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 Note about build and test times
23 Note that the build times can vary considerably. Perl 5.8.0 is about
24 twice the size of 5.6.1, and some source code files are quite large,
25 so your compiler might have hard time processing them. On a fast
26 modern system with lots of CPU and memory the build can be a matter
27 of ten minutes, but on slower/older/more heavily loaded systems it
28 can take up to eight hours, while half an hour to an hour being common.
30 Also testing times vary a lot. Perl 5.8.0 has almost six times the
31 the tests of Perl 5.6.1. Fifteen minutes to half an hour is quite
32 normal, but a slow system may easily take an hour or more.
34 =head1 Highlights In 5.8.0
40 Better Unicode support
48 New Thread Implementation
52 Better Numeric Accuracy
64 More Extensive Regression Testing
68 =head1 Incompatible Changes
70 =head2 Binary Incompatibility
72 B<Perl 5.8 is not binary compatible with earlier releases of Perl.>
74 B<You have to recompile your XS modules.>
76 (Pure Perl modules should continue to work.)
78 The major reason for the discontinuity is the new IO architecture
79 called PerlIO. PerlIO is the default configuration because without
80 it many new features of Perl 5.8 cannot be used. In other words:
81 you just have to recompile your modules containing XS code, sorry
84 In future releases of Perl, non-PerlIO aware XS modules may become
85 completely unsupported. This shouldn't be too difficult for module
86 authors, however: PerlIO has been designed as a drop-in replacement
87 (at the source code level) for the stdio interface.
89 Depending on your platform, there are also other reasons why
90 we decided to break binary compatibility, please read on.
92 =head2 64-bit platforms and malloc
94 If your pointers are 64 bits wide, the Perl malloc is no longer being
95 used because it does not work well with 8-byte pointers. Also,
96 usually the system mallocs on such platforms are much better optimized
97 for such large memory models than the Perl malloc. Some memory-hungry
98 Perl applications like the PDL don't work well with Perl's malloc.
99 Finally, other applications than Perl (such as mod_perl) tend to prefer
100 the system malloc. Such platforms include Alpha and 64-bit HPPA,
101 MIPS, PPC, and Sparc.
103 =head2 AIX Dynaloading
105 The AIX dynaloading now uses in AIX releases 4.3 and newer the native
106 dlopen interface of AIX instead of the old emulated interface. This
107 change will probably break backward compatibility with compiled
108 modules. The change was made to make Perl more compliant with other
109 applications like mod_perl which are using the AIX native interface.
111 =head2 Attributes for C<my> variables now handled at run-time
113 The C<my EXPR : ATTRS> syntax now applies variable attributes at
114 run-time. (Subroutine and C<our> variables still get attributes applied
115 at compile-time.) See L<attributes> for additional details. In particular,
116 however, this allows variable attributes to be useful for C<tie> interfaces,
117 which was a deficiency of earlier releases. Note that the new semantics
118 doesn't work with the Attribute::Handlers module (as of version 0.76).
120 =head2 Socket Extension Dynamic in VMS
122 The Socket extension is now dynamically loaded instead of being
123 statically built in. This may or may not be a problem with ancient
124 TCP/IP stacks of VMS: we do not know since we weren't able to test
125 Perl in such configurations.
127 =head2 IEEE-format Floating Point Default on OpenVMS Alpha
129 Perl now uses IEEE format (T_FLOAT) as the default internal floating
130 point format on OpenVMS Alpha, potentially breaking binary compatibility
131 with external libraries or existing data. G_FLOAT is still available as
132 a configuration option. The default on VAX (D_FLOAT) has not changed.
134 =head2 New Unicode Properties
136 Unicode I<scripts> are now supported. Scripts are similar to (and superior
137 to) Unicode I<blocks>. The difference between scripts and blocks is that
138 scripts are the glyphs used by a language or a group of languages, while
139 the blocks are more artificial groupings of (mostly) 256 characters based
140 on the Unicode numbering.
142 In general, scripts are more inclusive, but not universally so. For
143 example, while the script C<Latin> includes all the Latin characters and
144 their various diacritic-adorned versions, it does not include the various
145 punctuation or digits (since they are not solely C<Latin>).
147 A number of other properties are now supported, including C<\p{L&}>,
148 C<\p{Any}> C<\p{Assigned}>, C<\p{Unassigned}>, C<\p{Blank}> [561] and
149 C<\p{SpacePerl}> [561] (along with their C<\P{...}> versions, of course).
150 See L<perlunicode> for details, and more additions.
152 The C<In> or C<Is> prefix to names used with the C<\p{...}> and C<\P{...}>
153 are now almost always optional. The only exception is that a C<In> prefix
154 is required to signify a Unicode block when a block name conflicts with a
155 script name. For example, C<\p{Tibetan}> refers to the script, while
156 C<\p{InTibetan}> refers to the block. When there is no name conflict, you
157 can omit the C<In> from the block name (e.g. C<\p{BraillePatterns}>), but
158 to be safe, it's probably best to always use the C<In>).
160 =head2 REF(...) Instead Of SCALAR(...)
162 A reference to a reference now stringifies as "REF(0x81485ec)" instead
163 of "SCALAR(0x81485ec)" in order to be more consistent with the return
166 =head2 pack/unpack D/F recycled
168 The undocumented pack/unpack template letters D/F have been recycled
169 for better use: now they stand for long double (if supported by the
170 platform) and NV (Perl internal floating point type). (They used
171 to be aliases for d/f, but you never knew that.)
173 =head2 glob() now returns filenames in alphabetical order
175 The list of filenames from glob() (or <...>) is now by default sorted
176 alphabetically to be csh-compliant (which is what happened before
177 in most UNIX platforms). (bsd_glob() does still sort platform
178 natively, ASCII or EBCDIC, unless GLOB_ALPHASORT is specified.) [561]
186 The semantics of bless(REF, REF) were unclear and until someone proves
187 it to make some sense, it is forbidden.
191 The obsolete chat2 library that should never have been allowed
192 to escape the laboratory has been decommissioned.
196 The builtin dump() function has probably outlived most of its
197 usefulness. The core-dumping functionality will remain in future
198 available as an explicit call to C<CORE::dump()>, but in future
199 releases the behaviour of an unqualified C<dump()> call may change.
203 The very dusty examples in the eg/ directory have been removed.
204 Suggestions for new shiny examples welcome but the main issue is that
205 the examples need to be documented, tested and (most importantly)
210 The (bogus) escape sequences \8 and \9 now give an optional warning
211 ("Unrecognized escape passed through"). There is no need to \-escape
216 The *glob{FILEHANDLE} is deprecated, use *glob{IO} instead.
220 The C<package;> syntax (C<package> without an argument) has been
221 deprecated. Its semantics were never that clear and its
222 implementation even less so. If you have used that feature to
223 disallow all but fully qualified variables, C<use strict;> instead.
227 The unimplemented POSIX regex features [[.cc.]] and [[=c=]] are still
228 recognised but now cause fatal errors. The previous behaviour of
229 ignoring them by default and warning if requested was unacceptable
230 since it, in a way, falsely promised that the features could be used.
234 In future releases, non-PerlIO aware XS modules may become completely
235 unsupported. Since PerlIO is a drop-in replacement for stdio at the
236 source code level, this shouldn't be that drastic a change.
240 Previous versions of perl and some readings of some sections of Camel
241 III implied that the C<:raw> "discipline" was the inverse of C<:crlf>.
242 Turning off "clrfness" is no longer enough to make a stream truly
243 binary. So the PerlIO C<:raw> layer (or "discipline", to use the Camel
244 book's older terminology) is now formally defined as being equivalent
245 to binmode(FH) - which is in turn defined as doing whatever is
246 necessary to pass each byte as-is without any translation. In
247 particular binmode(FH) - and hence C<:raw> - will now turn off both
248 CRLF and UTF-8 translation and remove other layers (e.g. :encoding())
249 which would modify byte stream.
253 The current user-visible implementation of pseudo-hashes (the weird
254 use of the first array element) is deprecated starting from Perl 5.8.0
255 and will be removed in Perl 5.10.0, and the feature will be
256 implemented differently. Not only is the current interface rather
257 ugly, but the current implementation slows down normal array and hash
258 use quite noticeably. The C<fields> pragma interface will remain
259 available. The I<restricted hashes> interface is expected to
260 be the replacement interface (see L<Hash::Util>). If your existing
261 programs depends on the underlying implementation, consider using
262 L<Class::PseudoHash> from CPAN.
266 The syntaxes C<< @a->[...] >> and C<< %h->{...} >> have now been deprecated.
270 After years of trying, suidperl is considered to be too complex to
271 ever be considered truly secure. The suidperl functionality is likely
272 to be removed in a future release.
276 The 5.005 threads model (module C<Thread>) is deprecated and expected
277 to be removed in Perl 5.10. Multithreaded code should be migrated to
278 the new ithreads model (see L<threads>, L<threads::shared> and
283 The long deprecated uppercase aliases for the string comparison
284 operators (EQ, NE, LT, LE, GE, GT) have now been removed.
288 The tr///C and tr///U features have been removed and will not return;
289 the interface was a mistake. Sorry about that. For similar
290 functionality, see pack('U0', ...) and pack('C0', ...). [561]
294 Earlier Perls treated "sub foo (@bar)" as equivalent to "sub foo (@)".
295 The prototypes are now checked better at compile-time for invalid
296 syntax. An optional warning is generated ("Illegal character in
297 prototype...") but this may be upgraded to a fatal error in a future
302 The C<exec LIST> and C<system LIST> operations now produce warnings on
303 tainted data and in some future release they will produce fatal errors.
307 The existing behaviour when localising tied arrays and hashes is wrong,
308 and will be changed in a future release, so do not rely on the existing
309 behaviour. See L<"Localising Tied Arrays and Hashes Is Broken">.
313 =head1 Core Enhancements
315 =head2 PerlIO is Now The Default
321 IO is now by default done via PerlIO rather than system's "stdio".
322 PerlIO allows "layers" to be "pushed" onto a file handle to alter the
323 handle's behaviour. Layers can be specified at open time via 3-arg
326 open($fh,'>:crlf :utf8', $path) || ...
328 or on already opened handles via extended C<binmode>:
330 binmode($fh,':encoding(iso-8859-7)');
332 The built-in layers are: unix (low level read/write), stdio (as in
333 previous Perls), perlio (re-implementation of stdio buffering in a
334 portable manner), crlf (does CRLF <=> "\n" translation as on Win32,
335 but available on any platform). A mmap layer may be available if
336 platform supports it (mostly UNIXes).
338 Layers to be applied by default may be specified via the 'open' pragma.
340 See L</"Installation and Configuration Improvements"> for the effects
341 of PerlIO on your architecture name.
345 If your platform supports fork(), you can use the list form of C<open>
346 for pipes. For example:
348 open KID_PS, "-|", "ps", "aux" or die $!;
350 forks the ps(1) command (without spawning a shell, as there are more
351 than three arguments to open()), and reads its standard output via the
352 C<KID_PS> filehandle. See L<perlipc>.
356 File handles can be marked as accepting Perl's internal encoding of Unicode
357 (UTF-8 or UTF-EBCDIC depending on platform) by a pseudo layer ":utf8" :
359 open($fh,">:utf8","Uni.txt");
361 Note for EBCDIC users: the pseudo layer ":utf8" is erroneously named
362 for you since it's not UTF-8 what you will be getting but instead
363 UTF-EBCDIC. See L<perlunicode>, L<utf8>, and
364 http://www.unicode.org/unicode/reports/tr16/ for more information.
365 In future releases this naming may change. See L<perluniintro>
366 for more information about UTF-8.
370 If your environment variables (LC_ALL, LC_CTYPE, LANG, LANGUAGE) look
371 like you want to use UTF-8 (any of the the variables match C</utf-?8/i>),
372 your STDIN, STDOUT, STDERR handles and the default open layer
373 (see L<open>) are marked as UTF-8. (This feature, like other new
374 features that combine Unicode and I/O, work only if you are using
375 PerlIO, but that's the default.)
377 Note that after this Perl really does assume that everything is UTF-8:
378 for example if some input handle is not, Perl will probably very soon
379 complain about the input data like this "Malformed UTF-8 ..." since
380 any old eight-bit data is not legal UTF-8.
382 Note for code authors: if you want to enable your users to use UTF-8
383 as their default encoding but in your code still have eight-bit I/O streams
384 (such as images or zip files), you need to explicitly open() or binmode()
385 with C<:bytes> (see L<perlfunc/open> and L<perlfunc/binmode>), or you
386 can just use C<binmode(FH)> (nice for pre-5.8.0 backward compatibility).
390 File handles can translate character encodings from/to Perl's internal
391 Unicode form on read/write via the ":encoding()" layer.
395 File handles can be opened to "in memory" files held in Perl scalars via:
397 open($fh,'>', \$variable) || ...
401 Anonymous temporary files are available without need to
402 'use FileHandle' or other module via
404 open($fh,"+>", undef) || ...
406 That is a literal undef, not an undefined value.
410 The list form of C<open> is now implemented for pipes (at least on UNIX):
412 open($fh,"-|", 'cat', '/etc/motd')
414 creates a pipe, and runs the equivalent of exec('cat', '/etc/motd') in
419 If your locale environment variables (LANGUAGE, LC_ALL, LC_CTYPE, LANG)
420 contain the strings 'UTF-8' or 'UTF8' (case-insensitive matching),
421 the default encoding of your STDIN, STDOUT, and STDERR, and of
422 B<any subsequent file open>, is UTF-8.
426 =head2 Restricted Hashes
428 A restricted hash is restricted to a certain set of keys, no keys
429 outside the set can be added. Also individual keys can be restricted
430 so that the key cannot be deleted and the value cannot be changed.
