3 perldelta - what is new for perl v5.8.0
7 This document describes differences between the 5.6.0 release
10 Many of the bug fixes in 5.8.0 were already seen in the 5.6.1
11 maintenance release since the two releases were kept closely
14 If you are upgrading from Perl 5.005_03, you might also want
15 to read L<perl56delta>.
17 =head1 Highlights In 5.8.0
23 Better Unicode support
27 New Thread Implementation
35 Better Numeric Accuracy
43 More Extensive Regression Testing
47 =head1 Incompatible Changes
49 =head2 64-bit platforms and malloc
51 If your pointers are 64 bits wide, the Perl malloc is no longer being
52 used because it does not work well with 8-byte pointers. Also,
53 usually the system mallocs on such platforms are much better optimized
54 for such large memory models than the Perl malloc. Some memory-hungry
55 Perl applications like the PDL don't work well with Perl's malloc.
56 Finally, other applications than Perl (like modperl) tend to prefer
57 the system malloc. Such platforms include Alpha and 64-bit HPPA,
60 =head2 AIX Dynaloading
62 The AIX dynaloading now uses in AIX releases 4.3 and newer the native
63 dlopen interface of AIX instead of the old emulated interface. This
64 change will probably break backward compatibility with compiled
65 modules. The change was made to make Perl more compliant with other
66 applications like modperl which are using the AIX native interface.
68 =head2 Attributes for C<my> variables now handled at run-time.
70 The C<my EXPR : ATTRS> syntax now applies variable attributes at
71 run-time. (Subroutine and C<our> variables still get attributes applied
72 at compile-time.) See L<attributes> for additional details. In particular,
73 however, this allows variable attributes to be useful for C<tie> interfaces,
74 which was a deficiency of earlier releaes.
76 =head2 Socket Extension Dynamic in VMS
78 The Socket extension is now dynamically loaded instead of being
79 statically built in. This may or may not be a problem with ancient
80 TCP/IP stacks of VMS: we do not know since we weren't able to test
81 Perl in such configurations.
83 =head2 IEEE-format Floating Point Default on OpenVMS Alpha
85 Perl now uses IEEE format (T_FLOAT) as the default internal floating
86 point format on OpenVMS Alpha, potentially breaking binary compatibility
87 with external libraries or existing data. G_FLOAT is still available as
88 a configuration option. The default on VAX (D_FLOAT) has not changed.
90 =head2 Different Definition of the Unicode Character Classes \p{In...}
92 As suggested by the Unicode consortium, the Unicode character classes
93 now prefer I<scripts> as opposed to I<blocks> (as defined by Unicode);
94 in Perl, when the C<\p{In....}> and the C<\p{In....}> regular expression
95 constructs are used. This has changed the definition of some of those
98 The difference between scripts and blocks is that scripts are the
99 glyphs used by a language or a group of languages, while the blocks
100 are more artificial groupings of 256 characters based on the Unicode
103 In general this change results in more inclusive Unicode character
104 classes, but changes to the other direction also do take place:
105 for example while the script C<Latin> includes all the Latin
106 characters and their various diacritic-adorned versions, it
107 does not include the various punctuation or digits (since they
108 are not solely C<Latin>).
110 Changes in the character class semantics may have happened if a script
111 and a block happen to have the same name, for example C<Hebrew>.
112 In such cases the script wins and C<\p{InHebrew}> now means the script
113 definition of Hebrew. The block definition in still available,
114 though, by appending C<Block> to the name: C<\p{InHebrewBlock}> means
115 what C<\p{InHebrew}> meant in perl 5.6.0. For the full list
116 of affected character classes, see L<perlunicode/Blocks>.
118 =head2 Perl Parser Stress Tested
120 The Perl parser has been stress tested using both random input and
121 Markov chain input and the few found crashes and lockups have been
124 =head2 REF(...) Instead Of SCALAR(...)
126 A reference to a reference now stringifies as "REF(0x81485ec)" instead
127 of "SCALAR(0x81485ec)" in order to be more consistent with the return
136 The semantics of bless(REF, REF) were unclear and until someone proves
137 it to make some sense, it is forbidden.
141 The obsolete chat2 library that should never have been allowed
142 to escape the laboratory has been decommissioned.
146 The very dusty examples in the eg/ directory have been removed.
147 Suggestions for new shiny examples welcome but the main issue is that
148 the examples need to be documented, tested and (most importantly)
153 The (bogus) escape sequences \8 and \9 now give an optional warning
154 ("Unrecognized escape passed through"). There is no need to \-escape
159 The list of filenames from glob() (or <...>) is now by default sorted
160 alphabetically to be csh-compliant (which is what happened before
161 in most UNIX platforms). (bsd_glob() does still sort platform
162 natively, ASCII or EBCDIC, unless GLOB_ALPHASORT is specified.)
166 Spurious syntax errors generated in certain situations, when glob()
167 caused File::Glob to be loaded for the first time, have been fixed.
171 Although "you shouldn't do that", it was possible to write code that
172 depends on Perl's hashed key order (Data::Dumper does this). The new
173 algorithm "One-at-a-Time" produces a different hashed key order.
174 More details are in L</"Performance Enhancements">.
178 lstat(FILEHANDLE) now gives a warning because the operation makes no sense.
179 In future releases this may become a fatal error.
183 The C<package;> syntax (C<package> without an argument) has been
184 deprecated. Its semantics were never that clear and its
185 implementation even less so. If you have used that feature to
186 disallow all but fully qualified variables, C<use strict;> instead.
190 The unimplemented POSIX regex features [[.cc.]] and [[=c=]] are still
191 recognised but now cause fatal errors. The previous behaviour of
192 ignoring them by default and warning if requested was unacceptable
193 since it, in a way, falsely promised that the features could be used.
197 The current user-visible implementation of pseudo-hashes (the weird
198 use of the first array element) is deprecated starting from Perl 5.8.0
199 and will be removed in Perl 5.10.0, and the feature will be
200 implemented differently. Not only is the current interface rather
201 ugly, but the current implementation slows down normal array and hash
202 use quite noticeably. The C<fields> pragma interface will remain
207 The syntaxes C<< @a->[...] >> and C<< %h->{...} >> have now been deprecated.
211 After years of trying the suidperl is considered to be too complex to
212 ever be considered truly secure. The suidperl functionality is likely
213 to be removed in a future release.
217 The long deprecated uppercase aliases for the string comparison
218 operators (EQ, NE, LT, LE, GE, GT) have now been removed.
222 The tr///C and tr///U features have been removed and will not return;
223 the interface was a mistake. Sorry about that. For similar
224 functionality, see pack('U0', ...) and pack('C0', ...).
228 =head1 Core Enhancements
230 =head2 PerlIO is Now The Default
236 IO is now by default done via PerlIO rather than system's "stdio".
237 PerlIO allows "layers" to be "pushed" onto a file handle to alter the
238 handle's behaviour. Layers can be specified at open time via 3-arg
241 open($fh,'>:crlf :utf8', $path) || ...
243 or on already opened handles via extended C<binmode>:
245 binmode($fh,':encoding(iso-8859-7)');
247 The built-in layers are: unix (low level read/write), stdio (as in
248 previous Perls), perlio (re-implementation of stdio buffering in a
249 portable manner), crlf (does CRLF <=> "\n" translation as on Win32,
250 but available on any platform). A mmap layer may be available if
251 platform supports it (mostly UNIXes).
253 Layers to be applied by default may be specified via the 'open' pragma.
255 See L</"Installation and Configuration Improvements"> for the effects
256 of PerlIO on your architecture name.
260 File handles can be marked as accepting Perl's internal encoding of Unicode
261 (UTF-8 or UTF-EBCDIC depending on platform) by a pseudo layer ":utf8" :
263 open($fh,">:utf8","Uni.txt");
265 Note for EBCDIC users: the pseudo layer ":utf8" is erroneously named
266 for you since it's not UTF-8 what you will be getting but instead
267 UTF-EBCDIC. See L<perlunicode>, L<utf8>, and
268 http://www.unicode.org/unicode/reports/tr16/ for more information.
269 In future releases this naming may change.
273 File handles can translate character encodings from/to Perl's internal
274 Unicode form on read/write via the ":encoding()" layer.
278 File handles can be opened to "in memory" files held in Perl scalars via:
280 open($fh,'>', \$variable) || ...
284 Anonymous temporary files are available without need to
285 'use FileHandle' or other module via
287 open($fh,"+>", undef) || ...
289 That is a literal undef, not an undefined value.
293 The list form of C<open> is now implemented for pipes (at least on UNIX):
295 open($fh,"-|", 'cat', '/etc/motd')
297 creates a pipe, and runs the equivalent of exec('cat', '/etc/motd') in
302 =head2 Signals Are Now Safe
304 Perl used to be fragile in that signals arriving at inopportune moments
305 could corrupt Perl's internal state. Now Perl postpones handling of
306 signals until it's safe.
