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 In some cases the said bug/feature may have been further fixed/enhanced
18 You can see the list of changes in the 5.6.1 (from 5.6.0) by reading
21 If you are upgrading from Perl 5.005_03, you might also want
22 to read L<perl56delta>.
24 =head1 Highlights In 5.8.0
30 Better Unicode support
34 New Thread Implementation
42 Better Numeric Accuracy
50 More Extensive Regression Testing
54 =head1 Incompatible Changes
56 =head2 Binary Incompatibility
58 B<Perl 5.8 is not binary compatible with earlier releases of Perl.>
60 B<You have to recompile your XS modules.>
62 (Pure Perl modules should continue to work.)
64 The major reason for the discontinuity is the new IO architecture
65 called PerlIO. PerlIO is the default configuration because without
66 it many new features of Perl 5.8 cannot be used. In other words:
67 you just have to recompile your modules containing XS code, sorry
70 In future releases of Perl, non-PerlIO aware XS modules may become
71 completely unsupported. This shouldn't be too difficult for module
72 authors, however: PerlIO has been designed as a drop-in replacement
73 (at the source code level) for the stdio interface.
75 Depending on your platform, there are also other reasons why
76 we decided to break binary compatibility, please read on.
78 =head2 64-bit platforms and malloc
80 If your pointers are 64 bits wide, the Perl malloc is no longer being
81 used because it does not work well with 8-byte pointers. Also,
82 usually the system mallocs on such platforms are much better optimized
83 for such large memory models than the Perl malloc. Some memory-hungry
84 Perl applications like the PDL don't work well with Perl's malloc.
85 Finally, other applications than Perl (such as mod_perl) tend to prefer
86 the system malloc. Such platforms include Alpha and 64-bit HPPA,
89 =head2 AIX Dynaloading
91 The AIX dynaloading now uses in AIX releases 4.3 and newer the native
92 dlopen interface of AIX instead of the old emulated interface. This
93 change will probably break backward compatibility with compiled
94 modules. The change was made to make Perl more compliant with other
95 applications like mod_perl which are using the AIX native interface.
97 =head2 Attributes for C<my> variables now handled at run-time.
99 The C<my EXPR : ATTRS> syntax now applies variable attributes at
100 run-time. (Subroutine and C<our> variables still get attributes applied
101 at compile-time.) See L<attributes> for additional details. In particular,
102 however, this allows variable attributes to be useful for C<tie> interfaces,
103 which was a deficiency of earlier releases. Note that the new semantics
104 doesn't work with the Attribute::Handlers module (as of version 0.76).
106 =head2 Socket Extension Dynamic in VMS
108 The Socket extension is now dynamically loaded instead of being
109 statically built in. This may or may not be a problem with ancient
110 TCP/IP stacks of VMS: we do not know since we weren't able to test
111 Perl in such configurations.
113 =head2 IEEE-format Floating Point Default on OpenVMS Alpha
115 Perl now uses IEEE format (T_FLOAT) as the default internal floating
116 point format on OpenVMS Alpha, potentially breaking binary compatibility
117 with external libraries or existing data. G_FLOAT is still available as
118 a configuration option. The default on VAX (D_FLOAT) has not changed.
120 =head2 New Unicode Properties
122 Unicode I<scripts> are now supported. Scripts are similar to (and superior
123 to) Unicode I<blocks>. The difference between scripts and blocks is that
124 scripts are the glyphs used by a language or a group of languages, while
125 the blocks are more artificial groupings of (mostly) 256 characters based
126 on the Unicode numbering.
128 In general, scripts are more inclusive, but not universally so. For
129 example, while the script C<Latin> includes all the Latin characters and
130 their various diacritic-adorned versions, it does not include the various
131 punctuation or digits (since they are not solely C<Latin>).
133 A number of other properties are now supported, including C<\p{L&}>,
134 C<\p{Any}> C<\p{Assigned}>, C<\p{Unassigned}>, C<\p{Blank}> [561] and
135 C<\p{SpacePerl}> [561] (along with their C<\P{...}> versions, of course).
136 See L<perlunicode> for details, and more additions.
138 The C<In> or C<Is> prefix to names used with the C<\p{...}> and C<\P{...}>
139 are now almost always optional. The only exception is that a C<In> prefix
140 is required to signify a Unicode block when a block name conflicts with a
141 script name. For example, C<\p{Tibetan}> refers to the script, while
142 C<\p{InTibetan}> refers to the block. When there is no name conflict, you
143 can omit the C<In> from the block name (e.g. C<\p{BraillePatterns}>), but
144 to be safe, it's probably best to always use the C<In>).
146 =head2 REF(...) Instead Of SCALAR(...)
148 A reference to a reference now stringifies as "REF(0x81485ec)" instead
149 of "SCALAR(0x81485ec)" in order to be more consistent with the return
152 =head2 pack/unpack D/F recycled
154 The undocumented pack/unpack template letters D/F have been recycled
155 for better use: now they stand for long double (if supported by the
156 platform) and NV (Perl internal floating point type). (They used
157 to be aliases for d/f, but you never knew that.)
165 The semantics of bless(REF, REF) were unclear and until someone proves
166 it to make some sense, it is forbidden.
170 The obsolete chat2 library that should never have been allowed
171 to escape the laboratory has been decommissioned.
175 The builtin dump() function has probably outlived most of its
176 usefulness. The core-dumping functionality will remain in future
177 available as an explicit call to C<CORE::dump()>, but in future
178 releases the behaviour of an unqualified C<dump()> call may change.
182 The very dusty examples in the eg/ directory have been removed.
183 Suggestions for new shiny examples welcome but the main issue is that
184 the examples need to be documented, tested and (most importantly)
189 The (bogus) escape sequences \8 and \9 now give an optional warning
190 ("Unrecognized escape passed through"). There is no need to \-escape
195 The list of filenames from glob() (or <...>) is now by default sorted
196 alphabetically to be csh-compliant (which is what happened before
197 in most UNIX platforms). (bsd_glob() does still sort platform
198 natively, ASCII or EBCDIC, unless GLOB_ALPHASORT is specified.) [561]
202 Spurious syntax errors generated in certain situations, when glob()
203 caused File::Glob to be loaded for the first time, have been fixed. [561]
207 Although "you shouldn't do that", it was possible to write code that
208 depends on Perl's hashed key order (Data::Dumper does this). The new
209 algorithm "One-at-a-Time" produces a different hashed key order.
210 More details are in L</"Performance Enhancements">.
214 lstat(FILEHANDLE) now gives a warning because the operation makes no sense.
215 In future releases this may become a fatal error.
219 The C<package;> syntax (C<package> without an argument) has been
220 deprecated. Its semantics were never that clear and its
221 implementation even less so. If you have used that feature to
222 disallow all but fully qualified variables, C<use strict;> instead.
226 The unimplemented POSIX regex features [[.cc.]] and [[=c=]] are still
227 recognised but now cause fatal errors. The previous behaviour of
228 ignoring them by default and warning if requested was unacceptable
229 since it, in a way, falsely promised that the features could be used.
233 The current user-visible implementation of pseudo-hashes (the weird
234 use of the first array element) is deprecated starting from Perl 5.8.0
235 and will be removed in Perl 5.10.0, and the feature will be
236 implemented differently. Not only is the current interface rather
237 ugly, but the current implementation slows down normal array and hash
238 use quite noticeably. The C<fields> pragma interface will remain
239 available. The I<restricted hashes> interface is expected to
240 be the replacement interface (see L<Hash::Util>).
244 The syntaxes C<< @a->[...] >> and C<< %h->{...} >> have now been deprecated.
248 After years of trying, suidperl is considered to be too complex to
249 ever be considered truly secure. The suidperl functionality is likely
250 to be removed in a future release.
254 The 5.005 threads model (module C<Thread>) is deprecated and expected
255 to be removed in Perl 5.10. Multithreaded code should be migrated to
256 the new ithreads model (see L<threads>, L<threads::shared> and
261 The long deprecated uppercase aliases for the string comparison
262 operators (EQ, NE, LT, LE, GE, GT) have now been removed.
266 The tr///C and tr///U features have been removed and will not return;
267 the interface was a mistake. Sorry about that. For similar
268 functionality, see pack('U0', ...) and pack('C0', ...). [561]
272 Earlier Perls treated "sub foo (@bar)" as equivalent to "sub foo (@)".
273 The prototypes are now checked better at compile-time for invalid
274 syntax. An optional warning is generated ("Illegal character in
275 prototype...") but this may be upgraded to a fatal error in a future
280 The existing behaviour when localising tied arrays and hashes is wrong,
281 and will be changed in a future release, so do not rely on the existing
282 behaviour. See L<"Localising Tied Arrays and Hashes Is Broken">.
286 =head1 Core Enhancements
288 =head2 PerlIO is Now The Default
294 IO is now by default done via PerlIO rather than system's "stdio".
295 PerlIO allows "layers" to be "pushed" onto a file handle to alter the
296 handle's behaviour. Layers can be specified at open time via 3-arg
299 open($fh,'>:crlf :utf8', $path) || ...
301 or on already opened handles via extended C<binmode>:
303 binmode($fh,':encoding(iso-8859-7)');
305 The built-in layers are: unix (low level read/write), stdio (as in
306 previous Perls), perlio (re-implementation of stdio buffering in a
307 portable manner), crlf (does CRLF <=> "\n" translation as on Win32,
308 but available on any platform). A mmap layer may be available if
309 platform supports it (mostly UNIXes).
311 Layers to be applied by default may be specified via the 'open' pragma.
313 See L</"Installation and Configuration Improvements"> for the effects
314 of PerlIO on your architecture name.
318 File handles can be marked as accepting Perl's internal encoding of Unicode
319 (UTF-8 or UTF-EBCDIC depending on platform) by a pseudo layer ":utf8" :
321 open($fh,">:utf8","Uni.txt");
323 Note for EBCDIC users: the pseudo layer ":utf8" is erroneously named
324 for you since it's not UTF-8 what you will be getting but instead
325 UTF-EBCDIC. See L<perlunicode>, L<utf8>, and
326 http://www.unicode.org/unicode/reports/tr16/ for more information.
327 In future releases this naming may change.
331 File handles can translate character encodings from/to Perl's internal
332 Unicode form on read/write via the ":encoding()" layer.
336 File handles can be opened to "in memory" files held in Perl scalars via:
338 open($fh,'>', \$variable) || ...
342 Anonymous temporary files are available without need to
343 'use FileHandle' or other module via
345 open($fh,"+>", undef) || ...
347 That is a literal undef, not an undefined value.
351 The list form of C<open> is now implemented for pipes (at least on UNIX):
353 open($fh,"-|", 'cat', '/etc/motd')
355 creates a pipe, and runs the equivalent of exec('cat', '/etc/motd') in
360 If your locale environment variables (LANGUAGE, LC_ALL, LC_CTYPE, LANG)
361 contain the strings 'UTF-8' or 'UTF8' (case-insensitive matching),
362 the default encoding of your STDIN, STDOUT, and STDERR, and of
363 B<any subsequent file open>, is UTF-8.
367 =head2 Restricted Hashes
369 A restricted hash is restricted to a certain set of keys, no keys
370 outside the set can be added. Also individual keys can be restricted
371 so that the key cannot be deleted and the value cannot be changed.
372 No new syntax is involved: the Hash::Util module is the interface.
376 Perl used to be fragile in that signals arriving at inopportune moments
377 could corrupt Perl's internal state. Now Perl postpones handling of
378 signals until it's safe (between opcodes).
380 This change may have surprising side effects because signals no longer
381 interrupt Perl instantly. Perl will now first finish whatever it was
382 doing, like finishing an internal operation (like sort()) or an
383 external operation (like an I/O operation), and only then look at any
384 arrived signals (and before starting the next operation). No more corrupt
385 internal state since the current operation is always finished first,
386 but the signal may take more time to get heard. Note that breaking
387 out from potentially blocking operations should still work, though.
