3 perldelta - what is new for perl v5.8.0
7 This document describes differences between the 5.6.0 release and the
16 Better Unicode support
20 New Thread Implementation
28 Better Numeric Accuracy
36 More Extensive Regression Testing
40 =head1 Incompatible Changes
42 =head2 64-bit platforms and malloc
44 If your pointers are 64 bits wide, the Perl malloc is no longer being
45 used because it does not work well with 8-byte pointers. Also,
46 usually the system mallocs on such platforms are much better optimized
47 for such large memory models than the Perl malloc. Some memory-hungry
48 Perl applications like the PDL don't work well with Perl's malloc.
49 Finally, other applications than Perl (like modperl) tend to prefer
50 the system malloc. Such platforms include Alpha and 64-bit HPPA,
53 =head2 AIX Dynaloading
55 The AIX dynaloading now uses in AIX releases 4.3 and newer the native
56 dlopen interface of AIX instead of the old emulated interface. This
57 change will probably break backward compatibility with compiled
58 modules. The change was made to make Perl more compliant with other
59 applications like modperl which are using the AIX native interface.
61 =head2 Socket Extension Dynamic in VMS
63 The Socket extension is now dynamically loaded instead of being
64 statically built in. This may or may not be a problem with ancient
65 TCP/IP stacks of VMS: we do not know since we weren't able to test
66 Perl in such configurations.
68 =head2 IEEE-format Floating Point Default on OpenVMS Alpha
70 Perl now uses IEEE format (T_FLOAT) as the default internal floating
71 point format on OpenVMS Alpha, potentially breaking binary compatibility
72 with external libraries or existing data. G_FLOAT is still available as
73 a configuration option. The default on VAX (D_FLOAT) has not changed.
75 =head2 Different Definition of the Unicode Character Classes \p{In...}
77 As suggested by the Unicode consortium, the Unicode character classes
78 now prefer I<scripts> as opposed to I<blocks> (as defined by Unicode);
79 in Perl, when the C<\p{In....}> and the C<\p{In....}> regular expression
80 constructs are used. This has changed the definition of some of those
83 The difference between scripts and blocks is that scripts are the
84 glyphs used by a language or a group of languages, while the blocks
85 are more artificial groupings of 256 characters based on the Unicode
88 In general this change results in more inclusive Unicode character
89 classes, but changes to the other direction also do take place:
90 for example while the script C<Latin> includes all the Latin
91 characters and their various diacritic-adorned versions, it
92 does not include the various punctuation or digits (since they
93 are not solely C<Latin>).
95 Changes in the character class semantics may have happened if a script
96 and a block happen to have the same name, for example C<Hebrew>.
97 In such cases the script wins and C<\p{InHebrew}> now means the script
98 definition of Hebrew. The block definition in still available,
99 though, by appending C<Block> to the name: C<\p{InHebrewBlock}> means
100 what C<\p{InHebrew}> meant in perl 5.6.0. For the full list
101 of affected character classes, see L<perlunicode/Blocks>.
103 =head2 Perl Parser Stress Tested
105 The Perl parser has been stress tested using both random input and
106 Markov chain input and the few found crashes and lockups have been
109 =head2 REF(...) Instead Of SCALAR(...)
111 A reference to a reference now stringifies as "REF(0x81485ec)" instead
112 of "SCALAR(0x81485ec)" in order to be more consistent with the return
121 The semantics of bless(REF, REF) were unclear and until someone proves
122 it to make some sense, it is forbidden.
126 The obsolete chat2 library that should never have been allowed
127 to escape the laboratory has been decommissioned.
131 The very dusty examples in the eg/ directory have been removed.
132 Suggestions for new shiny examples welcome but the main issue is that
133 the examples need to be documented, tested and (most importantly)
138 The (bogus) escape sequences \8 and \9 now give an optional warning
139 ("Unrecognized escape passed through"). There is no need to \-escape
144 The list of filenames from glob() (or <...>) is now by default sorted
145 alphabetically to be csh-compliant. (bsd_glob() does still sort platform
146 natively, ASCII or EBCDIC, unless GLOB_ALPHASORT is specified.)
150 Although "you shouldn't do that", it was possible to write code that
151 depends on Perl's hashed key order (Data::Dumper does this). The new
152 algorithm "One-at-a-Time" produces a different hashed key order.
153 More details are in L</"Performance Enhancements">.
157 lstat(FILEHANDLE) now gives a warning because the operation makes no sense.
158 In future releases this may become a fatal error.
162 The C<package;> syntax (C<package> without an argument) has been
163 deprecated. Its semantics were never that clear and its
164 implementation even less so. If you have used that feature to
165 disallow all but fully qualified variables, C<use strict;> instead.
169 The unimplemented POSIX regex features [[.cc.]] and [[=c=]] are still
170 recognised but now cause fatal errors. The previous behaviour of
171 ignoring them by default and warning if requested was unacceptable
172 since it, in a way, falsely promised that the features could be used.
176 The current user-visible implementation of pseudo-hashes (the weird
177 use of the first array element) is deprecated starting from Perl 5.8.0
178 and will be removed in Perl 5.10.0, and the feature will be
179 implemented differently. Not only is the current interface rather
180 ugly, but the current implementation slows down normal array and hash
181 use quite noticeably. The C<fields> pragma interface will remain
186 The syntaxes C<< @a->[...] >> and C<< %h->{...} >> have now been deprecated.
190 After years of trying the suidperl is considered to be too complex to
191 ever be considered truly secure. The suidperl functionality is likely
192 to be removed in a future release.
196 The long deprecated uppercase aliases for the string comparison
197 operators (EQ, NE, LT, LE, GE, GT) have now been removed.
201 The tr///C and tr///U features have been removed and will not return;
202 the interface was a mistake. Sorry about that. For similar
203 functionality, see pack('U0', ...) and pack('C0', ...).
207 =head1 Core Enhancements
209 =head2 PerlIO is Now The Default
215 IO is now by default done via PerlIO rather than system's "stdio".
216 PerlIO allows "layers" to be "pushed" onto a file handle to alter the
217 handle's behaviour. Layers can be specified at open time via 3-arg
220 open($fh,'>:crlf :utf8', $path) || ...
222 or on already opened handles via extended C<binmode>:
224 binmode($fh,':encoding(iso-8859-7)');
226 The built-in layers are: unix (low level read/write), stdio (as in
227 previous Perls), perlio (re-implementation of stdio buffering in a
228 portable manner), crlf (does CRLF <=> "\n" translation as on Win32,
229 but available on any platform). A mmap layer may be available if
230 platform supports it (mostly UNIXes).
232 Layers to be applied by default may be specified via the 'open' pragma.
234 See L</"Installation and Configuration Improvements"> for the effects
235 of PerlIO on your architecture name.
239 File handles can be marked as accepting Perl's internal encoding of Unicode
240 (UTF-8 or UTF-EBCDIC depending on platform) by a pseudo layer ":utf8" :
242 open($fh,">:utf8","Uni.txt");
244 Note for EBCDIC users: the pseudo layer ":utf8" is erroneously named
245 for you since it's not UTF-8 what you will be getting but instead
246 UTF-EBCDIC. See L<perlunicode>, L<utf8>, and
247 http://www.unicode.org/unicode/reports/tr16/ for more information.
248 In future releases this naming may change.
252 File handles can translate character encodings from/to Perl's internal
253 Unicode form on read/write via the ":encoding()" layer.
257 File handles can be opened to "in memory" files held in Perl scalars via:
259 open($fh,'>', \$variable) || ...
263 Anonymous temporary files are available without need to
264 'use FileHandle' or other module via
266 open($fh,"+>", undef) || ...
268 That is a literal undef, not an undefined value.
272 The list form of C<open> is now implemented for pipes (at least on UNIX):
274 open($fh,"-|", 'cat', '/etc/motd')
276 creates a pipe, and runs the equivalent of exec('cat', '/etc/motd') in
281 =head2 Signals Are Now Safe
283 Perl used to be fragile in that signals arriving at inopportune moments
284 could corrupt Perl's internal state. Now Perl postpones handling of
285 signals until it's safe.
