3 perldelta - what is new for perl v5.8.0
7 This document describes differences between the 5.6.0 release and
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
12 coordinated (while 5.8.0 was still called 5.7.something).
14 Changes that were integrated into the 5.6.1 release are marked C<[561]>.
15 Many of these changes have been further developed since 5.6.1 was released,
16 those are marked C<[561+]>.
18 You can see the list of changes in the 5.6.1 release (both from the
19 5.005_03 release and the 5.6.0 release) by reading L<perl561delta>.
21 =head1 Highlights In 5.8.0
27 Better Unicode support
31 New Thread Implementation
39 Better Numeric Accuracy
47 More Extensive Regression Testing
51 =head1 Incompatible Changes
53 =head2 Binary Incompatibility
55 B<Perl 5.8 is not binary compatible with earlier releases of Perl.>
57 B<You have to recompile your XS modules.>
59 (Pure Perl modules should continue to work.)
61 The major reason for the discontinuity is the new IO architecture
62 called PerlIO. PerlIO is the default configuration because without
63 it many new features of Perl 5.8 cannot be used. In other words:
64 you just have to recompile your modules containing XS code, sorry
67 In future releases of Perl, non-PerlIO aware XS modules may become
68 completely unsupported. This shouldn't be too difficult for module
69 authors, however: PerlIO has been designed as a drop-in replacement
70 (at the source code level) for the stdio interface.
72 Depending on your platform, there are also other reasons why
73 we decided to break binary compatibility, please read on.
75 =head2 64-bit platforms and malloc
77 If your pointers are 64 bits wide, the Perl malloc is no longer being
78 used because it does not work well with 8-byte pointers. Also,
79 usually the system mallocs on such platforms are much better optimized
80 for such large memory models than the Perl malloc. Some memory-hungry
81 Perl applications like the PDL don't work well with Perl's malloc.
82 Finally, other applications than Perl (such as mod_perl) tend to prefer
83 the system malloc. Such platforms include Alpha and 64-bit HPPA,
86 =head2 AIX Dynaloading
88 The AIX dynaloading now uses in AIX releases 4.3 and newer the native
89 dlopen interface of AIX instead of the old emulated interface. This
90 change will probably break backward compatibility with compiled
91 modules. The change was made to make Perl more compliant with other
92 applications like mod_perl which are using the AIX native interface.
94 =head2 Attributes for C<my> variables now handled at run-time.
96 The C<my EXPR : ATTRS> syntax now applies variable attributes at
97 run-time. (Subroutine and C<our> variables still get attributes applied
98 at compile-time.) See L<attributes> for additional details. In particular,
99 however, this allows variable attributes to be useful for C<tie> interfaces,
100 which was a deficiency of earlier releases. Note that the new semantics
101 doesn't work with the Attribute::Handlers module (as of version 0.76).
103 =head2 Socket Extension Dynamic in VMS
105 The Socket extension is now dynamically loaded instead of being
106 statically built in. This may or may not be a problem with ancient
107 TCP/IP stacks of VMS: we do not know since we weren't able to test
108 Perl in such configurations.
110 =head2 IEEE-format Floating Point Default on OpenVMS Alpha
112 Perl now uses IEEE format (T_FLOAT) as the default internal floating
113 point format on OpenVMS Alpha, potentially breaking binary compatibility
114 with external libraries or existing data. G_FLOAT is still available as
115 a configuration option. The default on VAX (D_FLOAT) has not changed.
117 =head2 New Unicode Properties
119 Unicode I<scripts> are now supported. Scripts are similar to (and superior
120 to) Unicode I<blocks>. The difference between scripts and blocks is that
121 scripts are the glyphs used by a language or a group of languages, while
122 the blocks are more artificial groupings of (mostly) 256 characters based
123 on the Unicode numbering.
125 In general, scripts are more inclusive, but not universally so. For
126 example, while the script C<Latin> includes all the Latin characters and
127 their various diacritic-adorned versions, it does not include the various
128 punctuation or digits (since they are not solely C<Latin>).
130 A number of other properties are now supported, including C<\p{L&}>,
131 C<\p{Any}> C<\p{Assigned}>, C<\p{Unassigned}>, C<\p{Blank}> [561] and
132 C<\p{SpacePerl}> [561] (along with their C<\P{...}> versions, of course).
133 See L<perlunicode> for details, and more additions.
135 The C<In> or C<Is> prefix to names used with the C<\p{...}> and C<\P{...}>
136 are now almost always optional. The only exception is that a C<In> prefix
137 is required to signify a Unicode block when a block name conflicts with a
138 script name. For example, C<\p{Tibetan}> refers to the script, while
139 C<\p{InTibetan}> refers to the block. When there is no name conflict, you
140 can omit the C<In> from the block name (e.g. C<\p{BraillePatterns}>), but
141 to be safe, it's probably best to always use the C<In>).
143 =head2 REF(...) Instead Of SCALAR(...)
145 A reference to a reference now stringifies as "REF(0x81485ec)" instead
146 of "SCALAR(0x81485ec)" in order to be more consistent with the return
149 =head2 pack/unpack D/F recycled
151 The undocumented pack/unpack template letters D/F have been recycled
152 for better use: now they stand for long double (if supported by the
153 platform) and NV (Perl internal floating point type). (They used
154 to be aliases for d/f, but you never knew that.)
162 The semantics of bless(REF, REF) were unclear and until someone proves
163 it to make some sense, it is forbidden.
167 The obsolete chat2 library that should never have been allowed
168 to escape the laboratory has been decommissioned.
172 The builtin dump() function has probably outlived most of its
173 usefulness. The core-dumping functionality will remain in future
174 available as an explicit call to C<CORE::dump()>, but in future
175 releases the behaviour of an unqualified C<dump()> call may change.
179 The very dusty examples in the eg/ directory have been removed.
180 Suggestions for new shiny examples welcome but the main issue is that
181 the examples need to be documented, tested and (most importantly)
186 The (bogus) escape sequences \8 and \9 now give an optional warning
187 ("Unrecognized escape passed through"). There is no need to \-escape
192 The list of filenames from glob() (or <...>) is now by default sorted
193 alphabetically to be csh-compliant (which is what happened before
194 in most UNIX platforms). (bsd_glob() does still sort platform
195 natively, ASCII or EBCDIC, unless GLOB_ALPHASORT is specified.) [561]
199 Spurious syntax errors generated in certain situations, when glob()
200 caused File::Glob to be loaded for the first time, have been fixed. [561]
204 Although "you shouldn't do that", it was possible to write code that
205 depends on Perl's hashed key order (Data::Dumper does this). The new
206 algorithm "One-at-a-Time" produces a different hashed key order.
207 More details are in L</"Performance Enhancements">.
211 lstat(FILEHANDLE) now gives a warning because the operation makes no sense.
212 In future releases this may become a fatal error.
216 The C<package;> syntax (C<package> without an argument) has been
217 deprecated. Its semantics were never that clear and its
218 implementation even less so. If you have used that feature to
219 disallow all but fully qualified variables, C<use strict;> instead.
223 The unimplemented POSIX regex features [[.cc.]] and [[=c=]] are still
224 recognised but now cause fatal errors. The previous behaviour of
225 ignoring them by default and warning if requested was unacceptable
226 since it, in a way, falsely promised that the features could be used.
230 The current user-visible implementation of pseudo-hashes (the weird
231 use of the first array element) is deprecated starting from Perl 5.8.0
232 and will be removed in Perl 5.10.0, and the feature will be
233 implemented differently. Not only is the current interface rather
234 ugly, but the current implementation slows down normal array and hash
235 use quite noticeably. The C<fields> pragma interface will remain
236 available. The I<restricted hashes> interface is expected to
237 be the replacement interface (see L<Hash::Util>).
241 The syntaxes C<< @a->[...] >> and C<< %h->{...} >> have now been deprecated.
245 After years of trying, suidperl is considered to be too complex to
246 ever be considered truly secure. The suidperl functionality is likely
247 to be removed in a future release.
251 The 5.005 threads model (module C<Thread>) is deprecated and expected
252 to be removed in Perl 5.10. Multithreaded code should be migrated to
253 the new ithreads model (see L<threads>, L<threads::shared> and
258 The long deprecated uppercase aliases for the string comparison
259 operators (EQ, NE, LT, LE, GE, GT) have now been removed.
263 The tr///C and tr///U features have been removed and will not return;
264 the interface was a mistake. Sorry about that. For similar
265 functionality, see pack('U0', ...) and pack('C0', ...). [561]
269 Earlier Perls treated "sub foo (@bar)" as equivalent to "sub foo (@)".
270 The prototypes are now checked better at compile-time for invalid
271 syntax. An optional warning is generated ("Illegal character in
272 prototype...") but this may be upgraded to a fatal error in a future
277 The existing behaviour when localising tied arrays and hashes is wrong,
278 and will be changed in a future release, so do not rely on the existing
279 behaviour. See L<"Localising Tied Arrays and Hashes Is Broken">.
283 =head1 Core Enhancements
285 =head2 PerlIO is Now The Default
291 IO is now by default done via PerlIO rather than system's "stdio".
292 PerlIO allows "layers" to be "pushed" onto a file handle to alter the
293 handle's behaviour. Layers can be specified at open time via 3-arg
296 open($fh,'>:crlf :utf8', $path) || ...
298 or on already opened handles via extended C<binmode>:
300 binmode($fh,':encoding(iso-8859-7)');
302 The built-in layers are: unix (low level read/write), stdio (as in
303 previous Perls), perlio (re-implementation of stdio buffering in a
304 portable manner), crlf (does CRLF <=> "\n" translation as on Win32,
305 but available on any platform). A mmap layer may be available if
306 platform supports it (mostly UNIXes).
308 Layers to be applied by default may be specified via the 'open' pragma.
310 See L</"Installation and Configuration Improvements"> for the effects
311 of PerlIO on your architecture name.
315 File handles can be marked as accepting Perl's internal encoding of Unicode
316 (UTF-8 or UTF-EBCDIC depending on platform) by a pseudo layer ":utf8" :
318 open($fh,">:utf8","Uni.txt");
320 Note for EBCDIC users: the pseudo layer ":utf8" is erroneously named
321 for you since it's not UTF-8 what you will be getting but instead
322 UTF-EBCDIC. See L<perlunicode>, L<utf8>, and
323 http://www.unicode.org/unicode/reports/tr16/ for more information.
324 In future releases this naming may change.
328 File handles can translate character encodings from/to Perl's internal
329 Unicode form on read/write via the ":encoding()" layer.
333 File handles can be opened to "in memory" files held in Perl scalars via:
335 open($fh,'>', \$variable) || ...
339 Anonymous temporary files are available without need to
340 'use FileHandle' or other module via
342 open($fh,"+>", undef) || ...
344 That is a literal undef, not an undefined value.
348 The list form of C<open> is now implemented for pipes (at least on UNIX):
350 open($fh,"-|", 'cat', '/etc/motd')
352 creates a pipe, and runs the equivalent of exec('cat', '/etc/motd') in
357 If your locale environment variables (LANGUAGE, LC_ALL, LC_CTYPE, LANG)
358 contain the strings 'UTF-8' or 'UTF8' (case-insensitive matching),
359 the default encoding of your STDIN, STDOUT, and STDERR, and of
360 B<any subsequent file open>, is UTF-8.
364 =head2 Restricted Hashes
366 A restricted hash is restricted to a certain set of keys, no keys
367 outside the set can be added. Also individual keys can be restricted
368 so that the key cannot be deleted and the value cannot be changed.
369 No new syntax is involved: the Hash::Util module is the interface.
373 Perl used to be fragile in that signals arriving at inopportune moments
374 could corrupt Perl's internal state. Now Perl postpones handling of
375 signals until it's safe (between opcodes).
377 This change may have surprising side effects because signals no longer
378 interrupt Perl instantly. Perl will now first finish whatever it was
379 doing, like finishing an internal operation (like sort()) or an
380 external operation (like an I/O operation), and only then look at any
381 arrived signals (and before starting the next operation). No more corrupt
382 internal state since the current operation is always finished first,
383 but the signal may take more time to get heard. Note that breaking
384 out from potentially blocking operations should still work, though.
