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 Binary Incompatibility
51 B<Perl 5.8 is not binary compatible with earlier releases of Perl.>
53 B<You have to recompile your XS modules.>
55 (Pure Perl modules should continue to work.)
57 The major reason for the discontinity is the new IO architecture
58 called PerlIO. PerlIO is the default configuration because
59 without it many new features of Perl 5.8 cannot be used. In other
60 words: you just have to recompile your modules, sorry about that.
62 In future releases of Perl, non-PerlIO aware XS modules may become
63 completely unsupported. This shouldn't be too difficult for module
64 authors, however: PerlIO has been designed as a drop-in replacement
65 (at the source code level) for the stdio interface.
67 Depending on your platform, there are also other reasons why
68 we decided to break binary compatibility, please read on.
70 =head2 64-bit platforms and malloc
72 If your pointers are 64 bits wide, the Perl malloc is no longer being
73 used because it does not work well with 8-byte pointers. Also,
74 usually the system mallocs on such platforms are much better optimized
75 for such large memory models than the Perl malloc. Some memory-hungry
76 Perl applications like the PDL don't work well with Perl's malloc.
77 Finally, other applications than Perl (like modperl) tend to prefer
78 the system malloc. Such platforms include Alpha and 64-bit HPPA,
81 =head2 AIX Dynaloading
83 The AIX dynaloading now uses in AIX releases 4.3 and newer the native
84 dlopen interface of AIX instead of the old emulated interface. This
85 change will probably break backward compatibility with compiled
86 modules. The change was made to make Perl more compliant with other
87 applications like modperl which are using the AIX native interface.
89 =head2 Attributes for C<my> variables now handled at run-time.
91 The C<my EXPR : ATTRS> syntax now applies variable attributes at
92 run-time. (Subroutine and C<our> variables still get attributes applied
93 at compile-time.) See L<attributes> for additional details. In particular,
94 however, this allows variable attributes to be useful for C<tie> interfaces,
95 which was a deficiency of earlier releases. Note that the new semantics
96 doesn't work with the Attribute::Handlers module (as of version 0.76).
98 =head2 Socket Extension Dynamic in VMS
100 The Socket extension is now dynamically loaded instead of being
101 statically built in. This may or may not be a problem with ancient
102 TCP/IP stacks of VMS: we do not know since we weren't able to test
103 Perl in such configurations.
105 =head2 IEEE-format Floating Point Default on OpenVMS Alpha
107 Perl now uses IEEE format (T_FLOAT) as the default internal floating
108 point format on OpenVMS Alpha, potentially breaking binary compatibility
109 with external libraries or existing data. G_FLOAT is still available as
110 a configuration option. The default on VAX (D_FLOAT) has not changed.
112 =head2 New Unicode Properties
114 Unicode I<scripts> are now supported. Scripts are similar to (and superior
115 to) Unicode I<blocks>. The difference between scripts and blocks is that
116 scripts are the glyphs used by a language or a group of languages, while
117 the blocks are more artificial groupings of (mostly) 256 characters based
118 on the Unicode numbering.
120 In general, scripts are more inclusive, but not universally so. For
121 example, while the script C<Latin> includes all the Latin characters and
122 their various diacritic-adorned versions, it does not include the various
123 punctuation or digits (since they are not solely C<Latin>).
125 A number of other properties are now supported, including C<\p{L&}>,
126 C<\p{Any}> C<\p{Assigned}>, C<\p{Unassigned}>, C<\p{Blank}> and
127 C<\p{SpacePerl}> (along with their C<\P{...}> versions, of course).
128 See L<perlunicode> for details, and more additions.
130 The C<In> or C<Is> prefix to names used with the C<\p{...}> and C<\P{...}>
131 are now almost always optional. The only exception is that a C<In> prefix
132 is required to signify a Unicode block when a block name conflicts with a
133 script name. For example, C<\p{Tibetan}> refers to the script, while
134 C<\p{InTibetan}> refers to the block. When there is no name conflict, you
135 can omit the C<In> from the block name (e.g. C<\p{BraillePatterns}>), but
136 to be safe, it's probably best to always use the C<In>).
138 =head2 REF(...) Instead Of SCALAR(...)
140 A reference to a reference now stringifies as "REF(0x81485ec)" instead
141 of "SCALAR(0x81485ec)" in order to be more consistent with the return
144 =head2 pack/unpack D/F recycled
146 The undocumented pack/unpack template letters D/F have been recycled
147 for better use: now they stand for long double (if supported by the
148 platform) and NV (Perl internal floating point type). (They used
149 to be aliases for d/f, but you never knew that.)
157 The semantics of bless(REF, REF) were unclear and until someone proves
158 it to make some sense, it is forbidden.
162 The obsolete chat2 library that should never have been allowed
163 to escape the laboratory has been decommissioned.
167 The builtin dump() function has probably outlived most of its
168 usefulness. The core-dumping functionality will remain in future
169 available as an explicit call to C<CORE::dump()>, but in future
170 releases the behaviour of an unqualified C<dump()> call may change.
174 The very dusty examples in the eg/ directory have been removed.
175 Suggestions for new shiny examples welcome but the main issue is that
176 the examples need to be documented, tested and (most importantly)
181 The (bogus) escape sequences \8 and \9 now give an optional warning
182 ("Unrecognized escape passed through"). There is no need to \-escape
187 The list of filenames from glob() (or <...>) is now by default sorted
188 alphabetically to be csh-compliant (which is what happened before
189 in most UNIX platforms). (bsd_glob() does still sort platform
190 natively, ASCII or EBCDIC, unless GLOB_ALPHASORT is specified.)
194 Spurious syntax errors generated in certain situations, when glob()
195 caused File::Glob to be loaded for the first time, have been fixed.
199 Although "you shouldn't do that", it was possible to write code that
200 depends on Perl's hashed key order (Data::Dumper does this). The new
201 algorithm "One-at-a-Time" produces a different hashed key order.
202 More details are in L</"Performance Enhancements">.
206 lstat(FILEHANDLE) now gives a warning because the operation makes no sense.
207 In future releases this may become a fatal error.
211 The C<package;> syntax (C<package> without an argument) has been
212 deprecated. Its semantics were never that clear and its
213 implementation even less so. If you have used that feature to
214 disallow all but fully qualified variables, C<use strict;> instead.
218 The unimplemented POSIX regex features [[.cc.]] and [[=c=]] are still
219 recognised but now cause fatal errors. The previous behaviour of
220 ignoring them by default and warning if requested was unacceptable
221 since it, in a way, falsely promised that the features could be used.
225 The current user-visible implementation of pseudo-hashes (the weird
226 use of the first array element) is deprecated starting from Perl 5.8.0
227 and will be removed in Perl 5.10.0, and the feature will be
228 implemented differently. Not only is the current interface rather
229 ugly, but the current implementation slows down normal array and hash
230 use quite noticeably. The C<fields> pragma interface will remain
231 available. The I<restricted hashes> interface is expected to
232 be the replacement interface (see L<Hash::Util>).
236 The syntaxes C<< @a->[...] >> and C<< %h->{...} >> have now been deprecated.
240 After years of trying the suidperl is considered to be too complex to
241 ever be considered truly secure. The suidperl functionality is likely
242 to be removed in a future release.
246 The 5.005 threads model (module C<Thread>) is deprecated and expected
247 to be removed in Perl 5.10. Multithreaded code should be migrated to
248 the new ithreads model (see L<threads> and L<threads::shared>).
252 The long deprecated uppercase aliases for the string comparison
253 operators (EQ, NE, LT, LE, GE, GT) have now been removed.
257 The tr///C and tr///U features have been removed and will not return;
258 the interface was a mistake. Sorry about that. For similar
259 functionality, see pack('U0', ...) and pack('C0', ...).
263 Earlier Perls treated "sub foo (@bar)" as equivalent to "sub foo (@)".
264 The prototypes are now checked at compile-time for invalid characters.
265 An optional warning is generated ("Illegal character in prototype...")
266 but this may be upgraded to a fatal error in a future release.
270 =head1 Core Enhancements
272 =head2 PerlIO is Now The Default
278 IO is now by default done via PerlIO rather than system's "stdio".
279 PerlIO allows "layers" to be "pushed" onto a file handle to alter the
280 handle's behaviour. Layers can be specified at open time via 3-arg
283 open($fh,'>:crlf :utf8', $path) || ...
285 or on already opened handles via extended C<binmode>:
287 binmode($fh,':encoding(iso-8859-7)');
289 The built-in layers are: unix (low level read/write), stdio (as in
290 previous Perls), perlio (re-implementation of stdio buffering in a
291 portable manner), crlf (does CRLF <=> "\n" translation as on Win32,
292 but available on any platform). A mmap layer may be available if
293 platform supports it (mostly UNIXes).
295 Layers to be applied by default may be specified via the 'open' pragma.
297 See L</"Installation and Configuration Improvements"> for the effects
298 of PerlIO on your architecture name.
302 File handles can be marked as accepting Perl's internal encoding of Unicode
303 (UTF-8 or UTF-EBCDIC depending on platform) by a pseudo layer ":utf8" :
305 open($fh,">:utf8","Uni.txt");
307 Note for EBCDIC users: the pseudo layer ":utf8" is erroneously named
308 for you since it's not UTF-8 what you will be getting but instead
309 UTF-EBCDIC. See L<perlunicode>, L<utf8>, and
310 http://www.unicode.org/unicode/reports/tr16/ for more information.
311 In future releases this naming may change.
315 File handles can translate character encodings from/to Perl's internal
316 Unicode form on read/write via the ":encoding()" layer.
320 File handles can be opened to "in memory" files held in Perl scalars via:
322 open($fh,'>', \$variable) || ...
326 Anonymous temporary files are available without need to
327 'use FileHandle' or other module via
329 open($fh,"+>", undef) || ...
331 That is a literal undef, not an undefined value.
335 The list form of C<open> is now implemented for pipes (at least on UNIX):
337 open($fh,"-|", 'cat', '/etc/motd')
339 creates a pipe, and runs the equivalent of exec('cat', '/etc/motd') in
344 If your locale environment variables (LANGUAGE, LC_ALL, LC_CTYPE, LANG)
345 contain the strings 'UTF-8' or 'UTF8' (case-insensitive matching),
346 the default encoding of your STDIN, STDOUT, and STDERR, and of
347 B<any subsequent file open>, is UTF-8.
351 =head2 Restricted Hashes
353 A restricted hash is restricted to a certain set of keys, no keys
354 outside the set can be added. Also individual keys can be restricted
355 so that the key cannot be deleted and the value cannot be changed.