431 No new syntax is involved: the Hash::Util module is the interface.
435 Perl used to be fragile in that signals arriving at inopportune moments
436 could corrupt Perl's internal state. Now Perl postpones handling of
437 signals until it's safe (between opcodes).
439 This change may have surprising side effects because signals no longer
440 interrupt Perl instantly. Perl will now first finish whatever it was
441 doing, like finishing an internal operation (like sort()) or an
442 external operation (like an I/O operation), and only then look at any
443 arrived signals (and before starting the next operation). No more corrupt
444 internal state since the current operation is always finished first,
445 but the signal may take more time to get heard. Note that breaking
446 out from potentially blocking operations should still work, though.
448 =head2 Unicode Overhaul
450 Unicode in general should be now much more usable than in Perl 5.6.0
451 (or even in 5.6.1). Unicode can be used in hash keys, Unicode in
452 regular expressions should work now, Unicode in tr/// should work now,
453 Unicode in I/O should work now. See L<perluniintro> for introduction
454 and L<perlunicode> for details.
460 The Unicode Character Database coming with Perl has been upgraded
461 to Unicode 3.2.0. For more information, see http://www.unicode.org/ .
462 [561+] (5.6.1 has UCD 3.0.1.)
466 For developers interested in enhancing Perl's Unicode capabilities:
467 almost all the UCD files are included with the Perl distribution in
468 the F<lib/unicore> subdirectory. The most notable omission, for space
469 considerations, is the Unihan database.
473 The properties \p{Blank} and \p{SpacePerl} have been added. "Blank" is like
474 C isblank(), that is, it contains only "horizontal whitespace" (the space
475 character is, the newline isn't), and the "SpacePerl" is the Unicode
476 equivalent of C<\s> (\p{Space} isn't, since that includes the vertical
477 tabulator character, whereas C<\s> doesn't.)
479 See "New Unicode Properties" earlier in this document for additional
480 information on changes with Unicode properties.
484 =head2 Understanding of Numbers
486 In general a lot of fixing has happened in the area of Perl's
487 understanding of numbers, both integer and floating point. Since in
488 many systems the standard number parsing functions like C<strtoul()>
489 and C<atof()> seem to have bugs, Perl tries to work around their
490 deficiencies. This results hopefully in more accurate numbers.
492 Perl now tries internally to use integer values in numeric conversions
493 and basic arithmetics (+ - * /) if the arguments are integers, and
494 tries also to keep the results stored internally as integers.
495 This change leads to often slightly faster and always less lossy
496 arithmetics. (Previously Perl always preferred floating point numbers
499 =head2 Arrays now always interpolate into double-quoted strings [561]
501 In double-quoted strings, arrays now interpolate, no matter what. The
502 behavior in earlier versions of perl 5 was that arrays would interpolate
503 into strings if the array had been mentioned before the string was
504 compiled, and otherwise Perl would raise a fatal compile-time error.
505 In versions 5.000 through 5.003, the error was
507 Literal @example now requires backslash
509 In versions 5.004_01 through 5.6.0, the error was
511 In string, @example now must be written as \@example
513 The idea here was to get people into the habit of writing
514 C<"fred\@example.com"> when they wanted a literal C<@> sign, just as
515 they have always written C<"Give me back my \$5"> when they wanted a
518 Starting with 5.6.1, when Perl now sees an C<@> sign in a
519 double-quoted string, it I<always> attempts to interpolate an array,
520 regardless of whether or not the array has been used or declared
521 already. The fatal error has been downgraded to an optional warning:
523 Possible unintended interpolation of @example in string
525 This warns you that C<"fred@example.com"> is going to turn into
526 C<fred.com> if you don't backslash the C<@>.
527 See http://www.plover.com/~mjd/perl/at-error.html for more details
528 about the history here.
530 =head2 Miscellaneous Changes
536 AUTOLOAD is now lvaluable, meaning that you can add the :lvalue attribute
537 to AUTOLOAD subroutines and you can assign to the AUTOLOAD return value.
541 The $Config{byteorder} (and corresponding BYTEORDER in config.h) was
542 previously wrong in platforms if sizeof(long) was 4, but sizeof(IV)
543 was 8. The byteorder was only sizeof(long) bytes long (1234 or 4321),
544 but now it is correctly sizeof(IV) bytes long, (12345678 or 87654321).
545 (This problem didn't affect Windows platforms.)
547 Also, $Config{byteorder} is now computed dynamically--this is more
548 robust with "fat binaries" where an executable image contains binaries
549 for more than one binary platform, and when cross-compiling.
553 C<perl -d:Module=arg,arg,arg> now works (previously one couldn't pass
554 in multiple arguments.)
558 C<do> followed by a bareword now ensures that this bareword isn't
559 a keyword (to avoid a bug where C<do q(foo.pl)> tried to call a
560 subroutine called C<q>). This means that for example instead of
561 C<do format()> you must write C<do &format()>.
565 The builtin dump() now gives an optional warning
566 C<dump() better written as CORE::dump()>,
567 meaning that by default C<dump(...)> is resolved as the builtin
568 dump() which dumps core and aborts, not as (possibly) user-defined
569 C<sub dump>. To call the latter, qualify the call as C<&dump(...)>.
570 (The whole dump() feature is to considered deprecated, and possibly
571 removed/changed in future releases.)
575 chomp() and chop() are now overridable. Note, however, that their
576 prototype (as given by C<prototype("CORE::chomp")> is undefined,
577 because it cannot be expressed and therefore one cannot really write
578 replacements to override these builtins.
582 END blocks are now run even if you exit/die in a BEGIN block.
583 Internally, the execution of END blocks is now controlled by
584 PL_exit_flags & PERL_EXIT_DESTRUCT_END. This enables the new
585 behaviour for Perl embedders. This will default in 5.10. See
590 Formats now support zero-padded decimal fields.
594 Although "you shouldn't do that", it was possible to write code that
595 depends on Perl's hashed key order (Data::Dumper does this). The new
596 algorithm "One-at-a-Time" produces a different hashed key order.
597 More details are in L</"Performance Enhancements">.
601 lstat(FILEHANDLE) now gives a warning because the operation makes no sense.
602 In future releases this may become a fatal error.
606 Spurious syntax errors generated in certain situations, when glob()
607 caused File::Glob to be loaded for the first time, have been fixed. [561]
611 Lvalue subroutines can now return C<undef> in list context. However,
612 the lvalue subroutine feature still remains experimental. [561+]
616 A lost warning "Can't declare ... dereference in my" has been
617 restored (Perl had it earlier but it became lost in later releases.)
621 A new special regular expression variable has been introduced:
622 C<$^N>, which contains the most-recently closed group (submatch).
626 C<no Module;> does not produce an error even if Module does not have an
627 unimport() method. This parallels the behavior of C<use> vis-a-vis
632 The numerical comparison operators return C<undef> if either operand
633 is a NaN. Previously the behaviour was unspecified.
637 C<our> can now have an experimental optional attribute C<unique> that
638 affects how global variables are shared among multiple interpreters,
643 The following builtin functions are now overridable: each(), keys(),
644 pop(), push(), shift(), splice(), unshift(). [561]
648 C<pack() / unpack()> can now group template letters with C<()> and then
649 apply repetition/count modifiers on the groups.
653 C<pack() / unpack()> can now process the Perl internal numeric types:
654 IVs, UVs, NVs-- and also long doubles, if supported by the platform.
655 The template letters are C<j>, C<J>, C<F>, and C<D>.
659 C<pack('U0a*', ...)> can now be used to force a string to UTF8.
663 my __PACKAGE__ $obj now works. [561]
667 POSIX::sleep() now returns the number of I<unslept> seconds
668 (as the POSIX standard says), as opposed to CORE::sleep() which
669 returns the number of slept seconds.
673 The printf() and sprintf() now support parameter reordering using the
674 C<%\d+\$> and C<*\d+\$> syntaxes. For example
676 print "%2\$s %1\$s\n", "foo", "bar";
678 will print "bar foo\n". This feature helps in writing
679 internationalised software, and in general when the order
680 of the parameters can vary.
684 The (\&) prototype now works properly. [561]
688 prototype(\[$@%&]) is now available to implicitly create references
689 (useful for example if you want to emulate the tie() interface).
693 A new command-line option, C<-t> is available. It is the
694 little brother of C<-T>: instead of dying on taint violations,
695 lexical warnings are given. B<This is only meant as a temporary
696 debugging aid while securing the code of old legacy applications.
697 This is not a substitute for -T.>
701 In other taint news, the C<exec LIST> and C<system LIST> have now been
702 considered too risky (think C<exec @ARGV>: it can start any program
703 with any arguments), and now the said forms cause a warning under
704 lexical warnings. You should carefully launder the arguments to
705 guarantee their validity. In future releases of Perl the forms will
706 become fatal errors so consider starting laundering now.
710 Tied hash interfaces are now required to have the EXISTS and DELETE
711 methods (either own or inherited).
715 If tr/// is just counting characters, it doesn't attempt to
720 untie() will now call an UNTIE() hook if it exists. See L<perltie>
725 L<utime> now supports C<utime undef, undef, @files> to change the
726 file timestamps to the current time.
730 The rules for allowing underscores (underbars) in numeric constants
731 have been relaxed and simplified: now you can have an underscore
732 simply B<between digits>.
736 Rather than relying on C's argv[0] (which may not contain a full pathname)
737 where possible $^X is now set by asking the operating system.
738 (eg by reading F</proc/self/exe> on Linux, F</proc/curproc/file> on FreeBSD)
742 A new variable, C<${^TAINT}>, indicates whether taint mode is enabled.
746 You can now override the readline() builtin, and this overrides also
747 the <FILEHANDLE> angle bracket operator.
751 The command-line options -s and -F are now recognized on the shebang
756 Use of the C</c> match modifier without an accompanying C</g> modifier
757 elicits a new warning: C<Use of /c modifier is meaningless without /g>.
759 Use of C</c> in substitutions, even with C</g>, elicits
760 C<Use of /c modifier is meaningless in s///>.
762 Use of C</g> with C<split> elicits C<Use of /g modifier is meaningless
767 Support for the C<CLONE> special subroutine had been added.
768 With ithreads, when a new thread is created, all Perl data is cloned,
769 however non-Perl data cannot be cloned automatically. In C<CLONE> you
770 can do whatever you need to do, like for example handle the cloning of
771 non-Perl data, if necessary. C<CLONE> will be executed once for every
772 package that has it defined or inherited. It will be called in the
773 context of the new thread, so all modifications are made in the new area.
779 =head1 Modules and Pragmata
781 =head2 New Modules and Pragmata
787 C<Attribute::Handlers>, originally by Damian Conway and now maintained
788 by Arthur Bergman, allows a class to define attribute handlers.
791 use Attribute::Handlers;
792 sub Wolf :ATTR(SCALAR) { print "howl!\n" }
794 # later, in some package using or inheriting from MyPack...
796 my MyPack $Fluffy : Wolf; # the attribute handler Wolf will be called
798 Both variables and routines can have attribute handlers. Handlers can
799 be specific to type (SCALAR, ARRAY, HASH, or CODE), or specific to the
800 exact compilation phase (BEGIN, CHECK, INIT, or END).
801 See L<Attribute::Handlers>.
805 C<B::Concise>, by Stephen McCamant, is a new compiler backend for
806 walking the Perl syntax tree, printing concise info about ops.
807 The output is highly customisable. See L<B::Concise>. [561+]
811 The new bignum, bigint, and bigrat pragmas, by Tels, implement
812 transparent bignum support (using the Math::BigInt, Math::BigFloat,
813 and Math::BigRat backends).
817 C<Class::ISA>, by Sean Burke, is a module for reporting the search
818 path for a class's ISA tree. See L<Class::ISA>.
822 C<Cwd> now has a split personality: if possible, an XS extension is
823 used, (this will hopefully be faster, more secure, and more robust)
824 but if not possible, the familiar Perl implementation is used.
828 C<Devel::PPPort>, originally by Kenneth Albanowski and now
829 maintained by Paul Marquess, has been added. It is primarily used
830 by C<h2xs> to enhance portability of XS modules between different
831 versions of Perl. See L<Devel::PPPort>.
835 C<Digest>, frontend module for calculating digests (checksums), from
836 Gisle Aas, has been added. See L<Digest>.
840 C<Digest::MD5> for calculating MD5 digests (checksums) as defined in
841 RFC 1321, from Gisle Aas, has been added. See L<Digest::MD5>.
843 use Digest::MD5 'md5_hex';
845 $digest = md5_hex("Thirsty Camel");
847 print $digest, "\n"; # 01d19d9d2045e005c3f1b80e8b164de1
849 NOTE: the C<MD5> backward compatibility module is deliberately not
850 included since its further use is discouraged.
852 See also L<PerlIO::via::QuotedPrint>.
856 C<Encode>, originally by Nick Ing-Simmons and now maintained by Dan
857 Kogai, provides a mechanism to translate between different character
858 encodings. Support for Unicode, ISO-8859-1, and ASCII are compiled in
859 to the module. Several other encodings (like the rest of the
860 ISO-8859, CP*/Win*, Mac, KOI8-R, three variants EBCDIC, Chinese,
861 Japanese, and Korean encodings) are included and can be loaded at
862 runtime. (For space considerations, the largest Chinese encodings
863 have been separated into their own CPAN module, Encode::HanExtra,
864 which Encode will use if available). See L<Encode>.
