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 =head1 Highlights In 5.8.0
20 Better Unicode support
24 New Thread Implementation
32 Better Numeric Accuracy
40 More Extensive Regression Testing
44 =head1 Incompatible Changes
46 =head2 64-bit platforms and malloc
48 If your pointers are 64 bits wide, the Perl malloc is no longer being
49 used because it does not work well with 8-byte pointers. Also,
50 usually the system mallocs on such platforms are much better optimized
51 for such large memory models than the Perl malloc. Some memory-hungry
52 Perl applications like the PDL don't work well with Perl's malloc.
53 Finally, other applications than Perl (like modperl) tend to prefer
54 the system malloc. Such platforms include Alpha and 64-bit HPPA,
57 =head2 AIX Dynaloading
59 The AIX dynaloading now uses in AIX releases 4.3 and newer the native
60 dlopen interface of AIX instead of the old emulated interface. This
61 change will probably break backward compatibility with compiled
62 modules. The change was made to make Perl more compliant with other
63 applications like modperl which are using the AIX native interface.
65 =head2 Socket Extension Dynamic in VMS
67 The Socket extension is now dynamically loaded instead of being
68 statically built in. This may or may not be a problem with ancient
69 TCP/IP stacks of VMS: we do not know since we weren't able to test
70 Perl in such configurations.
72 =head2 IEEE-format Floating Point Default on OpenVMS Alpha
74 Perl now uses IEEE format (T_FLOAT) as the default internal floating
75 point format on OpenVMS Alpha, potentially breaking binary compatibility
76 with external libraries or existing data. G_FLOAT is still available as
77 a configuration option. The default on VAX (D_FLOAT) has not changed.
79 =head2 Different Definition of the Unicode Character Classes \p{In...}
81 As suggested by the Unicode consortium, the Unicode character classes
82 now prefer I<scripts> as opposed to I<blocks> (as defined by Unicode);
83 in Perl, when the C<\p{In....}> and the C<\p{In....}> regular expression
84 constructs are used. This has changed the definition of some of those
87 The difference between scripts and blocks is that scripts are the
88 glyphs used by a language or a group of languages, while the blocks
89 are more artificial groupings of 256 characters based on the Unicode
92 In general this change results in more inclusive Unicode character
93 classes, but changes to the other direction also do take place:
94 for example while the script C<Latin> includes all the Latin
95 characters and their various diacritic-adorned versions, it
96 does not include the various punctuation or digits (since they
97 are not solely C<Latin>).
99 Changes in the character class semantics may have happened if a script
100 and a block happen to have the same name, for example C<Hebrew>.
101 In such cases the script wins and C<\p{InHebrew}> now means the script
102 definition of Hebrew. The block definition in still available,
103 though, by appending C<Block> to the name: C<\p{InHebrewBlock}> means
104 what C<\p{InHebrew}> meant in perl 5.6.0. For the full list
105 of affected character classes, see L<perlunicode/Blocks>.
107 =head2 Perl Parser Stress Tested
109 The Perl parser has been stress tested using both random input and
110 Markov chain input and the few found crashes and lockups have been
113 =head2 REF(...) Instead Of SCALAR(...)
115 A reference to a reference now stringifies as "REF(0x81485ec)" instead
116 of "SCALAR(0x81485ec)" in order to be more consistent with the return
125 The semantics of bless(REF, REF) were unclear and until someone proves
126 it to make some sense, it is forbidden.
130 The obsolete chat2 library that should never have been allowed
131 to escape the laboratory has been decommissioned.
135 The very dusty examples in the eg/ directory have been removed.
136 Suggestions for new shiny examples welcome but the main issue is that
137 the examples need to be documented, tested and (most importantly)
142 The (bogus) escape sequences \8 and \9 now give an optional warning
143 ("Unrecognized escape passed through"). There is no need to \-escape
148 The list of filenames from glob() (or <...>) is now by default sorted
149 alphabetically to be csh-compliant (which is what happened before
150 in most UNIX platforms). (bsd_glob() does still sort platform
151 natively, ASCII or EBCDIC, unless GLOB_ALPHASORT is specified.)
155 Spurious syntax errors generated in certain situations, when glob()
156 caused File::Glob to be loaded for the first time, have been fixed.
160 Although "you shouldn't do that", it was possible to write code that
161 depends on Perl's hashed key order (Data::Dumper does this). The new
162 algorithm "One-at-a-Time" produces a different hashed key order.
163 More details are in L</"Performance Enhancements">.
167 lstat(FILEHANDLE) now gives a warning because the operation makes no sense.
168 In future releases this may become a fatal error.
172 The C<package;> syntax (C<package> without an argument) has been
173 deprecated. Its semantics were never that clear and its
174 implementation even less so. If you have used that feature to
175 disallow all but fully qualified variables, C<use strict;> instead.
179 The unimplemented POSIX regex features [[.cc.]] and [[=c=]] are still
180 recognised but now cause fatal errors. The previous behaviour of
181 ignoring them by default and warning if requested was unacceptable
182 since it, in a way, falsely promised that the features could be used.
186 The current user-visible implementation of pseudo-hashes (the weird
187 use of the first array element) is deprecated starting from Perl 5.8.0
188 and will be removed in Perl 5.10.0, and the feature will be
189 implemented differently. Not only is the current interface rather
190 ugly, but the current implementation slows down normal array and hash
191 use quite noticeably. The C<fields> pragma interface will remain
196 The syntaxes C<< @a->[...] >> and C<< %h->{...} >> have now been deprecated.
200 After years of trying the suidperl is considered to be too complex to
201 ever be considered truly secure. The suidperl functionality is likely
202 to be removed in a future release.
206 The long deprecated uppercase aliases for the string comparison
207 operators (EQ, NE, LT, LE, GE, GT) have now been removed.
211 The tr///C and tr///U features have been removed and will not return;
212 the interface was a mistake. Sorry about that. For similar
213 functionality, see pack('U0', ...) and pack('C0', ...).
217 =head1 Core Enhancements
219 =head2 PerlIO is Now The Default
225 IO is now by default done via PerlIO rather than system's "stdio".
226 PerlIO allows "layers" to be "pushed" onto a file handle to alter the
227 handle's behaviour. Layers can be specified at open time via 3-arg
230 open($fh,'>:crlf :utf8', $path) || ...
232 or on already opened handles via extended C<binmode>:
234 binmode($fh,':encoding(iso-8859-7)');
236 The built-in layers are: unix (low level read/write), stdio (as in
237 previous Perls), perlio (re-implementation of stdio buffering in a
238 portable manner), crlf (does CRLF <=> "\n" translation as on Win32,
239 but available on any platform). A mmap layer may be available if
240 platform supports it (mostly UNIXes).
242 Layers to be applied by default may be specified via the 'open' pragma.
244 See L</"Installation and Configuration Improvements"> for the effects
245 of PerlIO on your architecture name.
249 File handles can be marked as accepting Perl's internal encoding of Unicode
250 (UTF-8 or UTF-EBCDIC depending on platform) by a pseudo layer ":utf8" :
252 open($fh,">:utf8","Uni.txt");
254 Note for EBCDIC users: the pseudo layer ":utf8" is erroneously named
255 for you since it's not UTF-8 what you will be getting but instead
256 UTF-EBCDIC. See L<perlunicode>, L<utf8>, and
257 http://www.unicode.org/unicode/reports/tr16/ for more information.
258 In future releases this naming may change.
262 File handles can translate character encodings from/to Perl's internal
263 Unicode form on read/write via the ":encoding()" layer.
267 File handles can be opened to "in memory" files held in Perl scalars via:
269 open($fh,'>', \$variable) || ...
273 Anonymous temporary files are available without need to
274 'use FileHandle' or other module via
276 open($fh,"+>", undef) || ...
278 That is a literal undef, not an undefined value.
282 The list form of C<open> is now implemented for pipes (at least on UNIX):
284 open($fh,"-|", 'cat', '/etc/motd')
286 creates a pipe, and runs the equivalent of exec('cat', '/etc/motd') in
291 =head2 Signals Are Now Safe
293 Perl used to be fragile in that signals arriving at inopportune moments
294 could corrupt Perl's internal state. Now Perl postpones handling of
295 signals until it's safe.
297 =head2 Unicode Overhaul
299 Unicode in general should be now much more usable than in Perl 5.6.0
300 (or even in 5.6.1). Unicode can be used in hash keys, Unicode in
301 regular expressions should work now, Unicode in tr/// should work now,
302 Unicode in I/O should work now.
308 The Unicode Character Database coming with Perl has been upgraded
309 to Unicode 3.1.1. For more information, see http://www.unicode.org/.