308 =head2 Unicode Overhaul
310 Unicode in general should be now much more usable than in Perl 5.6.0
311 (or even in 5.6.1). Unicode can be used in hash keys, Unicode in
312 regular expressions should work now, Unicode in tr/// should work now,
313 Unicode in I/O should work now.
319 The Unicode Character Database coming with Perl has been upgraded
320 to Unicode 3.1.1. For more information, see http://www.unicode.org/.
324 For developers interested in enhancing Perl's Unicode capabilities:
325 almost all the UCD files are included with the Perl distribution in
326 the lib/unicore subdirectory. The most notable omission, for space
327 considerations, is the Unihan database.
331 The Unicode character classes \p{Blank} and \p{SpacePerl} have been
332 added. "Blank" is like C isblank(), that is, it contains only
333 "horizontal whitespace" (the space character is, the newline isn't),
334 and the "SpacePerl" is the Unicode equivalent of C<\s> (\p{Space}
335 isn't, since that includes the vertical tabulator character, whereas
340 =head2 Understanding of Numbers
342 In general a lot of fixing has happened in the area of Perl's
343 understanding of numbers, both integer and floating point. Since in
344 many systems the standard number parsing functions like C<strtoul()>
345 and C<atof()> seem to have bugs, Perl tries to work around their
346 deficiencies. This results hopefully in more accurate numbers.
348 Perl now tries internally to use integer values in numeric conversions
349 and basic arithmetics (+ - * /) if the arguments are integers, and
350 tries also to keep the results stored internally as integers.
351 This change leads to often slightly faster and always less lossy
352 arithmetics. (Previously Perl always preferred floating point numbers
355 =head2 Miscellaneous Enhancements
361 AUTOLOAD is now lvaluable, meaning that you can add the :lvalue attribute
362 to AUTOLOAD subroutines and you can assign to the AUTOLOAD return value.
366 C<perl -d:Module=arg,arg,arg> now works (previously one couldn't pass
367 in multiple arguments.)
371 END blocks are now run even if you exit/die in a BEGIN block.
372 Internally, the execution of END blocks is now controlled by
373 PL_exit_flags & PERL_EXIT_DESTRUCT_END. This enables the new
374 behaviour for Perl embedders. This will default in 5.10. See
379 Formats now support zero-padded decimal fields.
383 Lvalue subroutines can now return C<undef> in list context.
384 However, the lvalue subroutine feature still remains experimental.
388 A new special regular expression variable has been introduced:
389 C<$^N>, which contains the most-recently closed group (submatch).
393 C<no Module;> now works even if there is no "sub unimport" in the Module.
397 The numerical comparison operators return C<undef> if either operand
398 is a NaN. Previously the behaviour was unspecified.
402 The following builtin functions are now overridable: each(), keys(),
403 pop(), push(), shift(), splice(), unshift().
407 C<pack('U0a*', ...)> can now be used to force a string to UTF8.
411 my __PACKAGE__ $obj now works.
415 The printf() and sprintf() now support parameter reordering using the
416 C<%\d+\$> and C<*\d+\$> syntaxes. For example
418 print "%2\$s %1\$s\n", "foo", "bar";
420 will print "bar foo\n". This feature helps in writing
421 internationalised software, and in general when the order
422 of the parameters can vary.
426 prototype(\&) is now available.
430 prototype(\[$@%&]) is now available to implicitly create references
431 (useful for example if you want to emulate the tie() interface).
435 untie() will now call an UNTIE() hook if it exists. See L<perltie>
440 L<utime> now supports C<utime undef, undef, @files> to change the
441 file timestamps to the current time.
445 The rules for allowing underscores (underbars) in numeric constants
446 have been relaxed and simplified: now you can have an underscore
447 simply B<between digits>.
451 =head1 Modules and Pragmata
453 =head2 New Modules and Pragmata
459 C<Attribute::Handlers> allows a class to define attribute handlers.
462 use Attribute::Handlers;
463 sub Wolf :ATTR(SCALAR) { print "howl!\n" }
465 # later, in some package using or inheriting from MyPack...
467 my MyPack $Fluffy : Wolf; # the attribute handler Wolf will be called
469 Both variables and routines can have attribute handlers. Handlers can
470 be specific to type (SCALAR, ARRAY, HASH, or CODE), or specific to the
471 exact compilation phase (BEGIN, CHECK, INIT, or END).
475 B<B::Concise> is a new compiler backend for walking the Perl syntax
476 tree, printing concise info about ops, from Stephen McCamant. The
477 output is highly customisable. See L<B::Concise>.
481 C<Class::ISA> for reporting the search path for a class's ISA tree,
482 by Sean Burke, has been added. See L<Class::ISA>.
486 C<Cwd> has now a split personality: if possible, an XS extension is
487 used, (this will hopefully be faster, more secure, and more robust)
488 but if not possible, the familiar Perl implementation is used.
492 C<Devel::PPPort>, originally from Kenneth Albanowski and now
493 maintained by Paul Marquess, has been added. It is primarily used
494 by C<h2xs> to enhance portability of of XS modules between different
499 C<Digest>, frontend module for calculating digests (checksums), from
500 Gisle Aas, has been added. See L<Digest>.
504 C<Digest::MD5> for calculating MD5 digests (checksums) as defined in
505 RFC 1321, from Gisle Aas, has been added. See L<Digest::MD5>.
507 use Digest::MD5 'md5_hex';
509 $digest = md5_hex("Thirsty Camel");
511 print $digest, "\n"; # 01d19d9d2045e005c3f1b80e8b164de1
513 NOTE: the C<MD5> backward compatibility module is deliberately not
514 included since its further use is discouraged.
518 C<Encode>, by Nick Ing-Simmons, provides a mechanism to translate
519 between different character encodings. Support for Unicode,
520 ISO-8859-*, ASCII, CP*, KOI8-R, and three variants of EBCDIC are
521 compiled in to the module. Several other encodings (like Japanese,
522 Chinese, and MacIntosh encodings) are included and will be loaded at
523 runtime. See L<Encode>.
525 Any encoding supported by Encode module is also available to the
526 ":encoding()" layer if PerlIO is used.
530 C<I18N::Langinfo> can be use to query locale information.
531 See L<I18N::Langinfo>.
535 C<I18N::LangTags> has functions for dealing with RFC3066-style
536 language tags, by Sean Burke. See L<I18N::LangTags>.
540 C<ExtUtils::Constant> is a new tool for extension writers for
541 generating XS code to import C header constants, by Nicholas Clark.
542 See L<ExtUtils::Constant>.
546 C<Filter::Simple> is an easy-to-use frontend to Filter::Util::Call,
547 from Damian Conway. See L<Filter::Simple>.
553 use Filter::Simple sub {
554 while (my ($from, $to) = splice @_, 0, 2) {
563 use MyFilter qr/red/ => 'green';
565 print "red\n"; # this code is filtered, will print "green\n"
566 print "bored\n"; # this code is filtered, will print "bogreen\n"
570 print "red\n"; # this code is not filtered, will print "red\n"
574 C<File::Temp> allows one to create temporary files and directories in
575 an easy, portable, and secure way, by Tim Jenness. See L<File::Temp>.
579 C<Filter::Util::Call> provides you with the framework to write
580 I<Source Filters> in Perl, from Paul Marquess. For most uses the
581 frontend Filter::Simple is to be preferred. See L<Filter::Util::Call>.
585 L<libnet> is a collection of perl5 modules related to network
586 programming, from Graham Barr. See L<Net::FTP>, L<Net::NNTP>,
587 L<Net::Ping>, L<Net::POP3>, L<Net::SMTP>, and L<Net::Time>.
589 Perl installation leaves libnet unconfigured, use F<libnetcfg> to configure.
593 C<List::Util> is a selection of general-utility list subroutines, like
594 sum(), min(), first(), and shuffle(), by Graham Barr. See L<List::Util>.
598 C<Locale::Constants>, C<Locale::Country>, C<Locale::Currency>, and
599 C<Locale::Language>, from Neil Bowers, have been added. They provide the
600 codes for various locale standards, such as "fr" for France, "usd" for
601 US Dollar, and "jp" for Japanese.
605 $country = code2country('jp'); # $country gets 'Japan'
606 $code = country2code('Norway'); # $code gets 'no'
608 See L<Locale::Constants>, L<Locale::Country>, L<Locale::Currency>,
609 and L<Locale::Language>.