866 Any encoding supported by Encode module is also available to the
867 ":encoding()" layer if PerlIO is used.
871 C<Hash::Util> is the interface to the new I<restricted hashes>
872 feature. (Implemented by Jeffrey Friedl, Nick Ing-Simmons, and
873 Michael Schwern.) See L<Hash::Util>.
877 C<I18N::Langinfo> can be used to query locale information.
878 See L<I18N::Langinfo>.
882 C<I18N::LangTags>, by Sean Burke, has functions for dealing with
883 RFC3066-style language tags. See L<I18N::LangTags>.
887 C<ExtUtils::Constant>, by Nicholas Clark, is a new tool for extension
888 writers for generating XS code to import C header constants.
889 See L<ExtUtils::Constant>.
893 C<Filter::Simple>, by Damian Conway, is an easy-to-use frontend to
894 Filter::Util::Call. See L<Filter::Simple>.
900 use Filter::Simple sub {
901 while (my ($from, $to) = splice @_, 0, 2) {
910 use MyFilter qr/red/ => 'green';
912 print "red\n"; # this code is filtered, will print "green\n"
913 print "bored\n"; # this code is filtered, will print "bogreen\n"
917 print "red\n"; # this code is not filtered, will print "red\n"
921 C<File::Temp>, by Tim Jenness, allows one to create temporary files
922 and directories in an easy, portable, and secure way. See L<File::Temp>.
927 C<Filter::Util::Call>, by Paul Marquess, provides you with the
928 framework to write I<source filters> in Perl. For most uses, the
929 frontend Filter::Simple is to be preferred. See L<Filter::Util::Call>.
933 C<if>, by Ilya Zakharevich, is a new pragma for conditional inclusion
938 L<libnet>, by Graham Barr, is a collection of perl5 modules related
939 to network programming. See L<Net::FTP>, L<Net::NNTP>, L<Net::Ping>
940 (not part of libnet, but related), L<Net::POP3>, L<Net::SMTP>,
943 Perl installation leaves libnet unconfigured; use F<libnetcfg>
948 C<List::Util>, by Graham Barr, is a selection of general-utility
949 list subroutines, such as sum(), min(), first(), and shuffle().
954 C<Locale::Constants>, C<Locale::Country>, C<Locale::Currency>
955 C<Locale::Language>, and L<Locale::Script>, by Neil Bowers, have
956 been added. They provide the codes for various locale standards, such
957 as "fr" for France, "usd" for US Dollar, and "ja" for Japanese.
961 $country = code2country('jp'); # $country gets 'Japan'
962 $code = country2code('Norway'); # $code gets 'no'
964 See L<Locale::Constants>, L<Locale::Country>, L<Locale::Currency>,
965 and L<Locale::Language>.
969 C<Locale::Maketext>, by Sean Burke, is a localization framework. See
970 L<Locale::Maketext>, and L<Locale::Maketext::TPJ13>. The latter is an
971 article about software localization, originally published in The Perl
972 Journal #13, and republished here with kind permission.
976 C<Math::BigRat> for big rational numbers, to accompany Math::BigInt and
977 Math::BigFloat, from Tels. See L<Math::BigRat>.
981 C<Memoize> can make your functions faster by trading space for time,
982 from Mark-Jason Dominus. See L<Memoize>.
986 C<MIME::Base64>, by Gisle Aas, allows you to encode data in base64,
987 as defined in RFC 2045 - I<MIME (Multipurpose Internet Mail
992 $encoded = encode_base64('Aladdin:open sesame');
993 $decoded = decode_base64($encoded);
995 print $encoded, "\n"; # "QWxhZGRpbjpvcGVuIHNlc2FtZQ=="
1001 C<MIME::QuotedPrint>, by Gisle Aas, allows you to encode data
1002 in quoted-printable encoding, as defined in RFC 2045 - I<MIME
1003 (Multipurpose Internet Mail Extensions)>.
1005 use MIME::QuotedPrint;
1007 $encoded = encode_qp("Smiley in Unicode: \x{263a}");
1008 $decoded = decode_qp($encoded);
1010 print $encoded, "\n"; # "Smiley in Unicode: =263A"
1012 See also L<PerlIO::via::QuotedPrint>.
1016 C<NEXT>, by Damian Conway, is a pseudo-class for method redispatch.
1021 C<open> is a new pragma for setting the default I/O layers
1026 C<PerlIO::scalar>, by Nick Ing-Simmons, provides the implementation
1027 of IO to "in memory" Perl scalars as discussed above. It also serves
1028 as an example of a loadable PerlIO layer. Other future possibilities
1029 include PerlIO::Array and PerlIO::Code. See L<PerlIO::scalar>.
1033 C<PerlIO::via>, by Nick Ing-Simmons, acts as a PerlIO layer and wraps
1034 PerlIO layer functionality provided by a class (typically implemented
1039 C<PerlIO::via::QuotedPrint>, by Elizabeth Mattijsen, is an example
1040 of a C<PerlIO::via> class:
1042 use PerlIO::via::QuotedPrint;
1043 open($fh,">:via(QuotedPrint)",$path);
1045 This will automatically convert everything output to C<$fh> to
1046 Quoted-Printable. See L<PerlIO::via> and L<PerlIO::via::QuotedPrint>.
1050 C<Pod::ParseLink>, by Russ Allbery, has been added,
1051 to parse LZ<><> links in pods as described in the new
1056 C<Pod::Text::Overstrike>, by Joe Smith, has been added.
1057 It converts POD data to formatted overstrike text.
1058 See L<Pod::Text::Overstrike>. [561+]
1062 C<Scalar::Util> is a selection of general-utility scalar subroutines,
1063 such as blessed(), reftype(), and tainted(). See L<Scalar::Util>.
1067 C<sort> is a new pragma for controlling the behaviour of sort().
1071 C<Storable> gives persistence to Perl data structures by allowing the
1072 storage and retrieval of Perl data to and from files in a fast and
1073 compact binary format. Because in effect Storable does serialisation
1074 of Perl data structures, with it you can also clone deep, hierarchical
1075 datastructures. Storable was originally created by Raphael Manfredi,
1076 but it is now maintained by Abhijit Menon-Sen. Storable has been
1077 enhanced to understand the two new hash features, Unicode keys and
1078 restricted hashes. See L<Storable>.
1082 C<Switch>, by Damian Conway, has been added. Just by saying
1086 you have C<switch> and C<case> available in Perl.
1092 case 1 { print "number 1" }
1093 case "a" { print "string a" }
1094 case [1..10,42] { print "number in list" }
1095 case (@array) { print "number in list" }
1096 case /\w+/ { print "pattern" }
1097 case qr/\w+/ { print "pattern" }
1098 case (%hash) { print "entry in hash" }
1099 case (\%hash) { print "entry in hash" }
1100 case (\&sub) { print "arg to subroutine" }
1101 else { print "previous case not true" }
1108 C<Test::More>, by Michael Schwern, is yet another framework for writing
1109 test scripts, more extensive than Test::Simple. See L<Test::More>.
1113 C<Test::Simple>, by Michael Schwern, has basic utilities for writing
1114 tests. See L<Test::Simple>.
1118 C<Text::Balanced>, by Damian Conway, has been added, for extracting
1119 delimited text sequences from strings.
1121 use Text::Balanced 'extract_delimited';
1123 ($a, $b) = extract_delimited("'never say never', he never said", "'", '');
1125 $a will be "'never say never'", $b will be ', he never said'.
1127 In addition to extract_delimited(), there are also extract_bracketed(),
1128 extract_quotelike(), extract_codeblock(), extract_variable(),
1129 extract_tagged(), extract_multiple(), gen_delimited_pat(), and
1130 gen_extract_tagged(). With these, you can implement rather advanced
1131 parsing algorithms. See L<Text::Balanced>.
1135 C<threads>, by Arthur Bergman, is an interface to interpreter threads.
1136 Interpreter threads (ithreads) is the new thread model introduced in
1137 Perl 5.6 but only available as an internal interface for extension
1138 writers (and for Win32 Perl for C<fork()> emulation). See L<threads>,
1139 L<threads::shared>, and L<perlthrtut>.
1143 C<threads::shared>, by Arthur Bergman, allows data sharing for
1144 interpreter threads. In the ithreads model any data sharing between
1145 threads must be explicit, as opposed to the old 5.005 thread model
1146 where data sharing was implicit. See L<threads::shared>.
1150 C<Tie::File>, by Mark-Jason Dominus, associates a Perl array with the
1151 lines of a file. See L<Tie::File>.
1155 C<Tie::Memoize>, by Ilya Zakharevich, provides on-demand loaded hashes.
1156 See L<Tie::Memoize>.
1160 C<Tie::RefHash::Nestable>, by Edward Avis, allows storing hash
1161 references (unlike the standard Tie::RefHash) The module is contained
1162 within Tie::RefHash. See L<Tie::RefHash>.
1166 C<Time::HiRes>, by Douglas E. Wegscheid, provides high resolution
1167 timing (ualarm, usleep, and gettimeofday). See L<Time::HiRes>.
1171 C<Unicode::UCD> offers a querying interface to the Unicode Character
1172 Database. See L<Unicode::UCD>.
1176 C<Unicode::Collate>, by SADAHIRO Tomoyuki, implements the UCA
1177 (Unicode Collation Algorithm) for sorting Unicode strings.
1178 See L<Unicode::Collate>.
1182 C<Unicode::Normalize>, by SADAHIRO Tomoyuki, implements the various
1183 Unicode normalization forms. See L<Unicode::Normalize>.
1187 C<XS::APItest>, by Tim Jenness, is a test extension that exercises XS
1188 APIs. Currently only C<printf()> is tested: how to output various
1189 basic data types from XS.
1193 C<XS::Typemap>, by Tim Jenness, is a test extension that exercises
1194 XS typemaps. Nothing gets installed, but the code is worth studying
1195 for extension writers.
1199 =head2 Updated And Improved Modules and Pragmata
1205 The following independently supported modules have been updated to the
1206 newest versions from CPAN: CGI, CPAN, DB_File, File::Spec, File::Temp,
1207 Getopt::Long, Math::BigFloat, Math::BigInt, the podlators bundle
1208 (Pod::Man, Pod::Text), Pod::LaTeX [561+], Pod::Parser, Storable,
1209 Term::ANSIColor, Test, Text-Tabs+Wrap.
1213 attributes::reftype() now works on tied arguments.
1217 AutoLoader can now be disabled with C<no AutoLoader;>.
1221 B::Deparse has been significantly enhanced by Robin Houston. It can
1222 now deparse almost all of the standard test suite (so that the tests
1223 still succeed). There is a make target "test.deparse" for trying this
1228 Carp now has better interface documentation, and the @CARP_NOT
1229 interface has been added to get optional control over where errors
1230 are reported independently of @ISA, by Ben Tilly.
1234 Class::Struct can now define the classes in compile time.
1238 Class::Struct now assigns the array/hash element if the accessor
1239 is called with an array/hash element as the B<sole> argument.
1243 The return value of Cwd::fastcwd() is now tainted.
1247 Data::Dumper now has an option to sort hashes.
1251 Data::Dumper now has an option to dump code references
1256 DB_File now supports newer Berkeley DB versions, among
1261 Devel::Peek now has an interface for the Perl memory statistics
1262 (this works only if you are using perl's malloc, and if you have
1263 compiled with debugging).
1267 The English module can now be used without the infamous performance
1270 use English '-no_match_vars';
1272 (Assuming, of course, that you don't need the troublesome variables
1273 C<$`>, C<$&>, or C<$'>.) Also, introduced C<@LAST_MATCH_START> and
1274 C<@LAST_MATCH_END> English aliases for C<@-> and C<@+>.
1278 ExtUtils::MakeMaker has been significantly cleaned up and fixed.
1279 The enhanced version has also been backported to earlier releases
1280 of Perl and submitted to CPAN so that the earlier releases can
1285 The arguments of WriteMakefile() in Makefile.PL are now checked
1286 for sanity much more carefully than before. This may cause new
1287 warnings when modules are being installed. See L<ExtUtils::MakeMaker>
1292 ExtUtils::MakeMaker now uses File::Spec internally, which hopefully
1293 leads to better portability.
1297 Fcntl, Socket, and Sys::Syslog have been rewritten by Nicholas Clark
1298 to use the new-style constant dispatch section (see L<ExtUtils::Constant>).
1299 This means that they will be more robust and hopefully faster.
1303 File::Find now chdir()s correctly when chasing symbolic links. [561]
1307 File::Find now has pre- and post-processing callbacks. It also
1308 correctly changes directories when chasing symbolic links. Callbacks
1309 (naughtily) exiting with "next;" instead of "return;" now work.
1313 File::Find is now (again) reentrant. It also has been made
1318 The warnings issued by File::Find now belong to their own category.
1319 You can enable/disable them with C<use/no warnings 'File::Find';>.
1323 File::Glob::glob() has been renamed to File::Glob::bsd_glob()
1324 because the name clashes with the builtin glob(). The older
1325 name is still available for compatibility, but is deprecated. [561]
1329 File::Glob now supports C<GLOB_LIMIT> constant to limit the size of
1330 the returned list of filenames.
