3 perldelta - what is new for perl v5.8.0
7 This document describes differences between the 5.6.0 release and
10 Many of the bug fixes in 5.8.0 were already seen in the 5.6.1
11 maintenance release since the two releases were kept closely
12 coordinated (while 5.8.0 was still called 5.7.something).
14 Changes that were integrated into the 5.6.1 release are marked C<[561]>.
15 Many of these changes have been further developed since 5.6.1 was released,
16 those are marked C<[561+]>.
18 You can see the list of changes in the 5.6.1 release (both from the
19 5.005_03 release and the 5.6.0 release) by reading L<perl561delta>.
21 =head1 Highlights In 5.8.0
27 Better Unicode support
35 New Thread Implementation
39 Better Numeric Accuracy
51 More Extensive Regression Testing
55 =head1 Incompatible Changes
57 =head2 Binary Incompatibility
59 B<Perl 5.8 is not binary compatible with earlier releases of Perl.>
61 B<You have to recompile your XS modules.>
63 (Pure Perl modules should continue to work.)
65 The major reason for the discontinuity is the new IO architecture
66 called PerlIO. PerlIO is the default configuration because without
67 it many new features of Perl 5.8 cannot be used. In other words:
68 you just have to recompile your modules containing XS code, sorry
71 In future releases of Perl, non-PerlIO aware XS modules may become
72 completely unsupported. This shouldn't be too difficult for module
73 authors, however: PerlIO has been designed as a drop-in replacement
74 (at the source code level) for the stdio interface.
76 Depending on your platform, there are also other reasons why
77 we decided to break binary compatibility, please read on.
79 =head2 64-bit platforms and malloc
81 If your pointers are 64 bits wide, the Perl malloc is no longer being
82 used because it does not work well with 8-byte pointers. Also,
83 usually the system mallocs on such platforms are much better optimized
84 for such large memory models than the Perl malloc. Some memory-hungry
85 Perl applications like the PDL don't work well with Perl's malloc.
86 Finally, other applications than Perl (such as mod_perl) tend to prefer
87 the system malloc. Such platforms include Alpha and 64-bit HPPA,
90 =head2 AIX Dynaloading
92 The AIX dynaloading now uses in AIX releases 4.3 and newer the native
93 dlopen interface of AIX instead of the old emulated interface. This
94 change will probably break backward compatibility with compiled
95 modules. The change was made to make Perl more compliant with other
96 applications like mod_perl which are using the AIX native interface.
98 =head2 Attributes for C<my> variables now handled at run-time
100 The C<my EXPR : ATTRS> syntax now applies variable attributes at
101 run-time. (Subroutine and C<our> variables still get attributes applied
102 at compile-time.) See L<attributes> for additional details. In particular,
103 however, this allows variable attributes to be useful for C<tie> interfaces,
104 which was a deficiency of earlier releases. Note that the new semantics
105 doesn't work with the Attribute::Handlers module (as of version 0.76).
107 =head2 Socket Extension Dynamic in VMS
109 The Socket extension is now dynamically loaded instead of being
110 statically built in. This may or may not be a problem with ancient
111 TCP/IP stacks of VMS: we do not know since we weren't able to test
112 Perl in such configurations.
114 =head2 IEEE-format Floating Point Default on OpenVMS Alpha
116 Perl now uses IEEE format (T_FLOAT) as the default internal floating
117 point format on OpenVMS Alpha, potentially breaking binary compatibility
118 with external libraries or existing data. G_FLOAT is still available as
119 a configuration option. The default on VAX (D_FLOAT) has not changed.
121 =head2 New Unicode Semantics (no more C<use utf8>)
123 Previously in Perl 5.6 to use Unicode one would say "use utf8" and
124 then the operations (like string concatenation) were Unicode-aware
125 in that lexical scope.
127 This was found to be an inconvenient interface, and in Perl 5.8 the
128 Unicode model has completely changed: now the "Unicodeness" is bound
129 to the data itself, and for most of the time "use utf8" is not needed
130 at all. The only remaining use of "use utf8" is when the Perl script
131 itself has been written in the UTF-8 encoding of Unicode.
133 See L<perluniintro> for the explanation of the current model.
135 =head2 New Unicode Properties
137 Unicode I<scripts> are now supported. Scripts are similar to (and superior
138 to) Unicode I<blocks>. The difference between scripts and blocks is that
139 scripts are the glyphs used by a language or a group of languages, while
140 the blocks are more artificial groupings of (mostly) 256 characters based
141 on the Unicode numbering.
143 In general, scripts are more inclusive, but not universally so. For
144 example, while the script C<Latin> includes all the Latin characters and
145 their various diacritic-adorned versions, it does not include the various
146 punctuation or digits (since they are not solely C<Latin>).
148 A number of other properties are now supported, including C<\p{L&}>,
149 C<\p{Any}> C<\p{Assigned}>, C<\p{Unassigned}>, C<\p{Blank}> [561] and
150 C<\p{SpacePerl}> [561] (along with their C<\P{...}> versions, of course).
151 See L<perlunicode> for details, and more additions.
153 The C<In> or C<Is> prefix to names used with the C<\p{...}> and C<\P{...}>
154 are now almost always optional. The only exception is that a C<In> prefix
155 is required to signify a Unicode block when a block name conflicts with a
156 script name. For example, C<\p{Tibetan}> refers to the script, while
157 C<\p{InTibetan}> refers to the block. When there is no name conflict, you
158 can omit the C<In> from the block name (e.g. C<\p{BraillePatterns}>), but
159 to be safe, it's probably best to always use the C<In>).
161 =head2 REF(...) Instead Of SCALAR(...)
163 A reference to a reference now stringifies as "REF(0x81485ec)" instead
164 of "SCALAR(0x81485ec)" in order to be more consistent with the return
167 =head2 pack/unpack D/F recycled
169 The undocumented pack/unpack template letters D/F have been recycled
170 for better use: now they stand for long double (if supported by the
171 platform) and NV (Perl internal floating point type). (They used
172 to be aliases for d/f, but you never knew that.)
174 =head2 glob() now returns filenames in alphabetical order
176 The list of filenames from glob() (or <...>) is now by default sorted
177 alphabetically to be csh-compliant (which is what happened before
178 in most UNIX platforms). (bsd_glob() does still sort platform
179 natively, ASCII or EBCDIC, unless GLOB_ALPHASORT is specified.) [561]
187 The semantics of bless(REF, REF) were unclear and until someone proves
188 it to make some sense, it is forbidden.
192 The obsolete chat2 library that should never have been allowed
193 to escape the laboratory has been decommissioned.
197 The builtin dump() function has probably outlived most of its
198 usefulness. The core-dumping functionality will remain in future
199 available as an explicit call to C<CORE::dump()>, but in future
200 releases the behaviour of an unqualified C<dump()> call may change.
204 The very dusty examples in the eg/ directory have been removed.
205 Suggestions for new shiny examples welcome but the main issue is that
206 the examples need to be documented, tested and (most importantly)
211 The (bogus) escape sequences \8 and \9 now give an optional warning
212 ("Unrecognized escape passed through"). There is no need to \-escape
217 The *glob{FILEHANDLE} is deprecated, use *glob{IO} instead.
221 The C<package;> syntax (C<package> without an argument) has been
222 deprecated. Its semantics were never that clear and its
223 implementation even less so. If you have used that feature to
224 disallow all but fully qualified variables, C<use strict;> instead.
228 The unimplemented POSIX regex features [[.cc.]] and [[=c=]] are still
229 recognised but now cause fatal errors. The previous behaviour of
230 ignoring them by default and warning if requested was unacceptable
231 since it, in a way, falsely promised that the features could be used.
235 In future releases, non-PerlIO aware XS modules may become completely
236 unsupported. Since PerlIO is a drop-in replacement for stdio at the
237 source code level, this shouldn't be that drastic a change.
241 Previous versions of perl and some readings of some sections of Camel
242 III implied that the C<:raw> "discipline" was the inverse of C<:crlf>.
243 Turning off "clrfness" is no longer enough to make a stream truly
244 binary. So the PerlIO C<:raw> layer (or "discipline", to use the Camel
245 book's older terminology) is now formally defined as being equivalent
246 to binmode(FH) - which is in turn defined as doing whatever is
247 necessary to pass each byte as-is without any translation. In
248 particular binmode(FH) - and hence C<:raw> - will now turn off both
249 CRLF and UTF-8 translation and remove other layers (e.g. :encoding())
250 which would modify byte stream.
254 The current user-visible implementation of pseudo-hashes (the weird
255 use of the first array element) is deprecated starting from Perl 5.8.0
256 and will be removed in Perl 5.10.0, and the feature will be
257 implemented differently. Not only is the current interface rather
258 ugly, but the current implementation slows down normal array and hash
259 use quite noticeably. The C<fields> pragma interface will remain
260 available. The I<restricted hashes> interface is expected to
261 be the replacement interface (see L<Hash::Util>). If your existing
262 programs depends on the underlying implementation, consider using
263 L<Class::PseudoHash> from CPAN.
267 The syntaxes C<< @a->[...] >> and C<< %h->{...} >> have now been deprecated.
271 After years of trying, suidperl is considered to be too complex to
272 ever be considered truly secure. The suidperl functionality is likely
273 to be removed in a future release.
277 The 5.005 threads model (module C<Thread>) is deprecated and expected
278 to be removed in Perl 5.10. Multithreaded code should be migrated to
279 the new ithreads model (see L<threads>, L<threads::shared> and
284 The long deprecated uppercase aliases for the string comparison
285 operators (EQ, NE, LT, LE, GE, GT) have now been removed.
289 The tr///C and tr///U features have been removed and will not return;
290 the interface was a mistake. Sorry about that. For similar
291 functionality, see pack('U0', ...) and pack('C0', ...). [561]
295 Earlier Perls treated "sub foo (@bar)" as equivalent to "sub foo (@)".
296 The prototypes are now checked better at compile-time for invalid
297 syntax. An optional warning is generated ("Illegal character in
298 prototype...") but this may be upgraded to a fatal error in a future
303 The C<exec LIST> and C<system LIST> operations now produce warnings on
304 tainted data and in some future release they will produce fatal errors.
308 The existing behaviour when localising tied arrays and hashes is wrong,
309 and will be changed in a future release, so do not rely on the existing
310 behaviour. See L<"Localising Tied Arrays and Hashes Is Broken">.
314 =head1 Core Enhancements
316 =head2 Unicode Overhaul
318 Unicode in general should be now much more usable than in Perl 5.6.0
319 (or even in 5.6.1). Unicode can be used in hash keys, Unicode in
320 regular expressions should work now, Unicode in tr/// should work now,
321 Unicode in I/O should work now. See L<perluniintro> for introduction
322 and L<perlunicode> for details.
328 The Unicode Character Database coming with Perl has been upgraded
329 to Unicode 3.2.0. For more information, see http://www.unicode.org/ .
330 [561+] (5.6.1 has UCD 3.0.1.)
334 For developers interested in enhancing Perl's Unicode capabilities:
335 almost all the UCD files are included with the Perl distribution in
336 the F<lib/unicore> subdirectory. The most notable omission, for space
337 considerations, is the Unihan database.
341 The properties \p{Blank} and \p{SpacePerl} have been added. "Blank" is like
342 C isblank(), that is, it contains only "horizontal whitespace" (the space
343 character is, the newline isn't), and the "SpacePerl" is the Unicode
344 equivalent of C<\s> (\p{Space} isn't, since that includes the vertical
345 tabulator character, whereas C<\s> doesn't.)
347 See "New Unicode Properties" earlier in this document for additional
348 information on changes with Unicode properties.
352 =head2 PerlIO is Now The Default
358 IO is now by default done via PerlIO rather than system's "stdio".
359 PerlIO allows "layers" to be "pushed" onto a file handle to alter the
360 handle's behaviour. Layers can be specified at open time via 3-arg
363 open($fh,'>:crlf :utf8', $path) || ...
365 or on already opened handles via extended C<binmode>:
367 binmode($fh,':encoding(iso-8859-7)');
369 The built-in layers are: unix (low level read/write), stdio (as in
370 previous Perls), perlio (re-implementation of stdio buffering in a
371 portable manner), crlf (does CRLF <=> "\n" translation as on Win32,
372 but available on any platform). A mmap layer may be available if
373 platform supports it (mostly UNIXes).
375 Layers to be applied by default may be specified via the 'open' pragma.
377 See L</"Installation and Configuration Improvements"> for the effects
378 of PerlIO on your architecture name.
382 If your platform supports fork(), you can use the list form of C<open>
383 for pipes. For example:
385 open KID_PS, "-|", "ps", "aux" or die $!;
387 forks the ps(1) command (without spawning a shell, as there are more
388 than three arguments to open()), and reads its standard output via the
389 C<KID_PS> filehandle. See L<perlipc>.
393 File handles can be marked as accepting Perl's internal encoding of Unicode
394 (UTF-8 or UTF-EBCDIC depending on platform) by a pseudo layer ":utf8" :
396 open($fh,">:utf8","Uni.txt");
398 Note for EBCDIC users: the pseudo layer ":utf8" is erroneously named
399 for you since it's not UTF-8 what you will be getting but instead
400 UTF-EBCDIC. See L<perlunicode>, L<utf8>, and
401 http://www.unicode.org/unicode/reports/tr16/ for more information.
402 In future releases this naming may change. See L<perluniintro>
403 for more information about UTF-8.
407 If your environment variables (LC_ALL, LC_CTYPE, LANG, LANGUAGE) look
408 like you want to use UTF-8 (any of the the variables match C</utf-?8/i>),
409 your STDIN, STDOUT, STDERR handles and the default open layer
410 (see L<open>) are marked as UTF-8. (This feature, like other new
411 features that combine Unicode and I/O, work only if you are using
412 PerlIO, but that's the default.)
414 Note that after this Perl really does assume that everything is UTF-8:
415 for example if some input handle is not, Perl will probably very soon
416 complain about the input data like this "Malformed UTF-8 ..." since
417 any old eight-bit data is not legal UTF-8.
