summaryrefslogtreecommitdiff
path: root/pod/perl5111delta.pod
diff options
context:
space:
mode:
authorJesse Vincent <jesse@bestpractical.com>2009-10-18 16:21:32 -0400
committerJesse Vincent <jesse@bestpractical.com>2009-10-18 16:21:32 -0400
commit8f3d5996a665cf70e12a836b95e184e9ab628251 (patch)
tree5d91b8cc415eb6ba637bc1bf5fd6fb0084780202 /pod/perl5111delta.pod
parentc22d4cd038b9aba734088cd649559b23b16fc394 (diff)
downloadperl-8f3d5996a665cf70e12a836b95e184e9ab628251.tar.gz
Another pass at getting the perldelta for 5.11.1 in shape
Diffstat (limited to 'pod/perl5111delta.pod')
-rw-r--r--pod/perl5111delta.pod518
1 files changed, 162 insertions, 356 deletions
diff --git a/pod/perl5111delta.pod b/pod/perl5111delta.pod
index 1e9fd2c729..7e8dc8920e 100644
--- a/pod/perl5111delta.pod
+++ b/pod/perl5111delta.pod
@@ -15,32 +15,109 @@ XXX Notice
=head1 Incompatible Changes
-XXX For a release on a stable branch, this section aspires to be:
+=over
- There are no changes intentionally incompatible with 5.XXX.XXX. If any
- exist, they are bugs and reports are welcome.
+=item The boolkeys op moved to the group of hash ops. This breaks binary compatibility.
+=item C<Perl_deprecate()> was replaced with a macro that calls C<Perl_ck_warner()>
+
+C<Perl_deprecate> was not part of the public API, and did not have a C<deprecate()>
+shortcut macro defined without C<-DPERL_CORE>. Neither codesearch.google.com nor
+CPAN::Unpack show any users outside the Perl core.
+
+=back
+
=head1 Core Enhancements
-XXX New core language features go here. Summarise user-visible core language
-enhancements. Particularly prominent performance optimisations could go
-here, but most should go in the L</Performance Enhancements> section.
+=head2 Add C<package NAME VERSION> syntax
+
+ This patch adds support for setting the $VERSION of a namespace
+ when the namespace is declared with 'package'. It eliminates the
+ need for 'our $VERSION = ...' and similar constructs. E.g.
+
+ package Foo::Bar 1.23;
+ # $Foo::Bar::VERSION == 1.23
-=head1 New Platforms
+ There are several advantages to this:
+
+ * VERSION is parsed in *exactly* the same way as 'use NAME VERSION'
+
+ * $VERSION is set at compile time
+
+ * Eliminates '$VERSION = ...' and 'eval $VERSION' clutter
+
+ * As it requires VERSION to be a numeric literal or v-string
+ literal, it can be statically parsed by toolchain modules
+ without 'eval' the way MM->parse_version does for '$VERSION = ...'
+
+ * Alpha versions with underscores do not need to be quoted; static
+ parsing will preserve the underscore, but during compilation, Perl
+ will remove underscores as it does for all numeric literals
+
+ During development of this, there was discussion on #corehackers and
+ elsewhere that this should also allow other metadata to be set such as
+ "status" (stable/alpha) or "author/authority". On reflection, those
+ metadata are not very well defined yet and likely should never be
+ encoded into Perl core parsing so they can be freely changed in the
+ future. (They could perhaps be achieved via a comment on the same line
+ as 'package NAME VERSION'.)
+
+ Version numbers, however, already have a very specific definition and
+ use defined in the core through 'use NAME VERSION'. This patch merely
+ provides appropriate symmetry for setting $VERSION with the exact same
+ parsing and semantics as 'use'.
+
+ It does not break old code with only 'package NAME', but code that
+ uses 'package NAME VERSION' will need to be restricted to perl 5.11.X.
+ This is analogous to the change to open() from two-args to three-args.
+ Users requiring the latest Perl will benefit, and perhaps N years from
+ now it will become standard practice when Perl 5.12 is targeted the
+ way that 5.6 is today.
