3 perldelta - what is new for perl v5.8.0
7 This document describes differences between the 5.6.0 release
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
14 If you are upgrading from Perl 5.005_03, you might also want
15 to read L<perl56delta>.
17 =head1 Highlights In 5.8.0
23 Better Unicode support
27 New Thread Implementation
35 Better Numeric Accuracy
43 More Extensive Regression Testing
47 =head1 Incompatible Changes
49 =head2 Binary Incompatibility
51 Perl 5.8 has not been designed to be binary compatible with earlier
52 releases of Perl. While the compatibility has not been intentionally
53 broken, it has not been intentionally protected, either. The major
54 reason for the discontinity is the new IO architecture called PerlIO.
55 The PerlIO is the default configuration because without it many new
56 features of Perl 5.8 cannot be used. In other words: you just have
57 to recompile your modules, sorry about that.
59 In future releases of Perl non-PerlIO aware XS modules may become
60 completely unsupported. This shouldn't be too difficult for module
61 authors, however: PerlIO has been designed as a drop-in replacement
62 (at the source code level) for the stdio interface.
64 =head2 64-bit platforms and malloc
66 If your pointers are 64 bits wide, the Perl malloc is no longer being
67 used because it does not work well with 8-byte pointers. Also,
68 usually the system mallocs on such platforms are much better optimized
69 for such large memory models than the Perl malloc. Some memory-hungry
70 Perl applications like the PDL don't work well with Perl's malloc.
71 Finally, other applications than Perl (like modperl) tend to prefer
72 the system malloc. Such platforms include Alpha and 64-bit HPPA,
75 =head2 AIX Dynaloading
77 The AIX dynaloading now uses in AIX releases 4.3 and newer the native
78 dlopen interface of AIX instead of the old emulated interface. This
79 change will probably break backward compatibility with compiled
80 modules. The change was made to make Perl more compliant with other
81 applications like modperl which are using the AIX native interface.
83 =head2 Attributes for C<my> variables now handled at run-time.
85 The C<my EXPR : ATTRS> syntax now applies variable attributes at
86 run-time. (Subroutine and C<our> variables still get attributes applied
87 at compile-time.) See L<attributes> for additional details. In particular,
88 however, this allows variable attributes to be useful for C<tie> interfaces,
89 which was a deficiency of earlier releases. Note that the new semantics
90 doesn't work with the Attribute::Handlers module (as of version 0.76).
92 =head2 Socket Extension Dynamic in VMS
94 The Socket extension is now dynamically loaded instead of being
95 statically built in. This may or may not be a problem with ancient
96 TCP/IP stacks of VMS: we do not know since we weren't able to test
97 Perl in such configurations.
99 =head2 IEEE-format Floating Point Default on OpenVMS Alpha
101 Perl now uses IEEE format (T_FLOAT) as the default internal floating
102 point format on OpenVMS Alpha, potentially breaking binary compatibility
103 with external libraries or existing data. G_FLOAT is still available as
104 a configuration option. The default on VAX (D_FLOAT) has not changed.
106 =head2 New Unicode Properties
108 Unicode I<scripts> are now supported. Scripts are similar to (and superior
109 to) Unicode I<blocks>. The difference between scripts and blocks is that
110 scripts are the glyphs used by a language or a group of languages, while
111 the blocks are more artificial groupings of (mostly) 256 characters based
112 on the Unicode numbering.
114 In general, scripts are more inclusive, but not universally so. For
115 example, while the script C<Latin> includes all the Latin characters and
116 their various diacritic-adorned versions, it does not include the various
117 punctuation or digits (since they are not solely C<Latin>).
119 A number of other properties are now supported, including C<\p{L&}>,
120 C<\p{Any}> C<\p{Assigned}>, C<\p{Unassigned}>, C<\p{Blank}> and
121 C<\p{SpacePerl}> (along with their C<\P{...}> versions, of course).
122 See L<perlunicode> for details, and more additions.
124 The C<In> or C<Is> prefix to names used with the C<\p{...}> and C<\P{...}>
125 are now almost always optional. The only exception is that a C<In> prefix
126 is required to signify a Unicode block when a block name conflicts with a
127 script name. For example, C<\p{Tibetan}> refers to the script, while
128 C<\p{InTibetan}> refers to the block. When there is no name conflict, you
129 can omit the C<In> from the block name (e.g. C<\p{BraillePatterns}>), but
130 to be safe, it's probably best to always use the C<In>).
132 =head2 REF(...) Instead Of SCALAR(...)
134 A reference to a reference now stringifies as "REF(0x81485ec)" instead
135 of "SCALAR(0x81485ec)" in order to be more consistent with the return
138 =head2 pack/unpack D/F recycled
140 The undocumented pack/unpack template letters D/F have been recycled
141 for better use: now they stand for long double (if supported by the
142 platform) and NV (Perl internal floating point type). (They used
143 to be aliases for d/f, but you never knew that.)
151 The semantics of bless(REF, REF) were unclear and until someone proves
152 it to make some sense, it is forbidden.
156 The obsolete chat2 library that should never have been allowed
157 to escape the laboratory has been decommissioned.
161 The builtin dump() function has probably outlived most of its
162 usefulness. The core-dumping functionality will remain in future
163 available as an explicit call to C<CORE::dump()>, but in future
164 releases the behaviour of an unqualified C<dump()> call may change.
168 The very dusty examples in the eg/ directory have been removed.
169 Suggestions for new shiny examples welcome but the main issue is that
170 the examples need to be documented, tested and (most importantly)
175 The (bogus) escape sequences \8 and \9 now give an optional warning
176 ("Unrecognized escape passed through"). There is no need to \-escape
181 The list of filenames from glob() (or <...>) is now by default sorted
182 alphabetically to be csh-compliant (which is what happened before
183 in most UNIX platforms). (bsd_glob() does still sort platform
184 natively, ASCII or EBCDIC, unless GLOB_ALPHASORT is specified.)