419 Note for code authors: if you want to enable your users to use UTF-8
420 as their default encoding but in your code still have eight-bit I/O streams
421 (such as images or zip files), you need to explicitly open() or binmode()
422 with C<:bytes> (see L<perlfunc/open> and L<perlfunc/binmode>), or you
423 can just use C<binmode(FH)> (nice for pre-5.8.0 backward compatibility).
427 File handles can translate character encodings from/to Perl's internal
428 Unicode form on read/write via the ":encoding()" layer.
432 File handles can be opened to "in memory" files held in Perl scalars via:
434 open($fh,'>', \$variable) || ...
438 Anonymous temporary files are available without need to
439 'use FileHandle' or other module via
441 open($fh,"+>", undef) || ...
443 That is a literal undef, not an undefined value.
449 The new interpreter threads ("ithreads" for short) implementation of
450 multithreading, by Arthur Bergman, replaces the old "5.005 threads"
451 implementation. In the ithreads model any data sharing between
452 threads must be explicit, as opposed to the model where data sharing
453 was implicit. See L<threads> and L<threads::shared>, and
456 As a part of the ithreads implementation Perl will also use
457 any necessary and detectable reentrant libc interfaces.
459 =head2 Restricted Hashes
461 A restricted hash is restricted to a certain set of keys, no keys
462 outside the set can be added. Also individual keys can be restricted
463 so that the key cannot be deleted and the value cannot be changed.
464 No new syntax is involved: the Hash::Util module is the interface.
468 Perl used to be fragile in that signals arriving at inopportune moments
469 could corrupt Perl's internal state. Now Perl postpones handling of
470 signals until it's safe (between opcodes).
472 This change may have surprising side effects because signals no longer
473 interrupt Perl instantly. Perl will now first finish whatever it was
474 doing, like finishing an internal operation (like sort()) or an
475 external operation (like an I/O operation), and only then look at any
476 arrived signals (and before starting the next operation). No more corrupt
477 internal state since the current operation is always finished first,
478 but the signal may take more time to get heard. Note that breaking
479 out from potentially blocking operations should still work, though.
481 =head2 Understanding of Numbers
483 In general a lot of fixing has happened in the area of Perl's
484 understanding of numbers, both integer and floating point. Since in
485 many systems the standard number parsing functions like C<strtoul()>
486 and C<atof()> seem to have bugs, Perl tries to work around their
487 deficiencies. This results hopefully in more accurate numbers.
489 Perl now tries internally to use integer values in numeric conversions
490 and basic arithmetics (+ - * /) if the arguments are integers, and
491 tries also to keep the results stored internally as integers.
492 This change leads to often slightly faster and always less lossy
493 arithmetics. (Previously Perl always preferred floating point numbers
496 =head2 Arrays now always interpolate into double-quoted strings [561]
498 In double-quoted strings, arrays now interpolate, no matter what. The
499 behavior in earlier versions of perl 5 was that arrays would interpolate
500 into strings if the array had been mentioned before the string was
501 compiled, and otherwise Perl would raise a fatal compile-time error.
502 In versions 5.000 through 5.003, the error was
504 Literal @example now requires backslash
506 In versions 5.004_01 through 5.6.0, the error was
508 In string, @example now must be written as \@example
510 The idea here was to get people into the habit of writing
511 C<"fred\@example.com"> when they wanted a literal C<@> sign, just as
512 they have always written C<"Give me back my \$5"> when they wanted a
515 Starting with 5.6.1, when Perl now sees an C<@> sign in a
516 double-quoted string, it I<always> attempts to interpolate an array,
517 regardless of whether or not the array has been used or declared
518 already. The fatal error has been downgraded to an optional warning:
520 Possible unintended interpolation of @example in string
522 This warns you that C<"fred@example.com"> is going to turn into
523 C<fred.com> if you don't backslash the C<@>.
524 See http://www.plover.com/~mjd/perl/at-error.html for more details
525 about the history here.
527 =head2 Miscellaneous Changes
533 AUTOLOAD is now lvaluable, meaning that you can add the :lvalue attribute
534 to AUTOLOAD subroutines and you can assign to the AUTOLOAD return value.
538 The $Config{byteorder} (and corresponding BYTEORDER in config.h) was
539 previously wrong in platforms if sizeof(long) was 4, but sizeof(IV)
540 was 8. The byteorder was only sizeof(long) bytes long (1234 or 4321),
541 but now it is correctly sizeof(IV) bytes long, (12345678 or 87654321).
542 (This problem didn't affect Windows platforms.)
544 Also, $Config{byteorder} is now computed dynamically--this is more
545 robust with "fat binaries" where an executable image contains binaries
546 for more than one binary platform, and when cross-compiling.
550 C<perl -d:Module=arg,arg,arg> now works (previously one couldn't pass
551 in multiple arguments.)
555 C<do> followed by a bareword now ensures that this bareword isn't
556 a keyword (to avoid a bug where C<do q(foo.pl)> tried to call a
557 subroutine called C<q>). This means that for example instead of
558 C<do format()> you must write C<do &format()>.
562 The builtin dump() now gives an optional warning
563 C<dump() better written as CORE::dump()>,
564 meaning that by default C<dump(...)> is resolved as the builtin
565 dump() which dumps core and aborts, not as (possibly) user-defined
566 C<sub dump>. To call the latter, qualify the call as C<&dump(...)>.
567 (The whole dump() feature is to considered deprecated, and possibly
568 removed/changed in future releases.)
572 chomp() and chop() are now overridable. Note, however, that their
573 prototype (as given by C<prototype("CORE::chomp")> is undefined,
574 because it cannot be expressed and therefore one cannot really write
575 replacements to override these builtins.
579 END blocks are now run even if you exit/die in a BEGIN block.
580 Internally, the execution of END blocks is now controlled by
581 PL_exit_flags & PERL_EXIT_DESTRUCT_END. This enables the new
582 behaviour for Perl embedders. This will default in 5.10. See
587 Formats now support zero-padded decimal fields.
591 Although "you shouldn't do that", it was possible to write code that
592 depends on Perl's hashed key order (Data::Dumper does this). The new
593 algorithm "One-at-a-Time" produces a different hashed key order.
594 More details are in L</"Performance Enhancements">.
598 lstat(FILEHANDLE) now gives a warning because the operation makes no sense.
599 In future releases this may become a fatal error.
603 Spurious syntax errors generated in certain situations, when glob()
604 caused File::Glob to be loaded for the first time, have been fixed. [561]
608 Lvalue subroutines can now return C<undef> in list context. However,
609 the lvalue subroutine feature still remains experimental. [561+]
613 A lost warning "Can't declare ... dereference in my" has been
614 restored (Perl had it earlier but it became lost in later releases.)
618 A new special regular expression variable has been introduced:
619 C<$^N>, which contains the most-recently closed group (submatch).
623 C<no Module;> does not produce an error even if Module does not have an
624 unimport() method. This parallels the behavior of C<use> vis-a-vis
629 The numerical comparison operators return C<undef> if either operand
630 is a NaN. Previously the behaviour was unspecified.
634 C<our> can now have an experimental optional attribute C<unique> that
635 affects how global variables are shared among multiple interpreters,
640 The following builtin functions are now overridable: each(), keys(),
641 pop(), push(), shift(), splice(), unshift(). [561]
645 C<pack() / unpack()> can now group template letters with C<()> and then
646 apply repetition/count modifiers on the groups.
650 C<pack() / unpack()> can now process the Perl internal numeric types:
651 IVs, UVs, NVs-- and also long doubles, if supported by the platform.
652 The template letters are C<j>, C<J>, C<F>, and C<D>.
656 C<pack('U0a*', ...)> can now be used to force a string to UTF8.
660 my __PACKAGE__ $obj now works. [561]
664 POSIX::sleep() now returns the number of I<unslept> seconds
665 (as the POSIX standard says), as opposed to CORE::sleep() which
666 returns the number of slept seconds.
670 The printf() and sprintf() now support parameter reordering using the
671 C<%\d+\$> and C<*\d+\$> syntaxes. For example
673 print "%2\$s %1\$s\n", "foo", "bar";
675 will print "bar foo\n". This feature helps in writing
676 internationalised software, and in general when the order
677 of the parameters can vary.
681 The (\&) prototype now works properly. [561]
685 prototype(\[$@%&]) is now available to implicitly create references
686 (useful for example if you want to emulate the tie() interface).
690 A new command-line option, C<-t> is available. It is the
691 little brother of C<-T>: instead of dying on taint violations,
692 lexical warnings are given. B<This is only meant as a temporary
693 debugging aid while securing the code of old legacy applications.
694 This is not a substitute for -T.>
698 In other taint news, the C<exec LIST> and C<system LIST> have now been
699 considered too risky (think C<exec @ARGV>: it can start any program
700 with any arguments), and now the said forms cause a warning under
701 lexical warnings. You should carefully launder the arguments to
702 guarantee their validity. In future releases of Perl the forms will
703 become fatal errors so consider starting laundering now.
707 Tied hash interfaces are now required to have the EXISTS and DELETE
708 methods (either own or inherited).
712 If tr/// is just counting characters, it doesn't attempt to
717 untie() will now call an UNTIE() hook if it exists. See L<perltie>
722 L<utime> now supports C<utime undef, undef, @files> to change the
723 file timestamps to the current time.
727 The rules for allowing underscores (underbars) in numeric constants
728 have been relaxed and simplified: now you can have an underscore
729 simply B<between digits>.
733 Rather than relying on C's argv[0] (which may not contain a full pathname)
734 where possible $^X is now set by asking the operating system.
735 (eg by reading F</proc/self/exe> on Linux, F</proc/curproc/file> on FreeBSD)
739 A new variable, C<${^TAINT}>, indicates whether taint mode is enabled.
743 You can now override the readline() builtin, and this overrides also
744 the <FILEHANDLE> angle bracket operator.
748 The command-line options -s and -F are now recognized on the shebang
753 Use of the C</c> match modifier without an accompanying C</g> modifier
754 elicits a new warning: C<Use of /c modifier is meaningless without /g>.
756 Use of C</c> in substitutions, even with C</g>, elicits
757 C<Use of /c modifier is meaningless in s///>.
759 Use of C</g> with C<split> elicits C<Use of /g modifier is meaningless
764 Support for the C<CLONE> special subroutine had been added.
765 With ithreads, when a new thread is created, all Perl data is cloned,
766 however non-Perl data cannot be cloned automatically. In C<CLONE> you
767 can do whatever you need to do, like for example handle the cloning of
768 non-Perl data, if necessary. C<CLONE> will be executed once for every
769 package that has it defined or inherited. It will be called in the
770 context of the new thread, so all modifications are made in the new area.
776 =head1 Modules and Pragmata
778 =head2 New Modules and Pragmata
784 C<Attribute::Handlers>, originally by Damian Conway and now maintained
785 by Arthur Bergman, allows a class to define attribute handlers.
788 use Attribute::Handlers;
789 sub Wolf :ATTR(SCALAR) { print "howl!\n" }
791 # later, in some package using or inheriting from MyPack...
793 my MyPack $Fluffy : Wolf; # the attribute handler Wolf will be called
795 Both variables and routines can have attribute handlers. Handlers can
796 be specific to type (SCALAR, ARRAY, HASH, or CODE), or specific to the
797 exact compilation phase (BEGIN, CHECK, INIT, or END).
798 See L<Attribute::Handlers>.
802 C<B::Concise>, by Stephen McCamant, is a new compiler backend for
803 walking the Perl syntax tree, printing concise info about ops.
804 The output is highly customisable. See L<B::Concise>. [561+]
808 The new bignum, bigint, and bigrat pragmas, by Tels, implement
809 transparent bignum support (using the Math::BigInt, Math::BigFloat,
810 and Math::BigRat backends).
814 C<Class::ISA>, by Sean Burke, is a module for reporting the search
815 path for a class's ISA tree. See L<Class::ISA>.
819 C<Cwd> now has a split personality: if possible, an XS extension is
820 used, (this will hopefully be faster, more secure, and more robust)
821 but if not possible, the familiar Perl implementation is used.
825 C<Devel::PPPort>, originally by Kenneth Albanowski and now
826 maintained by Paul Marquess, has been added. It is primarily used
827 by C<h2xs> to enhance portability of XS modules between different
828 versions of Perl. See L<Devel::PPPort>.
832 C<Digest>, frontend module for calculating digests (checksums), from
833 Gisle Aas, has been added. See L<Digest>.
837 C<Digest::MD5> for calculating MD5 digests (checksums) as defined in
838 RFC 1321, from Gisle Aas, has been added. See L<Digest::MD5>.
840 use Digest::MD5 'md5_hex';
842 $digest = md5_hex("Thirsty Camel");
844 print $digest, "\n"; # 01d19d9d2045e005c3f1b80e8b164de1
846 NOTE: the C<MD5> backward compatibility module is deliberately not
847 included since its further use is discouraged.
849 See also L<PerlIO::via::QuotedPrint>.
853 C<Encode>, originally by Nick Ing-Simmons and now maintained by Dan
854 Kogai, provides a mechanism to translate between different character
855 encodings. Support for Unicode, ISO-8859-1, and ASCII are compiled in
856 to the module. Several other encodings (like the rest of the
857 ISO-8859, CP*/Win*, Mac, KOI8-R, three variants EBCDIC, Chinese,
858 Japanese, and Korean encodings) are included and can be loaded at
859 runtime. (For space considerations, the largest Chinese encodings
860 have been separated into their own CPAN module, Encode::HanExtra,
861 which Encode will use if available). See L<Encode>.
