From cxreg@pobox.com

The following code, executed repeatedly, will produce different results,
seemingly based on the ordering of the junction​:

$ perl6 -e 'class Foo { has $.bar; has $.baz; has $.biff }; Foo.^methods.join(", ").say; say ("biff" eq any(Foo.^methods) ?? "yay" !! "nay")'
baz, bar, biff, uniq, isa, map, classify, kv, min, max, pick, flat, first, minmax, does, rotate, join, Numeric, elems, end, reduce, all, Str, postcircumfix​:<[ ]>, postcircumfix​:<{ }>, list, at_key, pairs, one, ACCEPTS, at_pos, any, sort, Seq, grep, values, none, can, reverse, keys, notdef, , BUILDALL, new, Bool, say, print, item, WALK, defined, BUILD, clone, perl, WHICH, Capture, CREATE, PARROT, bless, WHENCE, WHERE
nay

$ perl6 -e 'class Foo { has $.bar; has $.baz; has $.biff }; Foo.^methods.join(", ").say; say ("biff" eq any(Foo.^methods) ?? "yay" !! "nay")'
biff, baz, bar, ACCEPTS, at_pos, any, sort, Seq, grep, none, values, can, reverse, keys, uniq, isa, map, classify, kv, min, max, pick, flat, first, minmax, does, rotate, join, Numeric, elems, end, reduce, Str, all, list, postcircumfix​:<{ }>, postcircumfix​:<[ ]>, at_key, pairs, one, BUILDALL, new, Bool, say, print, item, WALK, defined, BUILD, clone, perl, WHICH, Capture, CREATE, PARROT, bless, WHENCE, WHERE, notdef,
yay

@coke - Status changed from 'new' to 'open'

From @coke

On Thu Jul 22 00​:45​:24 2010, cxreg@​pobox.com wrote​:

The following code, executed repeatedly, will produce different
results,
seemingly based on the ordering of the junction​:

$ perl6 -e 'class Foo { has $.bar; has $.baz; has $.biff };
Foo.^methods.join(", ").say; say ("biff" eq any(Foo.^methods) ??
"yay" !! "nay")'
baz, bar, biff, uniq, isa, map, classify, kv, min, max, pick, flat,
first, minmax, does, rotate, join, Numeric, elems, end, reduce,
all, Str, postcircumfix​:<[ ]>, postcircumfix​:<{ }>, list, at_key,
pairs, one, ACCEPTS, at_pos, any, sort, Seq, grep, values, none,
can, reverse, keys, notdef, , BUILDALL, new, Bool, say, print,
item, WALK, defined, BUILD, clone, perl, WHICH, Capture, CREATE,
PARROT, bless, WHENCE, WHERE
nay

$ perl6 -e 'class Foo { has $.bar; has $.baz; has $.biff };
Foo.^methods.join(", ").say; say ("biff" eq any(Foo.^methods) ??
"yay" !! "nay")'
biff, baz, bar, ACCEPTS, at_pos, any, sort, Seq, grep, none, values,
can, reverse, keys, uniq, isa, map, classify, kv, min, max, pick,
flat, first, minmax, does, rotate, join, Numeric, elems, end,
reduce, Str, all, list, postcircumfix​:<{ }>, postcircumfix​:<[ ]>,
at_key, pairs, one, BUILDALL, new, Bool, say, print, item, WALK,
defined, BUILD, clone, perl, WHICH, Capture, CREATE, PARROT, bless,
WHENCE, WHERE, notdef,
yay

This is now pretty consistent​:

19​:55 < [Coke]> rakudo​: class Foo { has $.bar; has $.baz; has $.biff };
  Foo.^methods.join(", ").say;
19​:56 <+p6eval> rakudo 38165a​: OUTPUT«bar, baz, biff␤»

Is this closable?