356 No new syntax is involved: the Hash::Util module is the interface.
360 Perl used to be fragile in that signals arriving at inopportune moments
361 could corrupt Perl's internal state. Now Perl postpones handling of
362 signals until it's safe (between opcodes).
364 This change may have surprising side effects because signals no longer
365 interrupt Perl instantly. Perl will now first finish whatever it was
366 doing, like finishing an internal operation (like sort()) or an
367 external operation (like an I/O operation), and only then look at any
368 arrived signals (and before starting the next operation). No more corrupt
369 internal state since the current operation is always finished first,
370 but the signal may take more time to get heard. Note that breaking
371 out from potentially blocking operations should still work, though.
373 =head2 Unicode Overhaul
375 Unicode in general should be now much more usable than in Perl 5.6.0
376 (or even in 5.6.1). Unicode can be used in hash keys, Unicode in
377 regular expressions should work now, Unicode in tr/// should work now,
378 Unicode in I/O should work now. See L<perluniintro> for introduction
379 and L<perlunicode> for details.
385 The Unicode Character Database coming with Perl has been upgraded
386 to Unicode 3.2.0. For more information, see http://www.unicode.org/ .
390 For developers interested in enhancing Perl's Unicode capabilities:
391 almost all the UCD files are included with the Perl distribution in
392 the F<lib/unicore subdirectory>. The most notable omission, for space
393 considerations, is the Unihan database.
397 The properties \p{Blank} and \p{SpacePerl} have been added. "Blank" is like
398 C isblank(), that is, it contains only "horizontal whitespace" (the space
399 character is, the newline isn't), and the "SpacePerl" is the Unicode
400 equivalent of C<\s> (\p{Space} isn't, since that includes the vertical
401 tabulator character, whereas C<\s> doesn't.)
403 See "New Unicode Properties" earlier in this document for additional
404 information on changes with Unicode properties.
408 =head2 Understanding of Numbers
410 In general a lot of fixing has happened in the area of Perl's
411 understanding of numbers, both integer and floating point. Since in
412 many systems the standard number parsing functions like C<strtoul()>
413 and C<atof()> seem to have bugs, Perl tries to work around their
414 deficiencies. This results hopefully in more accurate numbers.
416 Perl now tries internally to use integer values in numeric conversions
417 and basic arithmetics (+ - * /) if the arguments are integers, and
418 tries also to keep the results stored internally as integers.
419 This change leads to often slightly faster and always less lossy
420 arithmetics. (Previously Perl always preferred floating point numbers
423 =head2 Miscellaneous Changes
429 AUTOLOAD is now lvaluable, meaning that you can add the :lvalue attribute
430 to AUTOLOAD subroutines and you can assign to the AUTOLOAD return value.
434 C<perl -d:Module=arg,arg,arg> now works (previously one couldn't pass
435 in multiple arguments.)
439 The builtin dump() now gives an optional warning
440 C<dump() better written as CORE::dump()>,
441 meaning that by default C<dump(...)> is resolved as the builtin
442 dump() which dumps core and aborts, not as (possibly) user-defined
443 C<sub dump>. To call the latter, qualify the call as C<&dump(...)>.
444 (The whole dump() feature is to considered deprecated, and possibly
445 removed/changed in future releases.)
449 chomp() and chop() are now overridable. Note, however, that their
450 prototype (as given by C<prototype("CORE::chomp")> is undefined,
451 because it cannot be expressed and therefore one cannot really write
452 replacements to override these builtins.
456 END blocks are now run even if you exit/die in a BEGIN block.
457 Internally, the execution of END blocks is now controlled by
458 PL_exit_flags & PERL_EXIT_DESTRUCT_END. This enables the new
459 behaviour for Perl embedders. This will default in 5.10. See
464 Formats now support zero-padded decimal fields.
468 Lvalue subroutines can now return C<undef> in list context.
469 However, the lvalue subroutine feature still remains experimental.
473 A lost warning "Can't declare ... dereference in my" has been
474 restored (Perl had it earlier but it became lost in later releases.)
478 A new special regular expression variable has been introduced:
479 C<$^N>, which contains the most-recently closed group (submatch).
483 C<no Module;> now works even if there is no "sub unimport" in the Module.
487 The numerical comparison operators return C<undef> if either operand
488 is a NaN. Previously the behaviour was unspecified.
492 The following builtin functions are now overridable: each(), keys(),
493 pop(), push(), shift(), splice(), unshift().
497 C<pack() / unpack()> now can group template letters with C<()> and then
498 apply repetition/count modifiers on the groups.
502 C<pack() / unpack()> can now process the Perl internal numeric types:
503 IVs, UVs, NVs-- and also long doubles, if supported by the platform.
504 The template letters are C<j>, C<J>, C<F>, and C<D>.
508 C<pack('U0a*', ...)> can now be used to force a string to UTF8.
512 my __PACKAGE__ $obj now works.
516 The printf() and sprintf() now support parameter reordering using the
517 C<%\d+\$> and C<*\d+\$> syntaxes. For example
519 print "%2\$s %1\$s\n", "foo", "bar";
521 will print "bar foo\n". This feature helps in writing
522 internationalised software, and in general when the order
523 of the parameters can vary.
527 prototype(\&) is now available.
531 prototype(\[$@%&]) is now available to implicitly create references
532 (useful for example if you want to emulate the tie() interface).
536 A new command-line option, C<-t> is available. It is the
537 little brother of C<-T>: instead of dying on taint violations,
538 lexical warnings are given. B<This is only meant as a temporary
539 debugging aid while securing the code of old legacy applications.
540 This is not a substitute for -T.>
544 In other taint news, the C<exec LIST> and C<system LIST> have now been
545 considered too risky (think C<exec @ARGV>: it can start any program
546 with any arguments), and now the said forms cause a warning.
547 You should carefully launder the arguments to guarantee their
548 validity. In future releases of Perl the forms will become fatal
549 errors so consider starting laundering now.
553 Tied hash interfaces are now required to have the EXISTS and DELETE
554 methods (either own or inherited).
558 If tr/// is just counting characters, it doesn't attempt to
563 untie() will now call an UNTIE() hook if it exists. See L<perltie>
568 L<utime> now supports C<utime undef, undef, @files> to change the
569 file timestamps to the current time.
573 The rules for allowing underscores (underbars) in numeric constants
574 have been relaxed and simplified: now you can have an underscore
575 simply B<between digits>.
579 Rather than relying on C's argv[0] (which may not contain a full pathname)
580 where possible $^X is now set by asking the operating system.
581 (eg by reading F</proc/self/exe> on Linux, F</proc/curproc/file> on FreeBSD)
585 A new variable, C<${^TAINT}>, indicates whether taint mode is enabled.
589 You can now override the readline() builtin, and this overrides also
590 the <FILEHANDLE> angle bracket operator.
594 The command-line options -s and -F are now recognized on the shebang
599 Use of the C</c> match modifier without an accompanying C</g> modifier
600 elicits a new warning: C<Use of /c modifier is meaningless without /g>.
602 Use of C</c> in substitutions, even with C</g>, elicits
603 C<Use of /c modifier is meaningless in s///>.
605 Use of C</g> with C<split> elicits C<Use of /g modifier is meaningless
610 =head1 Modules and Pragmata
612 =head2 New Modules and Pragmata
618 C<Attribute::Handlers> allows a class to define attribute handlers.
621 use Attribute::Handlers;
622 sub Wolf :ATTR(SCALAR) { print "howl!\n" }
624 # later, in some package using or inheriting from MyPack...
626 my MyPack $Fluffy : Wolf; # the attribute handler Wolf will be called
628 Both variables and routines can have attribute handlers. Handlers can
629 be specific to type (SCALAR, ARRAY, HASH, or CODE), or specific to the
630 exact compilation phase (BEGIN, CHECK, INIT, or END).
634 B<B::Concise> is a new compiler backend for walking the Perl syntax
635 tree, printing concise info about ops, from Stephen McCamant. The
636 output is highly customisable. See L<B::Concise>.
640 The new bignum, bigint, and bigrat pragmas implement transparent
641 bignum support (using the Math::BigInt, Math::BigFloat, and
642 Math::BigRat backends), by Tels.
646 C<Class::ISA> for reporting the search path for a class's ISA tree,
647 by Sean Burke, has been added. See L<Class::ISA>.
651 C<Cwd> has now a split personality: if possible, an XS extension is
652 used, (this will hopefully be faster, more secure, and more robust)
653 but if not possible, the familiar Perl implementation is used.
657 C<Devel::PPPort>, originally from Kenneth Albanowski and now
658 maintained by Paul Marquess, has been added. It is primarily used
659 by C<h2xs> to enhance portability of XS modules between different
664 C<Digest>, frontend module for calculating digests (checksums), from
665 Gisle Aas, has been added. See L<Digest>.
669 C<Digest::MD5> for calculating MD5 digests (checksums) as defined in
670 RFC 1321, from Gisle Aas, has been added. See L<Digest::MD5>.
672 use Digest::MD5 'md5_hex';
674 $digest = md5_hex("Thirsty Camel");
676 print $digest, "\n"; # 01d19d9d2045e005c3f1b80e8b164de1
678 NOTE: the C<MD5> backward compatibility module is deliberately not
679 included since its further use is discouraged.
683 C<Encode>, orginally by Nick Ing-Simmons and now maintained by Dan
684 Kogai, provides a mechanism to translate between different character
685 encodings. Support for Unicode, ISO-8859-1, and ASCII are compiled in
686 to the module. Several other encodings (like the rest of the
687 ISO-8859, CP*/Win*, Mac, KOI8-R, three variants EBCDIC, Chinese,
688 Japanese, and Korean encodings) are included and can be loaded at
689 runtime. (For space considerations, the largest Chinese encodings
690 have been separated into their own CPAN module, Encode::HanExtra,
691 which Encode will use if available). See L<Encode>.
693 Any encoding supported by Encode module is also available to the
694 ":encoding()" layer if PerlIO is used.
698 C<Hash::Util> is the interface to the new I<restricted hashes>
699 feature. (Implemented by Jeffrey Friedl, Nick Ing-Simmons, and
704 C<I18N::Langinfo> can be use to query locale information.
705 See L<I18N::Langinfo>.
709 C<I18N::LangTags> has functions for dealing with RFC3066-style
710 language tags, by Sean Burke. See L<I18N::LangTags>.