188 Spurious syntax errors generated in certain situations, when glob()
189 caused File::Glob to be loaded for the first time, have been fixed.
193 Although "you shouldn't do that", it was possible to write code that
194 depends on Perl's hashed key order (Data::Dumper does this). The new
195 algorithm "One-at-a-Time" produces a different hashed key order.
196 More details are in L</"Performance Enhancements">.
200 lstat(FILEHANDLE) now gives a warning because the operation makes no sense.
201 In future releases this may become a fatal error.
205 The C<package;> syntax (C<package> without an argument) has been
206 deprecated. Its semantics were never that clear and its
207 implementation even less so. If you have used that feature to
208 disallow all but fully qualified variables, C<use strict;> instead.
212 The unimplemented POSIX regex features [[.cc.]] and [[=c=]] are still
213 recognised but now cause fatal errors. The previous behaviour of
214 ignoring them by default and warning if requested was unacceptable
215 since it, in a way, falsely promised that the features could be used.
219 The current user-visible implementation of pseudo-hashes (the weird
220 use of the first array element) is deprecated starting from Perl 5.8.0
221 and will be removed in Perl 5.10.0, and the feature will be
222 implemented differently. Not only is the current interface rather
223 ugly, but the current implementation slows down normal array and hash
224 use quite noticeably. The C<fields> pragma interface will remain
225 available. The I<restricted hashes> interface is expected to
226 be the replacement interface (see L<Hash::Util>).
230 The syntaxes C<< @a->[...] >> and C<< %h->{...} >> have now been deprecated.
234 After years of trying the suidperl is considered to be too complex to
235 ever be considered truly secure. The suidperl functionality is likely
236 to be removed in a future release.
240 The long deprecated uppercase aliases for the string comparison
241 operators (EQ, NE, LT, LE, GE, GT) have now been removed.
245 The tr///C and tr///U features have been removed and will not return;
246 the interface was a mistake. Sorry about that. For similar
247 functionality, see pack('U0', ...) and pack('C0', ...).
251 Earlier Perls treated "sub foo (@bar)" as equivalent to "sub foo (@)".
252 The prototypes are now checked at compile-time for invalid characters.
253 An optional warning is generated ("Illegal character in prototype...")
254 but this may be upgraded to a fatal error in a future release.
258 =head1 Core Enhancements
260 =head2 PerlIO is Now The Default
266 IO is now by default done via PerlIO rather than system's "stdio".
267 PerlIO allows "layers" to be "pushed" onto a file handle to alter the
268 handle's behaviour. Layers can be specified at open time via 3-arg
271 open($fh,'>:crlf :utf8', $path) || ...
273 or on already opened handles via extended C<binmode>:
275 binmode($fh,':encoding(iso-8859-7)');
277 The built-in layers are: unix (low level read/write), stdio (as in
278 previous Perls), perlio (re-implementation of stdio buffering in a
279 portable manner), crlf (does CRLF <=> "\n" translation as on Win32,
280 but available on any platform). A mmap layer may be available if
281 platform supports it (mostly UNIXes).
283 Layers to be applied by default may be specified via the 'open' pragma.
285 See L</"Installation and Configuration Improvements"> for the effects
286 of PerlIO on your architecture name.
290 File handles can be marked as accepting Perl's internal encoding of Unicode
291 (UTF-8 or UTF-EBCDIC depending on platform) by a pseudo layer ":utf8" :
293 open($fh,">:utf8","Uni.txt");
295 Note for EBCDIC users: the pseudo layer ":utf8" is erroneously named
296 for you since it's not UTF-8 what you will be getting but instead
297 UTF-EBCDIC. See L<perlunicode>, L<utf8>, and
298 http://www.unicode.org/unicode/reports/tr16/ for more information.
299 In future releases this naming may change.
303 File handles can translate character encodings from/to Perl's internal
304 Unicode form on read/write via the ":encoding()" layer.
308 File handles can be opened to "in memory" files held in Perl scalars via:
310 open($fh,'>', \$variable) || ...
314 Anonymous temporary files are available without need to
315 'use FileHandle' or other module via
317 open($fh,"+>", undef) || ...
319 That is a literal undef, not an undefined value.
323 The list form of C<open> is now implemented for pipes (at least on UNIX):
325 open($fh,"-|", 'cat', '/etc/motd')
327 creates a pipe, and runs the equivalent of exec('cat', '/etc/motd') in
332 If your locale environment variables (LANGUAGE, LC_ALL, LC_CTYPE, LANG)
333 contain the strings 'UTF-8' or 'UTF8' (case-insensitive matching),
334 the default encoding of your STDIN, STDOUT, and STDERR, and of
335 B<any subsequent file open>, is UTF-8.
339 =head2 Restricted Hashes
341 A restricted hash is restricted to a certain set of keys, no keys
342 outside the set can be added. Also individual keys can be restricted
343 so that the key cannot be deleted and the value cannot be changed.
344 No new syntax is involved: the Hash::Util module is the interface.
348 Perl used to be fragile in that signals arriving at inopportune moments
349 could corrupt Perl's internal state. Now Perl postpones handling of
350 signals until it's safe (between opcodes).
352 This change may have surprising side effects because signals no longer
353 interrupt Perl instantly. Perl will now first finish whatever it was
354 doing, like finishing an internal operation (like sort()) or an
355 external operation (like an I/O operation), and only then look at any
356 arrived signals (and before starting the next operation). No more corrupt
357 internal state since the current operation is always finished first,
358 but the signal may take more time to get heard. Note that breaking
359 out from potentially blocking operations should still work, though.
