3 perldelta - what is new for perl v5.8.0
7 This document describes differences between the 5.6.0 release
10 Many of the bug fixes in 5.8.0 were already seen in the 5.6.1
11 maintenance release since the two releases were kept closely
14 If you are upgrading from Perl 5.005_03, you might also want
15 to read L<perl56delta>.
17 =head1 Highlights In 5.8.0
23 Better Unicode support
27 New Thread Implementation
35 Better Numeric Accuracy
43 More Extensive Regression Testing
47 =head1 Incompatible Changes
49 =head2 64-bit platforms and malloc
51 If your pointers are 64 bits wide, the Perl malloc is no longer being
52 used because it does not work well with 8-byte pointers. Also,
53 usually the system mallocs on such platforms are much better optimized
54 for such large memory models than the Perl malloc. Some memory-hungry
55 Perl applications like the PDL don't work well with Perl's malloc.
56 Finally, other applications than Perl (like modperl) tend to prefer
57 the system malloc. Such platforms include Alpha and 64-bit HPPA,
60 =head2 AIX Dynaloading
62 The AIX dynaloading now uses in AIX releases 4.3 and newer the native
63 dlopen interface of AIX instead of the old emulated interface. This
64 change will probably break backward compatibility with compiled
65 modules. The change was made to make Perl more compliant with other
66 applications like modperl which are using the AIX native interface.
68 =head2 Socket Extension Dynamic in VMS
70 The Socket extension is now dynamically loaded instead of being
71 statically built in. This may or may not be a problem with ancient
72 TCP/IP stacks of VMS: we do not know since we weren't able to test
73 Perl in such configurations.
75 =head2 IEEE-format Floating Point Default on OpenVMS Alpha
77 Perl now uses IEEE format (T_FLOAT) as the default internal floating
78 point format on OpenVMS Alpha, potentially breaking binary compatibility
79 with external libraries or existing data. G_FLOAT is still available as
80 a configuration option. The default on VAX (D_FLOAT) has not changed.
82 =head2 Different Definition of the Unicode Character Classes \p{In...}
84 As suggested by the Unicode consortium, the Unicode character classes
85 now prefer I<scripts> as opposed to I<blocks> (as defined by Unicode);
86 in Perl, when the C<\p{In....}> and the C<\p{In....}> regular expression
87 constructs are used. This has changed the definition of some of those
90 The difference between scripts and blocks is that scripts are the
91 glyphs used by a language or a group of languages, while the blocks
92 are more artificial groupings of 256 characters based on the Unicode
95 In general this change results in more inclusive Unicode character
96 classes, but changes to the other direction also do take place:
97 for example while the script C<Latin> includes all the Latin
98 characters and their various diacritic-adorned versions, it
99 does not include the various punctuation or digits (since they
100 are not solely C<Latin>).
102 Changes in the character class semantics may have happened if a script
103 and a block happen to have the same name, for example C<Hebrew>.
104 In such cases the script wins and C<\p{InHebrew}> now means the script
105 definition of Hebrew. The block definition in still available,
106 though, by appending C<Block> to the name: C<\p{InHebrewBlock}> means
107 what C<\p{InHebrew}> meant in perl 5.6.0. For the full list
108 of affected character classes, see L<perlunicode/Blocks>.
110 =head2 Perl Parser Stress Tested
112 The Perl parser has been stress tested using both random input and
113 Markov chain input and the few found crashes and lockups have been
116 =head2 REF(...) Instead Of SCALAR(...)
118 A reference to a reference now stringifies as "REF(0x81485ec)" instead
119 of "SCALAR(0x81485ec)" in order to be more consistent with the return
128 The semantics of bless(REF, REF) were unclear and until someone proves
129 it to make some sense, it is forbidden.
133 The obsolete chat2 library that should never have been allowed
134 to escape the laboratory has been decommissioned.
138 The very dusty examples in the eg/ directory have been removed.
139 Suggestions for new shiny examples welcome but the main issue is that
140 the examples need to be documented, tested and (most importantly)
145 The (bogus) escape sequences \8 and \9 now give an optional warning
146 ("Unrecognized escape passed through"). There is no need to \-escape
151 The list of filenames from glob() (or <...>) is now by default sorted
152 alphabetically to be csh-compliant (which is what happened before
153 in most UNIX platforms). (bsd_glob() does still sort platform
154 natively, ASCII or EBCDIC, unless GLOB_ALPHASORT is specified.)
158 Spurious syntax errors generated in certain situations, when glob()
159 caused File::Glob to be loaded for the first time, have been fixed.
163 Although "you shouldn't do that", it was possible to write code that
164 depends on Perl's hashed key order (Data::Dumper does this). The new
165 algorithm "One-at-a-Time" produces a different hashed key order.
166 More details are in L</"Performance Enhancements">.
170 lstat(FILEHANDLE) now gives a warning because the operation makes no sense.
171 In future releases this may become a fatal error.
175 The C<package;> syntax (C<package> without an argument) has been
176 deprecated. Its semantics were never that clear and its
177 implementation even less so. If you have used that feature to
178 disallow all but fully qualified variables, C<use strict;> instead.
182 The unimplemented POSIX regex features [[.cc.]] and [[=c=]] are still
183 recognised but now cause fatal errors. The previous behaviour of
184 ignoring them by default and warning if requested was unacceptable
185 since it, in a way, falsely promised that the features could be used.
189 The current user-visible implementation of pseudo-hashes (the weird
190 use of the first array element) is deprecated starting from Perl 5.8.0
191 and will be removed in Perl 5.10.0, and the feature will be
192 implemented differently. Not only is the current interface rather
193 ugly, but the current implementation slows down normal array and hash
194 use quite noticeably. The C<fields> pragma interface will remain
199 The syntaxes C<< @a->[...] >> and C<< %h->{...} >> have now been deprecated.
203 After years of trying the suidperl is considered to be too complex to
204 ever be considered truly secure. The suidperl functionality is likely
205 to be removed in a future release.
209 The long deprecated uppercase aliases for the string comparison
210 operators (EQ, NE, LT, LE, GE, GT) have now been removed.
214 The tr///C and tr///U features have been removed and will not return;
215 the interface was a mistake. Sorry about that. For similar
216 functionality, see pack('U0', ...) and pack('C0', ...).
220 =head1 Core Enhancements
222 =head2 PerlIO is Now The Default
228 IO is now by default done via PerlIO rather than system's "stdio".
229 PerlIO allows "layers" to be "pushed" onto a file handle to alter the
230 handle's behaviour. Layers can be specified at open time via 3-arg
233 open($fh,'>:crlf :utf8', $path) || ...
235 or on already opened handles via extended C<binmode>:
237 binmode($fh,':encoding(iso-8859-7)');
239 The built-in layers are: unix (low level read/write), stdio (as in
240 previous Perls), perlio (re-implementation of stdio buffering in a
241 portable manner), crlf (does CRLF <=> "\n" translation as on Win32,
242 but available on any platform). A mmap layer may be available if
243 platform supports it (mostly UNIXes).
245 Layers to be applied by default may be specified via the 'open' pragma.
247 See L</"Installation and Configuration Improvements"> for the effects
248 of PerlIO on your architecture name.
252 File handles can be marked as accepting Perl's internal encoding of Unicode
253 (UTF-8 or UTF-EBCDIC depending on platform) by a pseudo layer ":utf8" :
255 open($fh,">:utf8","Uni.txt");
257 Note for EBCDIC users: the pseudo layer ":utf8" is erroneously named
258 for you since it's not UTF-8 what you will be getting but instead
259 UTF-EBCDIC. See L<perlunicode>, L<utf8>, and
260 http://www.unicode.org/unicode/reports/tr16/ for more information.
261 In future releases this naming may change.
265 File handles can translate character encodings from/to Perl's internal
266 Unicode form on read/write via the ":encoding()" layer.
270 File handles can be opened to "in memory" files held in Perl scalars via:
272 open($fh,'>', \$variable) || ...
276 Anonymous temporary files are available without need to
277 'use FileHandle' or other module via
279 open($fh,"+>", undef) || ...
281 That is a literal undef, not an undefined value.
285 The list form of C<open> is now implemented for pipes (at least on UNIX):
287 open($fh,"-|", 'cat', '/etc/motd')
289 creates a pipe, and runs the equivalent of exec('cat', '/etc/motd') in
294 =head2 Signals Are Now Safe
296 Perl used to be fragile in that signals arriving at inopportune moments
297 could corrupt Perl's internal state. Now Perl postpones handling of
298 signals until it's safe.