287 =head2 Unicode Overhaul
289 Unicode in general should be now much more usable than in Perl 5.6.0
290 (or even in 5.6.1). Unicode can be used in hash keys, Unicode in
291 regular expressions should work now, Unicode in tr/// should work now,
292 Unicode in I/O should work now.
298 The Unicode Character Database coming with Perl has been upgraded
299 to Unicode 3.1.1. For more information, see http://www.unicode.org/.
303 For developers interested in enhancing Perl's Unicode capabilities:
304 almost all the UCD files are included with the Perl distribution in
305 the lib/unicore subdirectory. The most notable omission, for space
306 considerations, is the Unihan database.
310 The Unicode character classes \p{Blank} and \p{SpacePerl} have been
311 added. "Blank" is like C isblank(), that is, it contains only
312 "horizontal whitespace" (the space character is, the newline isn't),
313 and the "SpacePerl" is the Unicode equivalent of C<\s> (\p{Space}
314 isn't, since that includes the vertical tabulator character, whereas
319 =head2 Understanding of Numbers
321 In general a lot of fixing has happened in the area of Perl's
322 understanding of numbers, both integer and floating point. Since in
323 many systems the standard number parsing functions like C<strtoul()>
324 and C<atof()> seem to have bugs, Perl tries to work around their
325 deficiencies. This results hopefully in more accurate numbers.
327 Perl now tries internally to use integer values in numeric conversions
328 and basic arithmetics (+ - * /) if the arguments are integers, and
329 tries also to keep the results stored internally as integers.
330 This change leads to often slightly faster and always less lossy
331 arithmetics. (Previously Perl always preferred floating point numbers
334 =head2 Miscellaneous Enhancements
340 AUTOLOAD is now lvaluable, meaning that you can add the :lvalue attribute
341 to AUTOLOAD subroutines and you can assign to the AUTOLOAD return value.
345 C<perl -d:Module=arg,arg,arg> now works (previously one couldn't pass
346 in multiple arguments.)
350 END blocks are now run even if you exit/die in a BEGIN block.
351 Internally, the execution of END blocks is now controlled by
352 PL_exit_flags & PERL_EXIT_DESTRUCT_END. This enables the new
353 behaviour for Perl embedders. This will default in 5.10. See
358 Formats now support zero-padded decimal fields.
362 Lvalue subroutines can now return C<undef> in list context.
366 A new special regular expression variable has been introduced:
367 C<$^N>, which contains the most-recently closed group (submatch).
371 C<no Module;> now works even if there is no "sub unimport" in the Module.
375 The numerical comparison operators return C<undef> if either operand
376 is a NaN. Previously the behaviour was unspecified.
380 The following builtin functions are now overridable: each(), keys(),
381 pop(), push(), shift(), splice(), unshift().
385 C<pack('U0a*', ...)> can now be used to force a string to UTF8.
389 my __PACKAGE__ $obj now works.
393 The printf() and sprintf() now support parameter reordering using the
394 C<%\d+\$> and C<*\d+\$> syntaxes. For example
396 print "%2\$s %1\$s\n", "foo", "bar";
398 will print "bar foo\n". This feature helps in writing
399 internationalised software, and in general when the order
400 of the parameters can vary.
404 prototype(\&) is now available.
408 prototype(\[$@%&]) is now available to implicitly create references
409 (useful for example if you want to emulate the tie() interface).
413 UNTIE method is now recognised.
417 L<utime> now supports C<utime undef, undef, @files> to change the
418 file timestamps to the current time.
422 The rules for allowing underscores (underbars) in numeric constants
423 have been relaxed and simplified: now you can have an underscore
424 simply B<between digits>.
428 =head1 Modules and Pragmata
430 =head2 New Modules and Pragmata
436 C<Attribute::Handlers> allows a class to define attribute handlers.
439 use Attribute::Handlers;
440 sub Wolf :ATTR(SCALAR) { print "howl!\n" }
442 # later, in some package using or inheriting from MyPack...
444 my MyPack $Fluffy : Wolf; # the attribute handler Wolf will be called
446 Both variables and routines can have attribute handlers. Handlers can
447 be specific to type (SCALAR, ARRAY, HASH, or CODE), or specific to the
448 exact compilation phase (BEGIN, CHECK, INIT, or END).
452 B<B::Concise> is a new compiler backend for walking the Perl syntax
453 tree, printing concise info about ops, from Stephen McCamant. The
454 output is highly customisable. See L<B::Concise>.
458 C<Class::ISA> for reporting the search path for a class's ISA tree,
459 by Sean Burke, has been added. See L<Class::ISA>.
463 C<Cwd> has now a split personality: if possible, an XS extension is
464 used, (this will hopefully be faster, more secure, and more robust)
465 but if not possible, the familiar Perl implementation is used.
469 C<Devel::PPPort>, originally from Kenneth Albanowski and now
470 maintained by Paul Marquess, has been added. It is primarily used
471 by C<h2xs> to enhance portability of of XS modules between different
476 C<Digest>, frontend module for calculating digests (checksums), from
477 Gisle Aas, has been added. See L<Digest>.
481 C<Digest::MD5> for calculating MD5 digests (checksums) as defined in
482 RFC 1321, from Gisle Aas, has been added. See L<Digest::MD5>.
484 use Digest::MD5 'md5_hex';
486 $digest = md5_hex("Thirsty Camel");
488 print $digest, "\n"; # 01d19d9d2045e005c3f1b80e8b164de1
490 NOTE: the C<MD5> backward compatibility module is deliberately not
491 included since its further use is discouraged.
495 C<Encode>, by Nick Ing-Simmons, provides a mechanism to translate
496 between different character encodings. Support for Unicode,
497 ISO-8859-*, ASCII, CP*, KOI8-R, and three variants of EBCDIC are
498 compiled in to the module. Several other encodings (like Japanese,
499 Chinese, and MacIntosh encodings) are included and will be loaded at
500 runtime. See L<Encode>.
502 Any encoding supported by Encode module is also available to the
503 ":encoding()" layer if PerlIO is used.
507 C<I18N::Langinfo> can be use to query locale information.
508 See L<I18N::Langinfo>.
512 C<I18N::LangTags> has functions for dealing with RFC3066-style
513 language tags, by Sean Burke. See L<I18N::LangTags>.
517 C<ExtUtils::Constant> is a new tool for extension writers for
518 generating XS code to import C header constants, by Nicholas Clark.
519 See L<ExtUtils::Constant>.
523 C<Filter::Simple> is an easy-to-use frontend to Filter::Util::Call,
524 from Damian Conway. See L<Filter::Simple>.
530 use Filter::Simple sub {
531 while (my ($from, $to) = splice @_, 0, 2) {
540 use MyFilter qr/red/ => 'green';
542 print "red\n"; # this code is filtered, will print "green\n"
543 print "bored\n"; # this code is filtered, will print "bogreen\n"
547 print "red\n"; # this code is not filtered, will print "red\n"
551 C<File::Temp> allows one to create temporary files and directories in
552 an easy, portable, and secure way, by Tim Jenness. See L<File::Temp>.
556 C<Filter::Util::Call> provides you with the framework to write
557 I<Source Filters> in Perl, from Paul Marquess. For most uses the
558 frontend Filter::Simple is to be preferred. See L<Filter::Util::Call>.
562 L<libnet> is a collection of perl5 modules related to network
563 programming, from Graham Barr. See L<Net::FTP>, L<Net::NNTP>,
564 L<Net::Ping>, L<Net::POP3>, L<Net::SMTP>, and L<Net::Time>.
566 Perl installation leaves libnet unconfigured, use F<libnetcfg> to configure.
570 C<List::Util> is a selection of general-utility list subroutines, like
571 sum(), min(), first(), and shuffle(), by Graham Barr. See L<List::Util>.
575 C<Locale::Constants>, C<Locale::Country>, C<Locale::Currency>, and
576 C<Locale::Language>, from Neil Bowers, have been added. They provide the
577 codes for various locale standards, such as "fr" for France, "usd" for
578 US Dollar, and "jp" for Japanese.
582 $country = code2country('jp'); # $country gets 'Japan'
583 $code = country2code('Norway'); # $code gets 'no'
585 See L<Locale::Constants>, L<Locale::Country>, L<Locale::Currency>,
586 and L<Locale::Language>.
590 C<Locale::Maketext> is localization framework from Sean Burke. See
591 L<Locale::Maketext>, and L<Locale::Maketext::TPJ13>. The latter is an
592 article about software localization, originally published in The Perl
593 Journal #13, republished here with kind permission.
597 C<Memoize> can make your functions faster by trading space for time,
598 from Mark-Jason Dominus. See L<Memoize>.