313 For developers interested in enhancing Perl's Unicode capabilities:
314 almost all the UCD files are included with the Perl distribution in
315 the lib/unicore subdirectory. The most notable omission, for space
316 considerations, is the Unihan database.
320 The Unicode character classes \p{Blank} and \p{SpacePerl} have been
321 added. "Blank" is like C isblank(), that is, it contains only
322 "horizontal whitespace" (the space character is, the newline isn't),
323 and the "SpacePerl" is the Unicode equivalent of C<\s> (\p{Space}
324 isn't, since that includes the vertical tabulator character, whereas
329 =head2 Understanding of Numbers
331 In general a lot of fixing has happened in the area of Perl's
332 understanding of numbers, both integer and floating point. Since in
333 many systems the standard number parsing functions like C<strtoul()>
334 and C<atof()> seem to have bugs, Perl tries to work around their
335 deficiencies. This results hopefully in more accurate numbers.
337 Perl now tries internally to use integer values in numeric conversions
338 and basic arithmetics (+ - * /) if the arguments are integers, and
339 tries also to keep the results stored internally as integers.
340 This change leads to often slightly faster and always less lossy
341 arithmetics. (Previously Perl always preferred floating point numbers
344 =head2 Miscellaneous Enhancements
350 AUTOLOAD is now lvaluable, meaning that you can add the :lvalue attribute
351 to AUTOLOAD subroutines and you can assign to the AUTOLOAD return value.
355 C<perl -d:Module=arg,arg,arg> now works (previously one couldn't pass
356 in multiple arguments.)
360 END blocks are now run even if you exit/die in a BEGIN block.
361 Internally, the execution of END blocks is now controlled by
362 PL_exit_flags & PERL_EXIT_DESTRUCT_END. This enables the new
363 behaviour for Perl embedders. This will default in 5.10. See
368 Formats now support zero-padded decimal fields.
372 Lvalue subroutines can now return C<undef> in list context.
373 However, the lvalue subroutine feature still remains experimental.
377 A new special regular expression variable has been introduced:
378 C<$^N>, which contains the most-recently closed group (submatch).
382 C<no Module;> now works even if there is no "sub unimport" in the Module.
386 The numerical comparison operators return C<undef> if either operand
387 is a NaN. Previously the behaviour was unspecified.
391 The following builtin functions are now overridable: each(), keys(),
392 pop(), push(), shift(), splice(), unshift().
396 C<pack('U0a*', ...)> can now be used to force a string to UTF8.
400 my __PACKAGE__ $obj now works.
404 The printf() and sprintf() now support parameter reordering using the
405 C<%\d+\$> and C<*\d+\$> syntaxes. For example
407 print "%2\$s %1\$s\n", "foo", "bar";
409 will print "bar foo\n". This feature helps in writing
410 internationalised software, and in general when the order
411 of the parameters can vary.
415 prototype(\&) is now available.
419 prototype(\[$@%&]) is now available to implicitly create references
420 (useful for example if you want to emulate the tie() interface).
424 untie() will now call an UNTIE() hook if it exists. See L<perltie>
429 L<utime> now supports C<utime undef, undef, @files> to change the
430 file timestamps to the current time.
434 The rules for allowing underscores (underbars) in numeric constants
435 have been relaxed and simplified: now you can have an underscore
436 simply B<between digits>.
440 =head1 Modules and Pragmata
442 =head2 New Modules and Pragmata
448 C<Attribute::Handlers> allows a class to define attribute handlers.
451 use Attribute::Handlers;
452 sub Wolf :ATTR(SCALAR) { print "howl!\n" }
454 # later, in some package using or inheriting from MyPack...
456 my MyPack $Fluffy : Wolf; # the attribute handler Wolf will be called
458 Both variables and routines can have attribute handlers. Handlers can
459 be specific to type (SCALAR, ARRAY, HASH, or CODE), or specific to the
460 exact compilation phase (BEGIN, CHECK, INIT, or END).
464 B<B::Concise> is a new compiler backend for walking the Perl syntax
465 tree, printing concise info about ops, from Stephen McCamant. The
466 output is highly customisable. See L<B::Concise>.
470 C<Class::ISA> for reporting the search path for a class's ISA tree,
471 by Sean Burke, has been added. See L<Class::ISA>.
475 C<Cwd> has now a split personality: if possible, an XS extension is
476 used, (this will hopefully be faster, more secure, and more robust)
477 but if not possible, the familiar Perl implementation is used.
481 C<Devel::PPPort>, originally from Kenneth Albanowski and now
482 maintained by Paul Marquess, has been added. It is primarily used
483 by C<h2xs> to enhance portability of of XS modules between different
488 C<Digest>, frontend module for calculating digests (checksums), from
489 Gisle Aas, has been added. See L<Digest>.
493 C<Digest::MD5> for calculating MD5 digests (checksums) as defined in
494 RFC 1321, from Gisle Aas, has been added. See L<Digest::MD5>.
496 use Digest::MD5 'md5_hex';
498 $digest = md5_hex("Thirsty Camel");
500 print $digest, "\n"; # 01d19d9d2045e005c3f1b80e8b164de1
502 NOTE: the C<MD5> backward compatibility module is deliberately not
503 included since its further use is discouraged.
507 C<Encode>, by Nick Ing-Simmons, provides a mechanism to translate
508 between different character encodings. Support for Unicode,
509 ISO-8859-*, ASCII, CP*, KOI8-R, and three variants of EBCDIC are
510 compiled in to the module. Several other encodings (like Japanese,
511 Chinese, and MacIntosh encodings) are included and will be loaded at
512 runtime. See L<Encode>.
514 Any encoding supported by Encode module is also available to the
515 ":encoding()" layer if PerlIO is used.
519 C<I18N::Langinfo> can be use to query locale information.
520 See L<I18N::Langinfo>.
524 C<I18N::LangTags> has functions for dealing with RFC3066-style
525 language tags, by Sean Burke. See L<I18N::LangTags>.
529 C<ExtUtils::Constant> is a new tool for extension writers for
530 generating XS code to import C header constants, by Nicholas Clark.
531 See L<ExtUtils::Constant>.
535 C<Filter::Simple> is an easy-to-use frontend to Filter::Util::Call,
536 from Damian Conway. See L<Filter::Simple>.
542 use Filter::Simple sub {
543 while (my ($from, $to) = splice @_, 0, 2) {
552 use MyFilter qr/red/ => 'green';
554 print "red\n"; # this code is filtered, will print "green\n"
555 print "bored\n"; # this code is filtered, will print "bogreen\n"
559 print "red\n"; # this code is not filtered, will print "red\n"
563 C<File::Temp> allows one to create temporary files and directories in
564 an easy, portable, and secure way, by Tim Jenness. See L<File::Temp>.
568 C<Filter::Util::Call> provides you with the framework to write
569 I<Source Filters> in Perl, from Paul Marquess. For most uses the
570 frontend Filter::Simple is to be preferred. See L<Filter::Util::Call>.
574 L<libnet> is a collection of perl5 modules related to network
575 programming, from Graham Barr. See L<Net::FTP>, L<Net::NNTP>,
576 L<Net::Ping>, L<Net::POP3>, L<Net::SMTP>, and L<Net::Time>.
578 Perl installation leaves libnet unconfigured, use F<libnetcfg> to configure.
582 C<List::Util> is a selection of general-utility list subroutines, like
583 sum(), min(), first(), and shuffle(), by Graham Barr. See L<List::Util>.
587 C<Locale::Constants>, C<Locale::Country>, C<Locale::Currency>, and
588 C<Locale::Language>, from Neil Bowers, have been added. They provide the
589 codes for various locale standards, such as "fr" for France, "usd" for
590 US Dollar, and "jp" for Japanese.
594 $country = code2country('jp'); # $country gets 'Japan'
595 $code = country2code('Norway'); # $code gets 'no'
597 See L<Locale::Constants>, L<Locale::Country>, L<Locale::Currency>,
598 and L<Locale::Language>.
602 C<Locale::Maketext> is localization framework from Sean Burke. See
603 L<Locale::Maketext>, and L<Locale::Maketext::TPJ13>. The latter is an
604 article about software localization, originally published in The Perl
605 Journal #13, republished here with kind permission.
609 C<Memoize> can make your functions faster by trading space for time,
610 from Mark-Jason Dominus. See L<Memoize>.
