From d.ginev@jacobs-university.de

<LlamaRider> r​: my $rex='rex'; say "Rex" ~~ m​:i/$rex/; say "Rex" ~~
m​:i/<$rex>/;
<p6eval> rakudo 11157e​: OUTPUT«｢Rex｣␤␤#<failed match>␤»

I think both should match Rex here.
Which works if we assign $rex as​:

<LlamaRider> r​: my $rex=rx​:i/rex/; say "Rex" ~~ m​:i/$rex/; say "Rex" ~~
m​:i/<$rex>/;
<p6eval> rakudo 11157e​: OUTPUT«｢Rex｣␤␤｢Rex｣␤␤»

From @pmichaud

According to S05​:220​:

  The C<​:i> (or C<​:ignorecase>) modifier causes case distinctions to be
  ignored in its lexical scope, but not in its dynamic scope. That is,
  subrules always use their own case settings.

Since <$rex> is a subrule (indirect subrule call), it has its own case
settings instead of using the ones from the dynamic scope.

Rejecting ticket.

Pm

@pmichaud - Status changed from 'new' to 'rejected'

@pmichaud - Status changed from 'rejected' to 'open'

From @pmichaud

Reopening ticket; S05​:1303 seems to indicate that strings being compiled
should honor the :ignorecase setting.

Sorry for the confusion,

Pm

From @FROGGS

perl6-m -e 'my $rex="rex"; say "Rex" ~~ m​:i/$rex/; say "Rex" ~~ m​:i/<$rex>/;'
｢Rex｣
｢Rex｣

Test​: Raku/roast@4da54d891f

@FROGGS - Status changed from 'open' to 'resolved'