+
+ The patch does not prevent 'package NAME VERSION' from being used
+ multiple times for the same package with different version numbers, but
+ nothing prevents $VERSION from being modified arbitrarily at runtime,
+ either, so I see no urgen reason to add limitations or warnings so
+ long as Perl uses a global $VERSION variable for package version
+ numbers.
+
+ I am posting this patch to the p5p list for discussion and review. If
+ there seems to be general assent (or lack of dissent), I will go ahead
+ and commit the patch to blead.
-XXX List any platforms that this version of perl compiles on, that previous
-versions did not. These will either be enabled by new files in the F<hints/>
-directories, or new subdirectories and F<README> files at the top level of the
-source tree.
=head1 Modules and Pragmata
-XXX All changes to installed files in F<ext/> and F<lib/> go here, in a list
-ordered by distribution name. Minimally it should be the module version,
-but it's more useful to the end user to give a paragraph's summary of the
-module's changes. In an ideal world, dual-life modules would have a
-F<Changes> file that could be cribbed.
+=over 4
+
+=item Upgrade to Test-Simple-0.94
+
+=item Upgrade to Storable-2.21
+
+=item Upgrade to Pod-Simple-3.08
+
+=item Upgrade to Parse-CPAN-Meta-1.40
+
+=item Upgrade to ExtUtils-Manifest-1.57
+
+=item Upgrade to ExtUtils-CBuilder-0.260301
+
+=item Upgrade to CGI.pm-3.48
+
+=item Update CPANPLUS to CPAN version 0.89_02
+
+=item Upgrade to threads::shared 1.32
+
+=item Update ExtUtils::ParseXS to 2.21
+
+=item Upgrade File::Path to 2.08 (and add taint.t test)
+
+=item Upgrade Module::CoreList to 2.20
+
+=item Updated Object-Accessor to CPAN version 0.36
+
+=back
=head2 New Modules and Pragmata
@@ -87,21 +164,19 @@ XXX
=head1 New Documentation
-XXX Changes which create B<new> files in F<pod/> go here.
-
=over 4
-=item L<XXX>
-
-XXX
+=item L<pod/perlpolicy.pod> extends the "Social contract about contributed modules" into the beginnings of a document on Perl porting policies.
=back
=head1 Changes to Existing Documentation
-XXX Changes which significantly change existing files in F<pod/> go here.
-Any changes to F<pod/perldiag.pod> should go in L</New or Changed Diagnostics>.
+=over
+
+=item Documentation for C<$1> in perlvar.pod clarified
+=back
=head1 Performance Enhancements
@@ -110,7 +185,7 @@ may well be none in a stable release.
=over 4
-=item *
+=item C<if (%foo)> has been optimized to be faster than C<if (keys %foo)>
XXX
@@ -133,9 +208,34 @@ XXX
=over 4
-=item XXX-some-platform
+=item Darwin (Mac OS X)
+
+=over 4
+
+=item Skip testing the be_BY.CP1131 locale on Darwin 10 (Mac OS X 10.6),
+as it's still buggy.
+
+=item Correct infelicities in the regexp used to identify buggy locales
+on Darwin 8 and 9 (Mac OS X 10.4 and 10.5, respectively).
+
+=back
+
+=item DragonFly BSD
+
+=over 4
+
+=item Fix thread library selection [perl #69686]
+
+=back
+
+=item Win32
+
+=over 4
+
+=item Initial support for mingw64 is now available
+
+=back
-XXX
=back
@@ -147,9 +247,14 @@ L</Modules and Pragmata>.
=over 4
-=item *
+=item Perl now properly returns a syntax error instead of segfaulting
+if C<each>, C<keys> or C<values> is used without an argument
-XXX
+=item C<tell()> now fails properly if called without an argument and when no previous file was read
+
+C<tell()> now returns C<-1>, and sets errno to C<EBADF>, thus restoring the 5.8.x behaviour
+
+=item overload no longer implicitly unsets fallback on repeated 'use overload' lines
=back
@@ -159,9 +264,22 @@ XXX New or changed warnings emitted by the core's C<C> code go here.