389 =head2 Unicode Overhaul
391 Unicode in general should be now much more usable than in Perl 5.6.0
392 (or even in 5.6.1). Unicode can be used in hash keys, Unicode in
393 regular expressions should work now, Unicode in tr/// should work now,
394 Unicode in I/O should work now. See L<perluniintro> for introduction
395 and L<perlunicode> for details.
401 The Unicode Character Database coming with Perl has been upgraded
402 to Unicode 3.2.0. For more information, see http://www.unicode.org/ .
407 For developers interested in enhancing Perl's Unicode capabilities:
408 almost all the UCD files are included with the Perl distribution in
409 the F<lib/unicore> subdirectory. The most notable omission, for space
410 considerations, is the Unihan database.
414 The properties \p{Blank} and \p{SpacePerl} have been added. "Blank" is like
415 C isblank(), that is, it contains only "horizontal whitespace" (the space
416 character is, the newline isn't), and the "SpacePerl" is the Unicode
417 equivalent of C<\s> (\p{Space} isn't, since that includes the vertical
418 tabulator character, whereas C<\s> doesn't.)
420 See "New Unicode Properties" earlier in this document for additional
421 information on changes with Unicode properties.
425 =head2 Understanding of Numbers
427 In general a lot of fixing has happened in the area of Perl's
428 understanding of numbers, both integer and floating point. Since in
429 many systems the standard number parsing functions like C<strtoul()>
430 and C<atof()> seem to have bugs, Perl tries to work around their
431 deficiencies. This results hopefully in more accurate numbers.
433 Perl now tries internally to use integer values in numeric conversions
434 and basic arithmetics (+ - * /) if the arguments are integers, and
435 tries also to keep the results stored internally as integers.
436 This change leads to often slightly faster and always less lossy
437 arithmetics. (Previously Perl always preferred floating point numbers
440 =head2 Arrays now always interpolate into double-quoted strings [561]
442 In double-quoted strings, arrays now interpolate, no matter what. The
443 behavior in earlier versions of perl 5 was that arrays would interpolate
444 into strings if the array had been mentioned before the string was
445 compiled, and otherwise Perl would raise a fatal compile-time error.
446 In versions 5.000 through 5.003, the error was
448 Literal @example now requires backslash
450 In versions 5.004_01 through 5.6.0, the error was
452 In string, @example now must be written as \@example
454 The idea here was to get people into the habit of writing
455 C<"fred\@example.com"> when they wanted a literal C<@> sign, just as
456 they have always written C<"Give me back my \$5"> when they wanted a
459 Starting with 5.6.1, when Perl now sees an C<@> sign in a
460 double-quoted string, it I<always> attempts to interpolate an array,
461 regardless of whether or not the array has been used or declared
462 already. The fatal error has been downgraded to an optional warning:
464 Possible unintended interpolation of @example in string
466 This warns you that C<"fred@example.com"> is going to turn into
467 C<fred.com> if you don't backslash the C<@>.
468 See http://www.plover.com/~mjd/perl/at-error.html for more details
469 about the history here.
471 =head2 Miscellaneous Changes
477 AUTOLOAD is now lvaluable, meaning that you can add the :lvalue attribute
478 to AUTOLOAD subroutines and you can assign to the AUTOLOAD return value.
482 The $Config{byteorder} (and corresponding BYTEORDER in config.h) was
483 previously wrong in platforms if sizeof(long) was 4, but sizeof(IV)
484 was 8. The byteorder was only sizeof(long) bytes long (1234 or 4321),
485 but now it is correctly sizeof(IV) bytes long, (12345678 or 87654321).
486 (This problem didn't affect Windows platforms.)
488 Also, $Config{byteorder} is now computed dynamically--this is more
489 robust with "fat binaries" where an executable image contains binaries
490 for more than one binary platform, and when cross-compiling.
494 C<perl -d:Module=arg,arg,arg> now works (previously one couldn't pass
495 in multiple arguments.)
499 The builtin dump() now gives an optional warning
500 C<dump() better written as CORE::dump()>,
501 meaning that by default C<dump(...)> is resolved as the builtin
502 dump() which dumps core and aborts, not as (possibly) user-defined
503 C<sub dump>. To call the latter, qualify the call as C<&dump(...)>.
504 (The whole dump() feature is to considered deprecated, and possibly
505 removed/changed in future releases.)
509 chomp() and chop() are now overridable. Note, however, that their
510 prototype (as given by C<prototype("CORE::chomp")> is undefined,
511 because it cannot be expressed and therefore one cannot really write
512 replacements to override these builtins.
516 END blocks are now run even if you exit/die in a BEGIN block.
517 Internally, the execution of END blocks is now controlled by
518 PL_exit_flags & PERL_EXIT_DESTRUCT_END. This enables the new
519 behaviour for Perl embedders. This will default in 5.10. See
524 Formats now support zero-padded decimal fields.
528 Lvalue subroutines can now return C<undef> in list context. However,
529 the lvalue subroutine feature still remains experimental.
533 A lost warning "Can't declare ... dereference in my" has been
534 restored (Perl had it earlier but it became lost in later releases.)
538 A new special regular expression variable has been introduced:
539 C<$^N>, which contains the most-recently closed group (submatch).
543 C<no Module;> does not produce an error even if Module does not have an
544 unimport() method. This parallels the behavior of C<use> vis-a-vis
549 The numerical comparison operators return C<undef> if either operand
550 is a NaN. Previously the behaviour was unspecified.
554 The following builtin functions are now overridable: each(), keys(),
555 pop(), push(), shift(), splice(), unshift(). [561]
559 C<pack() / unpack()> can now group template letters with C<()> and then
560 apply repetition/count modifiers on the groups.
564 C<pack() / unpack()> can now process the Perl internal numeric types:
565 IVs, UVs, NVs-- and also long doubles, if supported by the platform.
566 The template letters are C<j>, C<J>, C<F>, and C<D>.
570 C<pack('U0a*', ...)> can now be used to force a string to UTF8.
574 my __PACKAGE__ $obj now works. [561]
578 POSIX::sleep() now returns the number of I<unslept> seconds
579 (as the POSIX standard says), as opposed to CORE::sleep() which
580 returns the number of slept seconds.
584 The printf() and sprintf() now support parameter reordering using the
585 C<%\d+\$> and C<*\d+\$> syntaxes. For example
587 print "%2\$s %1\$s\n", "foo", "bar";
589 will print "bar foo\n". This feature helps in writing
590 internationalised software, and in general when the order
591 of the parameters can vary.
595 The (\&) prototype now works properly. [561]
599 prototype(\[$@%&]) is now available to implicitly create references
600 (useful for example if you want to emulate the tie() interface).
604 A new command-line option, C<-t> is available. It is the
605 little brother of C<-T>: instead of dying on taint violations,
606 lexical warnings are given. B<This is only meant as a temporary
607 debugging aid while securing the code of old legacy applications.
608 This is not a substitute for -T.>
612 In other taint news, the C<exec LIST> and C<system LIST> have now been
613 considered too risky (think C<exec @ARGV>: it can start any program
614 with any arguments), and now the said forms cause a warning.
615 You should carefully launder the arguments to guarantee their
616 validity. In future releases of Perl the forms will become fatal
617 errors so consider starting laundering now.
621 Tied hash interfaces are now required to have the EXISTS and DELETE
622 methods (either own or inherited).
626 If tr/// is just counting characters, it doesn't attempt to
631 untie() will now call an UNTIE() hook if it exists. See L<perltie>
636 L<utime> now supports C<utime undef, undef, @files> to change the
637 file timestamps to the current time.
641 The rules for allowing underscores (underbars) in numeric constants
642 have been relaxed and simplified: now you can have an underscore
643 simply B<between digits>.
647 Rather than relying on C's argv[0] (which may not contain a full pathname)
648 where possible $^X is now set by asking the operating system.
649 (eg by reading F</proc/self/exe> on Linux, F</proc/curproc/file> on FreeBSD)
653 A new variable, C<${^TAINT}>, indicates whether taint mode is enabled.
657 You can now override the readline() builtin, and this overrides also
658 the <FILEHANDLE> angle bracket operator.
662 The command-line options -s and -F are now recognized on the shebang
667 Use of the C</c> match modifier without an accompanying C</g> modifier
668 elicits a new warning: C<Use of /c modifier is meaningless without /g>.
670 Use of C</c> in substitutions, even with C</g>, elicits
671 C<Use of /c modifier is meaningless in s///>.
673 Use of C</g> with C<split> elicits C<Use of /g modifier is meaningless
678 =head1 Modules and Pragmata
680 =head2 New Modules and Pragmata
686 C<Attribute::Handlers> allows a class to define attribute handlers.
689 use Attribute::Handlers;
690 sub Wolf :ATTR(SCALAR) { print "howl!\n" }
692 # later, in some package using or inheriting from MyPack...
694 my MyPack $Fluffy : Wolf; # the attribute handler Wolf will be called
696 Both variables and routines can have attribute handlers. Handlers can
697 be specific to type (SCALAR, ARRAY, HASH, or CODE), or specific to the
698 exact compilation phase (BEGIN, CHECK, INIT, or END).
699 See L<Attribute::Handlers>.
703 C<B::Concise>, by Stephen McCamant, is a new compiler backend for
704 walking the Perl syntax tree, printing concise info about ops.
705 The output is highly customisable. See L<B::Concise>.
709 The new bignum, bigint, and bigrat pragmas, by Tels, implement
710 transparent bignum support (using the Math::BigInt, Math::BigFloat,
711 and Math::BigRat backends).
715 C<Class::ISA>, by Sean Burke, is a module for reporting the search
716 path for a class's ISA tree. See L<Class::ISA>.
720 C<Cwd> now has a split personality: if possible, an XS extension is
721 used, (this will hopefully be faster, more secure, and more robust)
722 but if not possible, the familiar Perl implementation is used.
726 C<Devel::PPPort>, originally by Kenneth Albanowski and now
727 maintained by Paul Marquess, has been added. It is primarily used
728 by C<h2xs> to enhance portability of XS modules between different
729 versions of Perl. See L<Devel::PPPort>.
733 C<Digest>, frontend module for calculating digests (checksums), from
734 Gisle Aas, has been added. See L<Digest>.
738 C<Digest::MD5> for calculating MD5 digests (checksums) as defined in
739 RFC 1321, from Gisle Aas, has been added. See L<Digest::MD5>.
741 use Digest::MD5 'md5_hex';
743 $digest = md5_hex("Thirsty Camel");
745 print $digest, "\n"; # 01d19d9d2045e005c3f1b80e8b164de1
747 NOTE: the C<MD5> backward compatibility module is deliberately not
748 included since its further use is discouraged.
752 C<Encode>, originally by Nick Ing-Simmons and now maintained by Dan
753 Kogai, provides a mechanism to translate between different character
754 encodings. Support for Unicode, ISO-8859-1, and ASCII are compiled in
755 to the module. Several other encodings (like the rest of the
756 ISO-8859, CP*/Win*, Mac, KOI8-R, three variants EBCDIC, Chinese,
757 Japanese, and Korean encodings) are included and can be loaded at
758 runtime. (For space considerations, the largest Chinese encodings
759 have been separated into their own CPAN module, Encode::HanExtra,
760 which Encode will use if available). See L<Encode>.
762 Any encoding supported by Encode module is also available to the
763 ":encoding()" layer if PerlIO is used.
767 C<Hash::Util> is the interface to the new I<restricted hashes>
768 feature. (Implemented by Jeffrey Friedl, Nick Ing-Simmons, and
769 Michael Schwern.) See L<Hash::Util>.
773 C<I18N::Langinfo> can be used to query locale information.
774 See L<I18N::Langinfo>.
778 C<I18N::LangTags>, by Sean Burke, has functions for dealing with
779 RFC3066-style language tags. See L<I18N::LangTags>.
783 C<ExtUtils::Constant>, by Nicholas Clark, is a new tool for extension
784 writers for generating XS code to import C header constants.
