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 In future releases, non-PerlIO aware XS modules may become completely
231 unsupported. Since PerlIO is a drop-in replacement for stdio, this
232 shouldn't be that drastic a change.
236 The current user-visible implementation of pseudo-hashes (the weird
237 use of the first array element) is deprecated starting from Perl 5.8.0
238 and will be removed in Perl 5.10.0, and the feature will be
239 implemented differently. Not only is the current interface rather
240 ugly, but the current implementation slows down normal array and hash
241 use quite noticeably. The C<fields> pragma interface will remain
242 available. The I<restricted hashes> interface is expected to
243 be the replacement interface (see L<Hash::Util>).
247 The syntaxes C<< @a->[...] >> and C<< %h->{...} >> have now been deprecated.
251 After years of trying, suidperl is considered to be too complex to
252 ever be considered truly secure. The suidperl functionality is likely
253 to be removed in a future release.
257 The 5.005 threads model (module C<Thread>) is deprecated and expected
258 to be removed in Perl 5.10. Multithreaded code should be migrated to
259 the new ithreads model (see L<threads>, L<threads::shared> and
264 The long deprecated uppercase aliases for the string comparison
265 operators (EQ, NE, LT, LE, GE, GT) have now been removed.
269 The tr///C and tr///U features have been removed and will not return;
270 the interface was a mistake. Sorry about that. For similar
271 functionality, see pack('U0', ...) and pack('C0', ...). [561]
275 Earlier Perls treated "sub foo (@bar)" as equivalent to "sub foo (@)".
276 The prototypes are now checked better at compile-time for invalid
277 syntax. An optional warning is generated ("Illegal character in
278 prototype...") but this may be upgraded to a fatal error in a future
283 The C<exec LIST> and C<system LIST> will become fatal errors
284 under tainting in some future release.
288 The existing behaviour when localising tied arrays and hashes is wrong,
289 and will be changed in a future release, so do not rely on the existing
290 behaviour. See L<"Localising Tied Arrays and Hashes Is Broken">.
294 =head1 Core Enhancements
296 =head2 PerlIO is Now The Default
302 IO is now by default done via PerlIO rather than system's "stdio".
303 PerlIO allows "layers" to be "pushed" onto a file handle to alter the
304 handle's behaviour. Layers can be specified at open time via 3-arg
307 open($fh,'>:crlf :utf8', $path) || ...
309 or on already opened handles via extended C<binmode>:
311 binmode($fh,':encoding(iso-8859-7)');
313 The built-in layers are: unix (low level read/write), stdio (as in
314 previous Perls), perlio (re-implementation of stdio buffering in a
315 portable manner), crlf (does CRLF <=> "\n" translation as on Win32,
316 but available on any platform). A mmap layer may be available if
317 platform supports it (mostly UNIXes).
319 Layers to be applied by default may be specified via the 'open' pragma.
321 See L</"Installation and Configuration Improvements"> for the effects
322 of PerlIO on your architecture name.
326 File handles can be marked as accepting Perl's internal encoding of Unicode
327 (UTF-8 or UTF-EBCDIC depending on platform) by a pseudo layer ":utf8" :
329 open($fh,">:utf8","Uni.txt");
331 Note for EBCDIC users: the pseudo layer ":utf8" is erroneously named
332 for you since it's not UTF-8 what you will be getting but instead
333 UTF-EBCDIC. See L<perlunicode>, L<utf8>, and
334 http://www.unicode.org/unicode/reports/tr16/ for more information.
335 In future releases this naming may change.
339 File handles can translate character encodings from/to Perl's internal
340 Unicode form on read/write via the ":encoding()" layer.
344 File handles can be opened to "in memory" files held in Perl scalars via:
346 open($fh,'>', \$variable) || ...
350 Anonymous temporary files are available without need to
351 'use FileHandle' or other module via
353 open($fh,"+>", undef) || ...
355 That is a literal undef, not an undefined value.
359 The list form of C<open> is now implemented for pipes (at least on UNIX):
361 open($fh,"-|", 'cat', '/etc/motd')
363 creates a pipe, and runs the equivalent of exec('cat', '/etc/motd') in
368 If your locale environment variables (LANGUAGE, LC_ALL, LC_CTYPE, LANG)
369 contain the strings 'UTF-8' or 'UTF8' (case-insensitive matching),
370 the default encoding of your STDIN, STDOUT, and STDERR, and of
371 B<any subsequent file open>, is UTF-8.
375 =head2 Restricted Hashes
377 A restricted hash is restricted to a certain set of keys, no keys
378 outside the set can be added. Also individual keys can be restricted
379 so that the key cannot be deleted and the value cannot be changed.
380 No new syntax is involved: the Hash::Util module is the interface.
384 Perl used to be fragile in that signals arriving at inopportune moments
385 could corrupt Perl's internal state. Now Perl postpones handling of
386 signals until it's safe (between opcodes).
388 This change may have surprising side effects because signals no longer
389 interrupt Perl instantly. Perl will now first finish whatever it was
390 doing, like finishing an internal operation (like sort()) or an
391 external operation (like an I/O operation), and only then look at any
392 arrived signals (and before starting the next operation). No more corrupt
393 internal state since the current operation is always finished first,
394 but the signal may take more time to get heard. Note that breaking
395 out from potentially blocking operations should still work, though.
397 =head2 Unicode Overhaul
399 Unicode in general should be now much more usable than in Perl 5.6.0
400 (or even in 5.6.1). Unicode can be used in hash keys, Unicode in
401 regular expressions should work now, Unicode in tr/// should work now,
402 Unicode in I/O should work now. See L<perluniintro> for introduction
403 and L<perlunicode> for details.
409 The Unicode Character Database coming with Perl has been upgraded
410 to Unicode 3.2.0. For more information, see http://www.unicode.org/ .
411 [561+] (5.6.1 has UCD 3.0.1.)
415 For developers interested in enhancing Perl's Unicode capabilities:
416 almost all the UCD files are included with the Perl distribution in
417 the F<lib/unicore> subdirectory. The most notable omission, for space
418 considerations, is the Unihan database.
422 The properties \p{Blank} and \p{SpacePerl} have been added. "Blank" is like
423 C isblank(), that is, it contains only "horizontal whitespace" (the space
424 character is, the newline isn't), and the "SpacePerl" is the Unicode
425 equivalent of C<\s> (\p{Space} isn't, since that includes the vertical
426 tabulator character, whereas C<\s> doesn't.)
428 See "New Unicode Properties" earlier in this document for additional
429 information on changes with Unicode properties.
433 =head2 Understanding of Numbers
435 In general a lot of fixing has happened in the area of Perl's
436 understanding of numbers, both integer and floating point. Since in
437 many systems the standard number parsing functions like C<strtoul()>
438 and C<atof()> seem to have bugs, Perl tries to work around their
439 deficiencies. This results hopefully in more accurate numbers.
441 Perl now tries internally to use integer values in numeric conversions
442 and basic arithmetics (+ - * /) if the arguments are integers, and
443 tries also to keep the results stored internally as integers.
444 This change leads to often slightly faster and always less lossy
445 arithmetics. (Previously Perl always preferred floating point numbers
448 =head2 Arrays now always interpolate into double-quoted strings [561]
450 In double-quoted strings, arrays now interpolate, no matter what. The
451 behavior in earlier versions of perl 5 was that arrays would interpolate
452 into strings if the array had been mentioned before the string was
453 compiled, and otherwise Perl would raise a fatal compile-time error.
454 In versions 5.000 through 5.003, the error was
456 Literal @example now requires backslash
458 In versions 5.004_01 through 5.6.0, the error was
460 In string, @example now must be written as \@example
462 The idea here was to get people into the habit of writing
463 C<"fred\@example.com"> when they wanted a literal C<@> sign, just as
464 they have always written C<"Give me back my \$5"> when they wanted a
467 Starting with 5.6.1, when Perl now sees an C<@> sign in a
468 double-quoted string, it I<always> attempts to interpolate an array,
469 regardless of whether or not the array has been used or declared
470 already. The fatal error has been downgraded to an optional warning:
472 Possible unintended interpolation of @example in string
474 This warns you that C<"fred@example.com"> is going to turn into
475 C<fred.com> if you don't backslash the C<@>.
476 See http://www.plover.com/~mjd/perl/at-error.html for more details
477 about the history here.
479 =head2 Miscellaneous Changes
485 AUTOLOAD is now lvaluable, meaning that you can add the :lvalue attribute
486 to AUTOLOAD subroutines and you can assign to the AUTOLOAD return value.
490 The $Config{byteorder} (and corresponding BYTEORDER in config.h) was
491 previously wrong in platforms if sizeof(long) was 4, but sizeof(IV)
492 was 8. The byteorder was only sizeof(long) bytes long (1234 or 4321),
493 but now it is correctly sizeof(IV) bytes long, (12345678 or 87654321).
494 (This problem didn't affect Windows platforms.)
496 Also, $Config{byteorder} is now computed dynamically--this is more
497 robust with "fat binaries" where an executable image contains binaries
498 for more than one binary platform, and when cross-compiling.
502 C<perl -d:Module=arg,arg,arg> now works (previously one couldn't pass
503 in multiple arguments.)
507 C<do> followed by a bareword now ensures that this bareword isn't
508 a keyword (to avoid a bug where C<do q(foo.pl)> tried to call
509 subroutine called C<q>). This means that for example instead of
510 C<do format()> you must write C<do &format()>.
514 The builtin dump() now gives an optional warning
515 C<dump() better written as CORE::dump()>,
516 meaning that by default C<dump(...)> is resolved as the builtin
517 dump() which dumps core and aborts, not as (possibly) user-defined
518 C<sub dump>. To call the latter, qualify the call as C<&dump(...)>.
519 (The whole dump() feature is to considered deprecated, and possibly
520 removed/changed in future releases.)
524 chomp() and chop() are now overridable. Note, however, that their
525 prototype (as given by C<prototype("CORE::chomp")> is undefined,
526 because it cannot be expressed and therefore one cannot really write
527 replacements to override these builtins.
531 END blocks are now run even if you exit/die in a BEGIN block.
532 Internally, the execution of END blocks is now controlled by
533 PL_exit_flags & PERL_EXIT_DESTRUCT_END. This enables the new
534 behaviour for Perl embedders. This will default in 5.10. See
539 Formats now support zero-padded decimal fields.
543 Lvalue subroutines can now return C<undef> in list context. However,
544 the lvalue subroutine feature still remains experimental. [561+]
548 A lost warning "Can't declare ... dereference in my" has been
549 restored (Perl had it earlier but it became lost in later releases.)
553 A new special regular expression variable has been introduced:
554 C<$^N>, which contains the most-recently closed group (submatch).
558 C<no Module;> does not produce an error even if Module does not have an
559 unimport() method. This parallels the behavior of C<use> vis-a-vis
564 The numerical comparison operators return C<undef> if either operand
565 is a NaN. Previously the behaviour was unspecified.
569 The following builtin functions are now overridable: each(), keys(),
570 pop(), push(), shift(), splice(), unshift(). [561]
574 C<pack() / unpack()> can now group template letters with C<()> and then
575 apply repetition/count modifiers on the groups.
