3 perldelta - what is new for perl v5.8.0
7 This document describes differences between the 5.6.0 release and
10 Many of the bug fixes in 5.8.0 were already seen in the 5.6.1
11 maintenance release since the two releases were kept closely
12 coordinated (while 5.8.0 was still called 5.7.something).
14 Changes that were integrated into the 5.6.1 release are marked C<[561]>.
15 Many of these changes have been further developed since 5.6.1 was released,
16 those are marked C<[561+]>.
18 You can see the list of changes in the 5.6.1 release (both from the
19 5.005_03 release and the 5.6.0 release) by reading L<perl561delta>.
21 =head1 Highlights In 5.8.0
27 Better Unicode support
35 New Thread Implementation
39 Better Numeric Accuracy
51 More Extensive Regression Testing
55 =head1 Incompatible Changes
57 =head2 Binary Incompatibility
59 B<Perl 5.8 is not binary compatible with earlier releases of Perl.>
61 B<You have to recompile your XS modules.>
63 (Pure Perl modules should continue to work.)
65 The major reason for the discontinuity is the new IO architecture
66 called PerlIO. PerlIO is the default configuration because without
67 it many new features of Perl 5.8 cannot be used. In other words:
68 you just have to recompile your modules containing XS code, sorry
71 In future releases of Perl, non-PerlIO aware XS modules may become
72 completely unsupported. This shouldn't be too difficult for module
73 authors, however: PerlIO has been designed as a drop-in replacement
74 (at the source code level) for the stdio interface.
76 Depending on your platform, there are also other reasons why
77 we decided to break binary compatibility, please read on.
79 =head2 64-bit platforms and malloc
81 If your pointers are 64 bits wide, the Perl malloc is no longer being
82 used because it does not work well with 8-byte pointers. Also,
83 usually the system mallocs on such platforms are much better optimized
84 for such large memory models than the Perl malloc. Some memory-hungry
85 Perl applications like the PDL don't work well with Perl's malloc.
86 Finally, other applications than Perl (such as mod_perl) tend to prefer
87 the system malloc. Such platforms include Alpha and 64-bit HPPA,
90 =head2 AIX Dynaloading
92 The AIX dynaloading now uses in AIX releases 4.3 and newer the native
93 dlopen interface of AIX instead of the old emulated interface. This
94 change will probably break backward compatibility with compiled
95 modules. The change was made to make Perl more compliant with other
96 applications like mod_perl which are using the AIX native interface.
98 =head2 Attributes for C<my> variables now handled at run-time
100 The C<my EXPR : ATTRS> syntax now applies variable attributes at
101 run-time. (Subroutine and C<our> variables still get attributes applied
102 at compile-time.) See L<attributes> for additional details. In particular,
103 however, this allows variable attributes to be useful for C<tie> interfaces,
104 which was a deficiency of earlier releases. Note that the new semantics
105 doesn't work with the Attribute::Handlers module (as of version 0.76).
107 =head2 Socket Extension Dynamic in VMS
109 The Socket extension is now dynamically loaded instead of being
110 statically built in. This may or may not be a problem with ancient
111 TCP/IP stacks of VMS: we do not know since we weren't able to test
112 Perl in such configurations.
114 =head2 IEEE-format Floating Point Default on OpenVMS Alpha
116 Perl now uses IEEE format (T_FLOAT) as the default internal floating
117 point format on OpenVMS Alpha, potentially breaking binary compatibility
118 with external libraries or existing data. G_FLOAT is still available as
119 a configuration option. The default on VAX (D_FLOAT) has not changed.
121 =head2 New Unicode Semantics (no more C<use utf8>)
123 Previously in Perl 5.6 to use Unicode one would say "use utf8" and
124 then the operations (like string concatenation) were Unicode-aware
125 in that lexical scope.
127 This was found to be an inconvenient interface, and in Perl 5.8 the
128 Unicode model has completely changed: now the "Unicodeness" is bound
129 to the data itself, and for most of the time "use utf8" is not needed
130 at all. The only remaining use of "use utf8" is when the Perl script
131 itself has been written in the UTF-8 encoding of Unicode.
133 See L<perluniintro> for the explanation of the current model,
134 and L<utf8> for the current use of the utf8 pragma.
136 =head2 New Unicode Properties
138 Unicode I<scripts> are now supported. Scripts are similar to (and superior
139 to) Unicode I<blocks>. The difference between scripts and blocks is that
140 scripts are the glyphs used by a language or a group of languages, while
141 the blocks are more artificial groupings of (mostly) 256 characters based
142 on the Unicode numbering.
144 In general, scripts are more inclusive, but not universally so. For
145 example, while the script C<Latin> includes all the Latin characters and
146 their various diacritic-adorned versions, it does not include the various
147 punctuation or digits (since they are not solely C<Latin>).
149 A number of other properties are now supported, including C<\p{L&}>,
150 C<\p{Any}> C<\p{Assigned}>, C<\p{Unassigned}>, C<\p{Blank}> [561] and
151 C<\p{SpacePerl}> [561] (along with their C<\P{...}> versions, of course).
152 See L<perlunicode> for details, and more additions.
154 The C<In> or C<Is> prefix to names used with the C<\p{...}> and C<\P{...}>
155 are now almost always optional. The only exception is that a C<In> prefix
156 is required to signify a Unicode block when a block name conflicts with a
157 script name. For example, C<\p{Tibetan}> refers to the script, while
158 C<\p{InTibetan}> refers to the block. When there is no name conflict, you
159 can omit the C<In> from the block name (e.g. C<\p{BraillePatterns}>), but
160 to be safe, it's probably best to always use the C<In>).
162 =head2 REF(...) Instead Of SCALAR(...)
164 A reference to a reference now stringifies as "REF(0x81485ec)" instead
165 of "SCALAR(0x81485ec)" in order to be more consistent with the return
168 =head2 pack/unpack D/F recycled
170 The undocumented pack/unpack template letters D/F have been recycled
171 for better use: now they stand for long double (if supported by the
172 platform) and NV (Perl internal floating point type). (They used
173 to be aliases for d/f, but you never knew that.)
175 =head2 glob() now returns filenames in alphabetical order
177 The list of filenames from glob() (or <...>) is now by default sorted
178 alphabetically to be csh-compliant (which is what happened before
179 in most UNIX platforms). (bsd_glob() does still sort platform
180 natively, ASCII or EBCDIC, unless GLOB_ALPHASORT is specified.) [561]
188 The semantics of bless(REF, REF) were unclear and until someone proves
189 it to make some sense, it is forbidden.
193 The obsolete chat2 library that should never have been allowed
194 to escape the laboratory has been decommissioned.
198 The builtin dump() function has probably outlived most of its
199 usefulness. The core-dumping functionality will remain in future
200 available as an explicit call to C<CORE::dump()>, but in future
201 releases the behaviour of an unqualified C<dump()> call may change.
205 The very dusty examples in the eg/ directory have been removed.
206 Suggestions for new shiny examples welcome but the main issue is that
207 the examples need to be documented, tested and (most importantly)
212 The (bogus) escape sequences \8 and \9 now give an optional warning
213 ("Unrecognized escape passed through"). There is no need to \-escape
218 The *glob{FILEHANDLE} is deprecated, use *glob{IO} instead.
222 The C<package;> syntax (C<package> without an argument) has been
223 deprecated. Its semantics were never that clear and its
224 implementation even less so. If you have used that feature to
225 disallow all but fully qualified variables, C<use strict;> instead.
229 The unimplemented POSIX regex features [[.cc.]] and [[=c=]] are still
230 recognised but now cause fatal errors. The previous behaviour of
231 ignoring them by default and warning if requested was unacceptable
232 since it, in a way, falsely promised that the features could be used.
236 In future releases, non-PerlIO aware XS modules may become completely
237 unsupported. Since PerlIO is a drop-in replacement for stdio at the
238 source code level, this shouldn't be that drastic a change.
242 Previous versions of perl and some readings of some sections of Camel
243 III implied that the C<:raw> "discipline" was the inverse of C<:crlf>.
244 Turning off "clrfness" is no longer enough to make a stream truly
245 binary. So the PerlIO C<:raw> layer (or "discipline", to use the Camel
246 book's older terminology) is now formally defined as being equivalent
247 to binmode(FH) - which is in turn defined as doing whatever is
248 necessary to pass each byte as-is without any translation. In
249 particular binmode(FH) - and hence C<:raw> - will now turn off both
250 CRLF and UTF-8 translation and remove other layers (e.g. :encoding())
251 which would modify byte stream.
255 The current user-visible implementation of pseudo-hashes (the weird
256 use of the first array element) is deprecated starting from Perl 5.8.0
257 and will be removed in Perl 5.10.0, and the feature will be
258 implemented differently. Not only is the current interface rather
259 ugly, but the current implementation slows down normal array and hash
260 use quite noticeably. The C<fields> pragma interface will remain
261 available. The I<restricted hashes> interface is expected to
262 be the replacement interface (see L<Hash::Util>). If your existing
263 programs depends on the underlying implementation, consider using
264 L<Class::PseudoHash> from CPAN.
268 The syntaxes C<< @a->[...] >> and C<< %h->{...} >> have now been deprecated.
272 After years of trying, suidperl is considered to be too complex to
273 ever be considered truly secure. The suidperl functionality is likely
274 to be removed in a future release.
278 The 5.005 threads model (module C<Thread>) is deprecated and expected
279 to be removed in Perl 5.10. Multithreaded code should be migrated to
280 the new ithreads model (see L<threads>, L<threads::shared> and
285 The long deprecated uppercase aliases for the string comparison
286 operators (EQ, NE, LT, LE, GE, GT) have now been removed.
290 The tr///C and tr///U features have been removed and will not return;
291 the interface was a mistake. Sorry about that. For similar
292 functionality, see pack('U0', ...) and pack('C0', ...). [561]
296 Earlier Perls treated "sub foo (@bar)" as equivalent to "sub foo (@)".
297 The prototypes are now checked better at compile-time for invalid
298 syntax. An optional warning is generated ("Illegal character in
299 prototype...") but this may be upgraded to a fatal error in a future
304 The C<exec LIST> and C<system LIST> operations now produce warnings on
305 tainted data and in some future release they will produce fatal errors.
309 The existing behaviour when localising tied arrays and hashes is wrong,
310 and will be changed in a future release, so do not rely on the existing
311 behaviour. See L<"Localising Tied Arrays and Hashes Is Broken">.
315 =head1 Core Enhancements
317 =head2 Unicode Overhaul
319 Unicode in general should be now much more usable than in Perl 5.6.0
320 (or even in 5.6.1). Unicode can be used in hash keys, Unicode in
321 regular expressions should work now, Unicode in tr/// should work now,
322 Unicode in I/O should work now. See L<perluniintro> for introduction
323 and L<perlunicode> for details.
329 The Unicode Character Database coming with Perl has been upgraded
330 to Unicode 3.2.0. For more information, see http://www.unicode.org/ .
331 [561+] (5.6.1 has UCD 3.0.1.)
335 For developers interested in enhancing Perl's Unicode capabilities:
336 almost all the UCD files are included with the Perl distribution in
337 the F<lib/unicore> subdirectory. The most notable omission, for space
338 considerations, is the Unihan database.
342 The properties \p{Blank} and \p{SpacePerl} have been added. "Blank" is like
343 C isblank(), that is, it contains only "horizontal whitespace" (the space
344 character is, the newline isn't), and the "SpacePerl" is the Unicode
345 equivalent of C<\s> (\p{Space} isn't, since that includes the vertical
346 tabulator character, whereas C<\s> doesn't.)
348 See "New Unicode Properties" earlier in this document for additional
349 information on changes with Unicode properties.
353 =head2 PerlIO is Now The Default
359 IO is now by default done via PerlIO rather than system's "stdio".
360 PerlIO allows "layers" to be "pushed" onto a file handle to alter the
361 handle's behaviour. Layers can be specified at open time via 3-arg
364 open($fh,'>:crlf :utf8', $path) || ...
366 or on already opened handles via extended C<binmode>:
368 binmode($fh,':encoding(iso-8859-7)');
370 The built-in layers are: unix (low level read/write), stdio (as in
371 previous Perls), perlio (re-implementation of stdio buffering in a
372 portable manner), crlf (does CRLF <=> "\n" translation as on Win32,
373 but available on any platform). A mmap layer may be available if
374 platform supports it (mostly UNIXes).
376 Layers to be applied by default may be specified via the 'open' pragma.
378 See L</"Installation and Configuration Improvements"> for the effects
379 of PerlIO on your architecture name.
383 If your platform supports fork(), you can use the list form of C<open>
384 for pipes. For example:
386 open KID_PS, "-|", "ps", "aux" or die $!;
388 forks the ps(1) command (without spawning a shell, as there are more
389 than three arguments to open()), and reads its standard output via the
390 C<KID_PS> filehandle. See L<perlipc>.
394 File handles can be marked as accepting Perl's internal encoding of Unicode
395 (UTF-8 or UTF-EBCDIC depending on platform) by a pseudo layer ":utf8" :
397 open($fh,">:utf8","Uni.txt");
399 Note for EBCDIC users: the pseudo layer ":utf8" is erroneously named
400 for you since it's not UTF-8 what you will be getting but instead
401 UTF-EBCDIC. See L<perlunicode>, L<utf8>, and
402 http://www.unicode.org/unicode/reports/tr16/ for more information.
403 In future releases this naming may change. See L<perluniintro>
404 for more information about UTF-8.