--
Will "Coke" Coleda

From @coke

On Tue Jan 10 16​:58​:10 2012, coke wrote​:

On Thu Jul 22 00​:45​:24 2010, cxreg@​pobox.com wrote​:

The following code, executed repeatedly, will produce different
results,
seemingly based on the ordering of the junction​:

$ perl6 -e 'class Foo { has $.bar; has $.baz; has $.biff };
Foo.^methods.join(", ").say; say ("biff" eq any(Foo.^methods) ??
"yay" !! "nay")'
baz, bar, biff, uniq, isa, map, classify, kv, min, max, pick, flat,
first, minmax, does, rotate, join, Numeric, elems, end, reduce,
all, Str, postcircumfix​:<[ ]>, postcircumfix​:<{ }>, list, at_key,
pairs, one, ACCEPTS, at_pos, any, sort, Seq, grep, values, none,
can, reverse, keys, notdef, , BUILDALL, new, Bool, say, print,
item, WALK, defined, BUILD, clone, perl, WHICH, Capture, CREATE,
PARROT, bless, WHENCE, WHERE
nay

$ perl6 -e 'class Foo { has $.bar; has $.baz; has $.biff };
Foo.^methods.join(", ").say; say ("biff" eq any(Foo.^methods) ??
"yay" !! "nay")'
biff, baz, bar, ACCEPTS, at_pos, any, sort, Seq, grep, none, values,
can, reverse, keys, uniq, isa, map, classify, kv, min, max, pick,
flat, first, minmax, does, rotate, join, Numeric, elems, end,
reduce, Str, all, list, postcircumfix​:<{ }>, postcircumfix​:<[ ]>,
at_key, pairs, one, BUILDALL, new, Bool, say, print, item, WALK,
defined, BUILD, clone, perl, WHICH, Capture, CREATE, PARROT, bless,
WHENCE, WHERE, notdef,
yay

This is now pretty consistent​:

19​:55 < [Coke]> rakudo​: class Foo { has $.bar; has $.baz; has $.biff };
Foo.^methods.join(", ").say;
19​:56 <+p6eval> rakudo 38165a​: OUTPUT«bar, baz, biff␤»

Is this closable?

Marking testneeded.

--
Will "Coke" Coleda

From @coke

On Sun Apr 21 22​:27​:34 2013, coke wrote​:

On Tue Jan 10 16​:58​:10 2012, coke wrote​:

On Thu Jul 22 00​:45​:24 2010, cxreg@​pobox.com wrote​:

The following code, executed repeatedly, will produce different
results,
seemingly based on the ordering of the junction​:

$ perl6 -e 'class Foo { has $.bar; has $.baz; has $.biff };
Foo.^methods.join(", ").say; say ("biff" eq any(Foo.^methods) ??
"yay" !! "nay")'
baz, bar, biff, uniq, isa, map, classify, kv, min, max, pick, flat,
first, minmax, does, rotate, join, Numeric, elems, end, reduce,
all, Str, postcircumfix​:<[ ]>, postcircumfix​:<{ }>, list, at_key,
pairs, one, ACCEPTS, at_pos, any, sort, Seq, grep, values, none,
can, reverse, keys, notdef, , BUILDALL, new, Bool, say, print,
item, WALK, defined, BUILD, clone, perl, WHICH, Capture, CREATE,
PARROT, bless, WHENCE, WHERE
nay

$ perl6 -e 'class Foo { has $.bar; has $.baz; has $.biff };
Foo.^methods.join(", ").say; say ("biff" eq any(Foo.^methods) ??
"yay" !! "nay")'
biff, baz, bar, ACCEPTS, at_pos, any, sort, Seq, grep, none, values,
can, reverse, keys, uniq, isa, map, classify, kv, min, max, pick,
flat, first, minmax, does, rotate, join, Numeric, elems, end,
reduce, Str, all, list, postcircumfix​:<{ }>, postcircumfix​:<[ ]>,
at_key, pairs, one, BUILDALL, new, Bool, say, print, item, WALK,
defined, BUILD, clone, perl, WHICH, Capture, CREATE, PARROT, bless,
WHENCE, WHERE, notdef,
yay

This is now pretty consistent​:

19​:55 < [Coke]> rakudo​: class Foo { has $.bar; has $.baz; has $.biff };
Foo.^methods.join(", ").say;
19​:56 <+p6eval> rakudo 38165a​: OUTPUT«bar, baz, biff␤»

Is this closable?

Marking testneeded.

Tests added to S12-introspection/methods.t
--
Will "Coke" Coleda

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