r29144 - docs/Perl6/Spec

[pugs-commits]

Author: lwall
Date: 2009-11-20 09:39:12 +0100 (Fri, 20 Nov 2009)
New Revision: 29144

Modified:
   docs/Perl6/Spec/S02-bits.pod
   docs/Perl6/Spec/S03-operators.pod
   docs/Perl6/Spec/S04-control.pod
Log:
[Specs] more constant cleanups
add 'anon' declarator in place of 'my'/'our' misuse


Modified: docs/Perl6/Spec/S02-bits.pod
===
--- docs/Perl6/Spec/S02-bits.pod2009-11-20 07:10:40 UTC (rev 29143)
+++ docs/Perl6/Spec/S02-bits.pod2009-11-20 08:39:12 UTC (rev 29144)
@@ -1492,14 +1492,11 @@
 $lay = sub of Egg {...};# of type
 $lay = sub (--> Egg) {...}; # of type
 
-but you can use a scope modifier to introduce an C prefix type:
+but you can use the C scope declarator to introduce an C prefix type:
 
-$lay = my Egg sub {...};# of type
-$hat = my Rabbit sub {...}; # of type
+$lay = anon Egg sub {...};# of type
+$hat = anon Rabbit sub {...}; # of type
 
-Because they are anonymous, you can change the C modifier to C
-without affecting the meaning.
-
 The return type may also be specified after a C<< --> >> token within
 the signature.  This doesn't mean exactly the same thing as C.
 The C type is the "official" return type, and may therefore be

Modified: docs/Perl6/Spec/S03-operators.pod
===
--- docs/Perl6/Spec/S03-operators.pod   2009-11-20 07:10:40 UTC (rev 29143)
+++ docs/Perl6/Spec/S03-operators.pod   2009-11-20 08:39:12 UTC (rev 29144)
@@ -16,7 +16,7 @@
 Created: 8 Mar 2004
 
 Last Modified: 19 Nov 2009
-Version: 177
+Version: 178
 
 =head1 Overview
 
@@ -4273,7 +4273,6 @@
 our $foo# lexically scoped alias to package variable
 has $foo# object attribute
 state $foo  # persistent lexical (cloned with closures)
-constant $foo   # "our" scoped compile-time constant
 
 Variable declarators such as C now take a I as their
 argument.  (The syntax of function signatures is described more fully in S06.)
@@ -4282,11 +4281,20 @@
 simple declaration that declares a single variable, along with its
 associated type, traits and the initializer:
 
-constant Dog $foo is woof = 123;# okay: initializes $foo to 123
-constant (Dog $foo is woof = 123);  # same thing (with explicit parens)
-constant :(Dog $foo is woof = 123); # same thing (full Signature form)
-constant (Dog $foo is woof) = 123;  # wrong: constants cannot be assigned 
to
+my Dog $foo is woof = 123;# okay: initializes $foo to 123
+my (Dog $foo is woof = 123);  # same thing (with explicit parens)
+my :(Dog $foo is woof = 123); # same thing (full Signature form)
 
+The C declarator can declare either variables or names
+as compile-time constants:
+
+constant $foo = 1;  # compile-time constant variable
+constant bar = 2;   # compile-time constant symbol
+
+Because it can declare names in "type" space, the C
+declarator may not declare using the signature, which would be
+ambiguous.
+
 Each declarator can take an initializer following an equals
 sign (which should not be confused with a normal assignment, because
 the timing of the initialization depends on the natural lifetime of the
@@ -4315,17 +4323,17 @@
 List-context pseudo-assignment is supported for simple declarations but
 not for signature defaults:
 
-constant @foo = 1,2,3;  # okay: initializes @foo to (1,2,3)
-constant (@foo = 1,2,3);# wrong: 2 and 3 are not variable names
+my @foo = 1,2,3;  # okay: initializes @foo to (1,2,3)
+my (@foo = 1,2,3);# wrong: 2 and 3 are not variable names
 
 When parentheses are omitted, you may use any infix assignment operator
 instead of C<=> as the initializer.  In that case, the left hand side of
 the infix operator will be the variable's prototype object:
 
-constant Dog $fido .= new;  # okay: a constant Dog object
-constant Dog $fido = Dog.new;   # same thing
-constant Dog $fido = $fido.new; # wrong: invalid self-reference
-constant (Dog $fido .= new);# wrong: cannot use .= with parens
+my Dog $fido .= new;  # okay: a constant Dog object
+my Dog $fido = Dog.new;   # same thing
+my Dog $fido = $fido.new; # wrong: invalid self-reference
+my (Dog $fido .= new);# wrong: cannot use .= inside signature
 
 Note that very few mutating operators make sense on a type object, however,
 since type objects are a kind of undefined object.  (Those operators with
@@ -4396,6 +4404,8 @@
 class Foo
 role Foo
 subset Foo
+enum Foo
+constant Foo
 
 and code declarators:
 
@@ -4412,6 +4422,17 @@
 
 These all have their uses and are explained in subsequent Synopses.
 