614 C<MIME::Base64> allows you to encode data in base64, from Gisle Aas,
615 as defined in RFC 2045 - I<MIME (Multipurpose Internet Mail
620 $encoded = encode_base64('Aladdin:open sesame');
621 $decoded = decode_base64($encoded);
623 print $encoded, "\n"; # "QWxhZGRpbjpvcGVuIHNlc2FtZQ=="
629 C<MIME::QuotedPrint> allows you to encode data in quoted-printable
630 encoding, as defined in RFC 2045 - I<MIME (Multipurpose Internet Mail
631 Extensions)>, from Gisle Aas.
633 use MIME::QuotedPrint;
635 $encoded = encode_qp("Smiley in Unicode: \x{263a}");
636 $decoded = decode_qp($encoded);
638 print $encoded, "\n"; # "Smiley in Unicode: =263A"
640 MIME::QuotedPrint has been enhanced to provide the basic methods
641 necessary to use it with PerlIO::Via as in :
643 use MIME::QuotedPrint;
644 open($fh,">Via(MIME::QuotedPrint)",$path);
646 See L<MIME::QuotedPrint>.
650 C<NEXT> is pseudo-class for method redispatch, from Damian Conway.
655 C<open> is a new pragma for setting the default I/O disciplines
660 C<PerlIO::Scalar> provides the implementation of IO to "in memory"
661 Perl scalars as discussed above, from Nick Ing-Simmons. It also
662 serves as an example of a loadable PerlIO layer. Other future
663 possibilities include PerlIO::Array and PerlIO::Code.
664 See L<PerlIO::Scalar>.
668 C<PerlIO::Via> acts as a PerlIO layer and wraps PerlIO layer
669 functionality provided by a class (typically implemented in perl
670 code), from Nick Ing-Simmons.
672 use MIME::QuotedPrint;
673 open($fh,">Via(MIME::QuotedPrint)",$path);
675 This will automatically convert everything output to C<$fh>
676 to Quoted-Printable. See L<PerlIO::Via>.
680 C<Pod::ParseLink>, by Russ Allbery, has been added,
681 to parse L<> links in pods as described in the new
686 C<Pod::Text::Overstrike>, by Joe Smith, has been added.
687 It converts POD data to formatted overstrike text.
688 See L<Pod::Text::Overstrike>.
692 C<Scalar::Util> is a selection of general-utility scalar subroutines,
693 like blessed(), reftype(), and tainted(). See L<Scalar::Util>.
697 C<sort> is a new pragma for controlling the behaviour of sort().
701 C<Storable> gives persistence to Perl data structures by allowing the
702 storage and retrieval of Perl data to and from files in a fast and
703 compact binary format, from Raphael Manfredi. See L<Storable>.
707 C<Switch>, from Damian Conway, has been added. Just by saying
711 you have C<switch> and C<case> available in Perl.
717 case 1 { print "number 1" }
718 case "a" { print "string a" }
719 case [1..10,42] { print "number in list" }
720 case (@array) { print "number in list" }
721 case /\w+/ { print "pattern" }
722 case qr/\w+/ { print "pattern" }
723 case (%hash) { print "entry in hash" }
724 case (\%hash) { print "entry in hash" }
725 case (\&sub) { print "arg to subroutine" }
726 else { print "previous case not true" }
733 C<Test::More> is yet another framework for writing test scripts,
734 more extensive than Test::Simple, by Michael Schwern. See L<Test::More>.
738 C<Test::Simple> has basic utilities for writing tests, by Michael
739 Schwern. See L<Test::Simple>.
743 C<Text::Balanced> has been added, for extracting delimited text
744 sequences from strings, from Damian Conway.
746 use Text::Balanced 'extract_delimited';
748 ($a, $b) = extract_delimited("'never say never', he never said", "'", '');
750 $a will be "'never say never'", $b will be ', he never said'.
752 In addition to extract_delimited() there are also extract_bracketed(),
753 extract_quotelike(), extract_codeblock(), extract_variable(),
754 extract_tagged(), extract_multiple(), gen_delimited_pat(), and
755 gen_extract_tagged(). With these you can implement rather advanced
756 parsing algorithms. See L<Text::Balanced>.
760 C<threads> is an interface to interpreter threads, by Arthur Bergman.
761 Interpreter threads (ithreads) is the new thread model introduced in
762 Perl 5.6 but only available as an internal interface for extension
763 writers (and for Win32 Perl for C<fork()> emulation). See L<threads>.
767 C<threads::shared> allows data sharing for interpreter threads, from
768 Arthur Bergman. In the ithreads model any data sharing between
769 threads must be explicit, as opposed to the old 5.005 thread model
770 where data sharing was implicit. See L<threads::shared>.
774 C<Tie::RefHash::Nestable>, by Edward Avis, allows storing hash
775 references (unlike the standard Tie::RefHash) The module is contained
776 within Tie::RefHash, see L<Tie::RefHash>.
780 C<Time::HiRes> provides high resolution timing (ualarm, usleep,
781 and gettimeofday), from Douglas E. Wegscheid. See L<Time::HiRes>.
785 C<Unicode::UCD> offers a querying interface to the Unicode Character
786 Database. See L<Unicode::UCD>.
790 C<Unicode::Collate> implements the UCA (Unicode Collation Algorithm)
791 for sorting Unicode strings, by SADAHIRO Tomoyuki. See L<Unicode::Collate>.
795 C<Unicode::Normalize> implements the various Unicode normalization
796 forms, by SADAHIRO Tomoyuki. See L<Unicode::Normalize>.
800 C<XS::Typemap>, by Tim Jenness, is a test extension that exercises XS
801 typemaps. Nothing gets installed but for extension writers the code
806 =head2 Updated And Improved Modules and Pragmata
812 The following independently supported modules have been updated to the
813 newest versions from CPAN: CGI, CPAN, DB_File, File::Spec, File::Temp,
814 Getopt::Long, Math::BigFloat, Math::BigInt, the podlators bundle
815 (Pod::Man, Pod::Text), Pod::LaTeX, Pod::Parser, Storable,
816 Term::ANSIColor, Test, Text-Tabs+Wrap.
820 The attributes::reftype() now works on tied arguments.
824 AutoLoader can now be disabled with C<no AutoLoader;>.
828 B::Deparse has been significantly enhanced. It now can deparse almost
829 all of the standard test suite (so that the tests still succeed).
830 There is a make target "test.deparse" for trying this out.
834 Class::Struct can now define the classes in compile time.
838 Class::Struct now assigns the array/hash element if the accessor
839 is called with an array/hash element as the B<sole> argument.
843 Data::Dumper has now an option to sort hashes.
847 Data::Dumper has now an option to dump code references
852 DB_File now supports newer Berkeley DB versions, among
857 The English module can now be used without the infamous performance
860 use English '-no_performance_hit';
862 (Assuming, of course, that one doesn't need the troublesome variables
863 C<$`>, C<$&>, or C<$'>.) Also, introduced C<@LAST_MATCH_START> and
864 C<@LAST_MATCH_END> English aliases for C<@-> and C<@+>.
868 Fcntl, Socket, and Sys::Syslog have been rewritten to use the
869 new-style constant dispatch section (see L<ExtUtils::Constant>).
870 This means that they will be more robust and hopefully faster.
874 File::Find now chdir()s correctly when chasing symbolic links.
878 File::Find now has pre- and post-processing callbacks. It also
879 correctly changes directories when chasing symbolic links. Callbacks
880 (naughtily) exiting with "next;" instead of "return;" now work.
884 File::Find is now (again) reentrant. It also has been made
889 File::Glob::glob() renamed to File::Glob::bsd_glob() to avoid
890 prototype mismatch with CORE::glob().
894 File::Glob now supports C<GLOB_LIMIT> constant to limit the size of
895 the returned list of filenames.
899 Devel::Peek now has an interface for the Perl memory statistics
900 (this works only if you are using perl's malloc, and if you have
901 compiled with debugging).
905 IPC::Open3 now allows the use of numeric file descriptors.
909 IO::Socket has now atmark() method, which returns true if the socket
910 is positioned at the out-of-band mark. The method is also exportable
911 as a sockatmark() function.
915 IO::Socket::INET has support for ReusePort option (if your platform
916 supports it). The Reuse option now has an alias, ReuseAddr. For clarity
917 you may want to prefer ReuseAddr.
921 IO::Socket::INET now supports C<LocalPort> of zero (usually meaning
922 that the operating system will make one up.)
926 use lib now works identically to @INC. Removing directories
927 with 'no lib' now works.
931 Math::BigFloat and Math::BigInt have undergone a full rewrite.
932 They are now magnitudes faster, and they support various
933 bignum libraries such as GMP and PARI as their backends.
937 Math::Complex handles inf, NaN etc., better.
941 Net::Ping has been enhanced. There is now "external" protocol which
942 uses Net::Ping::External module which runs external ping(1) and parses
943 the output. A version of Net::Ping::External is available in CPAN.
