Geth roast: SirBogman++ created pull request #597:
Add a test for rakudo/rakudo#3299 Proc::Async memory corruption bug.		 04:00
|Tux| Rakudo version 2019.07.1-478-g75056dfa1 - MoarVM version 2019.07.1-317-g1b2f72e5e
csv-ip5xs0.699 - 0.705
csv-ip5xs-206.266 - 6.482
csv-parser21.001 - 21.539
csv-test-xs-200.417 - 0.437
test6.498 - 6.512
test-t1.729 - 1.829
test-t --race0.815 - 0.817
test-t-2029.040 - 29.513
test-t-20 --race8.951 - 9.270
 09:44
Geth rakudo/add-type-to-positional-in-usage: 183518a2cc | (Elizabeth Mattijsen)++ | src/core.c/Main.pm6
Following suggestion by vrurg++
 11:56
Geth rakudo/add-type-to-positional-in-usage: 8f43623c55 | (Elizabeth Mattijsen)++ | src/core.c/Main.pm6
Different approach: only mention type if there is pod
 12:35
vrurg jnthn: what causes 'Object not in set of those to guard against' error from nqp::speshguardtype? 15:31
jnthn vrurg: You're trying to put a guard on an object, but it's not one that the plugin has previously been told about, so it doesn't know what to do 15:33
vrurg jnthn: thanks! Looks like another code-digging for another couple of days for me. :) 15:35
jnthn vrurg: How did you encoutner this?
vrurg Trying to get FQN-calls work by lookups on concretizations. But the strange part is that it seemingly fails on a class taken from $?CLASS 15:37
Thep point is spesh is seeing first if it's called in a role and if not then falls back to class. In either case it must now precisely what type object the call belongs to in order to dispatch properly. 15:38
*know precisely 15:41
lucasb the same way 'perl' and 'perl5.30.x' are hard links of each other, I went to the parent directory of 'perl6' and created a 'raku' link there. Can't this be done in the 'make install' process, for this next release? This an opportunity to introduce this name earlier. Any precautions for not doing so? 16:56
people would be able to update their shebang lines earlier 16:57
there are advice out there to create a shell alias, but it doesn't work the same as a symbolic or hard link 16:58
AlexDaniel lucasb: I'd love to see something like that, yeah 18:01
lucasb: but I have no idea what's the right way to do it, especially when it comes to other systems like windows
.seen patrickb
tellable6 AlexDaniel, I saw patrickb 2019-11-12T10:26:17Z in #raku: <patrickb> Is there any documentation of how the p6c ecosystem works and is meant to work wrt versioning?
TreyHarris I'm looking again at github.com/rakudo/rakudo/issues/3296 and linked issues and something confuses the heck out of me: was there ever a version of Rakudo in 6.c or later that would have handled the 'is export' trait the same in a .pm6 file whether or not there was a package declaration of some kind? 18:40
B/c I can't see where the code as described would have ever worked 18:43
nine AlexDaniel: on Windows just copy the executable. It's just a couple of kilobytes anyway 20:16
AlexDaniel hm, indeed 20:17
ok, I filed this ticket: github.com/rakudo/rakudo/issues/3303 20:19
vrurg AlexDaniel: I'll see if can get to it this weekend. 20:24
vrurg Unless Patrick will be first. 20:24

Powered by Cro and the Raku® Programming Language.

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

Thank you!