+Note that since C is parsed as a type declarator (essentially
+declaring a type with a single value), it can actually take a scope
+declarator in front:
+
+my constant companion = '

Re: "deprecated" (was Re: r29143 ...)

[Larry Wall]

On Thu, Nov 19, 2009 at 11:22:17PM -0800, Darren Duncan wrote:
> pugs-comm...@feather.perl6.nl wrote:
>> @@ -1020,22 +1018,17 @@
>>   C<< prefix: >>
>>  -Coerces to type C.  Floor semantics are used for fractional
>> -values, including strings that appear to express fractional values.
>> -That is, C must have the same result as C in all
>> -cases.  All implicit conversions to integer use the same semantics.
>> +Deprecated, use the C coercion or the C function.
>
> Why would the Perl 6 spec contain anything deprecated rather than simply 
> not having it at all?

Only because it was already implemented in earlier versions of Perl 6,
and someone might look it up here.

> Or do you mean "deprecated" as in the parser will explicitly recognize it 
> and die with a helpful error message, such as it does for Perl 5's "=~" 
> to catch brainos?

Nope, that'll tend to either work anyway with int() or fail with two terms
in a row for int $x.  It might be worth a special message I suppose.

> If you mean the latter, is there some central place (say in Synopsis 2) 
> that talks about this, such as in a "terminology" section?  "When the 
> Perl 6 spec uses the word ... it means ..."

You must think I'm writing a *real* spec.  :)

Larry

Re: "deprecated" (was Re: r29143 ...)

[Darren Duncan]

Larry Wall wrote:

On Thu, Nov 19, 2009 at 11:22:17PM -0800, Darren Duncan wrote:

pugs-comm...@feather.perl6.nl wrote:

@@ -1020,22 +1018,17 @@
  C<< prefix: >>
 -Coerces to type C.  Floor semantics are used for fractional
-values, including strings that appear to express fractional values.
-That is, C must have the same result as C in all
-cases.  All implicit conversions to integer use the same semantics.
+Deprecated, use the C coercion or the C function.
Why would the Perl 6 spec contain anything deprecated rather than simply 
not having it at all?


Only because it was already implemented in earlier versions of Perl 6,
and someone might look it up here.


Maybe it would be a good idea to add a file to the Perl 6 spec to collect things 
like this.  The file would list features that the Perl 6 spec once had but no 
longer does *and* were supported by at least one Perl 6 implementation.  Mention 
of these things can be omitted from all the other Perl 6 spec files.  And so 
people still have somewhere to look for the question "where did that go?".


Generally I'm a believer in having details relevant to the current version of a 
project be the only ones in the main files, and have details relevant only to 
past versions of the project that no longer exist be shunted to separate files. 
 As for future plans, those also maybe should be separate but there are better 
excuses to sometimes make mention of them in the main files too, since in effect 
they are part of the current version that just haven't been fleshed out yet.


-- Darren Duncan

Re: r29129 - docs/Perl6/Spec

[Thom Boyer]

pugs-comm...@feather.perl6.nl wrote:

Author: lwall
Date: 2009-11-19 05:34:29 +0100 (Thu, 19 Nov 2009)
New Revision: 29129

Modified:
   docs/Perl6/Spec/S04-control.pod
Log:
[S04] as several folks have suggested, rename "blorst" to "blast"
I'm curious about this change. I quickly figured out that "blorst" was 
derived from "BLock OR STatement" (quoting the previous rev: "In fact, 
most of these phasers will take either a block or a statement (known as 
a I in the vernacular)).


The best that I can figure for "blast" is "BLock And STatement", which 
seems to me like a less-descriptive name. And "blast" is less likely to 
google up the results I need.


So, what exactly _is_ the derivation of "blast"?
=thom

Re: r29144 - docs/Perl6/Spec

[Edwin Steiner]

Hello!