714 C<ExtUtils::Constant> is a new tool for extension writers for
715 generating XS code to import C header constants, by Nicholas Clark.
716 See L<ExtUtils::Constant>.
720 C<Filter::Simple> is an easy-to-use frontend to Filter::Util::Call,
721 from Damian Conway. See L<Filter::Simple>.
727 use Filter::Simple sub {
728 while (my ($from, $to) = splice @_, 0, 2) {
737 use MyFilter qr/red/ => 'green';
739 print "red\n"; # this code is filtered, will print "green\n"
740 print "bored\n"; # this code is filtered, will print "bogreen\n"
744 print "red\n"; # this code is not filtered, will print "red\n"
748 C<File::Temp> allows one to create temporary files and directories in
749 an easy, portable, and secure way, by Tim Jenness. See L<File::Temp>.
753 C<Filter::Util::Call> provides you with the framework to write
754 I<Source Filters> in Perl, from Paul Marquess. For most uses the
755 frontend Filter::Simple is to be preferred. See L<Filter::Util::Call>.
759 C<if> is a new pragma for conditional inclusion of modules, from
764 L<libnet> is a collection of perl5 modules related to network
765 programming, from Graham Barr. See L<Net::FTP>, L<Net::NNTP>,
766 L<Net::Ping> (not part of libnet, but related), L<Net::POP3>,
767 L<Net::SMTP>, and L<Net::Time>.
769 Perl installation leaves libnet unconfigured, use F<libnetcfg> to configure.
773 C<List::Util> is a selection of general-utility list subroutines, like
774 sum(), min(), first(), and shuffle(), by Graham Barr. See L<List::Util>.
778 C<Locale::Constants>, C<Locale::Country>, C<Locale::Currency>
779 C<Locale::Language>, and L<Locale::Script>, from Neil Bowers, have
780 been added. They provide the codes for various locale standards, such
781 as "fr" for France, "usd" for US Dollar, and "ja" for Japanese.
785 $country = code2country('jp'); # $country gets 'Japan'
786 $code = country2code('Norway'); # $code gets 'no'
788 See L<Locale::Constants>, L<Locale::Country>, L<Locale::Currency>,
789 and L<Locale::Language>.
793 C<Locale::Maketext> is localization framework from Sean Burke. See
794 L<Locale::Maketext>, and L<Locale::Maketext::TPJ13>. The latter is an
795 article about software localization, originally published in The Perl
796 Journal #13, republished here with kind permission.
800 C<Math::BigRat> for big rational numbers, to accompany Math::BigInt and
801 Math::BigFloat, from Tels.
805 C<Memoize> can make your functions faster by trading space for time,
806 from Mark-Jason Dominus. See L<Memoize>.
810 C<MIME::Base64> allows you to encode data in base64, from Gisle Aas,
811 as defined in RFC 2045 - I<MIME (Multipurpose Internet Mail
816 $encoded = encode_base64('Aladdin:open sesame');
817 $decoded = decode_base64($encoded);
819 print $encoded, "\n"; # "QWxhZGRpbjpvcGVuIHNlc2FtZQ=="
825 C<MIME::QuotedPrint> allows you to encode data in quoted-printable
826 encoding, as defined in RFC 2045 - I<MIME (Multipurpose Internet Mail
827 Extensions)>, from Gisle Aas.
829 use MIME::QuotedPrint;
831 $encoded = encode_qp("Smiley in Unicode: \x{263a}");
832 $decoded = decode_qp($encoded);
834 print $encoded, "\n"; # "Smiley in Unicode: =263A"
836 MIME::QuotedPrint has been enhanced to provide the basic methods
837 necessary to use it with PerlIO::Via as in :
839 use MIME::QuotedPrint;
840 open($fh,">Via(MIME::QuotedPrint)",$path);
842 See L<MIME::QuotedPrint>.
846 C<NEXT> is pseudo-class for method redispatch, from Damian Conway.
851 C<open> is a new pragma for setting the default I/O disciplines
856 C<PerlIO::Scalar> provides the implementation of IO to "in memory"
857 Perl scalars as discussed above, from Nick Ing-Simmons. It also
858 serves as an example of a loadable PerlIO layer. Other future
859 possibilities include PerlIO::Array and PerlIO::Code.
860 See L<PerlIO::Scalar>.
864 C<PerlIO::Via> acts as a PerlIO layer and wraps PerlIO layer
865 functionality provided by a class (typically implemented in perl
866 code), from Nick Ing-Simmons.
868 use MIME::QuotedPrint;
869 open($fh,">Via(MIME::QuotedPrint)",$path);
871 This will automatically convert everything output to C<$fh>
872 to Quoted-Printable. See L<PerlIO::Via>.
876 C<Pod::ParseLink>, by Russ Allbery, has been added,
877 to parse LZ<><> links in pods as described in the new
882 C<Pod::Text::Overstrike>, by Joe Smith, has been added.
883 It converts POD data to formatted overstrike text.
884 See L<Pod::Text::Overstrike>.
888 C<Scalar::Util> is a selection of general-utility scalar subroutines,
889 like blessed(), reftype(), and tainted(). See L<Scalar::Util>.
893 C<sort> is a new pragma for controlling the behaviour of sort().
897 C<Storable> gives persistence to Perl data structures by allowing the
898 storage and retrieval of Perl data to and from files in a fast and
899 compact binary format. Because in effect Storable does serialisation
900 of Perl data structues, with it you can also clone deep, hierarchical
901 datastructures. Storable was created by Raphael Manfredi but it is
902 now maintained by the Perl development team. Storable has been
903 enhanced to understand the two new hash features, Unicode keys and
904 restricted hashes. See L<Storable>.
908 C<Switch>, from Damian Conway, has been added. Just by saying
912 you have C<switch> and C<case> available in Perl.
918 case 1 { print "number 1" }
919 case "a" { print "string a" }
920 case [1..10,42] { print "number in list" }
921 case (@array) { print "number in list" }
922 case /\w+/ { print "pattern" }
923 case qr/\w+/ { print "pattern" }
924 case (%hash) { print "entry in hash" }
925 case (\%hash) { print "entry in hash" }
926 case (\&sub) { print "arg to subroutine" }
927 else { print "previous case not true" }
934 C<Test::More> is yet another framework for writing test scripts,
935 more extensive than Test::Simple, by Michael Schwern. See L<Test::More>.
939 C<Test::Simple> has basic utilities for writing tests, by Michael
940 Schwern. See L<Test::Simple>.
944 C<Text::Balanced> has been added, for extracting delimited text
945 sequences from strings, from Damian Conway.
947 use Text::Balanced 'extract_delimited';
949 ($a, $b) = extract_delimited("'never say never', he never said", "'", '');
951 $a will be "'never say never'", $b will be ', he never said'.
953 In addition to extract_delimited() there are also extract_bracketed(),
954 extract_quotelike(), extract_codeblock(), extract_variable(),
955 extract_tagged(), extract_multiple(), gen_delimited_pat(), and
956 gen_extract_tagged(). With these you can implement rather advanced
957 parsing algorithms. See L<Text::Balanced>.
961 C<threads> is an interface to interpreter threads, by Arthur Bergman.
962 Interpreter threads (ithreads) is the new thread model introduced in
963 Perl 5.6 but only available as an internal interface for extension
964 writers (and for Win32 Perl for C<fork()> emulation). See L<threads>.
968 C<threads::shared> allows data sharing for interpreter threads, from
969 Arthur Bergman. In the ithreads model any data sharing between
970 threads must be explicit, as opposed to the old 5.005 thread model
971 where data sharing was implicit. See L<threads::shared>.
975 C<Tie::File>, by Mark-Jason Dominus, associates a Perl array with the
980 C<Tie::Memoize>, by Ilya Zakharevich, provides on-demand loaded hashes.
984 C<Tie::RefHash::Nestable>, by Edward Avis, allows storing hash
985 references (unlike the standard Tie::RefHash) The module is contained
986 within Tie::RefHash, see L<Tie::RefHash>.
990 C<Time::HiRes> provides high resolution timing (ualarm, usleep,
991 and gettimeofday), from Douglas E. Wegscheid. See L<Time::HiRes>.
995 C<Unicode::UCD> offers a querying interface to the Unicode Character
996 Database. See L<Unicode::UCD>.
1000 C<Unicode::Collate> implements the UCA (Unicode Collation Algorithm)
1001 for sorting Unicode strings, by SADAHIRO Tomoyuki. See L<Unicode::Collate>.
1005 C<Unicode::Normalize> implements the various Unicode normalization
1006 forms, by SADAHIRO Tomoyuki. See L<Unicode::Normalize>.
1010 C<XS::Typemap>, by Tim Jenness, is a test extension that exercises XS
1011 typemaps. Nothing gets installed but for extension writers the code
1016 =head2 Updated And Improved Modules and Pragmata
1022 The following independently supported modules have been updated to the
1023 newest versions from CPAN: CGI, CPAN, DB_File, File::Spec, File::Temp,
1024 Getopt::Long, Math::BigFloat, Math::BigInt, the podlators bundle
1025 (Pod::Man, Pod::Text), Pod::LaTeX, Pod::Parser, Storable,
1026 Term::ANSIColor, Test, Text-Tabs+Wrap.
1030 The attributes::reftype() now works on tied arguments.
1034 AutoLoader can now be disabled with C<no AutoLoader;>.
1038 B::Deparse has been significantly enhanced. It now can deparse almost
1039 all of the standard test suite (so that the tests still succeed).
1040 There is a make target "test.deparse" for trying this out.
1044 Class::Struct can now define the classes in compile time.
1048 Class::Struct now assigns the array/hash element if the accessor
1049 is called with an array/hash element as the B<sole> argument.
1053 Data::Dumper has now an option to sort hashes.
1057 Data::Dumper has now an option to dump code references
1062 DB_File now supports newer Berkeley DB versions, among
1067 The English module can now be used without the infamous performance
1070 use English '-no_match_vars';
1072 (Assuming, of course, that one doesn't need the troublesome variables
1073 C<$`>, C<$&>, or C<$'>.) Also, introduced C<@LAST_MATCH_START> and
1074 C<@LAST_MATCH_END> English aliases for C<@-> and C<@+>.
1078 Fcntl, Socket, and Sys::Syslog have been rewritten to use the
1079 new-style constant dispatch section (see L<ExtUtils::Constant>).