300 =head2 Unicode Overhaul
302 Unicode in general should be now much more usable than in Perl 5.6.0
303 (or even in 5.6.1). Unicode can be used in hash keys, Unicode in
304 regular expressions should work now, Unicode in tr/// should work now,
305 Unicode in I/O should work now.
311 The Unicode Character Database coming with Perl has been upgraded
312 to Unicode 3.1.1. For more information, see http://www.unicode.org/.
316 For developers interested in enhancing Perl's Unicode capabilities:
317 almost all the UCD files are included with the Perl distribution in
318 the lib/unicore subdirectory. The most notable omission, for space
319 considerations, is the Unihan database.
323 The Unicode character classes \p{Blank} and \p{SpacePerl} have been
324 added. "Blank" is like C isblank(), that is, it contains only
325 "horizontal whitespace" (the space character is, the newline isn't),
326 and the "SpacePerl" is the Unicode equivalent of C<\s> (\p{Space}
327 isn't, since that includes the vertical tabulator character, whereas
332 =head2 Understanding of Numbers
334 In general a lot of fixing has happened in the area of Perl's
335 understanding of numbers, both integer and floating point. Since in
336 many systems the standard number parsing functions like C<strtoul()>
337 and C<atof()> seem to have bugs, Perl tries to work around their
338 deficiencies. This results hopefully in more accurate numbers.
340 Perl now tries internally to use integer values in numeric conversions
341 and basic arithmetics (+ - * /) if the arguments are integers, and
342 tries also to keep the results stored internally as integers.
343 This change leads to often slightly faster and always less lossy
344 arithmetics. (Previously Perl always preferred floating point numbers
347 =head2 Miscellaneous Enhancements
353 AUTOLOAD is now lvaluable, meaning that you can add the :lvalue attribute
354 to AUTOLOAD subroutines and you can assign to the AUTOLOAD return value.
358 C<perl -d:Module=arg,arg,arg> now works (previously one couldn't pass
359 in multiple arguments.)
363 END blocks are now run even if you exit/die in a BEGIN block.
364 Internally, the execution of END blocks is now controlled by
365 PL_exit_flags & PERL_EXIT_DESTRUCT_END. This enables the new
366 behaviour for Perl embedders. This will default in 5.10. See
371 Formats now support zero-padded decimal fields.
375 Lvalue subroutines can now return C<undef> in list context.
376 However, the lvalue subroutine feature still remains experimental.
380 A new special regular expression variable has been introduced:
381 C<$^N>, which contains the most-recently closed group (submatch).
385 C<no Module;> now works even if there is no "sub unimport" in the Module.
389 The numerical comparison operators return C<undef> if either operand
390 is a NaN. Previously the behaviour was unspecified.
394 The following builtin functions are now overridable: each(), keys(),
395 pop(), push(), shift(), splice(), unshift().
399 C<pack('U0a*', ...)> can now be used to force a string to UTF8.
403 my __PACKAGE__ $obj now works.
407 The printf() and sprintf() now support parameter reordering using the
408 C<%\d+\$> and C<*\d+\$> syntaxes. For example
410 print "%2\$s %1\$s\n", "foo", "bar";
412 will print "bar foo\n". This feature helps in writing
413 internationalised software, and in general when the order
414 of the parameters can vary.
418 prototype(\&) is now available.
422 prototype(\[$@%&]) is now available to implicitly create references
423 (useful for example if you want to emulate the tie() interface).
427 untie() will now call an UNTIE() hook if it exists. See L<perltie>
432 L<utime> now supports C<utime undef, undef, @files> to change the
433 file timestamps to the current time.
437 The rules for allowing underscores (underbars) in numeric constants
438 have been relaxed and simplified: now you can have an underscore
439 simply B<between digits>.
443 =head1 Modules and Pragmata
445 =head2 New Modules and Pragmata
451 C<Attribute::Handlers> allows a class to define attribute handlers.
454 use Attribute::Handlers;
455 sub Wolf :ATTR(SCALAR) { print "howl!\n" }
457 # later, in some package using or inheriting from MyPack...
459 my MyPack $Fluffy : Wolf; # the attribute handler Wolf will be called
461 Both variables and routines can have attribute handlers. Handlers can
462 be specific to type (SCALAR, ARRAY, HASH, or CODE), or specific to the
463 exact compilation phase (BEGIN, CHECK, INIT, or END).
467 B<B::Concise> is a new compiler backend for walking the Perl syntax
468 tree, printing concise info about ops, from Stephen McCamant. The
469 output is highly customisable. See L<B::Concise>.
473 C<Class::ISA> for reporting the search path for a class's ISA tree,
474 by Sean Burke, has been added. See L<Class::ISA>.
478 C<Cwd> has now a split personality: if possible, an XS extension is
479 used, (this will hopefully be faster, more secure, and more robust)
480 but if not possible, the familiar Perl implementation is used.
484 C<Devel::PPPort>, originally from Kenneth Albanowski and now
485 maintained by Paul Marquess, has been added. It is primarily used
486 by C<h2xs> to enhance portability of of XS modules between different
491 C<Digest>, frontend module for calculating digests (checksums), from
492 Gisle Aas, has been added. See L<Digest>.
496 C<Digest::MD5> for calculating MD5 digests (checksums) as defined in
497 RFC 1321, from Gisle Aas, has been added. See L<Digest::MD5>.
499 use Digest::MD5 'md5_hex';
501 $digest = md5_hex("Thirsty Camel");
503 print $digest, "\n"; # 01d19d9d2045e005c3f1b80e8b164de1
505 NOTE: the C<MD5> backward compatibility module is deliberately not
506 included since its further use is discouraged.
510 C<Encode>, by Nick Ing-Simmons, provides a mechanism to translate
511 between different character encodings. Support for Unicode,
512 ISO-8859-*, ASCII, CP*, KOI8-R, and three variants of EBCDIC are
513 compiled in to the module. Several other encodings (like Japanese,
514 Chinese, and MacIntosh encodings) are included and will be loaded at
515 runtime. See L<Encode>.
517 Any encoding supported by Encode module is also available to the
518 ":encoding()" layer if PerlIO is used.
522 C<I18N::Langinfo> can be use to query locale information.
523 See L<I18N::Langinfo>.
527 C<I18N::LangTags> has functions for dealing with RFC3066-style
528 language tags, by Sean Burke. See L<I18N::LangTags>.
532 C<ExtUtils::Constant> is a new tool for extension writers for
533 generating XS code to import C header constants, by Nicholas Clark.
534 See L<ExtUtils::Constant>.
538 C<Filter::Simple> is an easy-to-use frontend to Filter::Util::Call,
539 from Damian Conway. See L<Filter::Simple>.
545 use Filter::Simple sub {
546 while (my ($from, $to) = splice @_, 0, 2) {
555 use MyFilter qr/red/ => 'green';
557 print "red\n"; # this code is filtered, will print "green\n"
558 print "bored\n"; # this code is filtered, will print "bogreen\n"
562 print "red\n"; # this code is not filtered, will print "red\n"
566 C<File::Temp> allows one to create temporary files and directories in
567 an easy, portable, and secure way, by Tim Jenness. See L<File::Temp>.
571 C<Filter::Util::Call> provides you with the framework to write
572 I<Source Filters> in Perl, from Paul Marquess. For most uses the
573 frontend Filter::Simple is to be preferred. See L<Filter::Util::Call>.
577 L<libnet> is a collection of perl5 modules related to network
578 programming, from Graham Barr. See L<Net::FTP>, L<Net::NNTP>,
579 L<Net::Ping>, L<Net::POP3>, L<Net::SMTP>, and L<Net::Time>.
581 Perl installation leaves libnet unconfigured, use F<libnetcfg> to configure.
585 C<List::Util> is a selection of general-utility list subroutines, like
586 sum(), min(), first(), and shuffle(), by Graham Barr. See L<List::Util>.
590 C<Locale::Constants>, C<Locale::Country>, C<Locale::Currency>, and
591 C<Locale::Language>, from Neil Bowers, have been added. They provide the
592 codes for various locale standards, such as "fr" for France, "usd" for
593 US Dollar, and "jp" for Japanese.
597 $country = code2country('jp'); # $country gets 'Japan'
598 $code = country2code('Norway'); # $code gets 'no'
600 See L<Locale::Constants>, L<Locale::Country>, L<Locale::Currency>,
601 and L<Locale::Language>.
605 C<Locale::Maketext> is localization framework from Sean Burke. See
606 L<Locale::Maketext>, and L<Locale::Maketext::TPJ13>. The latter is an
607 article about software localization, originally published in The Perl
608 Journal #13, republished here with kind permission.
612 C<Memoize> can make your functions faster by trading space for time,
613 from Mark-Jason Dominus. See L<Memoize>.
