🦋 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.
Geth File-Find: tbrowder++ created pull request #43:
Proposed update
00:21
[Coke] I'm still not able to install META6 on a fresh build of rakudo head 02:38
github.com/jonathanstowe/META6/issues/30 02:44
Geth rakudo/main: 66d53dcde2 | (Elizabeth Mattijsen)++ | src/Raku/ast/variable-declaration.rakumod
RakuAST: remove faulty debugging code

And +3 spectest files pass :-) nemokosch++ for nudge
13:49
lizmat bisectable6: my $v = Version.new("6.*"); say Version.new($v.parts.join(".")).Str 16:43
bisectable6 lizmat, Will bisect the whole range automagically because no endpoints were provided, hang tight
lizmat, ¦6c (68 commits): «6.*␤» 16:44
lizmat, Nothing to bisect!
lizmat m: my $v = Version.new("6.*"); say Version.new($v.parts.join(".")).Str
camelia 6.e.PREVIEW
lizmat meh 16:45
m: //
camelia ===SORRY!=== Error while compiling <tmp>
Null regex not allowed. Please use .comb if you wanted to produce a
sequence of characters from a string.
at <tmp>:1
------> //⏏<EOL>
lizmat bisectable: //
bisectable6 lizmat, Will bisect the whole range automagically because no endpoints were provided, hang tight
lizmat, ¦6c (68 commits): «4===SORRY!4=== Error while compiling /tmp/jS4UMFaUwI␤Null regex not allowed␤at /tmp/jS4UMFaUwI:1␤------> 3//8⏏4<EOL>␤ «exit code = 1»» 16:46
lizmat, Nothing to bisect!
lizmat looks like bisectable6 is not uptodate :-(
Geth rakudo/main: 8e394fad08 | (Elizabeth Mattijsen)++ | src/core.c/Version.pm6
Fix stringification of v6.*

Commit 4f07e0e1a9280595c9a7 removed the whatever attribute. This changed the stringification of v6.* to "v6.e.PREVIEW", which caused test issues while installing the META6 distribution.
Fix this by creating a special case for v6.*, which will compare as a v6.e.PREVIEW, but which will stringify as "v6.*". This fixes the test problem on the META6 distribution and is spectest clean.
   say v6.* cmp v6.e.PREVIEW; # Same
17:01
ugexe that seems... wrong? 17:05
why does it assume a user calling Version.new("6.*") is only ever talking about raku versions? 17:06
a module can have a version 6.9, and thus things like Version.new("6.*") cant be turned into some raku specific .PREVIEW thing 17:07
thats definitely a regression 17:09
nemokosch and anyway, will somebody keep tinkering that every time a Raku version does come out? 17:12
[Coke] ugexe: this was in response to github.com/jonathanstowe/META6/issues/30, btw. 17:22
... and now I have github.com/jonathanstowe/Test-META/issues/43 17:25
jdv: how often do you run blin? (Does anyone else run it?) 17:26
ugexe [Coke]: yeah, the commit that broke META6 and the fix are both regressions 17:29
they both turn 6.* into a version that only makes sense for internal rakudo use 17:30
m: say v5.* ~~ v5.7 17:37
camelia True
ugexe m: say v6.* ~~ v6.7
camelia False
ugexe m: say v6.7 ~~ v6.*
camelia False
ugexe m: say v5.7 ~~ v5.*
camelia True
nemokosch to be honest, v5.* ~~ v5.7 is also dubious 17:41
(*, *) ~~ (1, 2) is False, only (1, 2) ~~ (*, *) is True 17:43
ugexe github.com/rakudo/rakudo/issues/5323 19:19
Geth ¦ rakudo: lizmat self-assigned 6.* version object not represented correctly github.com/rakudo/rakudo/issues/5323 19:21
jdv [Coke]: just before release, maybe 1 to 3 times depending on breakage. i believe i am the lonely one:( 19:47
its a heavy long runner 19:49
iirc a clean run takes me 2h and its been up to 5h i think. 19:50
i run it on a aws c6i.12xlarge instance usually 19:52
coleman that's big!
how much space does the output take
jdv rakudo is not yet quite performant:( 19:53
nothing. its compute bound iirc.
jdv is on my phone away from convenient computerage 19:54
coleman copy that :)
Geth rakudo/main: 824a500b3e | (Elizabeth Mattijsen)++ | src/core.c/Version.pm6
Fix version issues when * was involved

  - define a constant star "*" so we can use =:= for comparison
  - remove use of Whatever from the Version internals, use star instead
  - add a constant candidate for Version.new("*"), same as .new(Whatever)
  - remove an unnecessary decont
  - use proper increment for clarity (no efficiency reason anymore)
Fixes #5323
20:02
[Coke] that fixed all my meta bugs, thanks. 20:27