947 POSIX::sigaction() is now much more flexible and robust.
948 You can now install coderef handlers, 'DEFAULT', and 'IGNORE'
949 handlers, installing new handlers was not atomic.
953 In Safe the C<%INC> now localised in a Safe compartment so that
958 In SDBM_File on dosish platforms, some keys went missing because of
959 lack of support for files with "holes". A workaround for the problem
964 In Search::Dict one can now have a pre-processing hook for the
965 lines being searched.
969 The Shell module now has an OO interface.
973 The Test module has been significantly enhanced.
977 The vars pragma now supports declaring fully qualified variables.
978 (Something that C<our()> does not and will not support.)
982 The utf8:: name space (as in the pragma) provides various
983 Perl-callable functions to provide low level access to Perl's
984 internal Unicode representation. At the moment only length()
985 has been implemented.
989 =head1 Utility Changes
995 Emacs perl mode (emacs/cperl-mode.el) has been updated to version
1000 F<emacs/e2ctags.pl> is now much faster.
1004 C<h2ph> now supports C trigraphs.
1008 C<h2xs> now produces a template README.
1012 C<h2xs> now uses C<Devel::PPort> for better portability between
1013 different versions of Perl.
1017 C<h2xs> uses the new L<ExtUtils::Constant> module which will affect
1018 newly created extensions that define constants. Since the new code is
1019 more correct (if you have two constants where the first one is a
1020 prefix of the second one, the first constant B<never> gets defined),
1021 less lossy (it uses integers for integer constant, as opposed to the
1022 old code that used floating point numbers even for integer constants),
1023 and slightly faster, you might want to consider regenerating your
1024 extension code (the new scheme makes regenerating easy).
1025 L<h2xs> now also supports C trigraphs.
1029 C<libnetcfg> has been added to configure the libnet.
1033 C<perlbug> is now much more robust. It also sends the bug report to
1034 perl.org, not perl.com.
1038 C<perlcc> has been rewritten and its user interface (that is,
1039 command line) is much more like that of the UNIX C compiler, cc.
1040 (The perlbc tools has been removed. Use C<perlcc -B> instead.)
1044 C<perlivp> is a new Installation Verification Procedure utility
1045 for running any time after installing Perl.
1049 C<pod2html> now allows specifying a cache directory.
1053 C<s2p> has been completely rewritten in Perl. (It is in fact a full
1054 implementation of sed in Perl: you can use the sed functionality by
1055 using the C<psed> utility.)
1059 C<xsubpp> now understands POD documentation embedded in the *.xs files.
1063 C<xsubpp> now supports OUT keyword.
1067 =head1 New Documentation
1073 perl56delta details the changes between the 5.005 release and the
1078 perlclib documents the internal replacements for standard C library
1079 functions. (Interesting only for extension writers and Perl core
1084 perldebtut is a Perl debugging tutorial.
1088 perlebcdic contains considerations for running Perl on EBCDIC platforms.
1092 perlintro is a gentle introduction to Perl.
1096 perliol documents the internals of PerlIO with layers.
1100 perlmodstyle is a style guide for writing modules.
1104 perlnewmod tells about writing and submitting a new module.
1108 perlpacktut is a pack() tutorial.
1112 perlpod has been rewritten to be clearer and to record the best
1113 practices gathered over the years.
1117 perlpodspec is a more formal specification of the pod format,
1118 mainly of interest for writers of pod applications, not to
1119 people writing in pod.
1123 perlretut is a regular expression tutorial.
1127 perlrequick is a regular expressions quick-start guide.
1128 Yes, much quicker than perlretut.
1132 perltodo has been updated.
1136 perltootc has been renamed as perltooc (to not to conflict
1137 with perltoot in filesystems restricted to "8.3" names)
1141 perluniintro is an introduction to using Unicode in Perl
1142 (perlunicode is more of a reference)
1146 perlutil explains the command line utilities packaged with the Perl
1151 The following platform-specific documents are available before
1152 the installation as README.I<platform>, and after the installation
1155 perlaix perlamiga perlapollo perlbeos perlbs2000
1156 perlce perlcygwin perldgux perldos perlepoc perlhpux
1157 perlhurd perlmachten perlmacos perlmint perlmpeix
1158 perlnetware perlos2 perlos390 perlplan9 perlqnx perlsolaris
1159 perltru64 perluts perlvmesa perlvms perlvos perlwin32
1165 The documentation for the POSIX-BC platform is called "BS2000", to avoid
1166 confusion with the Perl POSIX module.
1170 The documentation for the WinCE platform is called "CE", to avoid
1171 confusion with the perlwin32 documentation on 8.3-restricted filesystems.
1175 =head1 Performance Enhancements
1181 map() could get pathologically slow when the result list it generates
1182 is larger than the source list. The performance has been improved for
1187 sort() has been changed to use primarily mergesort internally as
1188 opposed to the earlier quicksort. For very small lists this may
1189 result in slightly slower sorting times, but in general the speedup
1190 should be at least 20%. Additional bonuses are that the worst case
1191 behaviour of sort() is now better (in computer science terms it now
1192 runs in time O(N log N), as opposed to quicksort's Theta(N**2)
1193 worst-case run time behaviour), and that sort() is now stable
1194 (meaning that elements with identical keys will stay ordered as they
1195 were before the sort). See the C<sort> pragma for information.
1197 The story in more detail: suppose you want to serve yourself a little
1200 @digits = ( 3,1,4,1,5,9 );
1202 A numerical sort of the digits will yield (1,1,3,4,5,9), as expected.
1203 Which C<1> comes first is hard to know, since one C<1> looks pretty
1204 much like any other. You can regard this as totally trivial,
1205 or somewhat profound. However, if you just want to sort the even
1206 digits ahead of the odd ones, then what will
1208 sort { ($a % 2) <=> ($b % 2) } @digits;
1210 yield? The only even digit, C<4>, will come first. But how about
1211 the odd numbers, which all compare equal? With the quicksort algorithm
1212 used to implement Perl 5.6 and earlier, the order of ties is left up
1213 to the sort. So, as you add more and more digits of Pi, the order
1214 in which the sorted even and odd digits appear will change.
1215 and, for sufficiently large slices of Pi, the quicksort algorithm
1216 in Perl 5.8 won't return the same results even if reinvoked with the
1217 same input. The justification for this rests with quicksort's
1218 worst case behavior. If you run
1220 sort { $a <=> $b } ( 1 .. $N , 1 .. $N );
1222 (something you might approximate if you wanted to merge two sorted
1223 arrays using sort), doubling $N doesn't just double the quicksort time,
1224 it I<quadruples> it. Quicksort has a worst case run time that can
1225 grow like N**2, so-called I<quadratic> behaviour, and it can happen
1226 on patterns that may well arise in normal use. You won't notice this
1227 for small arrays, but you I<will> notice it with larger arrays,
1228 and you may not live long enough for the sort to complete on arrays
1229 of a million elements. So the 5.8 quicksort scrambles large arrays
1230 before sorting them, as a statistical defence against quadratic behaviour.
1231 But that means if you sort the same large array twice, ties may be
1232 broken in different ways.
1234 Because of the unpredictability of tie-breaking order, and the quadratic
1235 worst-case behaviour, quicksort was I<almost> replaced completely with
1236 a stable mergesort. I<Stable> means that ties are broken to preserve
1237 the original order of appearance in the input array. So
1239 sort { ($a % 2) <=> ($b % 2) } (3,1,4,1,5,9);
1241 will yield (4,3,1,1,5,9), guaranteed. The even and odd numbers
1242 appear in the output in the same order they appeared in the input.
1243 Mergesort has worst case O(NlogN) behaviour, the best value
1244 attainable. And, ironically, this mergesort does particularly
1245 well where quicksort goes quadratic: mergesort sorts (1..$N, 1..$N)
1246 in O(N) time. But quicksort was rescued at the last moment because
1247 it is faster than mergesort on certain inputs and platforms.
1248 For example, if you really I<don't> care about the order of even
1249 and odd digits, quicksort will run in O(N) time; it's very good
1250 at sorting many repetitions of a small number of distinct elements.
1251 The quicksort divide and conquer strategy works well on platforms
1252 with relatively small, very fast, caches. Eventually, the problem gets
1253 whittled down to one that fits in the cache, from which point it
1254 benefits from the increased memory speed.
1256 Quicksort was rescued by implementing a sort pragma to control aspects
1257 of the sort. The B<stable> subpragma forces stable behaviour,
1258 regardless of algorithm. The B<_quicksort> and B<_mergesort>
1259 subpragmas are heavy-handed ways to select the underlying implementation.
1260 The leading C<_> is a reminder that these subpragmas may not survive
1261 beyond 5.8. More appropriate mechanisms for selecting the implementation
1262 exist, but they wouldn't have arrived in time to save quicksort.