613 C<Locale::Maketext> is localization framework from Sean Burke. See
614 L<Locale::Maketext>, and L<Locale::Maketext::TPJ13>. The latter is an
615 article about software localization, originally published in The Perl
616 Journal #13, republished here with kind permission.
620 C<Memoize> can make your functions faster by trading space for time,
621 from Mark-Jason Dominus. See L<Memoize>.
625 C<MIME::Base64> allows you to encode data in base64, from Gisle Aas,
626 as defined in RFC 2045 - I<MIME (Multipurpose Internet Mail
631 $encoded = encode_base64('Aladdin:open sesame');
632 $decoded = decode_base64($encoded);
634 print $encoded, "\n"; # "QWxhZGRpbjpvcGVuIHNlc2FtZQ=="
640 C<MIME::QuotedPrint> allows you to encode data in quoted-printable
641 encoding, as defined in RFC 2045 - I<MIME (Multipurpose Internet Mail
642 Extensions)>, from Gisle Aas.
644 use MIME::QuotedPrint;
646 $encoded = encode_qp("Smiley in Unicode: \x{263a}");
647 $decoded = decode_qp($encoded);
649 print $encoded, "\n"; # "Smiley in Unicode: =263A"
651 MIME::QuotedPrint has been enhanced to provide the basic methods
652 necessary to use it with PerlIO::Via as in :
654 use MIME::QuotedPrint;
655 open($fh,">Via(MIME::QuotedPrint)",$path);
657 See L<MIME::QuotedPrint>.
661 C<NEXT> is pseudo-class for method redispatch, from Damian Conway.
666 C<open> is a new pragma for setting the default I/O disciplines
671 C<PerlIO::Scalar> provides the implementation of IO to "in memory"
672 Perl scalars as discussed above, from Nick Ing-Simmons. It also
673 serves as an example of a loadable PerlIO layer. Other future
674 possibilities include PerlIO::Array and PerlIO::Code.
675 See L<PerlIO::Scalar>.
679 C<PerlIO::Via> acts as a PerlIO layer and wraps PerlIO layer
680 functionality provided by a class (typically implemented in perl
681 code), from Nick Ing-Simmons.
683 use MIME::QuotedPrint;
684 open($fh,">Via(MIME::QuotedPrint)",$path);
686 This will automatically convert everything output to C<$fh>
687 to Quoted-Printable. See L<PerlIO::Via>.
691 C<Pod::ParseLink>, by Russ Allbery, has been added,
692 to parse LZ<><> links in pods as described in the new
697 C<Pod::Text::Overstrike>, by Joe Smith, has been added.
698 It converts POD data to formatted overstrike text.
699 See L<Pod::Text::Overstrike>.
703 C<Scalar::Util> is a selection of general-utility scalar subroutines,
704 like blessed(), reftype(), and tainted(). See L<Scalar::Util>.
708 C<sort> is a new pragma for controlling the behaviour of sort().
712 C<Storable> gives persistence to Perl data structures by allowing the
713 storage and retrieval of Perl data to and from files in a fast and
714 compact binary format, from Raphael Manfredi. See L<Storable>.
718 C<Switch>, from Damian Conway, has been added. Just by saying
722 you have C<switch> and C<case> available in Perl.
728 case 1 { print "number 1" }
729 case "a" { print "string a" }
730 case [1..10,42] { print "number in list" }
731 case (@array) { print "number in list" }
732 case /\w+/ { print "pattern" }
733 case qr/\w+/ { print "pattern" }
734 case (%hash) { print "entry in hash" }
735 case (\%hash) { print "entry in hash" }
736 case (\&sub) { print "arg to subroutine" }
737 else { print "previous case not true" }
744 C<Test::More> is yet another framework for writing test scripts,
745 more extensive than Test::Simple, by Michael Schwern. See L<Test::More>.
749 C<Test::Simple> has basic utilities for writing tests, by Michael
750 Schwern. See L<Test::Simple>.
754 C<Text::Balanced> has been added, for extracting delimited text
755 sequences from strings, from Damian Conway.
757 use Text::Balanced 'extract_delimited';
759 ($a, $b) = extract_delimited("'never say never', he never said", "'", '');
761 $a will be "'never say never'", $b will be ', he never said'.
763 In addition to extract_delimited() there are also extract_bracketed(),
764 extract_quotelike(), extract_codeblock(), extract_variable(),
765 extract_tagged(), extract_multiple(), gen_delimited_pat(), and
766 gen_extract_tagged(). With these you can implement rather advanced
767 parsing algorithms. See L<Text::Balanced>.
771 C<threads> is an interface to interpreter threads, by Arthur Bergman.
772 Interpreter threads (ithreads) is the new thread model introduced in
773 Perl 5.6 but only available as an internal interface for extension
774 writers (and for Win32 Perl for C<fork()> emulation). See L<threads>.
778 C<threads::shared> allows data sharing for interpreter threads, from
779 Arthur Bergman. In the ithreads model any data sharing between
780 threads must be explicit, as opposed to the old 5.005 thread model
781 where data sharing was implicit. See L<threads::shared>.
785 C<Tie::RefHash::Nestable>, by Edward Avis, allows storing hash
786 references (unlike the standard Tie::RefHash) The module is contained
787 within Tie::RefHash, see L<Tie::RefHash>.
791 C<Time::HiRes> provides high resolution timing (ualarm, usleep,
792 and gettimeofday), from Douglas E. Wegscheid. See L<Time::HiRes>.
796 C<Unicode::UCD> offers a querying interface to the Unicode Character
797 Database. See L<Unicode::UCD>.
801 C<Unicode::Collate> implements the UCA (Unicode Collation Algorithm)
802 for sorting Unicode strings, by SADAHIRO Tomoyuki. See L<Unicode::Collate>.
806 C<Unicode::Normalize> implements the various Unicode normalization
807 forms, by SADAHIRO Tomoyuki. See L<Unicode::Normalize>.
811 C<XS::Typemap>, by Tim Jenness, is a test extension that exercises XS
812 typemaps. Nothing gets installed but for extension writers the code
817 =head2 Updated And Improved Modules and Pragmata
823 The following independently supported modules have been updated to the
824 newest versions from CPAN: CGI, CPAN, DB_File, File::Spec, File::Temp,
825 Getopt::Long, Math::BigFloat, Math::BigInt, the podlators bundle
826 (Pod::Man, Pod::Text), Pod::LaTeX, Pod::Parser, Storable,
827 Term::ANSIColor, Test, Text-Tabs+Wrap.
831 The attributes::reftype() now works on tied arguments.
835 AutoLoader can now be disabled with C<no AutoLoader;>.
839 B::Deparse has been significantly enhanced. It now can deparse almost
840 all of the standard test suite (so that the tests still succeed).
841 There is a make target "test.deparse" for trying this out.
845 Class::Struct can now define the classes in compile time.
849 Class::Struct now assigns the array/hash element if the accessor
850 is called with an array/hash element as the B<sole> argument.
854 Data::Dumper has now an option to sort hashes.
858 Data::Dumper has now an option to dump code references
863 DB_File now supports newer Berkeley DB versions, among
868 The English module can now be used without the infamous performance
871 use English '-no_performance_hit';
873 (Assuming, of course, that one doesn't need the troublesome variables
874 C<$`>, C<$&>, or C<$'>.) Also, introduced C<@LAST_MATCH_START> and
875 C<@LAST_MATCH_END> English aliases for C<@-> and C<@+>.
879 Fcntl, Socket, and Sys::Syslog have been rewritten to use the
880 new-style constant dispatch section (see L<ExtUtils::Constant>).
881 This means that they will be more robust and hopefully faster.
885 File::Find now chdir()s correctly when chasing symbolic links.
889 File::Find now has pre- and post-processing callbacks. It also
890 correctly changes directories when chasing symbolic links. Callbacks
891 (naughtily) exiting with "next;" instead of "return;" now work.
895 File::Find is now (again) reentrant. It also has been made
900 File::Glob::glob() renamed to File::Glob::bsd_glob() to avoid
901 prototype mismatch with CORE::glob().
905 File::Glob now supports C<GLOB_LIMIT> constant to limit the size of
906 the returned list of filenames.
910 Devel::Peek now has an interface for the Perl memory statistics
911 (this works only if you are using perl's malloc, and if you have
912 compiled with debugging).
916 IPC::Open3 now allows the use of numeric file descriptors.
920 IO::Socket has now atmark() method, which returns true if the socket
921 is positioned at the out-of-band mark. The method is also exportable
922 as a sockatmark() function.
926 IO::Socket::INET has support for ReusePort option (if your platform
927 supports it). The Reuse option now has an alias, ReuseAddr. For clarity
928 you may want to prefer ReuseAddr.
932 IO::Socket::INET now supports C<LocalPort> of zero (usually meaning
933 that the operating system will make one up.)
937 use lib now works identically to @INC. Removing directories
938 with 'no lib' now works.