1080 This means that they will be more robust and hopefully faster.
1084 File::Find now chdir()s correctly when chasing symbolic links.
1088 File::Find now has pre- and post-processing callbacks. It also
1089 correctly changes directories when chasing symbolic links. Callbacks
1090 (naughtily) exiting with "next;" instead of "return;" now work.
1094 File::Find is now (again) reentrant. It also has been made
1099 The warnings issued by File::Find now belong to their own category.
1100 You can enable/disable them with C<use/no warnings 'File::Find';>.
1104 File::Glob::glob() renamed to File::Glob::bsd_glob() to avoid
1105 prototype mismatch with CORE::glob().
1109 File::Glob now supports C<GLOB_LIMIT> constant to limit the size of
1110 the returned list of filenames.
1114 Devel::Peek now has an interface for the Perl memory statistics
1115 (this works only if you are using perl's malloc, and if you have
1116 compiled with debugging).
1120 IPC::Open3 now allows the use of numeric file descriptors.
1124 IO::Socket has now atmark() method, which returns true if the socket
1125 is positioned at the out-of-band mark. The method is also exportable
1126 as a sockatmark() function.
1130 IO::Socket::INET has support for ReusePort option (if your platform
1131 supports it). The Reuse option now has an alias, ReuseAddr. For clarity
1132 you may want to prefer ReuseAddr.
1136 IO::Socket::INET now supports C<LocalPort> of zero (usually meaning
1137 that the operating system will make one up.)
1141 use lib now works identically to @INC. Removing directories
1142 with 'no lib' now works.
1146 ExtUtils::MakeMaker now uses File::Spec internally, which hopefully
1147 leads into better portability.
1151 Math::BigFloat and Math::BigInt have undergone a full rewrite.
1152 They are now magnitudes faster, and they support various
1153 bignum libraries such as GMP and PARI as their backends.
1157 Math::Complex handles inf, NaN etc., better.
1161 Net::Ping has been muchly enhanced: multihoming is now supported,
1162 Win32 functionality is better, there is now time measuring
1163 functionality (optionally high-resolution using Time::HiRes),
1164 and there is now "external" protocol which uses Net::Ping::External
1165 module which runs your external ping utility and parses the output.
1166 A version of Net::Ping::External is available in CPAN.
1168 Note that some of the Net::Ping tests are disabled when running
1169 under the Perl distribution since one cannot assume one or more
1170 of the following: enabled echo port at localhost, full Internet
1171 connectivity, or sympathetic firewalls. You can set the environment
1172 variable PERL_TEST_Net_Ping to "1" (one) before running the Perl test
1173 suite to enable all the Net::Ping tests.
1177 POSIX::sigaction() is now much more flexible and robust.
1178 You can now install coderef handlers, 'DEFAULT', and 'IGNORE'
1179 handlers, installing new handlers was not atomic.
1183 In Safe the C<%INC> now localised in a Safe compartment so that
1188 In SDBM_File on dosish platforms, some keys went missing because of
1189 lack of support for files with "holes". A workaround for the problem
1194 In Search::Dict one can now have a pre-processing hook for the
1195 lines being searched.
1199 The Shell module now has an OO interface.
1203 In Sys::Syslog there is now a failover mechanism that will go
1204 through alternative connection mechanisms until the message
1205 is successfully logged.
1209 The Test module has been significantly enhanced.
1213 Time::Local::timelocal() does not handle fractional seconds anymore.
1214 The rationale is that neither does localtime(), and timelocal() and
1215 localtime() are supposed to be inverses of each other.
1219 The vars pragma now supports declaring fully qualified variables.
1220 (Something that C<our()> does not and will not support.)
1224 The C<utf8::> name space (as in the pragma) provides various
1225 Perl-callable functions to provide low level access to Perl's
1226 internal Unicode representation. At the moment only length()
1227 has been implemented.
1231 =head1 Utility Changes
1237 Emacs perl mode (emacs/cperl-mode.el) has been updated to version
1242 F<emacs/e2ctags.pl> is now much faster.
1246 C<enc2xs> is a tool for people adding their own encodings to the
1251 C<h2ph> now supports C trigraphs.
1255 C<h2xs> now produces a template README.
1259 C<h2xs> now uses C<Devel::PPort> for better portability between
1260 different versions of Perl.
1264 C<h2xs> uses the new L<ExtUtils::Constant> module which will affect
1265 newly created extensions that define constants. Since the new code is
1266 more correct (if you have two constants where the first one is a
1267 prefix of the second one, the first constant B<never> gets defined),
1268 less lossy (it uses integers for integer constant, as opposed to the
1269 old code that used floating point numbers even for integer constants),
1270 and slightly faster, you might want to consider regenerating your
1271 extension code (the new scheme makes regenerating easy).
1272 L<h2xs> now also supports C trigraphs.
1276 C<libnetcfg> has been added to configure the libnet.
1280 C<perlbug> is now much more robust. It also sends the bug report to
1281 perl.org, not perl.com.
1285 C<perlcc> has been rewritten and its user interface (that is,
1286 command line) is much more like that of the UNIX C compiler, cc.
1287 (The perlbc tools has been removed. Use C<perlcc -B> instead.)
1291 C<perlivp> is a new Installation Verification Procedure utility
1292 for running any time after installing Perl.
1296 C<piconv> is an implementation of the character conversion utility
1297 C<iconv>, demonstrating the new Encode module.
1301 C<pod2html> now allows specifying a cache directory.
1305 C<pod2html> now produces XHTML 1.0.
1309 C<pod2html> now understands POD written using different line endings
1310 (PC-like CRLF versus UNIX-like LF versus MacClassic-like CR).
1314 C<s2p> has been completely rewritten in Perl. (It is in fact a full
1315 implementation of sed in Perl: you can use the sed functionality by
1316 using the C<psed> utility.)
1320 C<xsubpp> now understands POD documentation embedded in the *.xs files.
1324 C<xsubpp> now supports OUT keyword.
1328 =head1 New Documentation
1334 perl56delta details the changes between the 5.005 release and the
1339 perlclib documents the internal replacements for standard C library
1340 functions. (Interesting only for extension writers and Perl core
1345 perldebtut is a Perl debugging tutorial.
1349 perlebcdic contains considerations for running Perl on EBCDIC platforms.
1353 perlintro is a gentle introduction to Perl.
1357 perliol documents the internals of PerlIO with layers.
1361 perlmodstyle is a style guide for writing modules.
1365 perlnewmod tells about writing and submitting a new module.
1369 perlpacktut is a pack() tutorial.
1373 perlpod has been rewritten to be clearer and to record the best
1374 practices gathered over the years.
1378 perlpodspec is a more formal specification of the pod format,
1379 mainly of interest for writers of pod applications, not to
1380 people writing in pod.
1384 perlretut is a regular expression tutorial.
1388 perlrequick is a regular expressions quick-start guide.
1389 Yes, much quicker than perlretut.
1393 perltodo has been updated.
1397 perltootc has been renamed as perltooc (to not to conflict
1398 with perltoot in filesystems restricted to "8.3" names)
1402 perluniintro is an introduction to using Unicode in Perl.
1403 (perlunicode is more of a detailed reference and background
1408 perlutil explains the command line utilities packaged with the Perl
1413 The following platform-specific documents are available before
1414 the installation as README.I<platform>, and after the installation
1417 perlaix perlamiga perlapollo perlbeos perlbs2000
1418 perlce perlcygwin perldgux perldos perlepoc perlhpux
1419 perlhurd perlmachten perlmacos perlmint perlmpeix
1420 perlnetware perlos2 perlos390 perlplan9 perlqnx perlsolaris
1421 perltru64 perluts perlvmesa perlvms perlvos perlwin32
1427 The documentation for the POSIX-BC platform is called "BS2000", to avoid
1428 confusion with the Perl POSIX module.
1432 The documentation for the WinCE platform is called perlce (README.ce
1433 in the source code kit), to avoid confusion with the perlwin32
1434 documentation on 8.3-restricted filesystems.
1438 =head1 Performance Enhancements
1444 map() could get pathologically slow when the result list it generates
1445 is larger than the source list. The performance has been improved for
1450 sort() has been changed to use primarily mergesort internally as
1451 opposed to the earlier quicksort. For very small lists this may
1452 result in slightly slower sorting times, but in general the speedup
1453 should be at least 20%. Additional bonuses are that the worst case
1454 behaviour of sort() is now better (in computer science terms it now
1455 runs in time O(N log N), as opposed to quicksort's Theta(N**2)
1456 worst-case run time behaviour), and that sort() is now stable
1457 (meaning that elements with identical keys will stay ordered as they
1458 were before the sort). See the C<sort> pragma for information.
1460 The story in more detail: suppose you want to serve yourself a little
1463 @digits = ( 3,1,4,1,5,9 );
1465 A numerical sort of the digits will yield (1,1,3,4,5,9), as expected.
1466 Which C<1> comes first is hard to know, since one C<1> looks pretty
1467 much like any other. You can regard this as totally trivial,
1468 or somewhat profound. However, if you just want to sort the even
1469 digits ahead of the odd ones, then what will
1471 sort { ($a % 2) <=> ($b % 2) } @digits;
1473 yield? The only even digit, C<4>, will come first. But how about
1474 the odd numbers, which all compare equal? With the quicksort algorithm
1475 used to implement Perl 5.6 and earlier, the order of ties is left up
1476 to the sort. So, as you add more and more digits of Pi, the order
1477 in which the sorted even and odd digits appear will change.
1478 and, for sufficiently large slices of Pi, the quicksort algorithm
1479 in Perl 5.8 won't return the same results even if reinvoked with the
1480 same input. The justification for this rests with quicksort's
1481 worst case behavior. If you run
1483 sort { $a <=> $b } ( 1 .. $N , 1 .. $N );
1485 (something you might approximate if you wanted to merge two sorted
1486 arrays using sort), doubling $N doesn't just double the quicksort time,
1487 it I<quadruples> it. Quicksort has a worst case run time that can
1488 grow like N**2, so-called I<quadratic> behaviour, and it can happen
1489 on patterns that may well arise in normal use. You won't notice this
1490 for small arrays, but you I<will> notice it with larger arrays,
1491 and you may not live long enough for the sort to complete on arrays
1492 of a million elements. So the 5.8 quicksort scrambles large arrays
1493 before sorting them, as a statistical defence against quadratic behaviour.
1494 But that means if you sort the same large array twice, ties may be
1495 broken in different ways.