408 If your environment variables (LC_ALL, LC_CTYPE, LANG, LANGUAGE) look
409 like you want to use UTF-8 (any of the the variables match C</utf-?8/i>),
410 your STDIN, STDOUT, STDERR handles and the default open layer
411 (see L<open>) are marked as UTF-8. (This feature, like other new
412 features that combine Unicode and I/O, work only if you are using
413 PerlIO, but that's the default.)
415 Note that after this Perl really does assume that everything is UTF-8:
416 for example if some input handle is not, Perl will probably very soon
417 complain about the input data like this "Malformed UTF-8 ..." since
418 any old eight-bit data is not legal UTF-8.
420 Note for code authors: if you want to enable your users to use UTF-8
421 as their default encoding but in your code still have eight-bit I/O streams
422 (such as images or zip files), you need to explicitly open() or binmode()
423 with C<:bytes> (see L<perlfunc/open> and L<perlfunc/binmode>), or you
424 can just use C<binmode(FH)> (nice for pre-5.8.0 backward compatibility).
428 File handles can translate character encodings from/to Perl's internal
429 Unicode form on read/write via the ":encoding()" layer.
433 File handles can be opened to "in memory" files held in Perl scalars via:
435 open($fh,'>', \$variable) || ...
439 Anonymous temporary files are available without need to
440 'use FileHandle' or other module via
442 open($fh,"+>", undef) || ...
444 That is a literal undef, not an undefined value.
450 The new interpreter threads ("ithreads" for short) implementation of
451 multithreading, by Arthur Bergman, replaces the old "5.005 threads"
452 implementation. In the ithreads model any data sharing between
453 threads must be explicit, as opposed to the model where data sharing
454 was implicit. See L<threads> and L<threads::shared>, and
457 As a part of the ithreads implementation Perl will also use
458 any necessary and detectable reentrant libc interfaces.
460 =head2 Restricted Hashes
462 A restricted hash is restricted to a certain set of keys, no keys
463 outside the set can be added. Also individual keys can be restricted
464 so that the key cannot be deleted and the value cannot be changed.
465 No new syntax is involved: the Hash::Util module is the interface.
469 Perl used to be fragile in that signals arriving at inopportune moments
470 could corrupt Perl's internal state. Now Perl postpones handling of
471 signals until it's safe (between opcodes).
473 This change may have surprising side effects because signals no longer
474 interrupt Perl instantly. Perl will now first finish whatever it was
475 doing, like finishing an internal operation (like sort()) or an
476 external operation (like an I/O operation), and only then look at any
477 arrived signals (and before starting the next operation). No more corrupt
478 internal state since the current operation is always finished first,
479 but the signal may take more time to get heard. Note that breaking
480 out from potentially blocking operations should still work, though.
482 =head2 Understanding of Numbers
484 In general a lot of fixing has happened in the area of Perl's
485 understanding of numbers, both integer and floating point. Since in
486 many systems the standard number parsing functions like C<strtoul()>
487 and C<atof()> seem to have bugs, Perl tries to work around their
488 deficiencies. This results hopefully in more accurate numbers.
490 Perl now tries internally to use integer values in numeric conversions
491 and basic arithmetics (+ - * /) if the arguments are integers, and
492 tries also to keep the results stored internally as integers.
493 This change leads to often slightly faster and always less lossy
494 arithmetics. (Previously Perl always preferred floating point numbers
497 =head2 Arrays now always interpolate into double-quoted strings [561]
499 In double-quoted strings, arrays now interpolate, no matter what. The
500 behavior in earlier versions of perl 5 was that arrays would interpolate
501 into strings if the array had been mentioned before the string was
502 compiled, and otherwise Perl would raise a fatal compile-time error.
503 In versions 5.000 through 5.003, the error was
505 Literal @example now requires backslash
507 In versions 5.004_01 through 5.6.0, the error was
509 In string, @example now must be written as \@example
511 The idea here was to get people into the habit of writing
512 C<"fred\@example.com"> when they wanted a literal C<@> sign, just as
513 they have always written C<"Give me back my \$5"> when they wanted a
516 Starting with 5.6.1, when Perl now sees an C<@> sign in a
517 double-quoted string, it I<always> attempts to interpolate an array,
518 regardless of whether or not the array has been used or declared
519 already. The fatal error has been downgraded to an optional warning:
521 Possible unintended interpolation of @example in string
523 This warns you that C<"fred@example.com"> is going to turn into
524 C<fred.com> if you don't backslash the C<@>.
525 See http://www.plover.com/~mjd/perl/at-error.html for more details
526 about the history here.
528 =head2 Miscellaneous Changes
534 AUTOLOAD is now lvaluable, meaning that you can add the :lvalue attribute
535 to AUTOLOAD subroutines and you can assign to the AUTOLOAD return value.
539 The $Config{byteorder} (and corresponding BYTEORDER in config.h) was
540 previously wrong in platforms if sizeof(long) was 4, but sizeof(IV)
541 was 8. The byteorder was only sizeof(long) bytes long (1234 or 4321),
542 but now it is correctly sizeof(IV) bytes long, (12345678 or 87654321).
543 (This problem didn't affect Windows platforms.)
545 Also, $Config{byteorder} is now computed dynamically--this is more
546 robust with "fat binaries" where an executable image contains binaries
547 for more than one binary platform, and when cross-compiling.
551 C<perl -d:Module=arg,arg,arg> now works (previously one couldn't pass
552 in multiple arguments.)
556 C<do> followed by a bareword now ensures that this bareword isn't
557 a keyword (to avoid a bug where C<do q(foo.pl)> tried to call a
558 subroutine called C<q>). This means that for example instead of
559 C<do format()> you must write C<do &format()>.
563 The builtin dump() now gives an optional warning
564 C<dump() better written as CORE::dump()>,
565 meaning that by default C<dump(...)> is resolved as the builtin
566 dump() which dumps core and aborts, not as (possibly) user-defined
567 C<sub dump>. To call the latter, qualify the call as C<&dump(...)>.
568 (The whole dump() feature is to considered deprecated, and possibly
569 removed/changed in future releases.)
573 chomp() and chop() are now overridable. Note, however, that their
574 prototype (as given by C<prototype("CORE::chomp")> is undefined,
575 because it cannot be expressed and therefore one cannot really write
576 replacements to override these builtins.
580 END blocks are now run even if you exit/die in a BEGIN block.
581 Internally, the execution of END blocks is now controlled by
582 PL_exit_flags & PERL_EXIT_DESTRUCT_END. This enables the new
583 behaviour for Perl embedders. This will default in 5.10. See
588 Formats now support zero-padded decimal fields.
592 Although "you shouldn't do that", it was possible to write code that
593 depends on Perl's hashed key order (Data::Dumper does this). The new
594 algorithm "One-at-a-Time" produces a different hashed key order.
595 More details are in L</"Performance Enhancements">.
599 lstat(FILEHANDLE) now gives a warning because the operation makes no sense.
600 In future releases this may become a fatal error.
604 Spurious syntax errors generated in certain situations, when glob()
605 caused File::Glob to be loaded for the first time, have been fixed. [561]
609 Lvalue subroutines can now return C<undef> in list context. However,
610 the lvalue subroutine feature still remains experimental. [561+]
614 A lost warning "Can't declare ... dereference in my" has been
615 restored (Perl had it earlier but it became lost in later releases.)
619 A new special regular expression variable has been introduced:
620 C<$^N>, which contains the most-recently closed group (submatch).
624 C<no Module;> does not produce an error even if Module does not have an
625 unimport() method. This parallels the behavior of C<use> vis-a-vis
630 The numerical comparison operators return C<undef> if either operand
631 is a NaN. Previously the behaviour was unspecified.
635 C<our> can now have an experimental optional attribute C<unique> that
636 affects how global variables are shared among multiple interpreters,
641 The following builtin functions are now overridable: each(), keys(),
642 pop(), push(), shift(), splice(), unshift(). [561]
646 C<pack() / unpack()> can now group template letters with C<()> and then
647 apply repetition/count modifiers on the groups.
651 C<pack() / unpack()> can now process the Perl internal numeric types:
652 IVs, UVs, NVs-- and also long doubles, if supported by the platform.
653 The template letters are C<j>, C<J>, C<F>, and C<D>.
657 C<pack('U0a*', ...)> can now be used to force a string to UTF8.
661 my __PACKAGE__ $obj now works. [561]
665 POSIX::sleep() now returns the number of I<unslept> seconds
666 (as the POSIX standard says), as opposed to CORE::sleep() which
667 returns the number of slept seconds.
671 The printf() and sprintf() now support parameter reordering using the
672 C<%\d+\$> and C<*\d+\$> syntaxes. For example
674 print "%2\$s %1\$s\n", "foo", "bar";
676 will print "bar foo\n". This feature helps in writing
677 internationalised software, and in general when the order
678 of the parameters can vary.
682 The (\&) prototype now works properly. [561]
686 prototype(\[$@%&]) is now available to implicitly create references
687 (useful for example if you want to emulate the tie() interface).
691 A new command-line option, C<-t> is available. It is the
692 little brother of C<-T>: instead of dying on taint violations,
693 lexical warnings are given. B<This is only meant as a temporary
694 debugging aid while securing the code of old legacy applications.
695 This is not a substitute for -T.>
699 In other taint news, the C<exec LIST> and C<system LIST> have now been
700 considered too risky (think C<exec @ARGV>: it can start any program
701 with any arguments), and now the said forms cause a warning under
702 lexical warnings. You should carefully launder the arguments to
703 guarantee their validity. In future releases of Perl the forms will
704 become fatal errors so consider starting laundering now.
708 Tied hash interfaces are now required to have the EXISTS and DELETE
709 methods (either own or inherited).
713 If tr/// is just counting characters, it doesn't attempt to
718 untie() will now call an UNTIE() hook if it exists. See L<perltie>
723 L<utime> now supports C<utime undef, undef, @files> to change the
724 file timestamps to the current time.
728 The rules for allowing underscores (underbars) in numeric constants
729 have been relaxed and simplified: now you can have an underscore
730 simply B<between digits>.
734 Rather than relying on C's argv[0] (which may not contain a full pathname)
735 where possible $^X is now set by asking the operating system.
736 (eg by reading F</proc/self/exe> on Linux, F</proc/curproc/file> on FreeBSD)
740 A new variable, C<${^TAINT}>, indicates whether taint mode is enabled.
744 You can now override the readline() builtin, and this overrides also
745 the <FILEHANDLE> angle bracket operator.
749 The command-line options -s and -F are now recognized on the shebang
754 Use of the C</c> match modifier without an accompanying C</g> modifier
755 elicits a new warning: C<Use of /c modifier is meaningless without /g>.
757 Use of C</c> in substitutions, even with C</g>, elicits
758 C<Use of /c modifier is meaningless in s///>.
760 Use of C</g> with C<split> elicits C<Use of /g modifier is meaningless
765 Support for the C<CLONE> special subroutine had been added.
766 With ithreads, when a new thread is created, all Perl data is cloned,
767 however non-Perl data cannot be cloned automatically. In C<CLONE> you
768 can do whatever you need to do, like for example handle the cloning of
769 non-Perl data, if necessary. C<CLONE> will be executed once for every
770 package that has it defined or inherited. It will be called in the
771 context of the new thread, so all modifications are made in the new area.
777 =head1 Modules and Pragmata
779 =head2 New Modules and Pragmata
785 C<Attribute::Handlers>, originally by Damian Conway and now maintained
786 by Arthur Bergman, allows a class to define attribute handlers.
789 use Attribute::Handlers;
790 sub Wolf :ATTR(SCALAR) { print "howl!\n" }
792 # later, in some package using or inheriting from MyPack...
794 my MyPack $Fluffy : Wolf; # the attribute handler Wolf will be called
796 Both variables and routines can have attribute handlers. Handlers can
797 be specific to type (SCALAR, ARRAY, HASH, or CODE), or specific to the
798 exact compilation phase (BEGIN, CHECK, INIT, or END).
799 See L<Attribute::Handlers>.
803 C<B::Concise>, by Stephen McCamant, is a new compiler backend for
804 walking the Perl syntax tree, printing concise info about ops.
805 The output is highly customisable. See L<B::Concise>. [561+]
809 The new bignum, bigint, and bigrat pragmas, by Tels, implement
810 transparent bignum support (using the Math::BigInt, Math::BigFloat,
811 and Math::BigRat backends).
815 C<Class::ISA>, by Sean Burke, is a module for reporting the search
816 path for a class's ISA tree. See L<Class::ISA>.
820 C<Cwd> now has a split personality: if possible, an XS extension is
821 used, (this will hopefully be faster, more secure, and more robust)
822 but if not possible, the familiar Perl implementation is used.
826 C<Devel::PPPort>, originally by Kenneth Albanowski and now
827 maintained by Paul Marquess, has been added. It is primarily used
828 by C<h2xs> to enhance portability of XS modules between different
829 versions of Perl. See L<Devel::PPPort>.
833 C<Digest>, frontend module for calculating digests (checksums), from
834 Gisle Aas, has been added. See L<Digest>.
838 C<Digest::MD5> for calculating MD5 digests (checksums) as defined in
839 RFC 1321, from Gisle Aas, has been added. See L<Digest::MD5>.
841 use Digest::MD5 'md5_hex';
843 $digest = md5_hex("Thirsty Camel");
845 print $digest, "\n"; # 01d19d9d2045e005c3f1b80e8b164de1
847 NOTE: the C<MD5> backward compatibility module is deliberately not
848 included since its further use is discouraged.
850 See also L<PerlIO::via::QuotedPrint>.
854 C<Encode>, originally by Nick Ing-Simmons and now maintained by Dan
855 Kogai, provides a mechanism to translate between different character
856 encodings. Support for Unicode, ISO-8859-1, and ASCII are compiled in
857 to the module. Several other encodings (like the rest of the
858 ISO-8859, CP*/Win*, Mac, KOI8-R, three variants EBCDIC, Chinese,
859 Japanese, and Korean encodings) are included and can be loaded at
860 runtime. (For space considerations, the largest Chinese encodings
861 have been separated into their own CPAN module, Encode::HanExtra,
862 which Encode will use if available). See L<Encode>.