361 =head2 Unicode Overhaul
363 Unicode in general should be now much more usable than in Perl 5.6.0
364 (or even in 5.6.1). Unicode can be used in hash keys, Unicode in
365 regular expressions should work now, Unicode in tr/// should work now,
366 Unicode in I/O should work now. See L<perluniintro> for introduction
367 and L<perlunicode> for details.
373 The Unicode Character Database coming with Perl has been upgraded
374 to Unicode 3.2.0. For more information, see http://www.unicode.org/ .
378 For developers interested in enhancing Perl's Unicode capabilities:
379 almost all the UCD files are included with the Perl distribution in
380 the F<lib/unicore subdirectory>. The most notable omission, for space
381 considerations, is the Unihan database.
385 The properties \p{Blank} and \p{SpacePerl} have been added. "Blank" is like
386 C isblank(), that is, it contains only "horizontal whitespace" (the space
387 character is, the newline isn't), and the "SpacePerl" is the Unicode
388 equivalent of C<\s> (\p{Space} isn't, since that includes the vertical
389 tabulator character, whereas C<\s> doesn't.)
391 See "New Unicode Properties" earlier in this document for additional
392 information on changes with Unicode properties.
396 =head2 Understanding of Numbers
398 In general a lot of fixing has happened in the area of Perl's
399 understanding of numbers, both integer and floating point. Since in
400 many systems the standard number parsing functions like C<strtoul()>
401 and C<atof()> seem to have bugs, Perl tries to work around their
402 deficiencies. This results hopefully in more accurate numbers.
404 Perl now tries internally to use integer values in numeric conversions
405 and basic arithmetics (+ - * /) if the arguments are integers, and
406 tries also to keep the results stored internally as integers.
407 This change leads to often slightly faster and always less lossy
408 arithmetics. (Previously Perl always preferred floating point numbers
411 =head2 Miscellaneous Changes
417 AUTOLOAD is now lvaluable, meaning that you can add the :lvalue attribute
418 to AUTOLOAD subroutines and you can assign to the AUTOLOAD return value.
422 C<perl -d:Module=arg,arg,arg> now works (previously one couldn't pass
423 in multiple arguments.)
427 The builtin dump() now gives an optional warning
428 C<dump() better written as CORE::dump()>,
429 meaning that by default C<dump(...)> is resolved as the builtin
430 dump() which dumps core and aborts, not as (possibly) user-defined
431 C<sub dump>. To call the latter, qualify the call as C<&dump(...)>.
432 (The whole dump() feature is to considered deprecated, and possibly
433 removed/changed in future releases.)
437 chomp() and chop() are now overridable. Note, however, that their
438 prototype (as given by C<prototype("CORE::chomp")> is undefined,
439 because it cannot be expressed and therefore one cannot really write
440 replacements to override these builtins.
444 END blocks are now run even if you exit/die in a BEGIN block.
445 Internally, the execution of END blocks is now controlled by
446 PL_exit_flags & PERL_EXIT_DESTRUCT_END. This enables the new
447 behaviour for Perl embedders. This will default in 5.10. See
452 Formats now support zero-padded decimal fields.
456 Lvalue subroutines can now return C<undef> in list context.
457 However, the lvalue subroutine feature still remains experimental.
461 A lost warning "Can't declare ... dereference in my" has been
462 restored (Perl had it earlier but it became lost in later releases.)
466 A new special regular expression variable has been introduced:
467 C<$^N>, which contains the most-recently closed group (submatch).
471 C<no Module;> now works even if there is no "sub unimport" in the Module.
475 The numerical comparison operators return C<undef> if either operand
476 is a NaN. Previously the behaviour was unspecified.
480 The following builtin functions are now overridable: each(), keys(),
481 pop(), push(), shift(), splice(), unshift().
485 C<pack() / unpack()> now can group template letters with C<()> and then
486 apply repetition/count modifiers on the groups.
490 C<pack() / unpack()> can now process the Perl internal numeric types:
491 IVs, UVs, NVs-- and also long doubles, if supported by the platform.
492 The template letters are C<j>, C<J>, C<F>, and C<D>.
496 C<pack('U0a*', ...)> can now be used to force a string to UTF8.
500 my __PACKAGE__ $obj now works.
504 The printf() and sprintf() now support parameter reordering using the
505 C<%\d+\$> and C<*\d+\$> syntaxes. For example
507 print "%2\$s %1\$s\n", "foo", "bar";
509 will print "bar foo\n". This feature helps in writing
510 internationalised software, and in general when the order
511 of the parameters can vary.
515 prototype(\&) is now available.
519 prototype(\[$@%&]) is now available to implicitly create references
520 (useful for example if you want to emulate the tie() interface).
524 A new command-line option, C<-t> is available. It is the
525 little brother of C<-T>: instead of dying on taint violations,
526 lexical warnings are given. B<This is only meant as a temporary
527 debugging aid while securing the code of old legacy applications.
528 This is not a substitute for -T.>
532 In other taint news, the C<exec LIST> and C<system LIST> have now been
533 considered too risky (think C<exec @ARGV>: it can start any program
534 with any arguments), and now the said forms cause a warning.
535 You should carefully launder the arguments to guarantee their
536 validity. In future releases of Perl the forms will become fatal
537 errors so consider starting laundering now.
541 If tr/// is just counting characters, it doesn't attempt to
546 untie() will now call an UNTIE() hook if it exists. See L<perltie>
551 L<utime> now supports C<utime undef, undef, @files> to change the
552 file timestamps to the current time.
556 The rules for allowing underscores (underbars) in numeric constants
557 have been relaxed and simplified: now you can have an underscore
558 simply B<between digits>.
562 Rather than relying on C's argv[0] (which may not contain a full pathname)
563 where possible $^X is now set by asking the operating system.
