MasterDuke17

Hm. After a rebase, now rakudo installs, but `MVM_JIT_EXPR_DISABLE=1` doesn't seem to make any difference for the tests (i.e., still a ton of fails).

I got distracted before I could test the other cases, but I do plan to get to that, thanks for the reminder. On Thu, Jan 26, 2023 at 4:52 AM...

It's still there with OpenJDK 20 (from the extra/jdk-openjdk 20.0.2.u9-3 package on Arch linux). ``` Stage start : 0.000 # # A fatal error has been detected by the Java...

Hm. Now the difference between my two examples above is quite a bit smaller. ~0.48s for the non-coercion signature version and ~0.55s for coercion signature version. Should I switch to...

> > Should I switch to the "cleaner" coercion signature version? > > I'd say it makes sense. Hm. Now I'm wondering if it should be `Int() $a, Int() $b`...

I realized there are already `Int:D $a, Int:D $b` and `Real:D $a, Real:D $b` multis. So it would probably make more sense to just make those coercing instead. But then...

Yeah, actually making these coercers is a decent bit slower. With this PR as is, `my $a; for ^1_000_000 -> Int $_ { $a = ~$_ div ~($_ + 1)...

The comment here ( https://github.com/Raku/nqp/blob/main/src/QRegex/NFA.nqp#L667-L673) seems relevant. On Mon, Jan 16, 2023 at 9:44 AM Patrick Böker ***@***.***> wrote: > I ran the above golf with a patched moar that...

What about if I work the regex so that a ')' isn't matched unless there's also a '(' after the 'https?://'? On Wed, Jul 27, 2016 at 1:15 PM, Moritz...

I'm sorry, I would like to, but there's no way I could get my employer to understand why they had to sign such a document.