864 Any encoding supported by Encode module is also available to the
865 ":encoding()" layer if PerlIO is used.
869 C<Hash::Util> is the interface to the new I<restricted hashes>
870 feature. (Implemented by Jeffrey Friedl, Nick Ing-Simmons, and
871 Michael Schwern.) See L<Hash::Util>.
875 C<I18N::Langinfo> can be used to query locale information.
876 See L<I18N::Langinfo>.
880 C<I18N::LangTags>, by Sean Burke, has functions for dealing with
881 RFC3066-style language tags. See L<I18N::LangTags>.
885 C<ExtUtils::Constant>, by Nicholas Clark, is a new tool for extension
886 writers for generating XS code to import C header constants.
887 See L<ExtUtils::Constant>.
891 C<Filter::Simple>, by Damian Conway, is an easy-to-use frontend to
892 Filter::Util::Call. See L<Filter::Simple>.
898 use Filter::Simple sub {
899 while (my ($from, $to) = splice @_, 0, 2) {
908 use MyFilter qr/red/ => 'green';
910 print "red\n"; # this code is filtered, will print "green\n"
911 print "bored\n"; # this code is filtered, will print "bogreen\n"
915 print "red\n"; # this code is not filtered, will print "red\n"
919 C<File::Temp>, by Tim Jenness, allows one to create temporary files
920 and directories in an easy, portable, and secure way. See L<File::Temp>.
925 C<Filter::Util::Call>, by Paul Marquess, provides you with the
926 framework to write I<source filters> in Perl. For most uses, the
927 frontend Filter::Simple is to be preferred. See L<Filter::Util::Call>.
931 C<if>, by Ilya Zakharevich, is a new pragma for conditional inclusion
936 L<libnet>, by Graham Barr, is a collection of perl5 modules related
937 to network programming. See L<Net::FTP>, L<Net::NNTP>, L<Net::Ping>
938 (not part of libnet, but related), L<Net::POP3>, L<Net::SMTP>,
941 Perl installation leaves libnet unconfigured; use F<libnetcfg>
946 C<List::Util>, by Graham Barr, is a selection of general-utility
947 list subroutines, such as sum(), min(), first(), and shuffle().
952 C<Locale::Constants>, C<Locale::Country>, C<Locale::Currency>
953 C<Locale::Language>, and L<Locale::Script>, by Neil Bowers, have
954 been added. They provide the codes for various locale standards, such
955 as "fr" for France, "usd" for US Dollar, and "ja" for Japanese.
959 $country = code2country('jp'); # $country gets 'Japan'
960 $code = country2code('Norway'); # $code gets 'no'
962 See L<Locale::Constants>, L<Locale::Country>, L<Locale::Currency>,
963 and L<Locale::Language>.
967 C<Locale::Maketext>, by Sean Burke, is a localization framework. See
968 L<Locale::Maketext>, and L<Locale::Maketext::TPJ13>. The latter is an
969 article about software localization, originally published in The Perl
970 Journal #13, and republished here with kind permission.
974 C<Math::BigRat> for big rational numbers, to accompany Math::BigInt and
975 Math::BigFloat, from Tels. See L<Math::BigRat>.
979 C<Memoize> can make your functions faster by trading space for time,
980 from Mark-Jason Dominus. See L<Memoize>.
984 C<MIME::Base64>, by Gisle Aas, allows you to encode data in base64,
985 as defined in RFC 2045 - I<MIME (Multipurpose Internet Mail
990 $encoded = encode_base64('Aladdin:open sesame');
991 $decoded = decode_base64($encoded);
993 print $encoded, "\n"; # "QWxhZGRpbjpvcGVuIHNlc2FtZQ=="
999 C<MIME::QuotedPrint>, by Gisle Aas, allows you to encode data
1000 in quoted-printable encoding, as defined in RFC 2045 - I<MIME
1001 (Multipurpose Internet Mail Extensions)>.
1003 use MIME::QuotedPrint;
1005 $encoded = encode_qp("Smiley in Unicode: \x{263a}");
1006 $decoded = decode_qp($encoded);
1008 print $encoded, "\n"; # "Smiley in Unicode: =263A"
1010 See also L<PerlIO::via::QuotedPrint>.
1014 C<NEXT>, by Damian Conway, is a pseudo-class for method redispatch.
1019 C<open> is a new pragma for setting the default I/O layers
1024 C<PerlIO::scalar>, by Nick Ing-Simmons, provides the implementation
1025 of IO to "in memory" Perl scalars as discussed above. It also serves
1026 as an example of a loadable PerlIO layer. Other future possibilities
1027 include PerlIO::Array and PerlIO::Code. See L<PerlIO::scalar>.
1031 C<PerlIO::via>, by Nick Ing-Simmons, acts as a PerlIO layer and wraps
1032 PerlIO layer functionality provided by a class (typically implemented
1037 C<PerlIO::via::QuotedPrint>, by Elizabeth Mattijsen, is an example
1038 of a C<PerlIO::via> class:
1040 use PerlIO::via::QuotedPrint;
1041 open($fh,">:via(QuotedPrint)",$path);
1043 This will automatically convert everything output to C<$fh> to
1044 Quoted-Printable. See L<PerlIO::via> and L<PerlIO::via::QuotedPrint>.
1048 C<Pod::ParseLink>, by Russ Allbery, has been added,
1049 to parse LZ<><> links in pods as described in the new
1054 C<Pod::Text::Overstrike>, by Joe Smith, has been added.
1055 It converts POD data to formatted overstrike text.
1056 See L<Pod::Text::Overstrike>. [561+]
1060 C<Scalar::Util> is a selection of general-utility scalar subroutines,
1061 such as blessed(), reftype(), and tainted(). See L<Scalar::Util>.
1065 C<sort> is a new pragma for controlling the behaviour of sort().
1069 C<Storable> gives persistence to Perl data structures by allowing the
1070 storage and retrieval of Perl data to and from files in a fast and
1071 compact binary format. Because in effect Storable does serialisation
1072 of Perl data structures, with it you can also clone deep, hierarchical
1073 datastructures. Storable was originally created by Raphael Manfredi,
1074 but it is now maintained by Abhijit Menon-Sen. Storable has been
1075 enhanced to understand the two new hash features, Unicode keys and
1076 restricted hashes. See L<Storable>.
1080 C<Switch>, by Damian Conway, has been added. Just by saying
1084 you have C<switch> and C<case> available in Perl.
1090 case 1 { print "number 1" }
1091 case "a" { print "string a" }
1092 case [1..10,42] { print "number in list" }
1093 case (@array) { print "number in list" }
1094 case /\w+/ { print "pattern" }
1095 case qr/\w+/ { print "pattern" }
1096 case (%hash) { print "entry in hash" }
1097 case (\%hash) { print "entry in hash" }
1098 case (\&sub) { print "arg to subroutine" }
1099 else { print "previous case not true" }
1106 C<Test::More>, by Michael Schwern, is yet another framework for writing
1107 test scripts, more extensive than Test::Simple. See L<Test::More>.
1111 C<Test::Simple>, by Michael Schwern, has basic utilities for writing
1112 tests. See L<Test::Simple>.
1116 C<Text::Balanced>, by Damian Conway, has been added, for extracting
1117 delimited text sequences from strings.
1119 use Text::Balanced 'extract_delimited';
1121 ($a, $b) = extract_delimited("'never say never', he never said", "'", '');
1123 $a will be "'never say never'", $b will be ', he never said'.
1125 In addition to extract_delimited(), there are also extract_bracketed(),
1126 extract_quotelike(), extract_codeblock(), extract_variable(),
1127 extract_tagged(), extract_multiple(), gen_delimited_pat(), and
1128 gen_extract_tagged(). With these, you can implement rather advanced
1129 parsing algorithms. See L<Text::Balanced>.
1133 C<threads>, by Arthur Bergman, is an interface to interpreter threads.
1134 Interpreter threads (ithreads) is the new thread model introduced in
1135 Perl 5.6 but only available as an internal interface for extension
1136 writers (and for Win32 Perl for C<fork()> emulation). See L<threads>,
1137 L<threads::shared>, and L<perlthrtut>.
1141 C<threads::shared>, by Arthur Bergman, allows data sharing for
1142 interpreter threads. See L<threads::shared>.
1146 C<Tie::File>, by Mark-Jason Dominus, associates a Perl array with the
1147 lines of a file. See L<Tie::File>.
1151 C<Tie::Memoize>, by Ilya Zakharevich, provides on-demand loaded hashes.
1152 See L<Tie::Memoize>.
1156 C<Tie::RefHash::Nestable>, by Edward Avis, allows storing hash
1157 references (unlike the standard Tie::RefHash) The module is contained
1158 within Tie::RefHash. See L<Tie::RefHash>.
1162 C<Time::HiRes>, by Douglas E. Wegscheid, provides high resolution
1163 timing (ualarm, usleep, and gettimeofday). See L<Time::HiRes>.
1167 C<Unicode::UCD> offers a querying interface to the Unicode Character
1168 Database. See L<Unicode::UCD>.
1172 C<Unicode::Collate>, by SADAHIRO Tomoyuki, implements the UCA
1173 (Unicode Collation Algorithm) for sorting Unicode strings.
1174 See L<Unicode::Collate>.
1178 C<Unicode::Normalize>, by SADAHIRO Tomoyuki, implements the various
1179 Unicode normalization forms. See L<Unicode::Normalize>.
1183 C<XS::APItest>, by Tim Jenness, is a test extension that exercises XS
1184 APIs. Currently only C<printf()> is tested: how to output various
1185 basic data types from XS.
1189 C<XS::Typemap>, by Tim Jenness, is a test extension that exercises
1190 XS typemaps. Nothing gets installed, but the code is worth studying
1191 for extension writers.
1195 =head2 Updated And Improved Modules and Pragmata
1201 The following independently supported modules have been updated to the
1202 newest versions from CPAN: CGI, CPAN, DB_File, File::Spec, File::Temp,
1203 Getopt::Long, Math::BigFloat, Math::BigInt, the podlators bundle
1204 (Pod::Man, Pod::Text), Pod::LaTeX [561+], Pod::Parser, Storable,
1205 Term::ANSIColor, Test, Text-Tabs+Wrap.
1209 attributes::reftype() now works on tied arguments.
1213 AutoLoader can now be disabled with C<no AutoLoader;>.
1217 B::Deparse has been significantly enhanced by Robin Houston. It can
1218 now deparse almost all of the standard test suite (so that the tests
1219 still succeed). There is a make target "test.deparse" for trying this
1224 Carp now has better interface documentation, and the @CARP_NOT
1225 interface has been added to get optional control over where errors
1226 are reported independently of @ISA, by Ben Tilly.
1230 Class::Struct can now define the classes in compile time.
1234 Class::Struct now assigns the array/hash element if the accessor
1235 is called with an array/hash element as the B<sole> argument.
1239 The return value of Cwd::fastcwd() is now tainted.
1243 Data::Dumper now has an option to sort hashes.
1247 Data::Dumper now has an option to dump code references
1252 DB_File now supports newer Berkeley DB versions, among
1257 Devel::Peek now has an interface for the Perl memory statistics
1258 (this works only if you are using perl's malloc, and if you have
1259 compiled with debugging).
1263 The English module can now be used without the infamous performance
1266 use English '-no_match_vars';
1268 (Assuming, of course, that you don't need the troublesome variables
1269 C<$`>, C<$&>, or C<$'>.) Also, introduced C<@LAST_MATCH_START> and
1270 C<@LAST_MATCH_END> English aliases for C<@-> and C<@+>.
1274 ExtUtils::MakeMaker has been significantly cleaned up and fixed.
1275 The enhanced version has also been backported to earlier releases
1276 of Perl and submitted to CPAN so that the earlier releases can
1281 The arguments of WriteMakefile() in Makefile.PL are now checked
1282 for sanity much more carefully than before. This may cause new
1283 warnings when modules are being installed. See L<ExtUtils::MakeMaker>
1288 ExtUtils::MakeMaker now uses File::Spec internally, which hopefully
1289 leads to better portability.
1293 Fcntl, Socket, and Sys::Syslog have been rewritten by Nicholas Clark
1294 to use the new-style constant dispatch section (see L<ExtUtils::Constant>).
1295 This means that they will be more robust and hopefully faster.
1299 File::Find now chdir()s correctly when chasing symbolic links. [561]
1303 File::Find now has pre- and post-processing callbacks. It also
1304 correctly changes directories when chasing symbolic links. Callbacks
1305 (naughtily) exiting with "next;" instead of "return;" now work.
1309 File::Find is now (again) reentrant. It also has been made
1314 The warnings issued by File::Find now belong to their own category.
1315 You can enable/disable them with C<use/no warnings 'File::Find';>.
1319 File::Glob::glob() has been renamed to File::Glob::bsd_glob()
1320 because the name clashes with the builtin glob(). The older
1321 name is still available for compatibility, but is deprecated. [561]
1325 File::Glob now supports C<GLOB_LIMIT> constant to limit the size of
1326 the returned list of filenames.
1330 IPC::Open3 now allows the use of numeric file descriptors.
1334 IO::Socket now has an atmark() method, which returns true if the socket
1335 is positioned at the out-of-band mark. The method is also exportable
1336 as a sockatmark() function.
1340 IO::Socket::INET failed to open the specified port if the service name
1341 was not known. It now correctly uses the supplied port number as is. [561]
1345 IO::Socket::INET has support for the ReusePort option (if your
1346 platform supports it). The Reuse option now has an alias, ReuseAddr.