863 Any encoding supported by Encode module is also available to the
864 ":encoding()" layer if PerlIO is used.
868 C<Hash::Util> is the interface to the new I<restricted hashes>
869 feature. (Implemented by Jeffrey Friedl, Nick Ing-Simmons, and
870 Michael Schwern.) See L<Hash::Util>.
874 C<I18N::Langinfo> can be used to query locale information.
875 See L<I18N::Langinfo>.
879 C<I18N::LangTags>, by Sean Burke, has functions for dealing with
880 RFC3066-style language tags. See L<I18N::LangTags>.
884 C<ExtUtils::Constant>, by Nicholas Clark, is a new tool for extension
885 writers for generating XS code to import C header constants.
886 See L<ExtUtils::Constant>.
890 C<Filter::Simple>, by Damian Conway, is an easy-to-use frontend to
891 Filter::Util::Call. See L<Filter::Simple>.
897 use Filter::Simple sub {
898 while (my ($from, $to) = splice @_, 0, 2) {
907 use MyFilter qr/red/ => 'green';
909 print "red\n"; # this code is filtered, will print "green\n"
910 print "bored\n"; # this code is filtered, will print "bogreen\n"
914 print "red\n"; # this code is not filtered, will print "red\n"
918 C<File::Temp>, by Tim Jenness, allows one to create temporary files
919 and directories in an easy, portable, and secure way. See L<File::Temp>.
924 C<Filter::Util::Call>, by Paul Marquess, provides you with the
925 framework to write I<source filters> in Perl. For most uses, the
926 frontend Filter::Simple is to be preferred. See L<Filter::Util::Call>.
930 C<if>, by Ilya Zakharevich, is a new pragma for conditional inclusion
935 L<libnet>, by Graham Barr, is a collection of perl5 modules related
936 to network programming. See L<Net::FTP>, L<Net::NNTP>, L<Net::Ping>
937 (not part of libnet, but related), L<Net::POP3>, L<Net::SMTP>,
940 Perl installation leaves libnet unconfigured; use F<libnetcfg>
945 C<List::Util>, by Graham Barr, is a selection of general-utility
946 list subroutines, such as sum(), min(), first(), and shuffle().
951 C<Locale::Constants>, C<Locale::Country>, C<Locale::Currency>
952 C<Locale::Language>, and L<Locale::Script>, by Neil Bowers, have
953 been added. They provide the codes for various locale standards, such
954 as "fr" for France, "usd" for US Dollar, and "ja" for Japanese.
958 $country = code2country('jp'); # $country gets 'Japan'
959 $code = country2code('Norway'); # $code gets 'no'
961 See L<Locale::Constants>, L<Locale::Country>, L<Locale::Currency>,
962 and L<Locale::Language>.
966 C<Locale::Maketext>, by Sean Burke, is a localization framework. See
967 L<Locale::Maketext>, and L<Locale::Maketext::TPJ13>. The latter is an
968 article about software localization, originally published in The Perl
969 Journal #13, and republished here with kind permission.
973 C<Math::BigRat> for big rational numbers, to accompany Math::BigInt and
974 Math::BigFloat, from Tels. See L<Math::BigRat>.
978 C<Memoize> can make your functions faster by trading space for time,
979 from Mark-Jason Dominus. See L<Memoize>.
983 C<MIME::Base64>, by Gisle Aas, allows you to encode data in base64,
984 as defined in RFC 2045 - I<MIME (Multipurpose Internet Mail
989 $encoded = encode_base64('Aladdin:open sesame');
990 $decoded = decode_base64($encoded);
992 print $encoded, "\n"; # "QWxhZGRpbjpvcGVuIHNlc2FtZQ=="
998 C<MIME::QuotedPrint>, by Gisle Aas, allows you to encode data
999 in quoted-printable encoding, as defined in RFC 2045 - I<MIME
1000 (Multipurpose Internet Mail Extensions)>.
1002 use MIME::QuotedPrint;
1004 $encoded = encode_qp("Smiley in Unicode: \x{263a}");
1005 $decoded = decode_qp($encoded);
1007 print $encoded, "\n"; # "Smiley in Unicode: =263A"
1009 See also L<PerlIO::via::QuotedPrint>.
1013 C<NEXT>, by Damian Conway, is a pseudo-class for method redispatch.
1018 C<open> is a new pragma for setting the default I/O layers
1023 C<PerlIO::scalar>, by Nick Ing-Simmons, provides the implementation
1024 of IO to "in memory" Perl scalars as discussed above. It also serves
1025 as an example of a loadable PerlIO layer. Other future possibilities
1026 include PerlIO::Array and PerlIO::Code. See L<PerlIO::scalar>.
1030 C<PerlIO::via>, by Nick Ing-Simmons, acts as a PerlIO layer and wraps
1031 PerlIO layer functionality provided by a class (typically implemented
1036 C<PerlIO::via::QuotedPrint>, by Elizabeth Mattijsen, is an example
1037 of a C<PerlIO::via> class:
1039 use PerlIO::via::QuotedPrint;
1040 open($fh,">:via(QuotedPrint)",$path);
1042 This will automatically convert everything output to C<$fh> to
1043 Quoted-Printable. See L<PerlIO::via> and L<PerlIO::via::QuotedPrint>.
1047 C<Pod::ParseLink>, by Russ Allbery, has been added,
1048 to parse LZ<><> links in pods as described in the new
1053 C<Pod::Text::Overstrike>, by Joe Smith, has been added.
1054 It converts POD data to formatted overstrike text.
1055 See L<Pod::Text::Overstrike>. [561+]
1059 C<Scalar::Util> is a selection of general-utility scalar subroutines,
1060 such as blessed(), reftype(), and tainted(). See L<Scalar::Util>.
1064 C<sort> is a new pragma for controlling the behaviour of sort().
1068 C<Storable> gives persistence to Perl data structures by allowing the
1069 storage and retrieval of Perl data to and from files in a fast and
1070 compact binary format. Because in effect Storable does serialisation
1071 of Perl data structures, with it you can also clone deep, hierarchical
1072 datastructures. Storable was originally created by Raphael Manfredi,
1073 but it is now maintained by Abhijit Menon-Sen. Storable has been
1074 enhanced to understand the two new hash features, Unicode keys and
1075 restricted hashes. See L<Storable>.
1079 C<Switch>, by Damian Conway, has been added. Just by saying
1083 you have C<switch> and C<case> available in Perl.
1089 case 1 { print "number 1" }
1090 case "a" { print "string a" }
1091 case [1..10,42] { print "number in list" }
1092 case (@array) { print "number in list" }
1093 case /\w+/ { print "pattern" }
1094 case qr/\w+/ { print "pattern" }
1095 case (%hash) { print "entry in hash" }
1096 case (\%hash) { print "entry in hash" }
1097 case (\&sub) { print "arg to subroutine" }
1098 else { print "previous case not true" }
1105 C<Test::More>, by Michael Schwern, is yet another framework for writing
1106 test scripts, more extensive than Test::Simple. See L<Test::More>.
1110 C<Test::Simple>, by Michael Schwern, has basic utilities for writing
1111 tests. See L<Test::Simple>.
1115 C<Text::Balanced>, by Damian Conway, has been added, for extracting
1116 delimited text sequences from strings.
1118 use Text::Balanced 'extract_delimited';
1120 ($a, $b) = extract_delimited("'never say never', he never said", "'", '');
1122 $a will be "'never say never'", $b will be ', he never said'.
1124 In addition to extract_delimited(), there are also extract_bracketed(),
1125 extract_quotelike(), extract_codeblock(), extract_variable(),
1126 extract_tagged(), extract_multiple(), gen_delimited_pat(), and
1127 gen_extract_tagged(). With these, you can implement rather advanced
1128 parsing algorithms. See L<Text::Balanced>.
1132 C<threads>, by Arthur Bergman, is an interface to interpreter threads.
1133 Interpreter threads (ithreads) is the new thread model introduced in
1134 Perl 5.6 but only available as an internal interface for extension
1135 writers (and for Win32 Perl for C<fork()> emulation). See L<threads>,
1136 L<threads::shared>, and L<perlthrtut>.
1140 C<threads::shared>, by Arthur Bergman, allows data sharing for
1141 interpreter threads. See L<threads::shared>.
1145 C<Tie::File>, by Mark-Jason Dominus, associates a Perl array with the
1146 lines of a file. See L<Tie::File>.
1150 C<Tie::Memoize>, by Ilya Zakharevich, provides on-demand loaded hashes.
1151 See L<Tie::Memoize>.
1155 C<Tie::RefHash::Nestable>, by Edward Avis, allows storing hash
1156 references (unlike the standard Tie::RefHash) The module is contained
1157 within Tie::RefHash. See L<Tie::RefHash>.
1161 C<Time::HiRes>, by Douglas E. Wegscheid, provides high resolution
1162 timing (ualarm, usleep, and gettimeofday). See L<Time::HiRes>.
1166 C<Unicode::UCD> offers a querying interface to the Unicode Character
1167 Database. See L<Unicode::UCD>.
1171 C<Unicode::Collate>, by SADAHIRO Tomoyuki, implements the UCA
1172 (Unicode Collation Algorithm) for sorting Unicode strings.
1173 See L<Unicode::Collate>.
1177 C<Unicode::Normalize>, by SADAHIRO Tomoyuki, implements the various
1178 Unicode normalization forms. See L<Unicode::Normalize>.
1182 C<XS::APItest>, by Tim Jenness, is a test extension that exercises XS
1183 APIs. Currently only C<printf()> is tested: how to output various
1184 basic data types from XS.
1188 C<XS::Typemap>, by Tim Jenness, is a test extension that exercises
1189 XS typemaps. Nothing gets installed, but the code is worth studying
1190 for extension writers.
1194 =head2 Updated And Improved Modules and Pragmata
1200 The following independently supported modules have been updated to the
1201 newest versions from CPAN: CGI, CPAN, DB_File, File::Spec, File::Temp,
1202 Getopt::Long, Math::BigFloat, Math::BigInt, the podlators bundle
1203 (Pod::Man, Pod::Text), Pod::LaTeX [561+], Pod::Parser, Storable,
1204 Term::ANSIColor, Test, Text-Tabs+Wrap.
1208 attributes::reftype() now works on tied arguments.
1212 AutoLoader can now be disabled with C<no AutoLoader;>.
1216 B::Deparse has been significantly enhanced by Robin Houston. It can
1217 now deparse almost all of the standard test suite (so that the tests
1218 still succeed). There is a make target "test.deparse" for trying this
1223 Carp now has better interface documentation, and the @CARP_NOT
1224 interface has been added to get optional control over where errors
1225 are reported independently of @ISA, by Ben Tilly.
1229 Class::Struct can now define the classes in compile time.
1233 Class::Struct now assigns the array/hash element if the accessor
1234 is called with an array/hash element as the B<sole> argument.
1238 The return value of Cwd::fastcwd() is now tainted.
1242 Data::Dumper now has an option to sort hashes.
1246 Data::Dumper now has an option to dump code references
1251 DB_File now supports newer Berkeley DB versions, among
1256 Devel::Peek now has an interface for the Perl memory statistics
1257 (this works only if you are using perl's malloc, and if you have
1258 compiled with debugging).
1262 The English module can now be used without the infamous performance
1265 use English '-no_match_vars';
1267 (Assuming, of course, that you don't need the troublesome variables
1268 C<$`>, C<$&>, or C<$'>.) Also, introduced C<@LAST_MATCH_START> and
1269 C<@LAST_MATCH_END> English aliases for C<@-> and C<@+>.
1273 ExtUtils::MakeMaker has been significantly cleaned up and fixed.
1274 The enhanced version has also been backported to earlier releases
1275 of Perl and submitted to CPAN so that the earlier releases can
1280 The arguments of WriteMakefile() in Makefile.PL are now checked
1281 for sanity much more carefully than before. This may cause new
1282 warnings when modules are being installed. See L<ExtUtils::MakeMaker>
1287 ExtUtils::MakeMaker now uses File::Spec internally, which hopefully
1288 leads to better portability.
1292 Fcntl, Socket, and Sys::Syslog have been rewritten by Nicholas Clark
1293 to use the new-style constant dispatch section (see L<ExtUtils::Constant>).
1294 This means that they will be more robust and hopefully faster.
1298 File::Find now chdir()s correctly when chasing symbolic links. [561]
1302 File::Find now has pre- and post-processing callbacks. It also
1303 correctly changes directories when chasing symbolic links. Callbacks
1304 (naughtily) exiting with "next;" instead of "return;" now work.
1308 File::Find is now (again) reentrant. It also has been made
1313 The warnings issued by File::Find now belong to their own category.
1314 You can enable/disable them with C<use/no warnings 'File::Find';>.
1318 File::Glob::glob() has been renamed to File::Glob::bsd_glob()
1319 because the name clashes with the builtin glob(). The older
1320 name is still available for compatibility, but is deprecated. [561]
1324 File::Glob now supports C<GLOB_LIMIT> constant to limit the size of
1325 the returned list of filenames.
1329 IPC::Open3 now allows the use of numeric file descriptors.
1333 IO::Socket now has an atmark() method, which returns true if the socket
1334 is positioned at the out-of-band mark. The method is also exportable
1335 as a sockatmark() function.
1339 IO::Socket::INET failed to open the specified port if the service name
1340 was not known. It now correctly uses the supplied port number as is. [561]
1344 IO::Socket::INET has support for the ReusePort option (if your
1345 platform supports it). The Reuse option now has an alias, ReuseAddr.
1346 For clarity, you may want to prefer ReuseAddr.
1350 IO::Socket::INET now supports a value of zero for C<LocalPort>
1351 (usually meaning that the operating system will make one up.)