617 C<MIME::Base64> allows you to encode data in base64, from Gisle Aas,
618 as defined in RFC 2045 - I<MIME (Multipurpose Internet Mail
623 $encoded = encode_base64('Aladdin:open sesame');
624 $decoded = decode_base64($encoded);
626 print $encoded, "\n"; # "QWxhZGRpbjpvcGVuIHNlc2FtZQ=="
632 C<MIME::QuotedPrint> allows you to encode data in quoted-printable
633 encoding, as defined in RFC 2045 - I<MIME (Multipurpose Internet Mail
634 Extensions)>, from Gisle Aas.
636 use MIME::QuotedPrint;
638 $encoded = encode_qp("Smiley in Unicode: \x{263a}");
639 $decoded = decode_qp($encoded);
641 print $encoded, "\n"; # "Smiley in Unicode: =263A"
643 MIME::QuotedPrint has been enhanced to provide the basic methods
644 necessary to use it with PerlIO::Via as in :
646 use MIME::QuotedPrint;
647 open($fh,">Via(MIME::QuotedPrint)",$path);
649 See L<MIME::QuotedPrint>.
653 C<NEXT> is pseudo-class for method redispatch, from Damian Conway.
658 C<open> is a new pragma for setting the default I/O disciplines
663 C<PerlIO::Scalar> provides the implementation of IO to "in memory"
664 Perl scalars as discussed above, from Nick Ing-Simmons. It also
665 serves as an example of a loadable PerlIO layer. Other future
666 possibilities include PerlIO::Array and PerlIO::Code.
667 See L<PerlIO::Scalar>.
671 C<PerlIO::Via> acts as a PerlIO layer and wraps PerlIO layer
672 functionality provided by a class (typically implemented in perl
673 code), from Nick Ing-Simmons.
675 use MIME::QuotedPrint;
676 open($fh,">Via(MIME::QuotedPrint)",$path);
678 This will automatically convert everything output to C<$fh>
679 to Quoted-Printable. See L<PerlIO::Via>.
683 C<Pod::ParseLink>, by Russ Allbery, has been added,
684 to parse L<> links in pods as described in the new
689 C<Pod::Text::Overstrike>, by Joe Smith, has been added.
690 It converts POD data to formatted overstrike text.
691 See L<Pod::Text::Overstrike>.
695 C<Scalar::Util> is a selection of general-utility scalar subroutines,
696 like blessed(), reftype(), and tainted(). See L<Scalar::Util>.
700 C<sort> is a new pragma for controlling the behaviour of sort().
704 C<Storable> gives persistence to Perl data structures by allowing the
705 storage and retrieval of Perl data to and from files in a fast and
706 compact binary format, from Raphael Manfredi. See L<Storable>.
710 C<Switch>, from Damian Conway, has been added. Just by saying
714 you have C<switch> and C<case> available in Perl.
720 case 1 { print "number 1" }
721 case "a" { print "string a" }
722 case [1..10,42] { print "number in list" }
723 case (@array) { print "number in list" }
724 case /\w+/ { print "pattern" }
725 case qr/\w+/ { print "pattern" }
726 case (%hash) { print "entry in hash" }
727 case (\%hash) { print "entry in hash" }
728 case (\&sub) { print "arg to subroutine" }
729 else { print "previous case not true" }
736 C<Test::More> is yet another framework for writing test scripts,
737 more extensive than Test::Simple, by Michael Schwern. See L<Test::More>.
741 C<Test::Simple> has basic utilities for writing tests, by Michael
742 Schwern. See L<Test::Simple>.
746 C<Text::Balanced> has been added, for extracting delimited text
747 sequences from strings, from Damian Conway.
749 use Text::Balanced 'extract_delimited';
751 ($a, $b) = extract_delimited("'never say never', he never said", "'", '');
753 $a will be "'never say never'", $b will be ', he never said'.
755 In addition to extract_delimited() there are also extract_bracketed(),
756 extract_quotelike(), extract_codeblock(), extract_variable(),
757 extract_tagged(), extract_multiple(), gen_delimited_pat(), and
758 gen_extract_tagged(). With these you can implement rather advanced
759 parsing algorithms. See L<Text::Balanced>.
763 C<threads> is an interface to interpreter threads, by Arthur Bergman.
764 Interpreter threads (ithreads) is the new thread model introduced in
765 Perl 5.6 but only available as an internal interface for extension
766 writers (and for Win32 Perl for C<fork()> emulation). See L<threads>.
770 C<threads::shared> allows data sharing for interpreter threads, from
771 Arthur Bergman. In the ithreads model any data sharing between
772 threads must be explicit, as opposed to the old 5.005 thread model
773 where data sharing was implicit. See L<threads::shared>.
777 C<Tie::RefHash::Nestable>, by Edward Avis, allows storing hash
778 references (unlike the standard Tie::RefHash) The module is contained
779 within Tie::RefHash, see L<Tie::RefHash>.
783 C<Time::HiRes> provides high resolution timing (ualarm, usleep,
784 and gettimeofday), from Douglas E. Wegscheid. See L<Time::HiRes>.
788 C<Unicode::UCD> offers a querying interface to the Unicode Character
789 Database. See L<Unicode::UCD>.
793 C<Unicode::Collate> implements the UCA (Unicode Collation Algorithm)
794 for sorting Unicode strings, by SADAHIRO Tomoyuki. See L<Unicode::Collate>.
798 C<Unicode::Normalize> implements the various Unicode normalization
799 forms, by SADAHIRO Tomoyuki. See L<Unicode::Normalize>.
803 C<XS::Typemap>, by Tim Jenness, is a test extension that exercises XS
804 typemaps. Nothing gets installed but for extension writers the code
809 =head2 Updated And Improved Modules and Pragmata
815 The following independently supported modules have been updated to the
816 newest versions from CPAN: CGI, CPAN, DB_File, File::Spec, File::Temp,
817 Getopt::Long, Math::BigFloat, Math::BigInt, the podlators bundle
818 (Pod::Man, Pod::Text), Pod::LaTeX, Pod::Parser, Storable,
819 Term::ANSIColor, Test, Text-Tabs+Wrap.
823 The attributes::reftype() now works on tied arguments.
827 AutoLoader can now be disabled with C<no AutoLoader;>.
831 B::Deparse has been significantly enhanced. It now can deparse almost
832 all of the standard test suite (so that the tests still succeed).
833 There is a make target "test.deparse" for trying this out.
837 Class::Struct can now define the classes in compile time.
841 Class::Struct now assigns the array/hash element if the accessor
842 is called with an array/hash element as the B<sole> argument.
846 Data::Dumper has now an option to sort hashes.
850 Data::Dumper has now an option to dump code references
855 DB_File now supports newer Berkeley DB versions, among
860 The English module can now be used without the infamous performance
863 use English '-no_performance_hit';
865 (Assuming, of course, that one doesn't need the troublesome variables
866 C<$`>, C<$&>, or C<$'>.) Also, introduced C<@LAST_MATCH_START> and
867 C<@LAST_MATCH_END> English aliases for C<@-> and C<@+>.
871 Fcntl, Socket, and Sys::Syslog have been rewritten to use the
872 new-style constant dispatch section (see L<ExtUtils::Constant>).
873 This means that they will be more robust and hopefully faster.
877 File::Find now chdir()s correctly when chasing symbolic links.
881 File::Find now has pre- and post-processing callbacks. It also
882 correctly changes directories when chasing symbolic links. Callbacks
883 (naughtily) exiting with "next;" instead of "return;" now work.
887 File::Find is now (again) reentrant. It also has been made
892 File::Glob::glob() renamed to File::Glob::bsd_glob() to avoid
893 prototype mismatch with CORE::glob().
897 File::Glob now supports C<GLOB_LIMIT> constant to limit the size of
898 the returned list of filenames.
902 Devel::Peek now has an interface for the Perl memory statistics
903 (this works only if you are using perl's malloc, and if you have
904 compiled with debugging).
908 IPC::Open3 now allows the use of numeric file descriptors.
912 IO::Socket has now atmark() method, which returns true if the socket
913 is positioned at the out-of-band mark. The method is also exportable
914 as a sockatmark() function.
918 IO::Socket::INET has support for ReusePort option (if your platform
919 supports it). The Reuse option now has an alias, ReuseAddr. For clarity
920 you may want to prefer ReuseAddr.
924 IO::Socket::INET now supports C<LocalPort> of zero (usually meaning
925 that the operating system will make one up.)
929 use lib now works identically to @INC. Removing directories
930 with 'no lib' now works.
934 Math::BigFloat and Math::BigInt have undergone a full rewrite.
935 They are now magnitudes faster, and they support various
936 bignum libraries such as GMP and PARI as their backends.
940 Math::Complex handles inf, NaN etc., better.
944 Net::Ping has been enhanced. There is now "external" protocol which
945 uses Net::Ping::External module which runs external ping(1) and parses
946 the output. A version of Net::Ping::External is available in CPAN.