602 C<MIME::Base64> allows you to encode data in base64, from Gisle Aas,
603 as defined in RFC 2045 - I<MIME (Multipurpose Internet Mail
608 $encoded = encode_base64('Aladdin:open sesame');
609 $decoded = decode_base64($encoded);
611 print $encoded, "\n"; # "QWxhZGRpbjpvcGVuIHNlc2FtZQ=="
617 C<MIME::QuotedPrint> allows you to encode data in quoted-printable
618 encoding, as defined in RFC 2045 - I<MIME (Multipurpose Internet Mail
619 Extensions)>, from Gisle Aas.
621 use MIME::QuotedPrint;
623 $encoded = encode_qp("Smiley in Unicode: \x{263a}");
624 $decoded = decode_qp($encoded);
626 print $encoded, "\n"; # "Smiley in Unicode: =263A"
628 MIME::QuotedPrint has been enhanced to provide the basic methods
629 necessary to use it with PerlIO::Via as in :
631 use MIME::QuotedPrint;
632 open($fh,">Via(MIME::QuotedPrint)",$path);
634 See L<MIME::QuotedPrint>.
638 C<NEXT> is pseudo-class for method redispatch, from Damian Conway.
643 C<open> is a new pragma for setting the default I/O disciplines
648 C<PerlIO::Scalar> provides the implementation of IO to "in memory"
649 Perl scalars as discussed above, from Nick Ing-Simmons. It also
650 serves as an example of a loadable PerlIO layer. Other future
651 possibilities include PerlIO::Array and PerlIO::Code.
652 See L<PerlIO::Scalar>.
656 C<PerlIO::Via> acts as a PerlIO layer and wraps PerlIO layer
657 functionality provided by a class (typically implemented in perl
658 code), from Nick Ing-Simmons.
660 use MIME::QuotedPrint;
661 open($fh,">Via(MIME::QuotedPrint)",$path);
663 This will automatically convert everything output to C<$fh>
664 to Quoted-Printable. See L<PerlIO::Via>.
668 C<Pod::ParseLink>, by Russ Allbery, has been added,
669 to parse L<> links in pods as described in the new
674 C<Pod::Text::Overstrike>, by Joe Smith, has been added.
675 It converts POD data to formatted overstrike text.
676 See L<Pod::Text::Overstrike>.
680 C<Scalar::Util> is a selection of general-utility scalar subroutines,
681 like blessed(), reftype(), and tainted(). See L<Scalar::Util>.
685 C<sort> is a new pragma for controlling the behaviour of sort().
689 C<Storable> gives persistence to Perl data structures by allowing the
690 storage and retrieval of Perl data to and from files in a fast and
691 compact binary format, from Raphael Manfredi. See L<Storable>.
695 C<Switch>, from Damian Conway, has been added. Just by saying
699 you have C<switch> and C<case> available in Perl.
705 case 1 { print "number 1" }
706 case "a" { print "string a" }
707 case [1..10,42] { print "number in list" }
708 case (@array) { print "number in list" }
709 case /\w+/ { print "pattern" }
710 case qr/\w+/ { print "pattern" }
711 case (%hash) { print "entry in hash" }
712 case (\%hash) { print "entry in hash" }
713 case (\&sub) { print "arg to subroutine" }
714 else { print "previous case not true" }
721 C<Test::More> is yet another framework for writing test scripts,
722 more extensive than Test::Simple, by Michael Schwern. See L<Test::More>.
726 C<Test::Simple> has basic utilities for writing tests, by Michael
727 Schwern. See L<Test::Simple>.
731 C<Text::Balanced> has been added, for extracting delimited text
732 sequences from strings, from Damian Conway.
734 use Text::Balanced 'extract_delimited';
736 ($a, $b) = extract_delimited("'never say never', he never said", "'", '');
738 $a will be "'never say never'", $b will be ', he never said'.
740 In addition to extract_delimited() there are also extract_bracketed(),
741 extract_quotelike(), extract_codeblock(), extract_variable(),
742 extract_tagged(), extract_multiple(), gen_delimited_pat(), and
743 gen_extract_tagged(). With these you can implement rather advanced
744 parsing algorithms. See L<Text::Balanced>.
748 C<threads> is an interface to interpreter threads, by Arthur Bergman.
749 Interpreter threads (ithreads) is the new thread model introduced in
750 Perl 5.6 but only available as an internal interface for extension
751 writers (and for Win32 Perl for C<fork()> emulation). See L<threads>.
755 C<threads::shared> allows data sharing for interpreter threads, from
756 Arthur Bergman. In the ithreads model any data sharing between
757 threads must be explicit, as opposed to the old 5.005 thread model
758 where data sharing was implicit. See L<threads::shared>.
762 C<Tie::RefHash::Nestable>, by Edward Avis, allows storing hash
763 references (unlike the standard Tie::RefHash) The module is contained
764 within Tie::RefHash, see L<Tie::RefHash>.
768 C<Time::HiRes> provides high resolution timing (ualarm, usleep,
769 and gettimeofday), from Douglas E. Wegscheid. See L<Time::HiRes>.
773 C<Unicode::UCD> offers a querying interface to the Unicode Character
774 Database. See L<Unicode::UCD>.
778 C<Unicode::Collate> implements the UCA (Unicode Collation Algorithm)
779 for sorting Unicode strings, by SADAHIRO Tomoyuki. See L<Unicode::Collate>.
783 C<Unicode::Normalize> implements the various Unicode normalization
784 forms, by SADAHIRO Tomoyuki. See L<Unicode::Normalize>.
788 C<XS::Typemap>, by Tim Jenness, is a test extension that exercises XS
789 typemaps. Nothing gets installed but for extension writers the code
794 =head2 Updated And Improved Modules and Pragmata
800 The following independently supported modules have been updated to the
801 newest versions from CPAN: CGI, CPAN, DB_File, File::Spec, File::Temp,
802 Getopt::Long, Math::BigFloat, Math::BigInt, the podlators bundle
803 (Pod::Man, Pod::Text), Pod::LaTeX, Pod::Parser, Storable,
804 Term::ANSIColor, Test, Text-Tabs+Wrap.
808 The attributes::reftype() now works on tied arguments.
812 AutoLoader can now be disabled with C<no AutoLoader;>.
816 B::Deparse has been significantly enhanced. It now can deparse almost
817 all of the standard test suite (so that the tests still succeed).
818 There is a make target "test.deparse" for trying this out.
822 Class::Struct can now define the classes in compile time.
826 Class::Struct now assigns the array/hash element if the accessor
827 is called with an array/hash element as the B<sole> argument.
831 Data::Dumper has now an option to sort hashes.
835 Data::Dumper has now an option to dump code references
840 The English module can now be used without the infamous performance
843 use English '-no_performance_hit';
845 (Assuming, of course, that one doesn't need the troublesome variables
846 C<$`>, C<$&>, or C<$'>.) Also, introduced C<@LAST_MATCH_START> and
847 C<@LAST_MATCH_END> English aliases for C<@-> and C<@+>.
851 Fcntl, Socket, and Sys::Syslog have been rewritten to use the
852 new-style constant dispatch section (see L<ExtUtils::Constant>).
853 This means that they will be more robust and hopefully faster.
857 File::Find now has pre- and post-processing callbacks. It also
858 correctly changes directories when chasing symbolic links. Callbacks
859 (naughtily) exiting with "next;" instead of "return;" now work.
863 File::Find is now (again) reentrant. It also has been made
868 File::Glob::glob() renamed to File::Glob::bsd_glob() to avoid
869 prototype mismatch with CORE::glob().
873 File::Glob now supports C<GLOB_LIMIT> constant to limit the size of
874 the returned list of filenames.
878 Devel::Peek now has an interface for the Perl memory statistics
879 (this works only if you are using perl's malloc, and if you have
880 compiled with debugging).
884 IPC::Open3 now allows the use of numeric file descriptors.
888 IO::Socket has now atmark() method, which returns true if the socket
889 is positioned at the out-of-band mark. The method is also exportable
890 as a sockatmark() function.