386 =head2 Unicode Overhaul
388 Unicode in general should be now much more usable than in Perl 5.6.0
389 (or even in 5.6.1). Unicode can be used in hash keys, Unicode in
390 regular expressions should work now, Unicode in tr/// should work now,
391 Unicode in I/O should work now. See L<perluniintro> for introduction
392 and L<perlunicode> for details.
398 The Unicode Character Database coming with Perl has been upgraded
399 to Unicode 3.2.0. For more information, see http://www.unicode.org/ .
400 [561+] (5.6.1 has UCD 3.0.1.)
404 For developers interested in enhancing Perl's Unicode capabilities:
405 almost all the UCD files are included with the Perl distribution in
406 the F<lib/unicore> subdirectory. The most notable omission, for space
407 considerations, is the Unihan database.
411 The properties \p{Blank} and \p{SpacePerl} have been added. "Blank" is like
412 C isblank(), that is, it contains only "horizontal whitespace" (the space
413 character is, the newline isn't), and the "SpacePerl" is the Unicode
414 equivalent of C<\s> (\p{Space} isn't, since that includes the vertical
415 tabulator character, whereas C<\s> doesn't.)
417 See "New Unicode Properties" earlier in this document for additional
418 information on changes with Unicode properties.
422 =head2 Understanding of Numbers
424 In general a lot of fixing has happened in the area of Perl's
425 understanding of numbers, both integer and floating point. Since in
426 many systems the standard number parsing functions like C<strtoul()>
427 and C<atof()> seem to have bugs, Perl tries to work around their
428 deficiencies. This results hopefully in more accurate numbers.
430 Perl now tries internally to use integer values in numeric conversions
431 and basic arithmetics (+ - * /) if the arguments are integers, and
432 tries also to keep the results stored internally as integers.
433 This change leads to often slightly faster and always less lossy
434 arithmetics. (Previously Perl always preferred floating point numbers
437 =head2 Arrays now always interpolate into double-quoted strings [561]
439 In double-quoted strings, arrays now interpolate, no matter what. The
440 behavior in earlier versions of perl 5 was that arrays would interpolate
441 into strings if the array had been mentioned before the string was
442 compiled, and otherwise Perl would raise a fatal compile-time error.
443 In versions 5.000 through 5.003, the error was
445 Literal @example now requires backslash
447 In versions 5.004_01 through 5.6.0, the error was
449 In string, @example now must be written as \@example
451 The idea here was to get people into the habit of writing
452 C<"fred\@example.com"> when they wanted a literal C<@> sign, just as
453 they have always written C<"Give me back my \$5"> when they wanted a
456 Starting with 5.6.1, when Perl now sees an C<@> sign in a
457 double-quoted string, it I<always> attempts to interpolate an array,
458 regardless of whether or not the array has been used or declared
459 already. The fatal error has been downgraded to an optional warning:
461 Possible unintended interpolation of @example in string
463 This warns you that C<"fred@example.com"> is going to turn into
464 C<fred.com> if you don't backslash the C<@>.
465 See http://www.plover.com/~mjd/perl/at-error.html for more details
466 about the history here.
468 =head2 Miscellaneous Changes
474 AUTOLOAD is now lvaluable, meaning that you can add the :lvalue attribute
475 to AUTOLOAD subroutines and you can assign to the AUTOLOAD return value.
479 The $Config{byteorder} (and corresponding BYTEORDER in config.h) was
480 previously wrong in platforms if sizeof(long) was 4, but sizeof(IV)
481 was 8. The byteorder was only sizeof(long) bytes long (1234 or 4321),
482 but now it is correctly sizeof(IV) bytes long, (12345678 or 87654321).
483 (This problem didn't affect Windows platforms.)
485 Also, $Config{byteorder} is now computed dynamically--this is more
486 robust with "fat binaries" where an executable image contains binaries
487 for more than one binary platform, and when cross-compiling.
491 C<perl -d:Module=arg,arg,arg> now works (previously one couldn't pass
492 in multiple arguments.)
496 C<do> followed by a bareword now ensures that this bareword isn't
497 a keyword (to avoid a bug where C<do q(foo.pl)> tried to call
498 subroutine called C<q>). This means that for example instead of
499 C<do format()> you must write C<do &format()>.
503 The builtin dump() now gives an optional warning
504 C<dump() better written as CORE::dump()>,
505 meaning that by default C<dump(...)> is resolved as the builtin
506 dump() which dumps core and aborts, not as (possibly) user-defined
507 C<sub dump>. To call the latter, qualify the call as C<&dump(...)>.
508 (The whole dump() feature is to considered deprecated, and possibly
509 removed/changed in future releases.)
513 chomp() and chop() are now overridable. Note, however, that their
514 prototype (as given by C<prototype("CORE::chomp")> is undefined,
515 because it cannot be expressed and therefore one cannot really write
516 replacements to override these builtins.
520 END blocks are now run even if you exit/die in a BEGIN block.
521 Internally, the execution of END blocks is now controlled by
522 PL_exit_flags & PERL_EXIT_DESTRUCT_END. This enables the new
523 behaviour for Perl embedders. This will default in 5.10. See
528 Formats now support zero-padded decimal fields.
532 Lvalue subroutines can now return C<undef> in list context. However,
533 the lvalue subroutine feature still remains experimental. [561+]
537 A lost warning "Can't declare ... dereference in my" has been
538 restored (Perl had it earlier but it became lost in later releases.)
542 A new special regular expression variable has been introduced:
543 C<$^N>, which contains the most-recently closed group (submatch).
547 C<no Module;> does not produce an error even if Module does not have an
548 unimport() method. This parallels the behavior of C<use> vis-a-vis
553 The numerical comparison operators return C<undef> if either operand
554 is a NaN. Previously the behaviour was unspecified.
558 The following builtin functions are now overridable: each(), keys(),
559 pop(), push(), shift(), splice(), unshift(). [561]
563 C<pack() / unpack()> can now group template letters with C<()> and then
564 apply repetition/count modifiers on the groups.
568 C<pack() / unpack()> can now process the Perl internal numeric types:
569 IVs, UVs, NVs-- and also long doubles, if supported by the platform.
570 The template letters are C<j>, C<J>, C<F>, and C<D>.
574 C<pack('U0a*', ...)> can now be used to force a string to UTF8.
578 my __PACKAGE__ $obj now works. [561]
582 POSIX::sleep() now returns the number of I<unslept> seconds
583 (as the POSIX standard says), as opposed to CORE::sleep() which
584 returns the number of slept seconds.
588 The printf() and sprintf() now support parameter reordering using the
589 C<%\d+\$> and C<*\d+\$> syntaxes. For example
591 print "%2\$s %1\$s\n", "foo", "bar";
593 will print "bar foo\n". This feature helps in writing
594 internationalised software, and in general when the order
595 of the parameters can vary.
599 The (\&) prototype now works properly. [561]
603 prototype(\[$@%&]) is now available to implicitly create references
604 (useful for example if you want to emulate the tie() interface).
608 A new command-line option, C<-t> is available. It is the
609 little brother of C<-T>: instead of dying on taint violations,
610 lexical warnings are given. B<This is only meant as a temporary
611 debugging aid while securing the code of old legacy applications.
612 This is not a substitute for -T.>
616 In other taint news, the C<exec LIST> and C<system LIST> have now been
617 considered too risky (think C<exec @ARGV>: it can start any program
618 with any arguments), and now the said forms cause a warning.
619 You should carefully launder the arguments to guarantee their
620 validity. In future releases of Perl the forms will become fatal
621 errors so consider starting laundering now.
625 Tied hash interfaces are now required to have the EXISTS and DELETE
626 methods (either own or inherited).
630 If tr/// is just counting characters, it doesn't attempt to
635 untie() will now call an UNTIE() hook if it exists. See L<perltie>
640 L<utime> now supports C<utime undef, undef, @files> to change the
641 file timestamps to the current time.
645 The rules for allowing underscores (underbars) in numeric constants
646 have been relaxed and simplified: now you can have an underscore
647 simply B<between digits>.
651 Rather than relying on C's argv[0] (which may not contain a full pathname)
652 where possible $^X is now set by asking the operating system.
653 (eg by reading F</proc/self/exe> on Linux, F</proc/curproc/file> on FreeBSD)
657 A new variable, C<${^TAINT}>, indicates whether taint mode is enabled.
661 You can now override the readline() builtin, and this overrides also
662 the <FILEHANDLE> angle bracket operator.
666 The command-line options -s and -F are now recognized on the shebang
671 Use of the C</c> match modifier without an accompanying C</g> modifier
672 elicits a new warning: C<Use of /c modifier is meaningless without /g>.
674 Use of C</c> in substitutions, even with C</g>, elicits
675 C<Use of /c modifier is meaningless in s///>.
677 Use of C</g> with C<split> elicits C<Use of /g modifier is meaningless
682 =head1 Modules and Pragmata
684 =head2 New Modules and Pragmata
690 C<Attribute::Handlers> allows a class to define attribute handlers.
693 use Attribute::Handlers;
694 sub Wolf :ATTR(SCALAR) { print "howl!\n" }
696 # later, in some package using or inheriting from MyPack...
698 my MyPack $Fluffy : Wolf; # the attribute handler Wolf will be called
700 Both variables and routines can have attribute handlers. Handlers can
701 be specific to type (SCALAR, ARRAY, HASH, or CODE), or specific to the
702 exact compilation phase (BEGIN, CHECK, INIT, or END).
703 See L<Attribute::Handlers>.
707 C<B::Concise>, by Stephen McCamant, is a new compiler backend for
708 walking the Perl syntax tree, printing concise info about ops.
709 The output is highly customisable. See L<B::Concise>. [561+]
713 The new bignum, bigint, and bigrat pragmas, by Tels, implement
714 transparent bignum support (using the Math::BigInt, Math::BigFloat,
715 and Math::BigRat backends).
719 C<Class::ISA>, by Sean Burke, is a module for reporting the search
720 path for a class's ISA tree. See L<Class::ISA>.
724 C<Cwd> now has a split personality: if possible, an XS extension is
725 used, (this will hopefully be faster, more secure, and more robust)
726 but if not possible, the familiar Perl implementation is used.
730 C<Devel::PPPort>, originally by Kenneth Albanowski and now
731 maintained by Paul Marquess, has been added. It is primarily used
732 by C<h2xs> to enhance portability of XS modules between different
733 versions of Perl. See L<Devel::PPPort>.
737 C<Digest>, frontend module for calculating digests (checksums), from
738 Gisle Aas, has been added. See L<Digest>.
742 C<Digest::MD5> for calculating MD5 digests (checksums) as defined in
743 RFC 1321, from Gisle Aas, has been added. See L<Digest::MD5>.
745 use Digest::MD5 'md5_hex';
747 $digest = md5_hex("Thirsty Camel");
749 print $digest, "\n"; # 01d19d9d2045e005c3f1b80e8b164de1
751 NOTE: the C<MD5> backward compatibility module is deliberately not
752 included since its further use is discouraged.
756 C<Encode>, originally by Nick Ing-Simmons and now maintained by Dan
757 Kogai, provides a mechanism to translate between different character
758 encodings. Support for Unicode, ISO-8859-1, and ASCII are compiled in
759 to the module. Several other encodings (like the rest of the
760 ISO-8859, CP*/Win*, Mac, KOI8-R, three variants EBCDIC, Chinese,
761 Japanese, and Korean encodings) are included and can be loaded at
762 runtime. (For space considerations, the largest Chinese encodings
763 have been separated into their own CPAN module, Encode::HanExtra,
764 which Encode will use if available). See L<Encode>.
766 Any encoding supported by Encode module is also available to the
767 ":encoding()" layer if PerlIO is used.