950 POSIX::sigaction() is now much more flexible and robust.
951 You can now install coderef handlers, 'DEFAULT', and 'IGNORE'
952 handlers, installing new handlers was not atomic.
956 In Safe the C<%INC> now localised in a Safe compartment so that
961 In SDBM_File on dosish platforms, some keys went missing because of
962 lack of support for files with "holes". A workaround for the problem
967 In Search::Dict one can now have a pre-processing hook for the
968 lines being searched.
972 The Shell module now has an OO interface.
976 The Test module has been significantly enhanced.
980 The vars pragma now supports declaring fully qualified variables.
981 (Something that C<our()> does not and will not support.)
985 The utf8:: name space (as in the pragma) provides various
986 Perl-callable functions to provide low level access to Perl's
987 internal Unicode representation. At the moment only length()
988 has been implemented.
992 =head1 Utility Changes
998 Emacs perl mode (emacs/cperl-mode.el) has been updated to version
1003 F<emacs/e2ctags.pl> is now much faster.
1007 C<h2ph> now supports C trigraphs.
1011 C<h2xs> now produces a template README.
1015 C<h2xs> now uses C<Devel::PPort> for better portability between
1016 different versions of Perl.
1020 C<h2xs> uses the new L<ExtUtils::Constant> module which will affect
1021 newly created extensions that define constants. Since the new code is
1022 more correct (if you have two constants where the first one is a
1023 prefix of the second one, the first constant B<never> gets defined),
1024 less lossy (it uses integers for integer constant, as opposed to the
1025 old code that used floating point numbers even for integer constants),
1026 and slightly faster, you might want to consider regenerating your
1027 extension code (the new scheme makes regenerating easy).
1028 L<h2xs> now also supports C trigraphs.
1032 C<libnetcfg> has been added to configure the libnet.
1036 C<perlbug> is now much more robust. It also sends the bug report to
1037 perl.org, not perl.com.
1041 C<perlcc> has been rewritten and its user interface (that is,
1042 command line) is much more like that of the UNIX C compiler, cc.
1043 (The perlbc tools has been removed. Use C<perlcc -B> instead.)
1047 C<perlivp> is a new Installation Verification Procedure utility
1048 for running any time after installing Perl.
1052 C<pod2html> now allows specifying a cache directory.
1056 C<s2p> has been completely rewritten in Perl. (It is in fact a full
1057 implementation of sed in Perl: you can use the sed functionality by
1058 using the C<psed> utility.)
1062 C<xsubpp> now understands POD documentation embedded in the *.xs files.
1066 C<xsubpp> now supports OUT keyword.
1070 =head1 New Documentation
1076 perl56delta details the changes between the 5.005 release and the
1081 perlclib documents the internal replacements for standard C library
1082 functions. (Interesting only for extension writers and Perl core
1087 perldebtut is a Perl debugging tutorial.
1091 perlebcdic contains considerations for running Perl on EBCDIC platforms.
1095 perlintro is a gentle introduction to Perl.
1099 perliol documents the internals of PerlIO with layers.
1103 perlmodstyle is a style guide for writing modules.
1107 perlnewmod tells about writing and submitting a new module.
1111 perlpacktut is a pack() tutorial.
1115 perlpod has been rewritten to be clearer and to record the best
1116 practices gathered over the years.
1120 perlpodspec is a more formal specification of the pod format,
1121 mainly of interest for writers of pod applications, not to
1122 people writing in pod.
1126 perlretut is a regular expression tutorial.
1130 perlrequick is a regular expressions quick-start guide.
1131 Yes, much quicker than perlretut.
1135 perltodo has been updated.
1139 perltootc has been renamed as perltooc (to not to conflict
1140 with perltoot in filesystems restricted to "8.3" names)
1144 perluniintro is an introduction to using Unicode in Perl
1145 (perlunicode is more of a reference)
1149 perlutil explains the command line utilities packaged with the Perl
1154 The following platform-specific documents are available before
1155 the installation as README.I<platform>, and after the installation
1158 perlaix perlamiga perlapollo perlbeos perlbs2000
1159 perlce perlcygwin perldgux perldos perlepoc perlhpux
1160 perlhurd perlmachten perlmacos perlmint perlmpeix
1161 perlnetware perlos2 perlos390 perlplan9 perlqnx perlsolaris
1162 perltru64 perluts perlvmesa perlvms perlvos perlwin32
1168 The documentation for the POSIX-BC platform is called "BS2000", to avoid
1169 confusion with the Perl POSIX module.
1173 The documentation for the WinCE platform is called "CE", to avoid
1174 confusion with the perlwin32 documentation on 8.3-restricted filesystems.
1178 =head1 Performance Enhancements
1184 map() could get pathologically slow when the result list it generates
1185 is larger than the source list. The performance has been improved for
1190 sort() has been changed to use primarily mergesort internally as
1191 opposed to the earlier quicksort. For very small lists this may
1192 result in slightly slower sorting times, but in general the speedup
1193 should be at least 20%. Additional bonuses are that the worst case
1194 behaviour of sort() is now better (in computer science terms it now
1195 runs in time O(N log N), as opposed to quicksort's Theta(N**2)
1196 worst-case run time behaviour), and that sort() is now stable
1197 (meaning that elements with identical keys will stay ordered as they
1198 were before the sort). See the C<sort> pragma for information.
1200 The story in more detail: suppose you want to serve yourself a little
1203 @digits = ( 3,1,4,1,5,9 );
1205 A numerical sort of the digits will yield (1,1,3,4,5,9), as expected.
1206 Which C<1> comes first is hard to know, since one C<1> looks pretty
1207 much like any other. You can regard this as totally trivial,
1208 or somewhat profound. However, if you just want to sort the even
1209 digits ahead of the odd ones, then what will
1211 sort { ($a % 2) <=> ($b % 2) } @digits;
1213 yield? The only even digit, C<4>, will come first. But how about
1214 the odd numbers, which all compare equal? With the quicksort algorithm
1215 used to implement Perl 5.6 and earlier, the order of ties is left up
1216 to the sort. So, as you add more and more digits of Pi, the order
1217 in which the sorted even and odd digits appear will change.
1218 and, for sufficiently large slices of Pi, the quicksort algorithm
1219 in Perl 5.8 won't return the same results even if reinvoked with the
1220 same input. The justification for this rests with quicksort's
1221 worst case behavior. If you run
1223 sort { $a <=> $b } ( 1 .. $N , 1 .. $N );
1225 (something you might approximate if you wanted to merge two sorted
1226 arrays using sort), doubling $N doesn't just double the quicksort time,
1227 it I<quadruples> it. Quicksort has a worst case run time that can
1228 grow like N**2, so-called I<quadratic> behaviour, and it can happen
1229 on patterns that may well arise in normal use. You won't notice this
1230 for small arrays, but you I<will> notice it with larger arrays,
1231 and you may not live long enough for the sort to complete on arrays
1232 of a million elements. So the 5.8 quicksort scrambles large arrays
1233 before sorting them, as a statistical defence against quadratic behaviour.
1234 But that means if you sort the same large array twice, ties may be
1235 broken in different ways.
1237 Because of the unpredictability of tie-breaking order, and the quadratic
1238 worst-case behaviour, quicksort was I<almost> replaced completely with
1239 a stable mergesort. I<Stable> means that ties are broken to preserve
1240 the original order of appearance in the input array. So
1242 sort { ($a % 2) <=> ($b % 2) } (3,1,4,1,5,9);
1244 will yield (4,3,1,1,5,9), guaranteed. The even and odd numbers
1245 appear in the output in the same order they appeared in the input.
1246 Mergesort has worst case O(NlogN) behaviour, the best value
1247 attainable. And, ironically, this mergesort does particularly
1248 well where quicksort goes quadratic: mergesort sorts (1..$N, 1..$N)
1249 in O(N) time. But quicksort was rescued at the last moment because
1250 it is faster than mergesort on certain inputs and platforms.
1251 For example, if you really I<don't> care about the order of even
1252 and odd digits, quicksort will run in O(N) time; it's very good
1253 at sorting many repetitions of a small number of distinct elements.
1254 The quicksort divide and conquer strategy works well on platforms
1255 with relatively small, very fast, caches. Eventually, the problem gets
1256 whittled down to one that fits in the cache, from which point it
1257 benefits from the increased memory speed.
1259 Quicksort was rescued by implementing a sort pragma to control aspects
1260 of the sort. The B<stable> subpragma forces stable behaviour,
1261 regardless of algorithm. The B<_quicksort> and B<_mergesort>
1262 subpragmas are heavy-handed ways to select the underlying implementation.