942 Math::BigFloat and Math::BigInt have undergone a full rewrite.
943 They are now magnitudes faster, and they support various
944 bignum libraries such as GMP and PARI as their backends.
948 Math::Complex handles inf, NaN etc., better.
952 Net::Ping has been enhanced. There is now "external" protocol which
953 uses Net::Ping::External module which runs external ping(1) and parses
954 the output. A version of Net::Ping::External is available in CPAN.
958 POSIX::sigaction() is now much more flexible and robust.
959 You can now install coderef handlers, 'DEFAULT', and 'IGNORE'
960 handlers, installing new handlers was not atomic.
964 In Safe the C<%INC> now localised in a Safe compartment so that
969 In SDBM_File on dosish platforms, some keys went missing because of
970 lack of support for files with "holes". A workaround for the problem
975 In Search::Dict one can now have a pre-processing hook for the
976 lines being searched.
980 The Shell module now has an OO interface.
984 The Test module has been significantly enhanced.
988 The vars pragma now supports declaring fully qualified variables.
989 (Something that C<our()> does not and will not support.)
993 The utf8:: name space (as in the pragma) provides various
994 Perl-callable functions to provide low level access to Perl's
995 internal Unicode representation. At the moment only length()
996 has been implemented.
1000 =head1 Utility Changes
1006 Emacs perl mode (emacs/cperl-mode.el) has been updated to version
1011 F<emacs/e2ctags.pl> is now much faster.
1015 C<h2ph> now supports C trigraphs.
1019 C<h2xs> now produces a template README.
1023 C<h2xs> now uses C<Devel::PPort> for better portability between
1024 different versions of Perl.
1028 C<h2xs> uses the new L<ExtUtils::Constant> module which will affect
1029 newly created extensions that define constants. Since the new code is
1030 more correct (if you have two constants where the first one is a
1031 prefix of the second one, the first constant B<never> gets defined),
1032 less lossy (it uses integers for integer constant, as opposed to the
1033 old code that used floating point numbers even for integer constants),
1034 and slightly faster, you might want to consider regenerating your
1035 extension code (the new scheme makes regenerating easy).
1036 L<h2xs> now also supports C trigraphs.
1040 C<libnetcfg> has been added to configure the libnet.
1044 C<perlbug> is now much more robust. It also sends the bug report to
1045 perl.org, not perl.com.
1049 C<perlcc> has been rewritten and its user interface (that is,
1050 command line) is much more like that of the UNIX C compiler, cc.
1051 (The perlbc tools has been removed. Use C<perlcc -B> instead.)
1055 C<perlivp> is a new Installation Verification Procedure utility
1056 for running any time after installing Perl.
1060 C<pod2html> now allows specifying a cache directory.
1064 C<s2p> has been completely rewritten in Perl. (It is in fact a full
1065 implementation of sed in Perl: you can use the sed functionality by
1066 using the C<psed> utility.)
1070 C<xsubpp> now understands POD documentation embedded in the *.xs files.
1074 C<xsubpp> now supports OUT keyword.
1078 =head1 New Documentation
1084 perl56delta details the changes between the 5.005 release and the
1089 perlclib documents the internal replacements for standard C library
1090 functions. (Interesting only for extension writers and Perl core
1095 perldebtut is a Perl debugging tutorial.
1099 perlebcdic contains considerations for running Perl on EBCDIC platforms.
1103 perlintro is a gentle introduction to Perl.
1107 perliol documents the internals of PerlIO with layers.
1111 perlmodstyle is a style guide for writing modules.
1115 perlnewmod tells about writing and submitting a new module.
1119 perlpacktut is a pack() tutorial.
1123 perlpod has been rewritten to be clearer and to record the best
1124 practices gathered over the years.
1128 perlpodspec is a more formal specification of the pod format,
1129 mainly of interest for writers of pod applications, not to
1130 people writing in pod.
1134 perlretut is a regular expression tutorial.
1138 perlrequick is a regular expressions quick-start guide.
1139 Yes, much quicker than perlretut.
1143 perltodo has been updated.
1147 perltootc has been renamed as perltooc (to not to conflict
1148 with perltoot in filesystems restricted to "8.3" names)
1152 perluniintro is an introduction to using Unicode in Perl
1153 (perlunicode is more of a reference)
1157 perlutil explains the command line utilities packaged with the Perl
1162 The following platform-specific documents are available before
1163 the installation as README.I<platform>, and after the installation
1166 perlaix perlamiga perlapollo perlbeos perlbs2000
1167 perlce perlcygwin perldgux perldos perlepoc perlhpux
1168 perlhurd perlmachten perlmacos perlmint perlmpeix
1169 perlnetware perlos2 perlos390 perlplan9 perlqnx perlsolaris
1170 perltru64 perluts perlvmesa perlvms perlvos perlwin32
1176 The documentation for the POSIX-BC platform is called "BS2000", to avoid
1177 confusion with the Perl POSIX module.
1181 The documentation for the WinCE platform is called "CE", to avoid
1182 confusion with the perlwin32 documentation on 8.3-restricted filesystems.
1186 =head1 Performance Enhancements
1192 map() could get pathologically slow when the result list it generates
1193 is larger than the source list. The performance has been improved for
1198 sort() has been changed to use primarily mergesort internally as
1199 opposed to the earlier quicksort. For very small lists this may
1200 result in slightly slower sorting times, but in general the speedup
1201 should be at least 20%. Additional bonuses are that the worst case
1202 behaviour of sort() is now better (in computer science terms it now
1203 runs in time O(N log N), as opposed to quicksort's Theta(N**2)
1204 worst-case run time behaviour), and that sort() is now stable
1205 (meaning that elements with identical keys will stay ordered as they
1206 were before the sort). See the C<sort> pragma for information.
1208 The story in more detail: suppose you want to serve yourself a little
1211 @digits = ( 3,1,4,1,5,9 );
1213 A numerical sort of the digits will yield (1,1,3,4,5,9), as expected.
1214 Which C<1> comes first is hard to know, since one C<1> looks pretty
1215 much like any other. You can regard this as totally trivial,
1216 or somewhat profound. However, if you just want to sort the even
1217 digits ahead of the odd ones, then what will
1219 sort { ($a % 2) <=> ($b % 2) } @digits;
1221 yield? The only even digit, C<4>, will come first. But how about
1222 the odd numbers, which all compare equal? With the quicksort algorithm
1223 used to implement Perl 5.6 and earlier, the order of ties is left up
1224 to the sort. So, as you add more and more digits of Pi, the order
1225 in which the sorted even and odd digits appear will change.
1226 and, for sufficiently large slices of Pi, the quicksort algorithm
1227 in Perl 5.8 won't return the same results even if reinvoked with the
1228 same input. The justification for this rests with quicksort's
1229 worst case behavior. If you run
1231 sort { $a <=> $b } ( 1 .. $N , 1 .. $N );
1233 (something you might approximate if you wanted to merge two sorted
1234 arrays using sort), doubling $N doesn't just double the quicksort time,
1235 it I<quadruples> it. Quicksort has a worst case run time that can
1236 grow like N**2, so-called I<quadratic> behaviour, and it can happen
1237 on patterns that may well arise in normal use. You won't notice this
1238 for small arrays, but you I<will> notice it with larger arrays,
1239 and you may not live long enough for the sort to complete on arrays
1240 of a million elements. So the 5.8 quicksort scrambles large arrays
1241 before sorting them, as a statistical defence against quadratic behaviour.
1242 But that means if you sort the same large array twice, ties may be
1243 broken in different ways.
1245 Because of the unpredictability of tie-breaking order, and the quadratic
1246 worst-case behaviour, quicksort was I<almost> replaced completely with
1247 a stable mergesort. I<Stable> means that ties are broken to preserve
1248 the original order of appearance in the input array. So
1250 sort { ($a % 2) <=> ($b % 2) } (3,1,4,1,5,9);
1252 will yield (4,3,1,1,5,9), guaranteed. The even and odd numbers
1253 appear in the output in the same order they appeared in the input.
1254 Mergesort has worst case O(NlogN) behaviour, the best value
1255 attainable. And, ironically, this mergesort does particularly
1256 well where quicksort goes quadratic: mergesort sorts (1..$N, 1..$N)
1257 in O(N) time. But quicksort was rescued at the last moment because
1258 it is faster than mergesort on certain inputs and platforms.
1259 For example, if you really I<don't> care about the order of even
1260 and odd digits, quicksort will run in O(N) time; it's very good
1261 at sorting many repetitions of a small number of distinct elements.
1262 The quicksort divide and conquer strategy works well on platforms
1263 with relatively small, very fast, caches. Eventually, the problem gets
1264 whittled down to one that fits in the cache, from which point it
1265 benefits from the increased memory speed.