=over 4
-=item C<XXX>
-XXX
+=item The 'syntax' category was removed from 5 warnings that should only be in 'deprecated'.
+
+=item Three fatal pack/unpack error messages have been normalized to "panic: %s"
+
+=item "Unicode character is illegal" has been rephrased to be more accurate
+
+It now reads C<Unicode non-character is illegal in interchange> and the
+perldiag documentation has been expanded a bit.
+
+=item Perl now defaults to issuing a warning if a deprecated language feature is used.
+
+To disable this feature in a given lexical scope, you should use C<no
+warnings 'deprecated';> For information about which language features
+are deprecated and explanations of various deprecation warnings, please
+see L<perldiag.pod>
=back
@@ -185,9 +303,14 @@ they represent may be.
=over 4
-=item Significant cleanups to core tests to ensure that language and interpreter features are not used before they're tested.
+=item Significant cleanups to core tests to ensure that language and
+interpreter features are not used before they're tested.
-XXX
+=item F<t/porting/podcheck.t> automatically checks the well-formedness of
+POD found in all .pl, .pm and .pod files in the F<MANIFEST>, other than in
+dual-lifed modules which are primarily maintained outside the Perl core.
+
+=item F<t/porting/manifest.t> now tests that all files listed in MANIFEST are present.
=back
@@ -229,6 +352,12 @@ stars here. It's probably best to group changes under the same section layout
as the main perldelta
+=head1 Errata
+
+
+k=item The Perl 5.11.0 release notes incorrectly described 'delete local'
+
+
=head1 Obituary
XXX If any significant core contributor has died, we've added a short obituary
@@ -278,103 +407,12 @@ The F<Artistic> and F<Copying> files for copyright information.
This is all changes through 704e1b1e
- Update CPANPLUS to CPAN version 0.89_02
-
- Upgrade to threads::shared 1.32
-
- Help ExtUtils::Install's tests find PERL_SRC on VMS.
-
-commit d1d15184c41c6ad4f16829561163cd118e5ae917
-Author: Nicholas Clark <nick@ccl4.org>
-Date: Tue Oct 13 16:04:18 2009 +0100
-
- Enable deprecation warnings by default.
-
- locked is deprecated, so use :lvalue instead.
-
- Can't use C<shift INC> to avoid @ in a commandline now, so use eval and octal.
-
- shift with barewords is deprecated, so this test from perl 1 needs updating.
-
- push and pop on barewords are deprecated, so need no warnings 'deprecated';
-
- Opening dirhandle DIR also as a file needs no warnings 'deprecated';
-
- defined @array and defined %hash need no warnings 'deprecated';
-
- localisation of $[ is deprecated, so needs no warnings 'deprecated';
-
- Tests for barewords and hash operators need no warnings 'deprecated';
-
- do subname() is deprecated, so tests for it need no warnings 'deprecated';
-
- Bracket deprecated features with no warnings 'deprecated';
-
- Move the test for the deprecated feature <<; out of t/base/lext.t
-
- Tests in base can't utilise pragmata, specifically no warnings 'deprecated';
-
- Add no warnings 'deprecated' to a test that assigns to $[
commit a44d0896a6c4bfe01ea532694b8c1c073ea6a2f1
Author: Nicholas Clark <nick@ccl4.org>
Date: Thu Oct 15 23:37:41 2009 +0100
- Skip testing the be_BY.CP1131 locale on Darwin 10, as it's still buggy.
-
- Correct infelicities in the regexp used to identify buggy locales on Darwin 8
- and 9.
-
- POSIX::strftime() should be able to handle Unicode characters in the format
- string.
-
-commit 2e0eeeaafce11cb0128a6d1e245f1a5b806e3a87
-Author: Nicholas Clark <nick@ccl4.org>
-Date: Thu Oct 15 15:12:49 2009 +0100
-
- Move the boolkeys op to the group of hash ops.