771 C<Hash::Util> is the interface to the new I<restricted hashes>
772 feature. (Implemented by Jeffrey Friedl, Nick Ing-Simmons, and
773 Michael Schwern.) See L<Hash::Util>.
777 C<I18N::Langinfo> can be used to query locale information.
778 See L<I18N::Langinfo>.
782 C<I18N::LangTags>, by Sean Burke, has functions for dealing with
783 RFC3066-style language tags. See L<I18N::LangTags>.
787 C<ExtUtils::Constant>, by Nicholas Clark, is a new tool for extension
788 writers for generating XS code to import C header constants.
789 See L<ExtUtils::Constant>.
793 C<Filter::Simple>, by Damian Conway, is an easy-to-use frontend to
794 Filter::Util::Call. See L<Filter::Simple>.
800 use Filter::Simple sub {
801 while (my ($from, $to) = splice @_, 0, 2) {
810 use MyFilter qr/red/ => 'green';
812 print "red\n"; # this code is filtered, will print "green\n"
813 print "bored\n"; # this code is filtered, will print "bogreen\n"
817 print "red\n"; # this code is not filtered, will print "red\n"
821 C<File::Temp>, by Tim Jenness, allows one to create temporary files
822 and directories in an easy, portable, and secure way. See L<File::Temp>.
827 C<Filter::Util::Call>, by Paul Marquess, provides you with the
828 framework to write I<source filters> in Perl. For most uses, the
829 frontend Filter::Simple is to be preferred. See L<Filter::Util::Call>.
833 C<if>, by Ilya Zakharevich, is a new pragma for conditional inclusion
838 L<libnet>, by Graham Barr, is a collection of perl5 modules related
839 to network programming. See L<Net::FTP>, L<Net::NNTP>, L<Net::Ping>
840 (not part of libnet, but related), L<Net::POP3>, L<Net::SMTP>,
843 Perl installation leaves libnet unconfigured; use F<libnetcfg>
848 C<List::Util>, by Graham Barr, is a selection of general-utility
849 list subroutines, such as sum(), min(), first(), and shuffle().
854 C<Locale::Constants>, C<Locale::Country>, C<Locale::Currency>
855 C<Locale::Language>, and L<Locale::Script>, by Neil Bowers, have
856 been added. They provide the codes for various locale standards, such
857 as "fr" for France, "usd" for US Dollar, and "ja" for Japanese.
861 $country = code2country('jp'); # $country gets 'Japan'
862 $code = country2code('Norway'); # $code gets 'no'
864 See L<Locale::Constants>, L<Locale::Country>, L<Locale::Currency>,
865 and L<Locale::Language>.
869 C<Locale::Maketext>, by Sean Burke, is a localization framework. See
870 L<Locale::Maketext>, and L<Locale::Maketext::TPJ13>. The latter is an
871 article about software localization, originally published in The Perl
872 Journal #13, and republished here with kind permission.
876 C<Math::BigRat> for big rational numbers, to accompany Math::BigInt and
877 Math::BigFloat, from Tels. See L<Math::BigRat>.
881 C<Memoize> can make your functions faster by trading space for time,
882 from Mark-Jason Dominus. See L<Memoize>.
886 C<MIME::Base64>, by Gisle Aas, allows you to encode data in base64,
887 as defined in RFC 2045 - I<MIME (Multipurpose Internet Mail
892 $encoded = encode_base64('Aladdin:open sesame');
893 $decoded = decode_base64($encoded);
895 print $encoded, "\n"; # "QWxhZGRpbjpvcGVuIHNlc2FtZQ=="
901 C<MIME::QuotedPrint>, by Gisle Aas, allows you to encode data
902 in quoted-printable encoding, as defined in RFC 2045 - I<MIME
903 (Multipurpose Internet Mail Extensions)>.
905 use MIME::QuotedPrint;
907 $encoded = encode_qp("Smiley in Unicode: \x{263a}");
908 $decoded = decode_qp($encoded);
910 print $encoded, "\n"; # "Smiley in Unicode: =263A"
912 MIME::QuotedPrint has been enhanced to provide the basic methods
913 necessary to use it with PerlIO::Via as in :
915 use MIME::QuotedPrint;
916 open($fh,">Via(MIME::QuotedPrint)",$path);
918 See L<MIME::QuotedPrint>.
922 C<NEXT>, by Damian Conway, is a pseudo-class for method redispatch.
927 C<open> is a new pragma for setting the default I/O disciplines
932 C<PerlIO::Scalar>, by Nick Ing-Simmons, provides the implementation
933 of IO to "in memory" Perl scalars as discussed above. It also serves
934 as an example of a loadable PerlIO layer. Other future possibilities
935 include PerlIO::Array and PerlIO::Code. See L<PerlIO::Scalar>.
939 C<PerlIO::Via>, by Nick Ing-Simmons, acts as a PerlIO layer and wraps
940 PerlIO layer functionality provided by a class (typically implemented
943 use MIME::QuotedPrint;
944 open($fh,">Via(MIME::QuotedPrint)",$path);
946 This will automatically convert everything output to C<$fh>
947 to Quoted-Printable. See L<PerlIO::Via>.
951 C<Pod::ParseLink>, by Russ Allbery, has been added,
952 to parse LZ<><> links in pods as described in the new
957 C<Pod::Text::Overstrike>, by Joe Smith, has been added.
958 It converts POD data to formatted overstrike text.
959 See L<Pod::Text::Overstrike>. [561+]
963 C<Scalar::Util> is a selection of general-utility scalar subroutines,
964 such as blessed(), reftype(), and tainted(). See L<Scalar::Util>.
968 C<sort> is a new pragma for controlling the behaviour of sort().
972 C<Storable> gives persistence to Perl data structures by allowing the
973 storage and retrieval of Perl data to and from files in a fast and
974 compact binary format. Because in effect Storable does serialisation
975 of Perl data structues, with it you can also clone deep, hierarchical
976 datastructures. Storable was originally created by Raphael Manfredi,
977 but it is now maintained by Abhijit Menon-Sen. Storable has been
978 enhanced to understand the two new hash features, Unicode keys and
979 restricted hashes. See L<Storable>.
983 C<Switch>, by Damian Conway, has been added. Just by saying
987 you have C<switch> and C<case> available in Perl.
993 case 1 { print "number 1" }
994 case "a" { print "string a" }
995 case [1..10,42] { print "number in list" }
996 case (@array) { print "number in list" }
997 case /\w+/ { print "pattern" }
998 case qr/\w+/ { print "pattern" }
999 case (%hash) { print "entry in hash" }
1000 case (\%hash) { print "entry in hash" }
1001 case (\&sub) { print "arg to subroutine" }
1002 else { print "previous case not true" }
1009 C<Test::More>, by Michael Schwern, is yet another framework for writing
1010 test scripts, more extensive than Test::Simple. See L<Test::More>.
1014 C<Test::Simple>, by Michael Schwern, has basic utilities for writing
1015 tests. See L<Test::Simple>.
1019 C<Text::Balanced>, by Damian Conway, has been added, for extracting
1020 delimited text sequences from strings.
1022 use Text::Balanced 'extract_delimited';
1024 ($a, $b) = extract_delimited("'never say never', he never said", "'", '');
1026 $a will be "'never say never'", $b will be ', he never said'.
1028 In addition to extract_delimited(), there are also extract_bracketed(),
1029 extract_quotelike(), extract_codeblock(), extract_variable(),
1030 extract_tagged(), extract_multiple(), gen_delimited_pat(), and
1031 gen_extract_tagged(). With these, you can implement rather advanced
1032 parsing algorithms. See L<Text::Balanced>.
1036 C<threads>, by Arthur Bergman, is an interface to interpreter threads.
1037 Interpreter threads (ithreads) is the new thread model introduced in
1038 Perl 5.6 but only available as an internal interface for extension
1039 writers (and for Win32 Perl for C<fork()> emulation). See L<threads>,
1040 L<threads::shared>, and L<perlthrtut>.
1044 C<threads::shared>, by Arthur Bergman, allows data sharing for
1045 interpreter threads. In the ithreads model any data sharing between
1046 threads must be explicit, as opposed to the old 5.005 thread model
1047 where data sharing was implicit. See L<threads::shared>.
1051 C<Tie::File>, by Mark-Jason Dominus, associates a Perl array with the
1052 lines of a file. See L<Tie::File>.
1056 C<Tie::Memoize>, by Ilya Zakharevich, provides on-demand loaded hashes.
1057 See L<Tie::Memoize>.
1061 C<Tie::RefHash::Nestable>, by Edward Avis, allows storing hash
1062 references (unlike the standard Tie::RefHash) The module is contained
1063 within Tie::RefHash. See L<Tie::RefHash>.
1067 C<Time::HiRes>, by Douglas E. Wegscheid, provides high resolution
1068 timing (ualarm, usleep, and gettimeofday). See L<Time::HiRes>.
1072 C<Unicode::UCD> offers a querying interface to the Unicode Character
1073 Database. See L<Unicode::UCD>.
1077 C<Unicode::Collate>, by SADAHIRO Tomoyuki, implements the UCA
1078 (Unicode Collation Algorithm) for sorting Unicode strings.
1079 See L<Unicode::Collate>.
1083 C<Unicode::Normalize>, by SADAHIRO Tomoyuki, implements the various
1084 Unicode normalization forms. See L<Unicode::Normalize>.
1088 C<XS::Typemap>, by Tim Jenness, is a test extension that exercises
1089 XS typemaps. Nothing gets installed, but the code is worth studying
1090 for extension writers.
1094 =head2 Updated And Improved Modules and Pragmata
1100 The following independently supported modules have been updated to the
1101 newest versions from CPAN: CGI, CPAN, DB_File, File::Spec, File::Temp,
1102 Getopt::Long, Math::BigFloat, Math::BigInt, the podlators bundle
1103 (Pod::Man, Pod::Text), Pod::LaTeX [561+], Pod::Parser, Storable,
1104 Term::ANSIColor, Test, Text-Tabs+Wrap.
1108 attributes::reftype() now works on tied arguments.
1112 AutoLoader can now be disabled with C<no AutoLoader;>.
1116 B::Deparse has been significantly enhanced by Robin Houston. It can
1117 now deparse almost all of the standard test suite (so that the tests
1118 still succeed). There is a make target "test.deparse" for trying this
1123 Carp now has better interface documentation, and the @CARP_NOT
1124 interface has been added to get optional control over where errors
1125 are reported independently of @ISA, by Ben Tilly.
1129 Class::Struct can now define the classes in compile time.
1133 Class::Struct now assigns the array/hash element if the accessor
1134 is called with an array/hash element as the B<sole> argument.
1138 The return value of Cwd::fastcwd() is now tainted.
1142 Data::Dumper now has an option to sort hashes.
1146 Data::Dumper now has an option to dump code references
1151 DB_File now supports newer Berkeley DB versions, among
1156 Devel::Peek now has an interface for the Perl memory statistics
1157 (this works only if you are using perl's malloc, and if you have
1158 compiled with debugging).
1162 The English module can now be used without the infamous performance
1165 use English '-no_match_vars';
1167 (Assuming, of course, that you don't need the troublesome variables
1168 C<$`>, C<$&>, or C<$'>.) Also, introduced C<@LAST_MATCH_START> and
1169 C<@LAST_MATCH_END> English aliases for C<@-> and C<@+>.
1173 ExtUtils::MakeMaker now uses File::Spec internally, which hopefully
1174 leads to better portability.
1178 Fcntl, Socket, and Sys::Syslog have been rewritten by Nicholas Clark
1179 to use the new-style constant dispatch section (see L<ExtUtils::Constant>).
1180 This means that they will be more robust and hopefully faster.
1184 File::Find now chdir()s correctly when chasing symbolic links. [561]
1188 File::Find now has pre- and post-processing callbacks. It also
1189 correctly changes directories when chasing symbolic links. Callbacks
1190 (naughtily) exiting with "next;" instead of "return;" now work.
1194 File::Find is now (again) reentrant. It also has been made
1199 The warnings issued by File::Find now belong to their own category.
1200 You can enable/disable them with C<use/no warnings 'File::Find';>.
1204 File::Glob::glob() has been renamed to File::Glob::bsd_glob()
1205 because the name clashes with the builtin glob(). The older
1206 name is still available for compatibility, but is deprecated. [561]
1210 File::Glob now supports C<GLOB_LIMIT> constant to limit the size of
1211 the returned list of filenames.