1334 IPC::Open3 now allows the use of numeric file descriptors.
1338 IO::Socket now has an atmark() method, which returns true if the socket
1339 is positioned at the out-of-band mark. The method is also exportable
1340 as a sockatmark() function.
1344 IO::Socket::INET failed to open the specified port if the service name
1345 was not known. It now correctly uses the supplied port number as is. [561]
1349 IO::Socket::INET has support for the ReusePort option (if your
1350 platform supports it). The Reuse option now has an alias, ReuseAddr.
1351 For clarity, you may want to prefer ReuseAddr.
1355 IO::Socket::INET now supports a value of zero for C<LocalPort>
1356 (usually meaning that the operating system will make one up.)
1360 'use lib' now works identically to @INC. Removing directories
1361 with 'no lib' now works.
1365 Math::BigFloat and Math::BigInt have undergone a full rewrite by Tels.
1366 They are now magnitudes faster, and they support various bignum
1367 libraries such as GMP and PARI as their backends.
1371 Math::Complex handles inf, NaN etc., better.
1375 Net::Ping has been considerably enhanced by Rob Brown: multihoming is
1376 now supported, Win32 functionality is better, there is now time
1377 measuring functionality (optionally high-resolution using
1378 Time::HiRes), and there is now "external" protocol which uses
1379 Net::Ping::External module which runs your external ping utility and
1380 parses the output. A version of Net::Ping::External is available in
1383 Note that some of the Net::Ping tests are disabled when running
1384 under the Perl distribution since one cannot assume one or more
1385 of the following: enabled echo port at localhost, full Internet
1386 connectivity, or sympathetic firewalls. You can set the environment
1387 variable PERL_TEST_Net_Ping to "1" (one) before running the Perl test
1388 suite to enable all the Net::Ping tests.
1392 POSIX::sigaction() is now much more flexible and robust.
1393 You can now install coderef handlers, 'DEFAULT', and 'IGNORE'
1394 handlers, installing new handlers was not atomic.
1398 In Safe, C<%INC> is now localised in a Safe compartment so that
1403 In SDBM_File on dosish platforms, some keys went missing because of
1404 lack of support for files with "holes". A workaround for the problem
1409 In Search::Dict one can now have a pre-processing hook for the
1410 lines being searched.
1414 The Shell module now has an OO interface.
1418 In Sys::Syslog there is now a failover mechanism that will go
1419 through alternative connection mechanisms until the message
1420 is successfully logged.
1424 The Test module has been significantly enhanced.
1428 Time::Local::timelocal() does not handle fractional seconds anymore.
1429 The rationale is that neither does localtime(), and timelocal() and
1430 localtime() are supposed to be inverses of each other.
1434 The vars pragma now supports declaring fully qualified variables.
1435 (Something that C<our()> does not and will not support.)
1439 The C<utf8::> name space (as in the pragma) provides various
1440 Perl-callable functions to provide low level access to Perl's
1441 internal Unicode representation. At the moment only length()
1442 has been implemented.
1446 =head1 Utility Changes
1452 Emacs perl mode (emacs/cperl-mode.el) has been updated to version
1457 F<emacs/e2ctags.pl> is now much faster.
1461 C<enc2xs> is a tool for people adding their own encodings to the
1466 C<h2ph> now supports C trigraphs.
1470 C<h2xs> now produces a template README.
1474 C<h2xs> now uses C<Devel::PPPort> for better portability between
1475 different versions of Perl.
1479 C<h2xs> uses the new L<ExtUtils::Constant|ExtUtils::Constant> module
1480 which will affect newly created extensions that define constants.
1481 Since the new code is more correct (if you have two constants where the
1482 first one is a prefix of the second one, the first constant B<never>
1483 got defined), less lossy (it uses integers for integer constant,
1484 as opposed to the old code that used floating point numbers even for
1485 integer constants), and slightly faster, you might want to consider
1486 regenerating your extension code (the new scheme makes regenerating
1487 easy). L<h2xs> now also supports C trigraphs.
1491 C<libnetcfg> has been added to configure libnet.
1495 C<perlbug> is now much more robust. It also sends the bug report to
1496 perl.org, not perl.com.
1500 C<perlcc> has been rewritten and its user interface (that is,
1501 command line) is much more like that of the UNIX C compiler, cc.
1502 (The perlbc tools has been removed. Use C<perlcc -B> instead.)
1503 B<Note that perlcc is still considered very experimental and
1508 C<perlivp> is a new Installation Verification Procedure utility
1509 for running any time after installing Perl.
1513 C<piconv> is an implementation of the character conversion utility
1514 C<iconv>, demonstrating the new Encode module.
1518 C<pod2html> now allows specifying a cache directory.
1522 C<pod2html> now produces XHTML 1.0.
1526 C<pod2html> now understands POD written using different line endings
1527 (PC-like CRLF versus UNIX-like LF versus MacClassic-like CR).
1531 C<s2p> has been completely rewritten in Perl. (It is in fact a full
1532 implementation of sed in Perl: you can use the sed functionality by
1533 using the C<psed> utility.)
1537 C<xsubpp> now understands POD documentation embedded in the *.xs
1542 C<xsubpp> now supports the OUT keyword.
1546 =head1 New Documentation
1552 perl56delta details the changes between the 5.005 release and the
1557 perlclib documents the internal replacements for standard C library
1558 functions. (Interesting only for extension writers and Perl core
1563 perldebtut is a Perl debugging tutorial. [561+]
1567 perlebcdic contains considerations for running Perl on EBCDIC
1572 perlintro is a gentle introduction to Perl.
1576 perliol documents the internals of PerlIO with layers.
1580 perlmodstyle is a style guide for writing modules.
1584 perlnewmod tells about writing and submitting a new module. [561+]
1588 perlpacktut is a pack() tutorial.
1592 perlpod has been rewritten to be clearer and to record the best
1593 practices gathered over the years.
1597 perlpodspec is a more formal specification of the pod format,
1598 mainly of interest for writers of pod applications, not to
1599 people writing in pod.
1603 perlretut is a regular expression tutorial. [561+]
1607 perlrequick is a regular expressions quick-start guide.
1608 Yes, much quicker than perlretut. [561]
1612 perltodo has been updated.
1616 perltootc has been renamed as perltooc (to not to conflict
1617 with perltoot in filesystems restricted to "8.3" names).
1621 perluniintro is an introduction to using Unicode in Perl.
1622 (perlunicode is more of a detailed reference and background
1627 perlutil explains the command line utilities packaged with the Perl
1628 distribution. [561+]
1632 The following platform-specific documents are available before
1633 the installation as README.I<platform>, and after the installation
1636 perlaix perlamiga perlapollo perlbeos perlbs2000
1637 perlce perlcygwin perldgux perldos perlepoc perlfreebsd perlhpux
1638 perlhurd perlirix perlmachten perlmacos perlmint perlmpeix
1639 perlnetware perlos2 perlos390 perlplan9 perlqnx perlsolaris
1640 perltru64 perluts perlvmesa perlvms perlvos perlwin32
1642 These documents usually detail one or more of the following subjects:
1643 configuring, building, testing, installing, and sometimes also using
1644 Perl on the said platform.
1646 Eastern Asian Perl users are now welcomed in their own languages:
1647 README.jp (Japanese), README.ko (Korean), README.cn (simplified
1648 Chinese) and README.tw (traditional Chinese), which are written in
1649 normal pod but encoded in EUC-JP, EUC-KR, EUC-CN and Big5. These
1650 will get installed as
1652 perljp perlko perlcn perltw
1658 The documentation for the POSIX-BC platform is called "BS2000", to avoid
1659 confusion with the Perl POSIX module.
1663 The documentation for the WinCE platform is called perlce (README.ce
1664 in the source code kit), to avoid confusion with the perlwin32
1665 documentation on 8.3-restricted filesystems.
1669 =head1 Performance Enhancements
1675 map() could get pathologically slow when the result list it generates
1676 is larger than the source list. The performance has been improved for
1677 common scenarios. [561]
1681 sort() is also fully reentrant, in the sense that the sort function
1682 can itself call sort(). This did not work reliably in previous
1687 sort() has been changed to use primarily mergesort internally as
1688 opposed to the earlier quicksort. For very small lists this may
1689 result in slightly slower sorting times, but in general the speedup
1690 should be at least 20%. Additional bonuses are that the worst case
1691 behaviour of sort() is now better (in computer science terms it now
1692 runs in time O(N log N), as opposed to quicksort's Theta(N**2)
1693 worst-case run time behaviour), and that sort() is now stable
1694 (meaning that elements with identical keys will stay ordered as they
1695 were before the sort). See the C<sort> pragma for information.
1697 The story in more detail: suppose you want to serve yourself a little
1700 @digits = ( 3,1,4,1,5,9 );
1702 A numerical sort of the digits will yield (1,1,3,4,5,9), as expected.
1703 Which C<1> comes first is hard to know, since one C<1> looks pretty
1704 much like any other. You can regard this as totally trivial,
1705 or somewhat profound. However, if you just want to sort the even
1706 digits ahead of the odd ones, then what will
1708 sort { ($a % 2) <=> ($b % 2) } @digits;
1710 yield? The only even digit, C<4>, will come first. But how about
1711 the odd numbers, which all compare equal? With the quicksort algorithm
1712 used to implement Perl 5.6 and earlier, the order of ties is left up
1713 to the sort. So, as you add more and more digits of Pi, the order
1714 in which the sorted even and odd digits appear will change.
1715 and, for sufficiently large slices of Pi, the quicksort algorithm
1716 in Perl 5.8 won't return the same results even if reinvoked with the
1717 same input. The justification for this rests with quicksort's
1718 worst case behavior. If you run
1720 sort { $a <=> $b } ( 1 .. $N , 1 .. $N );
1722 (something you might approximate if you wanted to merge two sorted
1723 arrays using sort), doubling $N doesn't just double the quicksort time,
1724 it I<quadruples> it. Quicksort has a worst case run time that can
1725 grow like N**2, so-called I<quadratic> behaviour, and it can happen
1726 on patterns that may well arise in normal use. You won't notice this
1727 for small arrays, but you I<will> notice it with larger arrays,
1728 and you may not live long enough for the sort to complete on arrays
1729 of a million elements. So the 5.8 quicksort scrambles large arrays
1730 before sorting them, as a statistical defence against quadratic behaviour.
1731 But that means if you sort the same large array twice, ties may be
1732 broken in different ways.
1734 Because of the unpredictability of tie-breaking order, and the quadratic
1735 worst-case behaviour, quicksort was I<almost> replaced completely with
1736 a stable mergesort. I<Stable> means that ties are broken to preserve
1737 the original order of appearance in the input array. So
1739 sort { ($a % 2) <=> ($b % 2) } (3,1,4,1,5,9);
1741 will yield (4,3,1,1,5,9), guaranteed. The even and odd numbers
1742 appear in the output in the same order they appeared in the input.
1743 Mergesort has worst case O(N log N) behaviour, the best value
1744 attainable. And, ironically, this mergesort does particularly
1745 well where quicksort goes quadratic: mergesort sorts (1..$N, 1..$N)
1746 in O(N) time. But quicksort was rescued at the last moment because
1747 it is faster than mergesort on certain inputs and platforms.
1748 For example, if you really I<don't> care about the order of even
1749 and odd digits, quicksort will run in O(N) time; it's very good
1750 at sorting many repetitions of a small number of distinct elements.
1751 The quicksort divide and conquer strategy works well on platforms
1752 with relatively small, very fast, caches. Eventually, the problem gets
1753 whittled down to one that fits in the cache, from which point it
1754 benefits from the increased memory speed.
1756 Quicksort was rescued by implementing a sort pragma to control aspects
1757 of the sort. The B<stable> subpragma forces stable behaviour,
1758 regardless of algorithm. The B<_quicksort> and B<_mergesort>
1759 subpragmas are heavy-handed ways to select the underlying implementation.
1760 The leading C<_> is a reminder that these subpragmas may not survive
1761 beyond 5.8. More appropriate mechanisms for selecting the implementation
1762 exist, but they wouldn't have arrived in time to save quicksort.
1766 Hashes now use Bob Jenkins "One-at-a-Time" hashing key algorithm
1767 ( http://burtleburtle.net/bob/hash/doobs.html ). This algorithm is
1768 reasonably fast while producing a much better spread of values than
1769 the old hashing algorithm (originally by Chris Torek, later tweaked by
1770 Ilya Zakharevich). Hash values output from the algorithm on a hash of
1771 all 3-char printable ASCII keys comes much closer to passing the
1772 DIEHARD random number generation tests. According to perlbench, this
1773 change has not affected the overall speed of Perl.
1777 unshift() should now be noticeably faster.
1781 =head1 Installation and Configuration Improvements
1783 =head2 Generic Improvements
1789 INSTALL now explains how you can configure Perl to use 64-bit
1790 integers even on non-64-bit platforms.
1794 Policy.sh policy change: if you are reusing a Policy.sh file
1795 (see INSTALL) and you use Configure -Dprefix=/foo/bar and in the old
1796 Policy $prefix eq $siteprefix and $prefix eq $vendorprefix, all of
1797 them will now be changed to the new prefix, /foo/bar. (Previously
1798 only $prefix changed.) If you do not like this new behaviour,
1799 specify prefix, siteprefix, and vendorprefix explicitly.
1803 A new optional location for Perl libraries, otherlibdirs, is available.
1804 It can be used for example for vendor add-ons without disturbing Perl's
1805 own library directories.