564 (eg by reading F</proc/self/exe> on Linux, F</proc/curproc/file> on FreeBSD)
568 A new variable, C<${^TAINT}>, indicates whether taint mode is enabled.
572 You can now override the readline() builtin, and this overrides also
573 the <FILEHANDLE> angle bracket operator.
577 The command-line options -s and -F are now recognized on the shebang
582 Use of the C</c> match modifier without an accompanying C</g> modifier
583 elicits a new warning: C<Use of /c modifier is meaningless without /g>.
585 Use of C</c> in substitutions, even with C</g>, elicits
586 C<Use of /c modifier is meaningless in s///>.
588 Use of C</g> with C<split> elicits C<Use of /g modifier is meaningless
593 =head1 Modules and Pragmata
595 =head2 New Modules and Pragmata
601 C<Attribute::Handlers> allows a class to define attribute handlers.
604 use Attribute::Handlers;
605 sub Wolf :ATTR(SCALAR) { print "howl!\n" }
607 # later, in some package using or inheriting from MyPack...
609 my MyPack $Fluffy : Wolf; # the attribute handler Wolf will be called
611 Both variables and routines can have attribute handlers. Handlers can
612 be specific to type (SCALAR, ARRAY, HASH, or CODE), or specific to the
613 exact compilation phase (BEGIN, CHECK, INIT, or END).
617 B<B::Concise> is a new compiler backend for walking the Perl syntax
618 tree, printing concise info about ops, from Stephen McCamant. The
619 output is highly customisable. See L<B::Concise>.
623 The new bignum, bigint, and bigrat pragmas implement transparent
624 bignum support (using the Math::BigInt, Math::BigFloat, and
625 Math::BigRat backends), by Tels.
629 C<Class::ISA> for reporting the search path for a class's ISA tree,
630 by Sean Burke, has been added. See L<Class::ISA>.
634 C<Cwd> has now a split personality: if possible, an XS extension is
635 used, (this will hopefully be faster, more secure, and more robust)
636 but if not possible, the familiar Perl implementation is used.
640 C<Devel::PPPort>, originally from Kenneth Albanowski and now
641 maintained by Paul Marquess, has been added. It is primarily used
642 by C<h2xs> to enhance portability of XS modules between different
647 C<Digest>, frontend module for calculating digests (checksums), from
648 Gisle Aas, has been added. See L<Digest>.
652 C<Digest::MD5> for calculating MD5 digests (checksums) as defined in
653 RFC 1321, from Gisle Aas, has been added. See L<Digest::MD5>.
655 use Digest::MD5 'md5_hex';
657 $digest = md5_hex("Thirsty Camel");
659 print $digest, "\n"; # 01d19d9d2045e005c3f1b80e8b164de1
661 NOTE: the C<MD5> backward compatibility module is deliberately not
662 included since its further use is discouraged.
666 C<Encode>, orginally by Nick Ing-Simmons and now maintained by Dan
667 Kogai, provides a mechanism to translate between different character
668 encodings. Support for Unicode, ISO-8859-1, and ASCII are compiled in
669 to the module. Several other encodings (like the rest of the
670 ISO-8859, CP*/Win*, Mac, KOI8-R, three variants EBCDIC, Chinese,
671 Japanese, and Korean encodings) are included and can be loaded at
672 runtime. (For space considerations, the largest Chinese encodings
673 have been separated into their own CPAN module, Encode::HanExtra,
674 which Encode will use if available). See L<Encode>.
676 Any encoding supported by Encode module is also available to the
677 ":encoding()" layer if PerlIO is used.
681 C<Hash::Util> is the interface to the new I<restricted hashes>
682 feature. (Implemented by Jeffrey Friedl, Nick Ing-Simmons, and
687 C<I18N::Langinfo> can be use to query locale information.
688 See L<I18N::Langinfo>.
692 C<I18N::LangTags> has functions for dealing with RFC3066-style
693 language tags, by Sean Burke. See L<I18N::LangTags>.
697 C<ExtUtils::Constant> is a new tool for extension writers for
698 generating XS code to import C header constants, by Nicholas Clark.
699 See L<ExtUtils::Constant>.
703 C<Filter::Simple> is an easy-to-use frontend to Filter::Util::Call,
704 from Damian Conway. See L<Filter::Simple>.
710 use Filter::Simple sub {
711 while (my ($from, $to) = splice @_, 0, 2) {
720 use MyFilter qr/red/ => 'green';
722 print "red\n"; # this code is filtered, will print "green\n"
723 print "bored\n"; # this code is filtered, will print "bogreen\n"
727 print "red\n"; # this code is not filtered, will print "red\n"
731 C<File::Temp> allows one to create temporary files and directories in
732 an easy, portable, and secure way, by Tim Jenness. See L<File::Temp>.
736 C<Filter::Util::Call> provides you with the framework to write
737 I<Source Filters> in Perl, from Paul Marquess. For most uses the
738 frontend Filter::Simple is to be preferred. See L<Filter::Util::Call>.
742 C<if> is a new pragma for conditional inclusion of modules, from
747 L<libnet> is a collection of perl5 modules related to network
748 programming, from Graham Barr. See L<Net::FTP>, L<Net::NNTP>,
749 L<Net::Ping> (not part of libnet, but related), L<Net::POP3>,
750 L<Net::SMTP>, and L<Net::Time>.
752 Perl installation leaves libnet unconfigured, use F<libnetcfg> to configure.
756 C<List::Util> is a selection of general-utility list subroutines, like
757 sum(), min(), first(), and shuffle(), by Graham Barr. See L<List::Util>.
761 C<Locale::Constants>, C<Locale::Country>, C<Locale::Currency>
762 C<Locale::Language>, and L<Locale::Script>, from Neil Bowers, have
763 been added. They provide the codes for various locale standards, such
764 as "fr" for France, "usd" for US Dollar, and "ja" for Japanese.