1497 Because of the unpredictability of tie-breaking order, and the quadratic
1498 worst-case behaviour, quicksort was I<almost> replaced completely with
1499 a stable mergesort. I<Stable> means that ties are broken to preserve
1500 the original order of appearance in the input array. So
1502 sort { ($a % 2) <=> ($b % 2) } (3,1,4,1,5,9);
1504 will yield (4,3,1,1,5,9), guaranteed. The even and odd numbers
1505 appear in the output in the same order they appeared in the input.
1506 Mergesort has worst case O(NlogN) behaviour, the best value
1507 attainable. And, ironically, this mergesort does particularly
1508 well where quicksort goes quadratic: mergesort sorts (1..$N, 1..$N)
1509 in O(N) time. But quicksort was rescued at the last moment because
1510 it is faster than mergesort on certain inputs and platforms.
1511 For example, if you really I<don't> care about the order of even
1512 and odd digits, quicksort will run in O(N) time; it's very good
1513 at sorting many repetitions of a small number of distinct elements.
1514 The quicksort divide and conquer strategy works well on platforms
1515 with relatively small, very fast, caches. Eventually, the problem gets
1516 whittled down to one that fits in the cache, from which point it
1517 benefits from the increased memory speed.
1519 Quicksort was rescued by implementing a sort pragma to control aspects
1520 of the sort. The B<stable> subpragma forces stable behaviour,
1521 regardless of algorithm. The B<_quicksort> and B<_mergesort>
1522 subpragmas are heavy-handed ways to select the underlying implementation.
1523 The leading C<_> is a reminder that these subpragmas may not survive
1524 beyond 5.8. More appropriate mechanisms for selecting the implementation
1525 exist, but they wouldn't have arrived in time to save quicksort.
1529 Hashes now use Bob Jenkins "One-at-a-Time" hashing key algorithm
1530 ( http://burtleburtle.net/bob/hash/doobs.html ). This algorithm is
1531 reasonably fast while producing a much better spread of values than
1532 the old hashing algorithm (originally by Chris Torek, later tweaked by
1533 Ilya Zakharevich). Hash values output from the algorithm on a hash of
1534 all 3-char printable ASCII keys comes much closer to passing the
1535 DIEHARD random number generation tests. According to perlbench, this
1536 change has not affected the overall speed of Perl.
1540 unshift() should now be noticeably faster.
1544 =head1 Installation and Configuration Improvements
1546 =head2 Generic Improvements
1552 INSTALL now explains how you can configure Perl to use 64-bit
1553 integers even on non-64-bit platforms.
1557 Policy.sh policy change: if you are reusing a Policy.sh file
1558 (see INSTALL) and you use Configure -Dprefix=/foo/bar and in the old
1559 Policy $prefix eq $siteprefix and $prefix eq $vendorprefix, all of
1560 them will now be changed to the new prefix, /foo/bar. (Previously
1561 only $prefix changed.) If you do not like this new behaviour,
1562 specify prefix, siteprefix, and vendorprefix explicitly.
1566 A new optional location for Perl libraries, otherlibdirs, is available.
1567 It can be used for example for vendor add-ons without disturbing Perl's
1568 own library directories.
1572 In many platforms the vendor-supplied 'cc' is too stripped-down to
1573 build Perl (basically, 'cc' doesn't do ANSI C). If this seems
1574 to be the case and 'cc' does not seem to be the GNU C compiler
1575 'gcc', an automatic attempt is made to find and use 'gcc' instead.
1579 gcc needs to closely track the operating system release to avoid
1580 build problems. If Configure finds that gcc was built for a different
1581 operating system release than is running, it now gives a clearly visible
1582 warning that there may be trouble ahead.
1586 If binary compatibility with the 5.005 release is not wanted, Configure
1587 no longer suggests including the 5.005 modules in @INC.
1591 Configure C<-S> can now run non-interactively.
1595 Configure support for pdp11-style memory models has been removed due
1600 configure.gnu now works with options with whitespace in them.
1604 installperl now outputs everything to STDERR.
1608 $Config{byteorder} is now computed dynamically (this is more robust
1609 with "fat binaries" where an executable image contains binaries for
1610 more than one binary platform.)
1614 Because PerlIO is now the default on most platforms, "-perlio" doesn't
1615 get appended to the $Config{archname} (also known as $^O) anymore.
1616 Instead, if you explicitly choose not to use perlio (Configure command
1617 line option -Uuseperlio), you will get "-stdio" appended.
1621 Another change related to the architecture name is that "-64all"
1622 (-Duse64bitall, or "maximally 64-bit") is appended only if your
1623 pointers are 64 bits wide. (To be exact, the use64bitall is ignored.)
1627 In AFS installations one can configure the root of the AFS to be
1628 somewhere else than the default F</afs> by using the Configure
1629 parameter C<-Dafsroot=/some/where/else>.
1633 APPLLIB_EXP, a less-know configuration-time definition, has been
1634 documented. It can be used to prepend site-specific directories
1635 to Perl's default search path (@INC), see INSTALL for information.
1639 The version of Berkeley DB used when the Perl (and, presumably, the
1640 DB_File extension) was built is now available as
1641 C<@Config{qw(db_version_major db_version_minor db_version_patch)}>
1642 from Perl and as C<DB_VERSION_MAJOR_CFG DB_VERSION_MINOR_CFG
1643 DB_VERSION_PATCH_CFG> from C.
1647 Building Berkeley DB3 for compatibility modes for DB, NDBM, and ODBM
1648 has been documented in INSTALL.
1652 If you have CPAN access (either network or a local copy such as a
1653 CD-ROM) you can during specify extra modules to Configure to build and
1654 install with Perl using the -Dextras=... option. See INSTALL for
1659 In addition to config.over a new override file, config.arch, is
1660 available. That is supposed to be used by hints file writers for
1661 architecture-wide changes (as opposed to config.over which is for
1666 If your file system supports symbolic links you can build Perl outside
1667 of the source directory by
1669 mkdir /tmp/perl/build/directory
1670 cd /tmp/perl/build/directory
1671 sh /path/to/perl/source/Configure -Dmksymlinks ...
1673 This will create in /tmp/perl/build/directory a tree of symbolic links
1674 pointing to files in /path/to/perl/source. The original files are left
1675 unaffected. After Configure has finished you can just say
1679 and Perl will be built and tested, all in /tmp/perl/build/directory.
1683 For Perl developers several new make targets for profiling
1684 and debugging have been added, see L<perlhack>.
1690 Use of the F<gprof> tool to profile Perl has been documented in
1691 L<perlhack>. There is a make target called "perl.gprof" for
1692 generating a gprofiled Perl executable.
1696 If you have GCC 3, there is a make target called "perl.gcov" for
1697 creating a gcoved Perl executable for coverage analysis. See
1702 If you are on IRIX or Tru64 platforms, new profiling/debugging options
1703 have been added, see L<perlhack> for more information about pixie and
1710 Guidelines of how to construct minimal Perl installations have
1711 been added to INSTALL.
1715 The Thread extension is now not built at all under ithreads
1716 (C<Configure -Duseithreads>) because it wouldn't work anyway (the
1717 Thread extension requires being Configured with C<-Duse5005threads>).
1719 But note that the Thread.pm interface is now shared by both
1724 The Gconvert macro ($Config{d_Gconvert}) used by perl for stringifying
1725 floating-point numbers is now more picky about using sprintf %.*g
1726 rules for the conversion. Some platforms that used to use gcvt may
1727 now resort to the slower sprintf.
1731 =head2 New Or Improved Platforms
1733 For the list of platforms known to support Perl,
1734 see L<perlport/"Supported Platforms">.
1740 AIX dynamic loading should be now better supported.
1744 AIX should now work better with gcc, threads, and 64-bitness. Also the
1745 long doubles support in AIX should be better now. See L<perlaix>.
1749 AtheOS ( http://www.atheos.cx/ ) is a new platform.
1753 BeOS has been reclaimed.
1757 DG/UX platform now supports the 5.005-style threads. See L<perldgux>.
1761 DYNIX/ptx platform (a.k.a. dynixptx) is supported at or near osvers 4.5.2.
1765 EBCDIC platforms (z/OS, also known as OS/390, POSIX-BC, and VM/ESA)
1766 have been regained. Many test suite tests still fail and the
1767 co-existence of Unicode and EBCDIC isn't quite settled, but the
1768 situation is much better than with Perl 5.6. See L<perlos390>,
1769 L<perlbs2000> (for POSIX-BC), and L<perlvmesa> for more information.
1773 Building perl with -Duseithreads or -Duse5005threads now works under
1774 HP-UX 10.20 (previously it only worked under 10.30 or later). You will
1775 need a thread library package installed. See README.hpux.
1779 MacOS Classic (MacPerl has of course been available since
1780 perl 5.004 but now the source code bases of standard Perl
1781 and MacPerl have been synchronised)
1785 MacOS X (or Darwin) should now be able to build Perl even on HFS+
1786 filesystems. (The case-insensitivity confused the Perl build process.)
1790 NCR MP-RAS is now supported.
1794 All the NetBSD specific patches (except for the installation
1795 specific ones) have been merged back to the main distribution.
1799 NetWare from Novell is now supported. See L<perlnetware>.
1803 NonStop-UX is now supported.
1807 NEC SUPER-UX is now supported.
1811 All the OpenBSD specific patches (except for the installation
1812 specific ones) have been merged back to the main distribution.
1816 Perl has been tested with the GNU pth userlevel thread package
1817 ( http://www.gnu.org/software/pth/pth.html ) . All but one thread
1818 test worked, and that one failure was because of test results arriving
1819 in unexpected order.
1823 Amdahl UTS UNIX mainframe platform is now supported.
1827 WinCE is now supported. See L<perlce>.
1831 z/OS (formerly known as OS/390, formerly known as MVS OE) has now
1832 support for dynamic loading. This is not selected by default,
1833 however, you must specify -Dusedl in the arguments of Configure.
1837 =head1 Selected Bug Fixes
1839 Numerous memory leaks and uninitialized memory accesses have been
1840 hunted down. Most importantly anonymous subs used to leak quite
1847 The autouse pragma didn't work for Multi::Part::Function::Names.
1851 caller() could cause core dumps in certain situations. Carp was sometimes
1852 affected by this problem.
1856 chop(@list) in list context returned the characters chopped in
1857 reverse order. This has been reversed to be in the right order.