-
- This breaks binary compatibility.
-
-commit 867fa1e2da145229b4db2c6e8d5b51700c15f114
-Author: demerphq <demerphq@gmail.com>
-Date: Thu Oct 15 14:27:30 2009 +0100
-
- Optimise if (%foo) to be faster than if(keys %foo)
-
- Thread was "[PATCH] Make if (%hash) {} act the same as if (keys %hash) {}"
- http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/2006-11/msg00432.html
- but the implementation evolved from the approach described in the subject, to
- instead add a new opcode pp_boolkeys, to exactly preserve the existing
- behaviour.
-
- Various conflicts with the passage of time resolved, 'register' removed, and a
- $VERSION bump.
-
-commit 1c85afcecc8ee030e2780aa5bfa85692c8db64df
-Author: demerphq <demerphq@gmail.com>
-Date: Thu Oct 15 14:22:47 2009 +0100
-
- Support for pp_boolkeys in B::Deparse.
-
- Part of "[PATCH] Make if (%hash) {} act the same as if (keys %hash) {}"
- http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/2006-11/msg00432.html
- which evolved from the approach described in the subject, to instead add a new
- opcode pp_boolkeys, to exactly preserve the existing behaviour.
-
- Plus a $VERSION bump.
-
-commit 55b77936aef50881a71470fd06e66edffd66d9de
-Author: Sisyphus <sisyphus1@optusnet.com.au>
-Date: Wed Oct 14 10:02:16 2009 -0400
-
- Add mingw64 support for win32
-
-commit d4c22fec77d7244882ce42a93a4ad25bdada2519
-Author: Sisyphus <sisyphus1@optusnet.com.au>
-Date: Wed Oct 14 06:58:49 2009 -0400
-
- Patch t/win32/system.t for mingw32/64
+ POSIX::strftime() should be able to handle Unicode characters in the format string.
commit e4d771f5006ebd70b76422437cce60e9ac40c830
Author: Jan Dubois <jand@activestate.com>
@@ -395,243 +433,11 @@ Date: Tue Oct 13 16:46:58 2009 -0700
-
- Podify the social contract about contributed module. Turn it into a policy document. Move the new "perl policy" document into pod/
-
-
-Author: Smylers <Smylers@stripey.com>
-Date: Tue Oct 13 14:14:46 2009 +0200
-
- perlvar $1 clarification
-
- $1 is currently documented as being set by the "last pattern match".
- But it is left alone by unsuccessful pattern match attempts (continuing
- to hold a value from an earlier successful match).
-
- Saying "last successful pattern match" clarifies this; it's also the
- phrase used to document $&.
-
- Second, the entry for $1 in perlvar doesn't actually contain the text
- "$1" anywhere. As such, doing man perlvar then using /\$1 to search for
- it in less doesn't locate it (though does match other places in that
- file where $1 happens to be used).
-
-
- =item *
-
- Replace Perl_deprecate() with a macro that calls Perl_ck_warner()
- Perl_deprecate was not part of the public API, and did not have a deprecate()
- shortcut macro defined without -DPERL_CORE. Neither codesearch.google.com nor
- CPAN::Unpack show any users outside the core.
-
-
-
- =item *
- Remove category 'syntax' from 5 warnings that should just be in 'deprecated'.
-
- None were documented as also being in 'syntax'. Effectively, this completes the
- reorganisation of commits 12bcd1a617c74d6e and 9014280dc8264580. See
- http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/2009-10/msg00601.html and
- http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/2002-03/msg00850.html
-
-
-
-commit c989e6a3e4b89b26d315693449c76cdcb754611f
-Author: Geoffrey T. Dairiki <dairiki at dairiki.org>
-Date: Tue Aug 4 17:54:34 2009 -0700
-
- overload no longer implicitly unsets fallback on repeated 'use overload' lines - Fix for RT#68916
-
-
-commit 5f5991a0d6d8ef99d2643b88a7d9285e35277331
-Author: Nicholas Clark <nick@ccl4.org>
-Date: Mon Oct 12 15:54:30 2009 +0100
-
- Normalise 3 fatal pack/unpack error messages to "panic: %s"
-
-commit 04e82a462b85b3d6265b04aa07a405316616dc66
-Author: Nicholas Clark <nick@ccl4.org>
-Date: Sun Oct 11 18:56:01 2009 +0100
-
- porting/manifest.t now tests that all files listed in MANIFEST are present.