894 IO::Socket::INET has support for ReusePort option (if your platform
895 supports it). The Reuse option now has an alias, ReuseAddr. For clarity
896 you may want to prefer ReuseAddr.
900 IO::Socket::INET now supports C<LocalPort> of zero (usually meaning
901 that the operating system will make one up.)
905 use lib now works identically to @INC. Removing directories
906 with 'no lib' now works.
910 Math::BigFloat and Math::BigInt have undergone a full rewrite.
911 They are now magnitudes faster, and they support various
912 bignum libraries such as GMP and PARI as their backends.
916 Net::Ping has been enhanced. There is now "external" protocol which
917 uses Net::Ping::External module which runs external ping(1) and parses
918 the output. An alpha version of Net::Ping::External is available in
919 CPAN and in 5.7.2 the Net::Ping::External may be integrated to Perl.
923 POSIX::sigaction() is now much more flexible and robust.
924 You can now install coderef handlers, 'DEFAULT', and 'IGNORE'
925 handlers, installing new handlers was not atomic.
929 In Safe the C<%INC> now localised in a Safe compartment so that
934 In Search::Dict one can now have a pre-processing hook for the
935 lines being searched.
939 The Shell module now has an OO interface.
943 The Test module has been significantly enhanced.
947 The vars pragma now supports declaring fully qualified variables.
948 (Something that C<our()> does not and will not support.)
952 The utf8:: name space (as in the pragma) provides various
953 Perl-callable functions to provide low level access to Perl's
954 internal Unicode representation. At the moment only length()
955 has been implemented.
959 =head1 Utility Changes
965 Emacs perl mode (emacs/cperl-mode.el) has been updated to version
970 F<emacs/e2ctags.pl> is now much faster.
974 C<h2ph> now supports C trigraphs.
978 C<h2xs> now produces a template README.
982 C<h2xs> now uses C<Devel::PPort> for better portability between
983 different versions of Perl.
987 C<h2xs> uses the new L<ExtUtils::Constant> module which will affect
988 newly created extensions that define constants. Since the new code is
989 more correct (if you have two constants where the first one is a
990 prefix of the second one, the first constant B<never> gets defined),
991 less lossy (it uses integers for integer constant, as opposed to the
992 old code that used floating point numbers even for integer constants),
993 and slightly faster, you might want to consider regenerating your
994 extension code (the new scheme makes regenerating easy).
995 L<h2xs> now also supports C trigraphs.
999 C<libnetcfg> has been added to configure the libnet.
1003 C<perlbug> is now much more robust. It also sends the bug report to
1004 perl.org, not perl.com.
1008 C<perlcc> has been rewritten and its user interface (that is,
1009 command line) is much more like that of the UNIX C compiler, cc.
1013 C<perlivp> is a new Installation Verification Procedure utility
1014 for running any time after installing Perl.
1018 C<pod2html> now allows specifying a cache directory.
1022 C<s2p> has been completely rewritten in Perl. (It is in fact a full
1023 implementation of sed in Perl: you can use the sed functionality by
1024 using the C<psed> utility.)
1028 C<xsubpp> now understands POD documentation embedded in the *.xs files.
1032 C<xsubpp> now supports OUT keyword.
1036 =head1 New Documentation
1042 perl56delta details the changes between the 5.005 release and the
1047 perlclib documents the internal replacements for standard C library
1048 functions. (Interesting only for extension writers and Perl core
1053 perldebtut is a Perl debugging tutorial.
1057 perlebcdic contains considerations for running Perl on EBCDIC platforms.
1061 perlintro is a gentle introduction to Perl.
1065 perliol documents the internals of PerlIO with layers.
1069 perlmodstyle is a style guide for writing modules.
1073 perlnewmod tells about writing and submitting a new module.
1077 perlpacktut is a pack() tutorial.
1081 perlpod has been rewritten to be clearer and to record the best
1082 practices gathered over the years.
1086 perlpodspec is a more formal specification of the pod format,
1087 mainly of interest for writers of pod applications, not to
1088 people writing in pod.
1092 perlretut is a regular expression tutorial.
1096 perlrequick is a regular expressions quick-start guide.
1097 Yes, much quicker than perlretut.
1101 perltodo has been updated.
1105 perltootc has been renamed as perltooc (to not to conflict
1106 with perltoot in filesystems restricted to "8.3" names)
1110 perluniintro is an introduction to using Unicode in Perl
1111 (perlunicode is more of a reference)
1115 perlutil explains the command line utilities packaged with the Perl
1120 The following platform-specific documents are available before
1121 the installation as README.I<platform>, and after the installation
1124 perlaix perlamiga perlapollo perlbeos perlbs2000
1125 perlce perlcygwin perldgux perldos perlepoc perlhpux
1126 perlhurd perlmachten perlmacos perlmint perlmpeix
1127 perlnetware perlos2 perlos390 perlplan9 perlqnx perlsolaris
1128 perltru64 perluts perlvmesa perlvms perlvos perlwin32
1134 The documentation for the POSIX-BC platform is called "BS2000", to avoid
1135 confusion with the Perl POSIX module.
1139 The documentation for the WinCE platform is called "CE", to avoid
1140 confusion with the perlwin32 documentation on 8.3-restricted filesystems.
1144 =head1 Performance Enhancements
1150 map() that changes the size of the list should now work faster.
1154 sort() has been changed to use primarily mergesort internally as
1155 opposed to the earlier quicksort. For very small lists this may
1156 result in slightly slower sorting times, but in general the speedup
1157 should be at least 20%. Additional bonuses are that the worst case
1158 behaviour of sort() is now better (in computer science terms it now
1159 runs in time O(N log N), as opposed to quicksort's Theta(N**2)
1160 worst-case run time behaviour), and that sort() is now stable
1161 (meaning that elements with identical keys will stay ordered as they
1162 were before the sort). See the C<sort> pragma for information.
1164 The story in more detail: suppose you want to serve yourself a little
1167 @digits = ( 3,1,4,1,5,9 );
1169 A numerical sort of the digits will yield (1,1,3,4,5,9), as expected.
1170 Which C<1> comes first is hard to know, since one C<1> looks pretty
1171 much like any other. You can regard this as totally trivial,
1172 or somewhat profound. However, if you just want to sort the even
1173 digits ahead of the odd ones, then what will
1175 sort { ($a % 2) <=> ($b % 2) } @digits;
1177 yield? The only even digit, C<4>, will come first. But how about
1178 the odd numbers, which all compare equal? With the quicksort algorithm
1179 used to implement Perl 5.6 and earlier, the order of ties is left up
1180 to the sort. So, as you add more and more digits of Pi, the order
1181 in which the sorted even and odd digits appear will change.
1182 and, for sufficiently large slices of Pi, the quicksort algorithm
1183 in Perl 5.8 won't return the same results even if reinvoked with the
1184 same input. The justification for this rests with quicksort's
1185 worst case behavior. If you run
1187 sort { $a <=> $b } ( 1 .. $N , 1 .. $N );
1189 (something you might approximate if you wanted to merge two sorted
1190 arrays using sort), doubling $N doesn't just double the quicksort time,
1191 it I<quadruples> it. Quicksort has a worst case run time that can
1192 grow like N**2, so-called I<quadratic> behaviour, and it can happen
1193 on patterns that may well arise in normal use. You won't notice this
1194 for small arrays, but you I<will> notice it with larger arrays,
1195 and you may not live long enough for the sort to complete on arrays
1196 of a million elements. So the 5.8 quicksort scrambles large arrays
1197 before sorting them, as a statistical defence against quadratic behaviour.
1198 But that means if you sort the same large array twice, ties may be
1199 broken in different ways.
1201 Because of the unpredictability of tie-breaking order, and the quadratic
1202 worst-case behaviour, quicksort was I<almost> replaced completely with
1203 a stable mergesort. I<Stable> means that ties are broken to preserve
1204 the original order of appearance in the input array. So
1206 sort { ($a % 2) <=> ($b % 2) } (3,1,4,1,5,9);
1208 will yield (4,3,1,1,5,9), guaranteed. The even and odd numbers
1209 appear in the output in the same order they appeared in the input.
1210 Mergesort has worst case O(NlogN) behaviour, the best value
1211 attainable. And, ironically, this mergesort does particularly
1212 well where quicksort goes quadratic: mergesort sorts (1..$N, 1..$N)
1213 in O(N) time. But quicksort was rescued at the last moment because
1214 it is faster than mergesort on certain inputs and platforms.