785 See L<ExtUtils::Constant>.
789 C<Filter::Simple>, by Damian Conway, is an easy-to-use frontend to
790 Filter::Util::Call. See L<Filter::Simple>.
796 use Filter::Simple sub {
797 while (my ($from, $to) = splice @_, 0, 2) {
806 use MyFilter qr/red/ => 'green';
808 print "red\n"; # this code is filtered, will print "green\n"
809 print "bored\n"; # this code is filtered, will print "bogreen\n"
813 print "red\n"; # this code is not filtered, will print "red\n"
817 C<File::Temp>, by Tim Jenness, allows one to create temporary files
818 and directories in an easy, portable, and secure way. See
823 C<Filter::Util::Call>, by Paul Marquess, provides you with the
824 framework to write I<source filters> in Perl. For most uses, the
825 frontend Filter::Simple is to be preferred. See L<Filter::Util::Call>.
829 C<if>, by Ilya Zakharevich, is a new pragma for conditional inclusion
834 L<libnet>, by Graham Barr, is a collection of perl5 modules related
835 to network programming. See L<Net::FTP>, L<Net::NNTP>, L<Net::Ping>
836 (not part of libnet, but related), L<Net::POP3>, L<Net::SMTP>,
839 Perl installation leaves libnet unconfigured; use F<libnetcfg>
844 C<List::Util>, by Graham Barr, is a selection of general-utility
845 list subroutines, such as sum(), min(), first(), and shuffle().
850 C<Locale::Constants>, C<Locale::Country>, C<Locale::Currency>
851 C<Locale::Language>, and L<Locale::Script>, by Neil Bowers, have
852 been added. They provide the codes for various locale standards, such
853 as "fr" for France, "usd" for US Dollar, and "ja" for Japanese.
857 $country = code2country('jp'); # $country gets 'Japan'
858 $code = country2code('Norway'); # $code gets 'no'
860 See L<Locale::Constants>, L<Locale::Country>, L<Locale::Currency>,
861 and L<Locale::Language>.
865 C<Locale::Maketext>, by Sean Burke, is a localization framework. See
866 L<Locale::Maketext>, and L<Locale::Maketext::TPJ13>. The latter is an
867 article about software localization, originally published in The Perl
868 Journal #13, and republished here with kind permission.
872 C<Math::BigRat> for big rational numbers, to accompany Math::BigInt and
873 Math::BigFloat, from Tels. See L<Math::BigRat>.
877 C<Memoize> can make your functions faster by trading space for time,
878 from Mark-Jason Dominus. See L<Memoize>.
882 C<MIME::Base64>, by Gisle Aas, allows you to encode data in base64,
883 as defined in RFC 2045 - I<MIME (Multipurpose Internet Mail
888 $encoded = encode_base64('Aladdin:open sesame');
889 $decoded = decode_base64($encoded);
891 print $encoded, "\n"; # "QWxhZGRpbjpvcGVuIHNlc2FtZQ=="
897 C<MIME::QuotedPrint>, by Gisle Aas, allows you to encode data
898 in quoted-printable encoding, as defined in RFC 2045 - I<MIME
899 (Multipurpose Internet Mail Extensions)>.
901 use MIME::QuotedPrint;
903 $encoded = encode_qp("Smiley in Unicode: \x{263a}");
904 $decoded = decode_qp($encoded);
906 print $encoded, "\n"; # "Smiley in Unicode: =263A"
908 MIME::QuotedPrint has been enhanced to provide the basic methods
909 necessary to use it with PerlIO::Via as in :
911 use MIME::QuotedPrint;
912 open($fh,">Via(MIME::QuotedPrint)",$path);
914 See L<MIME::QuotedPrint>.
918 C<NEXT>, by Damian Conway, is a pseudo-class for method redispatch.
923 C<open> is a new pragma for setting the default I/O disciplines
928 C<PerlIO::Scalar>, by Nick Ing-Simmons, provides the implementation
929 of IO to "in memory" Perl scalars as discussed above. It also serves
930 as an example of a loadable PerlIO layer. Other future possibilities
931 include PerlIO::Array and PerlIO::Code. See L<PerlIO::Scalar>.
935 C<PerlIO::Via>, by Nick Ing-Simmons, acts as a PerlIO layer and wraps
936 PerlIO layer functionality provided by a class (typically implemented
939 use MIME::QuotedPrint;
940 open($fh,">Via(MIME::QuotedPrint)",$path);
942 This will automatically convert everything output to C<$fh>
943 to Quoted-Printable. See L<PerlIO::Via>.
947 C<Pod::ParseLink>, by Russ Allbery, has been added,
948 to parse LZ<><> links in pods as described in the new
953 C<Pod::Text::Overstrike>, by Joe Smith, has been added.
954 It converts POD data to formatted overstrike text.
955 See L<Pod::Text::Overstrike>.
959 C<Scalar::Util> is a selection of general-utility scalar subroutines,
960 such as blessed(), reftype(), and tainted(). See L<Scalar::Util>.
964 C<sort> is a new pragma for controlling the behaviour of sort().
968 C<Storable> gives persistence to Perl data structures by allowing the
969 storage and retrieval of Perl data to and from files in a fast and
970 compact binary format. Because in effect Storable does serialisation
971 of Perl data structues, with it you can also clone deep, hierarchical
972 datastructures. Storable was originally created by Raphael Manfredi,
973 but it is now maintained by Abhijit Menon-Sen. Storable has been
974 enhanced to understand the two new hash features, Unicode keys and
975 restricted hashes. See L<Storable>.
979 C<Switch>, by Damian Conway, has been added. Just by saying
983 you have C<switch> and C<case> available in Perl.
989 case 1 { print "number 1" }
990 case "a" { print "string a" }
991 case [1..10,42] { print "number in list" }
992 case (@array) { print "number in list" }
993 case /\w+/ { print "pattern" }
994 case qr/\w+/ { print "pattern" }
995 case (%hash) { print "entry in hash" }
996 case (\%hash) { print "entry in hash" }
997 case (\&sub) { print "arg to subroutine" }
998 else { print "previous case not true" }
1005 C<Test::More>, by Michael Schwern, is yet another framework for writing
1006 test scripts, more extensive than Test::Simple. See L<Test::More>.
1010 C<Test::Simple>, by Michael Schwern, has basic utilities for writing
1011 tests. See L<Test::Simple>.
1015 C<Text::Balanced>, by Damian Conway, has been added, for extracting
1016 delimited text sequences from strings.
1018 use Text::Balanced 'extract_delimited';
1020 ($a, $b) = extract_delimited("'never say never', he never said", "'", '');
1022 $a will be "'never say never'", $b will be ', he never said'.
1024 In addition to extract_delimited(), there are also extract_bracketed(),
1025 extract_quotelike(), extract_codeblock(), extract_variable(),
1026 extract_tagged(), extract_multiple(), gen_delimited_pat(), and
1027 gen_extract_tagged(). With these, you can implement rather advanced
1028 parsing algorithms. See L<Text::Balanced>.
1032 C<threads>, by Arthur Bergman, is an interface to interpreter threads.
1033 Interpreter threads (ithreads) is the new thread model introduced in
1034 Perl 5.6 but only available as an internal interface for extension
1035 writers (and for Win32 Perl for C<fork()> emulation). See L<threads>,
1036 L<threads::shared>, and L<perlthrtut>.
1040 C<threads::shared>, by Arthur Bergman, allows data sharing for
1041 interpreter threads. In the ithreads model any data sharing between
1042 threads must be explicit, as opposed to the old 5.005 thread model
1043 where data sharing was implicit. See L<threads::shared>.
1047 C<Tie::File>, by Mark-Jason Dominus, associates a Perl array with the
1048 lines of a file. See L<Tie::File>.
1052 C<Tie::Memoize>, by Ilya Zakharevich, provides on-demand loaded hashes.
1053 See L<Tie::Memoize>.
1057 C<Tie::RefHash::Nestable>, by Edward Avis, allows storing hash
1058 references (unlike the standard Tie::RefHash) The module is contained
1059 within Tie::RefHash. See L<Tie::RefHash>.
1063 C<Time::HiRes>, by Douglas E. Wegscheid, provides high resolution
1064 timing (ualarm, usleep, and gettimeofday). See L<Time::HiRes>.
1068 C<Unicode::UCD> offers a querying interface to the Unicode Character
1069 Database. See L<Unicode::UCD>.
1073 C<Unicode::Collate>, by SADAHIRO Tomoyuki, implements the UCA
1074 (Unicode Collation Algorithm) for sorting Unicode strings.
1075 See L<Unicode::Collate>.
1079 C<Unicode::Normalize>, by SADAHIRO Tomoyuki, implements the various
1080 Unicode normalization forms. See L<Unicode::Normalize>.
1084 C<XS::Typemap>, by Tim Jenness, is a test extension that exercises
1085 XS typemaps. Nothing gets installed, but the code is worth studying
1086 for extension writers.
1090 =head2 Updated And Improved Modules and Pragmata
1096 The following independently supported modules have been updated to the
1097 newest versions from CPAN: CGI, CPAN, DB_File, File::Spec, File::Temp,
1098 Getopt::Long, Math::BigFloat, Math::BigInt, the podlators bundle
1099 (Pod::Man, Pod::Text), Pod::LaTeX, Pod::Parser, Storable,
1100 Term::ANSIColor, Test, Text-Tabs+Wrap.
1104 attributes::reftype() now works on tied arguments.
1108 AutoLoader can now be disabled with C<no AutoLoader;>.
1112 B::Deparse has been significantly enhanced by Robin Houston. It can
1113 now deparse almost all of the standard test suite (so that the tests
1114 still succeed). There is a make target "test.deparse" for trying this
1119 Carp now has better interface documentation, and the @CARP_NOT
1120 interface has been added to get optional control over where errors
1121 are reported independently of @ISA, by Ben Tilly.
1125 Class::Struct can now define the classes in compile time.
1129 Class::Struct now assigns the array/hash element if the accessor
1130 is called with an array/hash element as the B<sole> argument.
1134 The return value of Cwd::fastcwd() is now tainted.
1138 Data::Dumper now has an option to sort hashes.
1142 Data::Dumper now has an option to dump code references
1147 DB_File now supports newer Berkeley DB versions, among
1152 Devel::Peek now has an interface for the Perl memory statistics
1153 (this works only if you are using perl's malloc, and if you have
1154 compiled with debugging).
1158 The English module can now be used without the infamous performance
1161 use English '-no_match_vars';
1163 (Assuming, of course, that you don't need the troublesome variables
1164 C<$`>, C<$&>, or C<$'>.) Also, introduced C<@LAST_MATCH_START> and
1165 C<@LAST_MATCH_END> English aliases for C<@-> and C<@+>.
1169 ExtUtils::MakeMaker now uses File::Spec internally, which hopefully
1170 leads to better portability.
1174 Fcntl, Socket, and Sys::Syslog have been rewritten by Nicholas Clark
1175 to use the new-style constant dispatch section (see L<ExtUtils::Constant>).
1176 This means that they will be more robust and hopefully faster.
1180 File::Find now chdir()s correctly when chasing symbolic links. [561]
1184 File::Find now has pre- and post-processing callbacks. It also
1185 correctly changes directories when chasing symbolic links. Callbacks
1186 (naughtily) exiting with "next;" instead of "return;" now work.
1190 File::Find is now (again) reentrant. It also has been made
1195 The warnings issued by File::Find now belong to their own category.
1196 You can enable/disable them with C<use/no warnings 'File::Find';>.
1200 File::Glob::glob() has been renamed to File::Glob::bsd_glob()
1201 because the name clashes with the builtin glob(). The older
1202 name is still available for compatibility, but is deprecated. [561]
1206 File::Glob now supports C<GLOB_LIMIT> constant to limit the size of
1207 the returned list of filenames.