579 C<pack() / unpack()> can now process the Perl internal numeric types:
580 IVs, UVs, NVs-- and also long doubles, if supported by the platform.
581 The template letters are C<j>, C<J>, C<F>, and C<D>.
585 C<pack('U0a*', ...)> can now be used to force a string to UTF8.
589 my __PACKAGE__ $obj now works. [561]
593 POSIX::sleep() now returns the number of I<unslept> seconds
594 (as the POSIX standard says), as opposed to CORE::sleep() which
595 returns the number of slept seconds.
599 The printf() and sprintf() now support parameter reordering using the
600 C<%\d+\$> and C<*\d+\$> syntaxes. For example
602 print "%2\$s %1\$s\n", "foo", "bar";
604 will print "bar foo\n". This feature helps in writing
605 internationalised software, and in general when the order
606 of the parameters can vary.
610 The (\&) prototype now works properly. [561]
614 prototype(\[$@%&]) is now available to implicitly create references
615 (useful for example if you want to emulate the tie() interface).
619 A new command-line option, C<-t> is available. It is the
620 little brother of C<-T>: instead of dying on taint violations,
621 lexical warnings are given. B<This is only meant as a temporary
622 debugging aid while securing the code of old legacy applications.
623 This is not a substitute for -T.>
627 In other taint news, the C<exec LIST> and C<system LIST> have now been
628 considered too risky (think C<exec @ARGV>: it can start any program
629 with any arguments), and now the said forms cause a warning.
630 You should carefully launder the arguments to guarantee their
631 validity. In future releases of Perl the forms will become fatal
632 errors so consider starting laundering now.
636 Tied hash interfaces are now required to have the EXISTS and DELETE
637 methods (either own or inherited).
641 If tr/// is just counting characters, it doesn't attempt to
646 untie() will now call an UNTIE() hook if it exists. See L<perltie>
651 L<utime> now supports C<utime undef, undef, @files> to change the
652 file timestamps to the current time.
656 The rules for allowing underscores (underbars) in numeric constants
657 have been relaxed and simplified: now you can have an underscore
658 simply B<between digits>.
662 Rather than relying on C's argv[0] (which may not contain a full pathname)
663 where possible $^X is now set by asking the operating system.
664 (eg by reading F</proc/self/exe> on Linux, F</proc/curproc/file> on FreeBSD)
668 A new variable, C<${^TAINT}>, indicates whether taint mode is enabled.
672 You can now override the readline() builtin, and this overrides also
673 the <FILEHANDLE> angle bracket operator.
677 The command-line options -s and -F are now recognized on the shebang
682 Use of the C</c> match modifier without an accompanying C</g> modifier
683 elicits a new warning: C<Use of /c modifier is meaningless without /g>.
685 Use of C</c> in substitutions, even with C</g>, elicits
686 C<Use of /c modifier is meaningless in s///>.
688 Use of C</g> with C<split> elicits C<Use of /g modifier is meaningless
693 Support for the C<CLONE> special subroutine had been added.
694 With ithreads, when a new thread is created, all Perl data is cloned,
695 however non-Perl data cannot be cloned automatically. In C<CLONE> you
696 can do whatever you need to do, like for example handle the cloning of
697 non-Perl data, if necessary. C<CLONE> will be executed once for every
698 package that has it defined or inherited. It will be called in the
699 context of the new thread, so all modifications are made in the new area.
705 =head1 Modules and Pragmata
707 =head2 New Modules and Pragmata
713 C<Attribute::Handlers> allows a class to define attribute handlers.
716 use Attribute::Handlers;
717 sub Wolf :ATTR(SCALAR) { print "howl!\n" }
719 # later, in some package using or inheriting from MyPack...
721 my MyPack $Fluffy : Wolf; # the attribute handler Wolf will be called
723 Both variables and routines can have attribute handlers. Handlers can
724 be specific to type (SCALAR, ARRAY, HASH, or CODE), or specific to the
725 exact compilation phase (BEGIN, CHECK, INIT, or END).
726 See L<Attribute::Handlers>.
730 C<B::Concise>, by Stephen McCamant, is a new compiler backend for
731 walking the Perl syntax tree, printing concise info about ops.
732 The output is highly customisable. See L<B::Concise>. [561+]
736 The new bignum, bigint, and bigrat pragmas, by Tels, implement
737 transparent bignum support (using the Math::BigInt, Math::BigFloat,
738 and Math::BigRat backends).
742 C<Class::ISA>, by Sean Burke, is a module for reporting the search
743 path for a class's ISA tree. See L<Class::ISA>.
747 C<Cwd> now has a split personality: if possible, an XS extension is
748 used, (this will hopefully be faster, more secure, and more robust)
749 but if not possible, the familiar Perl implementation is used.
753 C<Devel::PPPort>, originally by Kenneth Albanowski and now
754 maintained by Paul Marquess, has been added. It is primarily used
755 by C<h2xs> to enhance portability of XS modules between different
756 versions of Perl. See L<Devel::PPPort>.
760 C<Digest>, frontend module for calculating digests (checksums), from
761 Gisle Aas, has been added. See L<Digest>.
765 C<Digest::MD5> for calculating MD5 digests (checksums) as defined in
766 RFC 1321, from Gisle Aas, has been added. See L<Digest::MD5>.
768 use Digest::MD5 'md5_hex';
770 $digest = md5_hex("Thirsty Camel");
772 print $digest, "\n"; # 01d19d9d2045e005c3f1b80e8b164de1
774 NOTE: the C<MD5> backward compatibility module is deliberately not
775 included since its further use is discouraged.
779 C<Encode>, originally by Nick Ing-Simmons and now maintained by Dan
780 Kogai, provides a mechanism to translate between different character
781 encodings. Support for Unicode, ISO-8859-1, and ASCII are compiled in
782 to the module. Several other encodings (like the rest of the
783 ISO-8859, CP*/Win*, Mac, KOI8-R, three variants EBCDIC, Chinese,
784 Japanese, and Korean encodings) are included and can be loaded at
785 runtime. (For space considerations, the largest Chinese encodings
786 have been separated into their own CPAN module, Encode::HanExtra,
787 which Encode will use if available). See L<Encode>.
789 Any encoding supported by Encode module is also available to the
790 ":encoding()" layer if PerlIO is used.
794 C<Hash::Util> is the interface to the new I<restricted hashes>
795 feature. (Implemented by Jeffrey Friedl, Nick Ing-Simmons, and
796 Michael Schwern.) See L<Hash::Util>.
800 C<I18N::Langinfo> can be used to query locale information.
801 See L<I18N::Langinfo>.
805 C<I18N::LangTags>, by Sean Burke, has functions for dealing with
806 RFC3066-style language tags. See L<I18N::LangTags>.
810 C<ExtUtils::Constant>, by Nicholas Clark, is a new tool for extension
811 writers for generating XS code to import C header constants.
812 See L<ExtUtils::Constant>.
816 C<Filter::Simple>, by Damian Conway, is an easy-to-use frontend to
817 Filter::Util::Call. See L<Filter::Simple>.
823 use Filter::Simple sub {
824 while (my ($from, $to) = splice @_, 0, 2) {
833 use MyFilter qr/red/ => 'green';
835 print "red\n"; # this code is filtered, will print "green\n"
836 print "bored\n"; # this code is filtered, will print "bogreen\n"
840 print "red\n"; # this code is not filtered, will print "red\n"
844 C<File::Temp>, by Tim Jenness, allows one to create temporary files
845 and directories in an easy, portable, and secure way. See L<File::Temp>.
850 C<Filter::Util::Call>, by Paul Marquess, provides you with the
851 framework to write I<source filters> in Perl. For most uses, the
852 frontend Filter::Simple is to be preferred. See L<Filter::Util::Call>.
856 C<if>, by Ilya Zakharevich, is a new pragma for conditional inclusion
861 L<libnet>, by Graham Barr, is a collection of perl5 modules related
862 to network programming. See L<Net::FTP>, L<Net::NNTP>, L<Net::Ping>
863 (not part of libnet, but related), L<Net::POP3>, L<Net::SMTP>,
866 Perl installation leaves libnet unconfigured; use F<libnetcfg>
871 C<List::Util>, by Graham Barr, is a selection of general-utility
872 list subroutines, such as sum(), min(), first(), and shuffle().
877 C<Locale::Constants>, C<Locale::Country>, C<Locale::Currency>
878 C<Locale::Language>, and L<Locale::Script>, by Neil Bowers, have
879 been added. They provide the codes for various locale standards, such
880 as "fr" for France, "usd" for US Dollar, and "ja" for Japanese.
884 $country = code2country('jp'); # $country gets 'Japan'
885 $code = country2code('Norway'); # $code gets 'no'
887 See L<Locale::Constants>, L<Locale::Country>, L<Locale::Currency>,
888 and L<Locale::Language>.
892 C<Locale::Maketext>, by Sean Burke, is a localization framework. See
893 L<Locale::Maketext>, and L<Locale::Maketext::TPJ13>. The latter is an
894 article about software localization, originally published in The Perl
895 Journal #13, and republished here with kind permission.
899 C<Math::BigRat> for big rational numbers, to accompany Math::BigInt and
900 Math::BigFloat, from Tels. See L<Math::BigRat>.
904 C<Memoize> can make your functions faster by trading space for time,
905 from Mark-Jason Dominus. See L<Memoize>.
909 C<MIME::Base64>, by Gisle Aas, allows you to encode data in base64,
910 as defined in RFC 2045 - I<MIME (Multipurpose Internet Mail
915 $encoded = encode_base64('Aladdin:open sesame');
916 $decoded = decode_base64($encoded);
918 print $encoded, "\n"; # "QWxhZGRpbjpvcGVuIHNlc2FtZQ=="
924 C<MIME::QuotedPrint>, by Gisle Aas, allows you to encode data
925 in quoted-printable encoding, as defined in RFC 2045 - I<MIME
926 (Multipurpose Internet Mail Extensions)>.
928 use MIME::QuotedPrint;
930 $encoded = encode_qp("Smiley in Unicode: \x{263a}");
931 $decoded = decode_qp($encoded);
933 print $encoded, "\n"; # "Smiley in Unicode: =263A"
935 MIME::QuotedPrint has been enhanced to provide the basic methods
936 necessary to use it with PerlIO::Via as in :
938 use MIME::QuotedPrint;
939 open($fh,">Via(MIME::QuotedPrint)",$path);
941 See L<MIME::QuotedPrint>.
945 C<NEXT>, by Damian Conway, is a pseudo-class for method redispatch.
950 C<open> is a new pragma for setting the default I/O disciplines
955 C<PerlIO::Scalar>, by Nick Ing-Simmons, provides the implementation
956 of IO to "in memory" Perl scalars as discussed above. It also serves
957 as an example of a loadable PerlIO layer. Other future possibilities
958 include PerlIO::Array and PerlIO::Code. See L<PerlIO::Scalar>.
962 C<PerlIO::Via>, by Nick Ing-Simmons, acts as a PerlIO layer and wraps
963 PerlIO layer functionality provided by a class (typically implemented
966 use MIME::QuotedPrint;
967 open($fh,">Via(MIME::QuotedPrint)",$path);
969 This will automatically convert everything output to C<$fh>
970 to Quoted-Printable. See L<PerlIO::Via>.