1215 For example, if you really I<don't> care about the order of even
1216 and odd digits, quicksort will run in O(N) time; it's very good
1217 at sorting many repetitions of a small number of distinct elements.
1218 The quicksort divide and conquer strategy works well on platforms
1219 with relatively small, very fast, caches. Eventually, the problem gets
1220 whittled down to one that fits in the cache, from which point it
1221 benefits from the increased memory speed.
1223 Quicksort was rescued by implementing a sort pragma to control aspects
1224 of the sort. The B<stable> subpragma forces stable behaviour,
1225 regardless of algorithm. The B<_quicksort> and B<_mergesort>
1226 subpragmas are heavy-handed ways to select the underlying implementation.
1227 The leading C<_> is a reminder that these subpragmas may not survive
1228 beyond 5.8. More appropriate mechanisms for selecting the implementation
1229 exist, but they wouldn't have arrived in time to save quicksort.
1233 Hashes now use Bob Jenkins "One-at-a-Time" hashing key algorithm
1234 (http://burtleburtle.net/bob/hash/doobs.html). This algorithm is
1235 reasonably fast while producing a much better spread of values than
1236 the old hashing algorithm (originally by Chris Torek, later tweaked by
1237 Ilya Zakharevich). Hash values output from the algorithm on a hash of
1238 all 3-char printable ASCII keys comes much closer to passing the
1239 DIEHARD random number generation tests. According to perlbench, this
1240 change has not affected the overall speed of Perl.
1244 unshift() should now be noticeably faster.
1248 =head1 Installation and Configuration Improvements
1250 =head2 Generic Improvements
1256 INSTALL now explains how you can configure Perl to use 64-bit
1257 integers even on non-64-bit platforms.
1261 Policy.sh policy change: if you are reusing a Policy.sh file
1262 (see INSTALL) and you use Configure -Dprefix=/foo/bar and in the old
1263 Policy $prefix eq $siteprefix and $prefix eq $vendorprefix, all of
1264 them will now be changed to the new prefix, /foo/bar. (Previously
1265 only $prefix changed.) If you do not like this new behaviour,
1266 specify prefix, siteprefix, and vendorprefix explicitly.
1270 A new optional location for Perl libraries, otherlibdirs, is available.
1271 It can be used for example for vendor add-ons without disturbing Perl's
1272 own library directories.
1276 In many platforms the vendor-supplied 'cc' is too stripped-down to
1277 build Perl (basically, 'cc' doesn't do ANSI C). If this seems
1278 to be the case and 'cc' does not seem to be the GNU C compiler
1279 'gcc', an automatic attempt is made to find and use 'gcc' instead.
1283 gcc needs to closely track the operating system release to avoid
1284 build problems. If Configure finds that gcc was built for a different
1285 operating system release than is running, it now gives a clearly visible
1286 warning that there may be trouble ahead.
1290 If binary compatibility with the 5.005 release is not wanted, Configure
1291 no longer suggests including the 5.005 modules in @INC.
1295 Configure C<-S> can now run non-interactively.
1299 configure.gnu now works with options with whitespace in them.
1303 installperl now outputs everything to STDERR.
1307 $Config{byteorder} is now computed dynamically (this is more robust
1308 with "fat binaries" where an executable image contains binaries for
1309 more than one binary platform.)
1313 Because PerlIO is now the default on most platforms, "-perlio" doesn't
1314 get appended to the $Config{archname} (also known as $^O) anymore.
1315 Instead, if you explicitly choose not to use perlio (Configure command
1316 line option -Uuseperlio), you will get "-stdio" appended.
1320 Another change related to the architecture name is that "-64all"
1321 (-Duse64bitall, or "maximally 64-bit") is appended only if your
1322 pointers are 64 bits wide. (To be exact, the use64bitall is ignored.)
1326 In AFS installations one can configure the root of the AFS to be
1327 somewhere else than the default F</afs> by using the Configure
1328 parameter C<-Dafsroot=/some/where/else>.
1332 APPLLIB_EXP, a less-know configuration-time definition, has been
1333 documented. It can be used to prepend site-specific directories
1334 to Perl's default search path (@INC), see INSTALL for information.
1338 The version of Berkeley DB used when the Perl (and, presumably, the
1339 DB_File extension) was built is now available as
1340 C<@Config{qw(db_version_major db_version_minor db_version_patch)}>
1341 from Perl and as C<DB_VERSION_MAJOR_CFG DB_VERSION_MINOR_CFG
1342 DB_VERSION_PATCH_CFG> from C.
1346 Building Berkeley DB3 for compatibility modes for DB, NDBM, and ODBM
1347 has been documented in INSTALL.
1351 If you have CPAN access (either network or a local copy such as a
1352 CD-ROM) you can during specify extra modules to Configure to build and
1353 install with Perl using the -Dextras=... option. See INSTALL for
1358 In addition to config.over a new override file, config.arch, is
1359 available. That is supposed to be used by hints file writers for
1360 architecture-wide changes (as opposed to config.over which is for
1365 If your file system supports symbolic links you can build Perl outside
1366 of the source directory by
1368 mkdir /tmp/perl/build/directory
1369 cd /tmp/perl/build/directory
1370 sh /path/to/perl/source/Configure -Dmksymlinks ...
1372 This will create in /tmp/perl/build/directory a tree of symbolic links
1373 pointing to files in /path/to/perl/source. The original files are left
1374 unaffected. After Configure has finished you can just say
1378 and Perl will be built and tested, all in /tmp/perl/build/directory.
1382 For Perl developers several new make targets for profiling
1383 and debugging have been added, see L<perlhack>.
1389 Use of the F<gprof> tool to profile Perl has been documented in
1390 L<perlhack>. There is a make target called "perl.gprof" for
1391 generating a gprofiled Perl executable.
1395 If you have GCC 3, there is a make target called "perl.gcov" for
1396 creating a gcoved Perl executable for coverage analysis. See
1401 If you are on IRIX or Tru64 platforms, new profiling/debugging options
1402 have been added, see L<perlhack> for more information about pixie and
1409 Guidelines of how to construct minimal Perl installations have
1410 been added to INSTALL.
1414 The Thread extension is now not built at all under ithreads
1415 (C<Configure -Duseithreads>) because it wouldn't work anyway (the
1416 Thread extension requires being Configured with C<-Duse5005threads>).
1418 But note that the Thread.pm interface is now shared by both
1423 =head2 New Or Improved Platforms
1425 For the list of platforms known to support Perl,
1426 see L<perlport/"Supported Platforms">.
1432 AIX dynamic loading should be now better supported.
1436 AIX should now work better with gcc, threads, and 64-bitness. Also the
1437 long doubles support in AIX should be better now. See L<perlaix>.
1441 After a long pause, AmigaOS has been verified to be happy with Perl.
1445 AtheOS (http://www.atheos.cx/) is a new platform.
1449 DG/UX platform now supports the 5.005-style threads. See L<perldgux>.
1453 DYNIX/ptx platform (a.k.a. dynixptx) is supported at or near osvers 4.5.2.
1457 EBCDIC platforms (z/OS, also known as OS/390, POSIX-BC, and VM/ESA)
1458 have been regained. Many test suite tests still fail and the
1459 co-existence of Unicode and EBCDIC isn't quite settled, but the
1460 situation is much better than with Perl 5.6. See L<perlos390>,
1461 L<perlbs2000> (for POSIX-BC), and L<perlvmesa> for more information.
1465 Building perl with -Duseithreads or -Duse5005threads now works under
1466 HP-UX 10.20 (previously it only worked under 10.30 or later). You will
1467 need a thread library package installed. See README.hpux.
1471 MacOS Classic (MacPerl has of course been available since
1472 perl 5.004 but now the source code bases of standard Perl
1473 and MacPerl have been synchronised)
1477 MacOS X (or Darwin) should now be able to build Perl even on HFS+
1478 filesystems. (The case-insensitivity confused the Perl build process.)
1482 NCR MP-RAS is now supported.
1486 NetWare from Novell is now supported. See L<perlnetware>.
1490 NonStop-UX is now supported.
1494 Amdahl UTS UNIX mainframe platform is now supported.
1498 WinCE is now supported. See L<perlce>.
1502 z/OS (formerly known as OS/390, formerly known as MVS OE) has now
1503 support for dynamic loading. This is not selected by default,
1504 however, you must specify -Dusedl in the arguments of Configure.
