[EXPERIMENT] the :const attribute on subroutines

[p5pRT]

Migrated from rt.perl.org#123630 (status was 'open')

Searchable as RT123630$

[p5pRT]

From @cpansprout

The ‘const’ attribute is being added as an experimental feature in perl 5.22.0.
This ticket will collect bugs and other acceptance-criteria tickets to
track before :cosnt can be considered accepted or failed in a
future version of perl.

--

Father Chrysostomos

[p5pRT]

From @cpansprout

On Mon Jan 19 20​:37​:06 2015, sprout wrote​:

The ‘const’ attribute is being added as an experimental feature in
perl 5.22.0.
This ticket will collect bugs and other acceptance-criteria tickets to
track before :cosnt can be considered accepted or failed in a
future version of perl.

This ticket has had no attention whatsoever. This means that either (a) the feature is bug-free and perfect, or that (b) nobody is using it, or that (c) those who are using it are very silent.

If a is the case, then this feature no longer needs to be experimental.

If b is the case, it should definitely remain experimental.

If c is the case, would somebody please speak up and give opinions on the feature (whether it works and is designed well, whether it is designed poorly, whether it seems like a good design but has bugs, etc.).

--

Father Chrysostomos

[p5pRT]

The RT System itself - Status changed from 'new' to 'open'

[p5pRT]

From @davidnicol

heavens, it's Warnock's dilemma, extended

On Fri, May 20, 2016 at 4​:36 PM, Father Chrysostomos via RT <perlbug-
<perlbug-followup@​perl.org>

This ticket has had no attention whatsoever. This means that either (a)
the feature is bug-free and perfect, or that (b) nobody is using it, or
that (c) those who are using it are very silent.

If a is the case, then this feature no longer needs to be experimental.

If b is the case, it should definitely remain experimental.

If c is the case, would somebody please speak up and give opinions on the
feature (whether it works and is designed well, whether it is designed
poorly, whether it seems like a good design but has bugs, etc.).

--
"Why would you want to be the last man alive on a sinking ship?" -- Elon
Musk

[p5pRT]

From @wolfsage

Sounds good to me. Though can we add something like the attached patch
to the documentation for clarity?

-- Matthew Horsfall (alh)

[p5pRT]

From @wolfsage

Inline Patch

diff --git a/pod/perlsub.pod b/pod/perlsub.pod
index 78de284..f95d5da 100644
--- a/pod/perlsub.pod
+++ b/pod/perlsub.pod
@@ -1707,11 +1707,14 @@ Perl 5.22 also introduces the experimental "const" attribute as an
 alternative.  (Disable the "experimental::const_attr" warnings if you want
 to use it.)  When applied to an anonymous subroutine, it forces the sub to
 be called when the C<sub> expression is evaluated.  The return value is
-captured and turned into a constant subroutine:
+captured and turned into a constant subroutine. If declared inside of
+a C<BEGIN> block, it may also be inlined:
 
-    my $x = 54321;
-    *INLINED = sub : const { $x };
-    $x++;
+    BEGIN {
+        my $x = 54321;
+        *INLINED = sub : const { $x };
+        $x++;
+    }

[jkeenan]

From @wolfsage

Inline Patch

diff --git a/pod/perlsub.pod b/pod/perlsub.pod
index 78de284..f95d5da 100644
--- a/pod/perlsub.pod
+++ b/pod/perlsub.pod
@@ -1707,11 +1707,14 @@ Perl 5.22 also introduces the experimental "const" attribute as an
 alternative.  (Disable the "experimental::const_attr" warnings if you want
 to use it.)  When applied to an anonymous subroutine, it forces the sub to
 be called when the C<sub> expression is evaluated.  The return value is
-captured and turned into a constant subroutine:
+captured and turned into a constant subroutine. If declared inside of
+a C<BEGIN> block, it may also be inlined:
 
-    my $x = 54321;
-    *INLINED = sub : const { $x };
-    $x++;
+    BEGIN {
+        my $x = 54321;
+        *INLINED = sub : const { $x };
+        $x++;
+    }

@wolfsage, do you still want to have this patch applied to pod/perlsub.pod? If so, could you create a p.r.? (I think that would expedite the de-experimentalize-ing of this feature.

Thank you very much.
Jim Keenan

[wolfsage]

I think this patch or something like it should be applied for sure. That whole documentation bit is a little messy IMO - it says "inlined" in a bunch of places where I don't think it means what it ... thinks it means. I do not know when I will have time to revisit it though, so expect no action from me here; apologies.

[leonerd]

Looking at this now there doesn't seem to be any reason to keep it experimental.

[rwp0]

I think this can be closed after @leonerd's 4b6ad51 commit.