974 C<Pod::ParseLink>, by Russ Allbery, has been added,
975 to parse LZ<><> links in pods as described in the new
980 C<Pod::Text::Overstrike>, by Joe Smith, has been added.
981 It converts POD data to formatted overstrike text.
982 See L<Pod::Text::Overstrike>. [561+]
986 C<Scalar::Util> is a selection of general-utility scalar subroutines,
987 such as blessed(), reftype(), and tainted(). See L<Scalar::Util>.
991 C<sort> is a new pragma for controlling the behaviour of sort().
995 C<Storable> gives persistence to Perl data structures by allowing the
996 storage and retrieval of Perl data to and from files in a fast and
997 compact binary format. Because in effect Storable does serialisation
998 of Perl data structues, with it you can also clone deep, hierarchical
999 datastructures. Storable was originally created by Raphael Manfredi,
1000 but it is now maintained by Abhijit Menon-Sen. Storable has been
1001 enhanced to understand the two new hash features, Unicode keys and
1002 restricted hashes. See L<Storable>.
1006 C<Switch>, by Damian Conway, has been added. Just by saying
1010 you have C<switch> and C<case> available in Perl.
1016 case 1 { print "number 1" }
1017 case "a" { print "string a" }
1018 case [1..10,42] { print "number in list" }
1019 case (@array) { print "number in list" }
1020 case /\w+/ { print "pattern" }
1021 case qr/\w+/ { print "pattern" }
1022 case (%hash) { print "entry in hash" }
1023 case (\%hash) { print "entry in hash" }
1024 case (\&sub) { print "arg to subroutine" }
1025 else { print "previous case not true" }
1032 C<Test::More>, by Michael Schwern, is yet another framework for writing
1033 test scripts, more extensive than Test::Simple. See L<Test::More>.
1037 C<Test::Simple>, by Michael Schwern, has basic utilities for writing
1038 tests. See L<Test::Simple>.
1042 C<Text::Balanced>, by Damian Conway, has been added, for extracting
1043 delimited text sequences from strings.
1045 use Text::Balanced 'extract_delimited';
1047 ($a, $b) = extract_delimited("'never say never', he never said", "'", '');
1049 $a will be "'never say never'", $b will be ', he never said'.
1051 In addition to extract_delimited(), there are also extract_bracketed(),
1052 extract_quotelike(), extract_codeblock(), extract_variable(),
1053 extract_tagged(), extract_multiple(), gen_delimited_pat(), and
1054 gen_extract_tagged(). With these, you can implement rather advanced
1055 parsing algorithms. See L<Text::Balanced>.
1059 C<threads>, by Arthur Bergman, is an interface to interpreter threads.
1060 Interpreter threads (ithreads) is the new thread model introduced in
1061 Perl 5.6 but only available as an internal interface for extension
1062 writers (and for Win32 Perl for C<fork()> emulation). See L<threads>,
1063 L<threads::shared>, and L<perlthrtut>.
1067 C<threads::shared>, by Arthur Bergman, allows data sharing for
1068 interpreter threads. In the ithreads model any data sharing between
1069 threads must be explicit, as opposed to the old 5.005 thread model
1070 where data sharing was implicit. See L<threads::shared>.
1074 C<Tie::File>, by Mark-Jason Dominus, associates a Perl array with the
1075 lines of a file. See L<Tie::File>.
1079 C<Tie::Memoize>, by Ilya Zakharevich, provides on-demand loaded hashes.
1080 See L<Tie::Memoize>.
1084 C<Tie::RefHash::Nestable>, by Edward Avis, allows storing hash
1085 references (unlike the standard Tie::RefHash) The module is contained
1086 within Tie::RefHash. See L<Tie::RefHash>.
1090 C<Time::HiRes>, by Douglas E. Wegscheid, provides high resolution
1091 timing (ualarm, usleep, and gettimeofday). See L<Time::HiRes>.
1095 C<Unicode::UCD> offers a querying interface to the Unicode Character
1096 Database. See L<Unicode::UCD>.
1100 C<Unicode::Collate>, by SADAHIRO Tomoyuki, implements the UCA
1101 (Unicode Collation Algorithm) for sorting Unicode strings.
1102 See L<Unicode::Collate>.
1106 C<Unicode::Normalize>, by SADAHIRO Tomoyuki, implements the various
1107 Unicode normalization forms. See L<Unicode::Normalize>.
1111 C<XS::Typemap>, by Tim Jenness, is a test extension that exercises
1112 XS typemaps. Nothing gets installed, but the code is worth studying
1113 for extension writers.
1117 =head2 Updated And Improved Modules and Pragmata
1123 The following independently supported modules have been updated to the
1124 newest versions from CPAN: CGI, CPAN, DB_File, File::Spec, File::Temp,
1125 Getopt::Long, Math::BigFloat, Math::BigInt, the podlators bundle
1126 (Pod::Man, Pod::Text), Pod::LaTeX [561+], Pod::Parser, Storable,
1127 Term::ANSIColor, Test, Text-Tabs+Wrap.
1131 attributes::reftype() now works on tied arguments.
1135 AutoLoader can now be disabled with C<no AutoLoader;>.
1139 B::Deparse has been significantly enhanced by Robin Houston. It can
1140 now deparse almost all of the standard test suite (so that the tests
1141 still succeed). There is a make target "test.deparse" for trying this
1146 Carp now has better interface documentation, and the @CARP_NOT
1147 interface has been added to get optional control over where errors
1148 are reported independently of @ISA, by Ben Tilly.
1152 Class::Struct can now define the classes in compile time.
1156 Class::Struct now assigns the array/hash element if the accessor
1157 is called with an array/hash element as the B<sole> argument.
1161 The return value of Cwd::fastcwd() is now tainted.
1165 Data::Dumper now has an option to sort hashes.
1169 Data::Dumper now has an option to dump code references
1174 DB_File now supports newer Berkeley DB versions, among
1179 Devel::Peek now has an interface for the Perl memory statistics
1180 (this works only if you are using perl's malloc, and if you have
1181 compiled with debugging).
1185 The English module can now be used without the infamous performance
1188 use English '-no_match_vars';
1190 (Assuming, of course, that you don't need the troublesome variables
1191 C<$`>, C<$&>, or C<$'>.) Also, introduced C<@LAST_MATCH_START> and
1192 C<@LAST_MATCH_END> English aliases for C<@-> and C<@+>.
1196 ExtUtils::MakeMaker now uses File::Spec internally, which hopefully
1197 leads to better portability.
1201 Fcntl, Socket, and Sys::Syslog have been rewritten by Nicholas Clark
1202 to use the new-style constant dispatch section (see L<ExtUtils::Constant>).
1203 This means that they will be more robust and hopefully faster.
1207 File::Find now chdir()s correctly when chasing symbolic links. [561]
1211 File::Find now has pre- and post-processing callbacks. It also
1212 correctly changes directories when chasing symbolic links. Callbacks
1213 (naughtily) exiting with "next;" instead of "return;" now work.
1217 File::Find is now (again) reentrant. It also has been made
1222 The warnings issued by File::Find now belong to their own category.
1223 You can enable/disable them with C<use/no warnings 'File::Find';>.
1227 File::Glob::glob() has been renamed to File::Glob::bsd_glob()
1228 because the name clashes with the builtin glob(). The older
1229 name is still available for compatibility, but is deprecated. [561]
1233 File::Glob now supports C<GLOB_LIMIT> constant to limit the size of
1234 the returned list of filenames.
1238 IPC::Open3 now allows the use of numeric file descriptors.
1242 IO::Socket now has an atmark() method, which returns true if the socket
1243 is positioned at the out-of-band mark. The method is also exportable
1244 as a sockatmark() function.
1248 IO::Socket::INET failed to open the specified port if the service name
1249 was not known. It now correctly uses the supplied port number as is. [561]
1253 IO::Socket::INET has support for the ReusePort option (if your
1254 platform supports it). The Reuse option now has an alias, ReuseAddr.
1255 For clarity, you may want to prefer ReuseAddr.
1259 IO::Socket::INET now supports a value of zero for C<LocalPort>
1260 (usually meaning that the operating system will make one up.)
1264 'use lib' now works identically to @INC. Removing directories
1265 with 'no lib' now works.
1269 Math::BigFloat and Math::BigInt have undergone a full rewrite by Tels.
1270 They are now magnitudes faster, and they support various bignum
1271 libraries such as GMP and PARI as their backends.
1275 Math::Complex handles inf, NaN etc., better.
1279 Net::Ping has been considerably enhanced by Rob Brown: multihoming is
1280 now supported, Win32 functionality is better, there is now time
1281 measuring functionality (optionally high-resolution using
1282 Time::HiRes), and there is now "external" protocol which uses
1283 Net::Ping::External module which runs your external ping utility and
1284 parses the output. A version of Net::Ping::External is available in
1287 Note that some of the Net::Ping tests are disabled when running
1288 under the Perl distribution since one cannot assume one or more
1289 of the following: enabled echo port at localhost, full Internet
1290 connectivity, or sympathetic firewalls. You can set the environment
1291 variable PERL_TEST_Net_Ping to "1" (one) before running the Perl test
1292 suite to enable all the Net::Ping tests.
1296 POSIX::sigaction() is now much more flexible and robust.
1297 You can now install coderef handlers, 'DEFAULT', and 'IGNORE'
1298 handlers, installing new handlers was not atomic.
1302 In Safe, C<%INC> is now localised in a Safe compartment so that
1307 In SDBM_File on dosish platforms, some keys went missing because of
1308 lack of support for files with "holes". A workaround for the problem
1313 In Search::Dict one can now have a pre-processing hook for the
1314 lines being searched.
1318 The Shell module now has an OO interface.
1322 In Sys::Syslog there is now a failover mechanism that will go
1323 through alternative connection mechanisms until the message
1324 is successfully logged.
1328 The Test module has been significantly enhanced.
1332 Time::Local::timelocal() does not handle fractional seconds anymore.
1333 The rationale is that neither does localtime(), and timelocal() and
1334 localtime() are supposed to be inverses of each other.
1338 The vars pragma now supports declaring fully qualified variables.
1339 (Something that C<our()> does not and will not support.)
1343 The C<utf8::> name space (as in the pragma) provides various
1344 Perl-callable functions to provide low level access to Perl's
1345 internal Unicode representation. At the moment only length()
1346 has been implemented.
1350 =head1 Utility Changes
1356 Emacs perl mode (emacs/cperl-mode.el) has been updated to version
1361 F<emacs/e2ctags.pl> is now much faster.
1365 C<enc2xs> is a tool for people adding their own encodings to the
1370 C<h2ph> now supports C trigraphs.
1374 C<h2xs> now produces a template README.
1378 C<h2xs> now uses C<Devel::PPPort> for better portability between
1379 different versions of Perl.
1383 C<h2xs> uses the new L<ExtUtils::Constant|ExtUtils::Constant> module
1384 which will affect newly created extensions that define constants.
1385 Since the new code is more correct (if you have two constants where the
1386 first one is a prefix of the second one, the first constant B<never>
1387 got defined), less lossy (it uses integers for integer constant,
1388 as opposed to the old code that used floating point numbers even for
1389 integer constants), and slightly faster, you might want to consider
1390 regenerating your extension code (the new scheme makes regenerating
1391 easy). L<h2xs> now also supports C trigraphs.