1211 IPC::Open3 now allows the use of numeric file descriptors.
1215 IO::Socket now has an atmark() method, which returns true if the socket
1216 is positioned at the out-of-band mark. The method is also exportable
1217 as a sockatmark() function.
1221 IO::Socket::INET failed to open the specified port if the service name
1222 was not known. It now correctly uses the supplied port number as is. [561]
1226 IO::Socket::INET has support for the ReusePort option (if your
1227 platform supports it). The Reuse option now has an alias, ReuseAddr.
1228 For clarity, you may want to prefer ReuseAddr.
1232 IO::Socket::INET now supports a value of zero for C<LocalPort>
1233 (usually meaning that the operating system will make one up.)
1237 'use lib' now works identically to @INC. Removing directories
1238 with 'no lib' now works.
1242 Math::BigFloat and Math::BigInt have undergone a full rewrite by Tels.
1243 They are now magnitudes faster, and they support various bignum
1244 libraries such as GMP and PARI as their backends.
1248 Math::Complex handles inf, NaN etc., better.
1252 Net::Ping has been considerably enhanced by Rob Brown: multihoming is
1253 now supported, Win32 functionality is better, there is now time
1254 measuring functionality (optionally high-resolution using
1255 Time::HiRes), and there is now "external" protocol which uses
1256 Net::Ping::External module which runs your external ping utility and
1257 parses the output. A version of Net::Ping::External is available in
1260 Note that some of the Net::Ping tests are disabled when running
1261 under the Perl distribution since one cannot assume one or more
1262 of the following: enabled echo port at localhost, full Internet
1263 connectivity, or sympathetic firewalls. You can set the environment
1264 variable PERL_TEST_Net_Ping to "1" (one) before running the Perl test
1265 suite to enable all the Net::Ping tests.
1269 POSIX::sigaction() is now much more flexible and robust.
1270 You can now install coderef handlers, 'DEFAULT', and 'IGNORE'
1271 handlers, installing new handlers was not atomic.
1275 In Safe, C<%INC> is now localised in a Safe compartment so that
1280 In SDBM_File on dosish platforms, some keys went missing because of
1281 lack of support for files with "holes". A workaround for the problem
1286 In Search::Dict one can now have a pre-processing hook for the
1287 lines being searched.
1291 The Shell module now has an OO interface.
1295 In Sys::Syslog there is now a failover mechanism that will go
1296 through alternative connection mechanisms until the message
1297 is successfully logged.
1301 The Test module has been significantly enhanced.
1305 Time::Local::timelocal() does not handle fractional seconds anymore.
1306 The rationale is that neither does localtime(), and timelocal() and
1307 localtime() are supposed to be inverses of each other.
1311 The vars pragma now supports declaring fully qualified variables.
1312 (Something that C<our()> does not and will not support.)
1316 The C<utf8::> name space (as in the pragma) provides various
1317 Perl-callable functions to provide low level access to Perl's
1318 internal Unicode representation. At the moment only length()
1319 has been implemented.
1323 =head1 Utility Changes
1329 Emacs perl mode (emacs/cperl-mode.el) has been updated to version
1334 F<emacs/e2ctags.pl> is now much faster.
1338 C<enc2xs> is a tool for people adding their own encodings to the
1343 C<h2ph> now supports C trigraphs.
1347 C<h2xs> now produces a template README.
1351 C<h2xs> now uses C<Devel::PPPort> for better portability between
1352 different versions of Perl.
1356 C<h2xs> uses the new L<ExtUtils::Constant|ExtUtils::Constant> module
1357 which will affect newly created extensions that define constants.
1358 Since the new code is more correct (if you have two constants where the
1359 first one is a prefix of the second one, the first constant B<never>
1360 got defined), less lossy (it uses integers for integer constant,
1361 as opposed to the old code that used floating point numbers even for
1362 integer constants), and slightly faster, you might want to consider
1363 regenerating your extension code (the new scheme makes regenerating
1364 easy). L<h2xs> now also supports C trigraphs.
1368 C<libnetcfg> has been added to configure libnet.
1372 C<perlbug> is now much more robust. It also sends the bug report to
1373 perl.org, not perl.com.
1377 C<perlcc> has been rewritten and its user interface (that is,
1378 command line) is much more like that of the UNIX C compiler, cc.
1379 (The perlbc tools has been removed. Use C<perlcc -B> instead.)
1380 B<Note that perlcc is still considered very experimental and
1385 C<perlivp> is a new Installation Verification Procedure utility
1386 for running any time after installing Perl.
1390 C<piconv> is an implementation of the character conversion utility
1391 C<iconv>, demonstrating the new Encode module.
1395 C<pod2html> now allows specifying a cache directory.
1399 C<pod2html> now produces XHTML 1.0.
1403 C<pod2html> now understands POD written using different line endings
1404 (PC-like CRLF versus UNIX-like LF versus MacClassic-like CR).
1408 C<s2p> has been completely rewritten in Perl. (It is in fact a full
1409 implementation of sed in Perl: you can use the sed functionality by
1410 using the C<psed> utility.)
1414 C<xsubpp> now understands POD documentation embedded in the *.xs
1419 C<xsubpp> now supports the OUT keyword.
1423 =head1 New Documentation
1429 perl56delta details the changes between the 5.005 release and the
1434 perlclib documents the internal replacements for standard C library
1435 functions. (Interesting only for extension writers and Perl core
1440 perldebtut is a Perl debugging tutorial.
1444 perlebcdic contains considerations for running Perl on EBCDIC
1449 perlintro is a gentle introduction to Perl.
1453 perliol documents the internals of PerlIO with layers.
1457 perlmodstyle is a style guide for writing modules.
1461 perlnewmod tells about writing and submitting a new module.
1465 perlpacktut is a pack() tutorial.
1469 perlpod has been rewritten to be clearer and to record the best
1470 practices gathered over the years.
1474 perlpodspec is a more formal specification of the pod format,
1475 mainly of interest for writers of pod applications, not to
1476 people writing in pod.
1480 perlretut is a regular expression tutorial.
1484 perlrequick is a regular expressions quick-start guide.
1485 Yes, much quicker than perlretut. [561]
1489 perltodo has been updated.
1493 perltootc has been renamed as perltooc (to not to conflict
1494 with perltoot in filesystems restricted to "8.3" names).
1498 perluniintro is an introduction to using Unicode in Perl.
1499 (perlunicode is more of a detailed reference and background
1504 perlutil explains the command line utilities packaged with the Perl
1509 The following platform-specific documents are available before
1510 the installation as README.I<platform>, and after the installation
1513 perlaix perlamiga perlapollo perlbeos perlbs2000
1514 perlce perlcygwin perldgux perldos perlepoc perlhpux
1515 perlhurd perlmachten perlmacos perlmint perlmpeix
1516 perlnetware perlos2 perlos390 perlplan9 perlqnx perlsolaris
1517 perltru64 perluts perlvmesa perlvms perlvos perlwin32
1519 Eastern Asian Perl users are now welcomed in their own languages:
1520 README.jp (Japanese), README.ko (Korean), README.cn (simplified
1521 Chinese) and README.tw (traditional Chinese), which are written in
1522 normal pod but encoded in EUC-JP, EUC-KR, EUC-CN and Big5. These
1523 will get installed as
1525 perljp perlko perlcn perltw
1531 The documentation for the POSIX-BC platform is called "BS2000", to avoid
1532 confusion with the Perl POSIX module.
1536 The documentation for the WinCE platform is called perlce (README.ce
1537 in the source code kit), to avoid confusion with the perlwin32
1538 documentation on 8.3-restricted filesystems.
1542 =head1 Performance Enhancements
1548 map() could get pathologically slow when the result list it generates
1549 is larger than the source list. The performance has been improved for
1550 common scenarios. [561]
1554 sort() is also fully reentrant, in the sense that the sort function
1555 can itself call sort(). This did not work reliably in previous
1560 sort() has been changed to use primarily mergesort internally as
1561 opposed to the earlier quicksort. For very small lists this may
1562 result in slightly slower sorting times, but in general the speedup
1563 should be at least 20%. Additional bonuses are that the worst case
1564 behaviour of sort() is now better (in computer science terms it now
1565 runs in time O(N log N), as opposed to quicksort's Theta(N**2)
1566 worst-case run time behaviour), and that sort() is now stable
1567 (meaning that elements with identical keys will stay ordered as they
1568 were before the sort). See the C<sort> pragma for information.
1570 The story in more detail: suppose you want to serve yourself a little
1573 @digits = ( 3,1,4,1,5,9 );
1575 A numerical sort of the digits will yield (1,1,3,4,5,9), as expected.
1576 Which C<1> comes first is hard to know, since one C<1> looks pretty
1577 much like any other. You can regard this as totally trivial,
1578 or somewhat profound. However, if you just want to sort the even
1579 digits ahead of the odd ones, then what will
1581 sort { ($a % 2) <=> ($b % 2) } @digits;
1583 yield? The only even digit, C<4>, will come first. But how about
1584 the odd numbers, which all compare equal? With the quicksort algorithm
1585 used to implement Perl 5.6 and earlier, the order of ties is left up
1586 to the sort. So, as you add more and more digits of Pi, the order
1587 in which the sorted even and odd digits appear will change.
1588 and, for sufficiently large slices of Pi, the quicksort algorithm
1589 in Perl 5.8 won't return the same results even if reinvoked with the
1590 same input. The justification for this rests with quicksort's
1591 worst case behavior. If you run
1593 sort { $a <=> $b } ( 1 .. $N , 1 .. $N );
1595 (something you might approximate if you wanted to merge two sorted
1596 arrays using sort), doubling $N doesn't just double the quicksort time,
1597 it I<quadruples> it. Quicksort has a worst case run time that can
1598 grow like N**2, so-called I<quadratic> behaviour, and it can happen
1599 on patterns that may well arise in normal use. You won't notice this
1600 for small arrays, but you I<will> notice it with larger arrays,
1601 and you may not live long enough for the sort to complete on arrays
1602 of a million elements. So the 5.8 quicksort scrambles large arrays
1603 before sorting them, as a statistical defence against quadratic behaviour.
1604 But that means if you sort the same large array twice, ties may be
1605 broken in different ways.
1607 Because of the unpredictability of tie-breaking order, and the quadratic
1608 worst-case behaviour, quicksort was I<almost> replaced completely with
1609 a stable mergesort. I<Stable> means that ties are broken to preserve
1610 the original order of appearance in the input array. So
1612 sort { ($a % 2) <=> ($b % 2) } (3,1,4,1,5,9);
1614 will yield (4,3,1,1,5,9), guaranteed. The even and odd numbers
1615 appear in the output in the same order they appeared in the input.
1616 Mergesort has worst case O(N log N) behaviour, the best value
1617 attainable. And, ironically, this mergesort does particularly
1618 well where quicksort goes quadratic: mergesort sorts (1..$N, 1..$N)
1619 in O(N) time. But quicksort was rescued at the last moment because
1620 it is faster than mergesort on certain inputs and platforms.
1621 For example, if you really I<don't> care about the order of even
1622 and odd digits, quicksort will run in O(N) time; it's very good
1623 at sorting many repetitions of a small number of distinct elements.
1624 The quicksort divide and conquer strategy works well on platforms
1625 with relatively small, very fast, caches. Eventually, the problem gets
1626 whittled down to one that fits in the cache, from which point it
1627 benefits from the increased memory speed.
1629 Quicksort was rescued by implementing a sort pragma to control aspects
1630 of the sort. The B<stable> subpragma forces stable behaviour,
1631 regardless of algorithm. The B<_quicksort> and B<_mergesort>
1632 subpragmas are heavy-handed ways to select the underlying implementation.
1633 The leading C<_> is a reminder that these subpragmas may not survive
1634 beyond 5.8. More appropriate mechanisms for selecting the implementation
1635 exist, but they wouldn't have arrived in time to save quicksort.