1215 IPC::Open3 now allows the use of numeric file descriptors.
1219 IO::Socket now has an atmark() method, which returns true if the socket
1220 is positioned at the out-of-band mark. The method is also exportable
1221 as a sockatmark() function.
1225 IO::Socket::INET failed to open the specified port if the service name
1226 was not known. It now correctly uses the supplied port number as is. [561]
1230 IO::Socket::INET has support for the ReusePort option (if your
1231 platform supports it). The Reuse option now has an alias, ReuseAddr.
1232 For clarity, you may want to prefer ReuseAddr.
1236 IO::Socket::INET now supports a value of zero for C<LocalPort>
1237 (usually meaning that the operating system will make one up.)
1241 'use lib' now works identically to @INC. Removing directories
1242 with 'no lib' now works.
1246 Math::BigFloat and Math::BigInt have undergone a full rewrite by Tels.
1247 They are now magnitudes faster, and they support various bignum
1248 libraries such as GMP and PARI as their backends.
1252 Math::Complex handles inf, NaN etc., better.
1256 Net::Ping has been considerably enhanced by Rob Brown: multihoming is
1257 now supported, Win32 functionality is better, there is now time
1258 measuring functionality (optionally high-resolution using
1259 Time::HiRes), and there is now "external" protocol which uses
1260 Net::Ping::External module which runs your external ping utility and
1261 parses the output. A version of Net::Ping::External is available in
1264 Note that some of the Net::Ping tests are disabled when running
1265 under the Perl distribution since one cannot assume one or more
1266 of the following: enabled echo port at localhost, full Internet
1267 connectivity, or sympathetic firewalls. You can set the environment
1268 variable PERL_TEST_Net_Ping to "1" (one) before running the Perl test
1269 suite to enable all the Net::Ping tests.
1273 POSIX::sigaction() is now much more flexible and robust.
1274 You can now install coderef handlers, 'DEFAULT', and 'IGNORE'
1275 handlers, installing new handlers was not atomic.
1279 In Safe, C<%INC> is now localised in a Safe compartment so that
1284 In SDBM_File on dosish platforms, some keys went missing because of
1285 lack of support for files with "holes". A workaround for the problem
1290 In Search::Dict one can now have a pre-processing hook for the
1291 lines being searched.
1295 The Shell module now has an OO interface.
1299 In Sys::Syslog there is now a failover mechanism that will go
1300 through alternative connection mechanisms until the message
1301 is successfully logged.
1305 The Test module has been significantly enhanced.
1309 Time::Local::timelocal() does not handle fractional seconds anymore.
1310 The rationale is that neither does localtime(), and timelocal() and
1311 localtime() are supposed to be inverses of each other.
1315 The vars pragma now supports declaring fully qualified variables.
1316 (Something that C<our()> does not and will not support.)
1320 The C<utf8::> name space (as in the pragma) provides various
1321 Perl-callable functions to provide low level access to Perl's
1322 internal Unicode representation. At the moment only length()
1323 has been implemented.
1327 =head1 Utility Changes
1333 Emacs perl mode (emacs/cperl-mode.el) has been updated to version
1338 F<emacs/e2ctags.pl> is now much faster.
1342 C<enc2xs> is a tool for people adding their own encodings to the
1347 C<h2ph> now supports C trigraphs.
1351 C<h2xs> now produces a template README.
1355 C<h2xs> now uses C<Devel::PPPort> for better portability between
1356 different versions of Perl.
1360 C<h2xs> uses the new L<ExtUtils::Constant|ExtUtils::Constant> module
1361 which will affect newly created extensions that define constants.
1362 Since the new code is more correct (if you have two constants where the
1363 first one is a prefix of the second one, the first constant B<never>
1364 got defined), less lossy (it uses integers for integer constant,
1365 as opposed to the old code that used floating point numbers even for
1366 integer constants), and slightly faster, you might want to consider
1367 regenerating your extension code (the new scheme makes regenerating
1368 easy). L<h2xs> now also supports C trigraphs.
1372 C<libnetcfg> has been added to configure libnet.
1376 C<perlbug> is now much more robust. It also sends the bug report to
1377 perl.org, not perl.com.
1381 C<perlcc> has been rewritten and its user interface (that is,
1382 command line) is much more like that of the UNIX C compiler, cc.
1383 (The perlbc tools has been removed. Use C<perlcc -B> instead.)
1384 B<Note that perlcc is still considered very experimental and
1389 C<perlivp> is a new Installation Verification Procedure utility
1390 for running any time after installing Perl.
1394 C<piconv> is an implementation of the character conversion utility
1395 C<iconv>, demonstrating the new Encode module.
1399 C<pod2html> now allows specifying a cache directory.
1403 C<pod2html> now produces XHTML 1.0.
1407 C<pod2html> now understands POD written using different line endings
1408 (PC-like CRLF versus UNIX-like LF versus MacClassic-like CR).
1412 C<s2p> has been completely rewritten in Perl. (It is in fact a full
1413 implementation of sed in Perl: you can use the sed functionality by
1414 using the C<psed> utility.)
1418 C<xsubpp> now understands POD documentation embedded in the *.xs
1423 C<xsubpp> now supports the OUT keyword.
1427 =head1 New Documentation
1433 perl56delta details the changes between the 5.005 release and the
1438 perlclib documents the internal replacements for standard C library
1439 functions. (Interesting only for extension writers and Perl core
1444 perldebtut is a Perl debugging tutorial. [561+]
1448 perlebcdic contains considerations for running Perl on EBCDIC
1453 perlintro is a gentle introduction to Perl.
1457 perliol documents the internals of PerlIO with layers.
1461 perlmodstyle is a style guide for writing modules.
1465 perlnewmod tells about writing and submitting a new module. [561+]
1469 perlpacktut is a pack() tutorial.
1473 perlpod has been rewritten to be clearer and to record the best
1474 practices gathered over the years.
1478 perlpodspec is a more formal specification of the pod format,
1479 mainly of interest for writers of pod applications, not to
1480 people writing in pod.
1484 perlretut is a regular expression tutorial. [561+]
1488 perlrequick is a regular expressions quick-start guide.
1489 Yes, much quicker than perlretut. [561]
1493 perltodo has been updated.
1497 perltootc has been renamed as perltooc (to not to conflict
1498 with perltoot in filesystems restricted to "8.3" names).
1502 perluniintro is an introduction to using Unicode in Perl.
1503 (perlunicode is more of a detailed reference and background
1508 perlutil explains the command line utilities packaged with the Perl
1509 distribution. [561+]
1513 The following platform-specific documents are available before
1514 the installation as README.I<platform>, and after the installation
1517 perlaix perlamiga perlapollo perlbeos perlbs2000
1518 perlce perlcygwin perldgux perldos perlepoc perlhpux
1519 perlhurd perlmachten perlmacos perlmint perlmpeix
1520 perlnetware perlos2 perlos390 perlplan9 perlqnx perlsolaris
1521 perltru64 perluts perlvmesa perlvms perlvos perlwin32
1523 These documents usually detail one or more of the following subjects:
1524 configuring, building, testing, installing, and sometimes also using
1525 Perl on the said platform.
1527 Eastern Asian Perl users are now welcomed in their own languages:
1528 README.jp (Japanese), README.ko (Korean), README.cn (simplified
1529 Chinese) and README.tw (traditional Chinese), which are written in
1530 normal pod but encoded in EUC-JP, EUC-KR, EUC-CN and Big5. These
1531 will get installed as
1533 perljp perlko perlcn perltw
1539 The documentation for the POSIX-BC platform is called "BS2000", to avoid
1540 confusion with the Perl POSIX module.
1544 The documentation for the WinCE platform is called perlce (README.ce
1545 in the source code kit), to avoid confusion with the perlwin32
1546 documentation on 8.3-restricted filesystems.
1550 =head1 Performance Enhancements
1556 map() could get pathologically slow when the result list it generates
1557 is larger than the source list. The performance has been improved for
1558 common scenarios. [561]
1562 sort() is also fully reentrant, in the sense that the sort function
1563 can itself call sort(). This did not work reliably in previous
1568 sort() has been changed to use primarily mergesort internally as
1569 opposed to the earlier quicksort. For very small lists this may
1570 result in slightly slower sorting times, but in general the speedup
1571 should be at least 20%. Additional bonuses are that the worst case
1572 behaviour of sort() is now better (in computer science terms it now
1573 runs in time O(N log N), as opposed to quicksort's Theta(N**2)
1574 worst-case run time behaviour), and that sort() is now stable
1575 (meaning that elements with identical keys will stay ordered as they
1576 were before the sort). See the C<sort> pragma for information.
1578 The story in more detail: suppose you want to serve yourself a little
1581 @digits = ( 3,1,4,1,5,9 );
1583 A numerical sort of the digits will yield (1,1,3,4,5,9), as expected.
1584 Which C<1> comes first is hard to know, since one C<1> looks pretty
1585 much like any other. You can regard this as totally trivial,
1586 or somewhat profound. However, if you just want to sort the even
1587 digits ahead of the odd ones, then what will
1589 sort { ($a % 2) <=> ($b % 2) } @digits;
1591 yield? The only even digit, C<4>, will come first. But how about
1592 the odd numbers, which all compare equal? With the quicksort algorithm
1593 used to implement Perl 5.6 and earlier, the order of ties is left up
1594 to the sort. So, as you add more and more digits of Pi, the order
1595 in which the sorted even and odd digits appear will change.
1596 and, for sufficiently large slices of Pi, the quicksort algorithm
1597 in Perl 5.8 won't return the same results even if reinvoked with the
1598 same input. The justification for this rests with quicksort's
1599 worst case behavior. If you run
1601 sort { $a <=> $b } ( 1 .. $N , 1 .. $N );
1603 (something you might approximate if you wanted to merge two sorted
1604 arrays using sort), doubling $N doesn't just double the quicksort time,
1605 it I<quadruples> it. Quicksort has a worst case run time that can
1606 grow like N**2, so-called I<quadratic> behaviour, and it can happen
1607 on patterns that may well arise in normal use. You won't notice this
1608 for small arrays, but you I<will> notice it with larger arrays,
1609 and you may not live long enough for the sort to complete on arrays
1610 of a million elements. So the 5.8 quicksort scrambles large arrays
1611 before sorting them, as a statistical defence against quadratic behaviour.
1612 But that means if you sort the same large array twice, ties may be
1613 broken in different ways.
1615 Because of the unpredictability of tie-breaking order, and the quadratic
1616 worst-case behaviour, quicksort was I<almost> replaced completely with
1617 a stable mergesort. I<Stable> means that ties are broken to preserve
1618 the original order of appearance in the input array. So
1620 sort { ($a % 2) <=> ($b % 2) } (3,1,4,1,5,9);
1622 will yield (4,3,1,1,5,9), guaranteed. The even and odd numbers
1623 appear in the output in the same order they appeared in the input.
1624 Mergesort has worst case O(N log N) behaviour, the best value
1625 attainable. And, ironically, this mergesort does particularly
1626 well where quicksort goes quadratic: mergesort sorts (1..$N, 1..$N)
1627 in O(N) time. But quicksort was rescued at the last moment because
1628 it is faster than mergesort on certain inputs and platforms.
1629 For example, if you really I<don't> care about the order of even
1630 and odd digits, quicksort will run in O(N) time; it's very good
1631 at sorting many repetitions of a small number of distinct elements.
1632 The quicksort divide and conquer strategy works well on platforms
1633 with relatively small, very fast, caches. Eventually, the problem gets
1634 whittled down to one that fits in the cache, from which point it
1635 benefits from the increased memory speed.
1637 Quicksort was rescued by implementing a sort pragma to control aspects
1638 of the sort. The B<stable> subpragma forces stable behaviour,
1639 regardless of algorithm. The B<_quicksort> and B<_mergesort>
1640 subpragmas are heavy-handed ways to select the underlying implementation.
