MyE { when (($s ~~ &R) && MyE) { "e2".say; $<v>.raku.say; } };
camelia e1
Nil
e2
Match.new(:orig("a"), :from(0), :pos(1))
thundergnat No, same issue. But simply by swapping around the the _order_ of the e1 and e2 tests... 13:49
m: my regex R { $<v> = ["a"] }; my $s = "a"; class MyE is Exception { }; given MyE { when (($s ~~ &R) && MyE) { "e2".say; $<v>.raku.say; } }; given MyE { when (MyE && ($s ~~ &R)) { "e1".say; $<v>.raku.say; } };
camelia e2
Match.new(:orig("a"), :from(0), :pos(1))
e1
Match.new(:orig("a"), :from(0), :pos(1))
thundergnat That looks buggy to me. 13:50
xinming thundergnat: No 13:52
thundergnat No to what? 13:53
xinming I just figured out, 'when' will evalute the whole result expression, then do $_ ~~ $expression-result
But if will just check the expression-result Bool status
thundergnat: no to "precedence problem" 13:54
But 'if' will just check the expression-result Bool status
This is the subtle difference between 'when' and 'if' on condiction check, The block executing logic is not hard to understand. 13:56
thundergnat True. It isn't a precedence problem. But it shouldn't matter the order in which you evaluate all of the conditions in a high precedence and. In the above example, if the regex is evaluated last, it doesn't make the match object available to the block. If it is evaluated first, it does. 14:00
holmdunc Thank you @librasteve for your stewardship of the Weekly! 14:06
14:07 thundergnat left
El_Che tbrowder: the rakudo-pkgs have been uploaded to all the repos 14:22
[Coke] If we need to be testing those variants, let's put them in the CI. 14:25
ww
14:56 hwj left 15:29 belluzj left 15:36 El_Che left 15:37 abraxxa left, apogee_ntv left 15:38 El_Che joined, apogee_ntv joined 16:14 belluzj joined 16:31 dakkar left 16:41 hwj joined 16:47 hwj left 17:18 wayland76 left, wayland76 joined 17:38 belluzj left
tbrowder El_Che: thanks! 18:58
ok, got it, now running 2026.07, thank you! i'll try not to worry for the next release... 19:03
19:11 human-blip left 19:13 human-blip joined 19:32 human-blip left 19:34 human-blip joined 20:03 hwj joined 20:07 hwj left 20:56 Geth left 20:57 Geth joined
tbrowder i'm now having a problem wth using App::Mi6 for testing. when i run "mi6 test" with 4 test files in dir ./t it skips one of the test files. i rename it but it still gets skipped. i've checked permissions...i can manually run the test file fine. weird. any ideas appreciated. source. i'm trying prove6 now...strange: it misses the same file!! 21:33
[Coke] nqp: class EEK {method Numeric() {return 3}}; my $a:= EEK.new(); say(+$a);
camelia
We trust you have received the usual lecture from the local System
Administrator. It usually boils down to these three things:

#1) Respect the privacy of others.
#2) Think before you type.
#3) With great power comes great responsibility…
[Coke] I'm kind of sad that bug has been open for 14 years. 21:35
(you get a real error running locally)
21:43 jjido joined
ugexe Remind me about it next week and I’ll take a look 21:50
Oh I wonder if anything can even be done 21:57
Presumably it’s overriding the numeric coercion
ugexe is one phone
21:58 jjido left
tbrowder aha, this works: prove6 --lib t/*t 21:58
note the added asterisk 21:59
[Coke] yah, if "this will never work like you expect", that's acceptable. 22:37
wayland76 lizmat: Thanks! 22:55
23:45 Sgeo joined
tbrowder issue filed... 23:59

Powered by Cro and the Raku® Programming Language.

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

Thank you!