1266 Hashes now use Bob Jenkins "One-at-a-Time" hashing key algorithm
1267 (http://burtleburtle.net/bob/hash/doobs.html). This algorithm is
1268 reasonably fast while producing a much better spread of values than
1269 the old hashing algorithm (originally by Chris Torek, later tweaked by
1270 Ilya Zakharevich). Hash values output from the algorithm on a hash of
1271 all 3-char printable ASCII keys comes much closer to passing the
1272 DIEHARD random number generation tests. According to perlbench, this
1273 change has not affected the overall speed of Perl.
1277 unshift() should now be noticeably faster.
1281 =head1 Installation and Configuration Improvements
1283 =head2 Generic Improvements
1289 INSTALL now explains how you can configure Perl to use 64-bit
1290 integers even on non-64-bit platforms.
1294 Policy.sh policy change: if you are reusing a Policy.sh file
1295 (see INSTALL) and you use Configure -Dprefix=/foo/bar and in the old
1296 Policy $prefix eq $siteprefix and $prefix eq $vendorprefix, all of
1297 them will now be changed to the new prefix, /foo/bar. (Previously
1298 only $prefix changed.) If you do not like this new behaviour,
1299 specify prefix, siteprefix, and vendorprefix explicitly.
1303 A new optional location for Perl libraries, otherlibdirs, is available.
1304 It can be used for example for vendor add-ons without disturbing Perl's
1305 own library directories.
1309 In many platforms the vendor-supplied 'cc' is too stripped-down to
1310 build Perl (basically, 'cc' doesn't do ANSI C). If this seems
1311 to be the case and 'cc' does not seem to be the GNU C compiler
1312 'gcc', an automatic attempt is made to find and use 'gcc' instead.
1316 gcc needs to closely track the operating system release to avoid
1317 build problems. If Configure finds that gcc was built for a different
1318 operating system release than is running, it now gives a clearly visible
1319 warning that there may be trouble ahead.
1323 If binary compatibility with the 5.005 release is not wanted, Configure
1324 no longer suggests including the 5.005 modules in @INC.
1328 Configure C<-S> can now run non-interactively.
1332 Configure support for pdp11-style memory models has been removed due
1337 configure.gnu now works with options with whitespace in them.
1341 installperl now outputs everything to STDERR.
1345 $Config{byteorder} is now computed dynamically (this is more robust
1346 with "fat binaries" where an executable image contains binaries for
1347 more than one binary platform.)
1351 Because PerlIO is now the default on most platforms, "-perlio" doesn't
1352 get appended to the $Config{archname} (also known as $^O) anymore.
1353 Instead, if you explicitly choose not to use perlio (Configure command
1354 line option -Uuseperlio), you will get "-stdio" appended.
1358 Another change related to the architecture name is that "-64all"
1359 (-Duse64bitall, or "maximally 64-bit") is appended only if your
1360 pointers are 64 bits wide. (To be exact, the use64bitall is ignored.)
1364 In AFS installations one can configure the root of the AFS to be
1365 somewhere else than the default F</afs> by using the Configure
1366 parameter C<-Dafsroot=/some/where/else>.
1370 APPLLIB_EXP, a less-know configuration-time definition, has been
1371 documented. It can be used to prepend site-specific directories
1372 to Perl's default search path (@INC), see INSTALL for information.
1376 The version of Berkeley DB used when the Perl (and, presumably, the
1377 DB_File extension) was built is now available as
1378 C<@Config{qw(db_version_major db_version_minor db_version_patch)}>
1379 from Perl and as C<DB_VERSION_MAJOR_CFG DB_VERSION_MINOR_CFG
1380 DB_VERSION_PATCH_CFG> from C.
1384 Building Berkeley DB3 for compatibility modes for DB, NDBM, and ODBM
1385 has been documented in INSTALL.
1389 If you have CPAN access (either network or a local copy such as a
1390 CD-ROM) you can during specify extra modules to Configure to build and
1391 install with Perl using the -Dextras=... option. See INSTALL for
1396 In addition to config.over a new override file, config.arch, is
1397 available. That is supposed to be used by hints file writers for
1398 architecture-wide changes (as opposed to config.over which is for
1403 If your file system supports symbolic links you can build Perl outside
1404 of the source directory by
1406 mkdir /tmp/perl/build/directory
1407 cd /tmp/perl/build/directory
1408 sh /path/to/perl/source/Configure -Dmksymlinks ...
1410 This will create in /tmp/perl/build/directory a tree of symbolic links
1411 pointing to files in /path/to/perl/source. The original files are left
1412 unaffected. After Configure has finished you can just say
1416 and Perl will be built and tested, all in /tmp/perl/build/directory.
1420 For Perl developers several new make targets for profiling
1421 and debugging have been added, see L<perlhack>.
1427 Use of the F<gprof> tool to profile Perl has been documented in
1428 L<perlhack>. There is a make target called "perl.gprof" for
1429 generating a gprofiled Perl executable.
1433 If you have GCC 3, there is a make target called "perl.gcov" for
1434 creating a gcoved Perl executable for coverage analysis. See
1439 If you are on IRIX or Tru64 platforms, new profiling/debugging options
1440 have been added, see L<perlhack> for more information about pixie and
1447 Guidelines of how to construct minimal Perl installations have
1448 been added to INSTALL.
1452 The Thread extension is now not built at all under ithreads
1453 (C<Configure -Duseithreads>) because it wouldn't work anyway (the
1454 Thread extension requires being Configured with C<-Duse5005threads>).
1456 But note that the Thread.pm interface is now shared by both
1461 =head2 New Or Improved Platforms
1463 For the list of platforms known to support Perl,
1464 see L<perlport/"Supported Platforms">.
1470 AIX dynamic loading should be now better supported.
1474 AIX should now work better with gcc, threads, and 64-bitness. Also the
1475 long doubles support in AIX should be better now. See L<perlaix>.
1479 After a long pause, AmigaOS has been verified to be happy with Perl.
1483 AtheOS (http://www.atheos.cx/) is a new platform.
1487 DG/UX platform now supports the 5.005-style threads. See L<perldgux>.
1491 DYNIX/ptx platform (a.k.a. dynixptx) is supported at or near osvers 4.5.2.
1495 EBCDIC platforms (z/OS, also known as OS/390, POSIX-BC, and VM/ESA)
1496 have been regained. Many test suite tests still fail and the
1497 co-existence of Unicode and EBCDIC isn't quite settled, but the
1498 situation is much better than with Perl 5.6. See L<perlos390>,
1499 L<perlbs2000> (for POSIX-BC), and L<perlvmesa> for more information.
1503 Building perl with -Duseithreads or -Duse5005threads now works under
1504 HP-UX 10.20 (previously it only worked under 10.30 or later). You will
1505 need a thread library package installed. See README.hpux.
1509 MacOS Classic (MacPerl has of course been available since
1510 perl 5.004 but now the source code bases of standard Perl
1511 and MacPerl have been synchronised)
1515 MacOS X (or Darwin) should now be able to build Perl even on HFS+
1516 filesystems. (The case-insensitivity confused the Perl build process.)
1520 NCR MP-RAS is now supported.
1524 NetWare from Novell is now supported. See L<perlnetware>.
1528 NonStop-UX is now supported.
1532 NEC SUPER-UX is now supported.
1536 Amdahl UTS UNIX mainframe platform is now supported.
1540 WinCE is now supported. See L<perlce>.
1544 z/OS (formerly known as OS/390, formerly known as MVS OE) has now
1545 support for dynamic loading. This is not selected by default,
1546 however, you must specify -Dusedl in the arguments of Configure.
1550 =head1 Selected Bug Fixes
1552 Numerous memory leaks and uninitialized memory accesses have been
1553 hunted down. Most importantly anonymous subs used to leak quite
1560 The autouse pragma didn't work for Multi::Part::Function::Names.
1564 caller() could cause core dumps in certain situations. Carp was sometimes
1565 affected by this problem.
1569 chop(@list) in list context returned the characters chopped in
1570 reverse order. This has been reversed to be in the right order.
1574 Configure no longer includes the DBM libraries (dbm, gdbm, db, ndbm)
1575 when building the Perl binary. The only exception to this is SunOS 4.x,
1580 The behaviour of non-decimal but numeric string constants such as
1581 "0x23" was platform-dependent: in some platforms that was seen as 35,
1582 in some as 0, in some as a floating point number (don't ask). This
1583 was caused by Perl using the operating system libraries in a situation
1584 where the result of the string to number conversion is undefined: now
1585 Perl consistently handles such strings as zero in numeric contexts.
1589 The order of DESTROYs has been made more predictable.
1593 Several debugger fixes: exit code now reflects the script exit code,
1594 condition C<"0"> now treated correctly, the C<d> command now checks
1595 line number, the C<$.> no longer gets corrupted, all debugger output
1596 now goes correctly to the socket if RemotePort is set.