1395 C<libnetcfg> has been added to configure libnet.
1399 C<perlbug> is now much more robust. It also sends the bug report to
1400 perl.org, not perl.com.
1404 C<perlcc> has been rewritten and its user interface (that is,
1405 command line) is much more like that of the UNIX C compiler, cc.
1406 (The perlbc tools has been removed. Use C<perlcc -B> instead.)
1407 B<Note that perlcc is still considered very experimental and
1412 C<perlivp> is a new Installation Verification Procedure utility
1413 for running any time after installing Perl.
1417 C<piconv> is an implementation of the character conversion utility
1418 C<iconv>, demonstrating the new Encode module.
1422 C<pod2html> now allows specifying a cache directory.
1426 C<pod2html> now produces XHTML 1.0.
1430 C<pod2html> now understands POD written using different line endings
1431 (PC-like CRLF versus UNIX-like LF versus MacClassic-like CR).
1435 C<s2p> has been completely rewritten in Perl. (It is in fact a full
1436 implementation of sed in Perl: you can use the sed functionality by
1437 using the C<psed> utility.)
1441 C<xsubpp> now understands POD documentation embedded in the *.xs
1446 C<xsubpp> now supports the OUT keyword.
1450 =head1 New Documentation
1456 perl56delta details the changes between the 5.005 release and the
1461 perlclib documents the internal replacements for standard C library
1462 functions. (Interesting only for extension writers and Perl core
1467 perldebtut is a Perl debugging tutorial. [561+]
1471 perlebcdic contains considerations for running Perl on EBCDIC
1476 perlintro is a gentle introduction to Perl.
1480 perliol documents the internals of PerlIO with layers.
1484 perlmodstyle is a style guide for writing modules.
1488 perlnewmod tells about writing and submitting a new module. [561+]
1492 perlpacktut is a pack() tutorial.
1496 perlpod has been rewritten to be clearer and to record the best
1497 practices gathered over the years.
1501 perlpodspec is a more formal specification of the pod format,
1502 mainly of interest for writers of pod applications, not to
1503 people writing in pod.
1507 perlretut is a regular expression tutorial. [561+]
1511 perlrequick is a regular expressions quick-start guide.
1512 Yes, much quicker than perlretut. [561]
1516 perltodo has been updated.
1520 perltootc has been renamed as perltooc (to not to conflict
1521 with perltoot in filesystems restricted to "8.3" names).
1525 perluniintro is an introduction to using Unicode in Perl.
1526 (perlunicode is more of a detailed reference and background
1531 perlutil explains the command line utilities packaged with the Perl
1532 distribution. [561+]
1536 The following platform-specific documents are available before
1537 the installation as README.I<platform>, and after the installation
1540 perlaix perlamiga perlapollo perlbeos perlbs2000
1541 perlce perlcygwin perldgux perldos perlepoc perlfreebsd perlhpux
1542 perlhurd perlirix perlmachten perlmacos perlmint perlmpeix
1543 perlnetware perlos2 perlos390 perlplan9 perlqnx perlsolaris
1544 perltru64 perluts perlvmesa perlvms perlvos perlwin32
1546 These documents usually detail one or more of the following subjects:
1547 configuring, building, testing, installing, and sometimes also using
1548 Perl on the said platform.
1550 Eastern Asian Perl users are now welcomed in their own languages:
1551 README.jp (Japanese), README.ko (Korean), README.cn (simplified
1552 Chinese) and README.tw (traditional Chinese), which are written in
1553 normal pod but encoded in EUC-JP, EUC-KR, EUC-CN and Big5. These
1554 will get installed as
1556 perljp perlko perlcn perltw
1562 The documentation for the POSIX-BC platform is called "BS2000", to avoid
1563 confusion with the Perl POSIX module.
1567 The documentation for the WinCE platform is called perlce (README.ce
1568 in the source code kit), to avoid confusion with the perlwin32
1569 documentation on 8.3-restricted filesystems.
1573 =head1 Performance Enhancements
1579 map() could get pathologically slow when the result list it generates
1580 is larger than the source list. The performance has been improved for
1581 common scenarios. [561]
1585 sort() is also fully reentrant, in the sense that the sort function
1586 can itself call sort(). This did not work reliably in previous
1591 sort() has been changed to use primarily mergesort internally as
1592 opposed to the earlier quicksort. For very small lists this may
1593 result in slightly slower sorting times, but in general the speedup
1594 should be at least 20%. Additional bonuses are that the worst case
1595 behaviour of sort() is now better (in computer science terms it now
1596 runs in time O(N log N), as opposed to quicksort's Theta(N**2)
1597 worst-case run time behaviour), and that sort() is now stable
1598 (meaning that elements with identical keys will stay ordered as they
1599 were before the sort). See the C<sort> pragma for information.
1601 The story in more detail: suppose you want to serve yourself a little
1604 @digits = ( 3,1,4,1,5,9 );
1606 A numerical sort of the digits will yield (1,1,3,4,5,9), as expected.
1607 Which C<1> comes first is hard to know, since one C<1> looks pretty
1608 much like any other. You can regard this as totally trivial,
1609 or somewhat profound. However, if you just want to sort the even
1610 digits ahead of the odd ones, then what will
1612 sort { ($a % 2) <=> ($b % 2) } @digits;
1614 yield? The only even digit, C<4>, will come first. But how about
1615 the odd numbers, which all compare equal? With the quicksort algorithm
1616 used to implement Perl 5.6 and earlier, the order of ties is left up
1617 to the sort. So, as you add more and more digits of Pi, the order
1618 in which the sorted even and odd digits appear will change.
1619 and, for sufficiently large slices of Pi, the quicksort algorithm
1620 in Perl 5.8 won't return the same results even if reinvoked with the
1621 same input. The justification for this rests with quicksort's
1622 worst case behavior. If you run
1624 sort { $a <=> $b } ( 1 .. $N , 1 .. $N );
1626 (something you might approximate if you wanted to merge two sorted
1627 arrays using sort), doubling $N doesn't just double the quicksort time,
1628 it I<quadruples> it. Quicksort has a worst case run time that can
1629 grow like N**2, so-called I<quadratic> behaviour, and it can happen
1630 on patterns that may well arise in normal use. You won't notice this
1631 for small arrays, but you I<will> notice it with larger arrays,
1632 and you may not live long enough for the sort to complete on arrays
1633 of a million elements. So the 5.8 quicksort scrambles large arrays
1634 before sorting them, as a statistical defence against quadratic behaviour.
1635 But that means if you sort the same large array twice, ties may be
1636 broken in different ways.
1638 Because of the unpredictability of tie-breaking order, and the quadratic
1639 worst-case behaviour, quicksort was I<almost> replaced completely with
1640 a stable mergesort. I<Stable> means that ties are broken to preserve
1641 the original order of appearance in the input array. So
1643 sort { ($a % 2) <=> ($b % 2) } (3,1,4,1,5,9);
1645 will yield (4,3,1,1,5,9), guaranteed. The even and odd numbers
1646 appear in the output in the same order they appeared in the input.
1647 Mergesort has worst case O(N log N) behaviour, the best value
1648 attainable. And, ironically, this mergesort does particularly
1649 well where quicksort goes quadratic: mergesort sorts (1..$N, 1..$N)
1650 in O(N) time. But quicksort was rescued at the last moment because
1651 it is faster than mergesort on certain inputs and platforms.
1652 For example, if you really I<don't> care about the order of even
1653 and odd digits, quicksort will run in O(N) time; it's very good
1654 at sorting many repetitions of a small number of distinct elements.
1655 The quicksort divide and conquer strategy works well on platforms
1656 with relatively small, very fast, caches. Eventually, the problem gets
1657 whittled down to one that fits in the cache, from which point it
1658 benefits from the increased memory speed.
1660 Quicksort was rescued by implementing a sort pragma to control aspects
1661 of the sort. The B<stable> subpragma forces stable behaviour,
1662 regardless of algorithm. The B<_quicksort> and B<_mergesort>
1663 subpragmas are heavy-handed ways to select the underlying implementation.
1664 The leading C<_> is a reminder that these subpragmas may not survive
1665 beyond 5.8. More appropriate mechanisms for selecting the implementation
1666 exist, but they wouldn't have arrived in time to save quicksort.
1670 Hashes now use Bob Jenkins "One-at-a-Time" hashing key algorithm
1671 ( http://burtleburtle.net/bob/hash/doobs.html ). This algorithm is
1672 reasonably fast while producing a much better spread of values than
1673 the old hashing algorithm (originally by Chris Torek, later tweaked by
1674 Ilya Zakharevich). Hash values output from the algorithm on a hash of
1675 all 3-char printable ASCII keys comes much closer to passing the
1676 DIEHARD random number generation tests. According to perlbench, this
1677 change has not affected the overall speed of Perl.
1681 unshift() should now be noticeably faster.
1685 =head1 Installation and Configuration Improvements
1687 =head2 Generic Improvements
1693 INSTALL now explains how you can configure Perl to use 64-bit
1694 integers even on non-64-bit platforms.
1698 Policy.sh policy change: if you are reusing a Policy.sh file
1699 (see INSTALL) and you use Configure -Dprefix=/foo/bar and in the old
1700 Policy $prefix eq $siteprefix and $prefix eq $vendorprefix, all of
1701 them will now be changed to the new prefix, /foo/bar. (Previously
1702 only $prefix changed.) If you do not like this new behaviour,
1703 specify prefix, siteprefix, and vendorprefix explicitly.
1707 A new optional location for Perl libraries, otherlibdirs, is available.
1708 It can be used for example for vendor add-ons without disturbing Perl's
1709 own library directories.
1713 In many platforms, the vendor-supplied 'cc' is too stripped-down to
1714 build Perl (basically, 'cc' doesn't do ANSI C). If this seems
1715 to be the case and 'cc' does not seem to be the GNU C compiler
1716 'gcc', an automatic attempt is made to find and use 'gcc' instead.
1720 gcc needs to closely track the operating system release to avoid
1721 build problems. If Configure finds that gcc was built for a different
1722 operating system release than is running, it now gives a clearly visible
1723 warning that there may be trouble ahead.
1727 Since Perl 5.8 is not binary-compatible with previous releases
1728 of Perl, Configure no longer suggests including the 5.005
1733 Configure C<-S> can now run non-interactively. [561]
1737 Configure support for pdp11-style memory models has been removed due
1738 to obsolescence. [561]
1742 configure.gnu now works with options with whitespace in them.
1746 installperl now outputs everything to STDERR.
1750 Because PerlIO is now the default on most platforms, "-perlio" doesn't
1751 get appended to the $Config{archname} (also known as $^O) anymore.
1752 Instead, if you explicitly choose not to use perlio (Configure command
1753 line option -Uuseperlio), you will get "-stdio" appended.
1757 Another change related to the architecture name is that "-64all"
1758 (-Duse64bitall, or "maximally 64-bit") is appended only if your
1759 pointers are 64 bits wide. (To be exact, the use64bitall is ignored.)
1763 In AFS installations, one can configure the root of the AFS to be
1764 somewhere else than the default F</afs> by using the Configure
1765 parameter C<-Dafsroot=/some/where/else>.
1769 APPLLIB_EXP, a lesser-known configuration-time definition, has been
1770 documented. It can be used to prepend site-specific directories
1771 to Perl's default search path (@INC); see INSTALL for information.