1861 Configure no longer includes the DBM libraries (dbm, gdbm, db, ndbm)
1862 when building the Perl binary. The only exception to this is SunOS 4.x,
1867 The behaviour of non-decimal but numeric string constants such as
1868 "0x23" was platform-dependent: in some platforms that was seen as 35,
1869 in some as 0, in some as a floating point number (don't ask). This
1870 was caused by Perl using the operating system libraries in a situation
1871 where the result of the string to number conversion is undefined: now
1872 Perl consistently handles such strings as zero in numeric contexts.
1876 The order of DESTROYs has been made more predictable.
1880 Several debugger fixes: exit code now reflects the script exit code,
1881 condition C<"0"> now treated correctly, the C<d> command now checks
1882 line number, the C<$.> no longer gets corrupted, all debugger output
1883 now goes correctly to the socket if RemotePort is set.
1887 Perl 5.6.0 could emit spurious warnings about redefinition of dl_error()
1888 when statically building extensions into perl. This has been corrected.
1892 L<dprofpp> -R didn't work.
1896 C<*foo{FORMAT}> now works.
1899 Infinity is now recognized as a number.
1903 UNIVERSAL::isa no longer caches methods incorrectly. (This broke
1904 the Tk extension with 5.6.0.)
1908 Lexicals I: lexicals outside an eval "" weren't resolved
1909 correctly inside a subroutine definition inside the eval "" if they
1910 were not already referenced in the top level of the eval""ed code.
1914 Lexicals II: lexicals leaked at file scope into subroutines that
1915 were declared before the lexicals.
1919 Lexical warnings now propagating correctly between scopes
1920 and into C<eval "...">.
1924 C<use warnings qw(FATAL all)> did not work as intended. This has been
1929 warnings::enabled() now reports the state of $^W correctly if the caller
1930 isn't using lexical warnings.
1934 Line renumbering with eval and C<#line> now works.
1938 Fixed numerous memory leaks, especially in eval "".
1942 Localised tied variables no more leak memory
1945 tie my %tied_hash => 'Tie::StdHash';
1949 # Used to leak memory every time local() was called,
1950 # in a loop this added up.
1951 local($tied_hash{Foo}) = 1;
1955 Localised hash elements (and %ENV) are correctly unlocalised to not to
1956 exist, if that's what they were.
1960 tie my %tied_hash => 'Tie::StdHash';
1964 # Nothing has set the FOO element so far
1966 { local $tied_hash{FOO} = 'Bar' }
1968 # Here the FOO element would have been C<undef>,
1971 As a side effect of this fix, tied hash interfaces B<must> define
1972 the EXISTS and DELETE methods.
1976 mkdir() now ignores trailing slashes in the directory name,
1977 as mandated by POSIX.
1981 Some versions of glibc have a broken modfl(). This affects builds
1982 with C<-Duselongdouble>. This version of Perl detects this brokenness
1983 and has a workaround for it. The glibc release 2.2.2 is known to have
1984 fixed the modfl() bug.
1988 Modulus of unsigned numbers now works (4063328477 % 65535 used to
1989 return 27406, instead of 27047).
1993 Some "not a number" warnings introduced in 5.6.0 eliminated to be
1994 more compatible with 5.005. Infinity is now recognised as a number.
1998 Numeric conversions did not recognize changes in the string value
1999 properly in certain circumstances.
2003 Attributes (like :shared) didn't work with our().
2007 our() variables will not cause "will not stay shared" warnings.
2011 "our" variables of the same name declared in two sibling blocks
2012 resulted in bogus warnings about "redeclaration" of the variables.
2013 The problem has been corrected.
2017 pack "Z" now correctly terminates the string with "\0".
2021 Fix password routines which in some shadow password platforms
2022 (e.g. HP-UX) caused getpwent() to return every other entry.
2026 The PERL5OPT environment variable (for passing command line arguments
2027 to Perl) didn't work for more than a single group of options.
2031 PERL5OPT with embedded spaces didn't work.
2035 printf() no longer resets the numeric locale to "C".
2039 C<qw(a\\b)> now parses correctly as C<'a\\b'>.
2043 pos() did not return the correct value within s///ge in earlier
2044 versions. This is now handled correctly.
2048 Printing quads (64-bit integers) with printf/sprintf now works
2049 without the q L ll prefixes (assuming you are on a quad-capable platform).
2053 Regular expressions on references and overloaded scalars now work.
2057 Right-hand side magic (GMAGIC) could in many cases such as string
2058 concatenation be invoked too many times.
2062 scalar() now forces scalar context even when used in void context.
2066 SOCKS support is now much more robust.
2070 sort() arguments are now compiled in the right wantarray context
2071 (they were accidentally using the context of the sort() itself).
2072 The comparison block is now run in scalar context, and the arguments
2073 to be sorted are always provided list context.
2077 Changed the POSIX character class C<[[:space:]]> to include the (very
2078 rarely used) vertical tab character. Added a new POSIX-ish character
2079 class C<[[:blank:]]> which stands for horizontal whitespace
2080 (currently, the space and the tab).
2084 The tainting behaviour of sprintf() has been rationalized. It does
2085 not taint the result of floating point formats anymore, making the
2086 behaviour consistent with that of string interpolation.
2090 Some cases of inconsistent taint propagation (such as within hash
2091 values) have been fixed.
2095 The RE engine found in Perl 5.6.0 accidentally pessimised certain kinds
2096 of simple pattern matches. These are now handled better.
2100 Regular expression debug output (whether through C<use re 'debug'>
2101 or via C<-Dr>) now looks better.
2105 Multi-line matches like C<"a\nxb\n" =~ /(?!\A)x/m> were flawed. The
2110 Use of $& could trigger a core dump under some situations. This
2115 The regular expression captured submatches ($1, $2, ...) are now
2116 more consistently unset if the match fails, instead of leaving false
2117 data lying around in them.
2121 readline() on files opened in "slurp" mode could return an extra "" at
2122 the end in certain situations. This has been corrected.
2126 Autovivification of symbolic references of special variables described
2127 in L<perlvar> (as in C<${$num}>) was accidentally disabled. This works
2132 Sys::Syslog ignored the C<LOG_AUTH> constant.
2136 All but the first argument of the IO syswrite() method are now optional.
2140 $AUTOLOAD, sort(), lock(), and spawning subprocesses
2141 in multiple threads simultaneously are now thread-safe.
2145 Tie::ARRAY SPLICE method was broken.
2149 Allow read-only string on left hand side of non-modifying tr///.
2153 If C<STDERR> is tied, warnings caused by C<warn> and C<die> now
2154 correctly pass to it.
2158 Several Unicode fixes.
2164 BOMs (byte order marks) in the beginning of Perl files
2165 (scripts, modules) should now be transparently skipped.
2166 UTF-16 (UCS-2) encoded Perl files should now be read correctly.
2170 The character tables have been updated to Unicode 3.2.0.
2174 Comparing with utf8 data does not magically upgrade non-utf8 data
2175 into utf8. (This was a problem for example if you were mixing data
2176 from I/O and Unicode data: your output might have got magically encoded
2181 Generating illegal Unicode code points like U+FFFE, or the UTF-16
2182 surrogates, now also generates an optional warning.
2186 C<IsAlnum>, C<IsAlpha>, and C<IsWord> now match titlecase.
2190 Concatenation with the C<.> operator or via variable interpolation,
2191 C<eq>, C<substr>, C<reverse>, C<quotemeta>, the C<x> operator,
2192 substitution with C<s///>, single-quoted UTF8, should now work.
2196 The C<tr///> operator now works. Note that the C<tr///CU>
2197 functionality has been removed (but see pack('U0', ...)).
2201 C<eval "v200"> now works.
2205 Perl 5.6.0 parsed m/\x{ab}/ incorrectly, leading to spurious warnings.
2206 This has been corrected.
2210 Zero entries were missing from the Unicode classes like C<IsDigit>.
2216 Large unsigned numbers (those above 2**31) could sometimes lose their
2217 unsignedness, causing bogus results in arithmetic operations.
2221 =head2 Platform Specific Changes and Fixes
2229 Perl now works on post-4.0 BSD/OSes.
2235 Setting C<$0> now works (as much as possible; see L<perlvar> for details).
2241 Numerous updates; currently synchronised with Cygwin 1.3.10.
2245 Previously DYNIX/ptx had problems in its Configure probe for non-blocking I/O.
2251 EPOC update after Perl 5.6.0. See README.epoc.
2257 Perl now works on post-3.0 FreeBSDs.
2263 README.hpux updated; C<Configure -Duse64bitall> now works.
2269 Numerous compilation flag and hint enhancements; accidental mixing
2270 of 32-bit and 64-bit libraries (a doomed attempt) made much harder.
2280 Long doubles should now work (see INSTALL).
2284 Linux previously had problems related to sockaddrlen when using
2285 accept(), revcfrom() (in Perl: recv()), getpeername(), and getsockname().
2293 Compilation of the standard Perl distribution in MacOS Classic should
2294 now work if you have the Metrowerks development environment and
2295 the missing Mac-specific toolkit bits. Contact the macperl mailing
2302 MPE/iX update after Perl 5.6.0. See README.mpeix.
2306 NetBSD/threads: try installing the GNU pth (should be in the
2307 packages collection, or http://www.gnu.org/software/pth/),
2308 and Configure with -Duseithreads.
2314 Perl now works on NetBSD/sparc.
2320 Now works with usethreads (see INSTALL).
2326 64-bitness using the Sun Workshop compiler now works.
2330 Tru64 (aka Digital UNIX, aka DEC OSF/1)
2332 The operating system version letter now recorded in $Config{osvers}.
2333 Allow compiling with gcc (previously explicitly forbidden). Compiling
2334 with gcc still not recommended because buggy code results, even with
2341 Fixed various alignment problems that lead into core dumps either
2342 during build or later; no longer dies on math errors at runtime;
2343 now using full quad integers (64 bits), previously was using
2344 only 46 bit integers for speed.
2350 chdir() now works better despite a CRT bug; now works with MULTIPLICITY
2351 (see INSTALL); now works with Perl's malloc.
2353 The tainting of C<%ENV> elements via C<keys> or C<values> was previously
2354 unimplemented. It now works as documented.
2356 The C<waitpid> emulation has been improved. The worst bug (now fixed)
2357 was that a pid of -1 would cause a wildcard search of all processes on
2358 the system. The most significant enhancement is that we can now
2359 usually get the completion status of a terminated process.
2361 POSIX-style signals are now emulated much better on VMS versions prior
2364 The C<system> function and backticks operator have improved
2365 functionality and better error handling.
