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 Properties
123 Unicode I<scripts> are now supported. Scripts are similar to (and superior
124 to) Unicode I<blocks>. The difference between scripts and blocks is that
125 scripts are the glyphs used by a language or a group of languages, while
126 the blocks are more artificial groupings of (mostly) 256 characters based
127 on the Unicode numbering.
129 In general, scripts are more inclusive, but not universally so. For
130 example, while the script C<Latin> includes all the Latin characters and
131 their various diacritic-adorned versions, it does not include the various
132 punctuation or digits (since they are not solely C<Latin>).
134 A number of other properties are now supported, including C<\p{L&}>,
135 C<\p{Any}> C<\p{Assigned}>, C<\p{Unassigned}>, C<\p{Blank}> [561] and
136 C<\p{SpacePerl}> [561] (along with their C<\P{...}> versions, of course).
137 See L<perlunicode> for details, and more additions.
139 The C<In> or C<Is> prefix to names used with the C<\p{...}> and C<\P{...}>
140 are now almost always optional. The only exception is that a C<In> prefix
141 is required to signify a Unicode block when a block name conflicts with a
142 script name. For example, C<\p{Tibetan}> refers to the script, while
143 C<\p{InTibetan}> refers to the block. When there is no name conflict, you
144 can omit the C<In> from the block name (e.g. C<\p{BraillePatterns}>), but
145 to be safe, it's probably best to always use the C<In>).
147 =head2 REF(...) Instead Of SCALAR(...)
149 A reference to a reference now stringifies as "REF(0x81485ec)" instead
150 of "SCALAR(0x81485ec)" in order to be more consistent with the return
153 =head2 pack/unpack D/F recycled
155 The undocumented pack/unpack template letters D/F have been recycled
156 for better use: now they stand for long double (if supported by the
157 platform) and NV (Perl internal floating point type). (They used
158 to be aliases for d/f, but you never knew that.)
160 =head2 glob() now returns filenames in alphabetical order
162 The list of filenames from glob() (or <...>) is now by default sorted
163 alphabetically to be csh-compliant (which is what happened before
164 in most UNIX platforms). (bsd_glob() does still sort platform
165 natively, ASCII or EBCDIC, unless GLOB_ALPHASORT is specified.) [561]
173 The semantics of bless(REF, REF) were unclear and until someone proves
174 it to make some sense, it is forbidden.
178 The obsolete chat2 library that should never have been allowed
179 to escape the laboratory has been decommissioned.
183 The builtin dump() function has probably outlived most of its
184 usefulness. The core-dumping functionality will remain in future
185 available as an explicit call to C<CORE::dump()>, but in future
186 releases the behaviour of an unqualified C<dump()> call may change.
190 The very dusty examples in the eg/ directory have been removed.
191 Suggestions for new shiny examples welcome but the main issue is that
192 the examples need to be documented, tested and (most importantly)
197 The (bogus) escape sequences \8 and \9 now give an optional warning
198 ("Unrecognized escape passed through"). There is no need to \-escape
203 The *glob{FILEHANDLE} is deprecated, use *glob{IO} instead.
207 The C<package;> syntax (C<package> without an argument) has been
208 deprecated. Its semantics were never that clear and its
209 implementation even less so. If you have used that feature to
210 disallow all but fully qualified variables, C<use strict;> instead.
214 The unimplemented POSIX regex features [[.cc.]] and [[=c=]] are still
215 recognised but now cause fatal errors. The previous behaviour of
216 ignoring them by default and warning if requested was unacceptable
217 since it, in a way, falsely promised that the features could be used.
221 In future releases, non-PerlIO aware XS modules may become completely
222 unsupported. Since PerlIO is a drop-in replacement for stdio at the
223 source code level, this shouldn't be that drastic a change.
227 Previous versions of perl and some readings of some sections of Camel
228 III implied that the C<:raw> "discipline" was the inverse of C<:crlf>.
229 Turning off "clrfness" is no longer enough to make a stream truly
230 binary. So the PerlIO C<:raw> layer (or "discipline", to use the Camel
231 book's older terminology) is now formally defined as being equivalent
232 to binmode(FH) - which is in turn defined as doing whatever is
233 necessary to pass each byte as-is without any translation. In
234 particular binmode(FH) - and hence C<:raw> - will now turn off both
235 CRLF and UTF-8 translation and remove other layers (e.g. :encoding())
236 which would modify byte stream.
240 The current user-visible implementation of pseudo-hashes (the weird
241 use of the first array element) is deprecated starting from Perl 5.8.0
242 and will be removed in Perl 5.10.0, and the feature will be
243 implemented differently. Not only is the current interface rather
244 ugly, but the current implementation slows down normal array and hash
245 use quite noticeably. The C<fields> pragma interface will remain
246 available. The I<restricted hashes> interface is expected to
247 be the replacement interface (see L<Hash::Util>). If your existing
248 programs depends on the underlying implementation, consider using
249 L<Class::PseudoHash> from CPAN.
253 The syntaxes C<< @a->[...] >> and C<< %h->{...} >> have now been deprecated.
257 After years of trying, suidperl is considered to be too complex to
258 ever be considered truly secure. The suidperl functionality is likely
259 to be removed in a future release.
263 The 5.005 threads model (module C<Thread>) is deprecated and expected
264 to be removed in Perl 5.10. Multithreaded code should be migrated to
265 the new ithreads model (see L<threads>, L<threads::shared> and
270 The long deprecated uppercase aliases for the string comparison
271 operators (EQ, NE, LT, LE, GE, GT) have now been removed.
275 The tr///C and tr///U features have been removed and will not return;
276 the interface was a mistake. Sorry about that. For similar
277 functionality, see pack('U0', ...) and pack('C0', ...). [561]
281 Earlier Perls treated "sub foo (@bar)" as equivalent to "sub foo (@)".
282 The prototypes are now checked better at compile-time for invalid
283 syntax. An optional warning is generated ("Illegal character in
284 prototype...") but this may be upgraded to a fatal error in a future
289 The C<exec LIST> and C<system LIST> operations now produce warnings on
290 tainted data and in some future release they will produce fatal errors.
294 The existing behaviour when localising tied arrays and hashes is wrong,
295 and will be changed in a future release, so do not rely on the existing
296 behaviour. See L<"Localising Tied Arrays and Hashes Is Broken">.
300 =head1 Core Enhancements
302 =head2 Unicode Overhaul
304 Unicode in general should be now much more usable than in Perl 5.6.0
305 (or even in 5.6.1). Unicode can be used in hash keys, Unicode in
306 regular expressions should work now, Unicode in tr/// should work now,
307 Unicode in I/O should work now. See L<perluniintro> for introduction
308 and L<perlunicode> for details.
314 The Unicode Character Database coming with Perl has been upgraded
315 to Unicode 3.2.0. For more information, see http://www.unicode.org/ .
316 [561+] (5.6.1 has UCD 3.0.1.)
320 For developers interested in enhancing Perl's Unicode capabilities:
321 almost all the UCD files are included with the Perl distribution in
322 the F<lib/unicore> subdirectory. The most notable omission, for space
323 considerations, is the Unihan database.
327 The properties \p{Blank} and \p{SpacePerl} have been added. "Blank" is like
328 C isblank(), that is, it contains only "horizontal whitespace" (the space
329 character is, the newline isn't), and the "SpacePerl" is the Unicode
330 equivalent of C<\s> (\p{Space} isn't, since that includes the vertical
331 tabulator character, whereas C<\s> doesn't.)
333 See "New Unicode Properties" earlier in this document for additional
334 information on changes with Unicode properties.
338 =head2 PerlIO is Now The Default
344 IO is now by default done via PerlIO rather than system's "stdio".
345 PerlIO allows "layers" to be "pushed" onto a file handle to alter the
346 handle's behaviour. Layers can be specified at open time via 3-arg
349 open($fh,'>:crlf :utf8', $path) || ...
351 or on already opened handles via extended C<binmode>:
353 binmode($fh,':encoding(iso-8859-7)');
355 The built-in layers are: unix (low level read/write), stdio (as in
356 previous Perls), perlio (re-implementation of stdio buffering in a
357 portable manner), crlf (does CRLF <=> "\n" translation as on Win32,
358 but available on any platform). A mmap layer may be available if
359 platform supports it (mostly UNIXes).
361 Layers to be applied by default may be specified via the 'open' pragma.
363 See L</"Installation and Configuration Improvements"> for the effects
364 of PerlIO on your architecture name.
368 If your platform supports fork(), you can use the list form of C<open>
369 for pipes. For example:
371 open KID_PS, "-|", "ps", "aux" or die $!;
373 forks the ps(1) command (without spawning a shell, as there are more
374 than three arguments to open()), and reads its standard output via the
375 C<KID_PS> filehandle. See L<perlipc>.
379 File handles can be marked as accepting Perl's internal encoding of Unicode
380 (UTF-8 or UTF-EBCDIC depending on platform) by a pseudo layer ":utf8" :
382 open($fh,">:utf8","Uni.txt");
384 Note for EBCDIC users: the pseudo layer ":utf8" is erroneously named
385 for you since it's not UTF-8 what you will be getting but instead
386 UTF-EBCDIC. See L<perlunicode>, L<utf8>, and
387 http://www.unicode.org/unicode/reports/tr16/ for more information.
388 In future releases this naming may change. See L<perluniintro>
389 for more information about UTF-8.
393 If your environment variables (LC_ALL, LC_CTYPE, LANG, LANGUAGE) look
394 like you want to use UTF-8 (any of the the variables match C</utf-?8/i>),
395 your STDIN, STDOUT, STDERR handles and the default open layer
396 (see L<open>) are marked as UTF-8. (This feature, like other new
397 features that combine Unicode and I/O, work only if you are using
398 PerlIO, but that's the default.)
400 Note that after this Perl really does assume that everything is UTF-8:
401 for example if some input handle is not, Perl will probably very soon
402 complain about the input data like this "Malformed UTF-8 ..." since
403 any old eight-bit data is not legal UTF-8.
405 Note for code authors: if you want to enable your users to use UTF-8
406 as their default encoding but in your code still have eight-bit I/O streams
407 (such as images or zip files), you need to explicitly open() or binmode()
408 with C<:bytes> (see L<perlfunc/open> and L<perlfunc/binmode>), or you
409 can just use C<binmode(FH)> (nice for pre-5.8.0 backward compatibility).
413 File handles can translate character encodings from/to Perl's internal
414 Unicode form on read/write via the ":encoding()" layer.
418 File handles can be opened to "in memory" files held in Perl scalars via:
420 open($fh,'>', \$variable) || ...
424 Anonymous temporary files are available without need to
425 'use FileHandle' or other module via
427 open($fh,"+>", undef) || ...
429 That is a literal undef, not an undefined value.
433 The list form of C<open> is now implemented for pipes (at least on UNIX):
435 open($fh,"-|", 'cat', '/etc/motd')
437 creates a pipe, and runs the equivalent of exec('cat', '/etc/motd') in
442 If your locale environment variables (LANGUAGE, LC_ALL, LC_CTYPE, LANG)
443 contain the strings 'UTF-8' or 'UTF8' (case-insensitive matching),
444 the default encoding of your STDIN, STDOUT, and STDERR, and of
445 B<any subsequent file open>, is UTF-8.
451 The new interpreter threads ("ithreads" for short) implementation of
452 multithreading, by Arthur Bergman, replaces the old "5.005 threads"
453 implementation. In the ithreads model any data sharing between
454 threads must be explicit, as opposed to the model where data sharing
455 was implicit. See L<threads> and L<threads::shared>, and
458 As a part of the ithreads implementation Perl will also use
459 any necessary and detectable reentrant libc interfaces.
461 =head2 Restricted Hashes
463 A restricted hash is restricted to a certain set of keys, no keys
464 outside the set can be added. Also individual keys can be restricted
465 so that the key cannot be deleted and the value cannot be changed.
466 No new syntax is involved: the Hash::Util module is the interface.
470 Perl used to be fragile in that signals arriving at inopportune moments
471 could corrupt Perl's internal state. Now Perl postpones handling of
472 signals until it's safe (between opcodes).
474 This change may have surprising side effects because signals no longer
475 interrupt Perl instantly. Perl will now first finish whatever it was
476 doing, like finishing an internal operation (like sort()) or an
477 external operation (like an I/O operation), and only then look at any
478 arrived signals (and before starting the next operation). No more corrupt
479 internal state since the current operation is always finished first,
480 but the signal may take more time to get heard. Note that breaking
481 out from potentially blocking operations should still work, though.
483 =head2 Understanding of Numbers
485 In general a lot of fixing has happened in the area of Perl's
486 understanding of numbers, both integer and floating point. Since in
487 many systems the standard number parsing functions like C<strtoul()>
488 and C<atof()> seem to have bugs, Perl tries to work around their
489 deficiencies. This results hopefully in more accurate numbers.
491 Perl now tries internally to use integer values in numeric conversions
492 and basic arithmetics (+ - * /) if the arguments are integers, and
493 tries also to keep the results stored internally as integers.
494 This change leads to often slightly faster and always less lossy
495 arithmetics. (Previously Perl always preferred floating point numbers
498 =head2 Arrays now always interpolate into double-quoted strings [561]
500 In double-quoted strings, arrays now interpolate, no matter what. The
501 behavior in earlier versions of perl 5 was that arrays would interpolate
502 into strings if the array had been mentioned before the string was
503 compiled, and otherwise Perl would raise a fatal compile-time error.