1347 For clarity, you may want to prefer ReuseAddr.
1351 IO::Socket::INET now supports a value of zero for C<LocalPort>
1352 (usually meaning that the operating system will make one up.)
1356 'use lib' now works identically to @INC. Removing directories
1357 with 'no lib' now works.
1361 Math::BigFloat and Math::BigInt have undergone a full rewrite by Tels.
1362 They are now magnitudes faster, and they support various bignum
1363 libraries such as GMP and PARI as their backends.
1367 Math::Complex handles inf, NaN etc., better.
1371 Net::Ping has been considerably enhanced by Rob Brown: multihoming is
1372 now supported, Win32 functionality is better, there is now time
1373 measuring functionality (optionally high-resolution using
1374 Time::HiRes), and there is now "external" protocol which uses
1375 Net::Ping::External module which runs your external ping utility and
1376 parses the output. A version of Net::Ping::External is available in
1379 Note that some of the Net::Ping tests are disabled when running
1380 under the Perl distribution since one cannot assume one or more
1381 of the following: enabled echo port at localhost, full Internet
1382 connectivity, or sympathetic firewalls. You can set the environment
1383 variable PERL_TEST_Net_Ping to "1" (one) before running the Perl test
1384 suite to enable all the Net::Ping tests.
1388 POSIX::sigaction() is now much more flexible and robust.
1389 You can now install coderef handlers, 'DEFAULT', and 'IGNORE'
1390 handlers, installing new handlers was not atomic.
1394 In Safe, C<%INC> is now localised in a Safe compartment so that
1399 In SDBM_File on dosish platforms, some keys went missing because of
1400 lack of support for files with "holes". A workaround for the problem
1405 In Search::Dict one can now have a pre-processing hook for the
1406 lines being searched.
1410 The Shell module now has an OO interface.
1414 In Sys::Syslog there is now a failover mechanism that will go
1415 through alternative connection mechanisms until the message
1416 is successfully logged.
1420 The Test module has been significantly enhanced.
1424 Time::Local::timelocal() does not handle fractional seconds anymore.
1425 The rationale is that neither does localtime(), and timelocal() and
1426 localtime() are supposed to be inverses of each other.
1430 The vars pragma now supports declaring fully qualified variables.
1431 (Something that C<our()> does not and will not support.)
1435 The C<utf8::> name space (as in the pragma) provides various
1436 Perl-callable functions to provide low level access to Perl's
1437 internal Unicode representation. At the moment only length()
1438 has been implemented.
1442 =head1 Utility Changes
1448 Emacs perl mode (emacs/cperl-mode.el) has been updated to version
1453 F<emacs/e2ctags.pl> is now much faster.
1457 C<enc2xs> is a tool for people adding their own encodings to the
1462 C<h2ph> now supports C trigraphs.
1466 C<h2xs> now produces a template README.
1470 C<h2xs> now uses C<Devel::PPPort> for better portability between
1471 different versions of Perl.
1475 C<h2xs> uses the new L<ExtUtils::Constant|ExtUtils::Constant> module
1476 which will affect newly created extensions that define constants.
1477 Since the new code is more correct (if you have two constants where the
1478 first one is a prefix of the second one, the first constant B<never>
1479 got defined), less lossy (it uses integers for integer constant,
1480 as opposed to the old code that used floating point numbers even for
1481 integer constants), and slightly faster, you might want to consider
1482 regenerating your extension code (the new scheme makes regenerating
1483 easy). L<h2xs> now also supports C trigraphs.
1487 C<libnetcfg> has been added to configure libnet.
1491 C<perlbug> is now much more robust. It also sends the bug report to
1492 perl.org, not perl.com.
1496 C<perlcc> has been rewritten and its user interface (that is,
1497 command line) is much more like that of the UNIX C compiler, cc.
1498 (The perlbc tools has been removed. Use C<perlcc -B> instead.)
1499 B<Note that perlcc is still considered very experimental and
1504 C<perlivp> is a new Installation Verification Procedure utility
1505 for running any time after installing Perl.
1509 C<piconv> is an implementation of the character conversion utility
1510 C<iconv>, demonstrating the new Encode module.
1514 C<pod2html> now allows specifying a cache directory.
1518 C<pod2html> now produces XHTML 1.0.
1522 C<pod2html> now understands POD written using different line endings
1523 (PC-like CRLF versus UNIX-like LF versus MacClassic-like CR).
1527 C<s2p> has been completely rewritten in Perl. (It is in fact a full
1528 implementation of sed in Perl: you can use the sed functionality by
1529 using the C<psed> utility.)
1533 C<xsubpp> now understands POD documentation embedded in the *.xs
1538 C<xsubpp> now supports the OUT keyword.
1542 =head1 New Documentation
1548 perl56delta details the changes between the 5.005 release and the
1553 perlclib documents the internal replacements for standard C library
1554 functions. (Interesting only for extension writers and Perl core
1559 perldebtut is a Perl debugging tutorial. [561+]
1563 perlebcdic contains considerations for running Perl on EBCDIC
1568 perlintro is a gentle introduction to Perl.
1572 perliol documents the internals of PerlIO with layers.
1576 perlmodstyle is a style guide for writing modules.
1580 perlnewmod tells about writing and submitting a new module. [561+]
1584 perlpacktut is a pack() tutorial.
1588 perlpod has been rewritten to be clearer and to record the best
1589 practices gathered over the years.
1593 perlpodspec is a more formal specification of the pod format,
1594 mainly of interest for writers of pod applications, not to
1595 people writing in pod.
1599 perlretut is a regular expression tutorial. [561+]
1603 perlrequick is a regular expressions quick-start guide.
1604 Yes, much quicker than perlretut. [561]
1608 perltodo has been updated.
1612 perltootc has been renamed as perltooc (to not to conflict
1613 with perltoot in filesystems restricted to "8.3" names).
1617 perluniintro is an introduction to using Unicode in Perl.
1618 (perlunicode is more of a detailed reference and background
1623 perlutil explains the command line utilities packaged with the Perl
1624 distribution. [561+]
1628 The following platform-specific documents are available before
1629 the installation as README.I<platform>, and after the installation
1632 perlaix perlamiga perlapollo perlbeos perlbs2000
1633 perlce perlcygwin perldgux perldos perlepoc perlfreebsd perlhpux
1634 perlhurd perlirix perlmachten perlmacos perlmint perlmpeix
1635 perlnetware perlos2 perlos390 perlplan9 perlqnx perlsolaris
1636 perltru64 perluts perlvmesa perlvms perlvos perlwin32
1638 These documents usually detail one or more of the following subjects:
1639 configuring, building, testing, installing, and sometimes also using
1640 Perl on the said platform.
1642 Eastern Asian Perl users are now welcomed in their own languages:
1643 README.jp (Japanese), README.ko (Korean), README.cn (simplified
1644 Chinese) and README.tw (traditional Chinese), which are written in
1645 normal pod but encoded in EUC-JP, EUC-KR, EUC-CN and Big5. These
1646 will get installed as
1648 perljp perlko perlcn perltw
1654 The documentation for the POSIX-BC platform is called "BS2000", to avoid
1655 confusion with the Perl POSIX module.
1659 The documentation for the WinCE platform is called perlce (README.ce
1660 in the source code kit), to avoid confusion with the perlwin32
1661 documentation on 8.3-restricted filesystems.
1665 =head1 Performance Enhancements
1671 map() could get pathologically slow when the result list it generates
1672 is larger than the source list. The performance has been improved for
1673 common scenarios. [561]
1677 sort() is also fully reentrant, in the sense that the sort function
1678 can itself call sort(). This did not work reliably in previous
1683 sort() has been changed to use primarily mergesort internally as
1684 opposed to the earlier quicksort. For very small lists this may
1685 result in slightly slower sorting times, but in general the speedup
1686 should be at least 20%. Additional bonuses are that the worst case
1687 behaviour of sort() is now better (in computer science terms it now
1688 runs in time O(N log N), as opposed to quicksort's Theta(N**2)
1689 worst-case run time behaviour), and that sort() is now stable
1690 (meaning that elements with identical keys will stay ordered as they
1691 were before the sort). See the C<sort> pragma for information.
1693 The story in more detail: suppose you want to serve yourself a little
1696 @digits = ( 3,1,4,1,5,9 );
1698 A numerical sort of the digits will yield (1,1,3,4,5,9), as expected.
1699 Which C<1> comes first is hard to know, since one C<1> looks pretty
1700 much like any other. You can regard this as totally trivial,
1701 or somewhat profound. However, if you just want to sort the even
1702 digits ahead of the odd ones, then what will
1704 sort { ($a % 2) <=> ($b % 2) } @digits;
1706 yield? The only even digit, C<4>, will come first. But how about
1707 the odd numbers, which all compare equal? With the quicksort algorithm
1708 used to implement Perl 5.6 and earlier, the order of ties is left up
1709 to the sort. So, as you add more and more digits of Pi, the order
1710 in which the sorted even and odd digits appear will change.
1711 and, for sufficiently large slices of Pi, the quicksort algorithm
1712 in Perl 5.8 won't return the same results even if reinvoked with the
1713 same input. The justification for this rests with quicksort's
1714 worst case behavior. If you run
1716 sort { $a <=> $b } ( 1 .. $N , 1 .. $N );
1718 (something you might approximate if you wanted to merge two sorted
1719 arrays using sort), doubling $N doesn't just double the quicksort time,
1720 it I<quadruples> it. Quicksort has a worst case run time that can
1721 grow like N**2, so-called I<quadratic> behaviour, and it can happen
1722 on patterns that may well arise in normal use. You won't notice this
1723 for small arrays, but you I<will> notice it with larger arrays,
1724 and you may not live long enough for the sort to complete on arrays
1725 of a million elements. So the 5.8 quicksort scrambles large arrays
1726 before sorting them, as a statistical defence against quadratic behaviour.
1727 But that means if you sort the same large array twice, ties may be
1728 broken in different ways.
1730 Because of the unpredictability of tie-breaking order, and the quadratic
1731 worst-case behaviour, quicksort was I<almost> replaced completely with
1732 a stable mergesort. I<Stable> means that ties are broken to preserve
1733 the original order of appearance in the input array. So
1735 sort { ($a % 2) <=> ($b % 2) } (3,1,4,1,5,9);
1737 will yield (4,3,1,1,5,9), guaranteed. The even and odd numbers
1738 appear in the output in the same order they appeared in the input.
1739 Mergesort has worst case O(N log N) behaviour, the best value
1740 attainable. And, ironically, this mergesort does particularly
1741 well where quicksort goes quadratic: mergesort sorts (1..$N, 1..$N)
1742 in O(N) time. But quicksort was rescued at the last moment because
1743 it is faster than mergesort on certain inputs and platforms.
1744 For example, if you really I<don't> care about the order of even
1745 and odd digits, quicksort will run in O(N) time; it's very good
1746 at sorting many repetitions of a small number of distinct elements.
1747 The quicksort divide and conquer strategy works well on platforms
1748 with relatively small, very fast, caches. Eventually, the problem gets
1749 whittled down to one that fits in the cache, from which point it
1750 benefits from the increased memory speed.
1752 Quicksort was rescued by implementing a sort pragma to control aspects
1753 of the sort. The B<stable> subpragma forces stable behaviour,
1754 regardless of algorithm. The B<_quicksort> and B<_mergesort>
1755 subpragmas are heavy-handed ways to select the underlying implementation.
1756 The leading C<_> is a reminder that these subpragmas may not survive
1757 beyond 5.8. More appropriate mechanisms for selecting the implementation
1758 exist, but they wouldn't have arrived in time to save quicksort.
1762 Hashes now use Bob Jenkins "One-at-a-Time" hashing key algorithm
1763 ( http://burtleburtle.net/bob/hash/doobs.html ). This algorithm is
1764 reasonably fast while producing a much better spread of values than
1765 the old hashing algorithm (originally by Chris Torek, later tweaked by
1766 Ilya Zakharevich). Hash values output from the algorithm on a hash of
1767 all 3-char printable ASCII keys comes much closer to passing the
1768 DIEHARD random number generation tests. According to perlbench, this
1769 change has not affected the overall speed of Perl.
1773 unshift() should now be noticeably faster.
1777 =head1 Installation and Configuration Improvements
1779 =head2 Generic Improvements
1785 INSTALL now explains how you can configure Perl to use 64-bit
1786 integers even on non-64-bit platforms.
1790 Policy.sh policy change: if you are reusing a Policy.sh file
1791 (see INSTALL) and you use Configure -Dprefix=/foo/bar and in the old
1792 Policy $prefix eq $siteprefix and $prefix eq $vendorprefix, all of
1793 them will now be changed to the new prefix, /foo/bar. (Previously
1794 only $prefix changed.) If you do not like this new behaviour,
1795 specify prefix, siteprefix, and vendorprefix explicitly.
1799 A new optional location for Perl libraries, otherlibdirs, is available.
1800 It can be used for example for vendor add-ons without disturbing Perl's
1801 own library directories.
1805 In many platforms, the vendor-supplied 'cc' is too stripped-down to
1806 build Perl (basically, 'cc' doesn't do ANSI C). If this seems
1807 to be the case and 'cc' does not seem to be the GNU C compiler
1808 'gcc', an automatic attempt is made to find and use 'gcc' instead.
1812 gcc needs to closely track the operating system release to avoid
1813 build problems. If Configure finds that gcc was built for a different
1814 operating system release than is running, it now gives a clearly visible
1815 warning that there may be trouble ahead.
1819 Since Perl 5.8 is not binary-compatible with previous releases
1820 of Perl, Configure no longer suggests including the 5.005
1825 Configure C<-S> can now run non-interactively. [561]
1829 Configure support for pdp11-style memory models has been removed due
1830 to obsolescence. [561]
1834 configure.gnu now works with options with whitespace in them.