1639 Hashes now use Bob Jenkins "One-at-a-Time" hashing key algorithm
1640 ( http://burtleburtle.net/bob/hash/doobs.html ). This algorithm is
1641 reasonably fast while producing a much better spread of values than
1642 the old hashing algorithm (originally by Chris Torek, later tweaked by
1643 Ilya Zakharevich). Hash values output from the algorithm on a hash of
1644 all 3-char printable ASCII keys comes much closer to passing the
1645 DIEHARD random number generation tests. According to perlbench, this
1646 change has not affected the overall speed of Perl.
1650 unshift() should now be noticeably faster.
1654 =head1 Installation and Configuration Improvements
1656 =head2 Generic Improvements
1662 INSTALL now explains how you can configure Perl to use 64-bit
1663 integers even on non-64-bit platforms.
1667 Policy.sh policy change: if you are reusing a Policy.sh file
1668 (see INSTALL) and you use Configure -Dprefix=/foo/bar and in the old
1669 Policy $prefix eq $siteprefix and $prefix eq $vendorprefix, all of
1670 them will now be changed to the new prefix, /foo/bar. (Previously
1671 only $prefix changed.) If you do not like this new behaviour,
1672 specify prefix, siteprefix, and vendorprefix explicitly.
1676 A new optional location for Perl libraries, otherlibdirs, is available.
1677 It can be used for example for vendor add-ons without disturbing Perl's
1678 own library directories.
1682 In many platforms, the vendor-supplied 'cc' is too stripped-down to
1683 build Perl (basically, 'cc' doesn't do ANSI C). If this seems
1684 to be the case and 'cc' does not seem to be the GNU C compiler
1685 'gcc', an automatic attempt is made to find and use 'gcc' instead.
1689 gcc needs to closely track the operating system release to avoid
1690 build problems. If Configure finds that gcc was built for a different
1691 operating system release than is running, it now gives a clearly visible
1692 warning that there may be trouble ahead.
1696 Since Perl 5.8 is not binary-compatible with previous releases
1697 of Perl, Configure no longer suggests including the 5.005
1702 Configure C<-S> can now run non-interactively. [561]
1706 Configure support for pdp11-style memory models has been removed due
1707 to obsolescence. [561]
1711 configure.gnu now works with options with whitespace in them.
1715 installperl now outputs everything to STDERR.
1719 Because PerlIO is now the default on most platforms, "-perlio" doesn't
1720 get appended to the $Config{archname} (also known as $^O) anymore.
1721 Instead, if you explicitly choose not to use perlio (Configure command
1722 line option -Uuseperlio), you will get "-stdio" appended.
1726 Another change related to the architecture name is that "-64all"
1727 (-Duse64bitall, or "maximally 64-bit") is appended only if your
1728 pointers are 64 bits wide. (To be exact, the use64bitall is ignored.)
1732 In AFS installations, one can configure the root of the AFS to be
1733 somewhere else than the default F</afs> by using the Configure
1734 parameter C<-Dafsroot=/some/where/else>.
1738 APPLLIB_EXP, a lesser-known configuration-time definition, has been
1739 documented. It can be used to prepend site-specific directories
1740 to Perl's default search path (@INC); see INSTALL for information.
1744 The version of Berkeley DB used when the Perl (and, presumably, the
1745 DB_File extension) was built is now available as
1746 C<@Config{qw(db_version_major db_version_minor db_version_patch)}>
1747 from Perl and as C<DB_VERSION_MAJOR_CFG DB_VERSION_MINOR_CFG
1748 DB_VERSION_PATCH_CFG> from C.
1752 Building Berkeley DB3 for compatibility modes for DB, NDBM, and ODBM
1753 has been documented in INSTALL.
1757 If you have CPAN access (either network or a local copy such as a
1758 CD-ROM) you can during specify extra modules to Configure to build and
1759 install with Perl using the -Dextras=... option. See INSTALL for
1764 In addition to config.over, a new override file, config.arch, is
1765 available. This file is supposed to be used by hints file writers
1766 for architecture-wide changes (as opposed to config.over which is
1767 for site-wide changes).
1771 If your file system supports symbolic links, you can build Perl outside
1772 of the source directory by
1774 mkdir /tmp/perl/build/directory
1775 cd /tmp/perl/build/directory
1776 sh /path/to/perl/source/Configure -Dmksymlinks ...
1778 This will create in /tmp/perl/build/directory a tree of symbolic links
1779 pointing to files in /path/to/perl/source. The original files are left
1780 unaffected. After Configure has finished, you can just say
1784 and Perl will be built and tested, all in /tmp/perl/build/directory.
1789 For Perl developers, several new make targets for profiling
1790 and debugging have been added; see L<perlhack>.
1796 Use of the F<gprof> tool to profile Perl has been documented in
1797 L<perlhack>. There is a make target called "perl.gprof" for
1798 generating a gprofiled Perl executable.
1802 If you have GCC 3, there is a make target called "perl.gcov" for
1803 creating a gcoved Perl executable for coverage analysis. See
1808 If you are on IRIX or Tru64 platforms, new profiling/debugging options
1809 have been added; see L<perlhack> for more information about pixie and
1816 Guidelines of how to construct minimal Perl installations have
1817 been added to INSTALL.
1821 The Thread extension is now not built at all under ithreads
1822 (C<Configure -Duseithreads>) because it wouldn't work anyway (the
1823 Thread extension requires being Configured with C<-Duse5005threads>).
1825 But note that the Thread.pm interface is now shared by both
1830 The Gconvert macro ($Config{d_Gconvert}) used by perl for stringifying
1831 floating-point numbers is now more picky about using sprintf %.*g
1832 rules for the conversion. Some platforms that used to use gcvt may
1833 now resort to the slower sprintf.
1837 The obsolete method of making a special (e.g., debugging) flavor
1840 make LIBPERL=libperld.a
1842 has been removed. Use -DDEBUGGING instead.
1846 =head2 New Or Improved Platforms
1848 For the list of platforms known to support Perl,
1849 see L<perlport/"Supported Platforms">.
1855 AIX dynamic loading should be now better supported.
1859 AIX should now work better with gcc, threads, and 64-bitness. Also the
1860 long doubles support in AIX should be better now. See L<perlaix>.
1864 AtheOS ( http://www.atheos.cx/ ) is a new platform.
1868 BeOS has been reclaimed.
1872 The DG/UX platform now supports 5.005-style threads.
1877 The DYNIX/ptx platform (a.k.a. dynixptx) is supported at or near
1882 EBCDIC platforms (z/OS (also known as OS/390), POSIX-BC, and VM/ESA)
1883 have been regained. Many test suite tests still fail and the
1884 co-existence of Unicode and EBCDIC isn't quite settled, but the
1885 situation is much better than with Perl 5.6. See L<perlos390>,
1886 L<perlbs2000> (for POSIX-BC), and L<perlvmesa> for more information.
1890 Building perl with -Duseithreads or -Duse5005threads now works under
1891 HP-UX 10.20 (previously it only worked under 10.30 or later). You will
1892 need a thread library package installed. See README.hpux. [561]
1896 Mac OS Classic is now supported in the mainstream source package
1897 (MacPerl has of course been available since perl 5.004 but now the
1898 source code bases of standard Perl and MacPerl have been synchronised)
1903 Mac OS X (or Darwin) should now be able to build Perl even on HFS+
1904 filesystems. (The case-insensitivity used to confuse the Perl build
1909 NCR MP-RAS is now supported. [561]
1913 All the NetBSD specific patches (except for the installation
1914 specific ones) have been merged back to the main distribution.
1918 NetWare from Novell is now supported. See L<perlnetware>.
1922 NonStop-UX is now supported. [561]
1926 NEC SUPER-UX is now supported.
1930 All the OpenBSD specific patches (except for the installation
1931 specific ones) have been merged back to the main distribution.
1935 Perl has been tested with the GNU pth userlevel thread package
1936 ( http://www.gnu.org/software/pth/pth.html ). All thread tests
1937 of Perl now work, but not without adding some yield()s to the tests,
1938 so while pth (and other userlevel thread implementations) can be
1939 considered to be "working" with Perl ithreads, keep in mind the
1940 possible non-preemptability of the underlying thread implementation.
1944 Stratus VOS is now supported using Perl's native build method
1945 (Configure). This is the recommended method to build Perl on
1946 VOS. The older methods, which build miniperl, are still
1947 available. See L<perlvos>.
1951 The Amdahl UTS UNIX mainframe platform is now supported. [561]
1955 WinCE is now supported. See L<perlce>.
1959 z/OS (formerly known as OS/390, formerly known as MVS OE) now has
1960 support for dynamic loading. This is not selected by default,
1961 however, you must specify -Dusedl in the arguments of Configure. [561]
1965 =head1 Selected Bug Fixes
1967 Numerous memory leaks and uninitialized memory accesses have been
1968 hunted down. Most importantly, anonymous subs used to leak quite
1975 The autouse pragma didn't work for Multi::Part::Function::Names.
1979 caller() could cause core dumps in certain situations. Carp was sometimes
1980 affected by this problem. In particular, caller() now returns a
1981 subroutine name of C<(unknown)> for subroutines that have been removed
1982 from the symbol table.
1986 chop(@list) in list context returned the characters chopped in
1987 reverse order. This has been reversed to be in the right order. [561]
1991 Configure no longer includes the DBM libraries (dbm, gdbm, db, ndbm)
1992 when building the Perl binary. The only exception to this is SunOS 4.x,
1993 which needs them. [561]
1997 The behaviour of non-decimal but numeric string constants such as
1998 "0x23" was platform-dependent: in some platforms that was seen as 35,
1999 in some as 0, in some as a floating point number (don't ask). This
2000 was caused by Perl's using the operating system libraries in a situation
2001 where the result of the string to number conversion is undefined: now
2002 Perl consistently handles such strings as zero in numeric contexts.
2006 The order of DESTROYs has been made more predictable.
2010 Several debugger fixes: exit code now reflects the script exit code,
2011 condition C<"0"> now treated correctly, the C<d> command now checks
2012 line number, C<$.> no longer gets corrupted, and all debugger output
2013 now goes correctly to the socket if RemotePort is set. [561]
2017 Perl 5.6.0 could emit spurious warnings about redefinition of
2018 dl_error() when statically building extensions into perl.
2019 This has been corrected. [561]
2023 L<dprofpp> -R didn't work.
2027 C<*foo{FORMAT}> now works.
2031 Infinity is now recognized as a number.
2035 UNIVERSAL::isa no longer caches methods incorrectly. (This broke
2036 the Tk extension with 5.6.0.) [561]
2040 Lexicals I: lexicals outside an eval "" weren't resolved
2041 correctly inside a subroutine definition inside the eval "" if they
2042 were not already referenced in the top level of the eval""ed code.
2046 Lexicals II: lexicals leaked at file scope into subroutines that
2047 were declared before the lexicals.
2051 Lexical warnings now propagating correctly between scopes
2052 and into C<eval "...">.
2056 C<use warnings qw(FATAL all)> did not work as intended. This has been
2061 warnings::enabled() now reports the state of $^W correctly if the caller
2062 isn't using lexical warnings. [561]
2066 Line renumbering with eval and C<#line> now works. [561]
2070 Fixed numerous memory leaks, especially in eval "".
2074 Localised tied variables no longer leak memory
2077 tie my %tied_hash => 'Tie::StdHash';
2081 # Used to leak memory every time local() was called;
2082 # in a loop, this added up.
2083 local($tied_hash{Foo}) = 1;
2087 Localised hash elements (and %ENV) are correctly unlocalised to not
2088 exist, if they didn't before they were localised.
2092 tie my %tied_hash => 'Tie::StdHash';
2096 # Nothing has set the FOO element so far
2098 { local $tied_hash{FOO} = 'Bar' }
2100 # This used to print, but not now.
2101 print "exists!\n" if exists $tied_hash{FOO};
2103 As a side effect of this fix, tied hash interfaces B<must> define
2104 the EXISTS and DELETE methods.