504 In versions 5.000 through 5.003, the error was
506 Literal @example now requires backslash
508 In versions 5.004_01 through 5.6.0, the error was
510 In string, @example now must be written as \@example
512 The idea here was to get people into the habit of writing
513 C<"fred\@example.com"> when they wanted a literal C<@> sign, just as
514 they have always written C<"Give me back my \$5"> when they wanted a
517 Starting with 5.6.1, when Perl now sees an C<@> sign in a
518 double-quoted string, it I<always> attempts to interpolate an array,
519 regardless of whether or not the array has been used or declared
520 already. The fatal error has been downgraded to an optional warning:
522 Possible unintended interpolation of @example in string
524 This warns you that C<"fred@example.com"> is going to turn into
525 C<fred.com> if you don't backslash the C<@>.
526 See http://www.plover.com/~mjd/perl/at-error.html for more details
527 about the history here.
529 =head2 Miscellaneous Changes
535 AUTOLOAD is now lvaluable, meaning that you can add the :lvalue attribute
536 to AUTOLOAD subroutines and you can assign to the AUTOLOAD return value.
540 The $Config{byteorder} (and corresponding BYTEORDER in config.h) was
541 previously wrong in platforms if sizeof(long) was 4, but sizeof(IV)
542 was 8. The byteorder was only sizeof(long) bytes long (1234 or 4321),
543 but now it is correctly sizeof(IV) bytes long, (12345678 or 87654321).
544 (This problem didn't affect Windows platforms.)
546 Also, $Config{byteorder} is now computed dynamically--this is more
547 robust with "fat binaries" where an executable image contains binaries
548 for more than one binary platform, and when cross-compiling.
552 C<perl -d:Module=arg,arg,arg> now works (previously one couldn't pass
553 in multiple arguments.)
557 C<do> followed by a bareword now ensures that this bareword isn't
558 a keyword (to avoid a bug where C<do q(foo.pl)> tried to call a
559 subroutine called C<q>). This means that for example instead of
560 C<do format()> you must write C<do &format()>.
564 The builtin dump() now gives an optional warning
565 C<dump() better written as CORE::dump()>,
566 meaning that by default C<dump(...)> is resolved as the builtin
567 dump() which dumps core and aborts, not as (possibly) user-defined
568 C<sub dump>. To call the latter, qualify the call as C<&dump(...)>.
569 (The whole dump() feature is to considered deprecated, and possibly
570 removed/changed in future releases.)
574 chomp() and chop() are now overridable. Note, however, that their
575 prototype (as given by C<prototype("CORE::chomp")> is undefined,
576 because it cannot be expressed and therefore one cannot really write
577 replacements to override these builtins.
581 END blocks are now run even if you exit/die in a BEGIN block.
582 Internally, the execution of END blocks is now controlled by
583 PL_exit_flags & PERL_EXIT_DESTRUCT_END. This enables the new
584 behaviour for Perl embedders. This will default in 5.10. See
589 Formats now support zero-padded decimal fields.
593 Although "you shouldn't do that", it was possible to write code that
594 depends on Perl's hashed key order (Data::Dumper does this). The new
595 algorithm "One-at-a-Time" produces a different hashed key order.
596 More details are in L</"Performance Enhancements">.
600 lstat(FILEHANDLE) now gives a warning because the operation makes no sense.
601 In future releases this may become a fatal error.
605 Spurious syntax errors generated in certain situations, when glob()
606 caused File::Glob to be loaded for the first time, have been fixed. [561]
610 Lvalue subroutines can now return C<undef> in list context. However,
611 the lvalue subroutine feature still remains experimental. [561+]
615 A lost warning "Can't declare ... dereference in my" has been
616 restored (Perl had it earlier but it became lost in later releases.)
620 A new special regular expression variable has been introduced:
621 C<$^N>, which contains the most-recently closed group (submatch).
625 C<no Module;> does not produce an error even if Module does not have an
626 unimport() method. This parallels the behavior of C<use> vis-a-vis
631 The numerical comparison operators return C<undef> if either operand
632 is a NaN. Previously the behaviour was unspecified.
636 C<our> can now have an experimental optional attribute C<unique> that
637 affects how global variables are shared among multiple interpreters,
642 The following builtin functions are now overridable: each(), keys(),
643 pop(), push(), shift(), splice(), unshift(). [561]
647 C<pack() / unpack()> can now group template letters with C<()> and then
648 apply repetition/count modifiers on the groups.
652 C<pack() / unpack()> can now process the Perl internal numeric types:
653 IVs, UVs, NVs-- and also long doubles, if supported by the platform.
654 The template letters are C<j>, C<J>, C<F>, and C<D>.
658 C<pack('U0a*', ...)> can now be used to force a string to UTF8.
662 my __PACKAGE__ $obj now works. [561]
666 POSIX::sleep() now returns the number of I<unslept> seconds
667 (as the POSIX standard says), as opposed to CORE::sleep() which
668 returns the number of slept seconds.
672 The printf() and sprintf() now support parameter reordering using the
673 C<%\d+\$> and C<*\d+\$> syntaxes. For example
675 print "%2\$s %1\$s\n", "foo", "bar";
677 will print "bar foo\n". This feature helps in writing
678 internationalised software, and in general when the order
679 of the parameters can vary.
683 The (\&) prototype now works properly. [561]
687 prototype(\[$@%&]) is now available to implicitly create references
688 (useful for example if you want to emulate the tie() interface).
692 A new command-line option, C<-t> is available. It is the
693 little brother of C<-T>: instead of dying on taint violations,
694 lexical warnings are given. B<This is only meant as a temporary
695 debugging aid while securing the code of old legacy applications.
696 This is not a substitute for -T.>
700 In other taint news, the C<exec LIST> and C<system LIST> have now been
701 considered too risky (think C<exec @ARGV>: it can start any program
702 with any arguments), and now the said forms cause a warning under
703 lexical warnings. You should carefully launder the arguments to
704 guarantee their validity. In future releases of Perl the forms will
705 become fatal errors so consider starting laundering now.
709 Tied hash interfaces are now required to have the EXISTS and DELETE
710 methods (either own or inherited).
714 If tr/// is just counting characters, it doesn't attempt to
719 untie() will now call an UNTIE() hook if it exists. See L<perltie>
724 L<utime> now supports C<utime undef, undef, @files> to change the
725 file timestamps to the current time.
729 The rules for allowing underscores (underbars) in numeric constants
730 have been relaxed and simplified: now you can have an underscore
731 simply B<between digits>.
735 Rather than relying on C's argv[0] (which may not contain a full pathname)
736 where possible $^X is now set by asking the operating system.
737 (eg by reading F</proc/self/exe> on Linux, F</proc/curproc/file> on FreeBSD)
741 A new variable, C<${^TAINT}>, indicates whether taint mode is enabled.
745 You can now override the readline() builtin, and this overrides also
746 the <FILEHANDLE> angle bracket operator.
750 The command-line options -s and -F are now recognized on the shebang
755 Use of the C</c> match modifier without an accompanying C</g> modifier
756 elicits a new warning: C<Use of /c modifier is meaningless without /g>.
758 Use of C</c> in substitutions, even with C</g>, elicits
759 C<Use of /c modifier is meaningless in s///>.
761 Use of C</g> with C<split> elicits C<Use of /g modifier is meaningless
766 Support for the C<CLONE> special subroutine had been added.
767 With ithreads, when a new thread is created, all Perl data is cloned,
768 however non-Perl data cannot be cloned automatically. In C<CLONE> you
769 can do whatever you need to do, like for example handle the cloning of
770 non-Perl data, if necessary. C<CLONE> will be executed once for every
771 package that has it defined or inherited. It will be called in the
772 context of the new thread, so all modifications are made in the new area.
778 =head1 Modules and Pragmata
780 =head2 New Modules and Pragmata
786 C<Attribute::Handlers>, originally by Damian Conway and now maintained
787 by Arthur Bergman, allows a class to define attribute handlers.
790 use Attribute::Handlers;
791 sub Wolf :ATTR(SCALAR) { print "howl!\n" }
793 # later, in some package using or inheriting from MyPack...
795 my MyPack $Fluffy : Wolf; # the attribute handler Wolf will be called
797 Both variables and routines can have attribute handlers. Handlers can
798 be specific to type (SCALAR, ARRAY, HASH, or CODE), or specific to the
799 exact compilation phase (BEGIN, CHECK, INIT, or END).
800 See L<Attribute::Handlers>.
804 C<B::Concise>, by Stephen McCamant, is a new compiler backend for
805 walking the Perl syntax tree, printing concise info about ops.
806 The output is highly customisable. See L<B::Concise>. [561+]
810 The new bignum, bigint, and bigrat pragmas, by Tels, implement
811 transparent bignum support (using the Math::BigInt, Math::BigFloat,
812 and Math::BigRat backends).
816 C<Class::ISA>, by Sean Burke, is a module for reporting the search
817 path for a class's ISA tree. See L<Class::ISA>.
821 C<Cwd> now has a split personality: if possible, an XS extension is
822 used, (this will hopefully be faster, more secure, and more robust)
823 but if not possible, the familiar Perl implementation is used.
827 C<Devel::PPPort>, originally by Kenneth Albanowski and now
828 maintained by Paul Marquess, has been added. It is primarily used
829 by C<h2xs> to enhance portability of XS modules between different
830 versions of Perl. See L<Devel::PPPort>.
834 C<Digest>, frontend module for calculating digests (checksums), from
835 Gisle Aas, has been added. See L<Digest>.
839 C<Digest::MD5> for calculating MD5 digests (checksums) as defined in
840 RFC 1321, from Gisle Aas, has been added. See L<Digest::MD5>.
842 use Digest::MD5 'md5_hex';
844 $digest = md5_hex("Thirsty Camel");
846 print $digest, "\n"; # 01d19d9d2045e005c3f1b80e8b164de1
848 NOTE: the C<MD5> backward compatibility module is deliberately not
849 included since its further use is discouraged.
851 See also L<PerlIO::via::QuotedPrint>.
855 C<Encode>, originally by Nick Ing-Simmons and now maintained by Dan
856 Kogai, provides a mechanism to translate between different character
857 encodings. Support for Unicode, ISO-8859-1, and ASCII are compiled in
858 to the module. Several other encodings (like the rest of the
859 ISO-8859, CP*/Win*, Mac, KOI8-R, three variants EBCDIC, Chinese,
860 Japanese, and Korean encodings) are included and can be loaded at
861 runtime. (For space considerations, the largest Chinese encodings
862 have been separated into their own CPAN module, Encode::HanExtra,
863 which Encode will use if available). See L<Encode>.
865 Any encoding supported by Encode module is also available to the
866 ":encoding()" layer if PerlIO is used.
870 C<Hash::Util> is the interface to the new I<restricted hashes>
871 feature. (Implemented by Jeffrey Friedl, Nick Ing-Simmons, and
872 Michael Schwern.) See L<Hash::Util>.
876 C<I18N::Langinfo> can be used to query locale information.
877 See L<I18N::Langinfo>.
881 C<I18N::LangTags>, by Sean Burke, has functions for dealing with
882 RFC3066-style language tags. See L<I18N::LangTags>.
886 C<ExtUtils::Constant>, by Nicholas Clark, is a new tool for extension
887 writers for generating XS code to import C header constants.
888 See L<ExtUtils::Constant>.
892 C<Filter::Simple>, by Damian Conway, is an easy-to-use frontend to
893 Filter::Util::Call. See L<Filter::Simple>.
899 use Filter::Simple sub {
900 while (my ($from, $to) = splice @_, 0, 2) {
909 use MyFilter qr/red/ => 'green';
911 print "red\n"; # this code is filtered, will print "green\n"
912 print "bored\n"; # this code is filtered, will print "bogreen\n"
916 print "red\n"; # this code is not filtered, will print "red\n"
920 C<File::Temp>, by Tim Jenness, allows one to create temporary files
921 and directories in an easy, portable, and secure way. See L<File::Temp>.
926 C<Filter::Util::Call>, by Paul Marquess, provides you with the
927 framework to write I<source filters> in Perl. For most uses, the
928 frontend Filter::Simple is to be preferred. See L<Filter::Util::Call>.
932 C<if>, by Ilya Zakharevich, is a new pragma for conditional inclusion
937 L<libnet>, by Graham Barr, is a collection of perl5 modules related
938 to network programming. See L<Net::FTP>, L<Net::NNTP>, L<Net::Ping>
939 (not part of libnet, but related), L<Net::POP3>, L<Net::SMTP>,
942 Perl installation leaves libnet unconfigured; use F<libnetcfg>
947 C<List::Util>, by Graham Barr, is a selection of general-utility
948 list subroutines, such as sum(), min(), first(), and shuffle().
953 C<Locale::Constants>, C<Locale::Country>, C<Locale::Currency>
954 C<Locale::Language>, and L<Locale::Script>, by Neil Bowers, have
955 been added. They provide the codes for various locale standards, such
956 as "fr" for France, "usd" for US Dollar, and "ja" for Japanese.
960 $country = code2country('jp'); # $country gets 'Japan'
961 $code = country2code('Norway'); # $code gets 'no'
963 See L<Locale::Constants>, L<Locale::Country>, L<Locale::Currency>,
964 and L<Locale::Language>.