1809 In many platforms, the vendor-supplied 'cc' is too stripped-down to
1810 build Perl (basically, 'cc' doesn't do ANSI C). If this seems
1811 to be the case and 'cc' does not seem to be the GNU C compiler
1812 'gcc', an automatic attempt is made to find and use 'gcc' instead.
1816 gcc needs to closely track the operating system release to avoid
1817 build problems. If Configure finds that gcc was built for a different
1818 operating system release than is running, it now gives a clearly visible
1819 warning that there may be trouble ahead.
1823 Since Perl 5.8 is not binary-compatible with previous releases
1824 of Perl, Configure no longer suggests including the 5.005
1829 Configure C<-S> can now run non-interactively. [561]
1833 Configure support for pdp11-style memory models has been removed due
1834 to obsolescence. [561]
1838 configure.gnu now works with options with whitespace in them.
1842 installperl now outputs everything to STDERR.
1846 Because PerlIO is now the default on most platforms, "-perlio" doesn't
1847 get appended to the $Config{archname} (also known as $^O) anymore.
1848 Instead, if you explicitly choose not to use perlio (Configure command
1849 line option -Uuseperlio), you will get "-stdio" appended.
1853 Another change related to the architecture name is that "-64all"
1854 (-Duse64bitall, or "maximally 64-bit") is appended only if your
1855 pointers are 64 bits wide. (To be exact, the use64bitall is ignored.)
1859 In AFS installations, one can configure the root of the AFS to be
1860 somewhere else than the default F</afs> by using the Configure
1861 parameter C<-Dafsroot=/some/where/else>.
1865 APPLLIB_EXP, a lesser-known configuration-time definition, has been
1866 documented. It can be used to prepend site-specific directories
1867 to Perl's default search path (@INC); see INSTALL for information.
1871 The version of Berkeley DB used when the Perl (and, presumably, the
1872 DB_File extension) was built is now available as
1873 C<@Config{qw(db_version_major db_version_minor db_version_patch)}>
1874 from Perl and as C<DB_VERSION_MAJOR_CFG DB_VERSION_MINOR_CFG
1875 DB_VERSION_PATCH_CFG> from C.
1879 Building Berkeley DB3 for compatibility modes for DB, NDBM, and ODBM
1880 has been documented in INSTALL.
1884 If you have CPAN access (either network or a local copy such as a
1885 CD-ROM) you can during specify extra modules to Configure to build and
1886 install with Perl using the -Dextras=... option. See INSTALL for
1891 In addition to config.over, a new override file, config.arch, is
1892 available. This file is supposed to be used by hints file writers
1893 for architecture-wide changes (as opposed to config.over which is
1894 for site-wide changes).
1898 If your file system supports symbolic links, you can build Perl outside
1899 of the source directory by
1901 mkdir /tmp/perl/build/directory
1902 cd /tmp/perl/build/directory
1903 sh /path/to/perl/source/Configure -Dmksymlinks ...
1905 This will create in /tmp/perl/build/directory a tree of symbolic links
1906 pointing to files in /path/to/perl/source. The original files are left
1907 unaffected. After Configure has finished, you can just say
1911 and Perl will be built and tested, all in /tmp/perl/build/directory.
1916 For Perl developers, several new make targets for profiling
1917 and debugging have been added; see L<perlhack>.
1923 Use of the F<gprof> tool to profile Perl has been documented in
1924 L<perlhack>. There is a make target called "perl.gprof" for
1925 generating a gprofiled Perl executable.
1929 If you have GCC 3, there is a make target called "perl.gcov" for
1930 creating a gcoved Perl executable for coverage analysis. See
1935 If you are on IRIX or Tru64 platforms, new profiling/debugging options
1936 have been added; see L<perlhack> for more information about pixie and
1943 Guidelines of how to construct minimal Perl installations have
1944 been added to INSTALL.
1948 The Thread extension is now not built at all under ithreads
1949 (C<Configure -Duseithreads>) because it wouldn't work anyway (the
1950 Thread extension requires being Configured with C<-Duse5005threads>).
1952 B<Note that the 5.005 threads are unsupported and deprecated: if you
1953 have code written for the old threads you should migrate it to the
1954 new ithreads model.>
1958 The Gconvert macro ($Config{d_Gconvert}) used by perl for stringifying
1959 floating-point numbers is now more picky about using sprintf %.*g
1960 rules for the conversion. Some platforms that used to use gcvt may
1961 now resort to the slower sprintf.
1965 The obsolete method of making a special (e.g., debugging) flavor
1968 make LIBPERL=libperld.a
1970 has been removed. Use -DDEBUGGING instead.
1974 =head2 New Or Improved Platforms
1976 For the list of platforms known to support Perl,
1977 see L<perlport/"Supported Platforms">.
1983 AIX dynamic loading should be now better supported.
1987 AIX should now work better with gcc, threads, and 64-bitness. Also the
1988 long doubles support in AIX should be better now. See L<perlaix>.
1992 AtheOS ( http://www.atheos.cx/ ) is a new platform.
1996 BeOS has been reclaimed.
2000 The DG/UX platform now supports 5.005-style threads.
2005 The DYNIX/ptx platform (also known as dynixptx) is supported at or
2010 EBCDIC platforms (z/OS (also known as OS/390), POSIX-BC, and VM/ESA)
2011 have been regained. Many test suite tests still fail and the
2012 co-existence of Unicode and EBCDIC isn't quite settled, but the
2013 situation is much better than with Perl 5.6. See L<perlos390>,
2014 L<perlbs2000> (for POSIX-BC), and L<perlvmesa> for more information.
2018 Building perl with -Duseithreads or -Duse5005threads now works under
2019 HP-UX 10.20 (previously it only worked under 10.30 or later). You will
2020 need a thread library package installed. See README.hpux. [561]
2024 Mac OS Classic is now supported in the mainstream source package
2025 (MacPerl has of course been available since perl 5.004 but now the
2026 source code bases of standard Perl and MacPerl have been synchronised)
2031 Mac OS X (or Darwin) should now be able to build Perl even on HFS+
2032 filesystems. (The case-insensitivity used to confuse the Perl build
2037 NCR MP-RAS is now supported. [561]
2041 All the NetBSD specific patches (except for the installation
2042 specific ones) have been merged back to the main distribution.
2046 NetWare from Novell is now supported. See L<perlnetware>.
2050 NonStop-UX is now supported. [561]
2054 NEC SUPER-UX is now supported.
2058 All the OpenBSD specific patches (except for the installation
2059 specific ones) have been merged back to the main distribution.
2063 Perl has been tested with the GNU pth userlevel thread package
2064 ( http://www.gnu.org/software/pth/pth.html ). All thread tests
2065 of Perl now work, but not without adding some yield()s to the tests,
2066 so while pth (and other userlevel thread implementations) can be
2067 considered to be "working" with Perl ithreads, keep in mind the
2068 possible non-preemptability of the underlying thread implementation.
2072 Stratus VOS is now supported using Perl's native build method
2073 (Configure). This is the recommended method to build Perl on
2074 VOS. The older methods, which build miniperl, are still
2075 available. See L<perlvos>. [561+]
2079 The Amdahl UTS UNIX mainframe platform is now supported. [561]
2083 WinCE is now supported. See L<perlce>.
2087 z/OS (formerly known as OS/390, formerly known as MVS OE) now has
2088 support for dynamic loading. This is not selected by default,
2089 however, you must specify -Dusedl in the arguments of Configure. [561]
2093 =head1 Selected Bug Fixes
2095 Numerous memory leaks and uninitialized memory accesses have been
2096 hunted down. Most importantly, anonymous subs used to leak quite
2103 The autouse pragma didn't work for Multi::Part::Function::Names.
2107 caller() could cause core dumps in certain situations. Carp was
2108 sometimes affected by this problem. In particular, caller() now
2109 returns a subroutine name of C<(unknown)> for subroutines that have
2110 been removed from the symbol table.
2114 chop(@list) in list context returned the characters chopped in
2115 reverse order. This has been reversed to be in the right order. [561]
2119 Configure no longer includes the DBM libraries (dbm, gdbm, db, ndbm)
2120 when building the Perl binary. The only exception to this is SunOS 4.x,
2121 which needs them. [561]
2125 The behaviour of non-decimal but numeric string constants such as
2126 "0x23" was platform-dependent: in some platforms that was seen as 35,
2127 in some as 0, in some as a floating point number (don't ask). This
2128 was caused by Perl's using the operating system libraries in a situation
2129 where the result of the string to number conversion is undefined: now
2130 Perl consistently handles such strings as zero in numeric contexts.
2134 Several debugger fixes: exit code now reflects the script exit code,
2135 condition C<"0"> now treated correctly, the C<d> command now checks
2136 line number, C<$.> no longer gets corrupted, and all debugger output
2137 now goes correctly to the socket if RemotePort is set. [561]
2141 The debugger (perl5db.pl) has been modified to present a more
2142 consistent commands interface, via (CommandSet=580). perl5db.t was
2143 also added to test the changes, and as a placeholder for further tests.
2149 The debugger has a new C<dumpDepth> option to control the maximum
2150 depth to which nested structures are dumped. The C<x> command has
2151 been extended so that C<x N EXPR> dumps out the value of I<EXPR> to a
2152 depth of at most I<N> levels.
2156 The debugger can now show lexical variables if you have the CPAN
2157 module PadWalker installed.
2161 The order of DESTROYs has been made more predictable.
2165 Perl 5.6.0 could emit spurious warnings about redefinition of
2166 dl_error() when statically building extensions into perl.
2167 This has been corrected. [561]
2171 L<dprofpp> -R didn't work.
2175 C<*foo{FORMAT}> now works.
2179 Infinity is now recognized as a number.
2183 UNIVERSAL::isa no longer caches methods incorrectly. (This broke
2184 the Tk extension with 5.6.0.) [561]
2188 Lexicals I: lexicals outside an eval "" weren't resolved
2189 correctly inside a subroutine definition inside the eval "" if they
2190 were not already referenced in the top level of the eval""ed code.
2194 Lexicals II: lexicals leaked at file scope into subroutines that
2195 were declared before the lexicals.
2199 Lexical warnings now propagating correctly between scopes
2200 and into C<eval "...">.
2204 C<use warnings qw(FATAL all)> did not work as intended. This has been
2209 warnings::enabled() now reports the state of $^W correctly if the caller
2210 isn't using lexical warnings. [561]
2214 Line renumbering with eval and C<#line> now works. [561]
2218 Fixed numerous memory leaks, especially in eval "".
2222 Localised tied variables no longer leak memory
2225 tie my %tied_hash => 'Tie::StdHash';
2229 # Used to leak memory every time local() was called;
2230 # in a loop, this added up.
2231 local($tied_hash{Foo}) = 1;
2235 Localised hash elements (and %ENV) are correctly unlocalised to not
2236 exist, if they didn't before they were localised.
2240 tie my %tied_hash => 'Tie::StdHash';
2244 # Nothing has set the FOO element so far
2246 { local $tied_hash{FOO} = 'Bar' }
2248 # This used to print, but not now.
2249 print "exists!\n" if exists $tied_hash{FOO};
2251 As a side effect of this fix, tied hash interfaces B<must> define
2252 the EXISTS and DELETE methods.
2256 mkdir() now ignores trailing slashes in the directory name,
2257 as mandated by POSIX.
2261 Some versions of glibc have a broken modfl(). This affects builds
2262 with C<-Duselongdouble>. This version of Perl detects this brokenness
2263 and has a workaround for it. The glibc release 2.2.2 is known to have
2264 fixed the modfl() bug.
2268 Modulus of unsigned numbers now works (4063328477 % 65535 used to
2269 return 27406, instead of 27047). [561]
2273 Some "not a number" warnings introduced in 5.6.0 eliminated to be
2274 more compatible with 5.005. Infinity is now recognised as a number. [561]
2278 Numeric conversions did not recognize changes in the string value
2279 properly in certain circumstances. [561]
2283 Attributes (such as :shared) didn't work with our().
2287 our() variables will not cause bogus "Variable will not stay shared"
2292 "our" variables of the same name declared in two sibling blocks
2293 resulted in bogus warnings about "redeclaration" of the variables.
2294 The problem has been corrected. [561]
2298 pack "Z" now correctly terminates the string with "\0".
2302 Fix password routines which in some shadow password platforms
2303 (e.g. HP-UX) caused getpwent() to return every other entry.
2307 The PERL5OPT environment variable (for passing command line arguments
2308 to Perl) didn't work for more than a single group of options. [561]
2312 PERL5OPT with embedded spaces didn't work.
2316 printf() no longer resets the numeric locale to "C".
2320 C<qw(a\\b)> now parses correctly as C<'a\\b'>: that is, as three
2321 characters, not four. [561]
2325 pos() did not return the correct value within s///ge in earlier
2326 versions. This is now handled correctly. [561]
2330 Printing quads (64-bit integers) with printf/sprintf now works
2331 without the q L ll prefixes (assuming you are on a quad-capable platform).
2335 Regular expressions on references and overloaded scalars now work. [561+]
2339 Right-hand side magic (GMAGIC) could in many cases such as string
2340 concatenation be invoked too many times.
2344 scalar() now forces scalar context even when used in void context.
2348 SOCKS support is now much more robust.
2352 sort() arguments are now compiled in the right wantarray context
2353 (they were accidentally using the context of the sort() itself).