2367 File access tests now use current process privileges rather than the
2368 user's default privileges, which could sometimes result in a mismatch
2369 between reported access and actual access.
2379 accept() no longer leaks memory.
2383 Borland C++ v5.5 is now a supported compiler that can build Perl.
2384 However, the generated binaries continue to be incompatible with those
2385 generated by the other supported compilers (GCC and Visual C++).
2389 Better chdir() return value for a non-existent directory.
2393 Duping socket handles with open(F, ">&MYSOCK") now works under Windows 9x.
2397 New %ENV entries now propagate to subprocesses.
2401 Current directory entries in %ENV are now correctly propagated to child
2406 $ENV{LIB} now used to search for libs under Visual C.
2410 fork() emulation has been improved in various ways, but still continues
2411 to be experimental. See L<perlfork> for known bugs and caveats.
2415 A failed (pseudo)fork now returns undef and sets errno to EAGAIN.
2419 Win32::GetCwd() correctly returns C:\ instead of C: when at the drive root.
2420 Other bugs in chdir() and Cwd::cwd() have also been fixed.
2424 HTML files will be installed in c:\perl\html instead of c:\perl\lib\pod\html
2428 The makefiles now provide a single switch to bulk-enable all the features
2429 enabled in ActiveState ActivePerl (a popular Win32 binary distribution).
2433 Allow REG_EXPAND_SZ keys in the registry.
2437 Can now send() from all threads, not just the first one.
2441 Fake signal handling reenabled, bugs and all.
2445 %SIG has been enabled under USE_ITHREADS, but its use is completely
2446 unsupported under all configurations.
2450 Less stack reserved per thread so that more threads can run
2451 concurrently. (Still 16M per thread.)
2455 C<File::Spec->tmpdir()> now prefers C:/temp over /tmp
2456 (works better when perl is running as service).
2460 Better UNC path handling under ithreads.
2464 wait(), waitpid() and backticks now return the correct exit status under
2469 winsock handle leak fixed.
2473 The Perl parser has been stress tested using both random input and
2474 Markov chain input and the few found crashes and lockups have been
2481 =head1 New or Changed Diagnostics
2487 The lexical warnings category "deprecated" is no longer a sub-category
2488 of the "syntax" category. It is now a top-level category in its own
2493 All regular expression compilation error messages are now hopefully
2494 easier to understand both because the error message now comes before
2495 the failed regex and because the point of failure is now clearly
2496 marked by a C<E<lt>-- HERE> marker.
2500 The various "opened only for", "on closed", "never opened" warnings
2501 drop the C<main::> prefix for filehandles in the C<main> package,
2502 for example C<STDIN> instead of C<main::STDIN>.
2506 The "Unrecognized escape" warning has been extended to include C<\8>,
2507 C<\9>, and C<\_>. There is no need to escape any of the C<\w> characters.
2511 Two new debugging options have been added: if you have compiled your
2512 Perl with debugging, you can use the -DT and -DR options to trace
2513 tokenising and to add reference counts to displaying variables,
2518 The debugger (perl5db.pl) has been modified to present a more
2519 consistent commands interface, via (CommandSet=580). perl5db.t was
2520 also added to test the changes, and as a placeholder for further tests.
2526 The debugger has a new C<dumpDepth> option to control the maximum
2527 depth to which nested structures are dumped. The C<x> command has
2528 been extended so that C<x N EXPR> dumps out the value of I<EXPR> to a
2529 depth of at most I<N> levels.
2533 The debugger can now show lexical variables if you have the CPAN
2534 module PadWalker installed.
2538 If an attempt to use a (non-blessed) reference as an array index
2539 is made, a warning is given.
2543 C<push @a;> and C<unshift @a;> (with no values to push or unshift)
2544 now give a warning. This may be a problem for generated and evaled
2549 If you try to L<perlfunc/pack> a number less than 0 or larger than 255
2550 using the C<"C"> format you will get an optional warning. Similarly
2551 for the C<"c"> format and a number less than -128 or more than 127.
2555 Certain regex modifiers such as C<(?o)> make sense only if applied to
2556 the entire regex. You will get an optional warning if you try to do
2561 Using arrays or hashes as references (e.g. C<< %foo->{bar} >>
2562 has been deprecated for a while. Now you will get an optional warning.
2566 Using C<sort> in scalar context now issues an optional warning.
2567 This didn't do anything useful, as the sort was not performed.
2571 =head1 Changed Internals
2577 perlapi.pod (a companion to perlguts) now attempts to document the
2582 You can now build a really minimal perl called microperl.
2583 Building microperl does not require even running Configure;
2584 C<make -f Makefile.micro> should be enough. Beware: microperl makes
2585 many assumptions, some of which may be too bold; the resulting
2586 executable may crash or otherwise misbehave in wondrous ways.
2587 For careful hackers only.
2591 Added rsignal(), whichsig(), do_join(), op_clear, op_null,
2592 ptr_table_clear(), ptr_table_free(), sv_setref_uv(), and several UTF-8
2593 interfaces to the publicised API. For the full list of the available
2594 APIs see L<perlapi>.
2598 Made possible to propagate customised exceptions via croak()ing.
2602 Now xsubs can have attributes just like subs. (Well, at least the
2603 built-in attributes.)
2607 dTHR and djSP have been obsoleted; the former removed (because it's
2608 a no-op) and the latter replaced with dSP.
2612 PERL_OBJECT has been completely removed.
2616 The MAGIC constants (e.g. C<'P'>) have been macrofied
2617 (e.g. C<PERL_MAGIC_TIED>) for better source code readability
2618 and maintainability.
2622 The regex compiler now maintains a structure that identifies nodes in
2623 the compiled bytecode with the corresponding syntactic features of the
2624 original regex expression. The information is attached to the new
2625 C<offsets> member of the C<struct regexp>. See L<perldebguts> for more
2626 complete information.
2630 The C code has been made much more C<gcc -Wall> clean. Some warning
2631 messages still remain in some platforms, so if you are compiling with
2632 gcc you may see some warnings about dubious practices. The warnings
2633 are being worked on.
2637 F<perly.c>, F<sv.c>, and F<sv.h> have now been extensively commented.
2641 Documentation on how to use the Perl source repository has been added
2642 to F<Porting/repository.pod>.
2646 There are now several profiling make targets.
2650 =head1 Security Vulnerability Closed
2652 (This change was already made in 5.7.0 but bears repeating here.)
2654 A potential security vulnerability in the optional suidperl component
2655 of Perl was identified in August 2000. suidperl is neither built nor
2656 installed by default. As of November 2001 the only known vulnerable
2657 platform is Linux, most likely all Linux distributions. CERT and
2658 various vendors and distributors have been alerted about the vulnerability.
2659 See http://www.cpan.org/src/5.0/sperl-2000-08-05/sperl-2000-08-05.txt
2660 for more information.
2662 The problem was caused by Perl trying to report a suspected security
2663 exploit attempt using an external program, /bin/mail. On Linux
2664 platforms the /bin/mail program had an undocumented feature which
2665 when combined with suidperl gave access to a root shell, resulting in
2666 a serious compromise instead of reporting the exploit attempt. If you
2667 don't have /bin/mail, or if you have 'safe setuid scripts', or if
2668 suidperl is not installed, you are safe.
2670 The exploit attempt reporting feature has been completely removed from
2671 Perl 5.8.0 (and the maintenance release 5.6.1, and it was removed also
2672 from all the Perl 5.7 releases), so that particular vulnerability
2673 isn't there anymore. However, further security vulnerabilities are,
2674 unfortunately, always possible. The suidperl functionality is most
2675 probably going to be removed in Perl 5.10. In any case, suidperl
2676 should only be used by security experts who know exactly what they are
2677 doing and why they are using suidperl instead of some other solution
2678 such as sudo ( see http://www.courtesan.com/sudo/ ).
2682 Several new tests have been added, especially for the F<lib>
2683 subsection. There are now about 56 000 individual tests (spread over
2684 about 620 test scripts), in the regression suite (5.6.1 has about
2685 11700 tests, in 258 test scripts) Many of the new tests are introduced
2686 by the new modules, but still in general Perl is now more thoroughly
2689 Because of the large number of tests, running the regression suite
2690 will take considerably longer time than it used to: expect the suite
2691 to take up to 4-5 times longer to run than in perl 5.6. In a really
2692 fast machine you can hope to finish the suite in about 6-8 minutes
2695 The tests are now reported in a different order than in earlier Perls.
2696 (This happens because the test scripts from under t/lib have been moved
2697 to be closer to the library/extension they are testing.)
2699 =head1 Known Problems
2707 In AIX 4.2 Perl extensions that use C++ functions that use statics
2708 may have problems in that the statics are not getting initialized.
2709 In newer AIX releases this has been solved by linking Perl with
2710 the libC_r library, but unfortunately in AIX 4.2 the said library
2711 has an obscure bug where the various functions related to time
2712 (such as time() and gettimeofday()) return broken values, and
2713 therefore in AIX 4.2 Perl is not linked against the libC_r.
2717 vac 5.0.0.0 May Produce Buggy Code For Perl
2719 The AIX C compiler vac version 5.0.0.0 may produce buggy code,
2720 resulting in few random tests failing, but when the failing tests
2721 are run by hand, they succeed. We suggest upgrading to at least
2722 vac version 5.0.1.0, that has been known to compile Perl correctly.
2723 "lslpp -L|grep vac.C" will tell you the vac version. See README.aix.
2727 If building threaded Perl, you may get compilation warning from pp_sys.c:
2729 "pp_sys.c", line 4651.39: 1506-280 (W) Function argument assignment between types "unsigned char*" and "const void*" is not allowed.
2731 This is harmless; it is caused by the getnetbyaddr() and getnetbyaddr_r()
2732 having slightly different types for their first argument.
2736 =head2 Amiga Perl Invoking Mystery
2738 One cannot call Perl using the C<volume:> syntax, that is, C<perl -v>
2739 works, but for example C<bin:perl -v> doesn't. The exact reason isn't
2740 known but the current suspect is the F<ixemul> library.
2742 =head2 ext/threads/t/libc
2744 If this test fails, it indicates that your libc (C library) is not
2745 threadsafe. This particular test stress tests the localtime() call to
2746 find out whether it is threadsafe. See L<perlthrtut> for more information.