1600 Perl 5.6.0 could emit spurious warnings about redefinition of dl_error()
1601 when statically building extensions into perl. This has been corrected.
1605 L<dprofpp> -R didn't work.
1609 C<*foo{FORMAT}> now works.
1612 Infinity is now recognized as a number.
1616 UNIVERSAL::isa no longer caches methods incorrectly. (This broke
1617 the Tk extension with 5.6.0.)
1621 Lexicals I: lexicals outside an eval "" weren't resolved
1622 correctly inside a subroutine definition inside the eval "" if they
1623 were not already referenced in the top level of the eval""ed code.
1627 Lexicals II: lexicals leaked at file scope into subroutines that
1628 were declared before the lexicals.
1632 Lexical warnings now propagating correctly between scopes
1633 and into C<eval "...">.
1637 C<use warnings qw(FATAL all)> did not work as intended. This has been
1642 warnings::enabled() now reports the state of $^W correctly if the caller
1643 isn't using lexical warnings.
1647 Line renumbering with eval and C<#line> now works.
1651 Fixed numerous memory leaks, especially in eval "".
1655 mkdir() now ignores trailing slashes in the directory name,
1656 as mandated by POSIX.
1660 Some versions of glibc have a broken modfl(). This affects builds
1661 with C<-Duselongdouble>. This version of Perl detects this brokenness
1662 and has a workaround for it. The glibc release 2.2.2 is known to have
1663 fixed the modfl() bug.
1667 Modulus of unsigned numbers now works (4063328477 % 65535 used to
1668 return 27406, instead of 27047).
1672 Some "not a number" warnings introduced in 5.6.0 eliminated to be
1673 more compatible with 5.005. Infinity is now recognised as a number.
1677 Numeric conversions did not recognize changes in the string value
1678 properly in certain circumstances.
1682 Attributes (like :shared) didn't work with our().
1686 our() variables will not cause "will not stay shared" warnings.
1690 "our" variables of the same name declared in two sibling blocks
1691 resulted in bogus warnings about "redeclaration" of the variables.
1692 The problem has been corrected.
1696 pack "Z" now correctly terminates the string with "\0".
1700 Fix password routines which in some shadow password platforms
1701 (e.g. HP-UX) caused getpwent() to return every other entry.
1705 The PERL5OPT environment variable (for passing command line arguments
1706 to Perl) didn't work for more than a single group of options.
1710 PERL5OPT with embedded spaces didn't work.
1714 printf() no longer resets the numeric locale to "C".
1718 C<qw(a\\b)> now parses correctly as C<'a\\b'>.
1722 pos() did not return the correct value within s///ge in earlier
1723 versions. This is now handled correctly.
1727 Printing quads (64-bit integers) with printf/sprintf now works
1728 without the q L ll prefixes (assuming you are on a quad-capable platform).
1732 Regular expressions on references and overloaded scalars now work.
1736 Right-hand side magic (GMAGIC) could in many cases such as string
1737 concatenation be invoked too many times.
1741 scalar() now forces scalar context even when used in void context.
1745 SOCKS support is now much more robust.
1749 sort() arguments are now compiled in the right wantarray context
1750 (they were accidentally using the context of the sort() itself).
1751 The comparison block is now run in scalar context, and the arguments
1752 to be sorted are always provided list context.
1756 Changed the POSIX character class C<[[:space:]]> to include the (very
1757 rarely used) vertical tab character. Added a new POSIX-ish character
1758 class C<[[:blank:]]> which stands for horizontal whitespace
1759 (currently, the space and the tab).
1763 The tainting behaviour of sprintf() has been rationalized. It does
1764 not taint the result of floating point formats anymore, making the
1765 behaviour consistent with that of string interpolation.
1769 Some cases of inconsistent taint propagation (such as within hash
1770 values) have been fixed.
1774 The RE engine found in Perl 5.6.0 accidentally pessimised certain kinds
1775 of simple pattern matches. These are now handled better.
1779 Regular expression debug output (whether through C<use re 'debug'>
1780 or via C<-Dr>) now looks better.
1784 Multi-line matches like C<"a\nxb\n" =~ /(?!\A)x/m> were flawed. The
1789 Use of $& could trigger a core dump under some situations. This
1794 The regular expression captured submatches ($1, $2, ...) are now
1795 more consistently unset if the match fails, instead of leaving false
1796 data lying around in them.
1800 readline() on files opened in "slurp" mode could return an extra "" at
1801 the end in certain situations. This has been corrected.
1805 Autovivification of symbolic references of special variables described
1806 in L<perlvar> (as in C<${$num}>) was accidentally disabled. This works
1811 Sys::Syslog ignored the C<LOG_AUTH> constant.
1815 All but the first argument of the IO syswrite() method are now optional.
1819 $AUTOLOAD, sort(), lock(), and spawning subprocesses
1820 in multiple threads simultaneously are now thread-safe.
1824 Tie::ARRAY SPLICE method was broken.
1828 Allow read-only string on left hand side of non-modifying tr///.
1832 Several Unicode fixes.
1838 BOMs (byte order marks) in the beginning of Perl files
1839 (scripts, modules) should now be transparently skipped.
1840 UTF-16 (UCS-2) encoded Perl files should now be read correctly.
1844 The character tables have been updated to Unicode 3.1.1.
1848 Comparing with utf8 data does not magically upgrade non-utf8 data
1853 C<IsAlnum>, C<IsAlpha>, and C<IsWord> now match titlecase.
1857 Concatenation with the C<.> operator or via variable interpolation,
1858 C<eq>, C<substr>, C<reverse>, C<quotemeta>, the C<x> operator,
1859 substitution with C<s///>, single-quoted UTF8, should now work.
1863 The C<tr///> operator now works. Note that the C<tr///CU>
1864 functionality has been removed (but see pack('U0', ...)).
1868 C<eval "v200"> now works.
1872 Perl 5.6.0 parsed m/\x{ab}/ incorrectly, leading to spurious warnings.
1873 This has been corrected.
1877 Zero entries were missing from the Unicode classes like C<IsDigit>.
1883 Large unsigned numbers (those above 2**31) could sometimes lose their
1884 unsignedness, causing bogus results in arithmetic operations.
1888 =head2 Platform Specific Changes and Fixes
1896 Perl now works on post-4.0 BSD/OSes.
1902 Setting C<$0> now works (as much as possible; see L<perlvar> for details).
1908 Numerous updates; currently synchronised with Cygwin 1.1.4.
1912 Previously DYNIX/ptx had problems in its Configure probe for non-blocking I/O.
1918 EPOC update after Perl 5.6.0. See README.epoc.
1924 Perl now works on post-3.0 FreeBSDs.
1930 README.hpux updated; C<Configure -Duse64bitall> now almost works.
1936 Numerous compilation flag and hint enhancements; accidental mixing
1937 of 32-bit and 64-bit libraries (a doomed attempt) made much harder.
1947 Long doubles should now work (see INSTALL).
1951 Linux previously had problems related to sockaddrlen when using
1952 accept(), revcfrom() (in Perl: recv()), getpeername(), and getsockname().
1960 Compilation of the standard Perl distribution in MacOS Classic should
1961 now work if you have the Metrowerks development environment and
1962 the missing Mac-specific toolkit bits. Contact the macperl mailing
1969 MPE/iX update after Perl 5.6.0. See README.mpeix.
1975 Perl now works on NetBSD/sparc.
1981 Now works with usethreads (see INSTALL).
1987 64-bitness using the Sun Workshop compiler now works.
1991 Tru64 (aka Digital UNIX, aka DEC OSF/1)
1993 The operating system version letter now recorded in $Config{osvers}.
1994 Allow compiling with gcc (previously explicitly forbidden). Compiling
1995 with gcc still not recommended because buggy code results, even with
2002 Fixed various alignment problems that lead into core dumps either
2003 during build or later; no longer dies on math errors at runtime;
2004 now using full quad integers (64 bits), previously was using
2005 only 46 bit integers for speed.
2011 chdir() now works better despite a CRT bug; now works with MULTIPLICITY
2012 (see INSTALL); now works with Perl's malloc.
2014 The tainting of C<%ENV> elements via C<keys> or C<values> was previously
2015 unimplemented. It now works as documented.
2017 The C<waitpid> emulation has been improved. The worst bug (now fixed)
2018 was that a pid of -1 would cause a wildcard search of all processes on
2019 the system. The most significant enhancement is that we can now
2020 usually get the completion status of a terminated process.