1263 The leading C<_> is a reminder that these subpragmas may not survive
1264 beyond 5.8. More appropriate mechanisms for selecting the implementation
1265 exist, but they wouldn't have arrived in time to save quicksort.
1269 Hashes now use Bob Jenkins "One-at-a-Time" hashing key algorithm
1270 (http://burtleburtle.net/bob/hash/doobs.html). This algorithm is
1271 reasonably fast while producing a much better spread of values than
1272 the old hashing algorithm (originally by Chris Torek, later tweaked by
1273 Ilya Zakharevich). Hash values output from the algorithm on a hash of
1274 all 3-char printable ASCII keys comes much closer to passing the
1275 DIEHARD random number generation tests. According to perlbench, this
1276 change has not affected the overall speed of Perl.
1280 unshift() should now be noticeably faster.
1284 =head1 Installation and Configuration Improvements
1286 =head2 Generic Improvements
1292 INSTALL now explains how you can configure Perl to use 64-bit
1293 integers even on non-64-bit platforms.
1297 Policy.sh policy change: if you are reusing a Policy.sh file
1298 (see INSTALL) and you use Configure -Dprefix=/foo/bar and in the old
1299 Policy $prefix eq $siteprefix and $prefix eq $vendorprefix, all of
1300 them will now be changed to the new prefix, /foo/bar. (Previously
1301 only $prefix changed.) If you do not like this new behaviour,
1302 specify prefix, siteprefix, and vendorprefix explicitly.
1306 A new optional location for Perl libraries, otherlibdirs, is available.
1307 It can be used for example for vendor add-ons without disturbing Perl's
1308 own library directories.
1312 In many platforms the vendor-supplied 'cc' is too stripped-down to
1313 build Perl (basically, 'cc' doesn't do ANSI C). If this seems
1314 to be the case and 'cc' does not seem to be the GNU C compiler
1315 'gcc', an automatic attempt is made to find and use 'gcc' instead.
1319 gcc needs to closely track the operating system release to avoid
1320 build problems. If Configure finds that gcc was built for a different
1321 operating system release than is running, it now gives a clearly visible
1322 warning that there may be trouble ahead.
1326 If binary compatibility with the 5.005 release is not wanted, Configure
1327 no longer suggests including the 5.005 modules in @INC.
1331 Configure C<-S> can now run non-interactively.
1335 Configure support for pdp11-style memory models has been removed due
1340 configure.gnu now works with options with whitespace in them.
1344 installperl now outputs everything to STDERR.
1348 $Config{byteorder} is now computed dynamically (this is more robust
1349 with "fat binaries" where an executable image contains binaries for
1350 more than one binary platform.)
1354 Because PerlIO is now the default on most platforms, "-perlio" doesn't
1355 get appended to the $Config{archname} (also known as $^O) anymore.
1356 Instead, if you explicitly choose not to use perlio (Configure command
1357 line option -Uuseperlio), you will get "-stdio" appended.
1361 Another change related to the architecture name is that "-64all"
1362 (-Duse64bitall, or "maximally 64-bit") is appended only if your
1363 pointers are 64 bits wide. (To be exact, the use64bitall is ignored.)
1367 In AFS installations one can configure the root of the AFS to be
1368 somewhere else than the default F</afs> by using the Configure
1369 parameter C<-Dafsroot=/some/where/else>.
1373 APPLLIB_EXP, a less-know configuration-time definition, has been
1374 documented. It can be used to prepend site-specific directories
1375 to Perl's default search path (@INC), see INSTALL for information.
1379 The version of Berkeley DB used when the Perl (and, presumably, the
1380 DB_File extension) was built is now available as
1381 C<@Config{qw(db_version_major db_version_minor db_version_patch)}>
1382 from Perl and as C<DB_VERSION_MAJOR_CFG DB_VERSION_MINOR_CFG
1383 DB_VERSION_PATCH_CFG> from C.
1387 Building Berkeley DB3 for compatibility modes for DB, NDBM, and ODBM
1388 has been documented in INSTALL.
1392 If you have CPAN access (either network or a local copy such as a
1393 CD-ROM) you can during specify extra modules to Configure to build and
1394 install with Perl using the -Dextras=... option. See INSTALL for
1399 In addition to config.over a new override file, config.arch, is
1400 available. That is supposed to be used by hints file writers for
1401 architecture-wide changes (as opposed to config.over which is for
1406 If your file system supports symbolic links you can build Perl outside
1407 of the source directory by
1409 mkdir /tmp/perl/build/directory
1410 cd /tmp/perl/build/directory
1411 sh /path/to/perl/source/Configure -Dmksymlinks ...
1413 This will create in /tmp/perl/build/directory a tree of symbolic links
1414 pointing to files in /path/to/perl/source. The original files are left
1415 unaffected. After Configure has finished you can just say
1419 and Perl will be built and tested, all in /tmp/perl/build/directory.
1423 For Perl developers several new make targets for profiling
1424 and debugging have been added, see L<perlhack>.
1430 Use of the F<gprof> tool to profile Perl has been documented in
1431 L<perlhack>. There is a make target called "perl.gprof" for
1432 generating a gprofiled Perl executable.
1436 If you have GCC 3, there is a make target called "perl.gcov" for
1437 creating a gcoved Perl executable for coverage analysis. See
1442 If you are on IRIX or Tru64 platforms, new profiling/debugging options
1443 have been added, see L<perlhack> for more information about pixie and
1450 Guidelines of how to construct minimal Perl installations have
1451 been added to INSTALL.
1455 The Thread extension is now not built at all under ithreads
1456 (C<Configure -Duseithreads>) because it wouldn't work anyway (the
1457 Thread extension requires being Configured with C<-Duse5005threads>).
1459 But note that the Thread.pm interface is now shared by both
1464 =head2 New Or Improved Platforms
1466 For the list of platforms known to support Perl,
1467 see L<perlport/"Supported Platforms">.
1473 AIX dynamic loading should be now better supported.
1477 AIX should now work better with gcc, threads, and 64-bitness. Also the
1478 long doubles support in AIX should be better now. See L<perlaix>.
1482 After a long pause, AmigaOS has been verified to be happy with Perl.
1486 AtheOS (http://www.atheos.cx/) is a new platform.
1490 DG/UX platform now supports the 5.005-style threads. See L<perldgux>.
1494 DYNIX/ptx platform (a.k.a. dynixptx) is supported at or near osvers 4.5.2.
1498 EBCDIC platforms (z/OS, also known as OS/390, POSIX-BC, and VM/ESA)
1499 have been regained. Many test suite tests still fail and the
1500 co-existence of Unicode and EBCDIC isn't quite settled, but the
1501 situation is much better than with Perl 5.6. See L<perlos390>,
1502 L<perlbs2000> (for POSIX-BC), and L<perlvmesa> for more information.
1506 Building perl with -Duseithreads or -Duse5005threads now works under
1507 HP-UX 10.20 (previously it only worked under 10.30 or later). You will
1508 need a thread library package installed. See README.hpux.
1512 MacOS Classic (MacPerl has of course been available since
1513 perl 5.004 but now the source code bases of standard Perl
1514 and MacPerl have been synchronised)
1518 MacOS X (or Darwin) should now be able to build Perl even on HFS+
1519 filesystems. (The case-insensitivity confused the Perl build process.)
1523 NCR MP-RAS is now supported.
1527 NetWare from Novell is now supported. See L<perlnetware>.
1531 NonStop-UX is now supported.
1535 NEC SUPER-UX is now supported.
1539 Amdahl UTS UNIX mainframe platform is now supported.
1543 WinCE is now supported. See L<perlce>.
1547 z/OS (formerly known as OS/390, formerly known as MVS OE) has now
1548 support for dynamic loading. This is not selected by default,
1549 however, you must specify -Dusedl in the arguments of Configure.
1553 =head1 Selected Bug Fixes
1555 Numerous memory leaks and uninitialized memory accesses have been
1556 hunted down. Most importantly anonymous subs used to leak quite
1563 The autouse pragma didn't work for Multi::Part::Function::Names.
1567 caller() could cause core dumps in certain situations. Carp was sometimes
1568 affected by this problem.
1572 chop(@list) in list context returned the characters chopped in
1573 reverse order. This has been reversed to be in the right order.
1577 Configure no longer includes the DBM libraries (dbm, gdbm, db, ndbm)
1578 when building the Perl binary. The only exception to this is SunOS 4.x,
1583 The behaviour of non-decimal but numeric string constants such as
1584 "0x23" was platform-dependent: in some platforms that was seen as 35,
1585 in some as 0, in some as a floating point number (don't ask). This
1586 was caused by Perl using the operating system libraries in a situation
1587 where the result of the string to number conversion is undefined: now
1588 Perl consistently handles such strings as zero in numeric contexts.
1592 The order of DESTROYs has been made more predictable.
1596 Several debugger fixes: exit code now reflects the script exit code,
1597 condition C<"0"> now treated correctly, the C<d> command now checks
1598 line number, the C<$.> no longer gets corrupted, all debugger output
1599 now goes correctly to the socket if RemotePort is set.