1355 'use lib' now works identically to @INC. Removing directories
1356 with 'no lib' now works.
1360 Math::BigFloat and Math::BigInt have undergone a full rewrite by Tels.
1361 They are now magnitudes faster, and they support various bignum
1362 libraries such as GMP and PARI as their backends.
1366 Math::Complex handles inf, NaN etc., better.
1370 Net::Ping has been considerably enhanced by Rob Brown: multihoming is
1371 now supported, Win32 functionality is better, there is now time
1372 measuring functionality (optionally high-resolution using
1373 Time::HiRes), and there is now "external" protocol which uses
1374 Net::Ping::External module which runs your external ping utility and
1375 parses the output. A version of Net::Ping::External is available in
1378 Note that some of the Net::Ping tests are disabled when running
1379 under the Perl distribution since one cannot assume one or more
1380 of the following: enabled echo port at localhost, full Internet
1381 connectivity, or sympathetic firewalls. You can set the environment
1382 variable PERL_TEST_Net_Ping to "1" (one) before running the Perl test
1383 suite to enable all the Net::Ping tests.
1387 POSIX::sigaction() is now much more flexible and robust.
1388 You can now install coderef handlers, 'DEFAULT', and 'IGNORE'
1389 handlers, installing new handlers was not atomic.
1393 In Safe, C<%INC> is now localised in a Safe compartment so that
1398 In SDBM_File on dosish platforms, some keys went missing because of
1399 lack of support for files with "holes". A workaround for the problem
1404 In Search::Dict one can now have a pre-processing hook for the
1405 lines being searched.
1409 The Shell module now has an OO interface.
1413 In Sys::Syslog there is now a failover mechanism that will go
1414 through alternative connection mechanisms until the message
1415 is successfully logged.
1419 The Test module has been significantly enhanced.
1423 Time::Local::timelocal() does not handle fractional seconds anymore.
1424 The rationale is that neither does localtime(), and timelocal() and
1425 localtime() are supposed to be inverses of each other.
1429 The vars pragma now supports declaring fully qualified variables.
1430 (Something that C<our()> does not and will not support.)
1434 The C<utf8::> name space (as in the pragma) provides various
1435 Perl-callable functions to provide low level access to Perl's
1436 internal Unicode representation. At the moment only length()
1437 has been implemented.
1441 =head1 Utility Changes
1447 Emacs perl mode (emacs/cperl-mode.el) has been updated to version
1452 F<emacs/e2ctags.pl> is now much faster.
1456 C<enc2xs> is a tool for people adding their own encodings to the
1461 C<h2ph> now supports C trigraphs.
1465 C<h2xs> now produces a template README.
1469 C<h2xs> now uses C<Devel::PPPort> for better portability between
1470 different versions of Perl.
1474 C<h2xs> uses the new L<ExtUtils::Constant|ExtUtils::Constant> module
1475 which will affect newly created extensions that define constants.
1476 Since the new code is more correct (if you have two constants where the
1477 first one is a prefix of the second one, the first constant B<never>
1478 got defined), less lossy (it uses integers for integer constant,
1479 as opposed to the old code that used floating point numbers even for
1480 integer constants), and slightly faster, you might want to consider
1481 regenerating your extension code (the new scheme makes regenerating
1482 easy). L<h2xs> now also supports C trigraphs.
1486 C<libnetcfg> has been added to configure libnet.
1490 C<perlbug> is now much more robust. It also sends the bug report to
1491 perl.org, not perl.com.
1495 C<perlcc> has been rewritten and its user interface (that is,
1496 command line) is much more like that of the UNIX C compiler, cc.
1497 (The perlbc tools has been removed. Use C<perlcc -B> instead.)
1498 B<Note that perlcc is still considered very experimental and
1503 C<perlivp> is a new Installation Verification Procedure utility
1504 for running any time after installing Perl.
1508 C<piconv> is an implementation of the character conversion utility
1509 C<iconv>, demonstrating the new Encode module.
1513 C<pod2html> now allows specifying a cache directory.
1517 C<pod2html> now produces XHTML 1.0.
1521 C<pod2html> now understands POD written using different line endings
1522 (PC-like CRLF versus UNIX-like LF versus MacClassic-like CR).
1526 C<s2p> has been completely rewritten in Perl. (It is in fact a full
1527 implementation of sed in Perl: you can use the sed functionality by
1528 using the C<psed> utility.)
1532 C<xsubpp> now understands POD documentation embedded in the *.xs
1537 C<xsubpp> now supports the OUT keyword.
1541 =head1 New Documentation
1547 perl56delta details the changes between the 5.005 release and the
1552 perlclib documents the internal replacements for standard C library
1553 functions. (Interesting only for extension writers and Perl core
1558 perldebtut is a Perl debugging tutorial. [561+]
1562 perlebcdic contains considerations for running Perl on EBCDIC
1567 perlintro is a gentle introduction to Perl.
1571 perliol documents the internals of PerlIO with layers.
1575 perlmodstyle is a style guide for writing modules.
1579 perlnewmod tells about writing and submitting a new module. [561+]
1583 perlpacktut is a pack() tutorial.
1587 perlpod has been rewritten to be clearer and to record the best
1588 practices gathered over the years.
1592 perlpodspec is a more formal specification of the pod format,
1593 mainly of interest for writers of pod applications, not to
1594 people writing in pod.
1598 perlretut is a regular expression tutorial. [561+]
1602 perlrequick is a regular expressions quick-start guide.
1603 Yes, much quicker than perlretut. [561]
1607 perltodo has been updated.
1611 perltootc has been renamed as perltooc (to not to conflict
1612 with perltoot in filesystems restricted to "8.3" names).
1616 perluniintro is an introduction to using Unicode in Perl.
1617 (perlunicode is more of a detailed reference and background
1622 perlutil explains the command line utilities packaged with the Perl
1623 distribution. [561+]
1627 The following platform-specific documents are available before
1628 the installation as README.I<platform>, and after the installation
1631 perlaix perlamiga perlapollo perlbeos perlbs2000
1632 perlce perlcygwin perldgux perldos perlepoc perlfreebsd perlhpux
1633 perlhurd perlirix perlmachten perlmacos perlmint perlmpeix
1634 perlnetware perlos2 perlos390 perlplan9 perlqnx perlsolaris
1635 perltru64 perluts perlvmesa perlvms perlvos perlwin32
1637 These documents usually detail one or more of the following subjects:
1638 configuring, building, testing, installing, and sometimes also using
1639 Perl on the said platform.
1641 Eastern Asian Perl users are now welcomed in their own languages:
1642 README.jp (Japanese), README.ko (Korean), README.cn (simplified
1643 Chinese) and README.tw (traditional Chinese), which are written in
1644 normal pod but encoded in EUC-JP, EUC-KR, EUC-CN and Big5. These
1645 will get installed as
1647 perljp perlko perlcn perltw
1653 The documentation for the POSIX-BC platform is called "BS2000", to avoid
1654 confusion with the Perl POSIX module.
1658 The documentation for the WinCE platform is called perlce (README.ce
1659 in the source code kit), to avoid confusion with the perlwin32
1660 documentation on 8.3-restricted filesystems.
1664 =head1 Performance Enhancements
1670 map() could get pathologically slow when the result list it generates
1671 is larger than the source list. The performance has been improved for
1672 common scenarios. [561]
1676 sort() is also fully reentrant, in the sense that the sort function
1677 can itself call sort(). This did not work reliably in previous
1682 sort() has been changed to use primarily mergesort internally as
1683 opposed to the earlier quicksort. For very small lists this may
1684 result in slightly slower sorting times, but in general the speedup
1685 should be at least 20%. Additional bonuses are that the worst case
1686 behaviour of sort() is now better (in computer science terms it now
1687 runs in time O(N log N), as opposed to quicksort's Theta(N**2)
1688 worst-case run time behaviour), and that sort() is now stable
1689 (meaning that elements with identical keys will stay ordered as they
1690 were before the sort). See the C<sort> pragma for information.
1692 The story in more detail: suppose you want to serve yourself a little
1695 @digits = ( 3,1,4,1,5,9 );
1697 A numerical sort of the digits will yield (1,1,3,4,5,9), as expected.
1698 Which C<1> comes first is hard to know, since one C<1> looks pretty
1699 much like any other. You can regard this as totally trivial,
1700 or somewhat profound. However, if you just want to sort the even
1701 digits ahead of the odd ones, then what will
1703 sort { ($a % 2) <=> ($b % 2) } @digits;
1705 yield? The only even digit, C<4>, will come first. But how about
1706 the odd numbers, which all compare equal? With the quicksort algorithm
1707 used to implement Perl 5.6 and earlier, the order of ties is left up
1708 to the sort. So, as you add more and more digits of Pi, the order
1709 in which the sorted even and odd digits appear will change.
1710 and, for sufficiently large slices of Pi, the quicksort algorithm
1711 in Perl 5.8 won't return the same results even if reinvoked with the
1712 same input. The justification for this rests with quicksort's
1713 worst case behavior. If you run
1715 sort { $a <=> $b } ( 1 .. $N , 1 .. $N );
1717 (something you might approximate if you wanted to merge two sorted
1718 arrays using sort), doubling $N doesn't just double the quicksort time,
1719 it I<quadruples> it. Quicksort has a worst case run time that can
1720 grow like N**2, so-called I<quadratic> behaviour, and it can happen
1721 on patterns that may well arise in normal use. You won't notice this
1722 for small arrays, but you I<will> notice it with larger arrays,
1723 and you may not live long enough for the sort to complete on arrays
1724 of a million elements. So the 5.8 quicksort scrambles large arrays
1725 before sorting them, as a statistical defence against quadratic behaviour.
1726 But that means if you sort the same large array twice, ties may be
1727 broken in different ways.
1729 Because of the unpredictability of tie-breaking order, and the quadratic
1730 worst-case behaviour, quicksort was I<almost> replaced completely with
1731 a stable mergesort. I<Stable> means that ties are broken to preserve
1732 the original order of appearance in the input array. So
1734 sort { ($a % 2) <=> ($b % 2) } (3,1,4,1,5,9);
1736 will yield (4,3,1,1,5,9), guaranteed. The even and odd numbers
1737 appear in the output in the same order they appeared in the input.
1738 Mergesort has worst case O(N log N) behaviour, the best value
1739 attainable. And, ironically, this mergesort does particularly
1740 well where quicksort goes quadratic: mergesort sorts (1..$N, 1..$N)
1741 in O(N) time. But quicksort was rescued at the last moment because
1742 it is faster than mergesort on certain inputs and platforms.
1743 For example, if you really I<don't> care about the order of even
1744 and odd digits, quicksort will run in O(N) time; it's very good
1745 at sorting many repetitions of a small number of distinct elements.
1746 The quicksort divide and conquer strategy works well on platforms
1747 with relatively small, very fast, caches. Eventually, the problem gets
1748 whittled down to one that fits in the cache, from which point it
1749 benefits from the increased memory speed.
1751 Quicksort was rescued by implementing a sort pragma to control aspects
1752 of the sort. The B<stable> subpragma forces stable behaviour,
1753 regardless of algorithm. The B<_quicksort> and B<_mergesort>
1754 subpragmas are heavy-handed ways to select the underlying implementation.
1755 The leading C<_> is a reminder that these subpragmas may not survive
1756 beyond 5.8. More appropriate mechanisms for selecting the implementation
1757 exist, but they wouldn't have arrived in time to save quicksort.
1761 Hashes now use Bob Jenkins "One-at-a-Time" hashing key algorithm
1762 ( http://burtleburtle.net/bob/hash/doobs.html ). This algorithm is
1763 reasonably fast while producing a much better spread of values than
1764 the old hashing algorithm (originally by Chris Torek, later tweaked by
1765 Ilya Zakharevich). Hash values output from the algorithm on a hash of
1766 all 3-char printable ASCII keys comes much closer to passing the
1767 DIEHARD random number generation tests. According to perlbench, this
1768 change has not affected the overall speed of Perl.
1772 unshift() should now be noticeably faster.
1776 =head1 Installation and Configuration Improvements
1778 =head2 Generic Improvements
1784 INSTALL now explains how you can configure Perl to use 64-bit
1785 integers even on non-64-bit platforms.
1789 Policy.sh policy change: if you are reusing a Policy.sh file
1790 (see INSTALL) and you use Configure -Dprefix=/foo/bar and in the old
1791 Policy $prefix eq $siteprefix and $prefix eq $vendorprefix, all of
1792 them will now be changed to the new prefix, /foo/bar. (Previously
1793 only $prefix changed.) If you do not like this new behaviour,
1794 specify prefix, siteprefix, and vendorprefix explicitly.
1798 A new optional location for Perl libraries, otherlibdirs, is available.
1799 It can be used for example for vendor add-ons without disturbing Perl's
1800 own library directories.
1804 In many platforms, the vendor-supplied 'cc' is too stripped-down to
1805 build Perl (basically, 'cc' doesn't do ANSI C). If this seems
1806 to be the case and 'cc' does not seem to be the GNU C compiler
1807 'gcc', an automatic attempt is made to find and use 'gcc' instead.
1811 gcc needs to closely track the operating system release to avoid
1812 build problems. If Configure finds that gcc was built for a different
1813 operating system release than is running, it now gives a clearly visible
1814 warning that there may be trouble ahead.
1818 Since Perl 5.8 is not binary-compatible with previous releases
1819 of Perl, Configure no longer suggests including the 5.005
1824 Configure C<-S> can now run non-interactively. [561]
1828 Configure support for pdp11-style memory models has been removed due
1829 to obsolescence. [561]
1833 configure.gnu now works with options with whitespace in them.
1837 installperl now outputs everything to STDERR.