1508 =head1 Selected Bug Fixes
1510 Numerous memory leaks and uninitialized memory accesses have been
1511 hunted down. Most importantly anonymous subs used to leak quite
1518 The autouse pragma didn't work for Multi::Part::Function::Names.
1522 chop(@list) in list context returned the characters chopped in
1523 reverse order. This has been reversed to be in the right order.
1527 Configure no longer includes the DBM libraries (dbm, gdbm, db, ndbm)
1528 when building the Perl binary. The only exception to this is SunOS 4.x,
1533 The behaviour of non-decimal but numeric string constants such as
1534 "0x23" was platform-dependent: in some platforms that was seen as 35,
1535 in some as 0, in some as a floating point number (don't ask). This
1536 was caused by Perl using the operating system libraries in a situation
1537 where the result of the string to number conversion is undefined: now
1538 Perl consistently handles such strings as zero in numeric contexts.
1542 The order of DESTROYs has been made more predictable.
1546 Several debugger fixes: exit code now reflects the script exit code,
1547 condition C<"0"> now treated correctly, the C<d> command now checks
1548 line number, the C<$.> no longer gets corrupted, all debugger output now
1549 goes correctly to the socket if RemotePort is set.
1553 L<dprofpp> -R didn't work.
1557 C<*foo{FORMAT}> now works.
1561 UNIVERSAL::isa no longer caches methods incorrectly. (This broke
1562 the Tk extension with 5.6.0.)
1566 Lexicals I: lexicals outside an eval "" weren't resolved
1567 correctly inside a subroutine definition inside the eval "" if they
1568 were not already referenced in the top level of the eval""ed code.
1572 Lexicals II: lexicals leaked at file scope into subroutines that
1573 were declared before the lexicals.
1577 Lexical warnings now propagating correctly between scopes.
1581 Line renumbering with eval and C<#line> now works.
1585 Fixed numerous memory leaks, especially in eval "".
1589 mkdir() now ignores trailing slashes in the directory name,
1590 as mandated by POSIX.
1594 Some versions of glibc have a broken modfl(). This affects builds
1595 with C<-Duselongdouble>. This version of Perl detects this brokenness
1596 and has a workaround for it. The glibc release 2.2.2 is known to have
1597 fixed the modfl() bug.
1601 Modulus of unsigned numbers now works (4063328477 % 65535 used to
1602 return 27406, instead of 27047).
1606 Some "not a number" warnings introduced in 5.6.0 eliminated to be
1607 more compatible with 5.005. Infinity is now recognised as a number.
1611 Attributes (like :shared) didn't work with our().
1615 our() variables will not cause "will not stay shared" warnings.
1619 pack "Z" now correctly terminates the string with "\0".
1623 Fix password routines which in some shadow password platforms
1624 (e.g. HP-UX) caused getpwent() to return every other entry.
1628 The PERL5OPT environment variable (for passing command line arguments
1629 to Perl) didn't work for more than a single group of options.
1633 PERL5OPT with embedded spaces didn't work.
1637 printf() no longer resets the numeric locale to "C".
1641 C<q(a\\b)> now parses correctly as C<'a\\b'>.
1645 Printing quads (64-bit integers) with printf/sprintf now works
1646 without the q L ll prefixes (assuming you are on a quad-capable platform).
1650 Regular expressions on references and overloaded scalars now work.
1654 Right-hand side magic (GMAGIC) could in many cases such as string
1655 concatenation be invoked too many times.
1659 scalar() now forces scalar context even when used in void context.
1663 SOCKS support is now much more robust.
1667 sort() arguments are now compiled in the right wantarray context
1668 (they were accidentally using the context of the sort() itself).
1672 Changed the POSIX character class C<[[:space:]]> to include the (very
1673 rarely used) vertical tab character. Added a new POSIX-ish character
1674 class C<[[:blank:]]> which stands for horizontal whitespace
1675 (currently, the space and the tab).
1679 The tainting behaviour of sprintf() has been rationalized. It does
1680 not taint the result of floating point formats anymore, making the
1681 behaviour consistent with that of string interpolation.
1685 The regular expression captured submatches ($1, $2, ...) are now
1686 more consistently unset if the match fails, instead of leaving false
1687 data lying around in them.
1691 Sys::Syslog ignored the C<LOG_AUTH> constant.
1695 All but the first argument of the IO syswrite() method are now optional.
1699 $AUTOLOAD, sort(), lock(), and spawning subprocesses
1700 in multiple threads simultaneously are now thread-safe.
1704 Tie::ARRAY SPLICE method was broken.
1708 Allow read-only string on left hand side of non-modifying tr///.
1712 Several Unicode fixes.
1718 BOMs (byte order marks) in the beginning of Perl files
1719 (scripts, modules) should now be transparently skipped.
1720 UTF-16 (UCS-2) encoded Perl files should now be read correctly.
1724 The character tables have been updated to Unicode 3.1.1.
1728 Comparing with utf8 data does not magically upgrade non-utf8 data
1733 C<IsAlnum>, C<IsAlpha>, and C<IsWord> now match titlecase.
1737 Concatenation with the C<.> operator or via variable interpolation,
1738 C<eq>, C<substr>, C<reverse>, C<quotemeta>, the C<x> operator,
1739 substitution with C<s///>, single-quoted UTF8, should now work.
1743 The C<tr///> operator now works. Note that the C<tr///CU>
1744 functionality has been removed (but see pack('U0', ...)).
1748 C<eval "v200"> now works.
1752 Zero entries were missing from the Unicode classes like C<IsDigit>.
1758 =head2 Platform Specific Changes and Fixes
1766 Perl now works on post-4.0 BSD/OSes.
1772 Setting C<$0> now works (as much as possible; see L<perlvar> for details).
1778 Numerous updates; currently synchronised with Cygwin 1.1.4.
1782 Previously DYNIX/ptx had problems in its Configure probe for non-blocking I/O.
1788 EPOC update after Perl 5.6.0. See README.epoc.
1794 Perl now works on post-3.0 FreeBSDs.
1800 README.hpux updated; C<Configure -Duse64bitall> now almost works.
1806 Numerous compilation flag and hint enhancements; accidental mixing
1807 of 32-bit and 64-bit libraries (a doomed attempt) made much harder.
1817 Long doubles should now work (see INSTALL).
1821 Linux previously had problems related to sockaddrlen when using
1822 accept(), revcfrom() (in Perl: recv()), getpeername(), and getsockname().
1830 Compilation of the standard Perl distribution in MacOS Classic should
1831 now work if you have the Metrowerks development environment and
1832 the missing Mac-specific toolkit bits. Contact the macperl mailing
1839 MPE/iX update after Perl 5.6.0. See README.mpeix.
1845 Perl now works on NetBSD/sparc.
1851 Now works with usethreads (see INSTALL).
1857 64-bitness using the Sun Workshop compiler now works.
1861 Tru64 (aka Digital UNIX, aka DEC OSF/1)
1863 The operating system version letter now recorded in $Config{osvers}.
1864 Allow compiling with gcc (previously explicitly forbidden). Compiling
1865 with gcc still not recommended because buggy code results, even with
1872 Fixed various alignment problems that lead into core dumps either
1873 during build or later; no longer dies on math errors at runtime;
1874 now using full quad integers (64 bits), previously was using
1875 only 46 bit integers for speed.
1881 chdir() now works better despite a CRT bug; now works with MULTIPLICITY
1882 (see INSTALL); now works with Perl's malloc.
1884 The tainting of C<%ENV> elements via C<keys> or C<values> was previously
1885 unimplemented. It now works as documented.
1887 The C<waitpid> emulation has been improved. The worst bug (now fixed)
1888 was that a pid of -1 would cause a wildcard search of all processes on
1889 the system. The most significant enhancement is that we can now
1890 usually get the completion status of a terminated process.
1892 POSIX-style signals are now emulated much better on VMS versions prior
1895 The C<system> function and backticks operator have improved
1896 functionality and better error handling.
1906 accept() no longer leaks memory.
1910 Borland C++ v5.5 is now a supported compiler that can build Perl.
1911 However, the generated binaries continue to be incompatible with those
1912 generated by the other supported compilers (GCC and Visual C++).