1267 Quicksort was rescued by implementing a sort pragma to control aspects
1268 of the sort. The B<stable> subpragma forces stable behaviour,
1269 regardless of algorithm. The B<_quicksort> and B<_mergesort>
1270 subpragmas are heavy-handed ways to select the underlying implementation.
1271 The leading C<_> is a reminder that these subpragmas may not survive
1272 beyond 5.8. More appropriate mechanisms for selecting the implementation
1273 exist, but they wouldn't have arrived in time to save quicksort.
1277 Hashes now use Bob Jenkins "One-at-a-Time" hashing key algorithm
1278 (http://burtleburtle.net/bob/hash/doobs.html). This algorithm is
1279 reasonably fast while producing a much better spread of values than
1280 the old hashing algorithm (originally by Chris Torek, later tweaked by
1281 Ilya Zakharevich). Hash values output from the algorithm on a hash of
1282 all 3-char printable ASCII keys comes much closer to passing the
1283 DIEHARD random number generation tests. According to perlbench, this
1284 change has not affected the overall speed of Perl.
1288 unshift() should now be noticeably faster.
1292 =head1 Installation and Configuration Improvements
1294 =head2 Generic Improvements
1300 INSTALL now explains how you can configure Perl to use 64-bit
1301 integers even on non-64-bit platforms.
1305 Policy.sh policy change: if you are reusing a Policy.sh file
1306 (see INSTALL) and you use Configure -Dprefix=/foo/bar and in the old
1307 Policy $prefix eq $siteprefix and $prefix eq $vendorprefix, all of
1308 them will now be changed to the new prefix, /foo/bar. (Previously
1309 only $prefix changed.) If you do not like this new behaviour,
1310 specify prefix, siteprefix, and vendorprefix explicitly.
1314 A new optional location for Perl libraries, otherlibdirs, is available.
1315 It can be used for example for vendor add-ons without disturbing Perl's
1316 own library directories.
1320 In many platforms the vendor-supplied 'cc' is too stripped-down to
1321 build Perl (basically, 'cc' doesn't do ANSI C). If this seems
1322 to be the case and 'cc' does not seem to be the GNU C compiler
1323 'gcc', an automatic attempt is made to find and use 'gcc' instead.
1327 gcc needs to closely track the operating system release to avoid
1328 build problems. If Configure finds that gcc was built for a different
1329 operating system release than is running, it now gives a clearly visible
1330 warning that there may be trouble ahead.
1334 If binary compatibility with the 5.005 release is not wanted, Configure
1335 no longer suggests including the 5.005 modules in @INC.
1339 Configure C<-S> can now run non-interactively.
1343 Configure support for pdp11-style memory models has been removed due
1348 configure.gnu now works with options with whitespace in them.
1352 installperl now outputs everything to STDERR.
1356 $Config{byteorder} is now computed dynamically (this is more robust
1357 with "fat binaries" where an executable image contains binaries for
1358 more than one binary platform.)
1362 Because PerlIO is now the default on most platforms, "-perlio" doesn't
1363 get appended to the $Config{archname} (also known as $^O) anymore.
1364 Instead, if you explicitly choose not to use perlio (Configure command
1365 line option -Uuseperlio), you will get "-stdio" appended.
1369 Another change related to the architecture name is that "-64all"
1370 (-Duse64bitall, or "maximally 64-bit") is appended only if your
1371 pointers are 64 bits wide. (To be exact, the use64bitall is ignored.)
1375 In AFS installations one can configure the root of the AFS to be
1376 somewhere else than the default F</afs> by using the Configure
1377 parameter C<-Dafsroot=/some/where/else>.
1381 APPLLIB_EXP, a less-know configuration-time definition, has been
1382 documented. It can be used to prepend site-specific directories
1383 to Perl's default search path (@INC), see INSTALL for information.
1387 The version of Berkeley DB used when the Perl (and, presumably, the
1388 DB_File extension) was built is now available as
1389 C<@Config{qw(db_version_major db_version_minor db_version_patch)}>
1390 from Perl and as C<DB_VERSION_MAJOR_CFG DB_VERSION_MINOR_CFG
1391 DB_VERSION_PATCH_CFG> from C.
1395 Building Berkeley DB3 for compatibility modes for DB, NDBM, and ODBM
1396 has been documented in INSTALL.
1400 If you have CPAN access (either network or a local copy such as a
1401 CD-ROM) you can during specify extra modules to Configure to build and
1402 install with Perl using the -Dextras=... option. See INSTALL for
1407 In addition to config.over a new override file, config.arch, is
1408 available. That is supposed to be used by hints file writers for
1409 architecture-wide changes (as opposed to config.over which is for
1414 If your file system supports symbolic links you can build Perl outside
1415 of the source directory by
1417 mkdir /tmp/perl/build/directory
1418 cd /tmp/perl/build/directory
1419 sh /path/to/perl/source/Configure -Dmksymlinks ...
1421 This will create in /tmp/perl/build/directory a tree of symbolic links
1422 pointing to files in /path/to/perl/source. The original files are left
1423 unaffected. After Configure has finished you can just say
1427 and Perl will be built and tested, all in /tmp/perl/build/directory.
1431 For Perl developers several new make targets for profiling
1432 and debugging have been added, see L<perlhack>.
1438 Use of the F<gprof> tool to profile Perl has been documented in
1439 L<perlhack>. There is a make target called "perl.gprof" for
1440 generating a gprofiled Perl executable.
1444 If you have GCC 3, there is a make target called "perl.gcov" for
1445 creating a gcoved Perl executable for coverage analysis. See
1450 If you are on IRIX or Tru64 platforms, new profiling/debugging options
1451 have been added, see L<perlhack> for more information about pixie and
1458 Guidelines of how to construct minimal Perl installations have
1459 been added to INSTALL.
1463 The Thread extension is now not built at all under ithreads
1464 (C<Configure -Duseithreads>) because it wouldn't work anyway (the
1465 Thread extension requires being Configured with C<-Duse5005threads>).
1467 But note that the Thread.pm interface is now shared by both
1472 =head2 New Or Improved Platforms
1474 For the list of platforms known to support Perl,
1475 see L<perlport/"Supported Platforms">.
1481 AIX dynamic loading should be now better supported.
1485 AIX should now work better with gcc, threads, and 64-bitness. Also the
1486 long doubles support in AIX should be better now. See L<perlaix>.
1490 After a long pause, AmigaOS has been verified to be happy with Perl.
1494 AtheOS (http://www.atheos.cx/) is a new platform.
1498 DG/UX platform now supports the 5.005-style threads. See L<perldgux>.
1502 DYNIX/ptx platform (a.k.a. dynixptx) is supported at or near osvers 4.5.2.
1506 EBCDIC platforms (z/OS, also known as OS/390, POSIX-BC, and VM/ESA)
1507 have been regained. Many test suite tests still fail and the
1508 co-existence of Unicode and EBCDIC isn't quite settled, but the
1509 situation is much better than with Perl 5.6. See L<perlos390>,
1510 L<perlbs2000> (for POSIX-BC), and L<perlvmesa> for more information.
1514 Building perl with -Duseithreads or -Duse5005threads now works under
1515 HP-UX 10.20 (previously it only worked under 10.30 or later). You will
1516 need a thread library package installed. See README.hpux.
1520 MacOS Classic (MacPerl has of course been available since
1521 perl 5.004 but now the source code bases of standard Perl
1522 and MacPerl have been synchronised)
1526 MacOS X (or Darwin) should now be able to build Perl even on HFS+
1527 filesystems. (The case-insensitivity confused the Perl build process.)
1531 NCR MP-RAS is now supported.
1535 NetWare from Novell is now supported. See L<perlnetware>.
1539 NonStop-UX is now supported.
1543 NEC SUPER-UX is now supported.
1547 Amdahl UTS UNIX mainframe platform is now supported.
1551 WinCE is now supported. See L<perlce>.
1555 z/OS (formerly known as OS/390, formerly known as MVS OE) has now
1556 support for dynamic loading. This is not selected by default,
1557 however, you must specify -Dusedl in the arguments of Configure.
1561 =head1 Selected Bug Fixes
1563 Numerous memory leaks and uninitialized memory accesses have been
1564 hunted down. Most importantly anonymous subs used to leak quite
1571 The autouse pragma didn't work for Multi::Part::Function::Names.
1575 caller() could cause core dumps in certain situations. Carp was sometimes
1576 affected by this problem.
1580 chop(@list) in list context returned the characters chopped in
1581 reverse order. This has been reversed to be in the right order.