1841 Because PerlIO is now the default on most platforms, "-perlio" doesn't
1842 get appended to the $Config{archname} (also known as $^O) anymore.
1843 Instead, if you explicitly choose not to use perlio (Configure command
1844 line option -Uuseperlio), you will get "-stdio" appended.
1848 Another change related to the architecture name is that "-64all"
1849 (-Duse64bitall, or "maximally 64-bit") is appended only if your
1850 pointers are 64 bits wide. (To be exact, the use64bitall is ignored.)
1854 In AFS installations, one can configure the root of the AFS to be
1855 somewhere else than the default F</afs> by using the Configure
1856 parameter C<-Dafsroot=/some/where/else>.
1860 APPLLIB_EXP, a lesser-known configuration-time definition, has been
1861 documented. It can be used to prepend site-specific directories
1862 to Perl's default search path (@INC); see INSTALL for information.
1866 The version of Berkeley DB used when the Perl (and, presumably, the
1867 DB_File extension) was built is now available as
1868 C<@Config{qw(db_version_major db_version_minor db_version_patch)}>
1869 from Perl and as C<DB_VERSION_MAJOR_CFG DB_VERSION_MINOR_CFG
1870 DB_VERSION_PATCH_CFG> from C.
1874 Building Berkeley DB3 for compatibility modes for DB, NDBM, and ODBM
1875 has been documented in INSTALL.
1879 If you have CPAN access (either network or a local copy such as a
1880 CD-ROM) you can during specify extra modules to Configure to build and
1881 install with Perl using the -Dextras=... option. See INSTALL for
1886 In addition to config.over, a new override file, config.arch, is
1887 available. This file is supposed to be used by hints file writers
1888 for architecture-wide changes (as opposed to config.over which is
1889 for site-wide changes).
1893 If your file system supports symbolic links, you can build Perl outside
1894 of the source directory by
1896 mkdir /tmp/perl/build/directory
1897 cd /tmp/perl/build/directory
1898 sh /path/to/perl/source/Configure -Dmksymlinks ...
1900 This will create in /tmp/perl/build/directory a tree of symbolic links
1901 pointing to files in /path/to/perl/source. The original files are left
1902 unaffected. After Configure has finished, you can just say
1906 and Perl will be built and tested, all in /tmp/perl/build/directory.
1911 For Perl developers, several new make targets for profiling
1912 and debugging have been added; see L<perlhack>.
1918 Use of the F<gprof> tool to profile Perl has been documented in
1919 L<perlhack>. There is a make target called "perl.gprof" for
1920 generating a gprofiled Perl executable.
1924 If you have GCC 3, there is a make target called "perl.gcov" for
1925 creating a gcoved Perl executable for coverage analysis. See
1930 If you are on IRIX or Tru64 platforms, new profiling/debugging options
1931 have been added; see L<perlhack> for more information about pixie and
1938 Guidelines of how to construct minimal Perl installations have
1939 been added to INSTALL.
1943 The Thread extension is now not built at all under ithreads
1944 (C<Configure -Duseithreads>) because it wouldn't work anyway (the
1945 Thread extension requires being Configured with C<-Duse5005threads>).
1947 B<Note that the 5.005 threads are unsupported and deprecated: if you
1948 have code written for the old threads you should migrate it to the
1949 new ithreads model.>
1953 The Gconvert macro ($Config{d_Gconvert}) used by perl for stringifying
1954 floating-point numbers is now more picky about using sprintf %.*g
1955 rules for the conversion. Some platforms that used to use gcvt may
1956 now resort to the slower sprintf.
1960 The obsolete method of making a special (e.g., debugging) flavor
1963 make LIBPERL=libperld.a
1965 has been removed. Use -DDEBUGGING instead.
1969 =head2 New Or Improved Platforms
1971 For the list of platforms known to support Perl,
1972 see L<perlport/"Supported Platforms">.
1978 AIX dynamic loading should be now better supported.
1982 AIX should now work better with gcc, threads, and 64-bitness. Also the
1983 long doubles support in AIX should be better now. See L<perlaix>.
1987 AtheOS ( http://www.atheos.cx/ ) is a new platform.
1991 BeOS has been reclaimed.
1995 The DG/UX platform now supports 5.005-style threads.
2000 The DYNIX/ptx platform (also known as dynixptx) is supported at or
2005 EBCDIC platforms (z/OS (also known as OS/390), POSIX-BC, and VM/ESA)
2006 have been regained. Many test suite tests still fail and the
2007 co-existence of Unicode and EBCDIC isn't quite settled, but the
2008 situation is much better than with Perl 5.6. See L<perlos390>,
2009 L<perlbs2000> (for POSIX-BC), and L<perlvmesa> for more information.
2013 Building perl with -Duseithreads or -Duse5005threads now works under
2014 HP-UX 10.20 (previously it only worked under 10.30 or later). You will
2015 need a thread library package installed. See README.hpux. [561]
2019 Mac OS Classic is now supported in the mainstream source package
2020 (MacPerl has of course been available since perl 5.004 but now the
2021 source code bases of standard Perl and MacPerl have been synchronised)
2026 Mac OS X (or Darwin) should now be able to build Perl even on HFS+
2027 filesystems. (The case-insensitivity used to confuse the Perl build
2032 NCR MP-RAS is now supported. [561]
2036 All the NetBSD specific patches (except for the installation
2037 specific ones) have been merged back to the main distribution.
2041 NetWare from Novell is now supported. See L<perlnetware>.
2045 NonStop-UX is now supported. [561]
2049 NEC SUPER-UX is now supported.
2053 All the OpenBSD specific patches (except for the installation
2054 specific ones) have been merged back to the main distribution.
2058 Perl has been tested with the GNU pth userlevel thread package
2059 ( http://www.gnu.org/software/pth/pth.html ). All thread tests
2060 of Perl now work, but not without adding some yield()s to the tests,
2061 so while pth (and other userlevel thread implementations) can be
2062 considered to be "working" with Perl ithreads, keep in mind the
2063 possible non-preemptability of the underlying thread implementation.
2067 Stratus VOS is now supported using Perl's native build method
2068 (Configure). This is the recommended method to build Perl on
2069 VOS. The older methods, which build miniperl, are still
2070 available. See L<perlvos>. [561+]
2074 The Amdahl UTS UNIX mainframe platform is now supported. [561]
2078 WinCE is now supported. See L<perlce>.
2082 z/OS (formerly known as OS/390, formerly known as MVS OE) now has
2083 support for dynamic loading. This is not selected by default,
2084 however, you must specify -Dusedl in the arguments of Configure. [561]
2088 =head1 Selected Bug Fixes
2090 Numerous memory leaks and uninitialized memory accesses have been
2091 hunted down. Most importantly, anonymous subs used to leak quite
2098 The autouse pragma didn't work for Multi::Part::Function::Names.
2102 caller() could cause core dumps in certain situations. Carp was
2103 sometimes affected by this problem. In particular, caller() now
2104 returns a subroutine name of C<(unknown)> for subroutines that have
2105 been removed from the symbol table.
2109 chop(@list) in list context returned the characters chopped in
2110 reverse order. This has been reversed to be in the right order. [561]
2114 Configure no longer includes the DBM libraries (dbm, gdbm, db, ndbm)
2115 when building the Perl binary. The only exception to this is SunOS 4.x,
2116 which needs them. [561]
2120 The behaviour of non-decimal but numeric string constants such as
2121 "0x23" was platform-dependent: in some platforms that was seen as 35,
2122 in some as 0, in some as a floating point number (don't ask). This
2123 was caused by Perl's using the operating system libraries in a situation
2124 where the result of the string to number conversion is undefined: now
2125 Perl consistently handles such strings as zero in numeric contexts.
2129 Several debugger fixes: exit code now reflects the script exit code,
2130 condition C<"0"> now treated correctly, the C<d> command now checks
2131 line number, C<$.> no longer gets corrupted, and all debugger output
2132 now goes correctly to the socket if RemotePort is set. [561]
2136 The debugger (perl5db.pl) has been modified to present a more
2137 consistent commands interface, via (CommandSet=580). perl5db.t was
2138 also added to test the changes, and as a placeholder for further tests.
2144 The debugger has a new C<dumpDepth> option to control the maximum
2145 depth to which nested structures are dumped. The C<x> command has
2146 been extended so that C<x N EXPR> dumps out the value of I<EXPR> to a
2147 depth of at most I<N> levels.
2151 The debugger can now show lexical variables if you have the CPAN
2152 module PadWalker installed.
2156 The order of DESTROYs has been made more predictable.
2160 Perl 5.6.0 could emit spurious warnings about redefinition of
2161 dl_error() when statically building extensions into perl.
2162 This has been corrected. [561]
2166 L<dprofpp> -R didn't work.
2170 C<*foo{FORMAT}> now works.
2174 Infinity is now recognized as a number.
2178 UNIVERSAL::isa no longer caches methods incorrectly. (This broke
2179 the Tk extension with 5.6.0.) [561]
2183 Lexicals I: lexicals outside an eval "" weren't resolved
2184 correctly inside a subroutine definition inside the eval "" if they
2185 were not already referenced in the top level of the eval""ed code.
2189 Lexicals II: lexicals leaked at file scope into subroutines that
2190 were declared before the lexicals.
2194 Lexical warnings now propagating correctly between scopes
2195 and into C<eval "...">.
2199 C<use warnings qw(FATAL all)> did not work as intended. This has been
2204 warnings::enabled() now reports the state of $^W correctly if the caller
2205 isn't using lexical warnings. [561]
2209 Line renumbering with eval and C<#line> now works. [561]
2213 Fixed numerous memory leaks, especially in eval "".
2217 Localised tied variables no longer leak memory
2220 tie my %tied_hash => 'Tie::StdHash';
2224 # Used to leak memory every time local() was called;
2225 # in a loop, this added up.
2226 local($tied_hash{Foo}) = 1;
2230 Localised hash elements (and %ENV) are correctly unlocalised to not
2231 exist, if they didn't before they were localised.
2235 tie my %tied_hash => 'Tie::StdHash';
2239 # Nothing has set the FOO element so far
2241 { local $tied_hash{FOO} = 'Bar' }
2243 # This used to print, but not now.
2244 print "exists!\n" if exists $tied_hash{FOO};
2246 As a side effect of this fix, tied hash interfaces B<must> define
2247 the EXISTS and DELETE methods.
2251 mkdir() now ignores trailing slashes in the directory name,
2252 as mandated by POSIX.
2256 Some versions of glibc have a broken modfl(). This affects builds
2257 with C<-Duselongdouble>. This version of Perl detects this brokenness
2258 and has a workaround for it. The glibc release 2.2.2 is known to have
2259 fixed the modfl() bug.
2263 Modulus of unsigned numbers now works (4063328477 % 65535 used to
2264 return 27406, instead of 27047). [561]
2268 Some "not a number" warnings introduced in 5.6.0 eliminated to be
2269 more compatible with 5.005. Infinity is now recognised as a number. [561]
2273 Numeric conversions did not recognize changes in the string value
2274 properly in certain circumstances. [561]
2278 Attributes (such as :shared) didn't work with our().
2282 our() variables will not cause bogus "Variable will not stay shared"
2287 "our" variables of the same name declared in two sibling blocks
2288 resulted in bogus warnings about "redeclaration" of the variables.
2289 The problem has been corrected. [561]
2293 pack "Z" now correctly terminates the string with "\0".
2297 Fix password routines which in some shadow password platforms
2298 (e.g. HP-UX) caused getpwent() to return every other entry.
2302 The PERL5OPT environment variable (for passing command line arguments
2303 to Perl) didn't work for more than a single group of options. [561]
2307 PERL5OPT with embedded spaces didn't work.
2311 printf() no longer resets the numeric locale to "C".
2315 C<qw(a\\b)> now parses correctly as C<'a\\b'>: that is, as three
2316 characters, not four. [561]
2320 pos() did not return the correct value within s///ge in earlier
2321 versions. This is now handled correctly. [561]
2325 Printing quads (64-bit integers) with printf/sprintf now works
2326 without the q L ll prefixes (assuming you are on a quad-capable platform).
2330 Regular expressions on references and overloaded scalars now work. [561+]
2334 Right-hand side magic (GMAGIC) could in many cases such as string
2335 concatenation be invoked too many times.
2339 scalar() now forces scalar context even when used in void context.
2343 SOCKS support is now much more robust.
2347 sort() arguments are now compiled in the right wantarray context
2348 (they were accidentally using the context of the sort() itself).
2349 The comparison block is now run in scalar context, and the arguments
2350 to be sorted are always provided list context. [561]
2354 Changed the POSIX character class C<[[:space:]]> to include the (very
2355 rarely used) vertical tab character. Added a new POSIX-ish character
2356 class C<[[:blank:]]> which stands for horizontal whitespace
2357 (currently, the space and the tab).
2361 The tainting behaviour of sprintf() has been rationalized. It does
2362 not taint the result of floating point formats anymore, making the
2363 behaviour consistent with that of string interpolation. [561]
2367 Some cases of inconsistent taint propagation (such as within hash
2368 values) have been fixed.
2372 The RE engine found in Perl 5.6.0 accidentally pessimised certain kinds
2373 of simple pattern matches. These are now handled better. [561]
2377 Regular expression debug output (whether through C<use re 'debug'>
2378 or via C<-Dr>) now looks better. [561]
2382 Multi-line matches like C<"a\nxb\n" =~ /(?!\A)x/m> were flawed. The
2383 bug has been fixed. [561]
2387 Use of $& could trigger a core dump under some situations. This
2388 is now avoided. [561]
2392 The regular expression captured submatches ($1, $2, ...) are now
2393 more consistently unset if the match fails, instead of leaving false
2394 data lying around in them. [561]
2398 readline() on files opened in "slurp" mode could return an extra
2399 "" (blank line) at the end in certain situations. This has been
2404 Autovivification of symbolic references of special variables described
2405 in L<perlvar> (as in C<${$num}>) was accidentally disabled. This works
2410 Sys::Syslog ignored the C<LOG_AUTH> constant.
2414 $AUTOLOAD, sort(), lock(), and spawning subprocesses
2415 in multiple threads simultaneously are now thread-safe.