-
-commit 41239ce77fcd273e18c4017d3d96a5f42e228594
-Author: Nicholas Clark <nick@ccl4.org>
-Date: Sun Oct 11 17:28:02 2009 +0100
-
- Default to running tests in cpan/ with relative paths for perl and @INC
-
- Explicitly turn paths absolute for the 33 extensions in cpan/ that fail tests
- with relative paths.
-
-commit fc5e5837c991d3d3224259ff5c1d728d4e0636e2
-Author: Nicholas Clark <nick@ccl4.org>
-Date: Sun Oct 11 15:05:58 2009 +0100
-
- MakeMaker::Test::Utils::perl_lib now copes with relative paths for core testing.
-
- In the core, @INC already contains the moral equivalent of blib/lib. However,
- it's a relative path (by default), so make it absolute. It's easier to KISS if
- this is done *before* any change of directory, so document this, and change the
- non-core case to add the absolute path of 'blib/lib' to @INC, rather than the
- absolute path of '../blib/lib'.
-
-
- Upgrade to Test-Simple-0.94
-
- Upgrade to Storable-2.21
-
- Upgrade to Pod-Simple-3.08
-
- Upgrade to Parse-CPAN-Meta-1.40
-
- Upgrade to ExtUtils-Manifest-1.57
-
- Upgrade to ExtUtils-CBuilder-0.260301
-
-
- Upgrade to CGI.pm-3.48
-
-commit 6c1b5ced18901286f16f6d5f6914ba4b5e3db601
-Author: Simon 'corecode' Schubert <corecode@fs.ei.tum.de>
-Date: Sat Oct 10 17:55:26 2009 +0100
-
- [perl #69686] Fix DragonFly thread lib selection
-
-
-
-Merge: 8e32470 e92f586
-Author: Rafael Garcia-Suarez <rgs@consttype.org>
-Date: Thu Oct 8 15:13:18 2009 +0200
-
- Merge branch 'feature/package-name-version' into blead
-
-commit f03173f2c707a804ec3e9c291d2ab1adb9db4abc
-Author: Rafael Garcia-Suarez <rgs@consttype.org>
-Date: Thu Oct 8 11:44:21 2009 +0200
-
- Make tell() fail properly if called without an argument and when no previous file was read
-
- This makes tell() return -1, and sets errno to EBADF, thus
- restoring the 5.8.x behaviour
-
-commit a916b30221d5aac718ed67c9a5bc9c0905daddd0
-Author: Rafael Garcia-Suarez <rgs@consttype.org>
-Date: Thu Oct 8 11:33:06 2009 +0200
-
- Properly return a syntax error instead of segfaulting if each/keys/values is used without an argument
-
- =item Updated Object-Accessor to CPAN version 0.36
-
-
Add perltodo: write an XS cookbook
- Porting/checkAUTHORS.pl now runs clean
-
-commit 6fa4d285bff5644bebb95aff09143322042282cc
-Author: David Golden <dagolden@cpan.org>
-Date: Tue Oct 6 06:48:48 2009 -0400
-
- Add 'package NAME VERSION' syntax
-
- This patch adds support for setting the $VERSION of a namespace
- when the namespace is declared with 'package'. It eliminates the
- need for 'our $VERSION = ...' and similar constructs. E.g.