1838 installperl now outputs everything to STDERR.
1842 Because PerlIO is now the default on most platforms, "-perlio" doesn't
1843 get appended to the $Config{archname} (also known as $^O) anymore.
1844 Instead, if you explicitly choose not to use perlio (Configure command
1845 line option -Uuseperlio), you will get "-stdio" appended.
1849 Another change related to the architecture name is that "-64all"
1850 (-Duse64bitall, or "maximally 64-bit") is appended only if your
1851 pointers are 64 bits wide. (To be exact, the use64bitall is ignored.)
1855 In AFS installations, one can configure the root of the AFS to be
1856 somewhere else than the default F</afs> by using the Configure
1857 parameter C<-Dafsroot=/some/where/else>.
1861 APPLLIB_EXP, a lesser-known configuration-time definition, has been
1862 documented. It can be used to prepend site-specific directories
1863 to Perl's default search path (@INC); see INSTALL for information.
1867 The version of Berkeley DB used when the Perl (and, presumably, the
1868 DB_File extension) was built is now available as
1869 C<@Config{qw(db_version_major db_version_minor db_version_patch)}>
1870 from Perl and as C<DB_VERSION_MAJOR_CFG DB_VERSION_MINOR_CFG
1871 DB_VERSION_PATCH_CFG> from C.
1875 Building Berkeley DB3 for compatibility modes for DB, NDBM, and ODBM
1876 has been documented in INSTALL.
1880 If you have CPAN access (either network or a local copy such as a
1881 CD-ROM) you can during specify extra modules to Configure to build and
1882 install with Perl using the -Dextras=... option. See INSTALL for
1887 In addition to config.over, a new override file, config.arch, is
1888 available. This file is supposed to be used by hints file writers
1889 for architecture-wide changes (as opposed to config.over which is
1890 for site-wide changes).
1894 If your file system supports symbolic links, you can build Perl outside
1895 of the source directory by
1897 mkdir /tmp/perl/build/directory
1898 cd /tmp/perl/build/directory
1899 sh /path/to/perl/source/Configure -Dmksymlinks ...
1901 This will create in /tmp/perl/build/directory a tree of symbolic links
1902 pointing to files in /path/to/perl/source. The original files are left
1903 unaffected. After Configure has finished, you can just say
1907 and Perl will be built and tested, all in /tmp/perl/build/directory.
1912 For Perl developers, several new make targets for profiling
1913 and debugging have been added; see L<perlhack>.
1919 Use of the F<gprof> tool to profile Perl has been documented in
1920 L<perlhack>. There is a make target called "perl.gprof" for
1921 generating a gprofiled Perl executable.
1925 If you have GCC 3, there is a make target called "perl.gcov" for
1926 creating a gcoved Perl executable for coverage analysis. See
1931 If you are on IRIX or Tru64 platforms, new profiling/debugging options
1932 have been added; see L<perlhack> for more information about pixie and
1939 Guidelines of how to construct minimal Perl installations have
1940 been added to INSTALL.
1944 The Thread extension is now not built at all under ithreads
1945 (C<Configure -Duseithreads>) because it wouldn't work anyway (the
1946 Thread extension requires being Configured with C<-Duse5005threads>).
1948 B<Note that the 5.005 threads are unsupported and deprecated: if you
1949 have code written for the old threads you should migrate it to the
1950 new ithreads model.>
1954 The Gconvert macro ($Config{d_Gconvert}) used by perl for stringifying
1955 floating-point numbers is now more picky about using sprintf %.*g
1956 rules for the conversion. Some platforms that used to use gcvt may
1957 now resort to the slower sprintf.
1961 The obsolete method of making a special (e.g., debugging) flavor
1964 make LIBPERL=libperld.a
1966 has been removed. Use -DDEBUGGING instead.
1970 =head2 New Or Improved Platforms
1972 For the list of platforms known to support Perl,
1973 see L<perlport/"Supported Platforms">.
1979 AIX dynamic loading should be now better supported.
1983 AIX should now work better with gcc, threads, and 64-bitness. Also the
1984 long doubles support in AIX should be better now. See L<perlaix>.
1988 AtheOS ( http://www.atheos.cx/ ) is a new platform.
1992 BeOS has been reclaimed.
1996 The DG/UX platform now supports 5.005-style threads.
2001 The DYNIX/ptx platform (also known as dynixptx) is supported at or
2006 EBCDIC platforms (z/OS (also known as OS/390), POSIX-BC, and VM/ESA)
2007 have been regained. Many test suite tests still fail and the
2008 co-existence of Unicode and EBCDIC isn't quite settled, but the
2009 situation is much better than with Perl 5.6. See L<perlos390>,
2010 L<perlbs2000> (for POSIX-BC), and L<perlvmesa> for more information.
2014 Building perl with -Duseithreads or -Duse5005threads now works under
2015 HP-UX 10.20 (previously it only worked under 10.30 or later). You will
2016 need a thread library package installed. See README.hpux. [561]
2020 Mac OS Classic is now supported in the mainstream source package
2021 (MacPerl has of course been available since perl 5.004 but now the
2022 source code bases of standard Perl and MacPerl have been synchronised)
2027 Mac OS X (or Darwin) should now be able to build Perl even on HFS+
2028 filesystems. (The case-insensitivity used to confuse the Perl build
2033 NCR MP-RAS is now supported. [561]
2037 All the NetBSD specific patches (except for the installation
2038 specific ones) have been merged back to the main distribution.
2042 NetWare from Novell is now supported. See L<perlnetware>.
2046 NonStop-UX is now supported. [561]
2050 NEC SUPER-UX is now supported.
2054 All the OpenBSD specific patches (except for the installation
2055 specific ones) have been merged back to the main distribution.
2059 Perl has been tested with the GNU pth userlevel thread package
2060 ( http://www.gnu.org/software/pth/pth.html ). All thread tests
2061 of Perl now work, but not without adding some yield()s to the tests,
2062 so while pth (and other userlevel thread implementations) can be
2063 considered to be "working" with Perl ithreads, keep in mind the
2064 possible non-preemptability of the underlying thread implementation.
2068 Stratus VOS is now supported using Perl's native build method
2069 (Configure). This is the recommended method to build Perl on
2070 VOS. The older methods, which build miniperl, are still
2071 available. See L<perlvos>. [561+]
2075 The Amdahl UTS UNIX mainframe platform is now supported. [561]
2079 WinCE is now supported. See L<perlce>.
2083 z/OS (formerly known as OS/390, formerly known as MVS OE) now has
2084 support for dynamic loading. This is not selected by default,
2085 however, you must specify -Dusedl in the arguments of Configure. [561]
2089 =head1 Selected Bug Fixes
2091 Numerous memory leaks and uninitialized memory accesses have been
2092 hunted down. Most importantly, anonymous subs used to leak quite
2099 The autouse pragma didn't work for Multi::Part::Function::Names.
2103 caller() could cause core dumps in certain situations. Carp was
2104 sometimes affected by this problem. In particular, caller() now
2105 returns a subroutine name of C<(unknown)> for subroutines that have
2106 been removed from the symbol table.
2110 chop(@list) in list context returned the characters chopped in
2111 reverse order. This has been reversed to be in the right order. [561]
2115 Configure no longer includes the DBM libraries (dbm, gdbm, db, ndbm)
2116 when building the Perl binary. The only exception to this is SunOS 4.x,
2117 which needs them. [561]
2121 The behaviour of non-decimal but numeric string constants such as
2122 "0x23" was platform-dependent: in some platforms that was seen as 35,
2123 in some as 0, in some as a floating point number (don't ask). This
2124 was caused by Perl's using the operating system libraries in a situation
2125 where the result of the string to number conversion is undefined: now
2126 Perl consistently handles such strings as zero in numeric contexts.
2130 Several debugger fixes: exit code now reflects the script exit code,
2131 condition C<"0"> now treated correctly, the C<d> command now checks
2132 line number, C<$.> no longer gets corrupted, and all debugger output
2133 now goes correctly to the socket if RemotePort is set. [561]
2137 The debugger (perl5db.pl) has been modified to present a more
2138 consistent commands interface, via (CommandSet=580). perl5db.t was
2139 also added to test the changes, and as a placeholder for further tests.
2145 The debugger has a new C<dumpDepth> option to control the maximum
2146 depth to which nested structures are dumped. The C<x> command has
2147 been extended so that C<x N EXPR> dumps out the value of I<EXPR> to a
2148 depth of at most I<N> levels.
2152 The debugger can now show lexical variables if you have the CPAN
2153 module PadWalker installed.
2157 The order of DESTROYs has been made more predictable.
2161 Perl 5.6.0 could emit spurious warnings about redefinition of
2162 dl_error() when statically building extensions into perl.
2163 This has been corrected. [561]
2167 L<dprofpp> -R didn't work.
2171 C<*foo{FORMAT}> now works.
2175 Infinity is now recognized as a number.
2179 UNIVERSAL::isa no longer caches methods incorrectly. (This broke
2180 the Tk extension with 5.6.0.) [561]
2184 Lexicals I: lexicals outside an eval "" weren't resolved
2185 correctly inside a subroutine definition inside the eval "" if they
2186 were not already referenced in the top level of the eval""ed code.
2190 Lexicals II: lexicals leaked at file scope into subroutines that
2191 were declared before the lexicals.
2195 Lexical warnings now propagating correctly between scopes
2196 and into C<eval "...">.
2200 C<use warnings qw(FATAL all)> did not work as intended. This has been
2205 warnings::enabled() now reports the state of $^W correctly if the caller
2206 isn't using lexical warnings. [561]
2210 Line renumbering with eval and C<#line> now works. [561]
2214 Fixed numerous memory leaks, especially in eval "".
2218 Localised tied variables no longer leak memory
2221 tie my %tied_hash => 'Tie::StdHash';
2225 # Used to leak memory every time local() was called;
2226 # in a loop, this added up.
2227 local($tied_hash{Foo}) = 1;
2231 Localised hash elements (and %ENV) are correctly unlocalised to not
2232 exist, if they didn't before they were localised.
2236 tie my %tied_hash => 'Tie::StdHash';
2240 # Nothing has set the FOO element so far
2242 { local $tied_hash{FOO} = 'Bar' }
2244 # This used to print, but not now.
2245 print "exists!\n" if exists $tied_hash{FOO};
2247 As a side effect of this fix, tied hash interfaces B<must> define
2248 the EXISTS and DELETE methods.
2252 mkdir() now ignores trailing slashes in the directory name,
2253 as mandated by POSIX.
2257 Some versions of glibc have a broken modfl(). This affects builds
2258 with C<-Duselongdouble>. This version of Perl detects this brokenness
2259 and has a workaround for it. The glibc release 2.2.2 is known to have
2260 fixed the modfl() bug.
2264 Modulus of unsigned numbers now works (4063328477 % 65535 used to
2265 return 27406, instead of 27047). [561]
2269 Some "not a number" warnings introduced in 5.6.0 eliminated to be
2270 more compatible with 5.005. Infinity is now recognised as a number. [561]
2274 Numeric conversions did not recognize changes in the string value
2275 properly in certain circumstances. [561]
2279 Attributes (such as :shared) didn't work with our().
2283 our() variables will not cause bogus "Variable will not stay shared"
2288 "our" variables of the same name declared in two sibling blocks
2289 resulted in bogus warnings about "redeclaration" of the variables.
2290 The problem has been corrected. [561]
2294 pack "Z" now correctly terminates the string with "\0".
2298 Fix password routines which in some shadow password platforms
2299 (e.g. HP-UX) caused getpwent() to return every other entry.
2303 The PERL5OPT environment variable (for passing command line arguments
2304 to Perl) didn't work for more than a single group of options. [561]
2308 PERL5OPT with embedded spaces didn't work.
2312 printf() no longer resets the numeric locale to "C".
2316 C<qw(a\\b)> now parses correctly as C<'a\\b'>: that is, as three
2317 characters, not four. [561]
2321 pos() did not return the correct value within s///ge in earlier
2322 versions. This is now handled correctly. [561]
2326 Printing quads (64-bit integers) with printf/sprintf now works
2327 without the q L ll prefixes (assuming you are on a quad-capable platform).
2331 Regular expressions on references and overloaded scalars now work. [561+]
2335 Right-hand side magic (GMAGIC) could in many cases such as string
2336 concatenation be invoked too many times.
2340 scalar() now forces scalar context even when used in void context.
2344 SOCKS support is now much more robust.
2348 sort() arguments are now compiled in the right wantarray context
2349 (they were accidentally using the context of the sort() itself).
2350 The comparison block is now run in scalar context, and the arguments
2351 to be sorted are always provided list context. [561]
2355 Changed the POSIX character class C<[[:space:]]> to include the (very
2356 rarely used) vertical tab character. Added a new POSIX-ish character
2357 class C<[[:blank:]]> which stands for horizontal whitespace
2358 (currently, the space and the tab).
2362 The tainting behaviour of sprintf() has been rationalized. It does
2363 not taint the result of floating point formats anymore, making the
2364 behaviour consistent with that of string interpolation. [561]
2368 Some cases of inconsistent taint propagation (such as within hash
2369 values) have been fixed.
2373 The RE engine found in Perl 5.6.0 accidentally pessimised certain kinds
2374 of simple pattern matches. These are now handled better. [561]
2378 Regular expression debug output (whether through C<use re 'debug'>
2379 or via C<-Dr>) now looks better. [561]
2383 Multi-line matches like C<"a\nxb\n" =~ /(?!\A)x/m> were flawed. The
2384 bug has been fixed. [561]
2388 Use of $& could trigger a core dump under some situations. This
2389 is now avoided. [561]
2393 The regular expression captured submatches ($1, $2, ...) are now
2394 more consistently unset if the match fails, instead of leaving false
2395 data lying around in them. [561]
2399 readline() on files opened in "slurp" mode could return an extra
2400 "" (blank line) at the end in certain situations. This has been
2405 Autovivification of symbolic references of special variables described
2406 in L<perlvar> (as in C<${$num}>) was accidentally disabled. This works
2411 Sys::Syslog ignored the C<LOG_AUTH> constant.
2415 $AUTOLOAD, sort(), lock(), and spawning subprocesses
2416 in multiple threads simultaneously are now thread-safe.