968 C<Locale::Maketext>, by Sean Burke, is a localization framework. See
969 L<Locale::Maketext>, and L<Locale::Maketext::TPJ13>. The latter is an
970 article about software localization, originally published in The Perl
971 Journal #13, and republished here with kind permission.
975 C<Math::BigRat> for big rational numbers, to accompany Math::BigInt and
976 Math::BigFloat, from Tels. See L<Math::BigRat>.
980 C<Memoize> can make your functions faster by trading space for time,
981 from Mark-Jason Dominus. See L<Memoize>.
985 C<MIME::Base64>, by Gisle Aas, allows you to encode data in base64,
986 as defined in RFC 2045 - I<MIME (Multipurpose Internet Mail
991 $encoded = encode_base64('Aladdin:open sesame');
992 $decoded = decode_base64($encoded);
994 print $encoded, "\n"; # "QWxhZGRpbjpvcGVuIHNlc2FtZQ=="
1000 C<MIME::QuotedPrint>, by Gisle Aas, allows you to encode data
1001 in quoted-printable encoding, as defined in RFC 2045 - I<MIME
1002 (Multipurpose Internet Mail Extensions)>.
1004 use MIME::QuotedPrint;
1006 $encoded = encode_qp("Smiley in Unicode: \x{263a}");
1007 $decoded = decode_qp($encoded);
1009 print $encoded, "\n"; # "Smiley in Unicode: =263A"
1011 See also L<PerlIO::via::QuotedPrint>.
1015 C<NEXT>, by Damian Conway, is a pseudo-class for method redispatch.
1020 C<open> is a new pragma for setting the default I/O layers
1025 C<PerlIO::scalar>, by Nick Ing-Simmons, provides the implementation
1026 of IO to "in memory" Perl scalars as discussed above. It also serves
1027 as an example of a loadable PerlIO layer. Other future possibilities
1028 include PerlIO::Array and PerlIO::Code. See L<PerlIO::scalar>.
1032 C<PerlIO::via>, by Nick Ing-Simmons, acts as a PerlIO layer and wraps
1033 PerlIO layer functionality provided by a class (typically implemented
1038 C<PerlIO::via::QuotedPrint>, by Elizabeth Mattijsen, is an example
1039 of a C<PerlIO::via> class:
1041 use PerlIO::via::QuotedPrint;
1042 open($fh,">:via(QuotedPrint)",$path);
1044 This will automatically convert everything output to C<$fh> to
1045 Quoted-Printable. See L<PerlIO::via> and L<PerlIO::via::QuotedPrint>.
1049 C<Pod::ParseLink>, by Russ Allbery, has been added,
1050 to parse LZ<><> links in pods as described in the new
1055 C<Pod::Text::Overstrike>, by Joe Smith, has been added.
1056 It converts POD data to formatted overstrike text.
1057 See L<Pod::Text::Overstrike>. [561+]
1061 C<Scalar::Util> is a selection of general-utility scalar subroutines,
1062 such as blessed(), reftype(), and tainted(). See L<Scalar::Util>.
1066 C<sort> is a new pragma for controlling the behaviour of sort().
1070 C<Storable> gives persistence to Perl data structures by allowing the
1071 storage and retrieval of Perl data to and from files in a fast and
1072 compact binary format. Because in effect Storable does serialisation
1073 of Perl data structures, with it you can also clone deep, hierarchical
1074 datastructures. Storable was originally created by Raphael Manfredi,
1075 but it is now maintained by Abhijit Menon-Sen. Storable has been
1076 enhanced to understand the two new hash features, Unicode keys and
1077 restricted hashes. See L<Storable>.
1081 C<Switch>, by Damian Conway, has been added. Just by saying
1085 you have C<switch> and C<case> available in Perl.
1091 case 1 { print "number 1" }
1092 case "a" { print "string a" }
1093 case [1..10,42] { print "number in list" }
1094 case (@array) { print "number in list" }
1095 case /\w+/ { print "pattern" }
1096 case qr/\w+/ { print "pattern" }
1097 case (%hash) { print "entry in hash" }
1098 case (\%hash) { print "entry in hash" }
1099 case (\&sub) { print "arg to subroutine" }
1100 else { print "previous case not true" }
1107 C<Test::More>, by Michael Schwern, is yet another framework for writing
1108 test scripts, more extensive than Test::Simple. See L<Test::More>.
1112 C<Test::Simple>, by Michael Schwern, has basic utilities for writing
1113 tests. See L<Test::Simple>.
1117 C<Text::Balanced>, by Damian Conway, has been added, for extracting
1118 delimited text sequences from strings.
1120 use Text::Balanced 'extract_delimited';
1122 ($a, $b) = extract_delimited("'never say never', he never said", "'", '');
1124 $a will be "'never say never'", $b will be ', he never said'.
1126 In addition to extract_delimited(), there are also extract_bracketed(),
1127 extract_quotelike(), extract_codeblock(), extract_variable(),
1128 extract_tagged(), extract_multiple(), gen_delimited_pat(), and
1129 gen_extract_tagged(). With these, you can implement rather advanced
1130 parsing algorithms. See L<Text::Balanced>.
1134 C<threads>, by Arthur Bergman, is an interface to interpreter threads.
1135 Interpreter threads (ithreads) is the new thread model introduced in
1136 Perl 5.6 but only available as an internal interface for extension
1137 writers (and for Win32 Perl for C<fork()> emulation). See L<threads>,
1138 L<threads::shared>, and L<perlthrtut>.
1142 C<threads::shared>, by Arthur Bergman, allows data sharing for
1143 interpreter threads. See L<threads::shared>.
1147 C<Tie::File>, by Mark-Jason Dominus, associates a Perl array with the
1148 lines of a file. See L<Tie::File>.
1152 C<Tie::Memoize>, by Ilya Zakharevich, provides on-demand loaded hashes.
1153 See L<Tie::Memoize>.
1157 C<Tie::RefHash::Nestable>, by Edward Avis, allows storing hash
1158 references (unlike the standard Tie::RefHash) The module is contained
1159 within Tie::RefHash. See L<Tie::RefHash>.
1163 C<Time::HiRes>, by Douglas E. Wegscheid, provides high resolution
1164 timing (ualarm, usleep, and gettimeofday). See L<Time::HiRes>.
1168 C<Unicode::UCD> offers a querying interface to the Unicode Character
1169 Database. See L<Unicode::UCD>.
1173 C<Unicode::Collate>, by SADAHIRO Tomoyuki, implements the UCA
1174 (Unicode Collation Algorithm) for sorting Unicode strings.
1175 See L<Unicode::Collate>.
1179 C<Unicode::Normalize>, by SADAHIRO Tomoyuki, implements the various
1180 Unicode normalization forms. See L<Unicode::Normalize>.
1184 C<XS::APItest>, by Tim Jenness, is a test extension that exercises XS
1185 APIs. Currently only C<printf()> is tested: how to output various
1186 basic data types from XS.
1190 C<XS::Typemap>, by Tim Jenness, is a test extension that exercises
1191 XS typemaps. Nothing gets installed, but the code is worth studying
1192 for extension writers.
1196 =head2 Updated And Improved Modules and Pragmata
1202 The following independently supported modules have been updated to the
1203 newest versions from CPAN: CGI, CPAN, DB_File, File::Spec, File::Temp,
1204 Getopt::Long, Math::BigFloat, Math::BigInt, the podlators bundle
1205 (Pod::Man, Pod::Text), Pod::LaTeX [561+], Pod::Parser, Storable,
1206 Term::ANSIColor, Test, Text-Tabs+Wrap.
1210 attributes::reftype() now works on tied arguments.
1214 AutoLoader can now be disabled with C<no AutoLoader;>.
1218 B::Deparse has been significantly enhanced by Robin Houston. It can
1219 now deparse almost all of the standard test suite (so that the tests
1220 still succeed). There is a make target "test.deparse" for trying this
1225 Carp now has better interface documentation, and the @CARP_NOT
1226 interface has been added to get optional control over where errors
1227 are reported independently of @ISA, by Ben Tilly.
1231 Class::Struct can now define the classes in compile time.
1235 Class::Struct now assigns the array/hash element if the accessor
1236 is called with an array/hash element as the B<sole> argument.
1240 The return value of Cwd::fastcwd() is now tainted.
1244 Data::Dumper now has an option to sort hashes.
1248 Data::Dumper now has an option to dump code references
1253 DB_File now supports newer Berkeley DB versions, among
1258 Devel::Peek now has an interface for the Perl memory statistics
1259 (this works only if you are using perl's malloc, and if you have
1260 compiled with debugging).
1264 The English module can now be used without the infamous performance
1267 use English '-no_match_vars';
1269 (Assuming, of course, that you don't need the troublesome variables
1270 C<$`>, C<$&>, or C<$'>.) Also, introduced C<@LAST_MATCH_START> and
1271 C<@LAST_MATCH_END> English aliases for C<@-> and C<@+>.
1275 ExtUtils::MakeMaker has been significantly cleaned up and fixed.
1276 The enhanced version has also been backported to earlier releases
1277 of Perl and submitted to CPAN so that the earlier releases can
1282 The arguments of WriteMakefile() in Makefile.PL are now checked
1283 for sanity much more carefully than before. This may cause new
1284 warnings when modules are being installed. See L<ExtUtils::MakeMaker>
1289 ExtUtils::MakeMaker now uses File::Spec internally, which hopefully
1290 leads to better portability.
1294 Fcntl, Socket, and Sys::Syslog have been rewritten by Nicholas Clark
1295 to use the new-style constant dispatch section (see L<ExtUtils::Constant>).
1296 This means that they will be more robust and hopefully faster.
1300 File::Find now chdir()s correctly when chasing symbolic links. [561]
1304 File::Find now has pre- and post-processing callbacks. It also
1305 correctly changes directories when chasing symbolic links. Callbacks
1306 (naughtily) exiting with "next;" instead of "return;" now work.
1310 File::Find is now (again) reentrant. It also has been made
1315 The warnings issued by File::Find now belong to their own category.
1316 You can enable/disable them with C<use/no warnings 'File::Find';>.
1320 File::Glob::glob() has been renamed to File::Glob::bsd_glob()
1321 because the name clashes with the builtin glob(). The older
1322 name is still available for compatibility, but is deprecated. [561]
1326 File::Glob now supports C<GLOB_LIMIT> constant to limit the size of
1327 the returned list of filenames.
1331 IPC::Open3 now allows the use of numeric file descriptors.
1335 IO::Socket now has an atmark() method, which returns true if the socket
1336 is positioned at the out-of-band mark. The method is also exportable
1337 as a sockatmark() function.
1341 IO::Socket::INET failed to open the specified port if the service name
1342 was not known. It now correctly uses the supplied port number as is. [561]
1346 IO::Socket::INET has support for the ReusePort option (if your
1347 platform supports it). The Reuse option now has an alias, ReuseAddr.
1348 For clarity, you may want to prefer ReuseAddr.
1352 IO::Socket::INET now supports a value of zero for C<LocalPort>
1353 (usually meaning that the operating system will make one up.)
1357 'use lib' now works identically to @INC. Removing directories
1358 with 'no lib' now works.
1362 Math::BigFloat and Math::BigInt have undergone a full rewrite by Tels.
1363 They are now magnitudes faster, and they support various bignum
1364 libraries such as GMP and PARI as their backends.
1368 Math::Complex handles inf, NaN etc., better.
1372 Net::Ping has been considerably enhanced by Rob Brown: multihoming is
1373 now supported, Win32 functionality is better, there is now time
1374 measuring functionality (optionally high-resolution using
1375 Time::HiRes), and there is now "external" protocol which uses
1376 Net::Ping::External module which runs your external ping utility and
1377 parses the output. A version of Net::Ping::External is available in
1380 Note that some of the Net::Ping tests are disabled when running
1381 under the Perl distribution since one cannot assume one or more
1382 of the following: enabled echo port at localhost, full Internet
1383 connectivity, or sympathetic firewalls. You can set the environment
1384 variable PERL_TEST_Net_Ping to "1" (one) before running the Perl test
1385 suite to enable all the Net::Ping tests.
1389 POSIX::sigaction() is now much more flexible and robust.
1390 You can now install coderef handlers, 'DEFAULT', and 'IGNORE'
1391 handlers, installing new handlers was not atomic.
1395 In Safe, C<%INC> is now localised in a Safe compartment so that
1400 In SDBM_File on dosish platforms, some keys went missing because of
1401 lack of support for files with "holes". A workaround for the problem
1406 In Search::Dict one can now have a pre-processing hook for the
1407 lines being searched.
1411 The Shell module now has an OO interface.
1415 In Sys::Syslog there is now a failover mechanism that will go
1416 through alternative connection mechanisms until the message
1417 is successfully logged.
1421 The Test module has been significantly enhanced.
1425 Time::Local::timelocal() does not handle fractional seconds anymore.
1426 The rationale is that neither does localtime(), and timelocal() and
1427 localtime() are supposed to be inverses of each other.
1431 The vars pragma now supports declaring fully qualified variables.
1432 (Something that C<our()> does not and will not support.)
1436 The C<utf8::> name space (as in the pragma) provides various
1437 Perl-callable functions to provide low level access to Perl's
1438 internal Unicode representation. At the moment only length()
1439 has been implemented.