1603 Perl 5.6.0 could emit spurious warnings about redefinition of dl_error()
1604 when statically building extensions into perl. This has been corrected.
1608 L<dprofpp> -R didn't work.
1612 C<*foo{FORMAT}> now works.
1615 Infinity is now recognized as a number.
1619 UNIVERSAL::isa no longer caches methods incorrectly. (This broke
1620 the Tk extension with 5.6.0.)
1624 Lexicals I: lexicals outside an eval "" weren't resolved
1625 correctly inside a subroutine definition inside the eval "" if they
1626 were not already referenced in the top level of the eval""ed code.
1630 Lexicals II: lexicals leaked at file scope into subroutines that
1631 were declared before the lexicals.
1635 Lexical warnings now propagating correctly between scopes
1636 and into C<eval "...">.
1640 C<use warnings qw(FATAL all)> did not work as intended. This has been
1645 warnings::enabled() now reports the state of $^W correctly if the caller
1646 isn't using lexical warnings.
1650 Line renumbering with eval and C<#line> now works.
1654 Fixed numerous memory leaks, especially in eval "".
1658 mkdir() now ignores trailing slashes in the directory name,
1659 as mandated by POSIX.
1663 Some versions of glibc have a broken modfl(). This affects builds
1664 with C<-Duselongdouble>. This version of Perl detects this brokenness
1665 and has a workaround for it. The glibc release 2.2.2 is known to have
1666 fixed the modfl() bug.
1670 Modulus of unsigned numbers now works (4063328477 % 65535 used to
1671 return 27406, instead of 27047).
1675 Some "not a number" warnings introduced in 5.6.0 eliminated to be
1676 more compatible with 5.005. Infinity is now recognised as a number.
1680 Numeric conversions did not recognize changes in the string value
1681 properly in certain circumstances.
1685 Attributes (like :shared) didn't work with our().
1689 our() variables will not cause "will not stay shared" warnings.
1693 "our" variables of the same name declared in two sibling blocks
1694 resulted in bogus warnings about "redeclaration" of the variables.
1695 The problem has been corrected.
1699 pack "Z" now correctly terminates the string with "\0".
1703 Fix password routines which in some shadow password platforms
1704 (e.g. HP-UX) caused getpwent() to return every other entry.
1708 The PERL5OPT environment variable (for passing command line arguments
1709 to Perl) didn't work for more than a single group of options.
1713 PERL5OPT with embedded spaces didn't work.
1717 printf() no longer resets the numeric locale to "C".
1721 C<qw(a\\b)> now parses correctly as C<'a\\b'>.
1725 pos() did not return the correct value within s///ge in earlier
1726 versions. This is now handled correctly.
1730 Printing quads (64-bit integers) with printf/sprintf now works
1731 without the q L ll prefixes (assuming you are on a quad-capable platform).
1735 Regular expressions on references and overloaded scalars now work.
1739 Right-hand side magic (GMAGIC) could in many cases such as string
1740 concatenation be invoked too many times.
1744 scalar() now forces scalar context even when used in void context.
1748 SOCKS support is now much more robust.
1752 sort() arguments are now compiled in the right wantarray context
1753 (they were accidentally using the context of the sort() itself).
1754 The comparison block is now run in scalar context, and the arguments
1755 to be sorted are always provided list context.
1759 Changed the POSIX character class C<[[:space:]]> to include the (very
1760 rarely used) vertical tab character. Added a new POSIX-ish character
1761 class C<[[:blank:]]> which stands for horizontal whitespace
1762 (currently, the space and the tab).
1766 The tainting behaviour of sprintf() has been rationalized. It does
1767 not taint the result of floating point formats anymore, making the
1768 behaviour consistent with that of string interpolation.
1772 Some cases of inconsistent taint propagation (such as within hash
1773 values) have been fixed.
1777 The RE engine found in Perl 5.6.0 accidentally pessimised certain kinds
1778 of simple pattern matches. These are now handled better.
1782 Regular expression debug output (whether through C<use re 'debug'>
1783 or via C<-Dr>) now looks better.
1787 Multi-line matches like C<"a\nxb\n" =~ /(?!\A)x/m> were flawed. The
1792 Use of $& could trigger a core dump under some situations. This
1797 The regular expression captured submatches ($1, $2, ...) are now
1798 more consistently unset if the match fails, instead of leaving false
1799 data lying around in them.
1803 readline() on files opened in "slurp" mode could return an extra "" at
1804 the end in certain situations. This has been corrected.
1808 Autovivification of symbolic references of special variables described
1809 in L<perlvar> (as in C<${$num}>) was accidentally disabled. This works
1814 Sys::Syslog ignored the C<LOG_AUTH> constant.
1818 All but the first argument of the IO syswrite() method are now optional.
1822 $AUTOLOAD, sort(), lock(), and spawning subprocesses
1823 in multiple threads simultaneously are now thread-safe.
1827 Tie::ARRAY SPLICE method was broken.
1831 Allow read-only string on left hand side of non-modifying tr///.
1835 Several Unicode fixes.
1841 BOMs (byte order marks) in the beginning of Perl files
1842 (scripts, modules) should now be transparently skipped.
1843 UTF-16 (UCS-2) encoded Perl files should now be read correctly.
1847 The character tables have been updated to Unicode 3.1.1.
1851 Comparing with utf8 data does not magically upgrade non-utf8 data
1856 C<IsAlnum>, C<IsAlpha>, and C<IsWord> now match titlecase.
1860 Concatenation with the C<.> operator or via variable interpolation,
1861 C<eq>, C<substr>, C<reverse>, C<quotemeta>, the C<x> operator,
1862 substitution with C<s///>, single-quoted UTF8, should now work.
1866 The C<tr///> operator now works. Note that the C<tr///CU>
1867 functionality has been removed (but see pack('U0', ...)).
1871 C<eval "v200"> now works.
1875 Perl 5.6.0 parsed m/\x{ab}/ incorrectly, leading to spurious warnings.
1876 This has been corrected.
1880 Zero entries were missing from the Unicode classes like C<IsDigit>.
1886 Large unsigned numbers (those above 2**31) could sometimes lose their
1887 unsignedness, causing bogus results in arithmetic operations.
1891 =head2 Platform Specific Changes and Fixes
1899 Perl now works on post-4.0 BSD/OSes.
1905 Setting C<$0> now works (as much as possible; see L<perlvar> for details).
1911 Numerous updates; currently synchronised with Cygwin 1.1.4.
1915 Previously DYNIX/ptx had problems in its Configure probe for non-blocking I/O.
1921 EPOC update after Perl 5.6.0. See README.epoc.
1927 Perl now works on post-3.0 FreeBSDs.
1933 README.hpux updated; C<Configure -Duse64bitall> now almost works.
1939 Numerous compilation flag and hint enhancements; accidental mixing
1940 of 32-bit and 64-bit libraries (a doomed attempt) made much harder.
1950 Long doubles should now work (see INSTALL).
1954 Linux previously had problems related to sockaddrlen when using
1955 accept(), revcfrom() (in Perl: recv()), getpeername(), and getsockname().
1963 Compilation of the standard Perl distribution in MacOS Classic should
1964 now work if you have the Metrowerks development environment and
1965 the missing Mac-specific toolkit bits. Contact the macperl mailing
1972 MPE/iX update after Perl 5.6.0. See README.mpeix.
1978 Perl now works on NetBSD/sparc.
1984 Now works with usethreads (see INSTALL).
1990 64-bitness using the Sun Workshop compiler now works.
1994 Tru64 (aka Digital UNIX, aka DEC OSF/1)
1996 The operating system version letter now recorded in $Config{osvers}.
1997 Allow compiling with gcc (previously explicitly forbidden). Compiling
1998 with gcc still not recommended because buggy code results, even with
2005 Fixed various alignment problems that lead into core dumps either
2006 during build or later; no longer dies on math errors at runtime;
2007 now using full quad integers (64 bits), previously was using
2008 only 46 bit integers for speed.
2014 chdir() now works better despite a CRT bug; now works with MULTIPLICITY
2015 (see INSTALL); now works with Perl's malloc.
2017 The tainting of C<%ENV> elements via C<keys> or C<values> was previously
2018 unimplemented. It now works as documented.
2020 The C<waitpid> emulation has been improved. The worst bug (now fixed)
2021 was that a pid of -1 would cause a wildcard search of all processes on
2022 the system. The most significant enhancement is that we can now
2023 usually get the completion status of a terminated process.