2748 =head2 FreeBSD Failing locale Test 117 For ISO8859-15 Locales
2750 The ISO8859-15 locales may fail the locale test 117 in FreeBSD.
2751 This is caused by the characters \xFF (y with diaeresis) and \xBE
2752 (Y with diaeresis) not behaving correctly when being matched
2755 =head2 Modifying $_ Inside for(..)
2759 works without complaint. It shouldn't. (You should be able to
2760 modify only lvalue elements inside the loops.) You can see the
2761 correct behaviour by replacing the 1..5 with 1, 2, 3, 4, 5.
2763 =head2 mod_perl 1.26 Doesn't Build With Threaded Perl
2765 Use mod_perl 1.27 or higher.
2767 =head2 lib/ftmp-security tests warn 'system possibly insecure'
2769 Don't panic. Read INSTALL 'make test' section instead.
2771 =head2 HP-UX lib/posix Subtest 9 Fails When LP64-Configured
2773 If perl is configured with -Duse64bitall, the successful result of the
2774 subtest 10 of lib/posix may arrive before the successful result of the
2775 subtest 9, which confuses the test harness so much that it thinks the
2778 =head2 Linux With Sfio Fails op/misc Test 48
2784 Please remember to set your environment variable LC_ALL to "C"
2785 (setenv LC_ALL C) before running "make test" to avoid a lot of
2786 warnings about the broken locales of Mac OS X.
2788 The following tests are known to fail:
2790 Failed Test Stat Wstat Total Fail Failed List of Failed
2791 -------------------------------------------------------------------------
2792 ../ext/DB_File/t/db-btree.t 0 11 ?? ?? % ??
2793 ../ext/DB_File/t/db-recno.t 149 3 2.01% 61 63 65
2795 If you are building on a UFS partition, you will also probably see
2796 t/op/stat.t subtest #9 fail. This is caused by Darwin's UFS not
2797 supporting inode change time.
2799 Also the ext/POSIX/t/posix.t subtest #10 fails but it is skipped for
2800 now because the failure is Apple's fault, not Perl's (blocked signals
2803 If you Configure with ithreads, ext/threads/t/libc.t will fail, again
2804 not Perl's fault-- the libc of Mac OS X is not threadsafe (in this
2805 particular test the localtime() call is found to be threadunsafe.)
2807 =head2 op/sprintf tests 91, 129, and 130
2809 The op/sprintf tests 91, 129, and 130 are known to fail on some platforms.
2810 Examples include any platform using sfio, and Compaq/Tandem's NonStop-UX.
2812 The test 91 is known to fail at QNX6 (nto), because C<sprintf '%e',0>
2813 incorrectly produces C<0.000000e+0> instead of C<0.000000e+00>.
2815 For the tests 129 and 130 the failing platforms do not comply with
2816 the ANSI C Standard, line 19ff on page 134 of ANSI X3.159 1989 to
2817 be exact. (They produce something other than "1" and "-1" when
2818 formatting 0.6 and -0.6 using the printf format "%.0f", most often
2819 they produce "0" and "-0".)
2823 In case you are still using Solaris 2.5 (aka SunOS 5.5), you may
2824 experience failures (the test core dumping) in lib/locale.t.
2825 The suggested cure is to upgrade your Solaris.
2827 =head2 Failure of Thread (5.005-style) tests
2829 B<Note that support for 5.005-style threading is deprecated,
2830 experimental and practically unsupported. In 5.10 it is expected
2833 The following tests are known to fail due to fundamental problems in
2834 the 5.005 threading implementation. These are not new failures--Perl
2835 5.005_0x has the same bugs, but didn't have these tests.
2837 ../ext/List/Util/t/first.t 255 65280 7 4 57.14% 2 5-7
2838 ../lib/English.t 2 512 54 2 3.70% 2-3
2839 ../lib/Filter/Simple/t/data.t 6 3 50.00% 1-3
2840 ../lib/Filter/Simple/t/filter_onl 9 3 33.33% 1-2 5
2841 ../lib/autouse.t 10 1 10.00% 4
2842 op/flip.t 15 1 6.67% 15
2844 These failures are unlikely to get fixed as the 5.005-style
2845 threads are considered fundamentally broken.
2849 Failed Test Stat Wstat Total Fail Failed List of Failed
2850 -------------------------------------------------------------------------
2851 ../ext/Socket/socketpair.t 1 256 45 1 2.22% 12
2852 ../lib/Math/Trig.t 26 1 3.85% 25
2853 ../lib/warnings.t 460 1 0.22% 425
2854 io/fs.t 36 1 2.78% 31
2855 op/numconvert.t 1440 13 0.90% 208 509-510
2856 657-658 665-666 829-830 989-990 1149-1150
2864 During Configure the test
2866 Guessing which symbols your C compiler and preprocessor define...
2868 will probably fail with error messages like
2870 CC-20 cc: ERROR File = try.c, Line = 3
2871 The identifier "bad" is undefined.
2873 bad switch yylook 79bad switch yylook 79bad switch yylook 79bad switch yylook 79#ifdef A29K
2876 CC-65 cc: ERROR File = try.c, Line = 3
2877 A semicolon is expected at this point.
2879 This is caused by a bug in awk utility of UNICOS/mk. You can ignore
2880 the error, but it does cause a slight problem: you cannot fully
2881 benefit from the h2ph utility (see L<h2ph>) that can be used to
2882 convert C headers to Perl libraries, mainly used to be able to access
2883 from Perl the constants defined using C preprocessor, cpp. Because of
2884 the above error parts of the converted headers will be invisible.
2885 Luckily, these days the need for h2ph is rare.
2889 If building Perl with the interpreter threads (ithreads), the
2890 getgrent(), getgrnam(), and getgrgid() functions cannot return the
2891 list of the group members due to a bug in the multithreaded support of
2892 UNICOS/mk. What this means that in list context the functions will
2893 return only three values, not four.
2899 There are a few known test failures, see L<perluts>.
2903 There should be no reported test failures with a default configuration,
2904 though there are a number of tests marked TODO that point to areas
2905 needing further debugging and/or porting work.
2909 In multi-CPU boxes there are some problems with the I/O buffering:
2910 some output may appear twice.
2912 =head2 XML::Parser not working
2914 Use XML::Parser 2.31 or later.
2916 =head2 z/OS (OS/390)
2918 z/OS has rather many test failures but the situation is actually
2919 better than it was in 5.6.0, it's just that so many new modules and
2920 tests have been added.
2922 Failed Test Stat Wstat Total Fail Failed List of Failed
2923 ---------------------------------------------------------------------------
2924 ../ext/Data/Dumper/t/dumper.t 357 8 2.24% 311 314 325 327
2926 ../ext/IO/lib/IO/t/io_unix.t 5 4 80.00% 2-5
2927 ../lib/ExtUtils/t/Constant.t 121 30976 48 48 100.00% 1-48
2928 ../lib/ExtUtils/t/Embed.t 9 9 100.00% 1-9
2929 op/pat.t 910 9 0.99% 665 776 785
2931 op/sprintf.t 224 3 1.34% 98 100 136
2932 op/tr.t 97 5 5.15% 63 71-74
2933 run/fresh_perl.t 95 3 3.16% 92-94
2934 uni/fold.t 780 6 0.77% 61 169 196 661
2937 =head2 Localising Tied Arrays and Hashes Is Broken
2941 doesn't work as one would expect: the old value is restored
2944 =head2 Self-tying of Arrays and Hashes Is Forbidden
2946 Self-tying of arrays and hashes is broken in rather deep and
2947 hard-to-fix ways. As a stop-gap measure to avoid people from getting
2948 frustrated at the mysterious results (core dumps, most often) it is
2949 for now forbidden (you will get a fatal error even from an attempt).
2951 =head2 Building Extensions Can Fail Because Of Largefiles
2953 Some extensions like mod_perl are known to have issues with
2954 `largefiles', a change brought by Perl 5.6.0 in which file offsets
2955 default to 64 bits wide, where supported. Modules may fail to compile
2956 at all or compile and work incorrectly. Currently there is no good
2957 solution for the problem, but Configure now provides appropriate
2958 non-largefile ccflags, ldflags, libswanted, and libs in the %Config
2959 hash (e.g., $Config{ccflags_nolargefiles}) so the extensions that are
2960 having problems can try configuring themselves without the
2961 largefileness. This is admittedly not a clean solution, and the
2962 solution may not even work at all. One potential failure is whether
2963 one can (or, if one can, whether it's a good idea) link together at
2964 all binaries with different ideas about file offsets, all this is
2967 =head2 Unicode Support on EBCDIC Still Spotty
2969 Though mostly working, Unicode support still has problem spots on
2970 EBCDIC platforms. One such known spot are the C<\p{}> and C<\P{}>
2971 regular expression constructs for code points less than 256: the
2972 pP are testing for Unicode code points, not knowing about EBCDIC.
2974 =head2 The Compiler Suite Is Still Experimental
2976 The compiler suite is slowly getting better but it continues to be
2977 highly experimental. Use in production environments is discouraged.
2979 =head2 The Long Double Support Is Still Experimental
2981 The ability to configure Perl's numbers to use "long doubles",
2982 floating point numbers of hopefully better accuracy, is still
2983 experimental. The implementations of long doubles are not yet
2984 widespread and the existing implementations are not quite mature
2985 or standardised, therefore trying to support them is a rare
2986 and moving target. The gain of more precision may also be offset
2987 by slowdown in computations (more bits to move around, and the
2988 operations are more likely to be executed by less optimised
2991 =head2 Seen In Perl 5.7 But Gone Now
2993 C<Time::Piece> (previously known as C<Time::Object>) was removed
2994 because it was felt that it didn't have enough value in it to be a
2995 core module. It is still a useful module, though, and is available
2998 =head1 Reporting Bugs
3000 If you find what you think is a bug, you might check the articles
3001 recently posted to the comp.lang.perl.misc newsgroup and the perl
3002 bug database at http://bugs.perl.org/ There may also be
3003 information at http://www.perl.com/ , the Perl Home Page.
3005 If you believe you have an unreported bug, please run the B<perlbug>
3006 program included with your release. Be sure to trim your bug down
3007 to a tiny but sufficient test case. Your bug report, along with the
3008 output of C<perl -V>, will be sent off to perlbug@perl.org to be
3009 analysed by the Perl porting team.
3013 The F<Changes> file for exhaustive details on what changed.
3015 The F<INSTALL> file for how to build Perl.
3017 The F<README> file for general stuff.
3019 The F<Artistic> and F<Copying> files for copyright information.
3023 Written by Jarkko Hietaniemi <F<jhi@iki.fi>>.