1443 =head1 Utility Changes
1449 Emacs perl mode (emacs/cperl-mode.el) has been updated to version
1454 F<emacs/e2ctags.pl> is now much faster.
1458 C<enc2xs> is a tool for people adding their own encodings to the
1463 C<h2ph> now supports C trigraphs.
1467 C<h2xs> now produces a template README.
1471 C<h2xs> now uses C<Devel::PPPort> for better portability between
1472 different versions of Perl.
1476 C<h2xs> uses the new L<ExtUtils::Constant|ExtUtils::Constant> module
1477 which will affect newly created extensions that define constants.
1478 Since the new code is more correct (if you have two constants where the
1479 first one is a prefix of the second one, the first constant B<never>
1480 got defined), less lossy (it uses integers for integer constant,
1481 as opposed to the old code that used floating point numbers even for
1482 integer constants), and slightly faster, you might want to consider
1483 regenerating your extension code (the new scheme makes regenerating
1484 easy). L<h2xs> now also supports C trigraphs.
1488 C<libnetcfg> has been added to configure libnet.
1492 C<perlbug> is now much more robust. It also sends the bug report to
1493 perl.org, not perl.com.
1497 C<perlcc> has been rewritten and its user interface (that is,
1498 command line) is much more like that of the UNIX C compiler, cc.
1499 (The perlbc tools has been removed. Use C<perlcc -B> instead.)
1500 B<Note that perlcc is still considered very experimental and
1505 C<perlivp> is a new Installation Verification Procedure utility
1506 for running any time after installing Perl.
1510 C<piconv> is an implementation of the character conversion utility
1511 C<iconv>, demonstrating the new Encode module.
1515 C<pod2html> now allows specifying a cache directory.
1519 C<pod2html> now produces XHTML 1.0.
1523 C<pod2html> now understands POD written using different line endings
1524 (PC-like CRLF versus UNIX-like LF versus MacClassic-like CR).
1528 C<s2p> has been completely rewritten in Perl. (It is in fact a full
1529 implementation of sed in Perl: you can use the sed functionality by
1530 using the C<psed> utility.)
1534 C<xsubpp> now understands POD documentation embedded in the *.xs
1539 C<xsubpp> now supports the OUT keyword.
1543 =head1 New Documentation
1549 perl56delta details the changes between the 5.005 release and the
1554 perlclib documents the internal replacements for standard C library
1555 functions. (Interesting only for extension writers and Perl core
1560 perldebtut is a Perl debugging tutorial. [561+]
1564 perlebcdic contains considerations for running Perl on EBCDIC
1569 perlintro is a gentle introduction to Perl.
1573 perliol documents the internals of PerlIO with layers.
1577 perlmodstyle is a style guide for writing modules.
1581 perlnewmod tells about writing and submitting a new module. [561+]
1585 perlpacktut is a pack() tutorial.
1589 perlpod has been rewritten to be clearer and to record the best
1590 practices gathered over the years.
1594 perlpodspec is a more formal specification of the pod format,
1595 mainly of interest for writers of pod applications, not to
1596 people writing in pod.
1600 perlretut is a regular expression tutorial. [561+]
1604 perlrequick is a regular expressions quick-start guide.
1605 Yes, much quicker than perlretut. [561]
1609 perltodo has been updated.
1613 perltootc has been renamed as perltooc (to not to conflict
1614 with perltoot in filesystems restricted to "8.3" names).
1618 perluniintro is an introduction to using Unicode in Perl.
1619 (perlunicode is more of a detailed reference and background
1624 perlutil explains the command line utilities packaged with the Perl
1625 distribution. [561+]
1629 The following platform-specific documents are available before
1630 the installation as README.I<platform>, and after the installation
1633 perlaix perlamiga perlapollo perlbeos perlbs2000
1634 perlce perlcygwin perldgux perldos perlepoc perlfreebsd perlhpux
1635 perlhurd perlirix perlmachten perlmacos perlmint perlmpeix
1636 perlnetware perlos2 perlos390 perlplan9 perlqnx perlsolaris
1637 perltru64 perluts perlvmesa perlvms perlvos perlwin32
1639 These documents usually detail one or more of the following subjects:
1640 configuring, building, testing, installing, and sometimes also using
1641 Perl on the said platform.
1643 Eastern Asian Perl users are now welcomed in their own languages:
1644 README.jp (Japanese), README.ko (Korean), README.cn (simplified
1645 Chinese) and README.tw (traditional Chinese), which are written in
1646 normal pod but encoded in EUC-JP, EUC-KR, EUC-CN and Big5. These
1647 will get installed as
1649 perljp perlko perlcn perltw
1655 The documentation for the POSIX-BC platform is called "BS2000", to avoid
1656 confusion with the Perl POSIX module.
1660 The documentation for the WinCE platform is called perlce (README.ce
1661 in the source code kit), to avoid confusion with the perlwin32
1662 documentation on 8.3-restricted filesystems.
1666 =head1 Performance Enhancements
1672 map() could get pathologically slow when the result list it generates
1673 is larger than the source list. The performance has been improved for
1674 common scenarios. [561]
1678 sort() is also fully reentrant, in the sense that the sort function
1679 can itself call sort(). This did not work reliably in previous
1684 sort() has been changed to use primarily mergesort internally as
1685 opposed to the earlier quicksort. For very small lists this may
1686 result in slightly slower sorting times, but in general the speedup
1687 should be at least 20%. Additional bonuses are that the worst case
1688 behaviour of sort() is now better (in computer science terms it now
1689 runs in time O(N log N), as opposed to quicksort's Theta(N**2)
1690 worst-case run time behaviour), and that sort() is now stable
1691 (meaning that elements with identical keys will stay ordered as they
1692 were before the sort). See the C<sort> pragma for information.
1694 The story in more detail: suppose you want to serve yourself a little
1697 @digits = ( 3,1,4,1,5,9 );
1699 A numerical sort of the digits will yield (1,1,3,4,5,9), as expected.
1700 Which C<1> comes first is hard to know, since one C<1> looks pretty
1701 much like any other. You can regard this as totally trivial,
1702 or somewhat profound. However, if you just want to sort the even
1703 digits ahead of the odd ones, then what will
1705 sort { ($a % 2) <=> ($b % 2) } @digits;
1707 yield? The only even digit, C<4>, will come first. But how about
1708 the odd numbers, which all compare equal? With the quicksort algorithm
1709 used to implement Perl 5.6 and earlier, the order of ties is left up
1710 to the sort. So, as you add more and more digits of Pi, the order
1711 in which the sorted even and odd digits appear will change.
1712 and, for sufficiently large slices of Pi, the quicksort algorithm
1713 in Perl 5.8 won't return the same results even if reinvoked with the
1714 same input. The justification for this rests with quicksort's
1715 worst case behavior. If you run
1717 sort { $a <=> $b } ( 1 .. $N , 1 .. $N );
1719 (something you might approximate if you wanted to merge two sorted
1720 arrays using sort), doubling $N doesn't just double the quicksort time,
1721 it I<quadruples> it. Quicksort has a worst case run time that can
1722 grow like N**2, so-called I<quadratic> behaviour, and it can happen
1723 on patterns that may well arise in normal use. You won't notice this
1724 for small arrays, but you I<will> notice it with larger arrays,
1725 and you may not live long enough for the sort to complete on arrays
1726 of a million elements. So the 5.8 quicksort scrambles large arrays
1727 before sorting them, as a statistical defence against quadratic behaviour.
1728 But that means if you sort the same large array twice, ties may be
1729 broken in different ways.
1731 Because of the unpredictability of tie-breaking order, and the quadratic
1732 worst-case behaviour, quicksort was I<almost> replaced completely with
1733 a stable mergesort. I<Stable> means that ties are broken to preserve
1734 the original order of appearance in the input array. So
1736 sort { ($a % 2) <=> ($b % 2) } (3,1,4,1,5,9);
1738 will yield (4,3,1,1,5,9), guaranteed. The even and odd numbers
1739 appear in the output in the same order they appeared in the input.
1740 Mergesort has worst case O(N log N) behaviour, the best value
1741 attainable. And, ironically, this mergesort does particularly
1742 well where quicksort goes quadratic: mergesort sorts (1..$N, 1..$N)
1743 in O(N) time. But quicksort was rescued at the last moment because
1744 it is faster than mergesort on certain inputs and platforms.
1745 For example, if you really I<don't> care about the order of even
1746 and odd digits, quicksort will run in O(N) time; it's very good
1747 at sorting many repetitions of a small number of distinct elements.
1748 The quicksort divide and conquer strategy works well on platforms
1749 with relatively small, very fast, caches. Eventually, the problem gets
1750 whittled down to one that fits in the cache, from which point it
1751 benefits from the increased memory speed.
1753 Quicksort was rescued by implementing a sort pragma to control aspects
1754 of the sort. The B<stable> subpragma forces stable behaviour,
1755 regardless of algorithm. The B<_quicksort> and B<_mergesort>
1756 subpragmas are heavy-handed ways to select the underlying implementation.
1757 The leading C<_> is a reminder that these subpragmas may not survive
1758 beyond 5.8. More appropriate mechanisms for selecting the implementation
1759 exist, but they wouldn't have arrived in time to save quicksort.
1763 Hashes now use Bob Jenkins "One-at-a-Time" hashing key algorithm
1764 ( http://burtleburtle.net/bob/hash/doobs.html ). This algorithm is
1765 reasonably fast while producing a much better spread of values than
1766 the old hashing algorithm (originally by Chris Torek, later tweaked by
1767 Ilya Zakharevich). Hash values output from the algorithm on a hash of
1768 all 3-char printable ASCII keys comes much closer to passing the
1769 DIEHARD random number generation tests. According to perlbench, this
1770 change has not affected the overall speed of Perl.
1774 unshift() should now be noticeably faster.
1778 =head1 Installation and Configuration Improvements
1780 =head2 Generic Improvements
1786 INSTALL now explains how you can configure Perl to use 64-bit
1787 integers even on non-64-bit platforms.
1791 Policy.sh policy change: if you are reusing a Policy.sh file
1792 (see INSTALL) and you use Configure -Dprefix=/foo/bar and in the old
1793 Policy $prefix eq $siteprefix and $prefix eq $vendorprefix, all of
1794 them will now be changed to the new prefix, /foo/bar. (Previously
1795 only $prefix changed.) If you do not like this new behaviour,
1796 specify prefix, siteprefix, and vendorprefix explicitly.
1800 A new optional location for Perl libraries, otherlibdirs, is available.
1801 It can be used for example for vendor add-ons without disturbing Perl's
1802 own library directories.
1806 In many platforms, the vendor-supplied 'cc' is too stripped-down to
1807 build Perl (basically, 'cc' doesn't do ANSI C). If this seems
1808 to be the case and 'cc' does not seem to be the GNU C compiler
1809 'gcc', an automatic attempt is made to find and use 'gcc' instead.
1813 gcc needs to closely track the operating system release to avoid
1814 build problems. If Configure finds that gcc was built for a different
1815 operating system release than is running, it now gives a clearly visible
1816 warning that there may be trouble ahead.
1820 Since Perl 5.8 is not binary-compatible with previous releases
1821 of Perl, Configure no longer suggests including the 5.005
1826 Configure C<-S> can now run non-interactively. [561]
1830 Configure support for pdp11-style memory models has been removed due
1831 to obsolescence. [561]
1835 configure.gnu now works with options with whitespace in them.
1839 installperl now outputs everything to STDERR.
1843 Because PerlIO is now the default on most platforms, "-perlio" doesn't
1844 get appended to the $Config{archname} (also known as $^O) anymore.
1845 Instead, if you explicitly choose not to use perlio (Configure command
1846 line option -Uuseperlio), you will get "-stdio" appended.
1850 Another change related to the architecture name is that "-64all"
1851 (-Duse64bitall, or "maximally 64-bit") is appended only if your
1852 pointers are 64 bits wide. (To be exact, the use64bitall is ignored.)
1856 In AFS installations, one can configure the root of the AFS to be
1857 somewhere else than the default F</afs> by using the Configure
1858 parameter C<-Dafsroot=/some/where/else>.
1862 APPLLIB_EXP, a lesser-known configuration-time definition, has been
1863 documented. It can be used to prepend site-specific directories
1864 to Perl's default search path (@INC); see INSTALL for information.
1868 The version of Berkeley DB used when the Perl (and, presumably, the
1869 DB_File extension) was built is now available as
1870 C<@Config{qw(db_version_major db_version_minor db_version_patch)}>
1871 from Perl and as C<DB_VERSION_MAJOR_CFG DB_VERSION_MINOR_CFG
1872 DB_VERSION_PATCH_CFG> from C.
1876 Building Berkeley DB3 for compatibility modes for DB, NDBM, and ODBM
1877 has been documented in INSTALL.
1881 If you have CPAN access (either network or a local copy such as a
1882 CD-ROM) you can during specify extra modules to Configure to build and
1883 install with Perl using the -Dextras=... option. See INSTALL for
1888 In addition to config.over, a new override file, config.arch, is
1889 available. This file is supposed to be used by hints file writers
1890 for architecture-wide changes (as opposed to config.over which is
1891 for site-wide changes).