2419 Tie::Array's SPLICE method was broken.
2423 Allow a read-only string on the left-hand side of a non-modifying tr///.
2427 If C<STDERR> is tied, warnings caused by C<warn> and C<die> now
2428 correctly pass to it.
2432 Several Unicode fixes.
2438 BOMs (byte order marks) at the beginning of Perl files
2439 (scripts, modules) should now be transparently skipped.
2440 UTF-16 and UCS-2 encoded Perl files should now be read correctly.
2444 The character tables have been updated to Unicode 3.2.0.
2448 Comparing with utf8 data does not magically upgrade non-utf8 data
2449 into utf8. (This was a problem for example if you were mixing data
2450 from I/O and Unicode data: your output might have got magically encoded
2455 Generating illegal Unicode code points such as U+FFFE, or the UTF-16
2456 surrogates, now also generates an optional warning.
2460 C<IsAlnum>, C<IsAlpha>, and C<IsWord> now match titlecase.
2464 Concatenation with the C<.> operator or via variable interpolation,
2465 C<eq>, C<substr>, C<reverse>, C<quotemeta>, the C<x> operator,
2466 substitution with C<s///>, single-quoted UTF8, should now work.
2470 The C<tr///> operator now works. Note that the C<tr///CU>
2471 functionality has been removed (but see pack('U0', ...)).
2475 C<eval "v200"> now works.
2479 Perl 5.6.0 parsed m/\x{ab}/ incorrectly, leading to spurious warnings.
2480 This has been corrected. [561]
2484 Zero entries were missing from the Unicode classes such as C<IsDigit>.
2490 Large unsigned numbers (those above 2**31) could sometimes lose their
2491 unsignedness, causing bogus results in arithmetic operations. [561]
2495 The Perl parser has been stress tested using both random input and
2496 Markov chain input and the few found crashes and lockups have been
2501 =head2 Platform Specific Changes and Fixes
2509 Perl now works on post-4.0 BSD/OSes.
2515 Setting C<$0> now works (as much as possible; see L<perlvar> for details).
2521 Numerous updates; currently synchronised with Cygwin 1.3.10.
2525 Previously DYNIX/ptx had problems in its Configure probe for non-blocking I/O.
2531 EPOC now better supported. See README.epoc. [561]
2537 Perl now works on post-3.0 FreeBSDs.
2543 README.hpux updated; C<Configure -Duse64bitall> now works;
2544 now uses HP-UX malloc instead of Perl malloc.
2550 Numerous compilation flag and hint enhancements; accidental mixing
2551 of 32-bit and 64-bit libraries (a doomed attempt) made much harder.
2561 Long doubles should now work (see INSTALL). [561]
2565 Linux previously had problems related to sockaddrlen when using
2566 accept(), recvfrom() (in Perl: recv()), getpeername(), and
2575 Compilation of the standard Perl distribution in Mac OS Classic should
2576 now work if you have the Metrowerks development environment and the
2577 missing Mac-specific toolkit bits. Contact the macperl mailing list
2584 MPE/iX update after Perl 5.6.0. See README.mpeix. [561]
2588 NetBSD/threads: try installing the GNU pth (should be in the
2589 packages collection, or http://www.gnu.org/software/pth/),
2590 and Configure with -Duseithreads.
2596 Perl now works on NetBSD/sparc.
2602 Now works with usethreads (see INSTALL). [561]
2608 64-bitness using the Sun Workshop compiler now works.
2614 The native build method requires at least VOS Release 14.5.0
2615 and GNU C++/GNU Tools 2.0.1 or later. The Perl pack function
2616 now maps overflowed values to +infinity and underflowed values
2621 Tru64 (aka Digital UNIX, aka DEC OSF/1)
2623 The operating system version letter now recorded in $Config{osvers}.
2624 Allow compiling with gcc (previously explicitly forbidden). Compiling
2625 with gcc still not recommended because buggy code results, even with
2632 Fixed various alignment problems that lead into core dumps either
2633 during build or later; no longer dies on math errors at runtime;
2634 now using full quad integers (64 bits), previously was using
2635 only 46 bit integers for speed.
2641 See L</"Socket Extension Dynamic in VMS"> and L</"IEEE-format Floating Point
2642 Default on OpenVMS Alpha"> for important changes not otherwise listed here.
2644 chdir() now works better despite a CRT bug; now works with MULTIPLICITY
2645 (see INSTALL); now works with Perl's malloc.
2647 The tainting of C<%ENV> elements via C<keys> or C<values> was previously
2648 unimplemented. It now works as documented.
2650 The C<waitpid> emulation has been improved. The worst bug (now fixed)
2651 was that a pid of -1 would cause a wildcard search of all processes on
2654 POSIX-style signals are now emulated much better on VMS versions prior
2657 The C<system> function and backticks operator have improved
2658 functionality and better error handling. [561]
2660 File access tests now use current process privileges rather than the
2661 user's default privileges, which could sometimes result in a mismatch
2662 between reported access and actual access. This improvement is only
2663 available on VMS v6.0 and later.
2665 There is a new C<kill> implementation based on C<sys$sigprc> that allows
2666 older VMS systems (pre-7.0) to use C<kill> to send signals rather than
2667 simply force exit. This implementation also allows later systems to
2668 call C<kill> from within a signal handler.
2670 Iterative logical name translations are now limited to 10 iterations in
2671 imitation of SHOW LOGICAL and other OpenVMS facilities.
2681 Signal handling now works better than it used to. It is now implemented
2682 using a Windows message loop, and is therefore less prone to random
2687 fork() emulation is now more robust, but still continues to have a few
2688 esoteric bugs and caveats. See L<perlfork> for details. [561+]
2692 A failed (pseudo)fork now returns undef and sets errno to EAGAIN. [561]
2696 The following modules now work on Windows:
2698 ExtUtils::Embed [561]
2705 IO::File::new_tmpfile() is no longer limited to 32767 invocations
2710 Better chdir() return value for a non-existent directory.
2714 Compiling perl using the 64-bit Platform SDK tools is now supported.
2718 The Win32::SetChildShowWindow() builtin can be used to control the
2719 visibility of windows created by child processes. See L<Win32> for
2724 Non-blocking waits for child processes (or pseudo-processes) are
2725 supported via C<waitpid($pid, &POSIX::WNOHANG)>.
2729 The behavior of system() with multiple arguments has been rationalized.
2730 Each unquoted argument will be automatically quoted to protect whitespace,
2731 and any existing whitespace in the arguments will be preserved. This
2732 improves the portability of system(@args) by avoiding the need for
2733 Windows C<cmd> shell specific quoting in perl programs.
2735 Note that this means that some scripts that may have relied on earlier
2736 buggy behavior may no longer work correctly. For example,
2737 C<system("nmake /nologo", @args)> will now attempt to run the file
2738 C<nmake /nologo> and will fail when such a file isn't found.
2739 On the other hand, perl will now execute code such as
2740 C<system("c:/Program Files/MyApp/foo.exe", @args)> correctly.
2744 The perl header files no longer suppress common warnings from the
2745 Microsoft Visual C++ compiler. This means that additional warnings may
2746 now show up when compiling XS code.
2750 Borland C++ v5.5 is now a supported compiler that can build Perl.
2751 However, the generated binaries continue to be incompatible with those
2752 generated by the other supported compilers (GCC and Visual C++). [561]
2756 Duping socket handles with open(F, ">&MYSOCK") now works under Windows 9x.
2761 Current directory entries in %ENV are now correctly propagated to child
2766 New %ENV entries now propagate to subprocesses. [561]
2770 Win32::GetCwd() correctly returns C:\ instead of C: when at the drive root.
2771 Other bugs in chdir() and Cwd::cwd() have also been fixed. [561]
2775 The makefiles now default to the features enabled in ActiveState ActivePerl
2776 (a popular Win32 binary distribution). [561]
2780 HTML files will now be installed in c:\perl\html instead of
2781 c:\perl\lib\pod\html
2785 REG_EXPAND_SZ keys are now allowed in registry settings used by perl. [561]
2789 Can now send() from all threads, not just the first one. [561]
2793 ExtUtils::MakeMaker now uses $ENV{LIB} to search for libraries. [561]
2797 Less stack reserved per thread so that more threads can run
2798 concurrently. (Still 16M per thread.) [561]
2802 C<< File::Spec->tmpdir() >> now prefers C:/temp over /tmp
2803 (works better when perl is running as service).
2807 Better UNC path handling under ithreads. [561]
2811 wait(), waitpid(), and backticks now return the correct exit status
2812 under Windows 9x. [561]
2816 A socket handle leak in accept() has been fixed. [561]
2822 =head1 New or Changed Diagnostics
2824 Please see L<perldiag> for more details.
2830 Ambiguous range in the transliteration operator (like a-z-9) now
2835 Two new debugging options have been added: if you have compiled your
2836 Perl with debugging, you can use the -DT [561] and -DR options to trace
2837 tokenising and to add reference counts to displaying variables,
2842 The lexical warnings category "deprecated" is no longer a sub-category
2843 of the "syntax" category. It is now a top-level category in its own
2848 Unadorned dump() will now give a warning suggesting to
2849 use explicit CORE::dump() if that's what really is meant.
2853 The "Unrecognized escape" warning has been extended to include C<\8>,
2854 C<\9>, and C<\_>. There is no need to escape any of the C<\w> characters.
2858 All regular expression compilation error messages are now hopefully
2859 easier to understand both because the error message now comes before
2860 the failed regex and because the point of failure is now clearly
2861 marked by a C<E<lt>-- HERE> marker.
2865 Various I/O (and socket) functions like binmode(), close(), and so
2866 forth now more consistently warn if they are used illogically either
2867 on a yet unopened or on an already closed filehandle (or socket).
2871 Using lstat() on a filehandle now gives a warning. (It's a non-sensical
2876 The C<-M> and C<-m> options now warn if you didn't supply the module name.
2880 If you in C<use> specify a required minimum version, modules matching
2881 the name and but not defining a $VERSION will cause a fatal failure.
2885 Using negative offset for vec() in lvalue context is now a warnable offense.
2889 Odd number of arguments to oveload::constant now elicits a warning.
2893 Odd number of elements to in anonymous hash now elicits a warning.
2897 The various "opened only for", "on closed", "never opened" warnings
2898 drop the C<main::> prefix for filehandles in the C<main> package,
2899 for example C<STDIN> instead of C<main::STDIN>.
2903 Subroutine prototypes are now checked more carefully, you may
2904 get warnings for example if you have used non-prototype characters.
2908 If an attempt to use a (non-blessed) reference as an array index
2909 is made, a warning is given.
2913 C<push @a;> and C<unshift @a;> (with no values to push or unshift)
2914 now give a warning. This may be a problem for generated and evaled
2919 If you try to L<perlfunc/pack> a number less than 0 or larger than 255
2920 using the C<"C"> format you will get an optional warning. Similarly
2921 for the C<"c"> format and a number less than -128 or more than 127.
2925 pack C<P> format now demands an explicit size.
2929 unpack C<w> now warns of unterminated compressed integers.
2933 Warnings relating to the use of PerlIO have been added.
2937 Certain regex modifiers such as C<(?o)> make sense only if applied to
2938 the entire regex. You will get an optional warning if you try to do
2943 Variable length lookbehind has not yet been implemented, trying to
2944 use it will tell that.
2948 Using arrays or hashes as references (e.g. C<< %foo->{bar} >>
2949 has been deprecated for a while. Now you will get an optional warning.
2953 Warnings relating to the use of the new restricted hashes feature
2958 Self-ties of arrays and hashes are not supported and fatal errors
2959 will happen even at an attempt to do so.
2963 Using C<sort> in scalar context now issues an optional warning.
2964 This didn't do anything useful, as the sort was not performed.
2968 Using the /g modifier in split() is meaningless and will cause a warning.
2972 Using splice() past the end of an array now causes a warning.
2976 Malformed Unicode encodings (UTF-8 and UTF-16) cause a lot of warnings,
2977 ad doestrying to use UTF-16 surrogates (which are unimplemented).
2981 Trying to use Unicode characters on an I/O stream without marking the
2982 stream's encoding (using open() or binmode()) will cause "Wide character"
2987 Use of v-strings in use/require causes a (backward) portability warning.
2991 Warnings relating to the use interpreter threads and their shared data
2996 =head1 Changed Internals
3002 PerlIO is now the default.
3006 perlapi.pod (a companion to perlguts) now attempts to document the
3011 You can now build a really minimal perl called microperl.
3012 Building microperl does not require even running Configure;
3013 C<make -f Makefile.micro> should be enough. Beware: microperl makes
3014 many assumptions, some of which may be too bold; the resulting
3015 executable may crash or otherwise misbehave in wondrous ways.
3016 For careful hackers only.
3020 Added rsignal(), whichsig(), do_join(), op_clear, op_null,
3021 ptr_table_clear(), ptr_table_free(), sv_setref_uv(), and several UTF-8
3022 interfaces to the publicised API. For the full list of the available
3023 APIs see L<perlapi>.
3027 Made possible to propagate customised exceptions via croak()ing.
3031 Now xsubs can have attributes just like subs. (Well, at least the
3032 built-in attributes.)
3036 dTHR and djSP have been obsoleted; the former removed (because it's
3037 a no-op) and the latter replaced with dSP.