2420 Tie::Array's SPLICE method was broken.
2424 Allow a read-only string on the left-hand side of a non-modifying tr///.
2428 If C<STDERR> is tied, warnings caused by C<warn> and C<die> now
2429 correctly pass to it.
2433 Several Unicode fixes.
2439 BOMs (byte order marks) at the beginning of Perl files
2440 (scripts, modules) should now be transparently skipped.
2441 UTF-16 and UCS-2 encoded Perl files should now be read correctly.
2445 The character tables have been updated to Unicode 3.2.0.
2449 Comparing with utf8 data does not magically upgrade non-utf8 data
2450 into utf8. (This was a problem for example if you were mixing data
2451 from I/O and Unicode data: your output might have got magically encoded
2456 Generating illegal Unicode code points such as U+FFFE, or the UTF-16
2457 surrogates, now also generates an optional warning.
2461 C<IsAlnum>, C<IsAlpha>, and C<IsWord> now match titlecase.
2465 Concatenation with the C<.> operator or via variable interpolation,
2466 C<eq>, C<substr>, C<reverse>, C<quotemeta>, the C<x> operator,
2467 substitution with C<s///>, single-quoted UTF8, should now work.
2471 The C<tr///> operator now works. Note that the C<tr///CU>
2472 functionality has been removed (but see pack('U0', ...)).
2476 C<eval "v200"> now works.
2480 Perl 5.6.0 parsed m/\x{ab}/ incorrectly, leading to spurious warnings.
2481 This has been corrected. [561]
2485 Zero entries were missing from the Unicode classes such as C<IsDigit>.
2491 Large unsigned numbers (those above 2**31) could sometimes lose their
2492 unsignedness, causing bogus results in arithmetic operations. [561]
2496 The Perl parser has been stress tested using both random input and
2497 Markov chain input and the few found crashes and lockups have been
2502 =head2 Platform Specific Changes and Fixes
2510 Perl now works on post-4.0 BSD/OSes.
2516 Setting C<$0> now works (as much as possible; see L<perlvar> for details).
2522 Numerous updates; currently synchronised with Cygwin 1.3.10.
2526 Previously DYNIX/ptx had problems in its Configure probe for non-blocking I/O.
2532 EPOC now better supported. See README.epoc. [561]
2538 Perl now works on post-3.0 FreeBSDs.
2544 README.hpux updated; C<Configure -Duse64bitall> now works;
2545 now uses HP-UX malloc instead of Perl malloc.
2551 Numerous compilation flag and hint enhancements; accidental mixing
2552 of 32-bit and 64-bit libraries (a doomed attempt) made much harder.
2562 Long doubles should now work (see INSTALL). [561]
2566 Linux previously had problems related to sockaddrlen when using
2567 accept(), recvfrom() (in Perl: recv()), getpeername(), and
2576 Compilation of the standard Perl distribution in Mac OS Classic should
2577 now work if you have the Metrowerks development environment and the
2578 missing Mac-specific toolkit bits. Contact the macperl mailing list
2585 MPE/iX update after Perl 5.6.0. See README.mpeix. [561]
2589 NetBSD/threads: try installing the GNU pth (should be in the
2590 packages collection, or http://www.gnu.org/software/pth/),
2591 and Configure with -Duseithreads.
2597 Perl now works on NetBSD/sparc.
2603 Now works with usethreads (see INSTALL). [561]
2609 64-bitness using the Sun Workshop compiler now works.
2615 The native build method requires at least VOS Release 14.5.0
2616 and GNU C++/GNU Tools 2.0.1 or later. The Perl pack function
2617 now maps overflowed values to +infinity and underflowed values
2622 Tru64 (aka Digital UNIX, aka DEC OSF/1)
2624 The operating system version letter now recorded in $Config{osvers}.
2625 Allow compiling with gcc (previously explicitly forbidden). Compiling
2626 with gcc still not recommended because buggy code results, even with
2633 Fixed various alignment problems that lead into core dumps either
2634 during build or later; no longer dies on math errors at runtime;
2635 now using full quad integers (64 bits), previously was using
2636 only 46 bit integers for speed.
2642 See L</"Socket Extension Dynamic in VMS"> and L</"IEEE-format Floating Point
2643 Default on OpenVMS Alpha"> for important changes not otherwise listed here.
2645 chdir() now works better despite a CRT bug; now works with MULTIPLICITY
2646 (see INSTALL); now works with Perl's malloc.
2648 The tainting of C<%ENV> elements via C<keys> or C<values> was previously
2649 unimplemented. It now works as documented.
2651 The C<waitpid> emulation has been improved. The worst bug (now fixed)
2652 was that a pid of -1 would cause a wildcard search of all processes on
2655 POSIX-style signals are now emulated much better on VMS versions prior
2658 The C<system> function and backticks operator have improved
2659 functionality and better error handling. [561]
2661 File access tests now use current process privileges rather than the
2662 user's default privileges, which could sometimes result in a mismatch
2663 between reported access and actual access. This improvement is only
2664 available on VMS v6.0 and later.
2666 There is a new C<kill> implementation based on C<sys$sigprc> that allows
2667 older VMS systems (pre-7.0) to use C<kill> to send signals rather than
2668 simply force exit. This implementation also allows later systems to
2669 call C<kill> from within a signal handler.
2671 Iterative logical name translations are now limited to 10 iterations in
2672 imitation of SHOW LOGICAL and other OpenVMS facilities.
2682 Signal handling now works better than it used to. It is now implemented
2683 using a Windows message loop, and is therefore less prone to random
2688 fork() emulation is now more robust, but still continues to have a few
2689 esoteric bugs and caveats. See L<perlfork> for details. [561+]
2693 A failed (pseudo)fork now returns undef and sets errno to EAGAIN. [561]
2697 The following modules now work on Windows:
2699 ExtUtils::Embed [561]
2706 IO::File::new_tmpfile() is no longer limited to 32767 invocations
2711 Better chdir() return value for a non-existent directory.
2715 Compiling perl using the 64-bit Platform SDK tools is now supported.
2719 The Win32::SetChildShowWindow() builtin can be used to control the
2720 visibility of windows created by child processes. See L<Win32> for
2725 Non-blocking waits for child processes (or pseudo-processes) are
2726 supported via C<waitpid($pid, &POSIX::WNOHANG)>.
2730 The behavior of system() with multiple arguments has been rationalized.
2731 Each unquoted argument will be automatically quoted to protect whitespace,
2732 and any existing whitespace in the arguments will be preserved. This
2733 improves the portability of system(@args) by avoiding the need for
2734 Windows C<cmd> shell specific quoting in perl programs.
2736 Note that this means that some scripts that may have relied on earlier
2737 buggy behavior may no longer work correctly. For example,
2738 C<system("nmake /nologo", @args)> will now attempt to run the file
2739 C<nmake /nologo> and will fail when such a file isn't found.
2740 On the other hand, perl will now execute code such as
2741 C<system("c:/Program Files/MyApp/foo.exe", @args)> correctly.
2745 The perl header files no longer suppress common warnings from the
2746 Microsoft Visual C++ compiler. This means that additional warnings may
2747 now show up when compiling XS code.
2751 Borland C++ v5.5 is now a supported compiler that can build Perl.
2752 However, the generated binaries continue to be incompatible with those
2753 generated by the other supported compilers (GCC and Visual C++). [561]
2757 Duping socket handles with open(F, ">&MYSOCK") now works under Windows 9x.
2762 Current directory entries in %ENV are now correctly propagated to child
2767 New %ENV entries now propagate to subprocesses. [561]
2771 Win32::GetCwd() correctly returns C:\ instead of C: when at the drive root.
2772 Other bugs in chdir() and Cwd::cwd() have also been fixed. [561]
2776 The makefiles now default to the features enabled in ActiveState ActivePerl
2777 (a popular Win32 binary distribution). [561]
2781 HTML files will now be installed in c:\perl\html instead of
2782 c:\perl\lib\pod\html
2786 REG_EXPAND_SZ keys are now allowed in registry settings used by perl. [561]
2790 Can now send() from all threads, not just the first one. [561]
2794 ExtUtils::MakeMaker now uses $ENV{LIB} to search for libraries. [561]
2798 Less stack reserved per thread so that more threads can run
2799 concurrently. (Still 16M per thread.) [561]
2803 C<< File::Spec->tmpdir() >> now prefers C:/temp over /tmp
2804 (works better when perl is running as service).
2808 Better UNC path handling under ithreads. [561]
2812 wait(), waitpid(), and backticks now return the correct exit status
2813 under Windows 9x. [561]
2817 A socket handle leak in accept() has been fixed. [561]
2823 =head1 New or Changed Diagnostics
2825 Please see L<perldiag> for more details.
2831 Ambiguous range in the transliteration operator (like a-z-9) now
2836 Two new debugging options have been added: if you have compiled your
2837 Perl with debugging, you can use the -DT [561] and -DR options to trace
2838 tokenising and to add reference counts to displaying variables,
2843 The lexical warnings category "deprecated" is no longer a sub-category
2844 of the "syntax" category. It is now a top-level category in its own
2849 Unadorned dump() will now give a warning suggesting to
2850 use explicit CORE::dump() if that's what really is meant.
2854 The "Unrecognized escape" warning has been extended to include C<\8>,
2855 C<\9>, and C<\_>. There is no need to escape any of the C<\w> characters.
2859 All regular expression compilation error messages are now hopefully
2860 easier to understand both because the error message now comes before
2861 the failed regex and because the point of failure is now clearly
2862 marked by a C<E<lt>-- HERE> marker.
2866 Various I/O (and socket) functions like binmode(), close(), and so
2867 forth now more consistently warn if they are used illogically either
2868 on a yet unopened or on an already closed filehandle (or socket).
2872 Using lstat() on a filehandle now gives a warning. (It's a non-sensical
2877 The C<-M> and C<-m> options now warn if you didn't supply the module name.
2881 If you in C<use> specify a required minimum version, modules matching
2882 the name and but not defining a $VERSION will cause a fatal failure.
2886 Using negative offset for vec() in lvalue context is now a warnable offense.
2890 Odd number of arguments to oveload::constant now elicits a warning.
2894 Odd number of elements to in anonymous hash now elicits a warning.
2898 The various "opened only for", "on closed", "never opened" warnings
2899 drop the C<main::> prefix for filehandles in the C<main> package,
2900 for example C<STDIN> instead of C<main::STDIN>.
2904 Subroutine prototypes are now checked more carefully, you may
2905 get warnings for example if you have used non-prototype characters.
2909 If an attempt to use a (non-blessed) reference as an array index
2910 is made, a warning is given.
2914 C<push @a;> and C<unshift @a;> (with no values to push or unshift)
2915 now give a warning. This may be a problem for generated and evaled
2920 If you try to L<perlfunc/pack> a number less than 0 or larger than 255
2921 using the C<"C"> format you will get an optional warning. Similarly
2922 for the C<"c"> format and a number less than -128 or more than 127.
2926 pack C<P> format now demands an explicit size.
2930 unpack C<w> now warns of unterminated compressed integers.
2934 Warnings relating to the use of PerlIO have been added.
2938 Certain regex modifiers such as C<(?o)> make sense only if applied to
2939 the entire regex. You will get an optional warning if you try to do
2944 Variable length lookbehind has not yet been implemented, trying to
2945 use it will tell that.
2949 Using arrays or hashes as references (e.g. C<< %foo->{bar} >>
2950 has been deprecated for a while. Now you will get an optional warning.
2954 Warnings relating to the use of the new restricted hashes feature
2959 Self-ties of arrays and hashes are not supported and fatal errors
2960 will happen even at an attempt to do so.
2964 Using C<sort> in scalar context now issues an optional warning.
2965 This didn't do anything useful, as the sort was not performed.
2969 Using the /g modifier in split() is meaningless and will cause a warning.
2973 Using splice() past the end of an array now causes a warning.
2977 Malformed Unicode encodings (UTF-8 and UTF-16) cause a lot of warnings,
2978 ad doestrying to use UTF-16 surrogates (which are unimplemented).
2982 Trying to use Unicode characters on an I/O stream without marking the
2983 stream's encoding (using open() or binmode()) will cause "Wide character"
2988 Use of v-strings in use/require causes a (backward) portability warning.
2992 Warnings relating to the use interpreter threads and their shared data
2997 =head1 Changed Internals
3003 PerlIO is now the default.
3007 perlapi.pod (a companion to perlguts) now attempts to document the
3012 You can now build a really minimal perl called microperl.
3013 Building microperl does not require even running Configure;
3014 C<make -f Makefile.micro> should be enough. Beware: microperl makes
3015 many assumptions, some of which may be too bold; the resulting
3016 executable may crash or otherwise misbehave in wondrous ways.
3017 For careful hackers only.
3021 Added rsignal(), whichsig(), do_join(), op_clear, op_null,
3022 ptr_table_clear(), ptr_table_free(), sv_setref_uv(), and several UTF-8
3023 interfaces to the publicised API. For the full list of the available
3024 APIs see L<perlapi>.
3028 Made possible to propagate customised exceptions via croak()ing.
3032 Now xsubs can have attributes just like subs. (Well, at least the
3033 built-in attributes.)
3037 dTHR and djSP have been obsoleted; the former removed (because it's
3038 a no-op) and the latter replaced with dSP.