2025 POSIX-style signals are now emulated much better on VMS versions prior
2028 The C<system> function and backticks operator have improved
2029 functionality and better error handling.
2039 accept() no longer leaks memory.
2043 Borland C++ v5.5 is now a supported compiler that can build Perl.
2044 However, the generated binaries continue to be incompatible with those
2045 generated by the other supported compilers (GCC and Visual C++).
2049 Better chdir() return value for a non-existent directory.
2053 Duping socket handles with open(F, ">&MYSOCK") now works under Windows 9x.
2057 New %ENV entries now propagate to subprocesses.
2061 Current directory entries in %ENV are now correctly propagated to child
2066 $ENV{LIB} now used to search for libs under Visual C.
2070 fork() emulation has been improved in various ways, but still continues
2071 to be experimental. See L<perlfork> for known bugs and caveats.
2075 A failed (pseudo)fork now returns undef and sets errno to EAGAIN.
2079 Win32::GetCwd() correctly returns C:\ instead of C: when at the drive root.
2080 Other bugs in chdir() and Cwd::cwd() have also been fixed.
2084 HTML files will be installed in c:\perl\html instead of c:\perl\lib\pod\html
2088 The makefiles now provide a single switch to bulk-enable all the features
2089 enabled in ActiveState ActivePerl (a popular Win32 binary distribution).
2093 Allow REG_EXPAND_SZ keys in the registry.
2097 Can now send() from all threads, not just the first one.
2101 Fake signal handling reenabled, bugs and all.
2105 %SIG has been enabled under USE_ITHREADS, but its use is completely
2106 unsupported under all configurations.
2110 Less stack reserved per thread so that more threads can run
2111 concurrently. (Still 16M per thread.)
2115 C<File::Spec->tmpdir()> now prefers C:/temp over /tmp
2116 (works better when perl is running as service).
2120 Better UNC path handling under ithreads.
2124 wait(), waitpid() and backticks now return the correct exit status under
2129 winsock handle leak fixed.
2135 =head1 New or Changed Diagnostics
2141 All regular expression compilation error messages are now hopefully
2142 easier to understand both because the error message now comes before
2143 the failed regex and because the point of failure is now clearly
2144 marked by a C<E<lt>-- HERE> marker.
2148 The various "opened only for", "on closed", "never opened" warnings
2149 drop the C<main::> prefix for filehandles in the C<main> package,
2150 for example C<STDIN> instead of C<main::STDIN>.
2154 The "Unrecognized escape" warning has been extended to include C<\8>,
2155 C<\9>, and C<\_>. There is no need to escape any of the C<\w> characters.
2159 Two new debugging options have been added: if you have compiled your
2160 Perl with debugging, you can use the -DT and -DR options to trace
2161 tokenising and to add reference counts to displaying variables,
2166 If an attempt to use a (non-blessed) reference as an array index
2167 is made, a warning is given.
2171 C<push @a;> and C<unshift @a;> (with no values to push or unshift)
2172 now give a warning. This may be a problem for generated and evaled
2177 If you try to L<perlfunc/pack> a number less than 0 or larger than 255
2178 using the C<"C"> format you will get an optional warning. Similarly
2179 for the C<"c"> format and a number less than -128 or more than 127.
2183 Certain regex modifiers such as C<(?o)> make sense only if applied to
2184 the entire regex. You will an optional warning if you try to do otherwise.
2188 Using arrays or hashes as references (e.g. C<< %foo->{bar} >>
2189 has been deprecated for a while. Now you will get an optional warning.
2193 =head1 Changed Internals
2199 perlapi.pod (a companion to perlguts) now attempts to document the
2204 You can now build a really minimal perl called microperl.
2205 Building microperl does not require even running Configure;
2206 C<make -f Makefile.micro> should be enough. Beware: microperl makes
2207 many assumptions, some of which may be too bold; the resulting
2208 executable may crash or otherwise misbehave in wondrous ways.
2209 For careful hackers only.
2213 Added rsignal(), whichsig(), do_join(), op_clear, op_null,
2214 ptr_table_clear(), ptr_table_free(), sv_setref_uv(), and several UTF-8
2215 interfaces to the publicised API. For the full list of the available
2216 APIs see L<perlapi>.
2220 Made possible to propagate customised exceptions via croak()ing.
2224 Now xsubs can have attributes just like subs.
2228 dTHR and djSP have been obsoleted; the former removed (because it's
2229 a no-op) and the latter replaced with dSP.
2233 PERL_OBJECT has been completely removed.
2237 The MAGIC constants (e.g. C<'P'>) have been macrofied
2238 (e.g. C<PERL_MAGIC_TIED>) for better source code readability
2239 and maintainability.
2243 The regex compiler now maintains a structure that identifies nodes in
2244 the compiled bytecode with the corresponding syntactic features of the
2245 original regex expression. The information is attached to the new
2246 C<offsets> member of the C<struct regexp>. See L<perldebguts> for more
2247 complete information.
2251 The C code has been made much more C<gcc -Wall> clean. Some warning
2252 messages still remain in some platforms, so if you are compiling with
2253 gcc you may see some warnings about dubious practices. The warnings
2254 are being worked on.
2258 F<perly.c>, F<sv.c>, and F<sv.h> have now been extensively commented.
2262 Documentation on how to use the Perl source repository has been added
2263 to F<Porting/repository.pod>.
2267 There are now several profiling make targets.
2271 =head1 Security Vulnerability Closed
2273 (This change was already made in 5.7.0 but bears repeating here.)
2275 A potential security vulnerability in the optional suidperl component
2276 of Perl was identified in August 2000. suidperl is neither built nor
2277 installed by default. As of November 2001 the only known vulnerable
2278 platform is Linux, most likely all Linux distributions. CERT and
2279 various vendors and distributors have been alerted about the vulnerability.
2280 See http://www.cpan.org/src/5.0/sperl-2000-08-05/sperl-2000-08-05.txt
2281 for more information.
2283 The problem was caused by Perl trying to report a suspected security
2284 exploit attempt using an external program, /bin/mail. On Linux
2285 platforms the /bin/mail program had an undocumented feature which
2286 when combined with suidperl gave access to a root shell, resulting in
2287 a serious compromise instead of reporting the exploit attempt. If you
2288 don't have /bin/mail, or if you have 'safe setuid scripts', or if
2289 suidperl is not installed, you are safe.
2291 The exploit attempt reporting feature has been completely removed from
2292 Perl 5.8.0 (and the maintenance release 5.6.1, and it was removed also
2293 from all the Perl 5.7 releases), so that particular vulnerability
2294 isn't there anymore. However, further security vulnerabilities are,
2295 unfortunately, always possible. The suidperl functionality is most
2296 probably going to be removed in Perl 5.10. In any case, suidperl
2297 should only be used by security experts who know exactly what they are
2298 doing and why they are using suidperl instead of some other solution
2299 such as sudo (see http://www.courtesan.com/sudo/).
2303 Several new tests have been added, especially for the F<lib>
2304 subsection. There are now about 34 000 individual tests (spread over
2305 about 530 test scripts), in the regression suite (5.6.1 has about
2306 11700 tests, in 258 test scripts) Many of the new tests are introduced
2307 by the new modules, but still in general Perl is now more thoroughly
2310 Because of the large number of tests, running the regression suite
2311 will take considerably longer time than it used to: expect the suite
2312 to take up to 4-5 times longer to run than in perl 5.6. In a really
2313 fast machine you can hope to finish the suite in about 5 minutes
2316 The tests are now reported in a different order than in earlier Perls.
2317 (This happens because the test scripts from under t/lib have been moved
2318 to be closer to the library/extension they are testing.)
2320 =head1 Known Problems
2328 In AIX 4.2 Perl extensions that use C++ functions that use statics
2329 may have problems in that the statics are not getting initialized.
2330 In newer AIX releases this has been solved by linking Perl with
2331 the libC_r library, but unfortunately in AIX 4.2 the said library
2332 has an obscure bug where the various functions related to time
2333 (such as time() and gettimeofday()) return broken values, and
2334 therefore in AIX 4.2 Perl is not linked against the libC_r.
2338 vac 5.0.0.0 May Produce Buggy Code For Perl
2340 The AIX C compiler vac version 5.0.0.0 may produce buggy code,
2341 resulting in few random tests failing, but when the failing tests
2342 are run by hand, they succeed. We suggest upgrading to at least
2343 vac version 5.0.1.0, that has been known to compile Perl correctly.
2344 "lslpp -L|grep vac.C" will tell you the vac version.
2348 =head2 Amiga Perl Invoking Mystery
2350 One cannot call Perl using the C<volume:> syntax, that is, C<perl -v>
2351 works, but for example C<bin:perl -v> doesn't. The exact reason isn't
2352 known but the current suspect is the F<ixemul> library.