768 $country = code2country('jp'); # $country gets 'Japan'
769 $code = country2code('Norway'); # $code gets 'no'
771 See L<Locale::Constants>, L<Locale::Country>, L<Locale::Currency>,
772 and L<Locale::Language>.
776 C<Locale::Maketext> is localization framework from Sean Burke. See
777 L<Locale::Maketext>, and L<Locale::Maketext::TPJ13>. The latter is an
778 article about software localization, originally published in The Perl
779 Journal #13, republished here with kind permission.
783 C<Math::BigRat> for big rational numbers, to accompany Math::BigInt and
784 Math::BigFloat, from Tels.
788 C<Memoize> can make your functions faster by trading space for time,
789 from Mark-Jason Dominus. See L<Memoize>.
793 C<MIME::Base64> allows you to encode data in base64, from Gisle Aas,
794 as defined in RFC 2045 - I<MIME (Multipurpose Internet Mail
799 $encoded = encode_base64('Aladdin:open sesame');
800 $decoded = decode_base64($encoded);
802 print $encoded, "\n"; # "QWxhZGRpbjpvcGVuIHNlc2FtZQ=="
808 C<MIME::QuotedPrint> allows you to encode data in quoted-printable
809 encoding, as defined in RFC 2045 - I<MIME (Multipurpose Internet Mail
810 Extensions)>, from Gisle Aas.
812 use MIME::QuotedPrint;
814 $encoded = encode_qp("Smiley in Unicode: \x{263a}");
815 $decoded = decode_qp($encoded);
817 print $encoded, "\n"; # "Smiley in Unicode: =263A"
819 MIME::QuotedPrint has been enhanced to provide the basic methods
820 necessary to use it with PerlIO::Via as in :
822 use MIME::QuotedPrint;
823 open($fh,">Via(MIME::QuotedPrint)",$path);
825 See L<MIME::QuotedPrint>.
829 C<NEXT> is pseudo-class for method redispatch, from Damian Conway.
834 C<open> is a new pragma for setting the default I/O disciplines
839 C<PerlIO::Scalar> provides the implementation of IO to "in memory"
840 Perl scalars as discussed above, from Nick Ing-Simmons. It also
841 serves as an example of a loadable PerlIO layer. Other future
842 possibilities include PerlIO::Array and PerlIO::Code.
843 See L<PerlIO::Scalar>.
847 C<PerlIO::Via> acts as a PerlIO layer and wraps PerlIO layer
848 functionality provided by a class (typically implemented in perl
849 code), from Nick Ing-Simmons.
851 use MIME::QuotedPrint;
852 open($fh,">Via(MIME::QuotedPrint)",$path);
854 This will automatically convert everything output to C<$fh>
855 to Quoted-Printable. See L<PerlIO::Via>.
859 C<Pod::ParseLink>, by Russ Allbery, has been added,
860 to parse LZ<><> links in pods as described in the new
865 C<Pod::Text::Overstrike>, by Joe Smith, has been added.
866 It converts POD data to formatted overstrike text.
867 See L<Pod::Text::Overstrike>.
871 C<Scalar::Util> is a selection of general-utility scalar subroutines,
872 like blessed(), reftype(), and tainted(). See L<Scalar::Util>.
876 C<sort> is a new pragma for controlling the behaviour of sort().
880 C<Storable> gives persistence to Perl data structures by allowing the
881 storage and retrieval of Perl data to and from files in a fast and
882 compact binary format, from Raphael Manfredi. See L<Storable>.
886 C<Switch>, from Damian Conway, has been added. Just by saying
890 you have C<switch> and C<case> available in Perl.
896 case 1 { print "number 1" }
897 case "a" { print "string a" }
898 case [1..10,42] { print "number in list" }
899 case (@array) { print "number in list" }
900 case /\w+/ { print "pattern" }
901 case qr/\w+/ { print "pattern" }
902 case (%hash) { print "entry in hash" }
903 case (\%hash) { print "entry in hash" }
904 case (\&sub) { print "arg to subroutine" }
905 else { print "previous case not true" }
912 C<Test::More> is yet another framework for writing test scripts,
913 more extensive than Test::Simple, by Michael Schwern. See L<Test::More>.
917 C<Test::Simple> has basic utilities for writing tests, by Michael
918 Schwern. See L<Test::Simple>.
922 C<Text::Balanced> has been added, for extracting delimited text
923 sequences from strings, from Damian Conway.
925 use Text::Balanced 'extract_delimited';
927 ($a, $b) = extract_delimited("'never say never', he never said", "'", '');
929 $a will be "'never say never'", $b will be ', he never said'.
931 In addition to extract_delimited() there are also extract_bracketed(),
932 extract_quotelike(), extract_codeblock(), extract_variable(),
933 extract_tagged(), extract_multiple(), gen_delimited_pat(), and
934 gen_extract_tagged(). With these you can implement rather advanced
935 parsing algorithms. See L<Text::Balanced>.
939 C<threads> is an interface to interpreter threads, by Arthur Bergman.
940 Interpreter threads (ithreads) is the new thread model introduced in
941 Perl 5.6 but only available as an internal interface for extension
942 writers (and for Win32 Perl for C<fork()> emulation). See L<threads>.
946 C<threads::shared> allows data sharing for interpreter threads, from
947 Arthur Bergman. In the ithreads model any data sharing between
948 threads must be explicit, as opposed to the old 5.005 thread model
949 where data sharing was implicit. See L<threads::shared>.
953 C<Tie::File>, by Mark-Jason Dominus, associates a Perl array with the
958 C<Tie::Memoize>, by Ilya Zakharevich, provides on-demand loaded hashes.
962 C<Tie::RefHash::Nestable>, by Edward Avis, allows storing hash
963 references (unlike the standard Tie::RefHash) The module is contained
964 within Tie::RefHash, see L<Tie::RefHash>.