3042 PERL_OBJECT has been completely removed.
3046 The MAGIC constants (e.g. C<'P'>) have been macrofied
3047 (e.g. C<PERL_MAGIC_TIED>) for better source code readability
3048 and maintainability.
3052 The regex compiler now maintains a structure that identifies nodes in
3053 the compiled bytecode with the corresponding syntactic features of the
3054 original regex expression. The information is attached to the new
3055 C<offsets> member of the C<struct regexp>. See L<perldebguts> for more
3056 complete information.
3060 The C code has been made much more C<gcc -Wall> clean. Some warning
3061 messages still remain in some platforms, so if you are compiling with
3062 gcc you may see some warnings about dubious practices. The warnings
3063 are being worked on.
3067 F<perly.c>, F<sv.c>, and F<sv.h> have now been extensively commented.
3071 Documentation on how to use the Perl source repository has been added
3072 to F<Porting/repository.pod>.
3076 There are now several profiling make targets.
3080 =head1 Security Vulnerability Closed [561]
3082 (This change was already made in 5.7.0 but bears repeating here.)
3083 (5.7.0 came out before 5.6.1: the development branch 5.7 released
3084 earlier than the maintenance branch 5.6)
3086 A potential security vulnerability in the optional suidperl component
3087 of Perl was identified in August 2000. suidperl is neither built nor
3088 installed by default. As of November 2001 the only known vulnerable
3089 platform is Linux, most likely all Linux distributions. CERT and
3090 various vendors and distributors have been alerted about the vulnerability.
3091 See http://www.cpan.org/src/5.0/sperl-2000-08-05/sperl-2000-08-05.txt
3092 for more information.
3094 The problem was caused by Perl trying to report a suspected security
3095 exploit attempt using an external program, /bin/mail. On Linux
3096 platforms the /bin/mail program had an undocumented feature which
3097 when combined with suidperl gave access to a root shell, resulting in
3098 a serious compromise instead of reporting the exploit attempt. If you
3099 don't have /bin/mail, or if you have 'safe setuid scripts', or if
3100 suidperl is not installed, you are safe.
3102 The exploit attempt reporting feature has been completely removed from
3103 Perl 5.8.0 (and the maintenance release 5.6.1, and it was removed also
3104 from all the Perl 5.7 releases), so that particular vulnerability
3105 isn't there anymore. However, further security vulnerabilities are,
3106 unfortunately, always possible. The suidperl functionality is most
3107 probably going to be removed in Perl 5.10. In any case, suidperl
3108 should only be used by security experts who know exactly what they are
3109 doing and why they are using suidperl instead of some other solution
3110 such as sudo ( see http://www.courtesan.com/sudo/ ).
3114 Several new tests have been added, especially for the F<lib> and
3115 F<ext> subsections. There are now about 69 000 individual tests
3116 (spread over about 700 test scripts), in the regression suite (5.6.1
3117 has about 11 700 tests, in 258 test scripts) The exact numbers depend
3118 on the platform and Perl configuration used. Many of the new tests
3119 are of course introduced by the new modules, but still in general Perl
3120 is now more thoroughly tested.
3122 Because of the large number of tests, running the regression suite
3123 will take considerably longer time than it used to: expect the suite
3124 to take up to 4-5 times longer to run than in perl 5.6. On a really
3125 fast machine you can hope to finish the suite in about 6-8 minutes
3128 The tests are now reported in a different order than in earlier Perls.
3129 (This happens because the test scripts from under t/lib have been moved
3130 to be closer to the library/extension they are testing.)
3132 =head1 Known Problems
3134 =head2 The Compiler Suite Is Still Very Experimental
3136 The compiler suite is slowly getting better but it continues to be
3137 highly experimental. Use in production environments is discouraged.
3139 =head2 Localising Tied Arrays and Hashes Is Broken
3143 doesn't work as one would expect: the old value is restored
3144 incorrectly. This will be changed in a future release, but we don't
3145 know yet what the new semantics will exactly be. In any case, the
3146 change will break existing code that relies on the current
3147 (ill-defined) semantics, so just avoid doing this in general.
3149 =head2 Building Extensions Can Fail Because Of Largefiles
3151 Some extensions like mod_perl are known to have issues with
3152 `largefiles', a change brought by Perl 5.6.0 in which file offsets
3153 default to 64 bits wide, where supported. Modules may fail to compile
3154 at all, or they may compile and work incorrectly. Currently, there
3155 is no good solution for the problem, but Configure now provides
3156 appropriate non-largefile ccflags, ldflags, libswanted, and libs
3157 in the %Config hash (e.g., $Config{ccflags_nolargefiles}) so the
3158 extensions that are having problems can try configuring themselves
3159 without the largefileness. This is admittedly not a clean solution,
3160 and the solution may not even work at all. One potential failure is
3161 whether one can (or, if one can, whether it's a good idea to) link
3162 together at all binaries with different ideas about file offsets;
3163 all this is platform-dependent.
3165 =head2 Modifying $_ Inside for(..)
3169 works without complaint. It shouldn't. (You should be able to
3170 modify only lvalue elements inside the loops.) You can see the
3171 correct behaviour by replacing the 1..5 with 1, 2, 3, 4, 5.
3173 =head2 mod_perl 1.26 Doesn't Build With Threaded Perl
3175 Use mod_perl 1.27 or higher.
3177 =head2 lib/ftmp-security tests warn 'system possibly insecure'
3179 Don't panic. Read the 'make test' section of INSTALL instead.
3181 =head2 libwww-perl (LWP) fails base/date #51
3183 Use libwww-perl 5.65 or later.
3185 =head2 PDL failing some tests
3187 Use PDL 2.3.4 or later.
3191 You may get errors like 'Undefined symbol "Perl_get_sv"' or "can't
3192 resolve symbol 'Perl_get_sv'", or the symbol may be "Perl_sv_2pv".
3193 This probably means that you are trying to use an older shared Perl
3194 library (or extensions linked with such) with Perl 5.8.0 executable.
3195 Perl used to have such a subroutine, but that is no more the case.
3196 Check your shared library path, and any shared Perl libraries in those
3199 Sometimes this problem may also indicate a partial Perl 5.8.0
3200 installation, see L</"Mac OS X dyld undefined symbols"> for an
3201 example and how to deal with it.
3203 =head2 Self-tying Problems
3205 Self-tying of arrays and hashes is broken in rather deep and
3206 hard-to-fix ways. As a stop-gap measure to avoid people from getting
3207 frustrated at the mysterious results (core dumps, most often), it is
3208 forbidden for now (you will get a fatal error even from an attempt).
3210 A change to self-tying of globs has caused them to be recursively
3211 referenced (see: L<perlobj/"Two-Phased Garbage Collection">). You
3212 will now need an explicit untie to destroy a self-tied glob. This
3213 behaviour may be fixed at a later date.
3215 Self-tying of scalars and IO thingies works.
3217 =head2 ext/threads/t/libc
3219 If this test fails, it indicates that your libc (C library) is not
3220 threadsafe. This particular test stress tests the localtime() call to
3221 find out whether it is threadsafe. See L<perlthrtut> for more information.
3223 =head2 Failure of Thread (5.005-style) tests
3225 B<Note that support for 5.005-style threading is deprecated,
3226 experimental and practically unsupported. In 5.10, it is expected
3227 to be removed. You should migrate your code to ithreads.>
3229 The following tests are known to fail due to fundamental problems in
3230 the 5.005 threading implementation. These are not new failures--Perl
3231 5.005_0x has the same bugs, but didn't have these tests.
3233 ../ext/B/t/xref.t 255 65280 14 12 85.71% 3-14
3234 ../ext/List/Util/t/first.t 255 65280 7 4 57.14% 2 5-7
3235 ../lib/English.t 2 512 54 2 3.70% 2-3
3236 ../lib/FileCache.t 5 1 20.00% 5
3237 ../lib/Filter/Simple/t/data.t 6 3 50.00% 1-3
3238 ../lib/Filter/Simple/t/filter_only. 9 3 33.33% 1-2 5
3239 ../lib/Math/BigInt/t/bare_mbf.t 1627 4 0.25% 8 11 1626-1627
3240 ../lib/Math/BigInt/t/bigfltpm.t 1629 4 0.25% 10 13 1628-
3242 ../lib/Math/BigInt/t/sub_mbf.t 1633 4 0.24% 8 11 1632-1633
3243 ../lib/Math/BigInt/t/with_sub.t 1628 4 0.25% 9 12 1627-1628
3244 ../lib/Tie/File/t/31_autodefer.t 255 65280 65 32 49.23% 34-65
3245 ../lib/autouse.t 10 1 10.00% 4
3246 op/flip.t 15 1 6.67% 15
3248 These failures are unlikely to get fixed as 5.005-style threads
3249 are considered fundamentally broken. (Basically what happens is that
3250 competing threads can corrupt shared global state, one good example
3251 being regular expression engine's state.)
3253 =head2 Timing problems
3255 The following tests may fail intermittently because of timing
3256 problems, for example if the system is heavily loaded.
3259 ext/Time/HiRes/HiRes.t
3261 lib/Memoize/t/expmod_t.t
3262 lib/Memoize/t/speed.t
3264 In case of failure please try running them manually, for example
3266 ./perl -Ilib ext/Time/HiRes/HiRes.t
3268 =head2 Tied/Magical Array/Hash Elements Do Not Autovivify
3270 For normal arrays C<$foo = \$bar[1]> will assign C<undef> to
3271 C<$bar[1]> (assuming that it didn't exist before), but for
3272 tied/magical arrays and hashes such autovivification does not happen
3273 because there is currently no way to catch the reference creation.
3274 The same problem affects slicing over non-existent indices/keys of
3275 a tied/magical array/hash.
3277 =head2 Unicode in package/class and subroutine names does not work
3279 One can have Unicode in identifier names, but not in package/class or
3280 subroutine names. While some limited functionality towards this does
3281 exist as of Perl 5.8.0, that is more accidental than designed; use of
3282 Unicode for the said purposes is unsupported.
3284 One reason of this unfinishedness is its (currently) inherent
3285 unportability: since both package names and subroutine names may
3286 need to be mapped to file and directory names, the Unicode capability
3287 of the filesystem becomes important-- and there unfortunately aren't
3290 =head1 Platform Specific Problems
3298 If using the AIX native make command, instead of just "make" issue
3299 "make all". In some setups the former has been known to spuriously
3300 also try to run "make install". Alternatively, you may want to use
3305 In AIX 4.2, Perl extensions that use C++ functions that use statics
3306 may have problems in that the statics are not getting initialized.
3307 In newer AIX releases, this has been solved by linking Perl with
3308 the libC_r library, but unfortunately in AIX 4.2 the said library
3309 has an obscure bug where the various functions related to time
3310 (such as time() and gettimeofday()) return broken values, and
3311 therefore in AIX 4.2 Perl is not linked against libC_r.
3315 vac 5.0.0.0 May Produce Buggy Code For Perl
3317 The AIX C compiler vac version 5.0.0.0 may produce buggy code,
3318 resulting in a few random tests failing when run as part of "make
3319 test", but when the failing tests are run by hand, they succeed.
3320 We suggest upgrading to at least vac version 5.0.1.0, that has been
3321 known to compile Perl correctly. "lslpp -L|grep vac.C" will tell
3322 you the vac version. See README.aix.
3326 If building threaded Perl, you may get compilation warning from pp_sys.c:
3328 "pp_sys.c", line 4651.39: 1506-280 (W) Function argument assignment between types "unsigned char*" and "const void*" is not allowed.
3330 This is harmless; it is caused by the getnetbyaddr() and getnetbyaddr_r()
3331 having slightly different types for their first argument.
3335 =head2 Alpha systems with old gccs fail several tests
3337 If you see op/pack, op/pat, op/regexp, or ext/Storable tests failing
3338 in a Linux/alpha or *BSD/Alpha, it's probably time to upgrade your gcc.
3339 gccs prior to 2.95.3 are definitely not good enough, and gcc 3.1 may
3340 be even better. (RedHat Linux/alpha with gcc 3.1 reported no problems,
3341 as did Linux 2.4.18 with gcc 2.95.4.) (In Tru64, it is preferable to
3342 use the bundled C compiler.)
3346 Perl 5.8.0 doesn't build in AmigaOS. It broke at some point during
3347 the ithreads work and we could not find Amiga experts to unbreak the
3348 problems. Perl 5.6.1 still works for AmigaOS (as does the the 5.7.2
3349 development release).
3353 The following tests fail on 5.8.0 Perl in BeOS Personal 5.03:
3355 t/op/lfs............................FAILED at test 17
3356 t/op/magic..........................FAILED at test 24
3357 ext/Fcntl/t/syslfs..................FAILED at test 17
3358 ext/File/Glob/t/basic...............FAILED at test 3
3359 ext/POSIX/t/sigaction...............FAILED at test 13
3360 ext/POSIX/t/waitpid.................FAILED at test 1
3362 See L<perlbeos> (README.beos) for more details.
3364 =head2 Cygwin "unable to remap"
3366 For example when building the Tk extension for Cygwin,
3367 you may get an error message saying "unable to remap".
3368 This is known problem with Cygwin, and a workaround is
3369 detailed in here: http://sources.redhat.com/ml/cygwin/2001-12/msg00894.html
3371 =head2 Cygwin ndbm tests fail on FAT
3373 One can build but not install (or test the build of) the NDBM_File
3374 on FAT filesystems. Installation (or build) on NTFS works fine.
3375 If one attempts the test on a FAT install (or build) the following
3376 failures are expected:
3378 ../ext/NDBM_File/ndbm.t 13 3328 71 59 83.10% 1-2 4 16-71
3379 ../ext/ODBM_File/odbm.t 255 65280 ?? ?? % ??