On Fri, Nov 20, 2009 at 09:39:13AM +0100, pugs-comm...@feather.perl6.nl wrote:
> -constant Dog $fido .= new;  # okay: a constant Dog object
> -constant Dog $fido = Dog.new;   # same thing
> -constant Dog $fido = $fido.new; # wrong: invalid self-reference
> -constant (Dog $fido .= new);# wrong: cannot use .= with parens
> +my Dog $fido .= new;  # okay: a constant Dog object
> +my Dog $fido = Dog.new;   # same thing
> +my Dog $fido = $fido.new; # wrong: invalid self-reference
> +my (Dog $fido .= new);# wrong: cannot use .= inside signature

s/constant Dog/Dog/? Or am I missing something?

-Edwin

r29145 - docs/Perl6/Spec/S32-setting-library

[pugs-commits]

Author: masak
Date: 2009-11-20 14:15:01 +0100 (Fri, 20 Nov 2009)
New Revision: 29145

Modified:
   docs/Perl6/Spec/S32-setting-library/Containers.pod
Log:
[S32/Containers] fixed copy/paste error

Modified: docs/Perl6/Spec/S32-setting-library/Containers.pod
===
--- docs/Perl6/Spec/S32-setting-library/Containers.pod  2009-11-20 08:39:12 UTC 
(rev 29144)
+++ docs/Perl6/Spec/S32-setting-library/Containers.pod  2009-11-20 13:15:01 UTC 
(rev 29145)
@@ -830,8 +830,10 @@
 A set of unique values.  When used as a hash always treats the set's
 values as the keys of the hash, returning C for set elements.
 See C for a container that can represent different sets
-as keys are added or deleted.
+as keys are added or deleted.  A C responds to hash operators as
+if it were a C.
 
+
 =over
 
 =item pick
@@ -839,11 +841,7 @@
  our multi method pick ( $set: Int $num = 1, Bool :$replace )
  our multi method pick ( $set: Whatever, Bool :$replace )
 
-Like an ordinary list pick, but returns keys of the bag weighted by
-values, as if the keys were replicated the number of times indicated
-by the corresponding value and then list pick used. C is the
-mutable form of C.  A C responds to hash operators as
-if it were a C.
+Works like an ordinaty list C.
 
 =back
 

r29148 - docs/Perl6/Spec

[pugs-commits]

Author: lwall
Date: 2009-11-20 16:53:45 +0100 (Fri, 20 Nov 2009)
New Revision: 29148

Modified:
   docs/Perl6/Spec/S03-operators.pod
Log:
[S03] fix fossil found by edwin.steiner++
nail down that constants are still constants


Modified: docs/Perl6/Spec/S03-operators.pod
===
--- docs/Perl6/Spec/S03-operators.pod   2009-11-20 15:41:57 UTC (rev 29147)
+++ docs/Perl6/Spec/S03-operators.pod   2009-11-20 15:53:45 UTC (rev 29148)
@@ -15,7 +15,7 @@
 
 Created: 8 Mar 2004
 
-Last Modified: 19 Nov 2009
+Last Modified: 20 Nov 2009
 Version: 178
 
 =head1 Overview
@@ -4330,7 +4330,7 @@
 instead of C<=> as the initializer.  In that case, the left hand side of
 the infix operator will be the variable's prototype object:
 
-my Dog $fido .= new;  # okay: a constant Dog object
+my Dog $fido .= new;  # okay: a Dog object
 my Dog $fido = Dog.new;   # same thing
 my Dog $fido = $fido.new; # wrong: invalid self-reference
 my (Dog $fido .= new);# wrong: cannot use .= inside signature
@@ -4428,11 +4428,20 @@
 
 my constant companion = 'Fido';
 has constant $.pi = 22/7;
-state constant $latch = snapshot();
+state constant $latch = snapshot(); # careful with this!
 
-In these cases the explicit scoping determines when the initializer is
-evaluated.
+However, the constant declarator is intended to create values the
+compiler can inline, so it always evaluates its value at BEGIN time.
+Thus, while the extra scope declarator may say where the value is
+stored and when that storage is initialized, it cannot change the value
+of that from instance to instance.  In general, if you want something
+that doesn't vary over the nomral lifetime of a scope declarator,
+initialize it to a readonly value using C<::=> rather than declaring
+it as a constant.  Then each time the scope declarator is used,
+it can initialize to a different readonly value:
 
+state $latch ::= snapshot();  # each clone gets its own value of $latch
+
 =head1 Argument List Interpolating
 