2354 The comparison block is now run in scalar context, and the arguments
2355 to be sorted are always provided list context. [561]
2359 Changed the POSIX character class C<[[:space:]]> to include the (very
2360 rarely used) vertical tab character. Added a new POSIX-ish character
2361 class C<[[:blank:]]> which stands for horizontal whitespace
2362 (currently, the space and the tab).
2366 The tainting behaviour of sprintf() has been rationalized. It does
2367 not taint the result of floating point formats anymore, making the
2368 behaviour consistent with that of string interpolation. [561]
2372 Some cases of inconsistent taint propagation (such as within hash
2373 values) have been fixed.
2377 The RE engine found in Perl 5.6.0 accidentally pessimised certain kinds
2378 of simple pattern matches. These are now handled better. [561]
2382 Regular expression debug output (whether through C<use re 'debug'>
2383 or via C<-Dr>) now looks better. [561]
2387 Multi-line matches like C<"a\nxb\n" =~ /(?!\A)x/m> were flawed. The
2388 bug has been fixed. [561]
2392 Use of $& could trigger a core dump under some situations. This
2393 is now avoided. [561]
2397 The regular expression captured submatches ($1, $2, ...) are now
2398 more consistently unset if the match fails, instead of leaving false
2399 data lying around in them. [561]
2403 readline() on files opened in "slurp" mode could return an extra
2404 "" (blank line) at the end in certain situations. This has been
2409 Autovivification of symbolic references of special variables described
2410 in L<perlvar> (as in C<${$num}>) was accidentally disabled. This works
2415 Sys::Syslog ignored the C<LOG_AUTH> constant.
2419 $AUTOLOAD, sort(), lock(), and spawning subprocesses
2420 in multiple threads simultaneously are now thread-safe.
2424 Tie::Array's SPLICE method was broken.
2428 Allow a read-only string on the left-hand side of a non-modifying tr///.
2432 If C<STDERR> is tied, warnings caused by C<warn> and C<die> now
2433 correctly pass to it.
2437 Several Unicode fixes.
2443 BOMs (byte order marks) at the beginning of Perl files
2444 (scripts, modules) should now be transparently skipped.
2445 UTF-16 and UCS-2 encoded Perl files should now be read correctly.
2449 The character tables have been updated to Unicode 3.2.0.
2453 Comparing with utf8 data does not magically upgrade non-utf8 data
2454 into utf8. (This was a problem for example if you were mixing data
2455 from I/O and Unicode data: your output might have got magically encoded
2460 Generating illegal Unicode code points such as U+FFFE, or the UTF-16
2461 surrogates, now also generates an optional warning.
2465 C<IsAlnum>, C<IsAlpha>, and C<IsWord> now match titlecase.
2469 Concatenation with the C<.> operator or via variable interpolation,
2470 C<eq>, C<substr>, C<reverse>, C<quotemeta>, the C<x> operator,
2471 substitution with C<s///>, single-quoted UTF8, should now work.
2475 The C<tr///> operator now works. Note that the C<tr///CU>
2476 functionality has been removed (but see pack('U0', ...)).
2480 C<eval "v200"> now works.
2484 Perl 5.6.0 parsed m/\x{ab}/ incorrectly, leading to spurious warnings.
2485 This has been corrected. [561]
2489 Zero entries were missing from the Unicode classes such as C<IsDigit>.
2495 Large unsigned numbers (those above 2**31) could sometimes lose their
2496 unsignedness, causing bogus results in arithmetic operations. [561]
2500 The Perl parser has been stress tested using both random input and
2501 Markov chain input and the few found crashes and lockups have been
2506 =head2 Platform Specific Changes and Fixes
2514 Perl now works on post-4.0 BSD/OSes.
2520 Setting C<$0> now works (as much as possible; see L<perlvar> for details).
2526 Numerous updates; currently synchronised with Cygwin 1.3.10.
2530 Previously DYNIX/ptx had problems in its Configure probe for non-blocking I/O.
2536 EPOC now better supported. See README.epoc. [561]
2542 Perl now works on post-3.0 FreeBSDs.
2548 README.hpux updated; C<Configure -Duse64bitall> now works;
2549 now uses HP-UX malloc instead of Perl malloc.
2555 Numerous compilation flag and hint enhancements; accidental mixing
2556 of 32-bit and 64-bit libraries (a doomed attempt) made much harder.
2566 Long doubles should now work (see INSTALL). [561]
2570 Linux previously had problems related to sockaddrlen when using
2571 accept(), recvfrom() (in Perl: recv()), getpeername(), and
2580 Compilation of the standard Perl distribution in Mac OS Classic should
2581 now work if you have the Metrowerks development environment and the
2582 missing Mac-specific toolkit bits. Contact the macperl mailing list
2589 MPE/iX update after Perl 5.6.0. See README.mpeix. [561]
2593 NetBSD/threads: try installing the GNU pth (should be in the
2594 packages collection, or http://www.gnu.org/software/pth/),
2595 and Configure with -Duseithreads.
2601 Perl now works on NetBSD/sparc.
2607 Now works with usethreads (see INSTALL). [561]
2613 64-bitness using the Sun Workshop compiler now works.
2619 The native build method requires at least VOS Release 14.5.0
2620 and GNU C++/GNU Tools 2.0.1 or later. The Perl pack function
2621 now maps overflowed values to +infinity and underflowed values
2626 Tru64 (aka Digital UNIX, aka DEC OSF/1)
2628 The operating system version letter now recorded in $Config{osvers}.
2629 Allow compiling with gcc (previously explicitly forbidden). Compiling
2630 with gcc still not recommended because buggy code results, even with
2637 Fixed various alignment problems that lead into core dumps either
2638 during build or later; no longer dies on math errors at runtime;
2639 now using full quad integers (64 bits), previously was using
2640 only 46 bit integers for speed.
2646 See L</"Socket Extension Dynamic in VMS"> and L</"IEEE-format Floating Point
2647 Default on OpenVMS Alpha"> for important changes not otherwise listed here.
2649 chdir() now works better despite a CRT bug; now works with MULTIPLICITY
2650 (see INSTALL); now works with Perl's malloc.
2652 The tainting of C<%ENV> elements via C<keys> or C<values> was previously
2653 unimplemented. It now works as documented.
2655 The C<waitpid> emulation has been improved. The worst bug (now fixed)
2656 was that a pid of -1 would cause a wildcard search of all processes on
2659 POSIX-style signals are now emulated much better on VMS versions prior
2662 The C<system> function and backticks operator have improved
2663 functionality and better error handling. [561]
2665 File access tests now use current process privileges rather than the
2666 user's default privileges, which could sometimes result in a mismatch
2667 between reported access and actual access. This improvement is only
2668 available on VMS v6.0 and later.
2670 There is a new C<kill> implementation based on C<sys$sigprc> that allows
2671 older VMS systems (pre-7.0) to use C<kill> to send signals rather than
2672 simply force exit. This implementation also allows later systems to
2673 call C<kill> from within a signal handler.
2675 Iterative logical name translations are now limited to 10 iterations in
2676 imitation of SHOW LOGICAL and other OpenVMS facilities.
2686 Signal handling now works better than it used to. It is now implemented
2687 using a Windows message loop, and is therefore less prone to random
2692 fork() emulation is now more robust, but still continues to have a few
2693 esoteric bugs and caveats. See L<perlfork> for details. [561+]
2697 A failed (pseudo)fork now returns undef and sets errno to EAGAIN. [561]
2701 The following modules now work on Windows:
2703 ExtUtils::Embed [561]
2710 IO::File::new_tmpfile() is no longer limited to 32767 invocations
2715 Better chdir() return value for a non-existent directory.
2719 Compiling perl using the 64-bit Platform SDK tools is now supported.
2723 The Win32::SetChildShowWindow() builtin can be used to control the
2724 visibility of windows created by child processes. See L<Win32> for
2729 Non-blocking waits for child processes (or pseudo-processes) are
2730 supported via C<waitpid($pid, &POSIX::WNOHANG)>.
2734 The behavior of system() with multiple arguments has been rationalized.
2735 Each unquoted argument will be automatically quoted to protect whitespace,
2736 and any existing whitespace in the arguments will be preserved. This
2737 improves the portability of system(@args) by avoiding the need for
2738 Windows C<cmd> shell specific quoting in perl programs.
2740 Note that this means that some scripts that may have relied on earlier
2741 buggy behavior may no longer work correctly. For example,
2742 C<system("nmake /nologo", @args)> will now attempt to run the file
2743 C<nmake /nologo> and will fail when such a file isn't found.
2744 On the other hand, perl will now execute code such as
2745 C<system("c:/Program Files/MyApp/foo.exe", @args)> correctly.
2749 The perl header files no longer suppress common warnings from the
2750 Microsoft Visual C++ compiler. This means that additional warnings may
2751 now show up when compiling XS code.
2755 Borland C++ v5.5 is now a supported compiler that can build Perl.
2756 However, the generated binaries continue to be incompatible with those
2757 generated by the other supported compilers (GCC and Visual C++). [561]
2761 Duping socket handles with open(F, ">&MYSOCK") now works under Windows 9x.
2766 Current directory entries in %ENV are now correctly propagated to child
2771 New %ENV entries now propagate to subprocesses. [561]
2775 Win32::GetCwd() correctly returns C:\ instead of C: when at the drive root.
2776 Other bugs in chdir() and Cwd::cwd() have also been fixed. [561]
2780 The makefiles now default to the features enabled in ActiveState ActivePerl
2781 (a popular Win32 binary distribution). [561]
2785 HTML files will now be installed in c:\perl\html instead of
2786 c:\perl\lib\pod\html
2790 REG_EXPAND_SZ keys are now allowed in registry settings used by perl. [561]
2794 Can now send() from all threads, not just the first one. [561]
2798 ExtUtils::MakeMaker now uses $ENV{LIB} to search for libraries. [561]
2802 Less stack reserved per thread so that more threads can run
2803 concurrently. (Still 16M per thread.) [561]
2807 C<< File::Spec->tmpdir() >> now prefers C:/temp over /tmp
2808 (works better when perl is running as service).
2812 Better UNC path handling under ithreads. [561]
2816 wait(), waitpid(), and backticks now return the correct exit status
2817 under Windows 9x. [561]
2821 A socket handle leak in accept() has been fixed. [561]
2827 =head1 New or Changed Diagnostics
2829 Please see L<perldiag> for more details.
2835 Ambiguous range in the transliteration operator (like a-z-9) now
2840 Two new debugging options have been added: if you have compiled your
2841 Perl with debugging, you can use the -DT [561] and -DR options to trace
2842 tokenising and to add reference counts to displaying variables,
2847 The lexical warnings category "deprecated" is no longer a sub-category
2848 of the "syntax" category. It is now a top-level category in its own
2853 Unadorned dump() will now give a warning suggesting to
2854 use explicit CORE::dump() if that's what really is meant.
2858 The "Unrecognized escape" warning has been extended to include C<\8>,
2859 C<\9>, and C<\_>. There is no need to escape any of the C<\w> characters.
2863 All regular expression compilation error messages are now hopefully
2864 easier to understand both because the error message now comes before
2865 the failed regex and because the point of failure is now clearly
2866 marked by a C<E<lt>-- HERE> marker.
2870 Various I/O (and socket) functions like binmode(), close(), and so
2871 forth now more consistently warn if they are used illogically either
2872 on a yet unopened or on an already closed filehandle (or socket).
2876 Using lstat() on a filehandle now gives a warning. (It's a non-sensical
2881 The C<-M> and C<-m> options now warn if you didn't supply the module name.
2885 If you in C<use> specify a required minimum version, modules matching
2886 the name and but not defining a $VERSION will cause a fatal failure.
2890 Using negative offset for vec() in lvalue context is now a warnable offense.
2894 Odd number of arguments to oveload::constant now elicits a warning.
2898 Odd number of elements to in anonymous hash now elicits a warning.
2902 The various "opened only for", "on closed", "never opened" warnings
2903 drop the C<main::> prefix for filehandles in the C<main> package,
2904 for example C<STDIN> instead of C<main::STDIN>.
2908 Subroutine prototypes are now checked more carefully, you may
2909 get warnings for example if you have used non-prototype characters.
2913 If an attempt to use a (non-blessed) reference as an array index
2914 is made, a warning is given.
2918 C<push @a;> and C<unshift @a;> (with no values to push or unshift)
2919 now give a warning. This may be a problem for generated and evaled
2924 If you try to L<perlfunc/pack> a number less than 0 or larger than 255
2925 using the C<"C"> format you will get an optional warning. Similarly
2926 for the C<"c"> format and a number less than -128 or more than 127.
2930 pack C<P> format now demands an explicit size.
2934 unpack C<w> now warns of unterminated compressed integers.
2938 Warnings relating to the use of PerlIO have been added.
2942 Certain regex modifiers such as C<(?o)> make sense only if applied to
2943 the entire regex. You will get an optional warning if you try to do
2948 Variable length lookbehind has not yet been implemented, trying to
2949 use it will tell that.
2953 Using arrays or hashes as references (e.g. C<< %foo->{bar} >>
2954 has been deprecated for a while. Now you will get an optional warning.
2958 Warnings relating to the use of the new restricted hashes feature
2963 Self-ties of arrays and hashes are not supported and fatal errors
2964 will happen even at an attempt to do so.