3041 PERL_OBJECT has been completely removed.
3045 The MAGIC constants (e.g. C<'P'>) have been macrofied
3046 (e.g. C<PERL_MAGIC_TIED>) for better source code readability
3047 and maintainability.
3051 The regex compiler now maintains a structure that identifies nodes in
3052 the compiled bytecode with the corresponding syntactic features of the
3053 original regex expression. The information is attached to the new
3054 C<offsets> member of the C<struct regexp>. See L<perldebguts> for more
3055 complete information.
3059 The C code has been made much more C<gcc -Wall> clean. Some warning
3060 messages still remain in some platforms, so if you are compiling with
3061 gcc you may see some warnings about dubious practices. The warnings
3062 are being worked on.
3066 F<perly.c>, F<sv.c>, and F<sv.h> have now been extensively commented.
3070 Documentation on how to use the Perl source repository has been added
3071 to F<Porting/repository.pod>.
3075 There are now several profiling make targets.
3079 =head1 Security Vulnerability Closed [561]
3081 (This change was already made in 5.7.0 but bears repeating here.)
3082 (5.7.0 came out before 5.6.1: the development branch 5.7 released
3083 earlier than the maintenance branch 5.6)
3085 A potential security vulnerability in the optional suidperl component
3086 of Perl was identified in August 2000. suidperl is neither built nor
3087 installed by default. As of November 2001 the only known vulnerable
3088 platform is Linux, most likely all Linux distributions. CERT and
3089 various vendors and distributors have been alerted about the vulnerability.
3090 See http://www.cpan.org/src/5.0/sperl-2000-08-05/sperl-2000-08-05.txt
3091 for more information.
3093 The problem was caused by Perl trying to report a suspected security
3094 exploit attempt using an external program, /bin/mail. On Linux
3095 platforms the /bin/mail program had an undocumented feature which
3096 when combined with suidperl gave access to a root shell, resulting in
3097 a serious compromise instead of reporting the exploit attempt. If you
3098 don't have /bin/mail, or if you have 'safe setuid scripts', or if
3099 suidperl is not installed, you are safe.
3101 The exploit attempt reporting feature has been completely removed from
3102 Perl 5.8.0 (and the maintenance release 5.6.1, and it was removed also
3103 from all the Perl 5.7 releases), so that particular vulnerability
3104 isn't there anymore. However, further security vulnerabilities are,
3105 unfortunately, always possible. The suidperl functionality is most
3106 probably going to be removed in Perl 5.10. In any case, suidperl
3107 should only be used by security experts who know exactly what they are
3108 doing and why they are using suidperl instead of some other solution
3109 such as sudo ( see http://www.courtesan.com/sudo/ ).
3113 Several new tests have been added, especially for the F<lib> and
3114 F<ext> subsections. There are now about 69 000 individual tests
3115 (spread over about 700 test scripts), in the regression suite (5.6.1
3116 has about 11 700 tests, in 258 test scripts) The exact numbers depend
3117 on the platform and Perl configuration used. Many of the new tests
3118 are of course introduced by the new modules, but still in general Perl
3119 is now more thoroughly tested.
3121 Because of the large number of tests, running the regression suite
3122 will take considerably longer time than it used to: expect the suite
3123 to take up to 4-5 times longer to run than in perl 5.6. On a really
3124 fast machine you can hope to finish the suite in about 6-8 minutes
3127 The tests are now reported in a different order than in earlier Perls.
3128 (This happens because the test scripts from under t/lib have been moved
3129 to be closer to the library/extension they are testing.)
3131 =head1 Known Problems
3133 =head2 The Compiler Suite Is Still Very Experimental
3135 The compiler suite is slowly getting better but it continues to be
3136 highly experimental. Use in production environments is discouraged.
3138 =head2 Localising Tied Arrays and Hashes Is Broken
3142 doesn't work as one would expect: the old value is restored
3143 incorrectly. This will be changed in a future release, but we don't
3144 know yet what the new semantics will exactly be. In any case, the
3145 change will break existing code that relies on the current
3146 (ill-defined) semantics, so just avoid doing this in general.
3148 =head2 Building Extensions Can Fail Because Of Largefiles
3150 Some extensions like mod_perl are known to have issues with
3151 `largefiles', a change brought by Perl 5.6.0 in which file offsets
3152 default to 64 bits wide, where supported. Modules may fail to compile
3153 at all, or they may compile and work incorrectly. Currently, there
3154 is no good solution for the problem, but Configure now provides
3155 appropriate non-largefile ccflags, ldflags, libswanted, and libs
3156 in the %Config hash (e.g., $Config{ccflags_nolargefiles}) so the
3157 extensions that are having problems can try configuring themselves
3158 without the largefileness. This is admittedly not a clean solution,
3159 and the solution may not even work at all. One potential failure is
3160 whether one can (or, if one can, whether it's a good idea to) link
3161 together at all binaries with different ideas about file offsets;
3162 all this is platform-dependent.
3164 =head2 Modifying $_ Inside for(..)
3168 works without complaint. It shouldn't. (You should be able to
3169 modify only lvalue elements inside the loops.) You can see the
3170 correct behaviour by replacing the 1..5 with 1, 2, 3, 4, 5.
3172 =head2 mod_perl 1.26 Doesn't Build With Threaded Perl
3174 Use mod_perl 1.27 or higher.
3176 =head2 lib/ftmp-security tests warn 'system possibly insecure'
3178 Don't panic. Read the 'make test' section of INSTALL instead.
3180 =head2 libwww-perl (LWP) fails base/date #51
3182 Use libwww-perl 5.65 or later.
3184 =head2 PDL failing some tests
3186 Use PDL 2.3.4 or later.
3190 You may get errors like 'Undefined symbol "Perl_get_sv"' or "can't
3191 resolve symbol 'Perl_get_sv'", or the symbol may be "Perl_sv_2pv".
3192 This probably means that you are trying to use an older shared Perl
3193 library (or extensions linked with such) with Perl 5.8.0 executable.
3194 Perl used to have such a subroutine, but that is no more the case.
3195 Check your shared library path, and any shared Perl libraries in those
3198 Sometimes this problem may also indicate a partial Perl 5.8.0
3199 installation, see L</"Mac OS X dyld undefined symbols"> for an
3200 example and how to deal with it.
3202 =head2 Self-tying Problems
3204 Self-tying of arrays and hashes is broken in rather deep and
3205 hard-to-fix ways. As a stop-gap measure to avoid people from getting
3206 frustrated at the mysterious results (core dumps, most often), it is
3207 forbidden for now (you will get a fatal error even from an attempt).
3209 A change to self-tying of globs has caused them to be recursively
3210 referenced (see: L<perlobj/"Two-Phased Garbage Collection">). You
3211 will now need an explicit untie to destroy a self-tied glob. This
3212 behaviour may be fixed at a later date.
3214 Self-tying of scalars and IO thingies works.
3216 =head2 ext/threads/t/libc
3218 If this test fails, it indicates that your libc (C library) is not
3219 threadsafe. This particular test stress tests the localtime() call to
3220 find out whether it is threadsafe. See L<perlthrtut> for more information.
3222 =head2 Failure of Thread (5.005-style) tests
3224 B<Note that support for 5.005-style threading is deprecated,
3225 experimental and practically unsupported. In 5.10, it is expected
3226 to be removed. You should migrate your code to ithreads.>
3228 The following tests are known to fail due to fundamental problems in
3229 the 5.005 threading implementation. These are not new failures--Perl
3230 5.005_0x has the same bugs, but didn't have these tests.
3232 ../ext/B/t/xref.t 255 65280 14 12 85.71% 3-14
3233 ../ext/List/Util/t/first.t 255 65280 7 4 57.14% 2 5-7
3234 ../lib/English.t 2 512 54 2 3.70% 2-3
3235 ../lib/FileCache.t 5 1 20.00% 5
3236 ../lib/Filter/Simple/t/data.t 6 3 50.00% 1-3
3237 ../lib/Filter/Simple/t/filter_only. 9 3 33.33% 1-2 5
3238 ../lib/Math/BigInt/t/bare_mbf.t 1627 4 0.25% 8 11 1626-1627
3239 ../lib/Math/BigInt/t/bigfltpm.t 1629 4 0.25% 10 13 1628-
3241 ../lib/Math/BigInt/t/sub_mbf.t 1633 4 0.24% 8 11 1632-1633
3242 ../lib/Math/BigInt/t/with_sub.t 1628 4 0.25% 9 12 1627-1628
3243 ../lib/Tie/File/t/31_autodefer.t 255 65280 65 32 49.23% 34-65
3244 ../lib/autouse.t 10 1 10.00% 4
3245 op/flip.t 15 1 6.67% 15
3247 These failures are unlikely to get fixed as 5.005-style threads
3248 are considered fundamentally broken. (Basically what happens is that
3249 competing threads can corrupt shared global state, one good example
3250 being regular expression engine's state.)
3252 =head2 Timing problems
3254 The following tests may fail intermittently because of timing
3255 problems, for example if the system is heavily loaded.
3258 ext/Time/HiRes/HiRes.t
3260 lib/Memoize/t/expmod_t.t
3261 lib/Memoize/t/speed.t
3263 In case of failure please try running them manually, for example
3265 ./perl -Ilib ext/Time/HiRes/HiRes.t
3267 =head2 Tied/Magical Array/Hash Elements Do Not Autovivify
3269 For normal arrays C<$foo = \$bar[1]> will assign C<undef> to
3270 C<$bar[1]> (assuming that it didn't exist before), but for
3271 tied/magical arrays and hashes such autovivification does not happen
3272 because there is currently no way to catch the reference creation.
3273 The same problem affects slicing over non-existent indices/keys of
3274 a tied/magical array/hash.
3276 =head2 Unicode in package/class and subroutine names does not work
3278 One can have Unicode in identifier names, but not in package/class or
3279 subroutine names. While some limited functionality towards this does
3280 exist as of Perl 5.8.0, that is more accidental than designed; use of
3281 Unicode for the said purposes is unsupported.
3283 One reason of this unfinishedness is its (currently) inherent
3284 unportability: since both package names and subroutine names may
3285 need to be mapped to file and directory names, the Unicode capability
3286 of the filesystem becomes important-- and there unfortunately aren't
3289 =head1 Platform Specific Problems
3297 If using the AIX native make command, instead of just "make" issue
3298 "make all". In some setups the former has been known to spuriously
3299 also try to run "make install". Alternatively, you may want to use
3304 In AIX 4.2, Perl extensions that use C++ functions that use statics
3305 may have problems in that the statics are not getting initialized.
3306 In newer AIX releases, this has been solved by linking Perl with
3307 the libC_r library, but unfortunately in AIX 4.2 the said library
3308 has an obscure bug where the various functions related to time
3309 (such as time() and gettimeofday()) return broken values, and
3310 therefore in AIX 4.2 Perl is not linked against libC_r.
3314 vac 5.0.0.0 May Produce Buggy Code For Perl
3316 The AIX C compiler vac version 5.0.0.0 may produce buggy code,
3317 resulting in a few random tests failing when run as part of "make
3318 test", but when the failing tests are run by hand, they succeed.
3319 We suggest upgrading to at least vac version 5.0.1.0, that has been
3320 known to compile Perl correctly. "lslpp -L|grep vac.C" will tell
3321 you the vac version. See README.aix.
3325 If building threaded Perl, you may get compilation warning from pp_sys.c:
3327 "pp_sys.c", line 4651.39: 1506-280 (W) Function argument assignment between types "unsigned char*" and "const void*" is not allowed.
3329 This is harmless; it is caused by the getnetbyaddr() and getnetbyaddr_r()
3330 having slightly different types for their first argument.
3334 =head2 Alpha systems with old gccs fail several tests
3336 If you see op/pack, op/pat, op/regexp, or ext/Storable tests failing
3337 in a Linux/alpha or *BSD/Alpha, it's probably time to upgrade your gcc.
3338 gccs prior to 2.95.3 are definitely not good enough, and gcc 3.1 may
3339 be even better. (RedHat Linux/alpha with gcc 3.1 reported no problems,
3340 as did Linux 2.4.18 with gcc 2.95.4.) (In Tru64, it is preferable to
3341 use the bundled C compiler.)
3345 Perl 5.8.0 doesn't build in AmigaOS. It broke at some point during
3346 the ithreads work and we could not find Amiga experts to unbreak the
3347 problems. Perl 5.6.1 still works for AmigaOS (as does the the 5.7.2
3348 development release).
3352 The following tests fail on 5.8.0 Perl in BeOS Personal 5.03:
3354 t/op/lfs............................FAILED at test 17
3355 t/op/magic..........................FAILED at test 24
3356 ext/Fcntl/t/syslfs..................FAILED at test 17
3357 ext/File/Glob/t/basic...............FAILED at test 3
3358 ext/POSIX/t/sigaction...............FAILED at test 13
3359 ext/POSIX/t/waitpid.................FAILED at test 1
3361 See L<perlbeos> (README.beos) for more details.
3363 =head2 Cygwin "unable to remap"
3365 For example when building the Tk extension for Cygwin,
3366 you may get an error message saying "unable to remap".
3367 This is known problem with Cygwin, and a workaround is
3368 detailed in here: http://sources.redhat.com/ml/cygwin/2001-12/msg00894.html
3370 =head2 Cygwin ndbm tests fail on FAT
3372 One can build but not install (or test the build of) the NDBM_File
3373 on FAT filesystems. Installation (or build) on NTFS works fine.
3374 If one attempts the test on a FAT install (or build) the following
3375 failures are expected:
3377 ../ext/NDBM_File/ndbm.t 13 3328 71 59 83.10% 1-2 4 16-71
3378 ../ext/ODBM_File/odbm.t 255 65280 ?? ?? % ??