 Perl 5 forced interpolation of a function's argument list by use of

r29149 - docs/Perl6/Spec/S32-setting-library

[pugs-commits]

Author: lwall
Date: 2009-11-20 17:06:23 +0100 (Fri, 20 Nov 2009)
New Revision: 29149

Modified:
   docs/Perl6/Spec/S32-setting-library/Containers.pod
Log:
[S32/Containers] minor typos


Modified: docs/Perl6/Spec/S32-setting-library/Containers.pod
===
--- docs/Perl6/Spec/S32-setting-library/Containers.pod  2009-11-20 15:53:45 UTC 
(rev 29148)
+++ docs/Perl6/Spec/S32-setting-library/Containers.pod  2009-11-20 16:06:23 UTC 
(rev 29149)
@@ -694,11 +694,12 @@
 
 =item push
 
- our Int multi method push ( @hash: *...@values ) is export
+ our Hash multi method push ( @hash: *...@values ) is export
 
 Like hash assignment insofar as it accepts either C objects or
-alternating keys and values; however, unlike assignment, when
-a duplicate key is detected, coerces the colliding entry's value to an
+alternating keys and values; also like in that it returns the new hash;
+However, unlike assignment, when a duplicate key is detected,
+C coerces the colliding entry's value to an
 array and pushes the Pair's value onto that array.  Hence to invert
 a hash containing duplicate values without losing (associative) information,
 say:
@@ -841,7 +842,7 @@
  our multi method pick ( $set: Int $num = 1, Bool :$replace )
  our multi method pick ( $set: Whatever, Bool :$replace )
 
-Works like an ordinaty list C.
+Works like an ordinary list C.
 
 =back
 
@@ -860,7 +861,7 @@
  our multi method pick ( $bag: Int $num = 1, Bool :$replace )
  our multi method pick ( $bag: Whatever, Bool :$replace )
 
-Like an ordinary list pick, but returns keys of the bag weighted by
+Like an ordinary list C, but returns keys of the bag weighted by
 values, as if the keys were replicated the number of times indicated
 by the corresponding value and then list pick used.  C is the
 mutable form of C.  A C responds to hash operators as

r29150 - docs/Perl6/Spec

[pugs-commits]

Author: lwall
Date: 2009-11-20 17:19:09 +0100 (Fri, 20 Nov 2009)
New Revision: 29150

Modified:
   docs/Perl6/Spec/S03-operators.pod
Log:
[S03] typo


Modified: docs/Perl6/Spec/S03-operators.pod
===
--- docs/Perl6/Spec/S03-operators.pod   2009-11-20 16:06:23 UTC (rev 29149)
+++ docs/Perl6/Spec/S03-operators.pod   2009-11-20 16:19:09 UTC (rev 29150)
@@ -4435,7 +4435,7 @@
 Thus, while the extra scope declarator may say where the value is
 stored and when that storage is initialized, it cannot change the value
 of that from instance to instance.  In general, if you want something
-that doesn't vary over the nomral lifetime of a scope declarator,
+that doesn't vary over the normal lifetime of a scope declarator,
 initialize it to a readonly value using C<::=> rather than declaring
 it as a constant.  Then each time the scope declarator is used,
 it can initialize to a different readonly value:

r29151 - docs/Perl6/Spec/S32-setting-library

[pugs-commits]

Author: lwall
Date: 2009-11-20 17:21:42 +0100 (Fri, 20 Nov 2009)
New Revision: 29151

Modified:
   docs/Perl6/Spec/S32-setting-library/Containers.pod
Log:
[S32/Containers] typo


Modified: docs/Perl6/Spec/S32-setting-library/Containers.pod
===
--- docs/Perl6/Spec/S32-setting-library/Containers.pod  2009-11-20 16:19:09 UTC 
(rev 29150)
+++ docs/Perl6/Spec/S32-setting-library/Containers.pod  2009-11-20 16:21:42 UTC 
(rev 29151)
@@ -697,7 +697,7 @@
  our Hash multi method push ( @hash: *...@values ) is export
 
 Like hash assignment insofar as it accepts either C objects or
-alternating keys and values; also like in that it returns the new hash;
+alternating keys and values; also like in that it returns the new hash.
 However, unlike assignment, when a duplicate key is detected,
 C coerces the colliding entry's value to an
 array and pushes the Pair's value onto that array.  Hence to invert

r29153 - docs/Perl6/Spec

[pugs-commits]

Author: lwall
Date: 2009-11-20 18:43:16 +0100 (Fri, 20 Nov 2009)
New Revision: 29153