1775 The version of Berkeley DB used when the Perl (and, presumably, the
1776 DB_File extension) was built is now available as
1777 C<@Config{qw(db_version_major db_version_minor db_version_patch)}>
1778 from Perl and as C<DB_VERSION_MAJOR_CFG DB_VERSION_MINOR_CFG
1779 DB_VERSION_PATCH_CFG> from C.
1783 Building Berkeley DB3 for compatibility modes for DB, NDBM, and ODBM
1784 has been documented in INSTALL.
1788 If you have CPAN access (either network or a local copy such as a
1789 CD-ROM) you can during specify extra modules to Configure to build and
1790 install with Perl using the -Dextras=... option. See INSTALL for
1795 In addition to config.over, a new override file, config.arch, is
1796 available. This file is supposed to be used by hints file writers
1797 for architecture-wide changes (as opposed to config.over which is
1798 for site-wide changes).
1802 If your file system supports symbolic links, you can build Perl outside
1803 of the source directory by
1805 mkdir /tmp/perl/build/directory
1806 cd /tmp/perl/build/directory
1807 sh /path/to/perl/source/Configure -Dmksymlinks ...
1809 This will create in /tmp/perl/build/directory a tree of symbolic links
1810 pointing to files in /path/to/perl/source. The original files are left
1811 unaffected. After Configure has finished, you can just say
1815 and Perl will be built and tested, all in /tmp/perl/build/directory.
1820 For Perl developers, several new make targets for profiling
1821 and debugging have been added; see L<perlhack>.
1827 Use of the F<gprof> tool to profile Perl has been documented in
1828 L<perlhack>. There is a make target called "perl.gprof" for
1829 generating a gprofiled Perl executable.
1833 If you have GCC 3, there is a make target called "perl.gcov" for
1834 creating a gcoved Perl executable for coverage analysis. See
1839 If you are on IRIX or Tru64 platforms, new profiling/debugging options
1840 have been added; see L<perlhack> for more information about pixie and
1847 Guidelines of how to construct minimal Perl installations have
1848 been added to INSTALL.
1852 The Thread extension is now not built at all under ithreads
1853 (C<Configure -Duseithreads>) because it wouldn't work anyway (the
1854 Thread extension requires being Configured with C<-Duse5005threads>).
1856 But note that the Thread.pm interface is now shared by both
1861 The Gconvert macro ($Config{d_Gconvert}) used by perl for stringifying
1862 floating-point numbers is now more picky about using sprintf %.*g
1863 rules for the conversion. Some platforms that used to use gcvt may
1864 now resort to the slower sprintf.
1868 The obsolete method of making a special (e.g., debugging) flavor
1871 make LIBPERL=libperld.a
1873 has been removed. Use -DDEBUGGING instead.
1877 =head2 New Or Improved Platforms
1879 For the list of platforms known to support Perl,
1880 see L<perlport/"Supported Platforms">.
1886 AIX dynamic loading should be now better supported.
1890 AIX should now work better with gcc, threads, and 64-bitness. Also the
1891 long doubles support in AIX should be better now. See L<perlaix>.
1895 AtheOS ( http://www.atheos.cx/ ) is a new platform.
1899 BeOS has been reclaimed.
1903 The DG/UX platform now supports 5.005-style threads.
1908 The DYNIX/ptx platform (a.k.a. dynixptx) is supported at or near
1913 EBCDIC platforms (z/OS (also known as OS/390), POSIX-BC, and VM/ESA)
1914 have been regained. Many test suite tests still fail and the
1915 co-existence of Unicode and EBCDIC isn't quite settled, but the
1916 situation is much better than with Perl 5.6. See L<perlos390>,
1917 L<perlbs2000> (for POSIX-BC), and L<perlvmesa> for more information.
1921 Building perl with -Duseithreads or -Duse5005threads now works under
1922 HP-UX 10.20 (previously it only worked under 10.30 or later). You will
1923 need a thread library package installed. See README.hpux. [561]
1927 Mac OS Classic is now supported in the mainstream source package
1928 (MacPerl has of course been available since perl 5.004 but now the
1929 source code bases of standard Perl and MacPerl have been synchronised)
1934 Mac OS X (or Darwin) should now be able to build Perl even on HFS+
1935 filesystems. (The case-insensitivity used to confuse the Perl build
1940 NCR MP-RAS is now supported. [561]
1944 All the NetBSD specific patches (except for the installation
1945 specific ones) have been merged back to the main distribution.
1949 NetWare from Novell is now supported. See L<perlnetware>.
1953 NonStop-UX is now supported. [561]
1957 NEC SUPER-UX is now supported.
1961 All the OpenBSD specific patches (except for the installation
1962 specific ones) have been merged back to the main distribution.
1966 Perl has been tested with the GNU pth userlevel thread package
1967 ( http://www.gnu.org/software/pth/pth.html ). All thread tests
1968 of Perl now work, but not without adding some yield()s to the tests,
1969 so while pth (and other userlevel thread implementations) can be
1970 considered to be "working" with Perl ithreads, keep in mind the
1971 possible non-preemptability of the underlying thread implementation.
1975 Stratus VOS is now supported using Perl's native build method
1976 (Configure). This is the recommended method to build Perl on
1977 VOS. The older methods, which build miniperl, are still
1978 available. See L<perlvos>. [561+]
1982 The Amdahl UTS UNIX mainframe platform is now supported. [561]
1986 WinCE is now supported. See L<perlce>.
1990 z/OS (formerly known as OS/390, formerly known as MVS OE) now has
1991 support for dynamic loading. This is not selected by default,
1992 however, you must specify -Dusedl in the arguments of Configure. [561]
1996 =head1 Selected Bug Fixes
1998 Numerous memory leaks and uninitialized memory accesses have been
1999 hunted down. Most importantly, anonymous subs used to leak quite
2006 The autouse pragma didn't work for Multi::Part::Function::Names.
2010 caller() could cause core dumps in certain situations. Carp was
2011 sometimes affected by this problem. In particular, caller() now
2012 returns a subroutine name of C<(unknown)> for subroutines that have
2013 been removed from the symbol table.
2017 chop(@list) in list context returned the characters chopped in
2018 reverse order. This has been reversed to be in the right order. [561]
2022 Configure no longer includes the DBM libraries (dbm, gdbm, db, ndbm)
2023 when building the Perl binary. The only exception to this is SunOS 4.x,
2024 which needs them. [561]
2028 The behaviour of non-decimal but numeric string constants such as
2029 "0x23" was platform-dependent: in some platforms that was seen as 35,
2030 in some as 0, in some as a floating point number (don't ask). This
2031 was caused by Perl's using the operating system libraries in a situation
2032 where the result of the string to number conversion is undefined: now
2033 Perl consistently handles such strings as zero in numeric contexts.
2037 The order of DESTROYs has been made more predictable.
2041 Several debugger fixes: exit code now reflects the script exit code,
2042 condition C<"0"> now treated correctly, the C<d> command now checks
2043 line number, C<$.> no longer gets corrupted, and all debugger output
2044 now goes correctly to the socket if RemotePort is set. [561]
2048 Perl 5.6.0 could emit spurious warnings about redefinition of
2049 dl_error() when statically building extensions into perl.
2050 This has been corrected. [561]
2054 L<dprofpp> -R didn't work.
2058 C<*foo{FORMAT}> now works.
2062 Infinity is now recognized as a number.
2066 UNIVERSAL::isa no longer caches methods incorrectly. (This broke
2067 the Tk extension with 5.6.0.) [561]
2071 Lexicals I: lexicals outside an eval "" weren't resolved
2072 correctly inside a subroutine definition inside the eval "" if they
2073 were not already referenced in the top level of the eval""ed code.
2077 Lexicals II: lexicals leaked at file scope into subroutines that
2078 were declared before the lexicals.
2082 Lexical warnings now propagating correctly between scopes
2083 and into C<eval "...">.
2087 C<use warnings qw(FATAL all)> did not work as intended. This has been
2092 warnings::enabled() now reports the state of $^W correctly if the caller
2093 isn't using lexical warnings. [561]
2097 Line renumbering with eval and C<#line> now works. [561]
2101 Fixed numerous memory leaks, especially in eval "".
2105 Localised tied variables no longer leak memory
2108 tie my %tied_hash => 'Tie::StdHash';
2112 # Used to leak memory every time local() was called;
2113 # in a loop, this added up.
2114 local($tied_hash{Foo}) = 1;
2118 Localised hash elements (and %ENV) are correctly unlocalised to not
2119 exist, if they didn't before they were localised.
2123 tie my %tied_hash => 'Tie::StdHash';
2127 # Nothing has set the FOO element so far
2129 { local $tied_hash{FOO} = 'Bar' }
2131 # This used to print, but not now.
2132 print "exists!\n" if exists $tied_hash{FOO};
2134 As a side effect of this fix, tied hash interfaces B<must> define
2135 the EXISTS and DELETE methods.
2139 mkdir() now ignores trailing slashes in the directory name,
2140 as mandated by POSIX.
2144 Some versions of glibc have a broken modfl(). This affects builds
2145 with C<-Duselongdouble>. This version of Perl detects this brokenness
2146 and has a workaround for it. The glibc release 2.2.2 is known to have
2147 fixed the modfl() bug.
2151 Modulus of unsigned numbers now works (4063328477 % 65535 used to
2152 return 27406, instead of 27047). [561]
2156 Some "not a number" warnings introduced in 5.6.0 eliminated to be
2157 more compatible with 5.005. Infinity is now recognised as a number. [561]
2161 Numeric conversions did not recognize changes in the string value
2162 properly in certain circumstances. [561]
2166 Attributes (such as :shared) didn't work with our().
2170 our() variables will not cause bogus "Variable will not stay shared"
2175 "our" variables of the same name declared in two sibling blocks
2176 resulted in bogus warnings about "redeclaration" of the variables.
2177 The problem has been corrected. [561]
2181 pack "Z" now correctly terminates the string with "\0".
2185 Fix password routines which in some shadow password platforms
2186 (e.g. HP-UX) caused getpwent() to return every other entry.
2190 The PERL5OPT environment variable (for passing command line arguments
2191 to Perl) didn't work for more than a single group of options. [561]
2195 PERL5OPT with embedded spaces didn't work.
2199 printf() no longer resets the numeric locale to "C".
2203 C<qw(a\\b)> now parses correctly as C<'a\\b'>: that is, as three
2204 characters, not four. [561]
2208 pos() did not return the correct value within s///ge in earlier
2209 versions. This is now handled correctly. [561]
2213 Printing quads (64-bit integers) with printf/sprintf now works
2214 without the q L ll prefixes (assuming you are on a quad-capable platform).
2218 Regular expressions on references and overloaded scalars now work. [561+]
2222 Right-hand side magic (GMAGIC) could in many cases such as string
2223 concatenation be invoked too many times.
2227 scalar() now forces scalar context even when used in void context.
2231 SOCKS support is now much more robust.
2235 sort() arguments are now compiled in the right wantarray context
2236 (they were accidentally using the context of the sort() itself).
2237 The comparison block is now run in scalar context, and the arguments
2238 to be sorted are always provided list context. [561]
2242 Changed the POSIX character class C<[[:space:]]> to include the (very
2243 rarely used) vertical tab character. Added a new POSIX-ish character
2244 class C<[[:blank:]]> which stands for horizontal whitespace
2245 (currently, the space and the tab).