1895 If your file system supports symbolic links, you can build Perl outside
1896 of the source directory by
1898 mkdir /tmp/perl/build/directory
1899 cd /tmp/perl/build/directory
1900 sh /path/to/perl/source/Configure -Dmksymlinks ...
1902 This will create in /tmp/perl/build/directory a tree of symbolic links
1903 pointing to files in /path/to/perl/source. The original files are left
1904 unaffected. After Configure has finished, you can just say
1908 and Perl will be built and tested, all in /tmp/perl/build/directory.
1913 For Perl developers, several new make targets for profiling
1914 and debugging have been added; see L<perlhack>.
1920 Use of the F<gprof> tool to profile Perl has been documented in
1921 L<perlhack>. There is a make target called "perl.gprof" for
1922 generating a gprofiled Perl executable.
1926 If you have GCC 3, there is a make target called "perl.gcov" for
1927 creating a gcoved Perl executable for coverage analysis. See
1932 If you are on IRIX or Tru64 platforms, new profiling/debugging options
1933 have been added; see L<perlhack> for more information about pixie and
1940 Guidelines of how to construct minimal Perl installations have
1941 been added to INSTALL.
1945 The Thread extension is now not built at all under ithreads
1946 (C<Configure -Duseithreads>) because it wouldn't work anyway (the
1947 Thread extension requires being Configured with C<-Duse5005threads>).
1949 B<Note that the 5.005 threads are unsupported and deprecated: if you
1950 have code written for the old threads you should migrate it to the
1951 new ithreads model.>
1955 The Gconvert macro ($Config{d_Gconvert}) used by perl for stringifying
1956 floating-point numbers is now more picky about using sprintf %.*g
1957 rules for the conversion. Some platforms that used to use gcvt may
1958 now resort to the slower sprintf.
1962 The obsolete method of making a special (e.g., debugging) flavor
1965 make LIBPERL=libperld.a
1967 has been removed. Use -DDEBUGGING instead.
1971 =head2 New Or Improved Platforms
1973 For the list of platforms known to support Perl,
1974 see L<perlport/"Supported Platforms">.
1980 AIX dynamic loading should be now better supported.
1984 AIX should now work better with gcc, threads, and 64-bitness. Also the
1985 long doubles support in AIX should be better now. See L<perlaix>.
1989 AtheOS ( http://www.atheos.cx/ ) is a new platform.
1993 BeOS has been reclaimed.
1997 The DG/UX platform now supports 5.005-style threads.
2002 The DYNIX/ptx platform (also known as dynixptx) is supported at or
2007 EBCDIC platforms (z/OS (also known as OS/390), POSIX-BC, and VM/ESA)
2008 have been regained. Many test suite tests still fail and the
2009 co-existence of Unicode and EBCDIC isn't quite settled, but the
2010 situation is much better than with Perl 5.6. See L<perlos390>,
2011 L<perlbs2000> (for POSIX-BC), and L<perlvmesa> for more information.
2015 Building perl with -Duseithreads or -Duse5005threads now works under
2016 HP-UX 10.20 (previously it only worked under 10.30 or later). You will
2017 need a thread library package installed. See README.hpux. [561]
2021 Mac OS Classic is now supported in the mainstream source package
2022 (MacPerl has of course been available since perl 5.004 but now the
2023 source code bases of standard Perl and MacPerl have been synchronised)
2028 Mac OS X (or Darwin) should now be able to build Perl even on HFS+
2029 filesystems. (The case-insensitivity used to confuse the Perl build
2034 NCR MP-RAS is now supported. [561]
2038 All the NetBSD specific patches (except for the installation
2039 specific ones) have been merged back to the main distribution.
2043 NetWare from Novell is now supported. See L<perlnetware>.
2047 NonStop-UX is now supported. [561]
2051 NEC SUPER-UX is now supported.
2055 All the OpenBSD specific patches (except for the installation
2056 specific ones) have been merged back to the main distribution.
2060 Perl has been tested with the GNU pth userlevel thread package
2061 ( http://www.gnu.org/software/pth/pth.html ). All thread tests
2062 of Perl now work, but not without adding some yield()s to the tests,
2063 so while pth (and other userlevel thread implementations) can be
2064 considered to be "working" with Perl ithreads, keep in mind the
2065 possible non-preemptability of the underlying thread implementation.
2069 Stratus VOS is now supported using Perl's native build method
2070 (Configure). This is the recommended method to build Perl on
2071 VOS. The older methods, which build miniperl, are still
2072 available. See L<perlvos>. [561+]
2076 The Amdahl UTS UNIX mainframe platform is now supported. [561]
2080 WinCE is now supported. See L<perlce>.
2084 z/OS (formerly known as OS/390, formerly known as MVS OE) now has
2085 support for dynamic loading. This is not selected by default,
2086 however, you must specify -Dusedl in the arguments of Configure. [561]
2090 =head1 Selected Bug Fixes
2092 Numerous memory leaks and uninitialized memory accesses have been
2093 hunted down. Most importantly, anonymous subs used to leak quite
2100 The autouse pragma didn't work for Multi::Part::Function::Names.
2104 caller() could cause core dumps in certain situations. Carp was
2105 sometimes affected by this problem. In particular, caller() now
2106 returns a subroutine name of C<(unknown)> for subroutines that have
2107 been removed from the symbol table.
2111 chop(@list) in list context returned the characters chopped in
2112 reverse order. This has been reversed to be in the right order. [561]
2116 Configure no longer includes the DBM libraries (dbm, gdbm, db, ndbm)
2117 when building the Perl binary. The only exception to this is SunOS 4.x,
2118 which needs them. [561]
2122 The behaviour of non-decimal but numeric string constants such as
2123 "0x23" was platform-dependent: in some platforms that was seen as 35,
2124 in some as 0, in some as a floating point number (don't ask). This
2125 was caused by Perl's using the operating system libraries in a situation
2126 where the result of the string to number conversion is undefined: now
2127 Perl consistently handles such strings as zero in numeric contexts.
2131 Several debugger fixes: exit code now reflects the script exit code,
2132 condition C<"0"> now treated correctly, the C<d> command now checks
2133 line number, C<$.> no longer gets corrupted, and all debugger output
2134 now goes correctly to the socket if RemotePort is set. [561]
2138 The debugger (perl5db.pl) has been modified to present a more
2139 consistent commands interface, via (CommandSet=580). perl5db.t was
2140 also added to test the changes, and as a placeholder for further tests.
2146 The debugger has a new C<dumpDepth> option to control the maximum
2147 depth to which nested structures are dumped. The C<x> command has
2148 been extended so that C<x N EXPR> dumps out the value of I<EXPR> to a
2149 depth of at most I<N> levels.
2153 The debugger can now show lexical variables if you have the CPAN
2154 module PadWalker installed.
2158 The order of DESTROYs has been made more predictable.
2162 Perl 5.6.0 could emit spurious warnings about redefinition of
2163 dl_error() when statically building extensions into perl.
2164 This has been corrected. [561]
2168 L<dprofpp> -R didn't work.
2172 C<*foo{FORMAT}> now works.
2176 Infinity is now recognized as a number.
2180 UNIVERSAL::isa no longer caches methods incorrectly. (This broke
2181 the Tk extension with 5.6.0.) [561]
2185 Lexicals I: lexicals outside an eval "" weren't resolved
2186 correctly inside a subroutine definition inside the eval "" if they
2187 were not already referenced in the top level of the eval""ed code.
2191 Lexicals II: lexicals leaked at file scope into subroutines that
2192 were declared before the lexicals.
2196 Lexical warnings now propagating correctly between scopes
2197 and into C<eval "...">.
2201 C<use warnings qw(FATAL all)> did not work as intended. This has been
2206 warnings::enabled() now reports the state of $^W correctly if the caller
2207 isn't using lexical warnings. [561]
2211 Line renumbering with eval and C<#line> now works. [561]
2215 Fixed numerous memory leaks, especially in eval "".
2219 Localised tied variables no longer leak memory
2222 tie my %tied_hash => 'Tie::StdHash';
2226 # Used to leak memory every time local() was called;
2227 # in a loop, this added up.
2228 local($tied_hash{Foo}) = 1;
2232 Localised hash elements (and %ENV) are correctly unlocalised to not
2233 exist, if they didn't before they were localised.
2237 tie my %tied_hash => 'Tie::StdHash';
2241 # Nothing has set the FOO element so far
2243 { local $tied_hash{FOO} = 'Bar' }
2245 # This used to print, but not now.
2246 print "exists!\n" if exists $tied_hash{FOO};
2248 As a side effect of this fix, tied hash interfaces B<must> define
2249 the EXISTS and DELETE methods.
2253 mkdir() now ignores trailing slashes in the directory name,
2254 as mandated by POSIX.
2258 Some versions of glibc have a broken modfl(). This affects builds
2259 with C<-Duselongdouble>. This version of Perl detects this brokenness
2260 and has a workaround for it. The glibc release 2.2.2 is known to have
2261 fixed the modfl() bug.
2265 Modulus of unsigned numbers now works (4063328477 % 65535 used to
2266 return 27406, instead of 27047). [561]
2270 Some "not a number" warnings introduced in 5.6.0 eliminated to be
2271 more compatible with 5.005. Infinity is now recognised as a number. [561]
2275 Numeric conversions did not recognize changes in the string value
2276 properly in certain circumstances. [561]
2280 Attributes (such as :shared) didn't work with our().
2284 our() variables will not cause bogus "Variable will not stay shared"
2289 "our" variables of the same name declared in two sibling blocks
2290 resulted in bogus warnings about "redeclaration" of the variables.
2291 The problem has been corrected. [561]
2295 pack "Z" now correctly terminates the string with "\0".
2299 Fix password routines which in some shadow password platforms
2300 (e.g. HP-UX) caused getpwent() to return every other entry.
2304 The PERL5OPT environment variable (for passing command line arguments
2305 to Perl) didn't work for more than a single group of options. [561]
2309 PERL5OPT with embedded spaces didn't work.
2313 printf() no longer resets the numeric locale to "C".
2317 C<qw(a\\b)> now parses correctly as C<'a\\b'>: that is, as three
2318 characters, not four. [561]
2322 pos() did not return the correct value within s///ge in earlier
2323 versions. This is now handled correctly. [561]
2327 Printing quads (64-bit integers) with printf/sprintf now works
2328 without the q L ll prefixes (assuming you are on a quad-capable platform).
2332 Regular expressions on references and overloaded scalars now work. [561+]
2336 Right-hand side magic (GMAGIC) could in many cases such as string
2337 concatenation be invoked too many times.
2341 scalar() now forces scalar context even when used in void context.
2345 SOCKS support is now much more robust.
2349 sort() arguments are now compiled in the right wantarray context
2350 (they were accidentally using the context of the sort() itself).
2351 The comparison block is now run in scalar context, and the arguments
2352 to be sorted are always provided list context. [561]
2356 Changed the POSIX character class C<[[:space:]]> to include the (very
2357 rarely used) vertical tab character. Added a new POSIX-ish character
2358 class C<[[:blank:]]> which stands for horizontal whitespace
2359 (currently, the space and the tab).
2363 The tainting behaviour of sprintf() has been rationalized. It does
2364 not taint the result of floating point formats anymore, making the
2365 behaviour consistent with that of string interpolation. [561]
2369 Some cases of inconsistent taint propagation (such as within hash
2370 values) have been fixed.
2374 The RE engine found in Perl 5.6.0 accidentally pessimised certain kinds
2375 of simple pattern matches. These are now handled better. [561]
2379 Regular expression debug output (whether through C<use re 'debug'>
2380 or via C<-Dr>) now looks better. [561]
2384 Multi-line matches like C<"a\nxb\n" =~ /(?!\A)x/m> were flawed. The
2385 bug has been fixed. [561]
2389 Use of $& could trigger a core dump under some situations. This
2390 is now avoided. [561]
2394 The regular expression captured submatches ($1, $2, ...) are now
2395 more consistently unset if the match fails, instead of leaving false
2396 data lying around in them. [561]
2400 readline() on files opened in "slurp" mode could return an extra
2401 "" (blank line) at the end in certain situations. This has been
2406 Autovivification of symbolic references of special variables described
2407 in L<perlvar> (as in C<${$num}>) was accidentally disabled. This works
2412 Sys::Syslog ignored the C<LOG_AUTH> constant.
2416 $AUTOLOAD, sort(), lock(), and spawning subprocesses
2417 in multiple threads simultaneously are now thread-safe.
2421 Tie::Array's SPLICE method was broken.
2425 Allow a read-only string on the left-hand side of a non-modifying tr///.
2429 If C<STDERR> is tied, warnings caused by C<warn> and C<die> now
2430 correctly pass to it.
2434 Several Unicode fixes.
2440 BOMs (byte order marks) at the beginning of Perl files
2441 (scripts, modules) should now be transparently skipped.
2442 UTF-16 and UCS-2 encoded Perl files should now be read correctly.
2446 The character tables have been updated to Unicode 3.2.0.
2450 Comparing with utf8 data does not magically upgrade non-utf8 data
2451 into utf8. (This was a problem for example if you were mixing data
2452 from I/O and Unicode data: your output might have got magically encoded
2457 Generating illegal Unicode code points such as U+FFFE, or the UTF-16
2458 surrogates, now also generates an optional warning.