2968 Using C<sort> in scalar context now issues an optional warning.
2969 This didn't do anything useful, as the sort was not performed.
2973 Using the /g modifier in split() is meaningless and will cause a warning.
2977 Using splice() past the end of an array now causes a warning.
2981 Malformed Unicode encodings (UTF-8 and UTF-16) cause a lot of warnings,
2982 ad doestrying to use UTF-16 surrogates (which are unimplemented).
2986 Trying to use Unicode characters on an I/O stream without marking the
2987 stream's encoding (using open() or binmode()) will cause "Wide character"
2992 Use of v-strings in use/require causes a (backward) portability warning.
2996 Warnings relating to the use interpreter threads and their shared data
3001 =head1 Changed Internals
3007 PerlIO is now the default.
3011 perlapi.pod (a companion to perlguts) now attempts to document the
3016 You can now build a really minimal perl called microperl.
3017 Building microperl does not require even running Configure;
3018 C<make -f Makefile.micro> should be enough. Beware: microperl makes
3019 many assumptions, some of which may be too bold; the resulting
3020 executable may crash or otherwise misbehave in wondrous ways.
3021 For careful hackers only.
3025 Added rsignal(), whichsig(), do_join(), op_clear, op_null,
3026 ptr_table_clear(), ptr_table_free(), sv_setref_uv(), and several UTF-8
3027 interfaces to the publicised API. For the full list of the available
3028 APIs see L<perlapi>.
3032 Made possible to propagate customised exceptions via croak()ing.
3036 Now xsubs can have attributes just like subs. (Well, at least the
3037 built-in attributes.)
3041 dTHR and djSP have been obsoleted; the former removed (because it's
3042 a no-op) and the latter replaced with dSP.
3046 PERL_OBJECT has been completely removed.
3050 The MAGIC constants (e.g. C<'P'>) have been macrofied
3051 (e.g. C<PERL_MAGIC_TIED>) for better source code readability
3052 and maintainability.
3056 The regex compiler now maintains a structure that identifies nodes in
3057 the compiled bytecode with the corresponding syntactic features of the
3058 original regex expression. The information is attached to the new
3059 C<offsets> member of the C<struct regexp>. See L<perldebguts> for more
3060 complete information.
3064 The C code has been made much more C<gcc -Wall> clean. Some warning
3065 messages still remain in some platforms, so if you are compiling with
3066 gcc you may see some warnings about dubious practices. The warnings
3067 are being worked on.
3071 F<perly.c>, F<sv.c>, and F<sv.h> have now been extensively commented.
3075 Documentation on how to use the Perl source repository has been added
3076 to F<Porting/repository.pod>.
3080 There are now several profiling make targets.
3084 =head1 Security Vulnerability Closed [561]
3086 (This change was already made in 5.7.0 but bears repeating here.)
3087 (5.7.0 came out before 5.6.1: the development branch 5.7 released
3088 earlier than the maintenance branch 5.6)
3090 A potential security vulnerability in the optional suidperl component
3091 of Perl was identified in August 2000. suidperl is neither built nor
3092 installed by default. As of November 2001 the only known vulnerable
3093 platform is Linux, most likely all Linux distributions. CERT and
3094 various vendors and distributors have been alerted about the vulnerability.
3095 See http://www.cpan.org/src/5.0/sperl-2000-08-05/sperl-2000-08-05.txt
3096 for more information.
3098 The problem was caused by Perl trying to report a suspected security
3099 exploit attempt using an external program, /bin/mail. On Linux
3100 platforms the /bin/mail program had an undocumented feature which
3101 when combined with suidperl gave access to a root shell, resulting in
3102 a serious compromise instead of reporting the exploit attempt. If you
3103 don't have /bin/mail, or if you have 'safe setuid scripts', or if
3104 suidperl is not installed, you are safe.
3106 The exploit attempt reporting feature has been completely removed from
3107 Perl 5.8.0 (and the maintenance release 5.6.1, and it was removed also
3108 from all the Perl 5.7 releases), so that particular vulnerability
3109 isn't there anymore. However, further security vulnerabilities are,
3110 unfortunately, always possible. The suidperl functionality is most
3111 probably going to be removed in Perl 5.10. In any case, suidperl
3112 should only be used by security experts who know exactly what they are
3113 doing and why they are using suidperl instead of some other solution
3114 such as sudo ( see http://www.courtesan.com/sudo/ ).
3118 Several new tests have been added, especially for the F<lib> and
3119 F<ext> subsections. There are now about 69 000 individual tests
3120 (spread over about 700 test scripts), in the regression suite (5.6.1
3121 has about 11 700 tests, in 258 test scripts) The exact numbers depend
3122 on the platform and Perl configuration used. Many of the new tests
3123 are of course introduced by the new modules, but still in general Perl
3124 is now more thoroughly tested.
3126 Because of the large number of tests, running the regression suite
3127 will take considerably longer time than it used to: expect the suite
3128 to take up to 4-5 times longer to run than in perl 5.6. On a really
3129 fast machine you can hope to finish the suite in about 6-8 minutes
3132 The tests are now reported in a different order than in earlier Perls.
3133 (This happens because the test scripts from under t/lib have been moved
3134 to be closer to the library/extension they are testing.)
3136 =head1 Known Problems
3144 If using the AIX native make command, instead of just "make" issue
3145 "make all". In some setups the former has been known to spuriously
3146 also try to run "make install". Alternatively, you may want to use
3151 In AIX 4.2, Perl extensions that use C++ functions that use statics
3152 may have problems in that the statics are not getting initialized.
3153 In newer AIX releases, this has been solved by linking Perl with
3154 the libC_r library, but unfortunately in AIX 4.2 the said library
3155 has an obscure bug where the various functions related to time
3156 (such as time() and gettimeofday()) return broken values, and
3157 therefore in AIX 4.2 Perl is not linked against libC_r.
3161 vac 5.0.0.0 May Produce Buggy Code For Perl
3163 The AIX C compiler vac version 5.0.0.0 may produce buggy code,
3164 resulting in a few random tests failing when run as part of "make
3165 test", but when the failing tests are run by hand, they succeed.
3166 We suggest upgrading to at least vac version 5.0.1.0, that has been
3167 known to compile Perl correctly. "lslpp -L|grep vac.C" will tell
3168 you the vac version. See README.aix.
3172 If building threaded Perl, you may get compilation warning from pp_sys.c:
3174 "pp_sys.c", line 4651.39: 1506-280 (W) Function argument assignment between types "unsigned char*" and "const void*" is not allowed.
3176 This is harmless; it is caused by the getnetbyaddr() and getnetbyaddr_r()
3177 having slightly different types for their first argument.
3181 =head2 Alpha systems with old gccs fail several tests
3183 If you see op/pack, op/pat, op/regexp, or ext/Storable tests failing
3184 in a Linux/alpha or *BSD/Alpha, it's probably time to upgrade your gcc.
3185 gccs prior to 2.95.3 are definitely not good enough, and gcc 3.1 may
3186 be even better. (RedHat Linux/alpha with gcc 3.1 reported no problems,
3187 as did Linux 2.4.18 with gcc 2.95.4.) (In Tru64, it is preferable to
3188 use the bundled C compiler.)
3192 Perl 5.8.0 doesn't build in AmigaOS. It broke at some point during
3193 the ithreads work and we could not find Amiga experts to unbreak the
3194 problems. Perl 5.6.1 still works for AmigaOS (as does the the 5.7.2
3195 development release).
3199 The following tests fail on 5.8.0 Perl in BeOS Personal 5.03:
3201 t/op/lfs............................FAILED at test 17
3202 t/op/magic..........................FAILED at test 24
3203 ext/Fcntl/t/syslfs..................FAILED at test 17
3204 ext/File/Glob/t/basic...............FAILED at test 3
3205 ext/POSIX/t/sigaction...............FAILED at test 13
3206 ext/POSIX/t/waitpid.................FAILED at test 1
3208 See L<perlbeos> (README.beos) for more details.
3210 =head2 Cygwin "unable to remap"
3212 For example when building the Tk extension for Cygwin,
3213 you may get an error message saying "unable to remap".
3214 This is known problem with Cygwin, and a workaround is
3215 detailed in here: http://sources.redhat.com/ml/cygwin/2001-12/msg00894.html
3217 =head2 Cygwin ndbm tests fail on FAT
3219 One can build but not install (or test the build of) the NDBM_File
3220 on FAT filesystems. Installation (or build) on NTFS works fine.
3222 =head2 DJGPP does not build
3224 Unfortunately DJGPP build broke somewhere after 5.7.1.
3226 =head2 ext/threads/t/libc
3228 If this test fails, it indicates that your libc (C library) is not
3229 threadsafe. This particular test stress tests the localtime() call to
3230 find out whether it is threadsafe. See L<perlthrtut> for more information.
3232 =head2 FreeBSD built with ithreads coredumps reading large directories
3234 This is a known bug in FreeBSD 4.5's readdir_r(), it has been fixed in
3235 FreeBSD 4.6 (see L<perlfreebsd> (README.freebsd)).
3237 =head2 FreeBSD Failing locale Test 117 For ISO 8859-15 Locales
3239 The ISO 8859-15 locales may fail the locale test 117 in FreeBSD.
3240 This is caused by the characters \xFF (y with diaeresis) and \xBE
3241 (Y with diaeresis) not behaving correctly when being matched
3242 case-insensitively. Apparently this problem has been fixed in
3243 the latest FreeBSD releases.
3244 ( http://www.freebsd.org/cgi/query-pr.cgi?pr=34308 )
3246 =head2 IRIX fails ext/List/Util/t/shuffle.t
3248 IRIX with MIPSpro 7.3.1.3m compiler may fail the said List::Util test
3249 by dumping core. This seems to be a compiler error since if compiled
3250 with gcc no core dump ensues, and no failures on the said test on any
3253 =head2 Modifying $_ Inside for(..)
3257 works without complaint. It shouldn't. (You should be able to
3258 modify only lvalue elements inside the loops.) You can see the
3259 correct behaviour by replacing the 1..5 with 1, 2, 3, 4, 5.
3261 =head2 mod_perl 1.26 Doesn't Build With Threaded Perl
3263 Use mod_perl 1.27 or higher.
3265 =head2 lib/ftmp-security tests warn 'system possibly insecure'
3267 Don't panic. Read the 'make test' section of INSTALL instead.
3269 =head2 HP-UX lib/posix Subtest 9 Fails When LP64-Configured
3271 If perl is configured with -Duse64bitall, the successful result of the
3272 subtest 10 of lib/posix may arrive before the successful result of the
3273 subtest 9, which confuses the test harness so much that it thinks the
3276 =head2 Linux with glibc 2.2.5 fails t/op/int subtest #6 with -Duse64bitint
3278 This is a known bug in the glibc 2.2.5 with long long integers.
3279 ( http://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=65612 )
3281 =head2 Linux With Sfio Fails op/misc Test 48
3285 =head2 libwww-perl (LWP) fails base/date #51
3287 Use libwww-perl 5.65 or later.
3291 Please remember to set your environment variable LC_ALL to "C"
3292 (setenv LC_ALL C) before running "make test" to avoid a lot of
3293 warnings about the broken locales of Mac OS X.
3295 The following tests are known to fail in Mac OS X 10.1.5 because of
3296 buggy (old) implementations of Berkeley DB included in Mac OS X:
3298 Failed Test Stat Wstat Total Fail Failed List of Failed
3299 -------------------------------------------------------------------------
3300 ../ext/DB_File/t/db-btree.t 0 11 ?? ?? % ??
3301 ../ext/DB_File/t/db-recno.t 149 3 2.01% 61 63 65
3303 If you are building on a UFS partition, you will also probably see
3304 t/op/stat.t subtest #9 fail. This is caused by Darwin's UFS not
3305 supporting inode change time.
3307 Also the ext/POSIX/t/posix.t subtest #10 fails but it is skipped for
3308 now because the failure is Apple's fault, not Perl's (blocked signals
3311 If you Configure with ithreads, ext/threads/t/libc.t will fail. Again,
3312 this is not Perl's fault-- the libc of Mac OS X is not threadsafe
3313 (in this particular test, the localtime() call is found to be
3316 =head2 OS/2 Test Failures
3318 The following tests are known to fail on OS/2 (for clarity
3319 only the failures are shown, not the full error messages):
3321 t/io/utf8............................FAILED at test 19
3322 t/op/grent...........................FAILED at test 2
3323 t/op/pwent...........................FAILED at test 1
3324 t/lib/os2_base.......................FAILED at test 13
3325 t/lib/os2_process....................FAILED at test 10
3326 t/lib/os2_process_kid................FAILED at test 10
3327 t/lib/rx_cmprt.......................FAILED at test 16
3328 ext/DB_File/t/db-btree...............FAILED at test 0
3329 ext/DB_File/t/db-hash................FAILED at test 0
3330 ext/DB_File/t/db-recno...............FAILED at test 0
3331 lib/ExtUtils/t/basic.................FAILED at test 14
3332 lib/ExtUtils/t/Constant..............FAILED at test 4
3333 lib/Memoize/t/errors.................FAILED at test 4
3335 =head2 op/sprintf tests 91, 129, and 130
3337 The op/sprintf tests 91, 129, and 130 are known to fail on some platforms.
3338 Examples include any platform using sfio, and Compaq/Tandem's NonStop-UX.
3340 Test 91 is known to fail on QNX6 (nto), because C<sprintf '%e',0>
3341 incorrectly produces C<0.000000e+0> instead of C<0.000000e+00>.