2249 The tainting behaviour of sprintf() has been rationalized. It does
2250 not taint the result of floating point formats anymore, making the
2251 behaviour consistent with that of string interpolation. [561]
2255 Some cases of inconsistent taint propagation (such as within hash
2256 values) have been fixed.
2260 The RE engine found in Perl 5.6.0 accidentally pessimised certain kinds
2261 of simple pattern matches. These are now handled better. [561]
2265 Regular expression debug output (whether through C<use re 'debug'>
2266 or via C<-Dr>) now looks better. [561]
2270 Multi-line matches like C<"a\nxb\n" =~ /(?!\A)x/m> were flawed. The
2271 bug has been fixed. [561]
2275 Use of $& could trigger a core dump under some situations. This
2276 is now avoided. [561]
2280 The regular expression captured submatches ($1, $2, ...) are now
2281 more consistently unset if the match fails, instead of leaving false
2282 data lying around in them. [561]
2286 readline() on files opened in "slurp" mode could return an extra
2287 "" (blank line) at the end in certain situations. This has been
2292 Autovivification of symbolic references of special variables described
2293 in L<perlvar> (as in C<${$num}>) was accidentally disabled. This works
2298 Sys::Syslog ignored the C<LOG_AUTH> constant.
2302 $AUTOLOAD, sort(), lock(), and spawning subprocesses
2303 in multiple threads simultaneously are now thread-safe.
2307 Tie::Array's SPLICE method was broken.
2311 Allow a read-only string on the left-hand side of a non-modifying tr///.
2315 If C<STDERR> is tied, warnings caused by C<warn> and C<die> now
2316 correctly pass to it.
2320 Several Unicode fixes.
2326 BOMs (byte order marks) at the beginning of Perl files
2327 (scripts, modules) should now be transparently skipped.
2328 UTF-16 and UCS-2 encoded Perl files should now be read correctly.
2332 The character tables have been updated to Unicode 3.2.0.
2336 Comparing with utf8 data does not magically upgrade non-utf8 data
2337 into utf8. (This was a problem for example if you were mixing data
2338 from I/O and Unicode data: your output might have got magically encoded
2343 Generating illegal Unicode code points such as U+FFFE, or the UTF-16
2344 surrogates, now also generates an optional warning.
2348 C<IsAlnum>, C<IsAlpha>, and C<IsWord> now match titlecase.
2352 Concatenation with the C<.> operator or via variable interpolation,
2353 C<eq>, C<substr>, C<reverse>, C<quotemeta>, the C<x> operator,
2354 substitution with C<s///>, single-quoted UTF8, should now work.
2358 The C<tr///> operator now works. Note that the C<tr///CU>
2359 functionality has been removed (but see pack('U0', ...)).
2363 C<eval "v200"> now works.
2367 Perl 5.6.0 parsed m/\x{ab}/ incorrectly, leading to spurious warnings.
2368 This has been corrected. [561]
2372 Zero entries were missing from the Unicode classes such as C<IsDigit>.
2378 Large unsigned numbers (those above 2**31) could sometimes lose their
2379 unsignedness, causing bogus results in arithmetic operations. [561]
2383 The Perl parser has been stress tested using both random input and
2384 Markov chain input and the few found crashes and lockups have been
2389 =head2 Platform Specific Changes and Fixes
2397 Perl now works on post-4.0 BSD/OSes.
2403 Setting C<$0> now works (as much as possible; see L<perlvar> for details).
2409 Numerous updates; currently synchronised with Cygwin 1.3.10.
2413 Previously DYNIX/ptx had problems in its Configure probe for non-blocking I/O.
2419 EPOC now better supported. See README.epoc. [561]
2425 Perl now works on post-3.0 FreeBSDs.
2431 README.hpux updated; C<Configure -Duse64bitall> now works;
2432 now uses HP-UX malloc instead of Perl malloc.
2438 Numerous compilation flag and hint enhancements; accidental mixing
2439 of 32-bit and 64-bit libraries (a doomed attempt) made much harder.
2449 Long doubles should now work (see INSTALL). [561]
2453 Linux previously had problems related to sockaddrlen when using
2454 accept(), recvfrom() (in Perl: recv()), getpeername(), and
2463 Compilation of the standard Perl distribution in Mac OS Classic should
2464 now work if you have the Metrowerks development environment and the
2465 missing Mac-specific toolkit bits. Contact the macperl mailing list
2472 MPE/iX update after Perl 5.6.0. See README.mpeix. [561]
2476 NetBSD/threads: try installing the GNU pth (should be in the
2477 packages collection, or http://www.gnu.org/software/pth/),
2478 and Configure with -Duseithreads.
2484 Perl now works on NetBSD/sparc.
2490 Now works with usethreads (see INSTALL). [561]
2496 64-bitness using the Sun Workshop compiler now works.
2502 The native build method requires at least VOS Release 14.5.0
2503 and GNU C++/GNU Tools 2.0.1 or later. The Perl pack function
2504 now maps overflowed values to +infinity and underflowed values
2509 Tru64 (aka Digital UNIX, aka DEC OSF/1)
2511 The operating system version letter now recorded in $Config{osvers}.
2512 Allow compiling with gcc (previously explicitly forbidden). Compiling
2513 with gcc still not recommended because buggy code results, even with
2520 Fixed various alignment problems that lead into core dumps either
2521 during build or later; no longer dies on math errors at runtime;
2522 now using full quad integers (64 bits), previously was using
2523 only 46 bit integers for speed.
2529 See L</"Socket Extension Dynamic in VMS"> and L</"IEEE-format Floating Point
2530 Default on OpenVMS Alpha"> for important changes not otherwise listed here.
2532 chdir() now works better despite a CRT bug; now works with MULTIPLICITY
2533 (see INSTALL); now works with Perl's malloc.
2535 The tainting of C<%ENV> elements via C<keys> or C<values> was previously
2536 unimplemented. It now works as documented.
2538 The C<waitpid> emulation has been improved. The worst bug (now fixed)
2539 was that a pid of -1 would cause a wildcard search of all processes on
2542 POSIX-style signals are now emulated much better on VMS versions prior
2545 The C<system> function and backticks operator have improved
2546 functionality and better error handling. [561]
2548 File access tests now use current process privileges rather than the
2549 user's default privileges, which could sometimes result in a mismatch
2550 between reported access and actual access.
2552 There is a new C<kill> implementation based on C<sys$sigprc> that allows
2553 older VMS systems (pre-7.0) to use C<kill> to send signals rather than
2554 simply force exit. This implementation also allows later systems to
2555 call C<kill> from within a signal handler.
2557 Iterative logical name translations are now limited to 10 iterations in
2558 imitation of SHOW LOGICAL and other OpenVMS facilities.
2568 accept() no longer leaks memory. [561]
2572 Borland C++ v5.5 is now a supported compiler that can build Perl.
2573 However, the generated binaries continue to be incompatible with those
2574 generated by the other supported compilers (GCC and Visual C++). [561]
2578 Better chdir() return value for a non-existent directory.
2582 Duping socket handles with open(F, ">&MYSOCK") now works under Windows
2587 New %ENV entries now propagate to subprocesses. [561]
2591 Current directory entries in %ENV are now correctly propagated to child
2596 $ENV{LIB} now used to search for libs under Visual C.
2600 fork() emulation has been improved in various ways, but still continues
2601 to be experimental. See L<perlfork> for known bugs and caveats. [561+]
2605 A failed (pseudo)fork now returns undef and sets errno to EAGAIN.
2609 Win32::GetCwd() correctly returns C:\ instead of C: when at the drive root.
2610 Other bugs in chdir() and Cwd::cwd() have also been fixed. [561]
2614 HTML files will be installed in c:\perl\html instead of c:\perl\lib\pod\html
2618 The makefiles now provide a single switch to bulk-enable all the
2619 features enabled in ActiveState ActivePerl (a popular Win32 binary
2620 distribution). [561]
2624 Allow REG_EXPAND_SZ keys in the registry.
2628 Can now send() from all threads, not just the first one. [561]
2632 ExtUtils::MakeMaker now uses $ENV{LIB} to search for libraries.
2636 Fake signal handling reenabled, bugs and all.
2640 %SIG has been enabled under USE_ITHREADS, but its use is completely
2641 unsupported under all configurations. [561]
2645 Less stack reserved per thread so that more threads can run
2646 concurrently. (Still 16M per thread.)
2650 C<< File::Spec->tmpdir() >> now prefers C:/temp over /tmp
2651 (works better when perl is running as service).
2655 Better UNC path handling under ithreads. [561]
2659 wait(), waitpid(), and backticks now return the correct exit status
2660 under Windows 9x. [561]
2664 Non-blocking waits for child processes (or pseudo-processes) are
2665 supported via C<waitpid($pid, &POSIX::WNOHANG)>.
2669 Win64 compilation is now supported.
2673 winsock handle leak fixed. [561]
2679 =head1 New or Changed Diagnostics
2685 The lexical warnings category "deprecated" is no longer a sub-category
2686 of the "syntax" category. It is now a top-level category in its own
2691 All regular expression compilation error messages are now hopefully
2692 easier to understand both because the error message now comes before
2693 the failed regex and because the point of failure is now clearly
2694 marked by a C<E<lt>-- HERE> marker.
2698 The various "opened only for", "on closed", "never opened" warnings
2699 drop the C<main::> prefix for filehandles in the C<main> package,
2700 for example C<STDIN> instead of C<main::STDIN>.
2704 The "Unrecognized escape" warning has been extended to include C<\8>,
2705 C<\9>, and C<\_>. There is no need to escape any of the C<\w> characters.
2709 Two new debugging options have been added: if you have compiled your
2710 Perl with debugging, you can use the -DT [561] and -DR options to trace
2711 tokenising and to add reference counts to displaying variables,
2716 The debugger (perl5db.pl) has been modified to present a more
2717 consistent commands interface, via (CommandSet=580). perl5db.t was
2718 also added to test the changes, and as a placeholder for further tests.
2724 The debugger has a new C<dumpDepth> option to control the maximum
2725 depth to which nested structures are dumped. The C<x> command has
2726 been extended so that C<x N EXPR> dumps out the value of I<EXPR> to a
2727 depth of at most I<N> levels.
2731 The debugger can now show lexical variables if you have the CPAN
2732 module PadWalker installed.
2736 If an attempt to use a (non-blessed) reference as an array index
2737 is made, a warning is given.
2741 C<push @a;> and C<unshift @a;> (with no values to push or unshift)
2742 now give a warning. This may be a problem for generated and evaled
2747 If you try to L<perlfunc/pack> a number less than 0 or larger than 255
2748 using the C<"C"> format you will get an optional warning. Similarly
2749 for the C<"c"> format and a number less than -128 or more than 127.
2753 Certain regex modifiers such as C<(?o)> make sense only if applied to
2754 the entire regex. You will get an optional warning if you try to do
2759 Using arrays or hashes as references (e.g. C<< %foo->{bar} >>
2760 has been deprecated for a while. Now you will get an optional warning.