1641 The leading C<_> is a reminder that these subpragmas may not survive
1642 beyond 5.8. More appropriate mechanisms for selecting the implementation
1643 exist, but they wouldn't have arrived in time to save quicksort.
1647 Hashes now use Bob Jenkins "One-at-a-Time" hashing key algorithm
1648 ( http://burtleburtle.net/bob/hash/doobs.html ). This algorithm is
1649 reasonably fast while producing a much better spread of values than
1650 the old hashing algorithm (originally by Chris Torek, later tweaked by
1651 Ilya Zakharevich). Hash values output from the algorithm on a hash of
1652 all 3-char printable ASCII keys comes much closer to passing the
1653 DIEHARD random number generation tests. According to perlbench, this
1654 change has not affected the overall speed of Perl.
1658 unshift() should now be noticeably faster.
1662 =head1 Installation and Configuration Improvements
1664 =head2 Generic Improvements
1670 INSTALL now explains how you can configure Perl to use 64-bit
1671 integers even on non-64-bit platforms.
1675 Policy.sh policy change: if you are reusing a Policy.sh file
1676 (see INSTALL) and you use Configure -Dprefix=/foo/bar and in the old
1677 Policy $prefix eq $siteprefix and $prefix eq $vendorprefix, all of
1678 them will now be changed to the new prefix, /foo/bar. (Previously
1679 only $prefix changed.) If you do not like this new behaviour,
1680 specify prefix, siteprefix, and vendorprefix explicitly.
1684 A new optional location for Perl libraries, otherlibdirs, is available.
1685 It can be used for example for vendor add-ons without disturbing Perl's
1686 own library directories.
1690 In many platforms, the vendor-supplied 'cc' is too stripped-down to
1691 build Perl (basically, 'cc' doesn't do ANSI C). If this seems
1692 to be the case and 'cc' does not seem to be the GNU C compiler
1693 'gcc', an automatic attempt is made to find and use 'gcc' instead.
1697 gcc needs to closely track the operating system release to avoid
1698 build problems. If Configure finds that gcc was built for a different
1699 operating system release than is running, it now gives a clearly visible
1700 warning that there may be trouble ahead.
1704 Since Perl 5.8 is not binary-compatible with previous releases
1705 of Perl, Configure no longer suggests including the 5.005
1710 Configure C<-S> can now run non-interactively. [561]
1714 Configure support for pdp11-style memory models has been removed due
1715 to obsolescence. [561]
1719 configure.gnu now works with options with whitespace in them.
1723 installperl now outputs everything to STDERR.
1727 Because PerlIO is now the default on most platforms, "-perlio" doesn't
1728 get appended to the $Config{archname} (also known as $^O) anymore.
1729 Instead, if you explicitly choose not to use perlio (Configure command
1730 line option -Uuseperlio), you will get "-stdio" appended.
1734 Another change related to the architecture name is that "-64all"
1735 (-Duse64bitall, or "maximally 64-bit") is appended only if your
1736 pointers are 64 bits wide. (To be exact, the use64bitall is ignored.)
1740 In AFS installations, one can configure the root of the AFS to be
1741 somewhere else than the default F</afs> by using the Configure
1742 parameter C<-Dafsroot=/some/where/else>.
1746 APPLLIB_EXP, a lesser-known configuration-time definition, has been
1747 documented. It can be used to prepend site-specific directories
1748 to Perl's default search path (@INC); see INSTALL for information.
1752 The version of Berkeley DB used when the Perl (and, presumably, the
1753 DB_File extension) was built is now available as
1754 C<@Config{qw(db_version_major db_version_minor db_version_patch)}>
1755 from Perl and as C<DB_VERSION_MAJOR_CFG DB_VERSION_MINOR_CFG
1756 DB_VERSION_PATCH_CFG> from C.
1760 Building Berkeley DB3 for compatibility modes for DB, NDBM, and ODBM
1761 has been documented in INSTALL.
1765 If you have CPAN access (either network or a local copy such as a
1766 CD-ROM) you can during specify extra modules to Configure to build and
1767 install with Perl using the -Dextras=... option. See INSTALL for
1772 In addition to config.over, a new override file, config.arch, is
1773 available. This file is supposed to be used by hints file writers
1774 for architecture-wide changes (as opposed to config.over which is
1775 for site-wide changes).
1779 If your file system supports symbolic links, you can build Perl outside
1780 of the source directory by
1782 mkdir /tmp/perl/build/directory
1783 cd /tmp/perl/build/directory
1784 sh /path/to/perl/source/Configure -Dmksymlinks ...
1786 This will create in /tmp/perl/build/directory a tree of symbolic links
1787 pointing to files in /path/to/perl/source. The original files are left
1788 unaffected. After Configure has finished, you can just say
1792 and Perl will be built and tested, all in /tmp/perl/build/directory.
1797 For Perl developers, several new make targets for profiling
1798 and debugging have been added; see L<perlhack>.
1804 Use of the F<gprof> tool to profile Perl has been documented in
1805 L<perlhack>. There is a make target called "perl.gprof" for
1806 generating a gprofiled Perl executable.
1810 If you have GCC 3, there is a make target called "perl.gcov" for
1811 creating a gcoved Perl executable for coverage analysis. See
1816 If you are on IRIX or Tru64 platforms, new profiling/debugging options
1817 have been added; see L<perlhack> for more information about pixie and
1824 Guidelines of how to construct minimal Perl installations have
1825 been added to INSTALL.
1829 The Thread extension is now not built at all under ithreads
1830 (C<Configure -Duseithreads>) because it wouldn't work anyway (the
1831 Thread extension requires being Configured with C<-Duse5005threads>).
1833 But note that the Thread.pm interface is now shared by both
1838 The Gconvert macro ($Config{d_Gconvert}) used by perl for stringifying
1839 floating-point numbers is now more picky about using sprintf %.*g
1840 rules for the conversion. Some platforms that used to use gcvt may
1841 now resort to the slower sprintf.
1845 The obsolete method of making a special (e.g., debugging) flavor
1848 make LIBPERL=libperld.a
1850 has been removed. Use -DDEBUGGING instead.
1854 =head2 New Or Improved Platforms
1856 For the list of platforms known to support Perl,
1857 see L<perlport/"Supported Platforms">.
1863 AIX dynamic loading should be now better supported.
1867 AIX should now work better with gcc, threads, and 64-bitness. Also the
1868 long doubles support in AIX should be better now. See L<perlaix>.
1872 AtheOS ( http://www.atheos.cx/ ) is a new platform.
1876 BeOS has been reclaimed.
1880 The DG/UX platform now supports 5.005-style threads.
1885 The DYNIX/ptx platform (a.k.a. dynixptx) is supported at or near
1890 EBCDIC platforms (z/OS (also known as OS/390), POSIX-BC, and VM/ESA)
1891 have been regained. Many test suite tests still fail and the
1892 co-existence of Unicode and EBCDIC isn't quite settled, but the
1893 situation is much better than with Perl 5.6. See L<perlos390>,
1894 L<perlbs2000> (for POSIX-BC), and L<perlvmesa> for more information.
1898 Building perl with -Duseithreads or -Duse5005threads now works under
1899 HP-UX 10.20 (previously it only worked under 10.30 or later). You will
1900 need a thread library package installed. See README.hpux. [561]
1904 Mac OS Classic is now supported in the mainstream source package
1905 (MacPerl has of course been available since perl 5.004 but now the
1906 source code bases of standard Perl and MacPerl have been synchronised)
1911 Mac OS X (or Darwin) should now be able to build Perl even on HFS+
1912 filesystems. (The case-insensitivity used to confuse the Perl build
1917 NCR MP-RAS is now supported. [561]
1921 All the NetBSD specific patches (except for the installation
1922 specific ones) have been merged back to the main distribution.
1926 NetWare from Novell is now supported. See L<perlnetware>.
1930 NonStop-UX is now supported. [561]
1934 NEC SUPER-UX is now supported.
1938 All the OpenBSD specific patches (except for the installation
1939 specific ones) have been merged back to the main distribution.
1943 Perl has been tested with the GNU pth userlevel thread package
1944 ( http://www.gnu.org/software/pth/pth.html ). All thread tests
1945 of Perl now work, but not without adding some yield()s to the tests,
1946 so while pth (and other userlevel thread implementations) can be
1947 considered to be "working" with Perl ithreads, keep in mind the
1948 possible non-preemptability of the underlying thread implementation.
1952 Stratus VOS is now supported using Perl's native build method
1953 (Configure). This is the recommended method to build Perl on
1954 VOS. The older methods, which build miniperl, are still
1955 available. See L<perlvos>. [561+]
1959 The Amdahl UTS UNIX mainframe platform is now supported. [561]
1963 WinCE is now supported. See L<perlce>.
1967 z/OS (formerly known as OS/390, formerly known as MVS OE) now has
1968 support for dynamic loading. This is not selected by default,
1969 however, you must specify -Dusedl in the arguments of Configure. [561]
1973 =head1 Selected Bug Fixes
1975 Numerous memory leaks and uninitialized memory accesses have been
1976 hunted down. Most importantly, anonymous subs used to leak quite
1983 The autouse pragma didn't work for Multi::Part::Function::Names.
1987 caller() could cause core dumps in certain situations. Carp was
1988 sometimes affected by this problem. In particular, caller() now
1989 returns a subroutine name of C<(unknown)> for subroutines that have
1990 been removed from the symbol table.
1994 chop(@list) in list context returned the characters chopped in
1995 reverse order. This has been reversed to be in the right order. [561]
1999 Configure no longer includes the DBM libraries (dbm, gdbm, db, ndbm)
2000 when building the Perl binary. The only exception to this is SunOS 4.x,
2001 which needs them. [561]
2005 The behaviour of non-decimal but numeric string constants such as
2006 "0x23" was platform-dependent: in some platforms that was seen as 35,
2007 in some as 0, in some as a floating point number (don't ask). This
2008 was caused by Perl's using the operating system libraries in a situation
2009 where the result of the string to number conversion is undefined: now
2010 Perl consistently handles such strings as zero in numeric contexts.
2014 The order of DESTROYs has been made more predictable.
2018 Several debugger fixes: exit code now reflects the script exit code,
2019 condition C<"0"> now treated correctly, the C<d> command now checks
2020 line number, C<$.> no longer gets corrupted, and all debugger output
2021 now goes correctly to the socket if RemotePort is set. [561]
2025 Perl 5.6.0 could emit spurious warnings about redefinition of
2026 dl_error() when statically building extensions into perl.
2027 This has been corrected. [561]
2031 L<dprofpp> -R didn't work.
2035 C<*foo{FORMAT}> now works.
2039 Infinity is now recognized as a number.
2043 UNIVERSAL::isa no longer caches methods incorrectly. (This broke
2044 the Tk extension with 5.6.0.) [561]
2048 Lexicals I: lexicals outside an eval "" weren't resolved
2049 correctly inside a subroutine definition inside the eval "" if they
2050 were not already referenced in the top level of the eval""ed code.
2054 Lexicals II: lexicals leaked at file scope into subroutines that
2055 were declared before the lexicals.
2059 Lexical warnings now propagating correctly between scopes
2060 and into C<eval "...">.
2064 C<use warnings qw(FATAL all)> did not work as intended. This has been
2069 warnings::enabled() now reports the state of $^W correctly if the caller
2070 isn't using lexical warnings. [561]
2074 Line renumbering with eval and C<#line> now works. [561]
2078 Fixed numerous memory leaks, especially in eval "".
2082 Localised tied variables no longer leak memory
2085 tie my %tied_hash => 'Tie::StdHash';
2089 # Used to leak memory every time local() was called;
2090 # in a loop, this added up.
2091 local($tied_hash{Foo}) = 1;
2095 Localised hash elements (and %ENV) are correctly unlocalised to not
2096 exist, if they didn't before they were localised.
2100 tie my %tied_hash => 'Tie::StdHash';
2104 # Nothing has set the FOO element so far
2106 { local $tied_hash{FOO} = 'Bar' }
2108 # This used to print, but not now.