1585 Configure no longer includes the DBM libraries (dbm, gdbm, db, ndbm)
1586 when building the Perl binary. The only exception to this is SunOS 4.x,
1591 The behaviour of non-decimal but numeric string constants such as
1592 "0x23" was platform-dependent: in some platforms that was seen as 35,
1593 in some as 0, in some as a floating point number (don't ask). This
1594 was caused by Perl using the operating system libraries in a situation
1595 where the result of the string to number conversion is undefined: now
1596 Perl consistently handles such strings as zero in numeric contexts.
1600 The order of DESTROYs has been made more predictable.
1604 Several debugger fixes: exit code now reflects the script exit code,
1605 condition C<"0"> now treated correctly, the C<d> command now checks
1606 line number, the C<$.> no longer gets corrupted, all debugger output
1607 now goes correctly to the socket if RemotePort is set.
1611 Perl 5.6.0 could emit spurious warnings about redefinition of dl_error()
1612 when statically building extensions into perl. This has been corrected.
1616 L<dprofpp> -R didn't work.
1620 C<*foo{FORMAT}> now works.
1623 Infinity is now recognized as a number.
1627 UNIVERSAL::isa no longer caches methods incorrectly. (This broke
1628 the Tk extension with 5.6.0.)
1632 Lexicals I: lexicals outside an eval "" weren't resolved
1633 correctly inside a subroutine definition inside the eval "" if they
1634 were not already referenced in the top level of the eval""ed code.
1638 Lexicals II: lexicals leaked at file scope into subroutines that
1639 were declared before the lexicals.
1643 Lexical warnings now propagating correctly between scopes
1644 and into C<eval "...">.
1648 C<use warnings qw(FATAL all)> did not work as intended. This has been
1653 warnings::enabled() now reports the state of $^W correctly if the caller
1654 isn't using lexical warnings.
1658 Line renumbering with eval and C<#line> now works.
1662 Fixed numerous memory leaks, especially in eval "".
1666 mkdir() now ignores trailing slashes in the directory name,
1667 as mandated by POSIX.
1671 Some versions of glibc have a broken modfl(). This affects builds
1672 with C<-Duselongdouble>. This version of Perl detects this brokenness
1673 and has a workaround for it. The glibc release 2.2.2 is known to have
1674 fixed the modfl() bug.
1678 Modulus of unsigned numbers now works (4063328477 % 65535 used to
1679 return 27406, instead of 27047).
1683 Some "not a number" warnings introduced in 5.6.0 eliminated to be
1684 more compatible with 5.005. Infinity is now recognised as a number.
1688 Numeric conversions did not recognize changes in the string value
1689 properly in certain circumstances.
1693 Attributes (like :shared) didn't work with our().
1697 our() variables will not cause "will not stay shared" warnings.
1701 "our" variables of the same name declared in two sibling blocks
1702 resulted in bogus warnings about "redeclaration" of the variables.
1703 The problem has been corrected.
1707 pack "Z" now correctly terminates the string with "\0".
1711 Fix password routines which in some shadow password platforms
1712 (e.g. HP-UX) caused getpwent() to return every other entry.
1716 The PERL5OPT environment variable (for passing command line arguments
1717 to Perl) didn't work for more than a single group of options.
1721 PERL5OPT with embedded spaces didn't work.
1725 printf() no longer resets the numeric locale to "C".
1729 C<qw(a\\b)> now parses correctly as C<'a\\b'>.
1733 pos() did not return the correct value within s///ge in earlier
1734 versions. This is now handled correctly.
1738 Printing quads (64-bit integers) with printf/sprintf now works
1739 without the q L ll prefixes (assuming you are on a quad-capable platform).
1743 Regular expressions on references and overloaded scalars now work.
1747 Right-hand side magic (GMAGIC) could in many cases such as string
1748 concatenation be invoked too many times.
1752 scalar() now forces scalar context even when used in void context.
1756 SOCKS support is now much more robust.
1760 sort() arguments are now compiled in the right wantarray context
1761 (they were accidentally using the context of the sort() itself).
1762 The comparison block is now run in scalar context, and the arguments
1763 to be sorted are always provided list context.
1767 Changed the POSIX character class C<[[:space:]]> to include the (very
1768 rarely used) vertical tab character. Added a new POSIX-ish character
1769 class C<[[:blank:]]> which stands for horizontal whitespace
1770 (currently, the space and the tab).
1774 The tainting behaviour of sprintf() has been rationalized. It does
1775 not taint the result of floating point formats anymore, making the
1776 behaviour consistent with that of string interpolation.
1780 Some cases of inconsistent taint propagation (such as within hash
1781 values) have been fixed.
1785 The RE engine found in Perl 5.6.0 accidentally pessimised certain kinds
1786 of simple pattern matches. These are now handled better.
1790 Regular expression debug output (whether through C<use re 'debug'>
1791 or via C<-Dr>) now looks better.
1795 Multi-line matches like C<"a\nxb\n" =~ /(?!\A)x/m> were flawed. The
1800 Use of $& could trigger a core dump under some situations. This
1805 The regular expression captured submatches ($1, $2, ...) are now
1806 more consistently unset if the match fails, instead of leaving false
1807 data lying around in them.
1811 readline() on files opened in "slurp" mode could return an extra "" at
1812 the end in certain situations. This has been corrected.
1816 Autovivification of symbolic references of special variables described
1817 in L<perlvar> (as in C<${$num}>) was accidentally disabled. This works
1822 Sys::Syslog ignored the C<LOG_AUTH> constant.
1826 All but the first argument of the IO syswrite() method are now optional.
1830 $AUTOLOAD, sort(), lock(), and spawning subprocesses
1831 in multiple threads simultaneously are now thread-safe.
1835 Tie::ARRAY SPLICE method was broken.
1839 Allow read-only string on left hand side of non-modifying tr///.
1843 Several Unicode fixes.
1849 BOMs (byte order marks) in the beginning of Perl files
1850 (scripts, modules) should now be transparently skipped.
1851 UTF-16 (UCS-2) encoded Perl files should now be read correctly.
1855 The character tables have been updated to Unicode 3.1.1.
1859 Comparing with utf8 data does not magically upgrade non-utf8 data
1864 C<IsAlnum>, C<IsAlpha>, and C<IsWord> now match titlecase.
1868 Concatenation with the C<.> operator or via variable interpolation,
1869 C<eq>, C<substr>, C<reverse>, C<quotemeta>, the C<x> operator,
1870 substitution with C<s///>, single-quoted UTF8, should now work.
1874 The C<tr///> operator now works. Note that the C<tr///CU>
1875 functionality has been removed (but see pack('U0', ...)).
1879 C<eval "v200"> now works.
1883 Perl 5.6.0 parsed m/\x{ab}/ incorrectly, leading to spurious warnings.
1884 This has been corrected.
1888 Zero entries were missing from the Unicode classes like C<IsDigit>.
1894 Large unsigned numbers (those above 2**31) could sometimes lose their
1895 unsignedness, causing bogus results in arithmetic operations.
1899 =head2 Platform Specific Changes and Fixes
1907 Perl now works on post-4.0 BSD/OSes.
1913 Setting C<$0> now works (as much as possible; see L<perlvar> for details).
1919 Numerous updates; currently synchronised with Cygwin 1.1.4.
1923 Previously DYNIX/ptx had problems in its Configure probe for non-blocking I/O.
1929 EPOC update after Perl 5.6.0. See README.epoc.
1935 Perl now works on post-3.0 FreeBSDs.
1941 README.hpux updated; C<Configure -Duse64bitall> now almost works.
1947 Numerous compilation flag and hint enhancements; accidental mixing
1948 of 32-bit and 64-bit libraries (a doomed attempt) made much harder.
1958 Long doubles should now work (see INSTALL).
1962 Linux previously had problems related to sockaddrlen when using
1963 accept(), revcfrom() (in Perl: recv()), getpeername(), and getsockname().
1971 Compilation of the standard Perl distribution in MacOS Classic should
1972 now work if you have the Metrowerks development environment and
1973 the missing Mac-specific toolkit bits. Contact the macperl mailing
1980 MPE/iX update after Perl 5.6.0. See README.mpeix.
1986 Perl now works on NetBSD/sparc.
1992 Now works with usethreads (see INSTALL).
1998 64-bitness using the Sun Workshop compiler now works.
2002 Tru64 (aka Digital UNIX, aka DEC OSF/1)
2004 The operating system version letter now recorded in $Config{osvers}.
2005 Allow compiling with gcc (previously explicitly forbidden). Compiling
2006 with gcc still not recommended because buggy code results, even with
2013 Fixed various alignment problems that lead into core dumps either
2014 during build or later; no longer dies on math errors at runtime;
2015 now using full quad integers (64 bits), previously was using
2016 only 46 bit integers for speed.
2022 chdir() now works better despite a CRT bug; now works with MULTIPLICITY
2023 (see INSTALL); now works with Perl's malloc.