2764 Using C<sort> in scalar context now issues an optional warning.
2765 This didn't do anything useful, as the sort was not performed.
2769 =head1 Changed Internals
2775 perlapi.pod (a companion to perlguts) now attempts to document the
2780 You can now build a really minimal perl called microperl.
2781 Building microperl does not require even running Configure;
2782 C<make -f Makefile.micro> should be enough. Beware: microperl makes
2783 many assumptions, some of which may be too bold; the resulting
2784 executable may crash or otherwise misbehave in wondrous ways.
2785 For careful hackers only.
2789 Added rsignal(), whichsig(), do_join(), op_clear, op_null,
2790 ptr_table_clear(), ptr_table_free(), sv_setref_uv(), and several UTF-8
2791 interfaces to the publicised API. For the full list of the available
2792 APIs see L<perlapi>.
2796 Made possible to propagate customised exceptions via croak()ing.
2800 Now xsubs can have attributes just like subs. (Well, at least the
2801 built-in attributes.)
2805 dTHR and djSP have been obsoleted; the former removed (because it's
2806 a no-op) and the latter replaced with dSP.
2810 PERL_OBJECT has been completely removed.
2814 The MAGIC constants (e.g. C<'P'>) have been macrofied
2815 (e.g. C<PERL_MAGIC_TIED>) for better source code readability
2816 and maintainability.
2820 The regex compiler now maintains a structure that identifies nodes in
2821 the compiled bytecode with the corresponding syntactic features of the
2822 original regex expression. The information is attached to the new
2823 C<offsets> member of the C<struct regexp>. See L<perldebguts> for more
2824 complete information.
2828 The C code has been made much more C<gcc -Wall> clean. Some warning
2829 messages still remain in some platforms, so if you are compiling with
2830 gcc you may see some warnings about dubious practices. The warnings
2831 are being worked on.
2835 F<perly.c>, F<sv.c>, and F<sv.h> have now been extensively commented.
2839 Documentation on how to use the Perl source repository has been added
2840 to F<Porting/repository.pod>.
2844 There are now several profiling make targets.
2848 =head1 Security Vulnerability Closed [561]
2850 (This change was already made in 5.7.0 but bears repeating here.)
2852 A potential security vulnerability in the optional suidperl component
2853 of Perl was identified in August 2000. suidperl is neither built nor
2854 installed by default. As of November 2001 the only known vulnerable
2855 platform is Linux, most likely all Linux distributions. CERT and
2856 various vendors and distributors have been alerted about the vulnerability.
2857 See http://www.cpan.org/src/5.0/sperl-2000-08-05/sperl-2000-08-05.txt
2858 for more information.
2860 The problem was caused by Perl trying to report a suspected security
2861 exploit attempt using an external program, /bin/mail. On Linux
2862 platforms the /bin/mail program had an undocumented feature which
2863 when combined with suidperl gave access to a root shell, resulting in
2864 a serious compromise instead of reporting the exploit attempt. If you
2865 don't have /bin/mail, or if you have 'safe setuid scripts', or if
2866 suidperl is not installed, you are safe.
2868 The exploit attempt reporting feature has been completely removed from
2869 Perl 5.8.0 (and the maintenance release 5.6.1, and it was removed also
2870 from all the Perl 5.7 releases), so that particular vulnerability
2871 isn't there anymore. However, further security vulnerabilities are,
2872 unfortunately, always possible. The suidperl functionality is most
2873 probably going to be removed in Perl 5.10. In any case, suidperl
2874 should only be used by security experts who know exactly what they are
2875 doing and why they are using suidperl instead of some other solution
2876 such as sudo ( see http://www.courtesan.com/sudo/ ).
2880 Several new tests have been added, especially for the F<lib> and F<ext>
2881 subsections. There are now about 65 000 individual tests (spread over
2882 about 700 test scripts), in the regression suite (5.6.1 has about
2883 11700 tests, in 258 test scripts) Many of the new tests are of course
2884 introduced by the new modules, but still in general Perl is now more
2887 Because of the large number of tests, running the regression suite
2888 will take considerably longer time than it used to: expect the suite
2889 to take up to 4-5 times longer to run than in perl 5.6. On a really
2890 fast machine you can hope to finish the suite in about 6-8 minutes
2893 The tests are now reported in a different order than in earlier Perls.
2894 (This happens because the test scripts from under t/lib have been moved
2895 to be closer to the library/extension they are testing.)
2897 =head1 Known Problems
2905 If using the AIX native make command, instead of just "make" issue
2906 "make all". In some setups the former has been known to spuriously
2907 also try to run "make install". Alternatively, you may want to use
2912 In AIX 4.2, Perl extensions that use C++ functions that use statics
2913 may have problems in that the statics are not getting initialized.
2914 In newer AIX releases, this has been solved by linking Perl with
2915 the libC_r library, but unfortunately in AIX 4.2 the said library
2916 has an obscure bug where the various functions related to time
2917 (such as time() and gettimeofday()) return broken values, and
2918 therefore in AIX 4.2 Perl is not linked against libC_r.
2922 vac 5.0.0.0 May Produce Buggy Code For Perl
2924 The AIX C compiler vac version 5.0.0.0 may produce buggy code,
2925 resulting in a few random tests failing when run as part of "make
2926 test", but when the failing tests are run by hand, they succeed.
2927 We suggest upgrading to at least vac version 5.0.1.0, that has been
2928 known to compile Perl correctly. "lslpp -L|grep vac.C" will tell
2929 you the vac version. See README.aix.
2933 If building threaded Perl, you may get compilation warning from pp_sys.c:
2935 "pp_sys.c", line 4651.39: 1506-280 (W) Function argument assignment between types "unsigned char*" and "const void*" is not allowed.
2937 This is harmless; it is caused by the getnetbyaddr() and getnetbyaddr_r()
2938 having slightly different types for their first argument.
2942 =head2 Alpha systems with old gccs fail several tests
2944 If you see op/pack, op/pat, op/regexp, or ext/Storable tests failing
2945 in a Linux/alpha or *BSD/Alpha, it's probably time to upgrade your gcc.
2946 gccs prior to 2.95.3 are definitely not good enough, and gcc 3.1 may
2947 be even better. (RedHat Linux/alpha with gcc 3.1 reported no problems,
2948 as did Linux 2.4.18 with gcc 2.95.4.) (In Tru64, it is preferable to
2949 use the bundled C compiler.)
2953 Perl 5.8.0 doesn't build in AmigaOS. It broke at some point during
2954 the ithreads work and we could not find Amiga experts to unbreak the
2955 problems. Perl 5.6.1 still works for AmigaOS (as does the the 5.7.2
2956 development release).
2960 The following tests fail on 5.8.0 Perl in BeOS Personal 5.03:
2962 t/op/lfs............................FAILED at test 17
2963 t/op/magic..........................FAILED at test 24
2964 ext/POSIX/t/sigaction...............FAILED at test 13
2965 ext/POSIX/t/waitpid.................FAILED at test 1
2967 See L<perlbeos> (README.beos) for more details.
2969 =head2 Cygwin "unable to remap"
2971 For example when building the Tk extension for Cygwin,
2972 you may get an error message saying "unable to remap".
2973 This is known problem with Cygwin, and a workaround is
2974 detailed in here: http://sources.redhat.com/ml/cygwin/2001-12/msg00894.html
2976 =head2 ext/threads/t/libc
2978 If this test fails, it indicates that your libc (C library) is not
2979 threadsafe. This particular test stress tests the localtime() call to
2980 find out whether it is threadsafe. See L<perlthrtut> for more information.
2982 =head2 FreeBSD built with ithreads coredumps reading large directories
2984 This is a known bug in FreeBSD's readdir_r() (see L<perlfreebsd>
2985 (README.freebsd)), which hopefully will be fixed in FreeBSD 4.6.
2987 =head2 FreeBSD Failing locale Test 117 For ISO8859-15 Locales
2989 The ISO8859-15 locales may fail the locale test 117 in FreeBSD.
2990 This is caused by the characters \xFF (y with diaeresis) and \xBE
2991 (Y with diaeresis) not behaving correctly when being matched
2994 =head2 IRIX fails ext/List/Util/t/shuffle.t
2996 IRIX with MIPSpro 7.3.1.3m compiler may fail the said List::Util test
2997 by dumping core. This seems to be a compiler error since if compiled
2998 with gcc no core dump ensues, and no failures on the said test on any
3001 =head2 Modifying $_ Inside for(..)
3005 works without complaint. It shouldn't. (You should be able to
3006 modify only lvalue elements inside the loops.) You can see the
3007 correct behaviour by replacing the 1..5 with 1, 2, 3, 4, 5.
3009 =head2 mod_perl 1.26 Doesn't Build With Threaded Perl
3011 Use mod_perl 1.27 or higher.
3013 =head2 lib/ftmp-security tests warn 'system possibly insecure'
3015 Don't panic. Read the 'make test' section of INSTALL instead.
3017 =head2 HP-UX lib/posix Subtest 9 Fails When LP64-Configured
3019 If perl is configured with -Duse64bitall, the successful result of the
3020 subtest 10 of lib/posix may arrive before the successful result of the
3021 subtest 9, which confuses the test harness so much that it thinks the
3024 =head2 Linux with glibc 2.2.5 fails t/op/int subtest #6 with -Duse64bitint
3026 This is a known bug in the glibc 2.2.5 with long long integers.
3027 ( http://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=65612 )
3029 =head2 Linux With Sfio Fails op/misc Test 48
3033 =head2 libwww-perl (LWP) fails base/date #51
3035 Use libwww-perl 5.65 or later.
3039 Please remember to set your environment variable LC_ALL to "C"
3040 (setenv LC_ALL C) before running "make test" to avoid a lot of
3041 warnings about the broken locales of Mac OS X.
3043 The following tests are known to fail in Mac OS X 10.1.5 because of
3044 buggy (old) implementations of Berkeley DB included in Mac OS X:
3046 Failed Test Stat Wstat Total Fail Failed List of Failed
3047 -------------------------------------------------------------------------
3048 ../ext/DB_File/t/db-btree.t 0 11 ?? ?? % ??
3049 ../ext/DB_File/t/db-recno.t 149 3 2.01% 61 63 65
3051 If you are building on a UFS partition, you will also probably see
3052 t/op/stat.t subtest #9 fail. This is caused by Darwin's UFS not
3053 supporting inode change time.
3055 Also the ext/POSIX/t/posix.t subtest #10 fails but it is skipped for
3056 now because the failure is Apple's fault, not Perl's (blocked signals
3059 If you Configure with ithreads, ext/threads/t/libc.t will fail. Again,
3060 this is not Perl's fault-- the libc of Mac OS X is not threadsafe
3061 (in this particular test, the localtime() call is found to be
3064 =head2 op/sprintf tests 91, 129, and 130
3066 The op/sprintf tests 91, 129, and 130 are known to fail on some platforms.
3067 Examples include any platform using sfio, and Compaq/Tandem's NonStop-UX.
3069 Test 91 is known to fail on QNX6 (nto), because C<sprintf '%e',0>
3070 incorrectly produces C<0.000000e+0> instead of C<0.000000e+00>.
3072 For tests 129 and 130, the failing platforms do not comply with
3073 the ANSI C Standard: lines 19ff on page 134 of ANSI X3.159 1989, to
3074 be exact. (They produce something other than "1" and "-1" when
3075 formatting 0.6 and -0.6 using the printf format "%.0f"; most often,
3076 they produce "0" and "-0".)