1916 Better chdir() return value for a non-existent directory.
1920 Duping socket handles with open(F, ">&MYSOCK") now works under Windows 9x.
1924 New %ENV entries now propagate to subprocesses.
1928 $ENV{LIB} now used to search for libs under Visual C.
1932 Win32::GetCwd() correctly returns C:\ instead of C: when at the drive root.
1933 Other bugs in chdir() and Cwd::cwd() have also been fixed.
1937 A failed (pseudo)fork now returns undef and sets errno to EAGAIN.
1941 HTML files will be installed in c:\perl\html instead of c:\perl\lib\pod\html
1945 The makefiles now provide a single switch to bulk-enable all the features
1946 enabled in ActiveState ActivePerl (a popular Win32 binary distribution).
1950 Allow REG_EXPAND_SZ keys in the registry.
1954 Can now send() from all threads, not just the first one.
1958 Fake signal handling reenabled, bugs and all.
1962 Less stack reserved per thread so that more threads can run
1963 concurrently. (Still 16M per thread.)
1967 C<File::Spec->tmpdir()> now prefers C:/temp over /tmp
1968 (works better when perl is running as service).
1972 Better UNC path handling under ithreads.
1976 wait() and waitpid() now work much better.
1980 winsock handle leak fixed.
1986 =head1 New or Changed Diagnostics
1992 All regular expression compilation error messages are now hopefully
1993 easier to understand both because the error message now comes before
1994 the failed regex and because the point of failure is now clearly
1995 marked by a C<E<lt>-- HERE> marker.
1999 The various "opened only for", "on closed", "never opened" warnings
2000 drop the C<main::> prefix for filehandles in the C<main> package,
2001 for example C<STDIN> instead of C<main::STDIN>.
2005 The "Unrecognized escape" warning has been extended to include C<\8>,
2006 C<\9>, and C<\_>. There is no need to escape any of the C<\w> characters.
2010 Two new debugging options have been added: if you have compiled your
2011 Perl with debugging, you can use the -DT and -DR options to trace
2012 tokenising and to add reference counts to displaying variables,
2017 If an attempt to use a (non-blessed) reference as an array index
2018 is made, a warning is given.
2022 C<push @a;> and C<unshift @a;> (with no values to push or unshift)
2023 now give a warning. This may be a problem for generated and evaled
2028 If you try to L<perlfunc/pack> a number less than 0 or larger than 255
2029 using the C<"C"> format you will get an optional warning. Similarly
2030 for the C<"c"> format and a number less than -128 or more than 127.
2034 Certain regex modifiers such as C<(?o)> make sense only if applied to
2035 the entire regex. You will an optional warning if you try to do otherwise.
2039 Using arrays or hashes as references (e.g. C<< %foo->{bar} >>
2040 has been deprecated for a while. Now you will get an optional warning.
2044 =head1 Changed Internals
2050 perlapi.pod (a companion to perlguts) now attempts to document the
2055 You can now build a really minimal perl called microperl.
2056 Building microperl does not require even running Configure;
2057 C<make -f Makefile.micro> should be enough. Beware: microperl makes
2058 many assumptions, some of which may be too bold; the resulting
2059 executable may crash or otherwise misbehave in wondrous ways.
2060 For careful hackers only.
2064 Added rsignal(), whichsig(), do_join(), op_clear, op_null,
2065 ptr_table_clear(), ptr_table_free(), sv_setref_uv(), and several UTF-8
2066 interfaces to the publicised API. For the full list of the available
2067 APIs see L<perlapi>.
2071 Made possible to propagate customised exceptions via croak()ing.
2075 Now xsubs can have attributes just like subs.
2079 dTHR and djSP have been obsoleted; the former removed (because it's
2080 a no-op) and the latter replaced with dSP.
2084 PERL_OBJECT has been completely removed.
2088 The MAGIC constants (e.g. C<'P'>) have been macrofied
2089 (e.g. C<PERL_MAGIC_TIED>) for better source code readability
2090 and maintainability.
2094 The regex compiler now maintains a structure that identifies nodes in
2095 the compiled bytecode with the corresponding syntactic features of the
2096 original regex expression. The information is attached to the new
2097 C<offsets> member of the C<struct regexp>. See L<perldebguts> for more
2098 complete information.
2102 The C code has been made much more C<gcc -Wall> clean. Some warning
2103 messages still remain in some platforms, so if you are compiling with
2104 gcc you may see some warnings about dubious practices. The warnings
2105 are being worked on.
2109 F<perly.c>, F<sv.c>, and F<sv.h> have now been extensively commented.
2113 Documentation on how to use the Perl source repository has been added
2114 to F<Porting/repository.pod>.
2118 There are now several profiling make targets.
2122 =head1 Security Vulnerability Closed
2124 (This change was already made in 5.7.0 but bears repeating here.)
2126 A potential security vulnerability in the optional suidperl component
2127 of Perl was identified in August 2000. suidperl is neither built nor
2128 installed by default. As of November 2001 the only known vulnerable
2129 platform is Linux, most likely all Linux distributions. CERT and
2130 various vendors and distributors have been alerted about the vulnerability.
2131 See http://www.cpan.org/src/5.0/sperl-2000-08-05/sperl-2000-08-05.txt
2132 for more information.
2134 The problem was caused by Perl trying to report a suspected security
2135 exploit attempt using an external program, /bin/mail. On Linux
2136 platforms the /bin/mail program had an undocumented feature which
2137 when combined with suidperl gave access to a root shell, resulting in
2138 a serious compromise instead of reporting the exploit attempt. If you
2139 don't have /bin/mail, or if you have 'safe setuid scripts', or if
2140 suidperl is not installed, you are safe.
2142 The exploit attempt reporting feature has been completely removed from
2143 Perl 5.8.0 (and the maintenance release 5.6.1, and it was removed also
2144 from all the Perl 5.7 releases), so that particular vulnerability
2145 isn't there anymore. However, further security vulnerabilities are,
2146 unfortunately, always possible. The suidperl functionality is most
2147 probably going to be removed in Perl 5.10. In any case, suidperl
2148 should only be used by security experts who know exactly what they are
2149 doing and why they are using suidperl instead of some other solution
2150 such as sudo (see http://www.courtesan.com/sudo/).
2154 Several new tests have been added, especially for the F<lib>
2155 subsection. There are now about 34 000 individual tests (spread over
2156 about 530 test scripts), in the regression suite (5.6.1 has about
2157 11700 tests, in 258 test scripts) Many of the new tests are introduced
2158 by the new modules, but still in general Perl is now more thoroughly
2161 Because of the large number of tests, running the regression suite
2162 will take considerably longer time than it used to: expect the suite
2163 to take up to 4-5 times longer to run than in perl 5.6. In a really
2164 fast machine you can hope to finish the suite in about 5 minutes
2167 The tests are now reported in a different order than in earlier Perls.
2168 (This happens because the test scripts from under t/lib have been moved
2169 to be closer to the library/extension they are testing.)
2171 =head1 Known Problems
2179 In AIX 4.2 Perl extensions that use C++ functions that use statics
2180 may have problems in that the statics are not getting initialized.
2181 In newer AIX releases this has been solved by linking Perl with
2182 the libC_r library, but unfortunately in AIX 4.2 the said library
2183 has an obscure bug where the various functions related to time
2184 (such as time() and gettimeofday()) return broken values, and
2185 therefore in AIX 4.2 Perl is not linked against the libC_r.
2189 vac 5.0.0.0 May Produce Buggy Code For Perl
2191 The AIX C compiler vac version 5.0.0.0 may produce buggy code,
2192 resulting in few random tests failing, but when the failing tests
2193 are run by hand, they succeed. We suggest upgrading to at least
2194 vac version 5.0.1.0, that has been known to compile Perl correctly.
2195 "lslpp -L|grep vac.C" will tell you the vac version.
2199 =head2 Amiga Perl Invoking Mystery
2201 One cannot call Perl using the C<volume:> syntax, that is, C<perl -v>
2202 works, but for example C<bin:perl -v> doesn't. The exact reason isn't
2203 known but the current suspect is the F<ixemul> library.