2354 =head2 lib/ftmp-security tests warn 'system possibly insecure'
2356 Don't panic. Read INSTALL 'make test' section instead.
2358 =head2 Cygwin intermittent failures of lib/Memoize/t/expire_file 11 and 12
2360 The subtests 11 and 12 sometimes fail and sometimes work.
2362 =head2 HP-UX lib/io_multihomed Fails When LP64-Configured
2364 The lib/io_multihomed test may hang in HP-UX if Perl has been
2365 configured to be 64-bit. Because other 64-bit platforms do not hang in
2366 this test, HP-UX is suspect. All other tests pass in 64-bit HP-UX. The
2367 test attempts to create and connect to "multihomed" sockets (sockets
2368 which have multiple IP addresses).
2370 =head2 HP-UX lib/posix Subtest 9 Fails When LP64-Configured
2372 If perl is configured with -Duse64bitall, the successful result of the
2373 subtest 10 of lib/posix may arrive before the successful result of the
2374 subtest 9, which confuses the test harness so much that it thinks the
2377 =head2 Linux With Sfio Fails op/misc Test 48
2383 The following tests are known to fail:
2385 Failed Test Stat Wstat Total Fail Failed List of Failed
2386 -------------------------------------------------------------------------
2387 ../ext/DB_File/t/db-btree.t 0 11 ?? ?? % ??
2388 ../ext/DB_File/t/db-recno.t 149 3 2.01% 61 63 65
2389 ../ext/POSIX/t/posix.t 31 1 3.23% 10
2390 ../lib/warnings.t 450 1 0.22% 316
2394 OS/390 has rather many test failures but the situation is actually
2395 better than it was in 5.6.0, it's just that so many new modules and
2396 tests have been added.
2398 Failed Test Stat Wstat Total Fail Failed List of Failed
2399 -----------------------------------------------------------------------------
2400 ../ext/B/Deparse.t 14 1 7.14% 14
2401 ../ext/B/Showlex.t 1 1 100.00% 1
2402 ../ext/Encode/Encode/Tcl.t 610 13 2.13% 592 594 596 598
2404 ../ext/IO/lib/IO/t/io_unix.t 113 28928 5 3 60.00% 3-5
2405 ../ext/POSIX/POSIX.t 29 1 3.45% 14
2406 ../ext/Storable/t/lock.t 255 65280 5 3 60.00% 3-5
2407 ../lib/locale.t 129 33024 117 19 16.24% 99-117
2408 ../lib/warnings.t 434 1 0.23% 75
2409 ../lib/ExtUtils.t 27 1 3.70% 25
2410 ../lib/Math/BigInt/t/bigintpm.t 1190 1 0.08% 1145
2411 ../lib/Unicode/UCD.t 81 48 59.26% 1-16 49-64 66-81
2412 ../lib/User/pwent.t 9 1 11.11% 4
2413 op/pat.t 660 6 0.91% 242-243 424-425
2415 op/split.t 0 9 ?? ?? % ??
2416 op/taint.t 174 3 1.72% 156 162 168
2417 op/tr.t 70 3 4.29% 50 58-59
2418 Failed 16/422 test scripts, 96.21% okay. 105/23251 subtests failed, 99.55% okay.
2420 =head2 op/sprintf tests 129 and 130
2422 The op/sprintf tests 129 and 130 are known to fail on some platforms.
2423 Examples include any platform using sfio, and Compaq/Tandem's NonStop-UX.
2424 The failing platforms do not comply with the ANSI C Standard, line
2425 19ff on page 134 of ANSI X3.159 1989 to be exact. (They produce
2426 something other than "1" and "-1" when formatting 0.6 and -0.6 using
2427 the printf format "%.0f", most often they produce "0" and "-0".)
2429 =head2 Failure of Thread tests
2431 B<Note that support for 5.005-style threading remains experimental
2432 and practically unsupported.>
2434 The following tests are known to fail due to fundamental problems in
2435 the 5.005 threading implementation. These are not new failures--Perl
2436 5.005_0x has the same bugs, but didn't have these tests.
2438 ext/List/Util/t/first 2
2440 ext/Thread/thr5005 19-20
2442 These failures are unlikely to get fixed.
2450 ext/POSIX/sigaction subtests 6 and 13 may fail.
2454 lib/ExtUtils may spuriously claim that subtest 28 failed,
2455 which is interesting since the test only has 27 tests.
2459 Numerous numerical test failures
2461 op/numconvert 209,210,217,218
2463 ext/Time/HiRes/HiRes 9
2464 lib/Math/BigInt/t/bigintpm 1145
2467 These tests fail because of yet unresolved floating point inaccuracies.
2473 There are a few known test failures, see L<perluts>.
2477 There is one known test failure with a default configuration:
2479 [.run]switches..........................FAILED on test 1
2483 In multi-CPU boxes there are some problems with the I/O buffering:
2484 some output may appear twice.
2486 =head2 Localising a Tied Variable Leaks Memory
2489 tie my %tie_hash => 'Tie::StdHash';
2493 local($tie_hash{Foo}) = 1; # leaks
2495 Code like the above is known to leak memory every time the local()
2498 =head2 Localising Tied Arrays and Hashes Is Broken
2502 doesn't work as one would expect: the old value is restored
2505 =head2 Self-tying of Arrays and Hashes Is Forbidden
2507 Self-tying of arrays and hashes is broken in rather deep and
2508 hard-to-fix ways. As a stop-gap measure to avoid people from getting
2509 frustrated at the mysterious results (core dumps, most often) it is
2510 for now forbidden (you will get a fatal error even from an attempt).
2512 =head2 Variable Attributes are not Currently Usable for Tieing
2514 This limitation will hopefully be fixed in future. (Subroutine
2515 attributes work fine for tieing, see L<Attribute::Handlers>).
2517 One way to run into this limitation is to have a loop variable with
2518 attributes within a loop: the tie is called only once, not for each
2519 iteration of the loop.
2521 =head2 Building Extensions Can Fail Because Of Largefiles
2523 Some extensions like mod_perl are known to have issues with
2524 `largefiles', a change brought by Perl 5.6.0 in which file offsets
2525 default to 64 bits wide, where supported. Modules may fail to compile
2526 at all or compile and work incorrectly. Currently there is no good
2527 solution for the problem, but Configure now provides appropriate
2528 non-largefile ccflags, ldflags, libswanted, and libs in the %Config
2529 hash (e.g., $Config{ccflags_nolargefiles}) so the extensions that are
2530 having problems can try configuring themselves without the
2531 largefileness. This is admittedly not a clean solution, and the
2532 solution may not even work at all. One potential failure is whether
2533 one can (or, if one can, whether it's a good idea) link together at
2534 all binaries with different ideas about file offsets, all this is
2537 =head2 Unicode Support on EBCDIC Still Spotty
2539 Though mostly working, Unicode support still has problem spots on
2540 EBCDIC platforms. One such known spot are the C<\p{}> and C<\P{}>
2541 regular expression constructs for code points less than 256: the
2542 pP are testing for Unicode code points, not knowing about EBCDIC.
2544 =head2 The Compiler Suite Is Still Experimental
2546 The compiler suite is slowly getting better but it continues to be
2547 highly experimental. Use in production environments is discouraged.
2549 =head2 The Long Double Support is Still Experimental
2551 The ability to configure Perl's numbers to use "long doubles",
2552 floating point numbers of hopefully better accuracy, is still
2553 experimental. The implementations of long doubles are not yet
2554 widespread and the existing implementations are not quite mature
2555 or standardised, therefore trying to support them is a rare
2556 and moving target. The gain of more precision may also be offset
2557 by slowdown in computations (more bits to move around, and the
2558 operations are more likely to be executed by less optimised
2561 =head1 Reporting Bugs
2563 If you find what you think is a bug, you might check the articles
2564 recently posted to the comp.lang.perl.misc newsgroup and the perl
2565 bug database at http://bugs.perl.org. There may also be
2566 information at http://www.perl.com/perl/, the Perl Home Page.
2568 If you believe you have an unreported bug, please run the B<perlbug>
2569 program included with your release. Be sure to trim your bug down
2570 to a tiny but sufficient test case. Your bug report, along with the
2571 output of C<perl -V>, will be sent off to perlbug@perl.org to be
2572 analysed by the Perl porting team.
2576 The F<Changes> file for exhaustive details on what changed.
2578 The F<INSTALL> file for how to build Perl.
2580 The F<README> file for general stuff.
2582 The F<Artistic> and F<Copying> files for copyright information.
2586 Written by Jarkko Hietaniemi <F<jhi@iki.fi>>.