2109 print "exists!\n" if exists $tied_hash{FOO};
2111 As a side effect of this fix, tied hash interfaces B<must> define
2112 the EXISTS and DELETE methods.
2116 mkdir() now ignores trailing slashes in the directory name,
2117 as mandated by POSIX.
2121 Some versions of glibc have a broken modfl(). This affects builds
2122 with C<-Duselongdouble>. This version of Perl detects this brokenness
2123 and has a workaround for it. The glibc release 2.2.2 is known to have
2124 fixed the modfl() bug.
2128 Modulus of unsigned numbers now works (4063328477 % 65535 used to
2129 return 27406, instead of 27047). [561]
2133 Some "not a number" warnings introduced in 5.6.0 eliminated to be
2134 more compatible with 5.005. Infinity is now recognised as a number. [561]
2138 Numeric conversions did not recognize changes in the string value
2139 properly in certain circumstances. [561]
2143 Attributes (such as :shared) didn't work with our().
2147 our() variables will not cause bogus "Variable will not stay shared"
2152 "our" variables of the same name declared in two sibling blocks
2153 resulted in bogus warnings about "redeclaration" of the variables.
2154 The problem has been corrected. [561]
2158 pack "Z" now correctly terminates the string with "\0".
2162 Fix password routines which in some shadow password platforms
2163 (e.g. HP-UX) caused getpwent() to return every other entry.
2167 The PERL5OPT environment variable (for passing command line arguments
2168 to Perl) didn't work for more than a single group of options. [561]
2172 PERL5OPT with embedded spaces didn't work.
2176 printf() no longer resets the numeric locale to "C".
2180 C<qw(a\\b)> now parses correctly as C<'a\\b'>: that is, as three
2181 characters, not four. [561]
2185 pos() did not return the correct value within s///ge in earlier
2186 versions. This is now handled correctly. [561]
2190 Printing quads (64-bit integers) with printf/sprintf now works
2191 without the q L ll prefixes (assuming you are on a quad-capable platform).
2195 Regular expressions on references and overloaded scalars now work. [561+]
2199 Right-hand side magic (GMAGIC) could in many cases such as string
2200 concatenation be invoked too many times.
2204 scalar() now forces scalar context even when used in void context.
2208 SOCKS support is now much more robust.
2212 sort() arguments are now compiled in the right wantarray context
2213 (they were accidentally using the context of the sort() itself).
2214 The comparison block is now run in scalar context, and the arguments
2215 to be sorted are always provided list context. [561]
2219 Changed the POSIX character class C<[[:space:]]> to include the (very
2220 rarely used) vertical tab character. Added a new POSIX-ish character
2221 class C<[[:blank:]]> which stands for horizontal whitespace
2222 (currently, the space and the tab).
2226 The tainting behaviour of sprintf() has been rationalized. It does
2227 not taint the result of floating point formats anymore, making the
2228 behaviour consistent with that of string interpolation. [561]
2232 Some cases of inconsistent taint propagation (such as within hash
2233 values) have been fixed.
2237 The RE engine found in Perl 5.6.0 accidentally pessimised certain kinds
2238 of simple pattern matches. These are now handled better. [561]
2242 Regular expression debug output (whether through C<use re 'debug'>
2243 or via C<-Dr>) now looks better. [561]
2247 Multi-line matches like C<"a\nxb\n" =~ /(?!\A)x/m> were flawed. The
2248 bug has been fixed. [561]
2252 Use of $& could trigger a core dump under some situations. This
2253 is now avoided. [561]
2257 The regular expression captured submatches ($1, $2, ...) are now
2258 more consistently unset if the match fails, instead of leaving false
2259 data lying around in them. [561]
2263 readline() on files opened in "slurp" mode could return an extra
2264 "" (blank line) at the end in certain situations. This has been
2269 Autovivification of symbolic references of special variables described
2270 in L<perlvar> (as in C<${$num}>) was accidentally disabled. This works
2275 Sys::Syslog ignored the C<LOG_AUTH> constant.
2279 All but the first argument of the IO syswrite() method are now optional.
2283 $AUTOLOAD, sort(), lock(), and spawning subprocesses
2284 in multiple threads simultaneously are now thread-safe.
2288 Tie::Array's SPLICE method was broken.
2292 Allow a read-only string on the left-hand side of a non-modifying tr///.
2296 If C<STDERR> is tied, warnings caused by C<warn> and C<die> now
2297 correctly pass to it.
2301 Several Unicode fixes.
2307 BOMs (byte order marks) at the beginning of Perl files
2308 (scripts, modules) should now be transparently skipped.
2309 UTF-16 and UCS-2 encoded Perl files should now be read correctly.
2313 The character tables have been updated to Unicode 3.2.0.
2317 Comparing with utf8 data does not magically upgrade non-utf8 data
2318 into utf8. (This was a problem for example if you were mixing data
2319 from I/O and Unicode data: your output might have got magically encoded
2324 Generating illegal Unicode code points such as U+FFFE, or the UTF-16
2325 surrogates, now also generates an optional warning.
2329 C<IsAlnum>, C<IsAlpha>, and C<IsWord> now match titlecase.
2333 Concatenation with the C<.> operator or via variable interpolation,
2334 C<eq>, C<substr>, C<reverse>, C<quotemeta>, the C<x> operator,
2335 substitution with C<s///>, single-quoted UTF8, should now work.
2339 The C<tr///> operator now works. Note that the C<tr///CU>
2340 functionality has been removed (but see pack('U0', ...)).
2344 C<eval "v200"> now works.
2348 Perl 5.6.0 parsed m/\x{ab}/ incorrectly, leading to spurious warnings.
2349 This has been corrected. [561]
2353 Zero entries were missing from the Unicode classes such as C<IsDigit>.
2359 Large unsigned numbers (those above 2**31) could sometimes lose their
2360 unsignedness, causing bogus results in arithmetic operations. [561]
2364 The Perl parser has been stress tested using both random input and
2365 Markov chain input and the few found crashes and lockups have been
2370 =head2 Platform Specific Changes and Fixes
2378 Perl now works on post-4.0 BSD/OSes.
2384 Setting C<$0> now works (as much as possible; see L<perlvar> for details).
2390 Numerous updates; currently synchronised with Cygwin 1.3.10.
2394 Previously DYNIX/ptx had problems in its Configure probe for non-blocking I/O.
2400 EPOC now better supported. See README.epoc. [561]
2406 Perl now works on post-3.0 FreeBSDs.
2412 README.hpux updated; C<Configure -Duse64bitall> now works;
2413 now uses HP-UX malloc instead of Perl malloc.
2419 Numerous compilation flag and hint enhancements; accidental mixing
2420 of 32-bit and 64-bit libraries (a doomed attempt) made much harder.
2430 Long doubles should now work (see INSTALL). [561]
2434 Linux previously had problems related to sockaddrlen when using
2435 accept(), recvfrom() (in Perl: recv()), getpeername(), and
2444 Compilation of the standard Perl distribution in Mac OS Classic should
2445 now work if you have the Metrowerks development environment and the
2446 missing Mac-specific toolkit bits. Contact the macperl mailing list
2453 MPE/iX update after Perl 5.6.0. See README.mpeix. [561]
2457 NetBSD/threads: try installing the GNU pth (should be in the
2458 packages collection, or http://www.gnu.org/software/pth/),
2459 and Configure with -Duseithreads.
2465 Perl now works on NetBSD/sparc.
2471 Now works with usethreads (see INSTALL). [561]
2477 64-bitness using the Sun Workshop compiler now works.
2483 The native build method requires at least VOS Release 14.5.0
2484 and GNU C++/GNU Tools 2.0.1 or later. The Perl pack function
2485 now maps overflowed values to +infinity and underflowed values
2490 Tru64 (aka Digital UNIX, aka DEC OSF/1)
2492 The operating system version letter now recorded in $Config{osvers}.
2493 Allow compiling with gcc (previously explicitly forbidden). Compiling
2494 with gcc still not recommended because buggy code results, even with
2501 Fixed various alignment problems that lead into core dumps either
2502 during build or later; no longer dies on math errors at runtime;
2503 now using full quad integers (64 bits), previously was using
2504 only 46 bit integers for speed.
2510 See L</"Socket Extension Dynamic in VMS"> and L</"IEEE-format Floating Point
2511 Default on OpenVMS Alpha"> for important changes not otherwise listed here.
2513 chdir() now works better despite a CRT bug; now works with MULTIPLICITY
2514 (see INSTALL); now works with Perl's malloc.
2516 The tainting of C<%ENV> elements via C<keys> or C<values> was previously
2517 unimplemented. It now works as documented.
2519 The C<waitpid> emulation has been improved. The worst bug (now fixed)
2520 was that a pid of -1 would cause a wildcard search of all processes on
2523 POSIX-style signals are now emulated much better on VMS versions prior
2526 The C<system> function and backticks operator have improved
2527 functionality and better error handling. [561]
2529 File access tests now use current process privileges rather than the
2530 user's default privileges, which could sometimes result in a mismatch
2531 between reported access and actual access.
2533 There is a new C<kill> implementation based on C<sys$sigprc> that allows
2534 older VMS systems (pre-7.0) to use C<kill> to send signals rather than
2535 simply force exit. This implementation also allows later systems to
2536 call C<kill> from within a signal handler.
2538 Iterative logical name translations are now limited to 10 iterations in
2539 imitation of SHOW LOGICAL and other OpenVMS facilities.
2549 accept() no longer leaks memory. [561]
2553 Borland C++ v5.5 is now a supported compiler that can build Perl.
2554 However, the generated binaries continue to be incompatible with those
2555 generated by the other supported compilers (GCC and Visual C++). [561]
2559 Better chdir() return value for a non-existent directory.
2563 Duping socket handles with open(F, ">&MYSOCK") now works under Windows
2568 New %ENV entries now propagate to subprocesses. [561]
2572 Current directory entries in %ENV are now correctly propagated to child
2577 $ENV{LIB} now used to search for libs under Visual C.
2581 fork() emulation has been improved in various ways, but still continues
2582 to be experimental. See L<perlfork> for known bugs and caveats. [561+]
2586 A failed (pseudo)fork now returns undef and sets errno to EAGAIN.
2590 Win32::GetCwd() correctly returns C:\ instead of C: when at the drive root.
2591 Other bugs in chdir() and Cwd::cwd() have also been fixed. [561]
2595 HTML files will be installed in c:\perl\html instead of c:\perl\lib\pod\html
2599 The makefiles now provide a single switch to bulk-enable all the
2600 features enabled in ActiveState ActivePerl (a popular Win32 binary
2601 distribution). [561]
2605 Allow REG_EXPAND_SZ keys in the registry.
2609 Can now send() from all threads, not just the first one. [561]
2613 ExtUtils::MakeMaker now uses $ENV{LIB} to search for libraries.
2617 Fake signal handling reenabled, bugs and all.
2621 %SIG has been enabled under USE_ITHREADS, but its use is completely
2622 unsupported under all configurations. [561]
2626 Less stack reserved per thread so that more threads can run
2627 concurrently. (Still 16M per thread.)
2631 C<< File::Spec->tmpdir() >> now prefers C:/temp over /tmp
2632 (works better when perl is running as service).
2636 Better UNC path handling under ithreads. [561]
2640 wait(), waitpid(), and backticks now return the correct exit status
2641 under Windows 9x. [561]
2645 Non-blocking waits for child processes (or pseudo-processes) are
2646 supported via C<waitpid($pid, &POSIX::WNOHANG)>.
2650 Win64 compilation is now supported.
2654 winsock handle leak fixed. [561]
2660 =head1 New or Changed Diagnostics
2666 The lexical warnings category "deprecated" is no longer a sub-category
2667 of the "syntax" category. It is now a top-level category in its own
2672 All regular expression compilation error messages are now hopefully
2673 easier to understand both because the error message now comes before
2674 the failed regex and because the point of failure is now clearly
2675 marked by a C<E<lt>-- HERE> marker.