2025 The tainting of C<%ENV> elements via C<keys> or C<values> was previously
2026 unimplemented. It now works as documented.
2028 The C<waitpid> emulation has been improved. The worst bug (now fixed)
2029 was that a pid of -1 would cause a wildcard search of all processes on
2030 the system. The most significant enhancement is that we can now
2031 usually get the completion status of a terminated process.
2033 POSIX-style signals are now emulated much better on VMS versions prior
2036 The C<system> function and backticks operator have improved
2037 functionality and better error handling.
2047 accept() no longer leaks memory.
2051 Borland C++ v5.5 is now a supported compiler that can build Perl.
2052 However, the generated binaries continue to be incompatible with those
2053 generated by the other supported compilers (GCC and Visual C++).
2057 Better chdir() return value for a non-existent directory.
2061 Duping socket handles with open(F, ">&MYSOCK") now works under Windows 9x.
2065 New %ENV entries now propagate to subprocesses.
2069 Current directory entries in %ENV are now correctly propagated to child
2074 $ENV{LIB} now used to search for libs under Visual C.
2078 fork() emulation has been improved in various ways, but still continues
2079 to be experimental. See L<perlfork> for known bugs and caveats.
2083 A failed (pseudo)fork now returns undef and sets errno to EAGAIN.
2087 Win32::GetCwd() correctly returns C:\ instead of C: when at the drive root.
2088 Other bugs in chdir() and Cwd::cwd() have also been fixed.
2092 HTML files will be installed in c:\perl\html instead of c:\perl\lib\pod\html
2096 The makefiles now provide a single switch to bulk-enable all the features
2097 enabled in ActiveState ActivePerl (a popular Win32 binary distribution).
2101 Allow REG_EXPAND_SZ keys in the registry.
2105 Can now send() from all threads, not just the first one.
2109 Fake signal handling reenabled, bugs and all.
2113 %SIG has been enabled under USE_ITHREADS, but its use is completely
2114 unsupported under all configurations.
2118 Less stack reserved per thread so that more threads can run
2119 concurrently. (Still 16M per thread.)
2123 C<File::Spec->tmpdir()> now prefers C:/temp over /tmp
2124 (works better when perl is running as service).
2128 Better UNC path handling under ithreads.
2132 wait(), waitpid() and backticks now return the correct exit status under
2137 winsock handle leak fixed.
2143 =head1 New or Changed Diagnostics
2149 All regular expression compilation error messages are now hopefully
2150 easier to understand both because the error message now comes before
2151 the failed regex and because the point of failure is now clearly
2152 marked by a C<E<lt>-- HERE> marker.
2156 The various "opened only for", "on closed", "never opened" warnings
2157 drop the C<main::> prefix for filehandles in the C<main> package,
2158 for example C<STDIN> instead of C<main::STDIN>.
2162 The "Unrecognized escape" warning has been extended to include C<\8>,
2163 C<\9>, and C<\_>. There is no need to escape any of the C<\w> characters.
2167 Two new debugging options have been added: if you have compiled your
2168 Perl with debugging, you can use the -DT and -DR options to trace
2169 tokenising and to add reference counts to displaying variables,
2174 If an attempt to use a (non-blessed) reference as an array index
2175 is made, a warning is given.
2179 C<push @a;> and C<unshift @a;> (with no values to push or unshift)
2180 now give a warning. This may be a problem for generated and evaled
2185 If you try to L<perlfunc/pack> a number less than 0 or larger than 255
2186 using the C<"C"> format you will get an optional warning. Similarly
2187 for the C<"c"> format and a number less than -128 or more than 127.
2191 Certain regex modifiers such as C<(?o)> make sense only if applied to
2192 the entire regex. You will an optional warning if you try to do otherwise.
2196 Using arrays or hashes as references (e.g. C<< %foo->{bar} >>
2197 has been deprecated for a while. Now you will get an optional warning.
2201 =head1 Changed Internals
2207 perlapi.pod (a companion to perlguts) now attempts to document the
2212 You can now build a really minimal perl called microperl.
2213 Building microperl does not require even running Configure;
2214 C<make -f Makefile.micro> should be enough. Beware: microperl makes
2215 many assumptions, some of which may be too bold; the resulting
2216 executable may crash or otherwise misbehave in wondrous ways.
2217 For careful hackers only.
2221 Added rsignal(), whichsig(), do_join(), op_clear, op_null,
2222 ptr_table_clear(), ptr_table_free(), sv_setref_uv(), and several UTF-8
2223 interfaces to the publicised API. For the full list of the available
2224 APIs see L<perlapi>.
2228 Made possible to propagate customised exceptions via croak()ing.
2232 Now xsubs can have attributes just like subs. (Well, at least the
2233 built-in attributes.)
2237 dTHR and djSP have been obsoleted; the former removed (because it's
2238 a no-op) and the latter replaced with dSP.
2242 PERL_OBJECT has been completely removed.
2246 The MAGIC constants (e.g. C<'P'>) have been macrofied
2247 (e.g. C<PERL_MAGIC_TIED>) for better source code readability
2248 and maintainability.
2252 The regex compiler now maintains a structure that identifies nodes in
2253 the compiled bytecode with the corresponding syntactic features of the
2254 original regex expression. The information is attached to the new
2255 C<offsets> member of the C<struct regexp>. See L<perldebguts> for more
2256 complete information.
2260 The C code has been made much more C<gcc -Wall> clean. Some warning
2261 messages still remain in some platforms, so if you are compiling with
2262 gcc you may see some warnings about dubious practices. The warnings
2263 are being worked on.
2267 F<perly.c>, F<sv.c>, and F<sv.h> have now been extensively commented.
2271 Documentation on how to use the Perl source repository has been added
2272 to F<Porting/repository.pod>.
2276 There are now several profiling make targets.
2280 =head1 Security Vulnerability Closed
2282 (This change was already made in 5.7.0 but bears repeating here.)
2284 A potential security vulnerability in the optional suidperl component
2285 of Perl was identified in August 2000. suidperl is neither built nor
2286 installed by default. As of November 2001 the only known vulnerable
2287 platform is Linux, most likely all Linux distributions. CERT and
2288 various vendors and distributors have been alerted about the vulnerability.
2289 See http://www.cpan.org/src/5.0/sperl-2000-08-05/sperl-2000-08-05.txt
2290 for more information.
2292 The problem was caused by Perl trying to report a suspected security
2293 exploit attempt using an external program, /bin/mail. On Linux
2294 platforms the /bin/mail program had an undocumented feature which
2295 when combined with suidperl gave access to a root shell, resulting in
2296 a serious compromise instead of reporting the exploit attempt. If you
2297 don't have /bin/mail, or if you have 'safe setuid scripts', or if
2298 suidperl is not installed, you are safe.
2300 The exploit attempt reporting feature has been completely removed from
2301 Perl 5.8.0 (and the maintenance release 5.6.1, and it was removed also
2302 from all the Perl 5.7 releases), so that particular vulnerability
2303 isn't there anymore. However, further security vulnerabilities are,
2304 unfortunately, always possible. The suidperl functionality is most
2305 probably going to be removed in Perl 5.10. In any case, suidperl
2306 should only be used by security experts who know exactly what they are
2307 doing and why they are using suidperl instead of some other solution
2308 such as sudo (see http://www.courtesan.com/sudo/).
2312 Several new tests have been added, especially for the F<lib>
2313 subsection. There are now about 34 000 individual tests (spread over
2314 about 530 test scripts), in the regression suite (5.6.1 has about
2315 11700 tests, in 258 test scripts) Many of the new tests are introduced
2316 by the new modules, but still in general Perl is now more thoroughly
2319 Because of the large number of tests, running the regression suite
2320 will take considerably longer time than it used to: expect the suite
2321 to take up to 4-5 times longer to run than in perl 5.6. In a really
2322 fast machine you can hope to finish the suite in about 5 minutes
2325 The tests are now reported in a different order than in earlier Perls.
2326 (This happens because the test scripts from under t/lib have been moved
2327 to be closer to the library/extension they are testing.)
2329 =head1 Known Problems
2337 In AIX 4.2 Perl extensions that use C++ functions that use statics
2338 may have problems in that the statics are not getting initialized.
2339 In newer AIX releases this has been solved by linking Perl with
2340 the libC_r library, but unfortunately in AIX 4.2 the said library
2341 has an obscure bug where the various functions related to time
2342 (such as time() and gettimeofday()) return broken values, and
2343 therefore in AIX 4.2 Perl is not linked against the libC_r.
2347 vac 5.0.0.0 May Produce Buggy Code For Perl
2349 The AIX C compiler vac version 5.0.0.0 may produce buggy code,
2350 resulting in few random tests failing, but when the failing tests
2351 are run by hand, they succeed. We suggest upgrading to at least
2352 vac version 5.0.1.0, that has been known to compile Perl correctly.