968 C<Time::HiRes> provides high resolution timing (ualarm, usleep,
969 and gettimeofday), from Douglas E. Wegscheid. See L<Time::HiRes>.
973 C<Unicode::UCD> offers a querying interface to the Unicode Character
974 Database. See L<Unicode::UCD>.
978 C<Unicode::Collate> implements the UCA (Unicode Collation Algorithm)
979 for sorting Unicode strings, by SADAHIRO Tomoyuki. See L<Unicode::Collate>.
983 C<Unicode::Normalize> implements the various Unicode normalization
984 forms, by SADAHIRO Tomoyuki. See L<Unicode::Normalize>.
988 C<XS::Typemap>, by Tim Jenness, is a test extension that exercises XS
989 typemaps. Nothing gets installed but for extension writers the code
994 =head2 Updated And Improved Modules and Pragmata
1000 The following independently supported modules have been updated to the
1001 newest versions from CPAN: CGI, CPAN, DB_File, File::Spec, File::Temp,
1002 Getopt::Long, Math::BigFloat, Math::BigInt, the podlators bundle
1003 (Pod::Man, Pod::Text), Pod::LaTeX, Pod::Parser, Storable,
1004 Term::ANSIColor, Test, Text-Tabs+Wrap.
1008 The attributes::reftype() now works on tied arguments.
1012 AutoLoader can now be disabled with C<no AutoLoader;>.
1016 B::Deparse has been significantly enhanced. It now can deparse almost
1017 all of the standard test suite (so that the tests still succeed).
1018 There is a make target "test.deparse" for trying this out.
1022 Class::Struct can now define the classes in compile time.
1026 Class::Struct now assigns the array/hash element if the accessor
1027 is called with an array/hash element as the B<sole> argument.
1031 Data::Dumper has now an option to sort hashes.
1035 Data::Dumper has now an option to dump code references
1040 DB_File now supports newer Berkeley DB versions, among
1045 The English module can now be used without the infamous performance
1048 use English '-no_match_vars';
1050 (Assuming, of course, that one doesn't need the troublesome variables
1051 C<$`>, C<$&>, or C<$'>.) Also, introduced C<@LAST_MATCH_START> and
1052 C<@LAST_MATCH_END> English aliases for C<@-> and C<@+>.
1056 Fcntl, Socket, and Sys::Syslog have been rewritten to use the
1057 new-style constant dispatch section (see L<ExtUtils::Constant>).
1058 This means that they will be more robust and hopefully faster.
1062 File::Find now chdir()s correctly when chasing symbolic links.
1066 File::Find now has pre- and post-processing callbacks. It also
1067 correctly changes directories when chasing symbolic links. Callbacks
1068 (naughtily) exiting with "next;" instead of "return;" now work.
1072 File::Find is now (again) reentrant. It also has been made
1077 The warnings issued by File::Find now belong to their own category.
1078 You can enable/disable them with C<use/no warnings 'File::Find';>.
1082 File::Glob::glob() renamed to File::Glob::bsd_glob() to avoid
1083 prototype mismatch with CORE::glob().
1087 File::Glob now supports C<GLOB_LIMIT> constant to limit the size of
1088 the returned list of filenames.
1092 Devel::Peek now has an interface for the Perl memory statistics
1093 (this works only if you are using perl's malloc, and if you have
1094 compiled with debugging).
1098 IPC::Open3 now allows the use of numeric file descriptors.
1102 IO::Socket has now atmark() method, which returns true if the socket
1103 is positioned at the out-of-band mark. The method is also exportable
1104 as a sockatmark() function.
1108 IO::Socket::INET has support for ReusePort option (if your platform
1109 supports it). The Reuse option now has an alias, ReuseAddr. For clarity
1110 you may want to prefer ReuseAddr.
1114 IO::Socket::INET now supports C<LocalPort> of zero (usually meaning
1115 that the operating system will make one up.)
1119 use lib now works identically to @INC. Removing directories
1120 with 'no lib' now works.
1124 ExtUtils::MakeMaker now uses File::Spec internally, which hopefully
1125 leads into better portability.
1129 Math::BigFloat and Math::BigInt have undergone a full rewrite.
1130 They are now magnitudes faster, and they support various
1131 bignum libraries such as GMP and PARI as their backends.
1135 Math::Complex handles inf, NaN etc., better.
1139 Net::Ping has been muchly enhanced: multihoming is now supported,
1140 Win32 functionality is better, there is now time measuring
1141 functionality (optionally high-resolution using Time::HiRes),
1142 and there is now "external" protocol which uses Net::Ping::External
1143 module which runs your external ping utility and parses the output.
1144 A version of Net::Ping::External is available in CPAN.
1146 Note that some of the Net::Ping tests are disabled when running
1147 under the Perl distribution since one cannot assume one or more
1148 of the following: enabled echo port at localhost, full Internet
1149 connectivity, or sympathetic firewalls. You can set the environment
1150 variable PERL_TEST_Net_Ping to "1" (one) before running the Perl test
1151 suite to enable all the Net::Ping tests.
1155 POSIX::sigaction() is now much more flexible and robust.
1156 You can now install coderef handlers, 'DEFAULT', and 'IGNORE'
1157 handlers, installing new handlers was not atomic.
1161 In Safe the C<%INC> now localised in a Safe compartment so that
1166 In SDBM_File on dosish platforms, some keys went missing because of
1167 lack of support for files with "holes". A workaround for the problem
1172 In Search::Dict one can now have a pre-processing hook for the
1173 lines being searched.