2022 POSIX-style signals are now emulated much better on VMS versions prior
2025 The C<system> function and backticks operator have improved
2026 functionality and better error handling.
2036 accept() no longer leaks memory.
2040 Borland C++ v5.5 is now a supported compiler that can build Perl.
2041 However, the generated binaries continue to be incompatible with those
2042 generated by the other supported compilers (GCC and Visual C++).
2046 Better chdir() return value for a non-existent directory.
2050 Duping socket handles with open(F, ">&MYSOCK") now works under Windows 9x.
2054 New %ENV entries now propagate to subprocesses.
2058 Current directory entries in %ENV are now correctly propagated to child
2063 $ENV{LIB} now used to search for libs under Visual C.
2067 fork() emulation has been improved in various ways, but still continues
2068 to be experimental. See L<perlfork> for known bugs and caveats.
2072 A failed (pseudo)fork now returns undef and sets errno to EAGAIN.
2076 Win32::GetCwd() correctly returns C:\ instead of C: when at the drive root.
2077 Other bugs in chdir() and Cwd::cwd() have also been fixed.
2081 HTML files will be installed in c:\perl\html instead of c:\perl\lib\pod\html
2085 The makefiles now provide a single switch to bulk-enable all the features
2086 enabled in ActiveState ActivePerl (a popular Win32 binary distribution).
2090 Allow REG_EXPAND_SZ keys in the registry.
2094 Can now send() from all threads, not just the first one.
2098 Fake signal handling reenabled, bugs and all.
2102 %SIG has been enabled under USE_ITHREADS, but its use is completely
2103 unsupported under all configurations.
2107 Less stack reserved per thread so that more threads can run
2108 concurrently. (Still 16M per thread.)
2112 C<File::Spec->tmpdir()> now prefers C:/temp over /tmp
2113 (works better when perl is running as service).
2117 Better UNC path handling under ithreads.
2121 wait(), waitpid() and backticks now return the correct exit status under
2126 winsock handle leak fixed.
2132 =head1 New or Changed Diagnostics
2138 All regular expression compilation error messages are now hopefully
2139 easier to understand both because the error message now comes before
2140 the failed regex and because the point of failure is now clearly
2141 marked by a C<E<lt>-- HERE> marker.
2145 The various "opened only for", "on closed", "never opened" warnings
2146 drop the C<main::> prefix for filehandles in the C<main> package,
2147 for example C<STDIN> instead of C<main::STDIN>.
2151 The "Unrecognized escape" warning has been extended to include C<\8>,
2152 C<\9>, and C<\_>. There is no need to escape any of the C<\w> characters.
2156 Two new debugging options have been added: if you have compiled your
2157 Perl with debugging, you can use the -DT and -DR options to trace
2158 tokenising and to add reference counts to displaying variables,
2163 If an attempt to use a (non-blessed) reference as an array index
2164 is made, a warning is given.
2168 C<push @a;> and C<unshift @a;> (with no values to push or unshift)
2169 now give a warning. This may be a problem for generated and evaled
2174 If you try to L<perlfunc/pack> a number less than 0 or larger than 255
2175 using the C<"C"> format you will get an optional warning. Similarly
2176 for the C<"c"> format and a number less than -128 or more than 127.
2180 Certain regex modifiers such as C<(?o)> make sense only if applied to
2181 the entire regex. You will an optional warning if you try to do otherwise.
2185 Using arrays or hashes as references (e.g. C<< %foo->{bar} >>
2186 has been deprecated for a while. Now you will get an optional warning.
2190 =head1 Changed Internals
2196 perlapi.pod (a companion to perlguts) now attempts to document the
2201 You can now build a really minimal perl called microperl.
2202 Building microperl does not require even running Configure;
2203 C<make -f Makefile.micro> should be enough. Beware: microperl makes
2204 many assumptions, some of which may be too bold; the resulting
2205 executable may crash or otherwise misbehave in wondrous ways.
2206 For careful hackers only.
2210 Added rsignal(), whichsig(), do_join(), op_clear, op_null,
2211 ptr_table_clear(), ptr_table_free(), sv_setref_uv(), and several UTF-8
2212 interfaces to the publicised API. For the full list of the available
2213 APIs see L<perlapi>.
2217 Made possible to propagate customised exceptions via croak()ing.
2221 Now xsubs can have attributes just like subs.
2225 dTHR and djSP have been obsoleted; the former removed (because it's
2226 a no-op) and the latter replaced with dSP.
2230 PERL_OBJECT has been completely removed.
2234 The MAGIC constants (e.g. C<'P'>) have been macrofied
2235 (e.g. C<PERL_MAGIC_TIED>) for better source code readability
2236 and maintainability.
2240 The regex compiler now maintains a structure that identifies nodes in
2241 the compiled bytecode with the corresponding syntactic features of the
2242 original regex expression. The information is attached to the new
2243 C<offsets> member of the C<struct regexp>. See L<perldebguts> for more
2244 complete information.
2248 The C code has been made much more C<gcc -Wall> clean. Some warning
2249 messages still remain in some platforms, so if you are compiling with
2250 gcc you may see some warnings about dubious practices. The warnings
2251 are being worked on.
2255 F<perly.c>, F<sv.c>, and F<sv.h> have now been extensively commented.
2259 Documentation on how to use the Perl source repository has been added
2260 to F<Porting/repository.pod>.
2264 There are now several profiling make targets.
2268 =head1 Security Vulnerability Closed
2270 (This change was already made in 5.7.0 but bears repeating here.)
2272 A potential security vulnerability in the optional suidperl component
2273 of Perl was identified in August 2000. suidperl is neither built nor
2274 installed by default. As of November 2001 the only known vulnerable
2275 platform is Linux, most likely all Linux distributions. CERT and
2276 various vendors and distributors have been alerted about the vulnerability.
2277 See http://www.cpan.org/src/5.0/sperl-2000-08-05/sperl-2000-08-05.txt
2278 for more information.
2280 The problem was caused by Perl trying to report a suspected security
2281 exploit attempt using an external program, /bin/mail. On Linux
2282 platforms the /bin/mail program had an undocumented feature which
2283 when combined with suidperl gave access to a root shell, resulting in
2284 a serious compromise instead of reporting the exploit attempt. If you
2285 don't have /bin/mail, or if you have 'safe setuid scripts', or if
2286 suidperl is not installed, you are safe.
2288 The exploit attempt reporting feature has been completely removed from
2289 Perl 5.8.0 (and the maintenance release 5.6.1, and it was removed also
2290 from all the Perl 5.7 releases), so that particular vulnerability
2291 isn't there anymore. However, further security vulnerabilities are,
2292 unfortunately, always possible. The suidperl functionality is most
2293 probably going to be removed in Perl 5.10. In any case, suidperl
2294 should only be used by security experts who know exactly what they are
2295 doing and why they are using suidperl instead of some other solution
2296 such as sudo (see http://www.courtesan.com/sudo/).
2300 Several new tests have been added, especially for the F<lib>
2301 subsection. There are now about 34 000 individual tests (spread over
2302 about 530 test scripts), in the regression suite (5.6.1 has about
2303 11700 tests, in 258 test scripts) Many of the new tests are introduced
2304 by the new modules, but still in general Perl is now more thoroughly
2307 Because of the large number of tests, running the regression suite
2308 will take considerably longer time than it used to: expect the suite
2309 to take up to 4-5 times longer to run than in perl 5.6. In a really
2310 fast machine you can hope to finish the suite in about 5 minutes
2313 The tests are now reported in a different order than in earlier Perls.
2314 (This happens because the test scripts from under t/lib have been moved
2315 to be closer to the library/extension they are testing.)
2317 =head1 Known Problems
2325 In AIX 4.2 Perl extensions that use C++ functions that use statics
2326 may have problems in that the statics are not getting initialized.
2327 In newer AIX releases this has been solved by linking Perl with
2328 the libC_r library, but unfortunately in AIX 4.2 the said library
2329 has an obscure bug where the various functions related to time
2330 (such as time() and gettimeofday()) return broken values, and
2331 therefore in AIX 4.2 Perl is not linked against the libC_r.
2335 vac 5.0.0.0 May Produce Buggy Code For Perl
2337 The AIX C compiler vac version 5.0.0.0 may produce buggy code,
2338 resulting in few random tests failing, but when the failing tests
2339 are run by hand, they succeed. We suggest upgrading to at least
2340 vac version 5.0.1.0, that has been known to compile Perl correctly.
2341 "lslpp -L|grep vac.C" will tell you the vac version.