Modified:
   docs/Perl6/Spec/S03-operators.pod
Log:
[S03] add infix: to match infix:, suggested by TheDamian++


Modified: docs/Perl6/Spec/S03-operators.pod
===
--- docs/Perl6/Spec/S03-operators.pod   2009-11-20 17:25:59 UTC (rev 29152)
+++ docs/Perl6/Spec/S03-operators.pod   2009-11-20 17:43:16 UTC (rev 29153)
@@ -16,7 +16,7 @@
 Created: 8 Mar 2004
 
 Last Modified: 20 Nov 2009
-Version: 178
+Version: 179
 
 =head1 Overview
 
@@ -963,12 +963,12 @@
 C<< infix:<&> >>, but with left-to-right evaluation guaranteed, for use
 in guarded patterns:
 
-$target ~~ MyType also .mytest1 also .mytest2
+$target ~~ MyType also *.mytest1 also *.mytest2
 
 This is useful when later tests might throw exceptions if earlier
 tests don't pass.  This cannot be guaranteed by:
 
-$target ~~ MyType & .mytest1 & .mytest2
+$target ~~ MyType & *.mytest1 & *.mytest2
 
 =back
 
@@ -984,6 +984,24 @@
 
 =item *
 
+C<< infix: >>, short-circuit junctional or operator
+
+EXPR else EXPR else EXPR ...
+
+Can be used to construct ORed patterns with the same semantics as
+C<< infix:<|> >>, but with left-to-right evaluation guaranteed, for use
+in guarded patterns where the left argument is much more easily
+falsifiable than the right:
+
+$target ~~ *.mycheaptest else *.myexpensivetest
+
+This is also useful when you want to perform tests in order
+of safety:
+
+$target ~~ MyType else *.mysafetest else *.mydangeroustest
+
+=item *
+
 C<< infix:<^> >>, one() operator
 
 $a ^ $b ^ $c ...

r29155 - docs/Perl6/Spec

[pugs-commits]

Author: lwall
Date: 2009-11-20 21:13:08 +0100 (Fri, 20 Nov 2009)
New Revision: 29155

Modified:
   docs/Perl6/Spec/S03-operators.pod
Log:
[S03] destroy two low-powered, obscure, short-circuiting operators
in favor of one high-powered, obscure-but-transparent, S metaoperator


Modified: docs/Perl6/Spec/S03-operators.pod
===
--- docs/Perl6/Spec/S03-operators.pod   2009-11-20 17:56:13 UTC (rev 29154)
+++ docs/Perl6/Spec/S03-operators.pod   2009-11-20 20:13:08 UTC (rev 29155)
@@ -16,7 +16,7 @@
 Created: 8 Mar 2004
 
 Last Modified: 20 Nov 2009
-Version: 179
+Version: 180
 
 =head1 Overview
 
@@ -40,7 +40,7 @@
 L  Additive  + - +| +^ ~| ~^ ?| ?^
 L  Replication   x xx
 X  Concatenation ~
-X  Junctive and  & also
+X  Junctive and  &
 X  Junctive or   | ^
 L  Named unary   sleep abs sin temp let
 N  Structural infix  but does <=> leg cmp .. ..^ ^.. ^..^
@@ -953,17 +953,14 @@
 
 $a & $b & $c ...
 
-=item *
-
-C<< infix: >>, short-circuit junctional and operator
-
-EXPR also EXPR also EXPR ...
-
-Can be used to construct ANDed patterns with the same semantics as
-C<< infix:<&> >>, but with left-to-right evaluation guaranteed, for use
+By default junctions are allowed to reorder the comparisons in any
+order that makes sense to the optimizer.  To suppress this, use
+the C metaoperator for force sequential evaluation, which will
+construct a list of ANDed patterns with the same semantics as C<<
+infix:<&> >>, but with left-to-right evaluation guaranteed, for use
 in guarded patterns:
 
-$target ~~ MyType also *.mytest1 also *.mytest2
+$target ~~ MyType S& *.mytest1 S& *.mytest2
 
 This is useful when later tests might throw exceptions if earlier
 tests don't pass.  This cannot be guaranteed by:
@@ -984,21 +981,20 @@
 