1177 The Shell module now has an OO interface.
1181 In Sys::Syslog there is now a failover mechanism that will go
1182 through alternative connection mechanisms until the message
1183 is successfully logged.
1187 The Test module has been significantly enhanced.
1191 Time::Local::timelocal() does not handle fractional seconds anymore.
1192 The rationale is that neither does localtime(), and timelocal() and
1193 localtime() are supposed to be inverses of each other.
1197 The vars pragma now supports declaring fully qualified variables.
1198 (Something that C<our()> does not and will not support.)
1202 The C<utf8::> name space (as in the pragma) provides various
1203 Perl-callable functions to provide low level access to Perl's
1204 internal Unicode representation. At the moment only length()
1205 has been implemented.
1209 =head1 Utility Changes
1215 Emacs perl mode (emacs/cperl-mode.el) has been updated to version
1220 F<emacs/e2ctags.pl> is now much faster.
1224 C<enc2xs> is a tool for people adding their own encodings to the
1229 C<h2ph> now supports C trigraphs.
1233 C<h2xs> now produces a template README.
1237 C<h2xs> now uses C<Devel::PPort> for better portability between
1238 different versions of Perl.
1242 C<h2xs> uses the new L<ExtUtils::Constant> module which will affect
1243 newly created extensions that define constants. Since the new code is
1244 more correct (if you have two constants where the first one is a
1245 prefix of the second one, the first constant B<never> gets defined),
1246 less lossy (it uses integers for integer constant, as opposed to the
1247 old code that used floating point numbers even for integer constants),
1248 and slightly faster, you might want to consider regenerating your
1249 extension code (the new scheme makes regenerating easy).
1250 L<h2xs> now also supports C trigraphs.
1254 C<libnetcfg> has been added to configure the libnet.
1258 C<perlbug> is now much more robust. It also sends the bug report to
1259 perl.org, not perl.com.
1263 C<perlcc> has been rewritten and its user interface (that is,
1264 command line) is much more like that of the UNIX C compiler, cc.
1265 (The perlbc tools has been removed. Use C<perlcc -B> instead.)
1269 C<perlivp> is a new Installation Verification Procedure utility
1270 for running any time after installing Perl.
1274 C<piconv> is an implementation of the character conversion utility
1275 C<iconv>, demonstrating the new Encode module.
1279 C<pod2html> now allows specifying a cache directory.
1283 C<pod2html> now produces XHTML 1.0.
1287 C<pod2html> now understands POD written using different line endings
1288 (PC-like CRLF versus UNIX-like LF versus MacClassic-like CR).
1292 C<s2p> has been completely rewritten in Perl. (It is in fact a full
1293 implementation of sed in Perl: you can use the sed functionality by
1294 using the C<psed> utility.)
1298 C<xsubpp> now understands POD documentation embedded in the *.xs files.
1302 C<xsubpp> now supports OUT keyword.
1306 =head1 New Documentation
1312 perl56delta details the changes between the 5.005 release and the
1317 perlclib documents the internal replacements for standard C library
1318 functions. (Interesting only for extension writers and Perl core
1323 perldebtut is a Perl debugging tutorial.
1327 perlebcdic contains considerations for running Perl on EBCDIC platforms.
1331 perlintro is a gentle introduction to Perl.
1335 perliol documents the internals of PerlIO with layers.
1339 perlmodstyle is a style guide for writing modules.
1343 perlnewmod tells about writing and submitting a new module.
1347 perlpacktut is a pack() tutorial.
1351 perlpod has been rewritten to be clearer and to record the best
1352 practices gathered over the years.
1356 perlpodspec is a more formal specification of the pod format,
1357 mainly of interest for writers of pod applications, not to
1358 people writing in pod.
1362 perlretut is a regular expression tutorial.
1366 perlrequick is a regular expressions quick-start guide.
1367 Yes, much quicker than perlretut.
1371 perltodo has been updated.
1375 perltootc has been renamed as perltooc (to not to conflict
1376 with perltoot in filesystems restricted to "8.3" names)
1380 perluniintro is an introduction to using Unicode in Perl.
1381 (perlunicode is more of a detailed reference and background
1386 perlutil explains the command line utilities packaged with the Perl
1391 The following platform-specific documents are available before
1392 the installation as README.I<platform>, and after the installation
1395 perlaix perlamiga perlapollo perlbeos perlbs2000
1396 perlce perlcygwin perldgux perldos perlepoc perlhpux
1397 perlhurd perlmachten perlmacos perlmint perlmpeix
1398 perlnetware perlos2 perlos390 perlplan9 perlqnx perlsolaris
1399 perltru64 perluts perlvmesa perlvms perlvos perlwin32
1405 The documentation for the POSIX-BC platform is called "BS2000", to avoid
1406 confusion with the Perl POSIX module.
1410 The documentation for the WinCE platform is called perlce (README.ce
1411 in the source code kit), to avoid confusion with the perlwin32
1412 documentation on 8.3-restricted filesystems.
1416 =head1 Performance Enhancements
1422 map() could get pathologically slow when the result list it generates
1423 is larger than the source list. The performance has been improved for
1428 sort() has been changed to use primarily mergesort internally as
1429 opposed to the earlier quicksort. For very small lists this may
1430 result in slightly slower sorting times, but in general the speedup
1431 should be at least 20%. Additional bonuses are that the worst case
1432 behaviour of sort() is now better (in computer science terms it now
1433 runs in time O(N log N), as opposed to quicksort's Theta(N**2)
1434 worst-case run time behaviour), and that sort() is now stable
1435 (meaning that elements with identical keys will stay ordered as they
1436 were before the sort). See the C<sort> pragma for information.