3379 ../lib/AnyDBM_File.t 2 512 12 2 16.67% 1 4
3380 ../lib/Memoize/t/errors.t 0 139 11 5 45.45% 7-11
3381 ../lib/Memoize/t/tie_ndbm.t 13 3328 4 4 100.00% 1-4
3382 run/fresh_perl.t 97 1 1.03% 91
3384 NDBM_File fails and ODBM_File just coredumps.
3386 =head2 DJGPP does not build
3388 Unfortunately DJGPP build broke somewhere after 5.7.3.
3390 =head2 FreeBSD built with ithreads coredumps reading large directories
3392 This is a known bug in FreeBSD 4.5's readdir_r(), it has been fixed in
3393 FreeBSD 4.6 (see L<perlfreebsd> (README.freebsd)).
3395 =head2 FreeBSD Failing locale Test 117 For ISO 8859-15 Locales
3397 The ISO 8859-15 locales may fail the locale test 117 in FreeBSD.
3398 This is caused by the characters \xFF (y with diaeresis) and \xBE
3399 (Y with diaeresis) not behaving correctly when being matched
3400 case-insensitively. Apparently this problem has been fixed in
3401 the latest FreeBSD releases.
3402 ( http://www.freebsd.org/cgi/query-pr.cgi?pr=34308 )
3404 =head2 IRIX fails ext/List/Util/t/shuffle.t or Digest::MD5
3406 IRIX with MIPSpro 7.3.1.2m or 7.3.1.3m compiler may fail the List::Util
3407 test ext/List/Util/t/shuffle.t by dumping core. This seems to be
3408 a compiler error since if compiled with gcc no core dump ensues, and
3409 no failures have been seen on the said test on any other platform.
3411 Similarly, building the Digest::MD5 extension has been
3412 known to fail with "*** Termination code 139 (bu21)".
3414 The cure is to drop optimization level (Configure -Doptimize=-O2).
3416 =head2 HP-UX lib/posix Subtest 9 Fails When LP64-Configured
3418 If perl is configured with -Duse64bitall, the successful result of the
3419 subtest 10 of lib/posix may arrive before the successful result of the
3420 subtest 9, which confuses the test harness so much that it thinks the
3423 =head2 Linux with glibc 2.2.5 fails t/op/int subtest #6 with -Duse64bitint
3425 This is a known bug in the glibc 2.2.5 with long long integers.
3426 ( http://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=65612 )
3428 =head2 Linux With Sfio Fails op/misc Test 48
3434 Please remember to set your environment variable LC_ALL to "C"
3435 (setenv LC_ALL C) before running "make test" to avoid a lot of
3436 warnings about the broken locales of Mac OS X.
3438 The following tests are known to fail in Mac OS X 10.1.5 because of
3439 buggy (old) implementations of Berkeley DB included in Mac OS X:
3441 Failed Test Stat Wstat Total Fail Failed List of Failed
3442 -------------------------------------------------------------------------
3443 ../ext/DB_File/t/db-btree.t 0 11 ?? ?? % ??
3444 ../ext/DB_File/t/db-recno.t 149 3 2.01% 61 63 65
3446 If you are building on a UFS partition, you will also probably see
3447 t/op/stat.t subtest #9 fail. This is caused by Darwin's UFS not
3448 supporting inode change time.
3450 Also the ext/POSIX/t/posix.t subtest #10 fails but it is skipped for
3451 now because the failure is Apple's fault, not Perl's (blocked signals
3454 If you Configure with ithreads, ext/threads/t/libc.t will fail. Again,
3455 this is not Perl's fault-- the libc of Mac OS X is not threadsafe
3456 (in this particular test, the localtime() call is found to be
3459 =head2 Mac OS X dyld undefined symbols
3461 If after installing Perl 5.8.0 you are getting warnings about missing
3462 symbols, for example
3464 dyld: perl Undefined symbols
3468 you probably have an old pre-Perl-5.8.0 installation (or parts of one)
3469 in /Library/Perl (the undefined symbols used to exist in pre-5.8.0 Perls).
3470 It seems that for some reason "make install" doesn't always completely
3471 overwrite the files in /Library/Perl. You can remove the Perl
3472 shared library like this:
3474 rm /Library/Perl/darwin/CORE/libperl.dylib
3476 and then reissue "make install". Note that the above of course is
3477 extremely disruptive for anything using the /usr/local/bin/perl.
3478 If that doesn't help, you may have to try removing all the .bundle
3479 files from beneath /Library/Perl, and again "make install"-ing.
3481 =head2 OS/2 Test Failures
3483 The following tests are known to fail on OS/2 (for clarity
3484 only the failures are shown, not the full error messages):
3486 t/io/utf8............................FAILED at test 19
3487 t/op/grent...........................FAILED at test 2
3488 t/op/pwent...........................FAILED at test 1
3489 t/lib/os2_base.......................FAILED at test 13
3490 t/lib/os2_process....................FAILED at test 10
3491 t/lib/os2_process_kid................FAILED at test 10
3492 t/lib/rx_cmprt.......................FAILED at test 16
3493 ext/DB_File/t/db-btree...............FAILED at test 0
3494 ext/DB_File/t/db-hash................FAILED at test 0
3495 ext/DB_File/t/db-recno...............FAILED at test 0
3496 lib/ExtUtils/t/basic.................FAILED at test 14
3497 lib/ExtUtils/t/Constant..............FAILED at test 4
3498 lib/Memoize/t/errors.................FAILED at test 4
3500 =head2 op/sprintf tests 91, 129, and 130
3502 The op/sprintf tests 91, 129, and 130 are known to fail on some platforms.
3503 Examples include any platform using sfio, and Compaq/Tandem's NonStop-UX.
3505 Test 91 is known to fail on QNX6 (nto), because C<sprintf '%e',0>
3506 incorrectly produces C<0.000000e+0> instead of C<0.000000e+00>.
3508 For tests 129 and 130, the failing platforms do not comply with
3509 the ANSI C Standard: lines 19ff on page 134 of ANSI X3.159 1989, to
3510 be exact. (They produce something other than "1" and "-1" when
3511 formatting 0.6 and -0.6 using the printf format "%.0f"; most often,
3512 they produce "0" and "-0".)
3516 In case you are still using Solaris 2.5 (aka SunOS 5.5), you may
3517 experience failures (the test core dumping) in lib/locale.t.
3518 The suggested cure is to upgrade your Solaris.
3520 =head2 Solaris x86 Fails Tests With -Duse64bitint
3522 The following tests are known to fail in Solaris x86 with Perl
3523 configured to use 64 bit integers:
3525 ext/Data/Dumper/t/dumper.............FAILED at test 268
3526 ext/Devel/Peek/Peek..................FAILED at test 7
3528 =head2 SUPER-UX (NEC SX)
3530 The following tests are known to fail on SUPER-UX:
3532 op/64bitint...........................FAILED tests 29-30, 32-33, 35-36
3533 op/arith..............................FAILED tests 128-130
3534 op/pack...............................FAILED tests 25-5625
3535 op/pow................................
3536 op/taint..............................# msgsnd failed
3537 ../ext/IO/lib/IO/t/io_poll............FAILED tests 3-4
3538 ../ext/IPC/SysV/ipcsysv...............FAILED tests 2, 5-6
3539 ../ext/IPC/SysV/t/msg.................FAILED tests 2, 4-6
3540 ../ext/Socket/socketpair..............FAILED tests 12
3541 ../lib/IPC/SysV.......................FAILED tests 2, 5-6
3542 ../lib/warnings.......................FAILED tests 115-116, 118-119
3544 The op/pack failure ("Cannot compress negative numbers at op/pack.t line 126")
3545 is serious but as of yet unsolved. It points at some problems with the
3546 signedness handling of the C compiler, as do the 64bitint, arith, and pow
3547 failures. Most of the rest point at problems with SysV IPC.
3549 =head2 Term::ReadKey not working on Win32
3551 Use Term::ReadKey 2.20 or later.
3559 During Configure, the test
3561 Guessing which symbols your C compiler and preprocessor define...
3563 will probably fail with error messages like
3565 CC-20 cc: ERROR File = try.c, Line = 3
3566 The identifier "bad" is undefined.
3568 bad switch yylook 79bad switch yylook 79bad switch yylook 79bad switch yylook 79#ifdef A29K
3571 CC-65 cc: ERROR File = try.c, Line = 3
3572 A semicolon is expected at this point.
3574 This is caused by a bug in the awk utility of UNICOS/mk. You can ignore
3575 the error, but it does cause a slight problem: you cannot fully
3576 benefit from the h2ph utility (see L<h2ph>) that can be used to
3577 convert C headers to Perl libraries, mainly used to be able to access
3578 from Perl the constants defined using C preprocessor, cpp. Because of
3579 the above error, parts of the converted headers will be invisible.
3580 Luckily, these days the need for h2ph is rare.
3584 If building Perl with interpreter threads (ithreads), the
3585 getgrent(), getgrnam(), and getgrgid() functions cannot return the
3586 list of the group members due to a bug in the multithreaded support of
3587 UNICOS/mk. What this means is that in list context the functions will
3588 return only three values, not four.
3594 There are a few known test failures, see L<perluts> (README.uts).
3596 =head2 VOS (Stratus)
3598 When Perl is built using the native build process on VOS Release
3599 14.5.0 and GNU C++/GNU Tools 2.0.1, all attempted tests either
3600 pass or result in TODO (ignored) failures.
3604 There should be no reported test failures with a default configuration,
3605 though there are a number of tests marked TODO that point to areas
3606 needing further debugging and/or porting work.
3610 In multi-CPU boxes, there are some problems with the I/O buffering:
3611 some output may appear twice.
3613 =head2 XML::Parser not working
3615 Use XML::Parser 2.31 or later.
3617 =head2 z/OS (OS/390)
3619 z/OS has rather many test failures but the situation is actually much
3620 better than it was in 5.6.0; it's just that so many new modules and
3621 tests have been added.
3623 Failed Test Stat Wstat Total Fail Failed List of Failed
3624 ---------------------------------------------------------------------------
3625 ../ext/Data/Dumper/t/dumper.t 357 8 2.24% 311 314 325 327
3627 ../ext/IO/lib/IO/t/io_unix.t 5 4 80.00% 2-5
3628 ../ext/Storable/t/downgrade.t 12 3072 169 12 7.10% 14-15 46-47 78-79
3630 ../lib/ExtUtils/t/Constant.t 121 30976 48 48 100.00% 1-48
3631 ../lib/ExtUtils/t/Embed.t 9 9 100.00% 1-9
3632 op/pat.t 922 7 0.76% 665 776 785 832-
3634 op/sprintf.t 224 3 1.34% 98 100 136
3635 op/tr.t 97 5 5.15% 63 71-74
3636 uni/fold.t 780 6 0.77% 61 169 196 661
3639 The failures in dumper.t and downgrade.t are problems in the tests,
3640 those in io_unix and sprintf are problems in the USS (UDP sockets and
3641 printf formats). The pat, tr, and fold failures are genuine Perl
3642 problems caused by EBCDIC (and in the pat and fold cases, combining
3643 that with Unicode). The Constant and Embed are probably problems in
3644 the tests (since they test Perl's ability to build extensions, and
3645 that seems to be working reasonably well.)
3647 =head2 Unicode Support on EBCDIC Still Spotty
3649 Though mostly working, Unicode support still has problem spots on
3650 EBCDIC platforms. One such known spot are the C<\p{}> and C<\P{}>
3651 regular expression constructs for code points less than 256: the
3652 C<pP> are testing for Unicode code points, not knowing about EBCDIC.
3654 =head2 Seen In Perl 5.7 But Gone Now
3656 C<Time::Piece> (previously known as C<Time::Object>) was removed
3657 because it was felt that it didn't have enough value in it to be a
3658 core module. It is still a useful module, though, and is available
3661 Perl 5.8 unfortunately does not build anymore on AmigaOS; this broke
3662 accidentally at some point. Since there are not that many Amiga
3663 developers available, we could not get this fixed and tested in time
3664 for 5.8.0. Perl 5.6.1 still works for AmigaOS (as does the the 5.7.2
3665 development release).
3667 The C<PerlIO::Scalar> and C<PerlIO::Via> (capitalised) were renamed as
3668 C<PerlIO::scalar> and C<PerlIO::via> (all lowercase) just before 5.8.0.
3669 The main rationale was to have all core PerlIO layers to have all
3670 lowercase names. The "plugins" are named as usual, for example
3671 C<PerlIO::via::QuotedPrint>.
3673 The C<threads::shared::queue> and C<threads::shared::semaphore> were
3674 renamed as C<Thread::Queue> and C<Thread::Semaphore> just before 5.8.0.
3675 The main rationale was to have thread modules to obey normal naming,
3676 C<Thread::> (the C<threads> and C<threads::shared> themselves are
3677 more pragma-like, they affect compile-time, so they stay lowercase).
3679 =head1 Reporting Bugs
3681 If you find what you think is a bug, you might check the articles
3682 recently posted to the comp.lang.perl.misc newsgroup and the perl
3683 bug database at http://bugs.perl.org/ . There may also be
3684 information at http://www.perl.com/ , the Perl Home Page.
3686 If you believe you have an unreported bug, please run the B<perlbug>
3687 program included with your release. Be sure to trim your bug down
3688 to a tiny but sufficient test case. Your bug report, along with the
3689 output of C<perl -V>, will be sent off to perlbug@perl.org to be
3690 analysed by the Perl porting team.
3694 The F<Changes> file for exhaustive details on what changed.
3696 The F<INSTALL> file for how to build Perl.
3698 The F<README> file for general stuff.
3700 The F<Artistic> and F<Copying> files for copyright information.
3704 Written by Jarkko Hietaniemi <F<jhi@iki.fi>>.