2679 The various "opened only for", "on closed", "never opened" warnings
2680 drop the C<main::> prefix for filehandles in the C<main> package,
2681 for example C<STDIN> instead of C<main::STDIN>.
2685 The "Unrecognized escape" warning has been extended to include C<\8>,
2686 C<\9>, and C<\_>. There is no need to escape any of the C<\w> characters.
2690 Two new debugging options have been added: if you have compiled your
2691 Perl with debugging, you can use the -DT [561] and -DR options to trace
2692 tokenising and to add reference counts to displaying variables,
2697 The debugger (perl5db.pl) has been modified to present a more
2698 consistent commands interface, via (CommandSet=580). perl5db.t was
2699 also added to test the changes, and as a placeholder for further tests.
2705 The debugger has a new C<dumpDepth> option to control the maximum
2706 depth to which nested structures are dumped. The C<x> command has
2707 been extended so that C<x N EXPR> dumps out the value of I<EXPR> to a
2708 depth of at most I<N> levels.
2712 The debugger can now show lexical variables if you have the CPAN
2713 module PadWalker installed.
2717 If an attempt to use a (non-blessed) reference as an array index
2718 is made, a warning is given.
2722 C<push @a;> and C<unshift @a;> (with no values to push or unshift)
2723 now give a warning. This may be a problem for generated and evaled
2728 If you try to L<perlfunc/pack> a number less than 0 or larger than 255
2729 using the C<"C"> format you will get an optional warning. Similarly
2730 for the C<"c"> format and a number less than -128 or more than 127.
2734 Certain regex modifiers such as C<(?o)> make sense only if applied to
2735 the entire regex. You will get an optional warning if you try to do
2740 Using arrays or hashes as references (e.g. C<< %foo->{bar} >>
2741 has been deprecated for a while. Now you will get an optional warning.
2745 Using C<sort> in scalar context now issues an optional warning.
2746 This didn't do anything useful, as the sort was not performed.
2750 =head1 Changed Internals
2756 perlapi.pod (a companion to perlguts) now attempts to document the
2761 You can now build a really minimal perl called microperl.
2762 Building microperl does not require even running Configure;
2763 C<make -f Makefile.micro> should be enough. Beware: microperl makes
2764 many assumptions, some of which may be too bold; the resulting
2765 executable may crash or otherwise misbehave in wondrous ways.
2766 For careful hackers only.
2770 Added rsignal(), whichsig(), do_join(), op_clear, op_null,
2771 ptr_table_clear(), ptr_table_free(), sv_setref_uv(), and several UTF-8
2772 interfaces to the publicised API. For the full list of the available
2773 APIs see L<perlapi>.
2777 Made possible to propagate customised exceptions via croak()ing.
2781 Now xsubs can have attributes just like subs. (Well, at least the
2782 built-in attributes.)
2786 dTHR and djSP have been obsoleted; the former removed (because it's
2787 a no-op) and the latter replaced with dSP.
2791 PERL_OBJECT has been completely removed.
2795 The MAGIC constants (e.g. C<'P'>) have been macrofied
2796 (e.g. C<PERL_MAGIC_TIED>) for better source code readability
2797 and maintainability.
2801 The regex compiler now maintains a structure that identifies nodes in
2802 the compiled bytecode with the corresponding syntactic features of the
2803 original regex expression. The information is attached to the new
2804 C<offsets> member of the C<struct regexp>. See L<perldebguts> for more
2805 complete information.
2809 The C code has been made much more C<gcc -Wall> clean. Some warning
2810 messages still remain in some platforms, so if you are compiling with
2811 gcc you may see some warnings about dubious practices. The warnings
2812 are being worked on.
2816 F<perly.c>, F<sv.c>, and F<sv.h> have now been extensively commented.
2820 Documentation on how to use the Perl source repository has been added
2821 to F<Porting/repository.pod>.
2825 There are now several profiling make targets.
2829 =head1 Security Vulnerability Closed [561]
2831 (This change was already made in 5.7.0 but bears repeating here.)
2833 A potential security vulnerability in the optional suidperl component
2834 of Perl was identified in August 2000. suidperl is neither built nor
2835 installed by default. As of November 2001 the only known vulnerable
2836 platform is Linux, most likely all Linux distributions. CERT and
2837 various vendors and distributors have been alerted about the vulnerability.
2838 See http://www.cpan.org/src/5.0/sperl-2000-08-05/sperl-2000-08-05.txt
2839 for more information.
2841 The problem was caused by Perl trying to report a suspected security
2842 exploit attempt using an external program, /bin/mail. On Linux
2843 platforms the /bin/mail program had an undocumented feature which
2844 when combined with suidperl gave access to a root shell, resulting in
2845 a serious compromise instead of reporting the exploit attempt. If you
2846 don't have /bin/mail, or if you have 'safe setuid scripts', or if
2847 suidperl is not installed, you are safe.
2849 The exploit attempt reporting feature has been completely removed from
2850 Perl 5.8.0 (and the maintenance release 5.6.1, and it was removed also
2851 from all the Perl 5.7 releases), so that particular vulnerability
2852 isn't there anymore. However, further security vulnerabilities are,
2853 unfortunately, always possible. The suidperl functionality is most
2854 probably going to be removed in Perl 5.10. In any case, suidperl
2855 should only be used by security experts who know exactly what they are
2856 doing and why they are using suidperl instead of some other solution
2857 such as sudo ( see http://www.courtesan.com/sudo/ ).
2861 Several new tests have been added, especially for the F<lib> and F<ext>
2862 subsections. There are now about 65 000 individual tests (spread over
2863 about 700 test scripts), in the regression suite (5.6.1 has about
2864 11700 tests, in 258 test scripts) Many of the new tests are of course
2865 introduced by the new modules, but still in general Perl is now more
2868 Because of the large number of tests, running the regression suite
2869 will take considerably longer time than it used to: expect the suite
2870 to take up to 4-5 times longer to run than in perl 5.6. On a really
2871 fast machine you can hope to finish the suite in about 6-8 minutes
2874 The tests are now reported in a different order than in earlier Perls.
2875 (This happens because the test scripts from under t/lib have been moved
2876 to be closer to the library/extension they are testing.)
2878 =head1 Known Problems
2886 If using the AIX native make command, instead of just "make" issue
2887 "make all". In some setups the former has been known to spuriously
2888 also try to run "make install". Alternatively, you may want to use
2893 In AIX 4.2, Perl extensions that use C++ functions that use statics
2894 may have problems in that the statics are not getting initialized.
2895 In newer AIX releases, this has been solved by linking Perl with
2896 the libC_r library, but unfortunately in AIX 4.2 the said library
2897 has an obscure bug where the various functions related to time
2898 (such as time() and gettimeofday()) return broken values, and
2899 therefore in AIX 4.2 Perl is not linked against libC_r.
2903 vac 5.0.0.0 May Produce Buggy Code For Perl
2905 The AIX C compiler vac version 5.0.0.0 may produce buggy code,
2906 resulting in a few random tests failing when run as part of "make
2907 test", but when the failing tests are run by hand, they succeed.
2908 We suggest upgrading to at least vac version 5.0.1.0, that has been
2909 known to compile Perl correctly. "lslpp -L|grep vac.C" will tell
2910 you the vac version. See README.aix.
2914 If building threaded Perl, you may get compilation warning from pp_sys.c:
2916 "pp_sys.c", line 4651.39: 1506-280 (W) Function argument assignment between types "unsigned char*" and "const void*" is not allowed.
2918 This is harmless; it is caused by the getnetbyaddr() and getnetbyaddr_r()
2919 having slightly different types for their first argument.
2923 =head2 Alpha systems with old gccs fail several tests
2925 If you see op/pack, op/pat, op/regexp, or ext/Storable tests failing
2926 in a Linux/alpha or *BSD/Alpha, it's probably time to upgrade your gcc.
2927 gccs prior to 2.95.3 are definitely not good enough, and gcc 3.1 may
2928 be even better. (RedHat Linux/alpha with gcc 3.1 reported no problems,
2929 as did Linux 2.4.18 with gcc 2.95.4.) (In Tru64, it is preferable to
2930 use the bundled C compiler.)
2934 Perl 5.8.0 doesn't build in AmigaOS. It broke at some point during
2935 the ithreads work and we could not find Amiga experts to unbreak the
2936 problems. Perl 5.6.1 still works for AmigaOS (as does the the 5.7.2
2937 development release).
2941 The following tests fail on 5.8.0 Perl in BeOS Personal 5.03:
2943 t/op/lfs............................FAILED at test 17
2944 t/op/magic..........................FAILED at test 24
2945 ext/POSIX/t/sigaction...............FAILED at test 13
2946 ext/POSIX/t/waitpid.................FAILED at test 1
2948 See L<perlbeos> (README.beos) for more details.
2950 =head2 Cygwin "unable to remap"
2952 For example when building the Tk extension for Cygwin,
2953 you may get an error message saying "unable to remap".
2954 This is known problem with Cygwin, and a workaround is
2955 detailed in here: http://sources.redhat.com/ml/cygwin/2001-12/msg00894.html
2957 =head2 ext/threads/t/libc
2959 If this test fails, it indicates that your libc (C library) is not
2960 threadsafe. This particular test stress tests the localtime() call to
2961 find out whether it is threadsafe. See L<perlthrtut> for more information.
2963 =head2 FreeBSD Failing locale Test 117 For ISO8859-15 Locales
2965 The ISO8859-15 locales may fail the locale test 117 in FreeBSD.
2966 This is caused by the characters \xFF (y with diaeresis) and \xBE
2967 (Y with diaeresis) not behaving correctly when being matched
2970 =head2 IRIX fails ext/List/Util/t/shuffle.t
2972 IRIX with MIPSpro 7.3.1.3m compiler may fail the said List::Util test
2973 by dumping core. This seems to be a compiler error since if compiled
2974 with gcc no core dump ensues, and no failures on the said test on any
2977 =head2 Modifying $_ Inside for(..)
2981 works without complaint. It shouldn't. (You should be able to
2982 modify only lvalue elements inside the loops.) You can see the
2983 correct behaviour by replacing the 1..5 with 1, 2, 3, 4, 5.
2985 =head2 mod_perl 1.26 Doesn't Build With Threaded Perl
2987 Use mod_perl 1.27 or higher.
2989 =head2 lib/ftmp-security tests warn 'system possibly insecure'
2991 Don't panic. Read the 'make test' section of INSTALL instead.
2993 =head2 HP-UX lib/posix Subtest 9 Fails When LP64-Configured
2995 If perl is configured with -Duse64bitall, the successful result of the
2996 subtest 10 of lib/posix may arrive before the successful result of the
2997 subtest 9, which confuses the test harness so much that it thinks the
3000 =head2 Linux with glibc 2.2.5 fails t/op/int subtest #6 with -Duse64bitint
3002 This is a known bug in the glibc 2.2.5 with long long integers.
3003 ( http://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=65612 )
3005 =head2 Linux With Sfio Fails op/misc Test 48
3009 =head2 libwww-perl (LWP) fails base/date #51
3011 Use libwww-perl 5.65 or later.
3015 Please remember to set your environment variable LC_ALL to "C"
3016 (setenv LC_ALL C) before running "make test" to avoid a lot of
3017 warnings about the broken locales of Mac OS X.
3019 The following tests are known to fail in Mac OS X 10.1.5 because of
3020 buggy (old) implementations of Berkeley DB included in Mac OS X:
3022 Failed Test Stat Wstat Total Fail Failed List of Failed
3023 -------------------------------------------------------------------------
3024 ../ext/DB_File/t/db-btree.t 0 11 ?? ?? % ??
3025 ../ext/DB_File/t/db-recno.t 149 3 2.01% 61 63 65
3027 If you are building on a UFS partition, you will also probably see
3028 t/op/stat.t subtest #9 fail. This is caused by Darwin's UFS not
3029 supporting inode change time.