2108 mkdir() now ignores trailing slashes in the directory name,
2109 as mandated by POSIX.
2113 Some versions of glibc have a broken modfl(). This affects builds
2114 with C<-Duselongdouble>. This version of Perl detects this brokenness
2115 and has a workaround for it. The glibc release 2.2.2 is known to have
2116 fixed the modfl() bug.
2120 Modulus of unsigned numbers now works (4063328477 % 65535 used to
2121 return 27406, instead of 27047). [561]
2125 Some "not a number" warnings introduced in 5.6.0 eliminated to be
2126 more compatible with 5.005. Infinity is now recognised as a number. [561]
2130 Numeric conversions did not recognize changes in the string value
2131 properly in certain circumstances. [561]
2135 Attributes (such as :shared) didn't work with our().
2139 our() variables will not cause bogus "Variable will not stay shared"
2144 "our" variables of the same name declared in two sibling blocks
2145 resulted in bogus warnings about "redeclaration" of the variables.
2146 The problem has been corrected. [561]
2150 pack "Z" now correctly terminates the string with "\0".
2154 Fix password routines which in some shadow password platforms
2155 (e.g. HP-UX) caused getpwent() to return every other entry.
2159 The PERL5OPT environment variable (for passing command line arguments
2160 to Perl) didn't work for more than a single group of options. [561]
2164 PERL5OPT with embedded spaces didn't work.
2168 printf() no longer resets the numeric locale to "C".
2172 C<qw(a\\b)> now parses correctly as C<'a\\b'>: that is, as three
2173 characters, not four. [561]
2177 pos() did not return the correct value within s///ge in earlier
2178 versions. This is now handled correctly. [561]
2182 Printing quads (64-bit integers) with printf/sprintf now works
2183 without the q L ll prefixes (assuming you are on a quad-capable platform).
2187 Regular expressions on references and overloaded scalars now work.
2191 Right-hand side magic (GMAGIC) could in many cases such as string
2192 concatenation be invoked too many times.
2196 scalar() now forces scalar context even when used in void context.
2200 SOCKS support is now much more robust.
2204 sort() arguments are now compiled in the right wantarray context
2205 (they were accidentally using the context of the sort() itself).
2206 The comparison block is now run in scalar context, and the arguments
2207 to be sorted are always provided list context. [561]
2211 Changed the POSIX character class C<[[:space:]]> to include the (very
2212 rarely used) vertical tab character. Added a new POSIX-ish character
2213 class C<[[:blank:]]> which stands for horizontal whitespace
2214 (currently, the space and the tab).
2218 The tainting behaviour of sprintf() has been rationalized. It does
2219 not taint the result of floating point formats anymore, making the
2220 behaviour consistent with that of string interpolation. [561]
2224 Some cases of inconsistent taint propagation (such as within hash
2225 values) have been fixed.
2229 The RE engine found in Perl 5.6.0 accidentally pessimised certain kinds
2230 of simple pattern matches. These are now handled better. [561]
2234 Regular expression debug output (whether through C<use re 'debug'>
2235 or via C<-Dr>) now looks better. [561]
2239 Multi-line matches like C<"a\nxb\n" =~ /(?!\A)x/m> were flawed. The
2240 bug has been fixed. [561]
2244 Use of $& could trigger a core dump under some situations. This
2245 is now avoided. [561]
2249 The regular expression captured submatches ($1, $2, ...) are now
2250 more consistently unset if the match fails, instead of leaving false
2251 data lying around in them. [561]
2255 readline() on files opened in "slurp" mode could return an extra
2256 "" (blank line) at the end in certain situations. This has been
2261 Autovivification of symbolic references of special variables described
2262 in L<perlvar> (as in C<${$num}>) was accidentally disabled. This works
2267 Sys::Syslog ignored the C<LOG_AUTH> constant.
2271 All but the first argument of the IO syswrite() method are now optional.
2275 $AUTOLOAD, sort(), lock(), and spawning subprocesses
2276 in multiple threads simultaneously are now thread-safe.
2280 Tie::Array's SPLICE method was broken.
2284 Allow a read-only string on the left-hand side of a non-modifying tr///.
2288 If C<STDERR> is tied, warnings caused by C<warn> and C<die> now
2289 correctly pass to it.
2293 Several Unicode fixes.
2299 BOMs (byte order marks) at the beginning of Perl files
2300 (scripts, modules) should now be transparently skipped.
2301 UTF-16 and UCS-2 encoded Perl files should now be read correctly.
2305 The character tables have been updated to Unicode 3.2.0.
2309 Comparing with utf8 data does not magically upgrade non-utf8 data
2310 into utf8. (This was a problem for example if you were mixing data
2311 from I/O and Unicode data: your output might have got magically encoded
2316 Generating illegal Unicode code points such as U+FFFE, or the UTF-16
2317 surrogates, now also generates an optional warning.
2321 C<IsAlnum>, C<IsAlpha>, and C<IsWord> now match titlecase.
2325 Concatenation with the C<.> operator or via variable interpolation,
2326 C<eq>, C<substr>, C<reverse>, C<quotemeta>, the C<x> operator,
2327 substitution with C<s///>, single-quoted UTF8, should now work.
2331 The C<tr///> operator now works. Note that the C<tr///CU>
2332 functionality has been removed (but see pack('U0', ...)).
2336 C<eval "v200"> now works.
2340 Perl 5.6.0 parsed m/\x{ab}/ incorrectly, leading to spurious warnings.
2341 This has been corrected. [561]
2345 Zero entries were missing from the Unicode classes such as C<IsDigit>.
2351 Large unsigned numbers (those above 2**31) could sometimes lose their
2352 unsignedness, causing bogus results in arithmetic operations. [561]
2356 The Perl parser has been stress tested using both random input and
2357 Markov chain input and the few found crashes and lockups have been
2362 =head2 Platform Specific Changes and Fixes
2370 Perl now works on post-4.0 BSD/OSes.
2376 Setting C<$0> now works (as much as possible; see L<perlvar> for details).
2382 Numerous updates; currently synchronised with Cygwin 1.3.10.
2386 Previously DYNIX/ptx had problems in its Configure probe for non-blocking I/O.
2392 EPOC now better supported. See README.epoc. [561]
2398 Perl now works on post-3.0 FreeBSDs.
2404 README.hpux updated; C<Configure -Duse64bitall> now works;
2405 now uses HP-UX malloc instead of Perl malloc.
2411 Numerous compilation flag and hint enhancements; accidental mixing
2412 of 32-bit and 64-bit libraries (a doomed attempt) made much harder.
2422 Long doubles should now work (see INSTALL). [561]
2426 Linux previously had problems related to sockaddrlen when using
2427 accept(), recvfrom() (in Perl: recv()), getpeername(), and
2436 Compilation of the standard Perl distribution in Mac OS Classic should
2437 now work if you have the Metrowerks development environment and the
2438 missing Mac-specific toolkit bits. Contact the macperl mailing list
2445 MPE/iX update after Perl 5.6.0. See README.mpeix. [561]
2449 NetBSD/threads: try installing the GNU pth (should be in the
2450 packages collection, or http://www.gnu.org/software/pth/),
2451 and Configure with -Duseithreads.
2457 Perl now works on NetBSD/sparc.
2463 Now works with usethreads (see INSTALL). [561]
2469 64-bitness using the Sun Workshop compiler now works.
2475 The native build method requires at least VOS Release 14.5.0
2476 and GNU C++/GNU Tools 2.0.1 or later. The Perl pack function
2477 now maps overflowed values to +infinity and underflowed values
2482 Tru64 (aka Digital UNIX, aka DEC OSF/1)
2484 The operating system version letter now recorded in $Config{osvers}.
2485 Allow compiling with gcc (previously explicitly forbidden). Compiling
2486 with gcc still not recommended because buggy code results, even with
2493 Fixed various alignment problems that lead into core dumps either
2494 during build or later; no longer dies on math errors at runtime;
2495 now using full quad integers (64 bits), previously was using
2496 only 46 bit integers for speed.
2502 chdir() now works better despite a CRT bug; now works with MULTIPLICITY
2503 (see INSTALL); now works with Perl's malloc.
2505 The tainting of C<%ENV> elements via C<keys> or C<values> was previously
2506 unimplemented. It now works as documented.
2508 The C<waitpid> emulation has been improved. The worst bug (now fixed)
2509 was that a pid of -1 would cause a wildcard search of all processes on
2512 POSIX-style signals are now emulated much better on VMS versions prior
2515 The C<system> function and backticks operator have improved
2516 functionality and better error handling. [561]
2518 File access tests now use current process privileges rather than the
2519 user's default privileges, which could sometimes result in a mismatch
2520 between reported access and actual access.
2522 There is a new C<kill> implementation based on C<sys$sigprc> that allows
2523 older VMS systems (pre-7.0) to use C<kill> to send signals rather than
2524 simply force exit. This implementation also allows later systems to
2525 call C<kill> from within a signal handler.
2527 Iterative logical name translations are now limited to 10 iterations in
2528 imitation of SHOW LOGICAL and other OpenVMS facilities.
2538 accept() no longer leaks memory. [561]
2542 Borland C++ v5.5 is now a supported compiler that can build Perl.
2543 However, the generated binaries continue to be incompatible with those
2544 generated by the other supported compilers (GCC and Visual C++). [561]
2548 Better chdir() return value for a non-existent directory.
2552 Duping socket handles with open(F, ">&MYSOCK") now works under Windows
2557 New %ENV entries now propagate to subprocesses. [561]
2561 Current directory entries in %ENV are now correctly propagated to child
2566 $ENV{LIB} now used to search for libs under Visual C.
2570 fork() emulation has been improved in various ways, but still continues
2571 to be experimental. See L<perlfork> for known bugs and caveats.
2575 A failed (pseudo)fork now returns undef and sets errno to EAGAIN.
2579 Win32::GetCwd() correctly returns C:\ instead of C: when at the drive root.
2580 Other bugs in chdir() and Cwd::cwd() have also been fixed. [561]
2584 HTML files will be installed in c:\perl\html instead of c:\perl\lib\pod\html
2588 The makefiles now provide a single switch to bulk-enable all the
2589 features enabled in ActiveState ActivePerl (a popular Win32 binary
2590 distribution). [561]
2594 Allow REG_EXPAND_SZ keys in the registry.
2598 Can now send() from all threads, not just the first one. [561]
2602 ExtUtils::MakeMaker now uses $ENV{LIB} to search for libraries.
2606 Fake signal handling reenabled, bugs and all.
2610 %SIG has been enabled under USE_ITHREADS, but its use is completely
2611 unsupported under all configurations. [561]
2615 Less stack reserved per thread so that more threads can run
2616 concurrently. (Still 16M per thread.)
2620 C<< File::Spec->tmpdir() >> now prefers C:/temp over /tmp
2621 (works better when perl is running as service).
2625 Better UNC path handling under ithreads. [561]
2629 wait(), waitpid(), and backticks now return the correct exit status
2630 under Windows 9x. [561]
2634 Non-blocking waits for child processes (or pseudo-processes) are
2635 supported via C<waitpid($pid, &POSIX::WNOHANG)>.
2639 Win64 compilation is now supported.
2643 winsock handle leak fixed. [561]
2649 =head1 New or Changed Diagnostics
2655 The lexical warnings category "deprecated" is no longer a sub-category
2656 of the "syntax" category. It is now a top-level category in its own
2661 All regular expression compilation error messages are now hopefully
2662 easier to understand both because the error message now comes before
2663 the failed regex and because the point of failure is now clearly
2664 marked by a C<E<lt>-- HERE> marker.
2668 The various "opened only for", "on closed", "never opened" warnings
2669 drop the C<main::> prefix for filehandles in the C<main> package,
2670 for example C<STDIN> instead of C<main::STDIN>.