3380 ../lib/AnyDBM_File.t 2 512 12 2 16.67% 1 4
3381 ../lib/Memoize/t/errors.t 0 139 11 5 45.45% 7-11
3382 ../lib/Memoize/t/tie_ndbm.t 13 3328 4 4 100.00% 1-4
3383 run/fresh_perl.t 97 1 1.03% 91
3385 NDBM_File fails and ODBM_File just coredumps.
3387 =head2 FreeBSD built with ithreads coredumps reading large directories
3389 This is a known bug in FreeBSD 4.5's readdir_r(), it has been fixed in
3390 FreeBSD 4.6 (see L<perlfreebsd> (README.freebsd)).
3392 =head2 FreeBSD Failing locale Test 117 For ISO 8859-15 Locales
3394 The ISO 8859-15 locales may fail the locale test 117 in FreeBSD.
3395 This is caused by the characters \xFF (y with diaeresis) and \xBE
3396 (Y with diaeresis) not behaving correctly when being matched
3397 case-insensitively. Apparently this problem has been fixed in
3398 the latest FreeBSD releases.
3399 ( http://www.freebsd.org/cgi/query-pr.cgi?pr=34308 )
3401 =head2 IRIX fails ext/List/Util/t/shuffle.t or Digest::MD5
3403 IRIX with MIPSpro 7.3.1.2m or 7.3.1.3m compiler may fail the List::Util
3404 test ext/List/Util/t/shuffle.t by dumping core. This seems to be
3405 a compiler error since if compiled with gcc no core dump ensues, and
3406 no failures have been seen on the said test on any other platform.
3408 Similarly, building the Digest::MD5 extension has been
3409 known to fail with "*** Termination code 139 (bu21)".
3411 The cure is to drop optimization level (Configure -Doptimize=-O2).
3413 =head2 HP-UX lib/posix Subtest 9 Fails When LP64-Configured
3415 If perl is configured with -Duse64bitall, the successful result of the
3416 subtest 10 of lib/posix may arrive before the successful result of the
3417 subtest 9, which confuses the test harness so much that it thinks the
3420 =head2 Linux with glibc 2.2.5 fails t/op/int subtest #6 with -Duse64bitint
3422 This is a known bug in the glibc 2.2.5 with long long integers.
3423 ( http://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=65612 )
3425 =head2 Linux With Sfio Fails op/misc Test 48
3431 Please remember to set your environment variable LC_ALL to "C"
3432 (setenv LC_ALL C) before running "make test" to avoid a lot of
3433 warnings about the broken locales of Mac OS X.
3435 The following tests are known to fail in Mac OS X 10.1.5 because of
3436 buggy (old) implementations of Berkeley DB included in Mac OS X:
3438 Failed Test Stat Wstat Total Fail Failed List of Failed
3439 -------------------------------------------------------------------------
3440 ../ext/DB_File/t/db-btree.t 0 11 ?? ?? % ??
3441 ../ext/DB_File/t/db-recno.t 149 3 2.01% 61 63 65
3443 If you are building on a UFS partition, you will also probably see
3444 t/op/stat.t subtest #9 fail. This is caused by Darwin's UFS not
3445 supporting inode change time.
3447 Also the ext/POSIX/t/posix.t subtest #10 fails but it is skipped for
3448 now because the failure is Apple's fault, not Perl's (blocked signals
3451 If you Configure with ithreads, ext/threads/t/libc.t will fail. Again,
3452 this is not Perl's fault-- the libc of Mac OS X is not threadsafe
3453 (in this particular test, the localtime() call is found to be
3456 =head2 Mac OS X dyld undefined symbols
3458 If after installing Perl 5.8.0 you are getting warnings about missing
3459 symbols, for example
3461 dyld: perl Undefined symbols
3465 you probably have an old pre-Perl-5.8.0 installation (or parts of one)
3466 in /Library/Perl (the undefined symbols used to exist in pre-5.8.0 Perls).
3467 It seems that for some reason "make install" doesn't always completely
3468 overwrite the files in /Library/Perl. You can move the old Perl
3469 shared library out of the way like this:
3471 cd /Library/Perl/darwin/CORE
3472 mv libperl.dylib libperlold.dylib
3474 and then reissue "make install". Note that the above of course is
3475 extremely disruptive for anything using the /usr/local/bin/perl.
3476 If that doesn't help, you may have to try removing all the .bundle
3477 files from beneath /Library/Perl, and again "make install"-ing.
3479 =head2 OS/2 Test Failures
3481 The following tests are known to fail on OS/2 (for clarity
3482 only the failures are shown, not the full error messages):
3484 t/io/utf8............................FAILED at test 19
3485 t/op/grent...........................FAILED at test 2
3486 t/op/pwent...........................FAILED at test 1
3487 t/lib/os2_base.......................FAILED at test 13
3488 t/lib/os2_process....................FAILED at test 10
3489 t/lib/os2_process_kid................FAILED at test 10
3490 t/lib/rx_cmprt.......................FAILED at test 16
3491 ext/DB_File/t/db-btree...............FAILED at test 0
3492 ext/DB_File/t/db-hash................FAILED at test 0
3493 ext/DB_File/t/db-recno...............FAILED at test 0
3494 lib/ExtUtils/t/basic.................FAILED at test 14
3495 lib/ExtUtils/t/Constant..............FAILED at test 4
3496 lib/Memoize/t/errors.................FAILED at test 4
3498 =head2 op/sprintf tests 91, 129, and 130
3500 The op/sprintf tests 91, 129, and 130 are known to fail on some platforms.
3501 Examples include any platform using sfio, and Compaq/Tandem's NonStop-UX.
3503 Test 91 is known to fail on QNX6 (nto), because C<sprintf '%e',0>
3504 incorrectly produces C<0.000000e+0> instead of C<0.000000e+00>.
3506 For tests 129 and 130, the failing platforms do not comply with
3507 the ANSI C Standard: lines 19ff on page 134 of ANSI X3.159 1989, to
3508 be exact. (They produce something other than "1" and "-1" when
3509 formatting 0.6 and -0.6 using the printf format "%.0f"; most often,
3510 they produce "0" and "-0".)
3514 In case you are still using Solaris 2.5 (aka SunOS 5.5), you may
3515 experience failures (the test core dumping) in lib/locale.t.
3516 The suggested cure is to upgrade your Solaris.
3518 =head2 Solaris x86 Fails Tests With -Duse64bitint
3520 The following tests are known to fail in Solaris x86 with Perl
3521 configured to use 64 bit integers:
3523 ext/Data/Dumper/t/dumper.............FAILED at test 268
3524 ext/Devel/Peek/Peek..................FAILED at test 7
3526 =head2 SUPER-UX (NEC SX)
3528 The following tests are known to fail on SUPER-UX:
3530 op/64bitint...........................FAILED tests 29-30, 32-33, 35-36
3531 op/arith..............................FAILED tests 128-130
3532 op/pack...............................FAILED tests 25-5625
3533 op/pow................................
3534 op/taint..............................# msgsnd failed
3535 ../ext/IO/lib/IO/t/io_poll............FAILED tests 3-4
3536 ../ext/IPC/SysV/ipcsysv...............FAILED tests 2, 5-6
3537 ../ext/IPC/SysV/t/msg.................FAILED tests 2, 4-6
3538 ../ext/Socket/socketpair..............FAILED tests 12
3539 ../lib/IPC/SysV.......................FAILED tests 2, 5-6
3540 ../lib/warnings.......................FAILED tests 115-116, 118-119
3542 The op/pack failure ("Cannot compress negative numbers at op/pack.t line 126")
3543 is serious but as of yet unsolved. It points at some problems with the
3544 signedness handling of the C compiler, as do the 64bitint, arith, and pow
3545 failures. Most of the rest point at problems with SysV IPC.
3547 =head2 Term::ReadKey not working on Win32
3549 Use Term::ReadKey 2.20 or later.
3557 During Configure, the test
3559 Guessing which symbols your C compiler and preprocessor define...
3561 will probably fail with error messages like
3563 CC-20 cc: ERROR File = try.c, Line = 3
3564 The identifier "bad" is undefined.
3566 bad switch yylook 79bad switch yylook 79bad switch yylook 79bad switch yylook 79#ifdef A29K
3569 CC-65 cc: ERROR File = try.c, Line = 3
3570 A semicolon is expected at this point.
3572 This is caused by a bug in the awk utility of UNICOS/mk. You can ignore
3573 the error, but it does cause a slight problem: you cannot fully
3574 benefit from the h2ph utility (see L<h2ph>) that can be used to
3575 convert C headers to Perl libraries, mainly used to be able to access
3576 from Perl the constants defined using C preprocessor, cpp. Because of
3577 the above error, parts of the converted headers will be invisible.
3578 Luckily, these days the need for h2ph is rare.
3582 If building Perl with interpreter threads (ithreads), the
3583 getgrent(), getgrnam(), and getgrgid() functions cannot return the
3584 list of the group members due to a bug in the multithreaded support of
3585 UNICOS/mk. What this means is that in list context the functions will
3586 return only three values, not four.
3592 There are a few known test failures, see L<perluts> (README.uts).
3594 =head2 VOS (Stratus)
3596 When Perl is built using the native build process on VOS Release
3597 14.5.0 and GNU C++/GNU Tools 2.0.1, all attempted tests either
3598 pass or result in TODO (ignored) failures.
3602 There should be no reported test failures with a default configuration,
3603 though there are a number of tests marked TODO that point to areas
3604 needing further debugging and/or porting work.
3608 In multi-CPU boxes, there are some problems with the I/O buffering:
3609 some output may appear twice.
3611 =head2 XML::Parser not working
3613 Use XML::Parser 2.31 or later.
3615 =head2 z/OS (OS/390)
3617 z/OS has rather many test failures but the situation is actually much
3618 better than it was in 5.6.0; it's just that so many new modules and
3619 tests have been added.
3621 Failed Test Stat Wstat Total Fail Failed List of Failed
3622 ---------------------------------------------------------------------------
3623 ../ext/Data/Dumper/t/dumper.t 357 8 2.24% 311 314 325 327
3625 ../ext/IO/lib/IO/t/io_unix.t 5 4 80.00% 2-5
3626 ../ext/Storable/t/downgrade.t 12 3072 169 12 7.10% 14-15 46-47 78-79
3628 ../lib/ExtUtils/t/Constant.t 121 30976 48 48 100.00% 1-48
3629 ../lib/ExtUtils/t/Embed.t 9 9 100.00% 1-9
3630 op/pat.t 922 7 0.76% 665 776 785 832-
3632 op/sprintf.t 224 3 1.34% 98 100 136
3633 op/tr.t 97 5 5.15% 63 71-74
3634 uni/fold.t 780 6 0.77% 61 169 196 661
3637 The failures in dumper.t and downgrade.t are problems in the tests,
3638 those in io_unix and sprintf are problems in the USS (UDP sockets and
3639 printf formats). The pat, tr, and fold failures are genuine Perl
3640 problems caused by EBCDIC (and in the pat and fold cases, combining
3641 that with Unicode). The Constant and Embed are probably problems in
3642 the tests (since they test Perl's ability to build extensions, and
3643 that seems to be working reasonably well.)
3645 =head2 Unicode Support on EBCDIC Still Spotty
3647 Though mostly working, Unicode support still has problem spots on
3648 EBCDIC platforms. One such known spot are the C<\p{}> and C<\P{}>
3649 regular expression constructs for code points less than 256: the
3650 C<pP> are testing for Unicode code points, not knowing about EBCDIC.
3652 =head2 Seen In Perl 5.7 But Gone Now
3654 C<Time::Piece> (previously known as C<Time::Object>) was removed
3655 because it was felt that it didn't have enough value in it to be a
3656 core module. It is still a useful module, though, and is available
3659 Perl 5.8 unfortunately does not build anymore on AmigaOS; this broke
3660 accidentally at some point. Since there are not that many Amiga
3661 developers available, we could not get this fixed and tested in time
3662 for 5.8.0. Perl 5.6.1 still works for AmigaOS (as does the the 5.7.2
3663 development release).
3665 The C<PerlIO::Scalar> and C<PerlIO::Via> (capitalised) were renamed as
3666 C<PerlIO::scalar> and C<PerlIO::via> (all lowercase) just before 5.8.0.
3667 The main rationale was to have all core PerlIO layers to have all
3668 lowercase names. The "plugins" are named as usual, for example
3669 C<PerlIO::via::QuotedPrint>.
3671 The C<threads::shared::queue> and C<threads::shared::semaphore> were
3672 renamed as C<Thread::Queue> and C<Thread::Semaphore> just before 5.8.0.
3673 The main rationale was to have thread modules to obey normal naming,
3674 C<Thread::> (the C<threads> and C<threads::shared> themselves are
3675 more pragma-like, they affect compile-time, so they stay lowercase).
3677 =head1 Reporting Bugs
3679 If you find what you think is a bug, you might check the articles
3680 recently posted to the comp.lang.perl.misc newsgroup and the perl
3681 bug database at http://bugs.perl.org/ . There may also be
3682 information at http://www.perl.com/ , the Perl Home Page.
3684 If you believe you have an unreported bug, please run the B<perlbug>
3685 program included with your release. Be sure to trim your bug down
3686 to a tiny but sufficient test case. Your bug report, along with the
3687 output of C<perl -V>, will be sent off to perlbug@perl.org to be
3688 analysed by the Perl porting team.
3692 The F<Changes> file for exhaustive details on what changed.
3694 The F<INSTALL> file for how to build Perl.
3696 The F<README> file for general stuff.
3698 The F<Artistic> and F<Copying> files for copyright information.
3702 Written by Jarkko Hietaniemi <F<jhi@iki.fi>>.