-
- package Foo::Bar 1.23;
- # $Foo::Bar::VERSION == 1.23
-
- There are several advantages to this:
-
- * VERSION is parsed in *exactly* the same way as 'use NAME VERSION'
-
- * $VERSION is set at compile time
-
- * Eliminates '$VERSION = ...' and 'eval $VERSION' clutter
-
- * As it requires VERSION to be a numeric literal or v-string
- literal, it can be statically parsed by toolchain modules
- without 'eval' the way MM->parse_version does for '$VERSION = ...'
-
- * Alpha versions with underscores do not need to be quoted; static
- parsing will preserve the underscore, but during compilation, Perl
- will remove underscores as it does for all numeric literals
-
- During development of this, there was discussion on #corehackers and
- elsewhere that this should also allow other metadata to be set such as
- "status" (stable/alpha) or "author/authority". On reflection, those
- metadata are not very well defined yet and likely should never be
- encoded into Perl core parsing so they can be freely changed in the
- future. (They could perhaps be achieved via a comment on the same line
- as 'package NAME VERSION'.)
-
- Version numbers, however, already have a very specific definition and
- use defined in the core through 'use NAME VERSION'. This patch merely
- provides appropriate symmetry for setting $VERSION with the exact same
- parsing and semantics as 'use'.
-
- It does not break old code with only 'package NAME', but code that
- uses 'package NAME VERSION' will need to be restricted to perl 5.11.X.
- This is analogous to the change to open() from two-args to three-args.
- Users requiring the latest Perl will benefit, and perhaps N years from
- now it will become standard practice when Perl 5.12 is targeted the
- way that 5.6 is today.
-
- The patch does not prevent 'package NAME VERSION' from being used
- multiple times for the same package with different version numbers, but
- nothing prevents $VERSION from being modified arbitrarily at runtime,
- either, so I see no urgen reason to add limitations or warnings so
- long as Perl uses a global $VERSION variable for package version
- numbers.
-
- I am posting this patch to the p5p list for discussion and review. If
- there seems to be general assent (or lack of dissent), I will go ahead
- and commit the patch to blead.
-
-commit a67b1afafddaaed84e79a867acc888c9ccfb6460
-Author: Max Maischein <corion@corion.net>
-Date: Mon Oct 5 22:49:09 2009 +0200
-
-New porting tests to automatically check POD in lib/, ext/ and pod/
-
-commit 2d5f1d01166a325b29ccc86102d26e68def13786
-Author: David Golden <dagolden@cpan.org>
-Date: Mon Oct 5 17:46:36 2009 -0400
-
Explain using git send-email for patches
- =item Update ExtUtils::ParseXS to 2.21
-
- =item Upgrade File::Path to 2.08 (and add taint.t test)
-commit b2680017d861a93d2d51b07bce2f1731086bc8c3
-Author: Yves Orton <demerphq@gmail.com>
-Date: Mon Oct 5 09:34:52 2009 +0200
-
- in regexec.c move the BOUND logic out of the way of the special CC logic
-
- This is a first step towards macroizing the special CC handler logic so
- it is easier to maintain them, for instance interestng optimisations are
- being used in one, but not all, even though the logic is sharable. By
- moving the BOUND logic out of the way the code repition is much clearer.
-
-commit e74a3e73f5e128a77b691fcfc83214f58419a493
-Author: Jesse Vincent <jesse@bestpractical.com>
-Date: Sun Oct 4 05:22:17 2009 +0900
-
- Correcting mistaken description of 'delete local' in perl5110delta
-
- Reported by nothingmuch++ and rjbs++
=TODO FOR RELENG GUIDE Make Module::Corelist recognise $] as a version number on 5.11.0
- =item Bump version to 2.20
-
-commit 6f6ac1dea8501596050bc974dc468632797d51eb
-Author: Rafael Garcia-Suarez <rgs@consttype.org>
-Date: Sat Oct 3 09:56:42 2009 +0200
-
- Change warning "Unicode character is illegal" to more accurate description
-
- That now reads "Unicode non-character is illegal in interchange" and the
- perldiag documentation is expanded a bit.