2462 C<IsAlnum>, C<IsAlpha>, and C<IsWord> now match titlecase.
2466 Concatenation with the C<.> operator or via variable interpolation,
2467 C<eq>, C<substr>, C<reverse>, C<quotemeta>, the C<x> operator,
2468 substitution with C<s///>, single-quoted UTF8, should now work.
2472 The C<tr///> operator now works. Note that the C<tr///CU>
2473 functionality has been removed (but see pack('U0', ...)).
2477 C<eval "v200"> now works.
2481 Perl 5.6.0 parsed m/\x{ab}/ incorrectly, leading to spurious warnings.
2482 This has been corrected. [561]
2486 Zero entries were missing from the Unicode classes such as C<IsDigit>.
2492 Large unsigned numbers (those above 2**31) could sometimes lose their
2493 unsignedness, causing bogus results in arithmetic operations. [561]
2497 The Perl parser has been stress tested using both random input and
2498 Markov chain input and the few found crashes and lockups have been
2503 =head2 Platform Specific Changes and Fixes
2511 Perl now works on post-4.0 BSD/OSes.
2517 Setting C<$0> now works (as much as possible; see L<perlvar> for details).
2523 Numerous updates; currently synchronised with Cygwin 1.3.10.
2527 Previously DYNIX/ptx had problems in its Configure probe for non-blocking I/O.
2533 EPOC now better supported. See README.epoc. [561]
2539 Perl now works on post-3.0 FreeBSDs.
2545 README.hpux updated; C<Configure -Duse64bitall> now works;
2546 now uses HP-UX malloc instead of Perl malloc.
2552 Numerous compilation flag and hint enhancements; accidental mixing
2553 of 32-bit and 64-bit libraries (a doomed attempt) made much harder.
2563 Long doubles should now work (see INSTALL). [561]
2567 Linux previously had problems related to sockaddrlen when using
2568 accept(), recvfrom() (in Perl: recv()), getpeername(), and
2577 Compilation of the standard Perl distribution in Mac OS Classic should
2578 now work if you have the Metrowerks development environment and the
2579 missing Mac-specific toolkit bits. Contact the macperl mailing list
2586 MPE/iX update after Perl 5.6.0. See README.mpeix. [561]
2590 NetBSD/threads: try installing the GNU pth (should be in the
2591 packages collection, or http://www.gnu.org/software/pth/),
2592 and Configure with -Duseithreads.
2598 Perl now works on NetBSD/sparc.
2604 Now works with usethreads (see INSTALL). [561]
2610 64-bitness using the Sun Workshop compiler now works.
2616 The native build method requires at least VOS Release 14.5.0
2617 and GNU C++/GNU Tools 2.0.1 or later. The Perl pack function
2618 now maps overflowed values to +infinity and underflowed values
2623 Tru64 (aka Digital UNIX, aka DEC OSF/1)
2625 The operating system version letter now recorded in $Config{osvers}.
2626 Allow compiling with gcc (previously explicitly forbidden). Compiling
2627 with gcc still not recommended because buggy code results, even with
2634 Fixed various alignment problems that lead into core dumps either
2635 during build or later; no longer dies on math errors at runtime;
2636 now using full quad integers (64 bits), previously was using
2637 only 46 bit integers for speed.
2643 See L</"Socket Extension Dynamic in VMS"> and L</"IEEE-format Floating Point
2644 Default on OpenVMS Alpha"> for important changes not otherwise listed here.
2646 chdir() now works better despite a CRT bug; now works with MULTIPLICITY
2647 (see INSTALL); now works with Perl's malloc.
2649 The tainting of C<%ENV> elements via C<keys> or C<values> was previously
2650 unimplemented. It now works as documented.
2652 The C<waitpid> emulation has been improved. The worst bug (now fixed)
2653 was that a pid of -1 would cause a wildcard search of all processes on
2656 POSIX-style signals are now emulated much better on VMS versions prior
2659 The C<system> function and backticks operator have improved
2660 functionality and better error handling. [561]
2662 File access tests now use current process privileges rather than the
2663 user's default privileges, which could sometimes result in a mismatch
2664 between reported access and actual access. This improvement is only
2665 available on VMS v6.0 and later.
2667 There is a new C<kill> implementation based on C<sys$sigprc> that allows
2668 older VMS systems (pre-7.0) to use C<kill> to send signals rather than
2669 simply force exit. This implementation also allows later systems to
2670 call C<kill> from within a signal handler.
2672 Iterative logical name translations are now limited to 10 iterations in
2673 imitation of SHOW LOGICAL and other OpenVMS facilities.
2683 Signal handling now works better than it used to. It is now implemented
2684 using a Windows message loop, and is therefore less prone to random
2689 fork() emulation is now more robust, but still continues to have a few
2690 esoteric bugs and caveats. See L<perlfork> for details. [561+]
2694 A failed (pseudo)fork now returns undef and sets errno to EAGAIN. [561]
2698 The following modules now work on Windows:
2700 ExtUtils::Embed [561]
2707 IO::File::new_tmpfile() is no longer limited to 32767 invocations
2712 Better chdir() return value for a non-existent directory.
2716 Compiling perl using the 64-bit Platform SDK tools is now supported.
2720 The Win32::SetChildShowWindow() builtin can be used to control the
2721 visibility of windows created by child processes. See L<Win32> for
2726 Non-blocking waits for child processes (or pseudo-processes) are
2727 supported via C<waitpid($pid, &POSIX::WNOHANG)>.
2731 The behavior of system() with multiple arguments has been rationalized.
2732 Each unquoted argument will be automatically quoted to protect whitespace,
2733 and any existing whitespace in the arguments will be preserved. This
2734 improves the portability of system(@args) by avoiding the need for
2735 Windows C<cmd> shell specific quoting in perl programs.
2737 Note that this means that some scripts that may have relied on earlier
2738 buggy behavior may no longer work correctly. For example,
2739 C<system("nmake /nologo", @args)> will now attempt to run the file
2740 C<nmake /nologo> and will fail when such a file isn't found.
2741 On the other hand, perl will now execute code such as
2742 C<system("c:/Program Files/MyApp/foo.exe", @args)> correctly.
2746 The perl header files no longer suppress common warnings from the
2747 Microsoft Visual C++ compiler. This means that additional warnings may
2748 now show up when compiling XS code.
2752 Borland C++ v5.5 is now a supported compiler that can build Perl.
2753 However, the generated binaries continue to be incompatible with those
2754 generated by the other supported compilers (GCC and Visual C++). [561]
2758 Duping socket handles with open(F, ">&MYSOCK") now works under Windows 9x.
2763 Current directory entries in %ENV are now correctly propagated to child
2768 New %ENV entries now propagate to subprocesses. [561]
2772 Win32::GetCwd() correctly returns C:\ instead of C: when at the drive root.
2773 Other bugs in chdir() and Cwd::cwd() have also been fixed. [561]
2777 The makefiles now default to the features enabled in ActiveState ActivePerl
2778 (a popular Win32 binary distribution). [561]
2782 HTML files will now be installed in c:\perl\html instead of
2783 c:\perl\lib\pod\html
2787 REG_EXPAND_SZ keys are now allowed in registry settings used by perl. [561]
2791 Can now send() from all threads, not just the first one. [561]
2795 ExtUtils::MakeMaker now uses $ENV{LIB} to search for libraries. [561]
2799 Less stack reserved per thread so that more threads can run
2800 concurrently. (Still 16M per thread.) [561]
2804 C<< File::Spec->tmpdir() >> now prefers C:/temp over /tmp
2805 (works better when perl is running as service).
2809 Better UNC path handling under ithreads. [561]
2813 wait(), waitpid(), and backticks now return the correct exit status
2814 under Windows 9x. [561]
2818 A socket handle leak in accept() has been fixed. [561]
2824 =head1 New or Changed Diagnostics
2826 Please see L<perldiag> for more details.
2832 Ambiguous range in the transliteration operator (like a-z-9) now
2837 Two new debugging options have been added: if you have compiled your
2838 Perl with debugging, you can use the -DT [561] and -DR options to trace
2839 tokenising and to add reference counts to displaying variables,
2844 The lexical warnings category "deprecated" is no longer a sub-category
2845 of the "syntax" category. It is now a top-level category in its own
2850 Unadorned dump() will now give a warning suggesting to
2851 use explicit CORE::dump() if that's what really is meant.
2855 The "Unrecognized escape" warning has been extended to include C<\8>,
2856 C<\9>, and C<\_>. There is no need to escape any of the C<\w> characters.
2860 All regular expression compilation error messages are now hopefully
2861 easier to understand both because the error message now comes before
2862 the failed regex and because the point of failure is now clearly
2863 marked by a C<E<lt>-- HERE> marker.
2867 Various I/O (and socket) functions like binmode(), close(), and so
2868 forth now more consistently warn if they are used illogically either
2869 on a yet unopened or on an already closed filehandle (or socket).
2873 Using lstat() on a filehandle now gives a warning. (It's a non-sensical
2878 The C<-M> and C<-m> options now warn if you didn't supply the module name.
2882 If you in C<use> specify a required minimum version, modules matching
2883 the name and but not defining a $VERSION will cause a fatal failure.
2887 Using negative offset for vec() in lvalue context is now a warnable offense.
2891 Odd number of arguments to oveload::constant now elicits a warning.
2895 Odd number of elements to in anonymous hash now elicits a warning.
2899 The various "opened only for", "on closed", "never opened" warnings
2900 drop the C<main::> prefix for filehandles in the C<main> package,
2901 for example C<STDIN> instead of C<main::STDIN>.
2905 Subroutine prototypes are now checked more carefully, you may
2906 get warnings for example if you have used non-prototype characters.
2910 If an attempt to use a (non-blessed) reference as an array index
2911 is made, a warning is given.
2915 C<push @a;> and C<unshift @a;> (with no values to push or unshift)
2916 now give a warning. This may be a problem for generated and evaled
2921 If you try to L<perlfunc/pack> a number less than 0 or larger than 255
2922 using the C<"C"> format you will get an optional warning. Similarly
2923 for the C<"c"> format and a number less than -128 or more than 127.
2927 pack C<P> format now demands an explicit size.
2931 unpack C<w> now warns of unterminated compressed integers.
2935 Warnings relating to the use of PerlIO have been added.
2939 Certain regex modifiers such as C<(?o)> make sense only if applied to
2940 the entire regex. You will get an optional warning if you try to do
2945 Variable length lookbehind has not yet been implemented, trying to
2946 use it will tell that.
2950 Using arrays or hashes as references (e.g. C<< %foo->{bar} >>
2951 has been deprecated for a while. Now you will get an optional warning.
2955 Warnings relating to the use of the new restricted hashes feature
2960 Self-ties of arrays and hashes are not supported and fatal errors
2961 will happen even at an attempt to do so.
2965 Using C<sort> in scalar context now issues an optional warning.
2966 This didn't do anything useful, as the sort was not performed.
2970 Using the /g modifier in split() is meaningless and will cause a warning.
2974 Using splice() past the end of an array now causes a warning.
2978 Malformed Unicode encodings (UTF-8 and UTF-16) cause a lot of warnings,
2979 ad doestrying to use UTF-16 surrogates (which are unimplemented).
2983 Trying to use Unicode characters on an I/O stream without marking the
2984 stream's encoding (using open() or binmode()) will cause "Wide character"
2989 Use of v-strings in use/require causes a (backward) portability warning.
2993 Warnings relating to the use interpreter threads and their shared data
2998 =head1 Changed Internals
3004 PerlIO is now the default.
3008 perlapi.pod (a companion to perlguts) now attempts to document the
3013 You can now build a really minimal perl called microperl.
3014 Building microperl does not require even running Configure;
3015 C<make -f Makefile.micro> should be enough. Beware: microperl makes
3016 many assumptions, some of which may be too bold; the resulting
3017 executable may crash or otherwise misbehave in wondrous ways.
3018 For careful hackers only.
3022 Added rsignal(), whichsig(), do_join(), op_clear, op_null,
3023 ptr_table_clear(), ptr_table_free(), sv_setref_uv(), and several UTF-8
3024 interfaces to the publicised API. For the full list of the available
3025 APIs see L<perlapi>.
3029 Made possible to propagate customised exceptions via croak()ing.
3033 Now xsubs can have attributes just like subs. (Well, at least the
3034 built-in attributes.)
3038 dTHR and djSP have been obsoleted; the former removed (because it's
3039 a no-op) and the latter replaced with dSP.
3043 PERL_OBJECT has been completely removed.
3047 The MAGIC constants (e.g. C<'P'>) have been macrofied
3048 (e.g. C<PERL_MAGIC_TIED>) for better source code readability
3049 and maintainability.
3053 The regex compiler now maintains a structure that identifies nodes in
3054 the compiled bytecode with the corresponding syntactic features of the
3055 original regex expression. The information is attached to the new
3056 C<offsets> member of the C<struct regexp>. See L<perldebguts> for more
3057 complete information.
3061 The C code has been made much more C<gcc -Wall> clean. Some warning
3062 messages still remain in some platforms, so if you are compiling with
3063 gcc you may see some warnings about dubious practices. The warnings
3064 are being worked on.
3068 F<perly.c>, F<sv.c>, and F<sv.h> have now been extensively commented.
3072 Documentation on how to use the Perl source repository has been added
3073 to F<Porting/repository.pod>.