2205 =head2 lib/ftmp-security tests warn 'system possibly insecure'
2207 Don't panic. Read INSTALL 'make test' section instead.
2209 =head2 Cygwin intermittent failures of lib/Memoize/t/expire_file 11 and 12
2211 The subtests 11 and 12 sometimes fail and sometimes work.
2213 =head2 HP-UX lib/io_multihomed Fails When LP64-Configured
2215 The lib/io_multihomed test may hang in HP-UX if Perl has been
2216 configured to be 64-bit. Because other 64-bit platforms do not hang in
2217 this test, HP-UX is suspect. All other tests pass in 64-bit HP-UX. The
2218 test attempts to create and connect to "multihomed" sockets (sockets
2219 which have multiple IP addresses).
2221 =head2 HP-UX lib/posix Subtest 9 Fails When LP64-Configured
2223 If perl is configured with -Duse64bitall, the successful result of the
2224 subtest 10 of lib/posix may arrive before the successful result of the
2225 subtest 9, which confuses the test harness so much that it thinks the
2228 =head2 Linux With Sfio Fails op/misc Test 48
2234 The following tests are known to fail:
2236 Failed Test Stat Wstat Total Fail Failed List of Failed
2237 -------------------------------------------------------------------------
2238 ../ext/DB_File/t/db-btree.t 0 11 ?? ?? % ??
2239 ../ext/DB_File/t/db-recno.t 149 3 2.01% 61 63 65
2240 ../ext/POSIX/t/posix.t 31 1 3.23% 10
2241 ../lib/warnings.t 450 1 0.22% 316
2245 OS/390 has rather many test failures but the situation is actually
2246 better than it was in 5.6.0, it's just that so many new modules and
2247 tests have been added.
2249 Failed Test Stat Wstat Total Fail Failed List of Failed
2250 -----------------------------------------------------------------------------
2251 ../ext/B/Deparse.t 14 1 7.14% 14
2252 ../ext/B/Showlex.t 1 1 100.00% 1
2253 ../ext/Encode/Encode/Tcl.t 610 13 2.13% 592 594 596 598
2255 ../ext/IO/lib/IO/t/io_unix.t 113 28928 5 3 60.00% 3-5
2256 ../ext/POSIX/POSIX.t 29 1 3.45% 14
2257 ../ext/Storable/t/lock.t 255 65280 5 3 60.00% 3-5
2258 ../lib/locale.t 129 33024 117 19 16.24% 99-117
2259 ../lib/warnings.t 434 1 0.23% 75
2260 ../lib/ExtUtils.t 27 1 3.70% 25
2261 ../lib/Math/BigInt/t/bigintpm.t 1190 1 0.08% 1145
2262 ../lib/Unicode/UCD.t 81 48 59.26% 1-16 49-64 66-81
2263 ../lib/User/pwent.t 9 1 11.11% 4
2264 op/pat.t 660 6 0.91% 242-243 424-425
2266 op/split.t 0 9 ?? ?? % ??
2267 op/taint.t 174 3 1.72% 156 162 168
2268 op/tr.t 70 3 4.29% 50 58-59
2269 Failed 16/422 test scripts, 96.21% okay. 105/23251 subtests failed, 99.55% okay.
2271 =head2 op/sprintf tests 129 and 130
2273 The op/sprintf tests 129 and 130 are known to fail on some platforms.
2274 Examples include any platform using sfio, and Compaq/Tandem's NonStop-UX.
2275 The failing platforms do not comply with the ANSI C Standard, line
2276 19ff on page 134 of ANSI X3.159 1989 to be exact. (They produce
2277 something other than "1" and "-1" when formatting 0.6 and -0.6 using
2278 the printf format "%.0f", most often they produce "0" and "-0".)
2280 =head2 Failure of Thread tests
2282 B<Note that support for 5.005-style threading remains experimental.>
2284 The following tests are known to fail due to fundamental problems in
2285 the 5.005 threading implementation. These are not new failures--Perl
2286 5.005_0x has the same bugs, but didn't have these tests.
2289 t/lib/thr5005.t 19-20
2297 ext/POSIX/sigaction subtests 6 and 13 may fail.
2301 lib/ExtUtils may spuriously claim that subtest 28 failed,
2302 which is interesting since the test only has 27 tests.
2306 Numerous numerical test failures
2308 op/numconvert 209,210,217,218
2310 ext/Time/HiRes/HiRes 9
2311 lib/Math/BigInt/t/bigintpm 1145
2314 These tests fail because of yet unresolved floating point inaccuracies.
2320 There are a few known test failures, see L<perluts>.
2324 There is one known test failure with a default configuration:
2326 [.run]switches..........................FAILED on test 1
2330 In multi-CPU boxes there are some problems with the I/O buffering:
2331 some output may appear twice.
2333 =head2 Localising a Tied Variable Leaks Memory
2336 tie my %tie_hash => 'Tie::StdHash';
2340 local($tie_hash{Foo}) = 1; # leaks
2342 Code like the above is known to leak memory every time the local()
2345 =head2 Localising Tied Arrays and Hashes Is Broken
2349 doesn't work as one would expect: the old value is restored
2352 =head2 Self-tying of Arrays and Hashes Is Forbidden
2354 Self-tying of arrays and hashes is broken in rather deep and
2355 hard-to-fix ways. As a stop-gap measure to avoid people from getting
2356 frustrated at the mysterious results (core dumps, most often) it is
2357 for now forbidden (you will get a fatal error even from an attempt).
2359 =head2 Variable Attributes are not Currently Usable for Tieing
2361 This limitation will hopefully be fixed in future. (Subroutine
2362 attributes work fine for tieing, see L<Attribute::Handlers>).
2364 One way to run into this limitation is to have a loop variable with
2365 attributes within a loop: the tie is called only once, not for each
2366 iteration of the loop.
2368 =head2 Building Extensions Can Fail Because Of Largefiles
2370 Some extensions like mod_perl are known to have issues with
2371 `largefiles', a change brought by Perl 5.6.0 in which file offsets
2372 default to 64 bits wide, where supported. Modules may fail to compile
2373 at all or compile and work incorrectly. Currently there is no good
2374 solution for the problem, but Configure now provides appropriate
2375 non-largefile ccflags, ldflags, libswanted, and libs in the %Config
2376 hash (e.g., $Config{ccflags_nolargefiles}) so the extensions that are
2377 having problems can try configuring themselves without the
2378 largefileness. This is admittedly not a clean solution, and the
2379 solution may not even work at all. One potential failure is whether
2380 one can (or, if one can, whether it's a good idea) link together at
2381 all binaries with different ideas about file offsets, all this is
2384 =head2 Unicode Support on EBCDIC Still Spotty
2386 Though mostly working, Unicode support still has problem spots on
2387 EBCDIC platforms. One such known spot are the C<\p{}> and C<\P{}>
2388 regular expression constructs for code points less than 256: the
2389 pP are testing for Unicode code points, not knowing about EBCDIC.
2391 =head2 The Compiler Suite Is Still Experimental
2393 The compiler suite is slowly getting better but is nowhere near
2396 =head2 The Long Double Support is Still Experimental
2398 The ability to configure Perl's numbers to use "long doubles",
2399 floating point numbers of hopefully better accuracy, is still
2400 experimental. The implementations of long doubles are not yet
2401 widespread and the existing implementations are not quite mature
2402 or standardised, therefore trying to support them is a rare
2403 and moving target. The gain of more precision may also be offset
2404 by slowdown in computations (more bits to move around, and the
2405 operations are more likely to be executed by less optimised
2408 =head1 Reporting Bugs
2410 If you find what you think is a bug, you might check the articles
2411 recently posted to the comp.lang.perl.misc newsgroup and the perl
2412 bug database at http://bugs.perl.org. There may also be
2413 information at http://www.perl.com/perl/, the Perl Home Page.
2415 If you believe you have an unreported bug, please run the B<perlbug>
2416 program included with your release. Be sure to trim your bug down
2417 to a tiny but sufficient test case. Your bug report, along with the
2418 output of C<perl -V>, will be sent off to perlbug@perl.org to be
2419 analysed by the Perl porting team.
2423 The F<Changes> file for exhaustive details on what changed.
2425 The F<INSTALL> file for how to build Perl.
2427 The F<README> file for general stuff.
2429 The F<Artistic> and F<Copying> files for copyright information.
2433 Written by Jarkko Hietaniemi <F<jhi@iki.fi>>.