 =item *
 
-C<< infix: >>, short-circuit junctional or operator
-
-EXPR else EXPR else EXPR ...
-
-Can be used to construct ORed patterns with the same semantics as
-C<< infix:<|> >>, but with left-to-right evaluation guaranteed, for use
-in guarded patterns where the left argument is much more easily
+By default junctions are allowed to reorder the comparisons in
+any order that makes sense to the optimizer.  To suppress this,
+use the C metaoperator for force sequential evaluation, which
+will construct a list of ORed patterns with the same semantics as
+C<< infix:<|> >>, but with left-to-right evaluation guaranteed, for
+use in guarded patterns where the left argument is much more easily
 falsifiable than the right:
 
-$target ~~ *.mycheaptest else *.myexpensivetest
+$target ~~ *.mycheaptest S| *.myexpensivetest
 
 This is also useful when you want to perform tests in order
 of safety:
 
-$target ~~ MyType else *.mysafetest else *.mydangeroustest
+$target ~~ MyType S| *.mysafetest S| *.mydangeroustest
 
 =item *
 
@@ -1006,6 +1002,9 @@
 
 $a ^ $b ^ $c ...
 
+The C variant guarantees left-to-right evaluation, and in boolean
+context short-circuits to false if it sees a second match.
+
 =back
 
 =head2 Named unary precedence
@@ -3640,7 +3639,7 @@
 operators yourself.  Similarly, the carets that exclude the endpoints
 on ranges are there by convention only.
 
-In contrast to that, Perl 6 has six standard metaoperators for
+In contrast to that, Perl 6 has seven standard metaoperators for
 turning a given existing operator into a related operator that is
 more powerful (or at least differently powerful).  These differ from a
 mere naming convention in that Perl automatically generates these new
@@ -4252,6 +4251,35 @@
 
 Multidimensional lists should be handled properly.
 
+=head2 Sequential operators
+
+The sequence metaoperator, C, may be followed by any non-fiddly infix 
operator.
+It suppresses any explicit or implicit parallelism, and prevents the optimizer
+from reordering the operand evaluations.  The 'S' can be thought of as standing
+for Sequential, Serial, Synchronous, Short-circuit, Single-thread, and Safe.
+Among other things.  In particular, we can have:
+
+a S& b S& c short-circuited AND junction
+a S| b S| c short-circuited OR junction
+a S^ b S^ c short-circuited XOR junction
+a S»op« b   single-threaded hyperop
+a SX* b single-threaded X*
+a SX[*] b   single-threaded X*
+a S[X*] b   single-threaded X*
+a S+ b  suppress arg reordering by ignorant optimizer 
+
+This metaoperator has the same precedence and associativity as its base 
operator.
+The compiler is free to discard any C metaoperator that is provably 
redundant,
+such as the one in C.  The compiler is free to intuit an C on any
+operator involving known volatile operands where that does not otherwise
+change the semantics of the operator.
+
+Conjectural: since metaoperators are notionally applied from inside to outside,
+serializing a reversed opera

r29157 - docs/Perl6/Spec

[pugs-commits]

Author: lwall
Date: 2009-11-20 21:19:44 +0100 (Fri, 20 Nov 2009)
New Revision: 29157

Modified:
   docs/Perl6/Spec/S03-operators.pod
Log:
[S03] tweaks, add conjectural RSR case


Modified: docs/Perl6/Spec/S03-operators.pod
===
--- docs/Perl6/Spec/S03-operators.pod   2009-11-20 20:16:12 UTC (rev 29156)
+++ docs/Perl6/Spec/S03-operators.pod   2009-11-20 20:19:44 UTC (rev 29157)
@@ -4274,12 +4274,16 @@
 operator involving known volatile operands where that does not otherwise
 change the semantics of the operator.
 
-Conjectural: since metaoperators are notionally applied from inside to outside,
-serializing a reversed operator depends on the order of the metaoperators:
+[Conjectural: since metaoperators are notionally applied from inside
+to outside, the semantics of serializing and reversing depends on
+the order of the metaoperators:
 
 a SR/ b evaluates b, then a, then does b/a
 a RS/ b evaluates a, then b, then does b/a
+a RSR/ bevaluates b, then a, then does a/b
 
+...maybe. Can argue it all the other way too...]
+
 =head2 Nesting of metaoperators
 
 Any ordinary infix operator may be enclosed in square brackets
@@ -4300,7 +4304,7 @@
 This is convenient for function application:
 
 1,1 ... &[+]   # fibonacci series
-sort &[Rleg], @list  # reverse sort as strings
+sort &[Rleg], @list# reverse sort as strings
 
 The C<&[op]> form always refers to a binary function of the operator,
 even if it is underlyingly defined as a variadic list-associative operator.