3077 There are now several profiling make targets.
3081 =head1 Security Vulnerability Closed [561]
3083 (This change was already made in 5.7.0 but bears repeating here.)
3084 (5.7.0 came out before 5.6.1: the development branch 5.7 released
3085 earlier than the maintenance branch 5.6)
3087 A potential security vulnerability in the optional suidperl component
3088 of Perl was identified in August 2000. suidperl is neither built nor
3089 installed by default. As of November 2001 the only known vulnerable
3090 platform is Linux, most likely all Linux distributions. CERT and
3091 various vendors and distributors have been alerted about the vulnerability.
3092 See http://www.cpan.org/src/5.0/sperl-2000-08-05/sperl-2000-08-05.txt
3093 for more information.
3095 The problem was caused by Perl trying to report a suspected security
3096 exploit attempt using an external program, /bin/mail. On Linux
3097 platforms the /bin/mail program had an undocumented feature which
3098 when combined with suidperl gave access to a root shell, resulting in
3099 a serious compromise instead of reporting the exploit attempt. If you
3100 don't have /bin/mail, or if you have 'safe setuid scripts', or if
3101 suidperl is not installed, you are safe.
3103 The exploit attempt reporting feature has been completely removed from
3104 Perl 5.8.0 (and the maintenance release 5.6.1, and it was removed also
3105 from all the Perl 5.7 releases), so that particular vulnerability
3106 isn't there anymore. However, further security vulnerabilities are,
3107 unfortunately, always possible. The suidperl functionality is most
3108 probably going to be removed in Perl 5.10. In any case, suidperl
3109 should only be used by security experts who know exactly what they are
3110 doing and why they are using suidperl instead of some other solution
3111 such as sudo ( see http://www.courtesan.com/sudo/ ).
3115 Several new tests have been added, especially for the F<lib> and
3116 F<ext> subsections. There are now about 69 000 individual tests
3117 (spread over about 700 test scripts), in the regression suite (5.6.1
3118 has about 11 700 tests, in 258 test scripts) The exact numbers depend
3119 on the platform and Perl configuration used. Many of the new tests
3120 are of course introduced by the new modules, but still in general Perl
3121 is now more thoroughly tested.
3123 Because of the large number of tests, running the regression suite
3124 will take considerably longer time than it used to: expect the suite
3125 to take up to 4-5 times longer to run than in perl 5.6. On a really
3126 fast machine you can hope to finish the suite in about 6-8 minutes
3129 The tests are now reported in a different order than in earlier Perls.
3130 (This happens because the test scripts from under t/lib have been moved
3131 to be closer to the library/extension they are testing.)
3133 =head1 Known Problems
3135 =head2 The Compiler Suite Is Still Very Experimental
3137 The compiler suite is slowly getting better but it continues to be
3138 highly experimental. Use in production environments is discouraged.
3140 =head2 Localising Tied Arrays and Hashes Is Broken
3144 doesn't work as one would expect: the old value is restored
3145 incorrectly. This will be changed in a future release, but we don't
3146 know yet what the new semantics will exactly be. In any case, the
3147 change will break existing code that relies on the current
3148 (ill-defined) semantics, so just avoid doing this in general.
3150 =head2 Building Extensions Can Fail Because Of Largefiles
3152 Some extensions like mod_perl are known to have issues with
3153 `largefiles', a change brought by Perl 5.6.0 in which file offsets
3154 default to 64 bits wide, where supported. Modules may fail to compile
3155 at all, or they may compile and work incorrectly. Currently, there
3156 is no good solution for the problem, but Configure now provides
3157 appropriate non-largefile ccflags, ldflags, libswanted, and libs
3158 in the %Config hash (e.g., $Config{ccflags_nolargefiles}) so the
3159 extensions that are having problems can try configuring themselves
3160 without the largefileness. This is admittedly not a clean solution,
3161 and the solution may not even work at all. One potential failure is
3162 whether one can (or, if one can, whether it's a good idea to) link
3163 together at all binaries with different ideas about file offsets;
3164 all this is platform-dependent.
3166 =head2 Modifying $_ Inside for(..)
3170 works without complaint. It shouldn't. (You should be able to
3171 modify only lvalue elements inside the loops.) You can see the
3172 correct behaviour by replacing the 1..5 with 1, 2, 3, 4, 5.
3174 =head2 mod_perl 1.26 Doesn't Build With Threaded Perl
3176 Use mod_perl 1.27 or higher.
3178 =head2 lib/ftmp-security tests warn 'system possibly insecure'
3180 Don't panic. Read the 'make test' section of INSTALL instead.
3182 =head2 libwww-perl (LWP) fails base/date #51
3184 Use libwww-perl 5.65 or later.
3186 =head2 PDL failing some tests
3188 Use PDL 2.3.4 or later.
3192 You may get errors like 'Undefined symbol "Perl_get_sv"' or "can't
3193 resolve symbol 'Perl_get_sv'", or the symbol may be "Perl_sv_2pv".
3194 This probably means that you are trying to use an older shared Perl
3195 library (or extensions linked with such) with Perl 5.8.0 executable.
3196 Perl used to have such a subroutine, but that is no more the case.
3197 Check your shared library path, and any shared Perl libraries in those
3200 Sometimes this problem may also indicate a partial Perl 5.8.0
3201 installation, see L</"Mac OS X dyld undefined symbols"> for an
3202 example and how to deal with it.
3204 =head2 Self-tying Problems
3206 Self-tying of arrays and hashes is broken in rather deep and
3207 hard-to-fix ways. As a stop-gap measure to avoid people from getting
3208 frustrated at the mysterious results (core dumps, most often), it is
3209 forbidden for now (you will get a fatal error even from an attempt).
3211 A change to self-tying of globs has caused them to be recursively
3212 referenced (see: L<perlobj/"Two-Phased Garbage Collection">). You
3213 will now need an explicit untie to destroy a self-tied glob. This
3214 behaviour may be fixed at a later date.
3216 Self-tying of scalars and IO thingies works.
3218 =head2 ext/threads/t/libc
3220 If this test fails, it indicates that your libc (C library) is not
3221 threadsafe. This particular test stress tests the localtime() call to
3222 find out whether it is threadsafe. See L<perlthrtut> for more information.
3224 =head2 Failure of Thread (5.005-style) tests
3226 B<Note that support for 5.005-style threading is deprecated,
3227 experimental and practically unsupported. In 5.10, it is expected
3228 to be removed. You should migrate your code to ithreads.>
3230 The following tests are known to fail due to fundamental problems in
3231 the 5.005 threading implementation. These are not new failures--Perl
3232 5.005_0x has the same bugs, but didn't have these tests.
3234 ../ext/B/t/xref.t 255 65280 14 12 85.71% 3-14
3235 ../ext/List/Util/t/first.t 255 65280 7 4 57.14% 2 5-7
3236 ../lib/English.t 2 512 54 2 3.70% 2-3
3237 ../lib/FileCache.t 5 1 20.00% 5
3238 ../lib/Filter/Simple/t/data.t 6 3 50.00% 1-3
3239 ../lib/Filter/Simple/t/filter_only. 9 3 33.33% 1-2 5
3240 ../lib/Math/BigInt/t/bare_mbf.t 1627 4 0.25% 8 11 1626-1627
3241 ../lib/Math/BigInt/t/bigfltpm.t 1629 4 0.25% 10 13 1628-
3243 ../lib/Math/BigInt/t/sub_mbf.t 1633 4 0.24% 8 11 1632-1633
3244 ../lib/Math/BigInt/t/with_sub.t 1628 4 0.25% 9 12 1627-1628
3245 ../lib/Tie/File/t/31_autodefer.t 255 65280 65 32 49.23% 34-65
3246 ../lib/autouse.t 10 1 10.00% 4
3247 op/flip.t 15 1 6.67% 15
3249 These failures are unlikely to get fixed as 5.005-style threads
3250 are considered fundamentally broken. (Basically what happens is that
3251 competing threads can corrupt shared global state, one good example
3252 being regular expression engine's state.)
3254 =head2 Timing problems
3256 The following tests may fail intermittently because of timing
3257 problems, for example if the system is heavily loaded.
3260 ext/Time/HiRes/HiRes.t
3262 lib/Memoize/t/expmod_t.t
3263 lib/Memoize/t/speed.t
3265 In case of failure please try running them manually, for example
3267 ./perl -Ilib ext/Time/HiRes/HiRes.t
3269 =head2 Tied/Magical Array/Hash Elements Do Not Autovivify
3271 For normal arrays C<$foo = \$bar[1]> will assign C<undef> to
3272 C<$bar[1]> (assuming that it didn't exist before), but for
3273 tied/magical arrays and hashes such autovivification does not happen
3274 because there is currently no way to catch the reference creation.
3275 The same problem affects slicing over non-existent indices/keys of
3276 a tied/magical array/hash.
3278 =head2 Unicode in package/class and subroutine names does not work
3280 One can have Unicode in identifier names, but not in package/class or
3281 subroutine names. While some limited functionality towards this does
3282 exist as of Perl 5.8.0, that is more accidental than designed; use of
3283 Unicode for the said purposes is unsupported.
3285 One reason of this unfinishedness is its (currently) inherent
3286 unportability: since both package names and subroutine names may
3287 need to be mapped to file and directory names, the Unicode capability
3288 of the filesystem becomes important-- and there unfortunately aren't
3291 =head1 Platform Specific Problems
3299 If using the AIX native make command, instead of just "make" issue
3300 "make all". In some setups the former has been known to spuriously
3301 also try to run "make install". Alternatively, you may want to use
3306 In AIX 4.2, Perl extensions that use C++ functions that use statics
3307 may have problems in that the statics are not getting initialized.
3308 In newer AIX releases, this has been solved by linking Perl with
3309 the libC_r library, but unfortunately in AIX 4.2 the said library
3310 has an obscure bug where the various functions related to time
3311 (such as time() and gettimeofday()) return broken values, and
3312 therefore in AIX 4.2 Perl is not linked against libC_r.
3316 vac 5.0.0.0 May Produce Buggy Code For Perl
3318 The AIX C compiler vac version 5.0.0.0 may produce buggy code,
3319 resulting in a few random tests failing when run as part of "make
3320 test", but when the failing tests are run by hand, they succeed.
3321 We suggest upgrading to at least vac version 5.0.1.0, that has been
3322 known to compile Perl correctly. "lslpp -L|grep vac.C" will tell
3323 you the vac version. See README.aix.
3327 If building threaded Perl, you may get compilation warning from pp_sys.c:
3329 "pp_sys.c", line 4651.39: 1506-280 (W) Function argument assignment between types "unsigned char*" and "const void*" is not allowed.
3331 This is harmless; it is caused by the getnetbyaddr() and getnetbyaddr_r()
3332 having slightly different types for their first argument.
3336 =head2 Alpha systems with old gccs fail several tests
3338 If you see op/pack, op/pat, op/regexp, or ext/Storable tests failing
3339 in a Linux/alpha or *BSD/Alpha, it's probably time to upgrade your gcc.
3340 gccs prior to 2.95.3 are definitely not good enough, and gcc 3.1 may
3341 be even better. (RedHat Linux/alpha with gcc 3.1 reported no problems,
3342 as did Linux 2.4.18 with gcc 2.95.4.) (In Tru64, it is preferable to
3343 use the bundled C compiler.)
3347 Perl 5.8.0 doesn't build in AmigaOS. It broke at some point during
3348 the ithreads work and we could not find Amiga experts to unbreak the
3349 problems. Perl 5.6.1 still works for AmigaOS (as does the the 5.7.2
3350 development release).
3354 The following tests fail on 5.8.0 Perl in BeOS Personal 5.03:
3356 t/op/lfs............................FAILED at test 17
3357 t/op/magic..........................FAILED at test 24
3358 ext/Fcntl/t/syslfs..................FAILED at test 17
3359 ext/File/Glob/t/basic...............FAILED at test 3
3360 ext/POSIX/t/sigaction...............FAILED at test 13
3361 ext/POSIX/t/waitpid.................FAILED at test 1
3363 See L<perlbeos> (README.beos) for more details.
3365 =head2 Cygwin "unable to remap"
3367 For example when building the Tk extension for Cygwin,
3368 you may get an error message saying "unable to remap".
3369 This is known problem with Cygwin, and a workaround is
3370 detailed in here: http://sources.redhat.com/ml/cygwin/2001-12/msg00894.html
3372 =head2 Cygwin ndbm tests fail on FAT
3374 One can build but not install (or test the build of) the NDBM_File
3375 on FAT filesystems. Installation (or build) on NTFS works fine.
3376 If one attempts the test on a FAT install (or build) the following
3377 failures are expected:
3379 ../ext/NDBM_File/ndbm.t 13 3328 71 59 83.10% 1-2 4 16-71
3380 ../ext/ODBM_File/odbm.t 255 65280 ?? ?? % ??