2353 "lslpp -L|grep vac.C" will tell you the vac version.
2357 =head2 Amiga Perl Invoking Mystery
2359 One cannot call Perl using the C<volume:> syntax, that is, C<perl -v>
2360 works, but for example C<bin:perl -v> doesn't. The exact reason isn't
2361 known but the current suspect is the F<ixemul> library.
2363 =head2 lib/ftmp-security tests warn 'system possibly insecure'
2365 Don't panic. Read INSTALL 'make test' section instead.
2367 =head2 Cygwin intermittent failures of lib/Memoize/t/expire_file 11 and 12
2369 The subtests 11 and 12 sometimes fail and sometimes work.
2371 =head2 HP-UX lib/io_multihomed Fails When LP64-Configured
2373 The lib/io_multihomed test may hang in HP-UX if Perl has been
2374 configured to be 64-bit. Because other 64-bit platforms do not hang in
2375 this test, HP-UX is suspect. All other tests pass in 64-bit HP-UX. The
2376 test attempts to create and connect to "multihomed" sockets (sockets
2377 which have multiple IP addresses).
2379 =head2 HP-UX lib/posix Subtest 9 Fails When LP64-Configured
2381 If perl is configured with -Duse64bitall, the successful result of the
2382 subtest 10 of lib/posix may arrive before the successful result of the
2383 subtest 9, which confuses the test harness so much that it thinks the
2386 =head2 Linux With Sfio Fails op/misc Test 48
2392 The following tests are known to fail:
2394 Failed Test Stat Wstat Total Fail Failed List of Failed
2395 -------------------------------------------------------------------------
2396 ../ext/DB_File/t/db-btree.t 0 11 ?? ?? % ??
2397 ../ext/DB_File/t/db-recno.t 149 3 2.01% 61 63 65
2398 ../ext/POSIX/t/posix.t 31 1 3.23% 10
2399 ../lib/warnings.t 450 1 0.22% 316
2403 OS/390 has rather many test failures but the situation is actually
2404 better than it was in 5.6.0, it's just that so many new modules and
2405 tests have been added.
2407 Failed Test Stat Wstat Total Fail Failed List of Failed
2408 -----------------------------------------------------------------------------
2409 ../ext/B/Deparse.t 14 1 7.14% 14
2410 ../ext/B/Showlex.t 1 1 100.00% 1
2411 ../ext/Encode/Encode/Tcl.t 610 13 2.13% 592 594 596 598
2413 ../ext/IO/lib/IO/t/io_unix.t 113 28928 5 3 60.00% 3-5
2414 ../ext/POSIX/POSIX.t 29 1 3.45% 14
2415 ../ext/Storable/t/lock.t 255 65280 5 3 60.00% 3-5
2416 ../lib/locale.t 129 33024 117 19 16.24% 99-117
2417 ../lib/warnings.t 434 1 0.23% 75
2418 ../lib/ExtUtils.t 27 1 3.70% 25
2419 ../lib/Math/BigInt/t/bigintpm.t 1190 1 0.08% 1145
2420 ../lib/Unicode/UCD.t 81 48 59.26% 1-16 49-64 66-81
2421 ../lib/User/pwent.t 9 1 11.11% 4
2422 op/pat.t 660 6 0.91% 242-243 424-425
2424 op/split.t 0 9 ?? ?? % ??
2425 op/taint.t 174 3 1.72% 156 162 168
2426 op/tr.t 70 3 4.29% 50 58-59
2427 Failed 16/422 test scripts, 96.21% okay. 105/23251 subtests failed, 99.55% okay.
2429 =head2 op/sprintf tests 129 and 130
2431 The op/sprintf tests 129 and 130 are known to fail on some platforms.
2432 Examples include any platform using sfio, and Compaq/Tandem's NonStop-UX.
2433 The failing platforms do not comply with the ANSI C Standard, line
2434 19ff on page 134 of ANSI X3.159 1989 to be exact. (They produce
2435 something other than "1" and "-1" when formatting 0.6 and -0.6 using
2436 the printf format "%.0f", most often they produce "0" and "-0".)
2438 =head2 Failure of Thread tests
2440 B<Note that support for 5.005-style threading remains experimental
2441 and practically unsupported.>
2443 The following tests are known to fail due to fundamental problems in
2444 the 5.005 threading implementation. These are not new failures--Perl
2445 5.005_0x has the same bugs, but didn't have these tests.
2447 ext/List/Util/t/first 2
2449 ext/Thread/thr5005 19-20
2451 These failures are unlikely to get fixed.
2459 ext/POSIX/sigaction subtests 6 and 13 may fail.
2463 lib/ExtUtils may spuriously claim that subtest 28 failed,
2464 which is interesting since the test only has 27 tests.
2468 Numerous numerical test failures
2470 op/numconvert 209,210,217,218
2472 ext/Time/HiRes/HiRes 9
2473 lib/Math/BigInt/t/bigintpm 1145
2476 These tests fail because of yet unresolved floating point inaccuracies.
2482 There are a few known test failures, see L<perluts>.
2486 There is one known test failure with a default configuration:
2488 [.run]switches..........................FAILED on test 1
2492 In multi-CPU boxes there are some problems with the I/O buffering:
2493 some output may appear twice.
2495 =head2 Localising a Tied Variable Leaks Memory
2498 tie my %tie_hash => 'Tie::StdHash';
2502 local($tie_hash{Foo}) = 1; # leaks
2504 Code like the above is known to leak memory every time the local()
2507 =head2 Localising Tied Arrays and Hashes Is Broken
2511 doesn't work as one would expect: the old value is restored
2514 =head2 Self-tying of Arrays and Hashes Is Forbidden
2516 Self-tying of arrays and hashes is broken in rather deep and
2517 hard-to-fix ways. As a stop-gap measure to avoid people from getting
2518 frustrated at the mysterious results (core dumps, most often) it is
2519 for now forbidden (you will get a fatal error even from an attempt).
2521 =head2 Building Extensions Can Fail Because Of Largefiles
2523 Some extensions like mod_perl are known to have issues with
2524 `largefiles', a change brought by Perl 5.6.0 in which file offsets
2525 default to 64 bits wide, where supported. Modules may fail to compile
2526 at all or compile and work incorrectly. Currently there is no good
2527 solution for the problem, but Configure now provides appropriate
2528 non-largefile ccflags, ldflags, libswanted, and libs in the %Config
2529 hash (e.g., $Config{ccflags_nolargefiles}) so the extensions that are
2530 having problems can try configuring themselves without the
2531 largefileness. This is admittedly not a clean solution, and the
2532 solution may not even work at all. One potential failure is whether
2533 one can (or, if one can, whether it's a good idea) link together at
2534 all binaries with different ideas about file offsets, all this is
2537 =head2 Unicode Support on EBCDIC Still Spotty
2539 Though mostly working, Unicode support still has problem spots on
2540 EBCDIC platforms. One such known spot are the C<\p{}> and C<\P{}>
2541 regular expression constructs for code points less than 256: the
2542 pP are testing for Unicode code points, not knowing about EBCDIC.
2544 =head2 The Compiler Suite Is Still Experimental
2546 The compiler suite is slowly getting better but it continues to be
2547 highly experimental. Use in production environments is discouraged.
2549 =head2 The Long Double Support is Still Experimental
2551 The ability to configure Perl's numbers to use "long doubles",
2552 floating point numbers of hopefully better accuracy, is still
2553 experimental. The implementations of long doubles are not yet
2554 widespread and the existing implementations are not quite mature
2555 or standardised, therefore trying to support them is a rare
2556 and moving target. The gain of more precision may also be offset
2557 by slowdown in computations (more bits to move around, and the
2558 operations are more likely to be executed by less optimised
2561 =head1 Reporting Bugs
2563 If you find what you think is a bug, you might check the articles
2564 recently posted to the comp.lang.perl.misc newsgroup and the perl
2565 bug database at http://bugs.perl.org. There may also be
2566 information at http://www.perl.com/perl/, the Perl Home Page.
2568 If you believe you have an unreported bug, please run the B<perlbug>
2569 program included with your release. Be sure to trim your bug down
2570 to a tiny but sufficient test case. Your bug report, along with the
2571 output of C<perl -V>, will be sent off to perlbug@perl.org to be
2572 analysed by the Perl porting team.
2576 The F<Changes> file for exhaustive details on what changed.
2578 The F<INSTALL> file for how to build Perl.
2580 The F<README> file for general stuff.
2582 The F<Artistic> and F<Copying> files for copyright information.
2586 Written by Jarkko Hietaniemi <F<jhi@iki.fi>>.