2674 The "Unrecognized escape" warning has been extended to include C<\8>,
2675 C<\9>, and C<\_>. There is no need to escape any of the C<\w> characters.
2679 Two new debugging options have been added: if you have compiled your
2680 Perl with debugging, you can use the -DT [561] and -DR options to trace
2681 tokenising and to add reference counts to displaying variables,
2686 The debugger (perl5db.pl) has been modified to present a more
2687 consistent commands interface, via (CommandSet=580). perl5db.t was
2688 also added to test the changes, and as a placeholder for further tests.
2694 The debugger has a new C<dumpDepth> option to control the maximum
2695 depth to which nested structures are dumped. The C<x> command has
2696 been extended so that C<x N EXPR> dumps out the value of I<EXPR> to a
2697 depth of at most I<N> levels.
2701 The debugger can now show lexical variables if you have the CPAN
2702 module PadWalker installed.
2706 If an attempt to use a (non-blessed) reference as an array index
2707 is made, a warning is given.
2711 C<push @a;> and C<unshift @a;> (with no values to push or unshift)
2712 now give a warning. This may be a problem for generated and evaled
2717 If you try to L<perlfunc/pack> a number less than 0 or larger than 255
2718 using the C<"C"> format you will get an optional warning. Similarly
2719 for the C<"c"> format and a number less than -128 or more than 127.
2723 Certain regex modifiers such as C<(?o)> make sense only if applied to
2724 the entire regex. You will get an optional warning if you try to do
2729 Using arrays or hashes as references (e.g. C<< %foo->{bar} >>
2730 has been deprecated for a while. Now you will get an optional warning.
2734 Using C<sort> in scalar context now issues an optional warning.
2735 This didn't do anything useful, as the sort was not performed.
2739 =head1 Changed Internals
2745 perlapi.pod (a companion to perlguts) now attempts to document the
2750 You can now build a really minimal perl called microperl.
2751 Building microperl does not require even running Configure;
2752 C<make -f Makefile.micro> should be enough. Beware: microperl makes
2753 many assumptions, some of which may be too bold; the resulting
2754 executable may crash or otherwise misbehave in wondrous ways.
2755 For careful hackers only.
2759 Added rsignal(), whichsig(), do_join(), op_clear, op_null,
2760 ptr_table_clear(), ptr_table_free(), sv_setref_uv(), and several UTF-8
2761 interfaces to the publicised API. For the full list of the available
2762 APIs see L<perlapi>.
2766 Made possible to propagate customised exceptions via croak()ing.
2770 Now xsubs can have attributes just like subs. (Well, at least the
2771 built-in attributes.)
2775 dTHR and djSP have been obsoleted; the former removed (because it's
2776 a no-op) and the latter replaced with dSP.
2780 PERL_OBJECT has been completely removed.
2784 The MAGIC constants (e.g. C<'P'>) have been macrofied
2785 (e.g. C<PERL_MAGIC_TIED>) for better source code readability
2786 and maintainability.
2790 The regex compiler now maintains a structure that identifies nodes in
2791 the compiled bytecode with the corresponding syntactic features of the
2792 original regex expression. The information is attached to the new
2793 C<offsets> member of the C<struct regexp>. See L<perldebguts> for more
2794 complete information.
2798 The C code has been made much more C<gcc -Wall> clean. Some warning
2799 messages still remain in some platforms, so if you are compiling with
2800 gcc you may see some warnings about dubious practices. The warnings
2801 are being worked on.
2805 F<perly.c>, F<sv.c>, and F<sv.h> have now been extensively commented.
2809 Documentation on how to use the Perl source repository has been added
2810 to F<Porting/repository.pod>.
2814 There are now several profiling make targets.
2818 =head1 Security Vulnerability Closed [561]
2820 (This change was already made in 5.7.0 but bears repeating here.)
2822 A potential security vulnerability in the optional suidperl component
2823 of Perl was identified in August 2000. suidperl is neither built nor
2824 installed by default. As of November 2001 the only known vulnerable
2825 platform is Linux, most likely all Linux distributions. CERT and
2826 various vendors and distributors have been alerted about the vulnerability.
2827 See http://www.cpan.org/src/5.0/sperl-2000-08-05/sperl-2000-08-05.txt
2828 for more information.
2830 The problem was caused by Perl trying to report a suspected security
2831 exploit attempt using an external program, /bin/mail. On Linux
2832 platforms the /bin/mail program had an undocumented feature which
2833 when combined with suidperl gave access to a root shell, resulting in
2834 a serious compromise instead of reporting the exploit attempt. If you
2835 don't have /bin/mail, or if you have 'safe setuid scripts', or if
2836 suidperl is not installed, you are safe.
2838 The exploit attempt reporting feature has been completely removed from
2839 Perl 5.8.0 (and the maintenance release 5.6.1, and it was removed also
2840 from all the Perl 5.7 releases), so that particular vulnerability
2841 isn't there anymore. However, further security vulnerabilities are,
2842 unfortunately, always possible. The suidperl functionality is most
2843 probably going to be removed in Perl 5.10. In any case, suidperl
2844 should only be used by security experts who know exactly what they are
2845 doing and why they are using suidperl instead of some other solution
2846 such as sudo ( see http://www.courtesan.com/sudo/ ).
2850 Several new tests have been added, especially for the F<lib> and F<ext>
2851 subsections. There are now about 65 000 individual tests (spread over
2852 about 700 test scripts), in the regression suite (5.6.1 has about
2853 11700 tests, in 258 test scripts) Many of the new tests are of course
2854 introduced by the new modules, but still in general Perl is now more
2857 Because of the large number of tests, running the regression suite
2858 will take considerably longer time than it used to: expect the suite
2859 to take up to 4-5 times longer to run than in perl 5.6. On a really
2860 fast machine you can hope to finish the suite in about 6-8 minutes
2863 The tests are now reported in a different order than in earlier Perls.
2864 (This happens because the test scripts from under t/lib have been moved
2865 to be closer to the library/extension they are testing.)
2867 =head1 Known Problems
2875 If using the AIX native make command, instead of just "make" issue
2876 "make all". In some setups the former has been known to spuriously
2877 also try to run "make install". Alternatively, you may want to use
2882 In AIX 4.2, Perl extensions that use C++ functions that use statics
2883 may have problems in that the statics are not getting initialized.
2884 In newer AIX releases, this has been solved by linking Perl with
2885 the libC_r library, but unfortunately in AIX 4.2 the said library
2886 has an obscure bug where the various functions related to time
2887 (such as time() and gettimeofday()) return broken values, and
2888 therefore in AIX 4.2 Perl is not linked against libC_r.
2892 vac 5.0.0.0 May Produce Buggy Code For Perl
2894 The AIX C compiler vac version 5.0.0.0 may produce buggy code,
2895 resulting in a few random tests failing when run as part of "make
2896 test", but when the failing tests are run by hand, they succeed.
2897 We suggest upgrading to at least vac version 5.0.1.0, that has been
2898 known to compile Perl correctly. "lslpp -L|grep vac.C" will tell
2899 you the vac version. See README.aix.
2903 If building threaded Perl, you may get compilation warning from pp_sys.c:
2905 "pp_sys.c", line 4651.39: 1506-280 (W) Function argument assignment between types "unsigned char*" and "const void*" is not allowed.
2907 This is harmless; it is caused by the getnetbyaddr() and getnetbyaddr_r()
2908 having slightly different types for their first argument.
2912 =head2 Alpha systems with old gccs fail several tests
2914 If you see op/pack, op/pat, op/regexp, or ext/Storable tests failing
2915 in a Linux/alpha or *BSD/Alpha, it's probably time to upgrade your gcc.
2916 gccs prior to 2.95.3 are definitely not good enough, and gcc 3.1 may
2917 be even better. (RedHat Linux/alpha with gcc 3.1 reported no problems,
2918 as did Linux 2.4.18 with gcc 2.95.4.) (In Tru64, it is preferable to
2919 use the bundled C compiler.)
2923 Perl 5.8.0 doesn't build in AmigaOS. It broke at some point during
2924 the ithreads work and we could not find Amiga experts to unbreak the
2925 problems. Perl 5.6.1 still works for AmigaOS (as does the the 5.7.2
2926 development release).
2930 The following tests fail on 5.8.0 Perl in BeOS Personal 5.03:
2932 t/op/lfs............................FAILED at test 17
2933 t/op/magic..........................FAILED at test 24
2934 ext/POSIX/t/sigaction...............FAILED at test 13
2935 ext/POSIX/t/waitpid.................FAILED at test 1
2937 See L<perlbeos> (README.beos) for more details.
2939 =head2 Cygwin "unable to remap"
2941 For example when building the Tk extension for Cygwin,
2942 you may get an error message saying "unable to remap".
2943 This is known problem with Cygwin, and a workaround is
2944 detailed in here: http://sources.redhat.com/ml/cygwin/2001-12/msg00894.html
2946 =head2 ext/threads/t/libc
2948 If this test fails, it indicates that your libc (C library) is not
2949 threadsafe. This particular test stress tests the localtime() call to
2950 find out whether it is threadsafe. See L<perlthrtut> for more information.
2952 =head2 FreeBSD Failing locale Test 117 For ISO8859-15 Locales
2954 The ISO8859-15 locales may fail the locale test 117 in FreeBSD.
2955 This is caused by the characters \xFF (y with diaeresis) and \xBE
2956 (Y with diaeresis) not behaving correctly when being matched
2959 =head2 IRIX fails ext/List/Util/t/shuffle.t
2961 IRIX with MIPSpro 7.3.1.3m compiler may fail the said List::Util test
2962 by dumping core. This seems to be a compiler error since if compiled
2963 with gcc no core dump ensues, and no failures on the said test on any
2966 =head2 Modifying $_ Inside for(..)
2970 works without complaint. It shouldn't. (You should be able to
2971 modify only lvalue elements inside the loops.) You can see the
2972 correct behaviour by replacing the 1..5 with 1, 2, 3, 4, 5.
2974 =head2 mod_perl 1.26 Doesn't Build With Threaded Perl
2976 Use mod_perl 1.27 or higher.
2978 =head2 lib/ftmp-security tests warn 'system possibly insecure'
2980 Don't panic. Read the 'make test' section of INSTALL instead.
2982 =head2 HP-UX lib/posix Subtest 9 Fails When LP64-Configured
2984 If perl is configured with -Duse64bitall, the successful result of the
2985 subtest 10 of lib/posix may arrive before the successful result of the
2986 subtest 9, which confuses the test harness so much that it thinks the
2989 =head2 Linux with glibc 2.2.5 fails t/op/int subtest #6 with -Duse64bitint
2991 This is a known bug in the glibc 2.2.5 with long long integers.
2992 ( http://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=65612 )
2994 =head2 Linux With Sfio Fails op/misc Test 48
2998 =head2 libwww-perl (LWP) fails base/date #51
3000 Use libwww-perl 5.65 or later.
3004 Please remember to set your environment variable LC_ALL to "C"
3005 (setenv LC_ALL C) before running "make test" to avoid a lot of
3006 warnings about the broken locales of Mac OS X.
3008 The following tests are known to fail in Mac OS X 10.1.5 because of
3009 buggy (old) implementations of Berkeley DB included in Mac OS X:
3011 Failed Test Stat Wstat Total Fail Failed List of Failed
3012 -------------------------------------------------------------------------
3013 ../ext/DB_File/t/db-btree.t 0 11 ?? ?? % ??
3014 ../ext/DB_File/t/db-recno.t 149 3 2.01% 61 63 65
3016 If you are building on a UFS partition, you will also probably see
3017 t/op/stat.t subtest #9 fail. This is caused by Darwin's UFS not
3018 supporting inode change time.