3080 In case you are still using Solaris 2.5 (aka SunOS 5.5), you may
3081 experience failures (the test core dumping) in lib/locale.t.
3082 The suggested cure is to upgrade your Solaris.
3084 =head2 SUPER-UX (NEC SX)
3086 The following tests are known to fail on SUPER-UX:
3088 op/64bitint...........................FAILED tests 29-30, 32-33, 35-36
3089 op/arith..............................FAILED tests 128-130
3090 op/pack...............................FAILED tests 25-5625
3091 op/pow................................
3092 op/taint..............................# msgsnd failed
3093 ../ext/IO/lib/IO/t/io_poll............FAILED tests 3-4
3094 ../ext/IPC/SysV/ipcsysv...............FAILED tests 2, 5-6
3095 ../ext/IPC/SysV/t/msg.................FAILED tests 2, 4-6
3096 ../ext/Socket/socketpair..............FAILED tests 12
3097 ../lib/IPC/SysV.......................FAILED tests 2, 5-6
3098 ../lib/warnings.......................FAILED tests 115-116, 118-119
3100 The op/pack failure ("Cannot compress negative numbers at op/pack.t line 126")
3101 is serious but as of yet unsolved. It points at some problems with the
3102 signedness handling of the C compiler, as do the 64bitint, arith, and pow
3103 failures. Most of the rest point at problems with SysV IPC.
3105 =head2 Term::ReadKey not working on Win32
3107 Use Term::ReadKey 2.20 or later.
3109 =head2 Failure of Thread (5.005-style) tests
3111 B<Note that support for 5.005-style threading is deprecated,
3112 experimental and practically unsupported. In 5.10, it is expected
3115 The following tests are known to fail due to fundamental problems in
3116 the 5.005 threading implementation. These are not new failures--Perl
3117 5.005_0x has the same bugs, but didn't have these tests.
3119 ../ext/B/t/xref.t 255 65280 14 12 85.71% 3-14
3120 ../ext/List/Util/t/first.t 255 65280 7 4 57.14% 2 5-7
3121 ../lib/English.t 2 512 54 2 3.70% 2-3
3122 ../lib/ExtUtils/t/basic.t 1 256 17 1 5.88% 14
3123 ../lib/FileCache.t 5 1 20.00% 5
3124 ../lib/Filter/Simple/t/data.t 6 3 50.00% 1-3
3125 ../lib/Filter/Simple/t/filter_onl 9 3 33.33% 1-2 5
3126 ../lib/Tie/File/t/31_autodefer.t 255 65280 65 32 49.23% 34-65
3127 ../lib/autouse.t 10 1 10.00% 4
3128 op/flip.t 15 1 6.67% 15
3130 These failures are unlikely to get fixed as 5.005-style threads
3131 are considered fundamentally broken. (Basically what happens is that
3132 competing threads can corrupt shared global state.)
3134 =head2 Timing problems
3136 The following tests may fail intermittently because of timing
3137 problems, for example if the system is heavily loaded.
3140 ext/Time/HiRes/HiRes.t
3142 lib/Memoize/t/expmod_t.t
3143 lib/Memoize/t/speed.t
3145 In case of failure please try running them manually, for example
3147 ./perl -Ilib ext/Time/HiRes/HiRes.t
3151 ../lib/Math/Trig.t 26 1 3.85% 25
3152 ../lib/warnings.t 470 1 0.21% 429
3154 The Trig.t failure is caused by the slighly differing (from IEEE)
3155 floating point implementation of UNICOS. The warnings.t failure is
3156 also related: the test assumes a certain floating point output format;
3157 this assumption fails in UNICOS.
3165 During Configure, the test
3167 Guessing which symbols your C compiler and preprocessor define...
3169 will probably fail with error messages like
3171 CC-20 cc: ERROR File = try.c, Line = 3
3172 The identifier "bad" is undefined.
3174 bad switch yylook 79bad switch yylook 79bad switch yylook 79bad switch yylook 79#ifdef A29K
3177 CC-65 cc: ERROR File = try.c, Line = 3
3178 A semicolon is expected at this point.
3180 This is caused by a bug in the awk utility of UNICOS/mk. You can ignore
3181 the error, but it does cause a slight problem: you cannot fully
3182 benefit from the h2ph utility (see L<h2ph>) that can be used to
3183 convert C headers to Perl libraries, mainly used to be able to access
3184 from Perl the constants defined using C preprocessor, cpp. Because of
3185 the above error, parts of the converted headers will be invisible.
3186 Luckily, these days the need for h2ph is rare.
3190 If building Perl with interpreter threads (ithreads), the
3191 getgrent(), getgrnam(), and getgrgid() functions cannot return the
3192 list of the group members due to a bug in the multithreaded support of
3193 UNICOS/mk. What this means is that in list context the functions will
3194 return only three values, not four.
3200 There are a few known test failures, see L<perluts> (README.uts).
3202 =head2 VOS (Stratus)
3204 When Perl is built using the native build process on VOS Release
3205 14.5.0 and GNU C++/GNU Tools 2.0.1, all attempted tests either
3206 pass or result in TODO (ignored) failures.
3210 There should be no reported test failures with a default configuration,
3211 though there are a number of tests marked TODO that point to areas
3212 needing further debugging and/or porting work.
3216 In multi-CPU boxes, there are some problems with the I/O buffering:
3217 some output may appear twice.
3219 =head2 XML::Parser not working
3221 Use XML::Parser 2.31 or later.
3223 =head2 z/OS (OS/390)
3225 z/OS has rather many test failures but the situation is actually
3226 better than it was in 5.6.0; it's just that so many new modules and
3227 tests have been added.
3229 Failed Test Stat Wstat Total Fail Failed List of Failed
3230 ---------------------------------------------------------------------------
3231 ../ext/Data/Dumper/t/dumper.t 357 8 2.24% 311 314 325 327
3233 ../ext/IO/lib/IO/t/io_unix.t 5 4 80.00% 2-5
3234 ../ext/Storable/t/downgrade.t 12 3072 169 12 7.10% 14-15 46-47 78-79
3236 ../lib/ExtUtils/t/Constant.t 121 30976 48 48 100.00% 1-48
3237 ../lib/ExtUtils/t/Embed.t 9 9 100.00% 1-9
3238 op/pat.t 910 7 0.77% 665 776 785 832-
3240 op/sprintf.t 224 3 1.34% 98 100 136
3241 op/tr.t 97 5 5.15% 63 71-74
3242 uni/fold.t 780 6 0.77% 61 169 196 661
3245 The failures in dumper.t and downgrade.t are problems in the tests,
3246 those in io_unix and sprintf are problems in the USS (UDP sockets
3247 and printf formats). The pat, tr, and fold failures are genuine Perl
3248 problems caused by EBCDIC (and in the pat and fold cases, combining
3249 that with Unicode). The Constant and Embed are probably problems
3250 in the tests (since they test Perl's ability to build extensions,
3251 and that seems to be working reasonably well.)
3253 =head2 Localising Tied Arrays and Hashes Is Broken
3257 doesn't work as one would expect: the old value is restored
3258 incorrectly. This will be changed in a future release, but we don't
3259 know yet what the new semantics will exactly be. In any case, the
3260 change will break existing code that relies on the current
3261 (ill-defined) semantics, so just avoid doing this in general.
3263 =head2 Self-tying Problems
3265 Self-tying of arrays and hashes is broken in rather deep and
3266 hard-to-fix ways. As a stop-gap measure to avoid people from getting
3267 frustrated at the mysterious results (core dumps, most often), it is
3268 forbidden for now (you will get a fatal error even from an attempt).
3270 A change to self-tying of globs has caused them to be recursively
3271 referenced (see: L<perlobj/"Two-Phased Garbage Collection">). You
3272 will now need an explicit untie to destroy a self-tied glob. This
3273 behaviour may be fixed at a later date.
3275 Self-tying of scalars and IO thingies works.
3277 =head2 Building Extensions Can Fail Because Of Largefiles
3279 Some extensions like mod_perl are known to have issues with
3280 `largefiles', a change brought by Perl 5.6.0 in which file offsets
3281 default to 64 bits wide, where supported. Modules may fail to compile
3282 at all, or they may compile and work incorrectly. Currently, there
3283 is no good solution for the problem, but Configure now provides
3284 appropriate non-largefile ccflags, ldflags, libswanted, and libs
3285 in the %Config hash (e.g., $Config{ccflags_nolargefiles}) so the
3286 extensions that are having problems can try configuring themselves
3287 without the largefileness. This is admittedly not a clean solution,
3288 and the solution may not even work at all. One potential failure is
3289 whether one can (or, if one can, whether it's a good idea to) link
3290 together at all binaries with different ideas about file offsets;
3291 all this is platform-dependent.
3293 =head2 Unicode Support on EBCDIC Still Spotty
3295 Though mostly working, Unicode support still has problem spots on
3296 EBCDIC platforms. One such known spot are the C<\p{}> and C<\P{}>
3297 regular expression constructs for code points less than 256: the
3298 C<pP> are testing for Unicode code points, not knowing about EBCDIC.
3300 =head2 The Compiler Suite Is Still Very Experimental
3302 The compiler suite is slowly getting better but it continues to be
3303 highly experimental. Use in production environments is discouraged.
3305 =head2 The Long Double Support Is Still Experimental
3307 The ability to configure Perl's numbers to use "long doubles",
3308 floating point numbers of hopefully better accuracy, is still
3309 experimental. The implementations of long doubles are not yet
3310 widespread and the existing implementations are not quite mature
3311 or standardised, therefore trying to support them is a rare
3312 and moving target. The gain of more precision may also be offset
3313 by slowdown in computations (more bits to move around, and the
3314 operations are more likely to be executed by less optimised
3317 =head2 Seen In Perl 5.7 But Gone Now
3319 C<Time::Piece> (previously known as C<Time::Object>) was removed
3320 because it was felt that it didn't have enough value in it to be a
3321 core module. It is still a useful module, though, and is available
3324 Perl 5.8 unfortunately does not build anymore on AmigaOS; this broke
3325 accidentally at some point. Since there are not that many Amiga
3326 developers available, we could not get this fixed and tested in time
3327 for 5.8.0. Perl 5.6.1 still works for AmigaOS (as does the the 5.7.2
3328 development release).
3330 =head1 Reporting Bugs
3332 If you find what you think is a bug, you might check the articles
3333 recently posted to the comp.lang.perl.misc newsgroup and the perl
3334 bug database at http://bugs.perl.org/ . There may also be
3335 information at http://www.perl.com/ , the Perl Home Page.
3337 If you believe you have an unreported bug, please run the B<perlbug>
3338 program included with your release. Be sure to trim your bug down
3339 to a tiny but sufficient test case. Your bug report, along with the
3340 output of C<perl -V>, will be sent off to perlbug@perl.org to be
3341 analysed by the Perl porting team.
3345 The F<Changes> file for exhaustive details on what changed.
3347 The F<INSTALL> file for how to build Perl.
3349 The F<README> file for general stuff.
3351 The F<Artistic> and F<Copying> files for copyright information.
3355 Written by Jarkko Hietaniemi <F<jhi@iki.fi>>.