3343 For tests 129 and 130, the failing platforms do not comply with
3344 the ANSI C Standard: lines 19ff on page 134 of ANSI X3.159 1989, to
3345 be exact. (They produce something other than "1" and "-1" when
3346 formatting 0.6 and -0.6 using the printf format "%.0f"; most often,
3347 they produce "0" and "-0".)
3351 In case you are still using Solaris 2.5 (aka SunOS 5.5), you may
3352 experience failures (the test core dumping) in lib/locale.t.
3353 The suggested cure is to upgrade your Solaris.
3355 =head2 Solaris x86 Fails Tests With -Duse64bitint
3357 The following tests are known to fail in Solaris x86 with Perl
3358 configured to use 64 bit integers:
3360 ext/Data/Dumper/t/dumper.............FAILED at test 268
3361 ext/Devel/Peek/Peek..................FAILED at test 7
3363 =head2 SUPER-UX (NEC SX)
3365 The following tests are known to fail on SUPER-UX:
3367 op/64bitint...........................FAILED tests 29-30, 32-33, 35-36
3368 op/arith..............................FAILED tests 128-130
3369 op/pack...............................FAILED tests 25-5625
3370 op/pow................................
3371 op/taint..............................# msgsnd failed
3372 ../ext/IO/lib/IO/t/io_poll............FAILED tests 3-4
3373 ../ext/IPC/SysV/ipcsysv...............FAILED tests 2, 5-6
3374 ../ext/IPC/SysV/t/msg.................FAILED tests 2, 4-6
3375 ../ext/Socket/socketpair..............FAILED tests 12
3376 ../lib/IPC/SysV.......................FAILED tests 2, 5-6
3377 ../lib/warnings.......................FAILED tests 115-116, 118-119
3379 The op/pack failure ("Cannot compress negative numbers at op/pack.t line 126")
3380 is serious but as of yet unsolved. It points at some problems with the
3381 signedness handling of the C compiler, as do the 64bitint, arith, and pow
3382 failures. Most of the rest point at problems with SysV IPC.
3384 =head2 PDL failing some tests
3386 Use PDL 2.3.4 or later.
3390 You may get errors like 'Undefined symbol "Perl_get_sv"' or
3391 "can't resolve symbol 'Perl_get_sv'". This probably means that
3392 you are trying to use an older shared Perl library with Perl 5.8.0
3393 executable. Perl used to have such a subroutine, but that is no more
3394 the case. Check your shared library path, and any shared Perl
3395 libraries in those directories.
3397 =head2 Term::ReadKey not working on Win32
3399 Use Term::ReadKey 2.20 or later.
3401 =head2 Failure of Thread (5.005-style) tests
3403 B<Note that support for 5.005-style threading is deprecated,
3404 experimental and practically unsupported. In 5.10, it is expected
3405 to be removed. You should migrate your code to ithreads.>
3407 The following tests are known to fail due to fundamental problems in
3408 the 5.005 threading implementation. These are not new failures--Perl
3409 5.005_0x has the same bugs, but didn't have these tests.
3411 ../ext/B/t/xref.t 255 65280 14 12 85.71% 3-14
3412 ../ext/List/Util/t/first.t 255 65280 7 4 57.14% 2 5-7
3413 ../lib/English.t 2 512 54 2 3.70% 2-3
3414 ../lib/FileCache.t 5 1 20.00% 5
3415 ../lib/Filter/Simple/t/data.t 6 3 50.00% 1-3
3416 ../lib/Filter/Simple/t/filter_only. 9 3 33.33% 1-2 5
3417 ../lib/Math/BigInt/t/bare_mbf.t 1627 4 0.25% 8 11 1626-1627
3418 ../lib/Math/BigInt/t/bigfltpm.t 1629 4 0.25% 10 13 1628-
3420 ../lib/Math/BigInt/t/sub_mbf.t 1633 4 0.24% 8 11 1632-1633
3421 ../lib/Math/BigInt/t/with_sub.t 1628 4 0.25% 9 12 1627-1628
3422 ../lib/Tie/File/t/31_autodefer.t 255 65280 65 32 49.23% 34-65
3423 ../lib/autouse.t 10 1 10.00% 4
3424 op/flip.t 15 1 6.67% 15
3426 These failures are unlikely to get fixed as 5.005-style threads
3427 are considered fundamentally broken. (Basically what happens is that
3428 competing threads can corrupt shared global state, one good example
3429 being regular expression engine's state.)
3431 =head2 Timing problems
3433 The following tests may fail intermittently because of timing
3434 problems, for example if the system is heavily loaded.
3437 ext/Time/HiRes/HiRes.t
3439 lib/Memoize/t/expmod_t.t
3440 lib/Memoize/t/speed.t
3442 In case of failure please try running them manually, for example
3444 ./perl -Ilib ext/Time/HiRes/HiRes.t
3446 =head2 Unicode in package/class and subroutine names does not work
3448 One can have Unicode in identifier names, but not in package/class or
3449 subroutine names. While some limited functionality towards this does
3450 exist as of Perl 5.8.0, that is more accidental than designed; use of
3451 Unicode for the said purposes is unsupported.
3453 One reason of this unfinishedness is its (currently) inherent
3454 unportability: since both package names and subroutine names may
3455 need to be mapped to file and directory names, the Unicode capability
3456 of the filesystem becomes important-- and there unfortunately aren't
3465 During Configure, the test
3467 Guessing which symbols your C compiler and preprocessor define...
3469 will probably fail with error messages like
3471 CC-20 cc: ERROR File = try.c, Line = 3
3472 The identifier "bad" is undefined.
3474 bad switch yylook 79bad switch yylook 79bad switch yylook 79bad switch yylook 79#ifdef A29K
3477 CC-65 cc: ERROR File = try.c, Line = 3
3478 A semicolon is expected at this point.
3480 This is caused by a bug in the awk utility of UNICOS/mk. You can ignore
3481 the error, but it does cause a slight problem: you cannot fully
3482 benefit from the h2ph utility (see L<h2ph>) that can be used to
3483 convert C headers to Perl libraries, mainly used to be able to access
3484 from Perl the constants defined using C preprocessor, cpp. Because of
3485 the above error, parts of the converted headers will be invisible.
3486 Luckily, these days the need for h2ph is rare.
3490 If building Perl with interpreter threads (ithreads), the
3491 getgrent(), getgrnam(), and getgrgid() functions cannot return the
3492 list of the group members due to a bug in the multithreaded support of
3493 UNICOS/mk. What this means is that in list context the functions will
3494 return only three values, not four.
3500 There are a few known test failures, see L<perluts> (README.uts).
3502 =head2 VOS (Stratus)
3504 When Perl is built using the native build process on VOS Release
3505 14.5.0 and GNU C++/GNU Tools 2.0.1, all attempted tests either
3506 pass or result in TODO (ignored) failures.
3510 There should be no reported test failures with a default configuration,
3511 though there are a number of tests marked TODO that point to areas
3512 needing further debugging and/or porting work.
3516 In multi-CPU boxes, there are some problems with the I/O buffering:
3517 some output may appear twice.
3519 =head2 XML::Parser not working
3521 Use XML::Parser 2.31 or later.
3523 =head2 z/OS (OS/390)
3525 z/OS has rather many test failures but the situation is actually much
3526 better than it was in 5.6.0; it's just that so many new modules and
3527 tests have been added.
3529 Failed Test Stat Wstat Total Fail Failed List of Failed
3530 ---------------------------------------------------------------------------
3531 ../ext/Data/Dumper/t/dumper.t 357 8 2.24% 311 314 325 327
3533 ../ext/IO/lib/IO/t/io_unix.t 5 4 80.00% 2-5
3534 ../ext/Storable/t/downgrade.t 12 3072 169 12 7.10% 14-15 46-47 78-79
3536 ../lib/ExtUtils/t/Constant.t 121 30976 48 48 100.00% 1-48
3537 ../lib/ExtUtils/t/Embed.t 9 9 100.00% 1-9
3538 op/pat.t 922 7 0.76% 665 776 785 832-
3540 op/sprintf.t 224 3 1.34% 98 100 136
3541 op/tr.t 97 5 5.15% 63 71-74
3542 uni/fold.t 780 6 0.77% 61 169 196 661
3545 The failures in dumper.t and downgrade.t are problems in the tests,
3546 those in io_unix and sprintf are problems in the USS (UDP sockets and
3547 printf formats). The pat, tr, and fold failures are genuine Perl
3548 problems caused by EBCDIC (and in the pat and fold cases, combining
3549 that with Unicode). The Constant and Embed are probably problems in
3550 the tests (since they test Perl's ability to build extensions, and
3551 that seems to be working reasonably well.)
3553 =head2 Localising Tied Arrays and Hashes Is Broken
3557 doesn't work as one would expect: the old value is restored
3558 incorrectly. This will be changed in a future release, but we don't
3559 know yet what the new semantics will exactly be. In any case, the
3560 change will break existing code that relies on the current
3561 (ill-defined) semantics, so just avoid doing this in general.
3563 =head2 Self-tying Problems
3565 Self-tying of arrays and hashes is broken in rather deep and
3566 hard-to-fix ways. As a stop-gap measure to avoid people from getting
3567 frustrated at the mysterious results (core dumps, most often), it is
3568 forbidden for now (you will get a fatal error even from an attempt).
3570 A change to self-tying of globs has caused them to be recursively
3571 referenced (see: L<perlobj/"Two-Phased Garbage Collection">). You
3572 will now need an explicit untie to destroy a self-tied glob. This
3573 behaviour may be fixed at a later date.
3575 Self-tying of scalars and IO thingies works.
3577 =head2 Tied/Magical Array/Hash Elements Do Not Autovivify
3579 For normal arrays C<$foo = \$bar[1]> will assign C<undef> to
3580 C<$bar[1]> (assuming that it didn't exist before), but for
3581 tied/magical arrays and hashes such autovivification does not happen
3582 because there is currently no way to catch the reference creation.
3583 The same problem affects slicing over non-existent indices/keys of
3584 a tied/magical array/hash.
3586 =head2 Building Extensions Can Fail Because Of Largefiles
3588 Some extensions like mod_perl are known to have issues with
3589 `largefiles', a change brought by Perl 5.6.0 in which file offsets
3590 default to 64 bits wide, where supported. Modules may fail to compile
3591 at all, or they may compile and work incorrectly. Currently, there
3592 is no good solution for the problem, but Configure now provides
3593 appropriate non-largefile ccflags, ldflags, libswanted, and libs
3594 in the %Config hash (e.g., $Config{ccflags_nolargefiles}) so the
3595 extensions that are having problems can try configuring themselves
3596 without the largefileness. This is admittedly not a clean solution,
3597 and the solution may not even work at all. One potential failure is
3598 whether one can (or, if one can, whether it's a good idea to) link
3599 together at all binaries with different ideas about file offsets;
3600 all this is platform-dependent.
3602 =head2 Unicode Support on EBCDIC Still Spotty
3604 Though mostly working, Unicode support still has problem spots on
3605 EBCDIC platforms. One such known spot are the C<\p{}> and C<\P{}>
3606 regular expression constructs for code points less than 256: the
3607 C<pP> are testing for Unicode code points, not knowing about EBCDIC.
3609 =head2 The Compiler Suite Is Still Very Experimental
3611 The compiler suite is slowly getting better but it continues to be
3612 highly experimental. Use in production environments is discouraged.
3614 =head2 Seen In Perl 5.7 But Gone Now
3616 C<Time::Piece> (previously known as C<Time::Object>) was removed
3617 because it was felt that it didn't have enough value in it to be a
3618 core module. It is still a useful module, though, and is available
3621 Perl 5.8 unfortunately does not build anymore on AmigaOS; this broke
3622 accidentally at some point. Since there are not that many Amiga
3623 developers available, we could not get this fixed and tested in time
3624 for 5.8.0. Perl 5.6.1 still works for AmigaOS (as does the the 5.7.2
3625 development release).
3627 The C<PerlIO::Scalar> and C<PerlIO::Via> (capitalised) were renamed as
3628 C<PerlIO::scalar> and C<PerlIO::via> (all lowercase) just before 5.8.0.
3629 The main rationale was to have all core PerlIO layers to have all
3630 lowercase names. The "plugins" are named as usual, for example
3631 C<PerlIO::via::QuotedPrint>.
3633 The C<threads::shared::queue> and C<threads::shared::semaphore> were
3634 renamed as C<Thread::Queue> and C<Thread::Semaphore> just before 5.8.0.
3635 The main rationale was to have thread modules to obey normal naming,
3636 C<Thread::> (the C<threads> and C<threads::shared> themselves are
3637 more pragma-like, they affect compile-time, so they stay lowercase).
3639 =head1 Reporting Bugs
3641 If you find what you think is a bug, you might check the articles
3642 recently posted to the comp.lang.perl.misc newsgroup and the perl
3643 bug database at http://bugs.perl.org/ . There may also be
3644 information at http://www.perl.com/ , the Perl Home Page.
3646 If you believe you have an unreported bug, please run the B<perlbug>
3647 program included with your release. Be sure to trim your bug down
3648 to a tiny but sufficient test case. Your bug report, along with the
3649 output of C<perl -V>, will be sent off to perlbug@perl.org to be
3650 analysed by the Perl porting team.
3654 The F<Changes> file for exhaustive details on what changed.
3656 The F<INSTALL> file for how to build Perl.
3658 The F<README> file for general stuff.
3660 The F<Artistic> and F<Copying> files for copyright information.
3664 Written by Jarkko Hietaniemi <F<jhi@iki.fi>>.