2345 =head2 Amiga Perl Invoking Mystery
2347 One cannot call Perl using the C<volume:> syntax, that is, C<perl -v>
2348 works, but for example C<bin:perl -v> doesn't. The exact reason isn't
2349 known but the current suspect is the F<ixemul> library.
2351 =head2 lib/ftmp-security tests warn 'system possibly insecure'
2353 Don't panic. Read INSTALL 'make test' section instead.
2355 =head2 Cygwin intermittent failures of lib/Memoize/t/expire_file 11 and 12
2357 The subtests 11 and 12 sometimes fail and sometimes work.
2359 =head2 HP-UX lib/io_multihomed Fails When LP64-Configured
2361 The lib/io_multihomed test may hang in HP-UX if Perl has been
2362 configured to be 64-bit. Because other 64-bit platforms do not hang in
2363 this test, HP-UX is suspect. All other tests pass in 64-bit HP-UX. The
2364 test attempts to create and connect to "multihomed" sockets (sockets
2365 which have multiple IP addresses).
2367 =head2 HP-UX lib/posix Subtest 9 Fails When LP64-Configured
2369 If perl is configured with -Duse64bitall, the successful result of the
2370 subtest 10 of lib/posix may arrive before the successful result of the
2371 subtest 9, which confuses the test harness so much that it thinks the
2374 =head2 Linux With Sfio Fails op/misc Test 48
2380 The following tests are known to fail:
2382 Failed Test Stat Wstat Total Fail Failed List of Failed
2383 -------------------------------------------------------------------------
2384 ../ext/DB_File/t/db-btree.t 0 11 ?? ?? % ??
2385 ../ext/DB_File/t/db-recno.t 149 3 2.01% 61 63 65
2386 ../ext/POSIX/t/posix.t 31 1 3.23% 10
2387 ../lib/warnings.t 450 1 0.22% 316
2391 OS/390 has rather many test failures but the situation is actually
2392 better than it was in 5.6.0, it's just that so many new modules and
2393 tests have been added.
2395 Failed Test Stat Wstat Total Fail Failed List of Failed
2396 -----------------------------------------------------------------------------
2397 ../ext/B/Deparse.t 14 1 7.14% 14
2398 ../ext/B/Showlex.t 1 1 100.00% 1
2399 ../ext/Encode/Encode/Tcl.t 610 13 2.13% 592 594 596 598
2401 ../ext/IO/lib/IO/t/io_unix.t 113 28928 5 3 60.00% 3-5
2402 ../ext/POSIX/POSIX.t 29 1 3.45% 14
2403 ../ext/Storable/t/lock.t 255 65280 5 3 60.00% 3-5
2404 ../lib/locale.t 129 33024 117 19 16.24% 99-117
2405 ../lib/warnings.t 434 1 0.23% 75
2406 ../lib/ExtUtils.t 27 1 3.70% 25
2407 ../lib/Math/BigInt/t/bigintpm.t 1190 1 0.08% 1145
2408 ../lib/Unicode/UCD.t 81 48 59.26% 1-16 49-64 66-81
2409 ../lib/User/pwent.t 9 1 11.11% 4
2410 op/pat.t 660 6 0.91% 242-243 424-425
2412 op/split.t 0 9 ?? ?? % ??
2413 op/taint.t 174 3 1.72% 156 162 168
2414 op/tr.t 70 3 4.29% 50 58-59
2415 Failed 16/422 test scripts, 96.21% okay. 105/23251 subtests failed, 99.55% okay.
2417 =head2 op/sprintf tests 129 and 130
2419 The op/sprintf tests 129 and 130 are known to fail on some platforms.
2420 Examples include any platform using sfio, and Compaq/Tandem's NonStop-UX.
2421 The failing platforms do not comply with the ANSI C Standard, line
2422 19ff on page 134 of ANSI X3.159 1989 to be exact. (They produce
2423 something other than "1" and "-1" when formatting 0.6 and -0.6 using
2424 the printf format "%.0f", most often they produce "0" and "-0".)
2426 =head2 Failure of Thread tests
2428 B<Note that support for 5.005-style threading remains experimental.>
2430 The following tests are known to fail due to fundamental problems in
2431 the 5.005 threading implementation. These are not new failures--Perl
2432 5.005_0x has the same bugs, but didn't have these tests.
2435 t/lib/thr5005.t 19-20
2443 ext/POSIX/sigaction subtests 6 and 13 may fail.
2447 lib/ExtUtils may spuriously claim that subtest 28 failed,
2448 which is interesting since the test only has 27 tests.
2452 Numerous numerical test failures
2454 op/numconvert 209,210,217,218
2456 ext/Time/HiRes/HiRes 9
2457 lib/Math/BigInt/t/bigintpm 1145
2460 These tests fail because of yet unresolved floating point inaccuracies.
2466 There are a few known test failures, see L<perluts>.
2470 There is one known test failure with a default configuration:
2472 [.run]switches..........................FAILED on test 1
2476 In multi-CPU boxes there are some problems with the I/O buffering:
2477 some output may appear twice.
2479 =head2 Localising a Tied Variable Leaks Memory
2482 tie my %tie_hash => 'Tie::StdHash';
2486 local($tie_hash{Foo}) = 1; # leaks
2488 Code like the above is known to leak memory every time the local()
2491 =head2 Localising Tied Arrays and Hashes Is Broken
2495 doesn't work as one would expect: the old value is restored
2498 =head2 Self-tying of Arrays and Hashes Is Forbidden
2500 Self-tying of arrays and hashes is broken in rather deep and
2501 hard-to-fix ways. As a stop-gap measure to avoid people from getting
2502 frustrated at the mysterious results (core dumps, most often) it is
2503 for now forbidden (you will get a fatal error even from an attempt).
2505 =head2 Variable Attributes are not Currently Usable for Tieing
2507 This limitation will hopefully be fixed in future. (Subroutine
2508 attributes work fine for tieing, see L<Attribute::Handlers>).
2510 One way to run into this limitation is to have a loop variable with
2511 attributes within a loop: the tie is called only once, not for each
2512 iteration of the loop.
2514 =head2 Building Extensions Can Fail Because Of Largefiles
2516 Some extensions like mod_perl are known to have issues with
2517 `largefiles', a change brought by Perl 5.6.0 in which file offsets
2518 default to 64 bits wide, where supported. Modules may fail to compile
2519 at all or compile and work incorrectly. Currently there is no good
2520 solution for the problem, but Configure now provides appropriate
2521 non-largefile ccflags, ldflags, libswanted, and libs in the %Config
2522 hash (e.g., $Config{ccflags_nolargefiles}) so the extensions that are
2523 having problems can try configuring themselves without the
2524 largefileness. This is admittedly not a clean solution, and the
2525 solution may not even work at all. One potential failure is whether
2526 one can (or, if one can, whether it's a good idea) link together at
2527 all binaries with different ideas about file offsets, all this is
2530 =head2 Unicode Support on EBCDIC Still Spotty
2532 Though mostly working, Unicode support still has problem spots on
2533 EBCDIC platforms. One such known spot are the C<\p{}> and C<\P{}>
2534 regular expression constructs for code points less than 256: the
2535 pP are testing for Unicode code points, not knowing about EBCDIC.
2537 =head2 The Compiler Suite Is Still Experimental
2539 The compiler suite is slowly getting better but it continues to be
2540 highly experimental. Use in production environments is discouraged.
2542 =head2 The Long Double Support is Still Experimental
2544 The ability to configure Perl's numbers to use "long doubles",
2545 floating point numbers of hopefully better accuracy, is still
2546 experimental. The implementations of long doubles are not yet
2547 widespread and the existing implementations are not quite mature
2548 or standardised, therefore trying to support them is a rare
2549 and moving target. The gain of more precision may also be offset
2550 by slowdown in computations (more bits to move around, and the
2551 operations are more likely to be executed by less optimised
2554 =head1 Reporting Bugs
2556 If you find what you think is a bug, you might check the articles
2557 recently posted to the comp.lang.perl.misc newsgroup and the perl
2558 bug database at http://bugs.perl.org. There may also be
2559 information at http://www.perl.com/perl/, the Perl Home Page.
2561 If you believe you have an unreported bug, please run the B<perlbug>
2562 program included with your release. Be sure to trim your bug down
2563 to a tiny but sufficient test case. Your bug report, along with the
2564 output of C<perl -V>, will be sent off to perlbug@perl.org to be
2565 analysed by the Perl porting team.
2569 The F<Changes> file for exhaustive details on what changed.
2571 The F<INSTALL> file for how to build Perl.
2573 The F<README> file for general stuff.
2575 The F<Artistic> and F<Copying> files for copyright information.
2579 Written by Jarkko Hietaniemi <F<jhi@iki.fi>>.