3381 ../lib/AnyDBM_File.t 2 512 12 2 16.67% 1 4
3382 ../lib/Memoize/t/errors.t 0 139 11 5 45.45% 7-11
3383 ../lib/Memoize/t/tie_ndbm.t 13 3328 4 4 100.00% 1-4
3384 run/fresh_perl.t 97 1 1.03% 91
3386 NDBM_File fails and ODBM_File just coredumps.
3388 =head2 DJGPP does not build
3390 Unfortunately DJGPP build broke somewhere after 5.7.3.
3392 =head2 FreeBSD built with ithreads coredumps reading large directories
3394 This is a known bug in FreeBSD 4.5's readdir_r(), it has been fixed in
3395 FreeBSD 4.6 (see L<perlfreebsd> (README.freebsd)).
3397 =head2 FreeBSD Failing locale Test 117 For ISO 8859-15 Locales
3399 The ISO 8859-15 locales may fail the locale test 117 in FreeBSD.
3400 This is caused by the characters \xFF (y with diaeresis) and \xBE
3401 (Y with diaeresis) not behaving correctly when being matched
3402 case-insensitively. Apparently this problem has been fixed in
3403 the latest FreeBSD releases.
3404 ( http://www.freebsd.org/cgi/query-pr.cgi?pr=34308 )
3406 =head2 IRIX fails ext/List/Util/t/shuffle.t or Digest::MD5
3408 IRIX with MIPSpro 7.3.1.2m or 7.3.1.3m compiler may fail the List::Util
3409 test ext/List/Util/t/shuffle.t by dumping core. This seems to be
3410 a compiler error since if compiled with gcc no core dump ensues, and
3411 no failures have been seen on the said test on any other platform.
3413 Similarly, building the Digest::MD5 extension has been
3414 known to fail with "*** Termination code 139 (bu21)".
3416 The cure is to drop optimization level (Configure -Doptimize=-O2).
3418 =head2 HP-UX lib/posix Subtest 9 Fails When LP64-Configured
3420 If perl is configured with -Duse64bitall, the successful result of the
3421 subtest 10 of lib/posix may arrive before the successful result of the
3422 subtest 9, which confuses the test harness so much that it thinks the
3425 =head2 Linux with glibc 2.2.5 fails t/op/int subtest #6 with -Duse64bitint
3427 This is a known bug in the glibc 2.2.5 with long long integers.
3428 ( http://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=65612 )
3430 =head2 Linux With Sfio Fails op/misc Test 48
3436 Please remember to set your environment variable LC_ALL to "C"
3437 (setenv LC_ALL C) before running "make test" to avoid a lot of
3438 warnings about the broken locales of Mac OS X.
3440 The following tests are known to fail in Mac OS X 10.1.5 because of
3441 buggy (old) implementations of Berkeley DB included in Mac OS X:
3443 Failed Test Stat Wstat Total Fail Failed List of Failed
3444 -------------------------------------------------------------------------
3445 ../ext/DB_File/t/db-btree.t 0 11 ?? ?? % ??
3446 ../ext/DB_File/t/db-recno.t 149 3 2.01% 61 63 65
3448 If you are building on a UFS partition, you will also probably see
3449 t/op/stat.t subtest #9 fail. This is caused by Darwin's UFS not
3450 supporting inode change time.
3452 Also the ext/POSIX/t/posix.t subtest #10 fails but it is skipped for
3453 now because the failure is Apple's fault, not Perl's (blocked signals
3456 If you Configure with ithreads, ext/threads/t/libc.t will fail. Again,
3457 this is not Perl's fault-- the libc of Mac OS X is not threadsafe
3458 (in this particular test, the localtime() call is found to be
3461 =head2 Mac OS X dyld undefined symbols
3463 If after installing Perl 5.8.0 you are getting warnings about missing
3464 symbols, for example
3466 dyld: perl Undefined symbols
3470 you probably have an old pre-Perl-5.8.0 installation (or parts of one)
3471 in /Library/Perl (the undefined symbols used to exist in pre-5.8.0 Perls).
3472 It seems that for some reason "make install" doesn't always completely
3473 overwrite the files in /Library/Perl. You can remove the Perl
3474 shared library like this:
3476 rm /Library/Perl/darwin/CORE/libperl.dylib
3478 and then reissue "make install". Note that the above of course is
3479 extremely disruptive for anything using the /usr/local/bin/perl.
3480 If that doesn't help, you may have to try removing all the .bundle
3481 files from beneath /Library/Perl, and again "make install"-ing.
3483 =head2 OS/2 Test Failures
3485 The following tests are known to fail on OS/2 (for clarity
3486 only the failures are shown, not the full error messages):
3488 t/io/utf8............................FAILED at test 19
3489 t/op/grent...........................FAILED at test 2
3490 t/op/pwent...........................FAILED at test 1
3491 t/lib/os2_base.......................FAILED at test 13
3492 t/lib/os2_process....................FAILED at test 10
3493 t/lib/os2_process_kid................FAILED at test 10
3494 t/lib/rx_cmprt.......................FAILED at test 16
3495 ext/DB_File/t/db-btree...............FAILED at test 0
3496 ext/DB_File/t/db-hash................FAILED at test 0
3497 ext/DB_File/t/db-recno...............FAILED at test 0
3498 lib/ExtUtils/t/basic.................FAILED at test 14
3499 lib/ExtUtils/t/Constant..............FAILED at test 4
3500 lib/Memoize/t/errors.................FAILED at test 4
3502 =head2 op/sprintf tests 91, 129, and 130
3504 The op/sprintf tests 91, 129, and 130 are known to fail on some platforms.
3505 Examples include any platform using sfio, and Compaq/Tandem's NonStop-UX.
3507 Test 91 is known to fail on QNX6 (nto), because C<sprintf '%e',0>
3508 incorrectly produces C<0.000000e+0> instead of C<0.000000e+00>.
3510 For tests 129 and 130, the failing platforms do not comply with
3511 the ANSI C Standard: lines 19ff on page 134 of ANSI X3.159 1989, to
3512 be exact. (They produce something other than "1" and "-1" when
3513 formatting 0.6 and -0.6 using the printf format "%.0f"; most often,
3514 they produce "0" and "-0".)
3518 In case you are still using Solaris 2.5 (aka SunOS 5.5), you may
3519 experience failures (the test core dumping) in lib/locale.t.
3520 The suggested cure is to upgrade your Solaris.
3522 =head2 Solaris x86 Fails Tests With -Duse64bitint
3524 The following tests are known to fail in Solaris x86 with Perl
3525 configured to use 64 bit integers:
3527 ext/Data/Dumper/t/dumper.............FAILED at test 268
3528 ext/Devel/Peek/Peek..................FAILED at test 7
3530 =head2 SUPER-UX (NEC SX)
3532 The following tests are known to fail on SUPER-UX:
3534 op/64bitint...........................FAILED tests 29-30, 32-33, 35-36
3535 op/arith..............................FAILED tests 128-130
3536 op/pack...............................FAILED tests 25-5625
3537 op/pow................................
3538 op/taint..............................# msgsnd failed
3539 ../ext/IO/lib/IO/t/io_poll............FAILED tests 3-4
3540 ../ext/IPC/SysV/ipcsysv...............FAILED tests 2, 5-6
3541 ../ext/IPC/SysV/t/msg.................FAILED tests 2, 4-6
3542 ../ext/Socket/socketpair..............FAILED tests 12
3543 ../lib/IPC/SysV.......................FAILED tests 2, 5-6
3544 ../lib/warnings.......................FAILED tests 115-116, 118-119
3546 The op/pack failure ("Cannot compress negative numbers at op/pack.t line 126")
3547 is serious but as of yet unsolved. It points at some problems with the
3548 signedness handling of the C compiler, as do the 64bitint, arith, and pow
3549 failures. Most of the rest point at problems with SysV IPC.
3551 =head2 Term::ReadKey not working on Win32
3553 Use Term::ReadKey 2.20 or later.
3561 During Configure, the test
3563 Guessing which symbols your C compiler and preprocessor define...
3565 will probably fail with error messages like
3567 CC-20 cc: ERROR File = try.c, Line = 3
3568 The identifier "bad" is undefined.
3570 bad switch yylook 79bad switch yylook 79bad switch yylook 79bad switch yylook 79#ifdef A29K
3573 CC-65 cc: ERROR File = try.c, Line = 3
3574 A semicolon is expected at this point.
3576 This is caused by a bug in the awk utility of UNICOS/mk. You can ignore
3577 the error, but it does cause a slight problem: you cannot fully
3578 benefit from the h2ph utility (see L<h2ph>) that can be used to
3579 convert C headers to Perl libraries, mainly used to be able to access
3580 from Perl the constants defined using C preprocessor, cpp. Because of
3581 the above error, parts of the converted headers will be invisible.
3582 Luckily, these days the need for h2ph is rare.
3586 If building Perl with interpreter threads (ithreads), the
3587 getgrent(), getgrnam(), and getgrgid() functions cannot return the
3588 list of the group members due to a bug in the multithreaded support of
3589 UNICOS/mk. What this means is that in list context the functions will
3590 return only three values, not four.
3596 There are a few known test failures, see L<perluts> (README.uts).
3598 =head2 VOS (Stratus)
3600 When Perl is built using the native build process on VOS Release
3601 14.5.0 and GNU C++/GNU Tools 2.0.1, all attempted tests either
3602 pass or result in TODO (ignored) failures.
3606 There should be no reported test failures with a default configuration,
3607 though there are a number of tests marked TODO that point to areas
3608 needing further debugging and/or porting work.
3612 In multi-CPU boxes, there are some problems with the I/O buffering:
3613 some output may appear twice.
3615 =head2 XML::Parser not working
3617 Use XML::Parser 2.31 or later.
3619 =head2 z/OS (OS/390)
3621 z/OS has rather many test failures but the situation is actually much
3622 better than it was in 5.6.0; it's just that so many new modules and
3623 tests have been added.
3625 Failed Test Stat Wstat Total Fail Failed List of Failed
3626 ---------------------------------------------------------------------------
3627 ../ext/Data/Dumper/t/dumper.t 357 8 2.24% 311 314 325 327
3629 ../ext/IO/lib/IO/t/io_unix.t 5 4 80.00% 2-5
3630 ../ext/Storable/t/downgrade.t 12 3072 169 12 7.10% 14-15 46-47 78-79
3632 ../lib/ExtUtils/t/Constant.t 121 30976 48 48 100.00% 1-48
3633 ../lib/ExtUtils/t/Embed.t 9 9 100.00% 1-9
3634 op/pat.t 922 7 0.76% 665 776 785 832-
3636 op/sprintf.t 224 3 1.34% 98 100 136
3637 op/tr.t 97 5 5.15% 63 71-74
3638 uni/fold.t 780 6 0.77% 61 169 196 661
3641 The failures in dumper.t and downgrade.t are problems in the tests,
3642 those in io_unix and sprintf are problems in the USS (UDP sockets and
3643 printf formats). The pat, tr, and fold failures are genuine Perl
3644 problems caused by EBCDIC (and in the pat and fold cases, combining
3645 that with Unicode). The Constant and Embed are probably problems in
3646 the tests (since they test Perl's ability to build extensions, and
3647 that seems to be working reasonably well.)
3649 =head2 Unicode Support on EBCDIC Still Spotty
3651 Though mostly working, Unicode support still has problem spots on
3652 EBCDIC platforms. One such known spot are the C<\p{}> and C<\P{}>
3653 regular expression constructs for code points less than 256: the
3654 C<pP> are testing for Unicode code points, not knowing about EBCDIC.
3656 =head2 Seen In Perl 5.7 But Gone Now
3658 C<Time::Piece> (previously known as C<Time::Object>) was removed
3659 because it was felt that it didn't have enough value in it to be a
3660 core module. It is still a useful module, though, and is available
3663 Perl 5.8 unfortunately does not build anymore on AmigaOS; this broke
3664 accidentally at some point. Since there are not that many Amiga
3665 developers available, we could not get this fixed and tested in time
3666 for 5.8.0. Perl 5.6.1 still works for AmigaOS (as does the the 5.7.2
3667 development release).
3669 The C<PerlIO::Scalar> and C<PerlIO::Via> (capitalised) were renamed as
3670 C<PerlIO::scalar> and C<PerlIO::via> (all lowercase) just before 5.8.0.
3671 The main rationale was to have all core PerlIO layers to have all
3672 lowercase names. The "plugins" are named as usual, for example
3673 C<PerlIO::via::QuotedPrint>.
3675 The C<threads::shared::queue> and C<threads::shared::semaphore> were
3676 renamed as C<Thread::Queue> and C<Thread::Semaphore> just before 5.8.0.
3677 The main rationale was to have thread modules to obey normal naming,
3678 C<Thread::> (the C<threads> and C<threads::shared> themselves are
3679 more pragma-like, they affect compile-time, so they stay lowercase).
3681 =head1 Reporting Bugs
3683 If you find what you think is a bug, you might check the articles
3684 recently posted to the comp.lang.perl.misc newsgroup and the perl
3685 bug database at http://bugs.perl.org/ . There may also be
3686 information at http://www.perl.com/ , the Perl Home Page.
3688 If you believe you have an unreported bug, please run the B<perlbug>
3689 program included with your release. Be sure to trim your bug down
3690 to a tiny but sufficient test case. Your bug report, along with the
3691 output of C<perl -V>, will be sent off to perlbug@perl.org to be
3692 analysed by the Perl porting team.
3696 The F<Changes> file for exhaustive details on what changed.
3698 The F<INSTALL> file for how to build Perl.
3700 The F<README> file for general stuff.
3702 The F<Artistic> and F<Copying> files for copyright information.
3706 Written by Jarkko Hietaniemi <F<jhi@iki.fi>>.