3031 Also the ext/POSIX/t/posix.t subtest #10 fails but it is skipped for
3032 now because the failure is Apple's fault, not Perl's (blocked signals
3035 If you Configure with ithreads, ext/threads/t/libc.t will fail. Again,
3036 this is not Perl's fault-- the libc of Mac OS X is not threadsafe
3037 (in this particular test, the localtime() call is found to be
3040 =head2 op/sprintf tests 91, 129, and 130
3042 The op/sprintf tests 91, 129, and 130 are known to fail on some platforms.
3043 Examples include any platform using sfio, and Compaq/Tandem's NonStop-UX.
3045 Test 91 is known to fail on QNX6 (nto), because C<sprintf '%e',0>
3046 incorrectly produces C<0.000000e+0> instead of C<0.000000e+00>.
3048 For tests 129 and 130, the failing platforms do not comply with
3049 the ANSI C Standard: lines 19ff on page 134 of ANSI X3.159 1989, to
3050 be exact. (They produce something other than "1" and "-1" when
3051 formatting 0.6 and -0.6 using the printf format "%.0f"; most often,
3052 they produce "0" and "-0".)
3056 In case you are still using Solaris 2.5 (aka SunOS 5.5), you may
3057 experience failures (the test core dumping) in lib/locale.t.
3058 The suggested cure is to upgrade your Solaris.
3060 =head2 SUPER-UX (NEC SX)
3062 The following tests are known to fail on SUPER-UX:
3064 op/64bitint...........................FAILED tests 29-30, 32-33, 35-36
3065 op/arith..............................FAILED tests 128-130
3066 op/pack...............................FAILED tests 25-5625
3067 op/pow................................
3068 op/taint..............................# msgsnd failed
3069 ../ext/IO/lib/IO/t/io_poll............FAILED tests 3-4
3070 ../ext/IPC/SysV/ipcsysv...............FAILED tests 2, 5-6
3071 ../ext/IPC/SysV/t/msg.................FAILED tests 2, 4-6
3072 ../ext/Socket/socketpair..............FAILED tests 12
3073 ../lib/IPC/SysV.......................FAILED tests 2, 5-6
3074 ../lib/warnings.......................FAILED tests 115-116, 118-119
3076 The op/pack failure ("Cannot compress negative numbers at op/pack.t line 126")
3077 is serious but as of yet unsolved. It points at some problems with the
3078 signedness handling of the C compiler, as do the 64bitint, arith, and pow
3079 failures. Most of the rest point at problems with SysV IPC.
3081 =head2 Term::ReadKey not working on Win32
3083 Use Term::ReadKey 2.20 or later.
3085 =head2 Failure of Thread (5.005-style) tests
3087 B<Note that support for 5.005-style threading is deprecated,
3088 experimental and practically unsupported. In 5.10, it is expected
3091 The following tests are known to fail due to fundamental problems in
3092 the 5.005 threading implementation. These are not new failures--Perl
3093 5.005_0x has the same bugs, but didn't have these tests.
3095 ../ext/B/t/xref.t 255 65280 14 12 85.71% 3-14
3096 ../ext/List/Util/t/first.t 255 65280 7 4 57.14% 2 5-7
3097 ../lib/English.t 2 512 54 2 3.70% 2-3
3098 ../lib/ExtUtils/t/basic.t 1 256 17 1 5.88% 14
3099 ../lib/FileCache.t 5 1 20.00% 5
3100 ../lib/Filter/Simple/t/data.t 6 3 50.00% 1-3
3101 ../lib/Filter/Simple/t/filter_onl 9 3 33.33% 1-2 5
3102 ../lib/Tie/File/t/31_autodefer.t 255 65280 65 32 49.23% 34-65
3103 ../lib/autouse.t 10 1 10.00% 4
3104 op/flip.t 15 1 6.67% 15
3106 These failures are unlikely to get fixed as 5.005-style threads
3107 are considered fundamentally broken. (Basically what happens is that
3108 competing threads can corrupt shared global state.)
3110 =head2 Timing problems
3112 The following tests may fail intermittently because of timing
3113 problems, for example if the system is heavily loaded.
3116 ext/Time/HiRes/HiRes.t
3118 lib/Memoize/t/expmod_t.t
3119 lib/Memoize/t/speed.t
3121 In case of failure please try running them manually, for example
3123 ./perl -Ilib ext/Time/HiRes/HiRes.t
3127 ../lib/Math/Trig.t 26 1 3.85% 25
3128 ../lib/warnings.t 470 1 0.21% 429
3130 The Trig.t failure is caused by the slighly differing (from IEEE)
3131 floating point implementation of UNICOS. The warnings.t failure is
3132 also related: the test assumes a certain floating point output format;
3133 this assumption fails in UNICOS.
3141 During Configure, the test
3143 Guessing which symbols your C compiler and preprocessor define...
3145 will probably fail with error messages like
3147 CC-20 cc: ERROR File = try.c, Line = 3
3148 The identifier "bad" is undefined.
3150 bad switch yylook 79bad switch yylook 79bad switch yylook 79bad switch yylook 79#ifdef A29K
3153 CC-65 cc: ERROR File = try.c, Line = 3
3154 A semicolon is expected at this point.
3156 This is caused by a bug in the awk utility of UNICOS/mk. You can ignore
3157 the error, but it does cause a slight problem: you cannot fully
3158 benefit from the h2ph utility (see L<h2ph>) that can be used to
3159 convert C headers to Perl libraries, mainly used to be able to access
3160 from Perl the constants defined using C preprocessor, cpp. Because of
3161 the above error, parts of the converted headers will be invisible.
3162 Luckily, these days the need for h2ph is rare.
3166 If building Perl with interpreter threads (ithreads), the
3167 getgrent(), getgrnam(), and getgrgid() functions cannot return the
3168 list of the group members due to a bug in the multithreaded support of
3169 UNICOS/mk. What this means is that in list context the functions will
3170 return only three values, not four.
3176 There are a few known test failures, see L<perluts> (README.uts).
3178 =head2 VOS (Stratus)
3180 When Perl is built using the native build process on VOS Release
3181 14.5.0 and GNU C++/GNU Tools 2.0.1, all attempted tests either
3182 pass or result in TODO (ignored) failures.
3186 There should be no reported test failures with a default configuration,
3187 though there are a number of tests marked TODO that point to areas
3188 needing further debugging and/or porting work.
3192 In multi-CPU boxes, there are some problems with the I/O buffering:
3193 some output may appear twice.
3195 =head2 XML::Parser not working
3197 Use XML::Parser 2.31 or later.
3199 =head2 z/OS (OS/390)
3201 z/OS has rather many test failures but the situation is actually
3202 better than it was in 5.6.0; it's just that so many new modules and
3203 tests have been added.
3205 Failed Test Stat Wstat Total Fail Failed List of Failed
3206 ---------------------------------------------------------------------------
3207 ../ext/Data/Dumper/t/dumper.t 357 8 2.24% 311 314 325 327
3209 ../ext/IO/lib/IO/t/io_unix.t 5 4 80.00% 2-5
3210 ../ext/Storable/t/downgrade.t 12 3072 169 12 7.10% 14-15 46-47 78-79
3212 ../lib/ExtUtils/t/Constant.t 121 30976 48 48 100.00% 1-48
3213 ../lib/ExtUtils/t/Embed.t 9 9 100.00% 1-9
3214 op/pat.t 910 7 0.77% 665 776 785 832-
3216 op/sprintf.t 224 3 1.34% 98 100 136
3217 op/tr.t 97 5 5.15% 63 71-74
3218 uni/fold.t 780 6 0.77% 61 169 196 661
3221 The failures in dumper.t and downgrade.t are problems in the tests,
3222 those in io_unix and sprintf are problems in the USS (UDP sockets
3223 and printf formats). The pat, tr, and fold failures are genuine Perl
3224 problems caused by EBCDIC (and in the pat and fold cases, combining
3225 that with Unicode). The Constant and Embed are probably problems
3226 in the tests (since they test Perl's ability to build extensions,
3227 and that seems to be working reasonably well.)
3229 =head2 Localising Tied Arrays and Hashes Is Broken
3233 doesn't work as one would expect: the old value is restored
3234 incorrectly. This will be changed in a future release, but we don't
3235 know yet what the new semantics will exactly be. In any case, the
3236 change will break existing code that relies on the current
3237 (ill-defined) semantics, so just avoid doing this in general.
3239 =head2 Self-tying Problems
3241 Self-tying of arrays and hashes is broken in rather deep and
3242 hard-to-fix ways. As a stop-gap measure to avoid people from getting
3243 frustrated at the mysterious results (core dumps, most often), it is
3244 forbidden for now (you will get a fatal error even from an attempt).
3246 A change to self-tying of globs has caused them to be recursively
3247 referenced (see: L<perlobj/"Two-Phased Garbage Collection">). You
3248 will now need an explicit untie to destroy a self-tied glob. This
3249 behaviour may be fixed at a later date.
3251 Self-tying of scalars and IO thingies works.
3253 =head2 Building Extensions Can Fail Because Of Largefiles
3255 Some extensions like mod_perl are known to have issues with
3256 `largefiles', a change brought by Perl 5.6.0 in which file offsets
3257 default to 64 bits wide, where supported. Modules may fail to compile
3258 at all, or they may compile and work incorrectly. Currently, there
3259 is no good solution for the problem, but Configure now provides
3260 appropriate non-largefile ccflags, ldflags, libswanted, and libs
3261 in the %Config hash (e.g., $Config{ccflags_nolargefiles}) so the
3262 extensions that are having problems can try configuring themselves
3263 without the largefileness. This is admittedly not a clean solution,
3264 and the solution may not even work at all. One potential failure is
3265 whether one can (or, if one can, whether it's a good idea to) link
3266 together at all binaries with different ideas about file offsets;
3267 all this is platform-dependent.
3269 =head2 Unicode Support on EBCDIC Still Spotty
3271 Though mostly working, Unicode support still has problem spots on
3272 EBCDIC platforms. One such known spot are the C<\p{}> and C<\P{}>
3273 regular expression constructs for code points less than 256: the
3274 C<pP> are testing for Unicode code points, not knowing about EBCDIC.
3276 =head2 The Compiler Suite Is Still Very Experimental
3278 The compiler suite is slowly getting better but it continues to be
3279 highly experimental. Use in production environments is discouraged.
3281 =head2 The Long Double Support Is Still Experimental
3283 The ability to configure Perl's numbers to use "long doubles",
3284 floating point numbers of hopefully better accuracy, is still
3285 experimental. The implementations of long doubles are not yet
3286 widespread and the existing implementations are not quite mature
3287 or standardised, therefore trying to support them is a rare
3288 and moving target. The gain of more precision may also be offset
3289 by slowdown in computations (more bits to move around, and the
3290 operations are more likely to be executed by less optimised
3293 =head2 Seen In Perl 5.7 But Gone Now
3295 C<Time::Piece> (previously known as C<Time::Object>) was removed
3296 because it was felt that it didn't have enough value in it to be a
3297 core module. It is still a useful module, though, and is available
3300 Perl 5.8 unfortunately does not build anymore on AmigaOS; this broke
3301 accidentally at some point. Since there are not that many Amiga
3302 developers available, we could not get this fixed and tested in time
3303 for 5.8.0. Perl 5.6.1 still works for AmigaOS (as does the the 5.7.2
3304 development release).
3306 =head1 Reporting Bugs
3308 If you find what you think is a bug, you might check the articles
3309 recently posted to the comp.lang.perl.misc newsgroup and the perl
3310 bug database at http://bugs.perl.org/ . There may also be
3311 information at http://www.perl.com/ , the Perl Home Page.
3313 If you believe you have an unreported bug, please run the B<perlbug>
3314 program included with your release. Be sure to trim your bug down
3315 to a tiny but sufficient test case. Your bug report, along with the
3316 output of C<perl -V>, will be sent off to perlbug@perl.org to be
3317 analysed by the Perl porting team.
3321 The F<Changes> file for exhaustive details on what changed.
3323 The F<INSTALL> file for how to build Perl.
3325 The F<README> file for general stuff.
3327 The F<Artistic> and F<Copying> files for copyright information.
3331 Written by Jarkko Hietaniemi <F<jhi@iki.fi>>.