3020 Also the ext/POSIX/t/posix.t subtest #10 fails but it is skipped for
3021 now because the failure is Apple's fault, not Perl's (blocked signals
3024 If you Configure with ithreads, ext/threads/t/libc.t will fail. Again,
3025 this is not Perl's fault-- the libc of Mac OS X is not threadsafe
3026 (in this particular test, the localtime() call is found to be
3029 =head2 op/sprintf tests 91, 129, and 130
3031 The op/sprintf tests 91, 129, and 130 are known to fail on some platforms.
3032 Examples include any platform using sfio, and Compaq/Tandem's NonStop-UX.
3034 Test 91 is known to fail on QNX6 (nto), because C<sprintf '%e',0>
3035 incorrectly produces C<0.000000e+0> instead of C<0.000000e+00>.
3037 For tests 129 and 130, the failing platforms do not comply with
3038 the ANSI C Standard: lines 19ff on page 134 of ANSI X3.159 1989, to
3039 be exact. (They produce something other than "1" and "-1" when
3040 formatting 0.6 and -0.6 using the printf format "%.0f"; most often,
3041 they produce "0" and "-0".)
3045 In case you are still using Solaris 2.5 (aka SunOS 5.5), you may
3046 experience failures (the test core dumping) in lib/locale.t.
3047 The suggested cure is to upgrade your Solaris.
3049 =head2 SUPER-UX (NEC SX)
3051 The following tests are known to fail on SUPER-UX:
3053 op/64bitint...........................FAILED tests 29-30, 32-33, 35-36
3054 op/arith..............................FAILED tests 128-130
3055 op/pack...............................FAILED tests 25-5625
3056 op/pow................................
3057 op/taint..............................# msgsnd failed
3058 ../ext/IO/lib/IO/t/io_poll............FAILED tests 3-4
3059 ../ext/IPC/SysV/ipcsysv...............FAILED tests 2, 5-6
3060 ../ext/IPC/SysV/t/msg.................FAILED tests 2, 4-6
3061 ../ext/Socket/socketpair..............FAILED tests 12
3062 ../lib/IPC/SysV.......................FAILED tests 2, 5-6
3063 ../lib/warnings.......................FAILED tests 115-116, 118-119
3065 The op/pack failure ("Cannot compress negative numbers at op/pack.t line 126")
3066 is serious but as of yet unsolved. It points at some problems with the
3067 signedness handling of the C compiler, as do the 64bitint, arith, and pow
3068 failures. Most of the rest point at problems with SysV IPC.
3070 =head2 Term::ReadKey not working on Win32
3072 Use Term::ReadKey 2.20 or later.
3074 =head2 Failure of Thread (5.005-style) tests
3076 B<Note that support for 5.005-style threading is deprecated,
3077 experimental and practically unsupported. In 5.10, it is expected
3080 The following tests are known to fail due to fundamental problems in
3081 the 5.005 threading implementation. These are not new failures--Perl
3082 5.005_0x has the same bugs, but didn't have these tests.
3084 ../ext/B/t/xref.t 255 65280 14 12 85.71% 3-14
3085 ../ext/List/Util/t/first.t 255 65280 7 4 57.14% 2 5-7
3086 ../lib/English.t 2 512 54 2 3.70% 2-3
3087 ../lib/ExtUtils/t/basic.t 1 256 17 1 5.88% 14
3088 ../lib/FileCache.t 5 1 20.00% 5
3089 ../lib/Filter/Simple/t/data.t 6 3 50.00% 1-3
3090 ../lib/Filter/Simple/t/filter_onl 9 3 33.33% 1-2 5
3091 ../lib/Tie/File/t/31_autodefer.t 255 65280 65 32 49.23% 34-65
3092 ../lib/autouse.t 10 1 10.00% 4
3093 op/flip.t 15 1 6.67% 15
3095 These failures are unlikely to get fixed as 5.005-style threads
3096 are considered fundamentally broken. (Basically what happens is that
3097 competing threads can corrupt shared global state.)
3099 =head2 Timing problems
3101 The following tests may fail intermittently because of timing
3102 problems, for example if the system is heavily loaded.
3105 ext/Time/HiRes/HiRes.t
3107 lib/Memoize/t/expmod_t.t
3108 lib/Memoize/t/speed.t
3110 In case of failure please try running them manually, for example
3112 ./perl -Ilib ext/Time/HiRes/HiRes.t
3116 ../lib/Math/Trig.t 26 1 3.85% 25
3117 ../lib/warnings.t 470 1 0.21% 429
3119 The Trig.t failure is caused by the slighly differing (from IEEE)
3120 floating point implementation of UNICOS. The warnings.t failure is
3121 also related: the test assumes a certain floating point output format;
3122 this assumption fails in UNICOS.
3130 During Configure, the test
3132 Guessing which symbols your C compiler and preprocessor define...
3134 will probably fail with error messages like
3136 CC-20 cc: ERROR File = try.c, Line = 3
3137 The identifier "bad" is undefined.
3139 bad switch yylook 79bad switch yylook 79bad switch yylook 79bad switch yylook 79#ifdef A29K
3142 CC-65 cc: ERROR File = try.c, Line = 3
3143 A semicolon is expected at this point.
3145 This is caused by a bug in the awk utility of UNICOS/mk. You can ignore
3146 the error, but it does cause a slight problem: you cannot fully
3147 benefit from the h2ph utility (see L<h2ph>) that can be used to
3148 convert C headers to Perl libraries, mainly used to be able to access
3149 from Perl the constants defined using C preprocessor, cpp. Because of
3150 the above error, parts of the converted headers will be invisible.
3151 Luckily, these days the need for h2ph is rare.
3155 If building Perl with interpreter threads (ithreads), the
3156 getgrent(), getgrnam(), and getgrgid() functions cannot return the
3157 list of the group members due to a bug in the multithreaded support of
3158 UNICOS/mk. What this means is that in list context the functions will
3159 return only three values, not four.
3165 There are a few known test failures, see L<perluts> (README.uts).
3167 =head2 VOS (Stratus)
3169 When Perl is built using the native build process on VOS Release
3170 14.5.0 and GNU C++/GNU Tools 2.0.1, all attempted tests either
3171 pass or result in TODO (ignored) failures.
3175 There should be no reported test failures with a default configuration,
3176 though there are a number of tests marked TODO that point to areas
3177 needing further debugging and/or porting work.
3181 In multi-CPU boxes, there are some problems with the I/O buffering:
3182 some output may appear twice.
3184 =head2 XML::Parser not working
3186 Use XML::Parser 2.31 or later.
3188 =head2 z/OS (OS/390)
3190 z/OS has rather many test failures but the situation is actually
3191 better than it was in 5.6.0; it's just that so many new modules and
3192 tests have been added.
3194 Failed Test Stat Wstat Total Fail Failed List of Failed
3195 ---------------------------------------------------------------------------
3196 ../ext/Data/Dumper/t/dumper.t 357 8 2.24% 311 314 325 327
3198 ../ext/IO/lib/IO/t/io_unix.t 5 4 80.00% 2-5
3199 ../ext/Storable/t/downgrade.t 12 3072 169 12 7.10% 14-15 46-47 78-79
3201 ../lib/ExtUtils/t/Constant.t 121 30976 48 48 100.00% 1-48
3202 ../lib/ExtUtils/t/Embed.t 9 9 100.00% 1-9
3203 op/pat.t 910 7 0.77% 665 776 785 832-
3205 op/sprintf.t 224 3 1.34% 98 100 136
3206 op/tr.t 97 5 5.15% 63 71-74
3207 uni/fold.t 780 6 0.77% 61 169 196 661
3210 The failures in dumper.t and downgrade.t are problems in the tests,
3211 those in io_unix and sprintf are problems in the USS (UDP sockets
3212 and printf formats). The pat, tr, and fold failures are genuine Perl
3213 problems caused by EBCDIC (and in the pat and fold cases, combining
3214 that with Unicode). The Constant and Embed are probably problems
3215 in the tests (since they test Perl's ability to build extensions,
3216 and that seems to be working reasonably well.)
3218 =head2 Localising Tied Arrays and Hashes Is Broken
3222 doesn't work as one would expect: the old value is restored
3223 incorrectly. This will be changed in a future release, but we don't
3224 know yet what the new semantics will exactly be. In any case, the
3225 change will break existing code that relies on the current
3226 (ill-defined) semantics, so just avoid doing this in general.
3228 =head2 Self-tying Problems
3230 Self-tying of arrays and hashes is broken in rather deep and
3231 hard-to-fix ways. As a stop-gap measure to avoid people from getting
3232 frustrated at the mysterious results (core dumps, most often), it is
3233 forbidden for now (you will get a fatal error even from an attempt).
3235 A change to self-tying of globs has caused them to be recursively
3236 referenced (see: L<perlobj/"Two-Phased Garbage Collection">). You
3237 will now need an explicit untie to destroy a self-tied glob. This
3238 behaviour may be fixed at a later date.
3240 Self-tying of scalars and IO thingies works.
3242 =head2 Building Extensions Can Fail Because Of Largefiles
3244 Some extensions like mod_perl are known to have issues with
3245 `largefiles', a change brought by Perl 5.6.0 in which file offsets
3246 default to 64 bits wide, where supported. Modules may fail to compile
3247 at all, or they may compile and work incorrectly. Currently, there
3248 is no good solution for the problem, but Configure now provides
3249 appropriate non-largefile ccflags, ldflags, libswanted, and libs
3250 in the %Config hash (e.g., $Config{ccflags_nolargefiles}) so the
3251 extensions that are having problems can try configuring themselves
3252 without the largefileness. This is admittedly not a clean solution,
3253 and the solution may not even work at all. One potential failure is
3254 whether one can (or, if one can, whether it's a good idea to) link
3255 together at all binaries with different ideas about file offsets;
3256 all this is platform-dependent.
3258 =head2 Unicode Support on EBCDIC Still Spotty
3260 Though mostly working, Unicode support still has problem spots on
3261 EBCDIC platforms. One such known spot are the C<\p{}> and C<\P{}>
3262 regular expression constructs for code points less than 256: the
3263 C<pP> are testing for Unicode code points, not knowing about EBCDIC.
3265 =head2 The Compiler Suite Is Still Very Experimental
3267 The compiler suite is slowly getting better but it continues to be
3268 highly experimental. Use in production environments is discouraged.
3270 =head2 The Long Double Support Is Still Experimental
3272 The ability to configure Perl's numbers to use "long doubles",
3273 floating point numbers of hopefully better accuracy, is still
3274 experimental. The implementations of long doubles are not yet
3275 widespread and the existing implementations are not quite mature
3276 or standardised, therefore trying to support them is a rare
3277 and moving target. The gain of more precision may also be offset
3278 by slowdown in computations (more bits to move around, and the
3279 operations are more likely to be executed by less optimised
3282 =head2 Seen In Perl 5.7 But Gone Now
3284 C<Time::Piece> (previously known as C<Time::Object>) was removed
3285 because it was felt that it didn't have enough value in it to be a
3286 core module. It is still a useful module, though, and is available
3289 Perl 5.8 unfortunately does not build anymore on AmigaOS; this broke
3290 accidentally at some point. Since there are not that many Amiga
3291 developers available, we could not get this fixed and tested in time
3292 for 5.8.0. Perl 5.6.1 still works for AmigaOS (as does the the 5.7.2
3293 development release).
3295 =head1 Reporting Bugs
3297 If you find what you think is a bug, you might check the articles
3298 recently posted to the comp.lang.perl.misc newsgroup and the perl
3299 bug database at http://bugs.perl.org/ . There may also be
3300 information at http://www.perl.com/ , the Perl Home Page.
3302 If you believe you have an unreported bug, please run the B<perlbug>
3303 program included with your release. Be sure to trim your bug down
3304 to a tiny but sufficient test case. Your bug report, along with the
3305 output of C<perl -V>, will be sent off to perlbug@perl.org to be
3306 analysed by the Perl porting team.
3310 The F<Changes> file for exhaustive details on what changed.
3312 The F<INSTALL> file for how to build Perl.
3314 The F<README> file for general stuff.
3316 The F<Artistic> and F<Copying> files for copyright information.
3320 Written by Jarkko Hietaniemi <F<jhi@iki.fi>>.