1438 The story in more detail: suppose you want to serve yourself a little
1441 @digits = ( 3,1,4,1,5,9 );
1443 A numerical sort of the digits will yield (1,1,3,4,5,9), as expected.
1444 Which C<1> comes first is hard to know, since one C<1> looks pretty
1445 much like any other. You can regard this as totally trivial,
1446 or somewhat profound. However, if you just want to sort the even
1447 digits ahead of the odd ones, then what will
1449 sort { ($a % 2) <=> ($b % 2) } @digits;
1451 yield? The only even digit, C<4>, will come first. But how about
1452 the odd numbers, which all compare equal? With the quicksort algorithm
1453 used to implement Perl 5.6 and earlier, the order of ties is left up
1454 to the sort. So, as you add more and more digits of Pi, the order
1455 in which the sorted even and odd digits appear will change.
1456 and, for sufficiently large slices of Pi, the quicksort algorithm
1457 in Perl 5.8 won't return the same results even if reinvoked with the
1458 same input. The justification for this rests with quicksort's
1459 worst case behavior. If you run
1461 sort { $a <=> $b } ( 1 .. $N , 1 .. $N );
1463 (something you might approximate if you wanted to merge two sorted
1464 arrays using sort), doubling $N doesn't just double the quicksort time,
1465 it I<quadruples> it. Quicksort has a worst case run time that can
1466 grow like N**2, so-called I<quadratic> behaviour, and it can happen
1467 on patterns that may well arise in normal use. You won't notice this
1468 for small arrays, but you I<will> notice it with larger arrays,
1469 and you may not live long enough for the sort to complete on arrays
1470 of a million elements. So the 5.8 quicksort scrambles large arrays
1471 before sorting them, as a statistical defence against quadratic behaviour.
1472 But that means if you sort the same large array twice, ties may be
1473 broken in different ways.
1475 Because of the unpredictability of tie-breaking order, and the quadratic
1476 worst-case behaviour, quicksort was I<almost> replaced completely with
1477 a stable mergesort. I<Stable> means that ties are broken to preserve
1478 the original order of appearance in the input array. So
1480 sort { ($a % 2) <=> ($b % 2) } (3,1,4,1,5,9);
1482 will yield (4,3,1,1,5,9), guaranteed. The even and odd numbers
1483 appear in the output in the same order they appeared in the input.
1484 Mergesort has worst case O(NlogN) behaviour, the best value
1485 attainable. And, ironically, this mergesort does particularly
1486 well where quicksort goes quadratic: mergesort sorts (1..$N, 1..$N)
1487 in O(N) time. But quicksort was rescued at the last moment because
1488 it is faster than mergesort on certain inputs and platforms.
1489 For example, if you really I<don't> care about the order of even
1490 and odd digits, quicksort will run in O(N) time; it's very good
1491 at sorting many repetitions of a small number of distinct elements.
1492 The quicksort divide and conquer strategy works well on platforms
1493 with relatively small, very fast, caches. Eventually, the problem gets
1494 whittled down to one that fits in the cache, from which point it
1495 benefits from the increased memory speed.
1497 Quicksort was rescued by implementing a sort pragma to control aspects
1498 of the sort. The B<stable> subpragma forces stable behaviour,
1499 regardless of algorithm. The B<_quicksort> and B<_mergesort>
1500 subpragmas are heavy-handed ways to select the underlying implementation.
1501 The leading C<_> is a reminder that these subpragmas may not survive
1502 beyond 5.8. More appropriate mechanisms for selecting the implementation
1503 exist, but they wouldn't have arrived in time to save quicksort.
1507 Hashes now use Bob Jenkins "One-at-a-Time" hashing key algorithm
1508 ( http://burtleburtle.net/bob/hash/doobs.html ). This algorithm is
1509 reasonably fast while producing a much better spread of values than
1510 the old hashing algorithm (originally by Chris Torek, later tweaked by
1511 Ilya Zakharevich). Hash values output from the algorithm on a hash of
1512 all 3-char printable ASCII keys comes much closer to passing the
1513 DIEHARD random number generation tests. According to perlbench, this
1514 change has not affected the overall speed of Perl.
1518 unshift() should now be noticeably faster.
1522 =head1 Installation and Configuration Improvements
1524 =head2 Generic Improvements
1530 INSTALL now explains how you can configure Perl to use 64-bit
1531 integers even on non-64-bit platforms.
1535 Policy.sh policy change: if you are reusing a Policy.sh file
1536 (see INSTALL) and you use Configure -Dprefix=/foo/bar and in the old
1537 Policy $prefix eq $siteprefix and $prefix eq $vendorprefix, all of
1538 them will now be changed to the new prefix, /foo/bar. (Previously
1539 only $prefix changed.) If you do not like this new behaviour,
1540 specify prefix, siteprefix, and vendorprefix explicitly.
1544 A new optional location for Perl libraries, otherlibdirs, is available.
1545 It can be used for example for vendor add-ons without disturbing Perl's
1546 own library directories.
1550 In many platforms the vendor-supplied 'cc' is too stripped-down to
1551 build Perl (basically, 'cc' doesn't do ANSI C). If this seems
1552 to be the case and 'cc' does not seem to be the GNU C compiler
1553 'gcc', an automatic attempt is made to find and use 'gcc' instead.
1557 gcc needs to closely track the operating system release to avoid
1558 build problems. If Configure finds that gcc was built for a different
1559 operating system release than is running, it now gives a clearly visible
1560 warning that there may be trouble ahead.
1564 If binary compatibility with the 5.005 release is not wanted, Configure
1565 no longer suggests including the 5.005 modules in @INC.
1569 Configure C<-S> can now run non-interactively.
1573 Configure support for pdp11-style memory models has been removed due
1578 configure.gnu now works with options with whitespace in them.
1582 installperl now outputs everything to STDERR.
1586 $Config{byteorder} is now computed dynamically (this is more robust
1587 with "fat binaries" where an executable image contains binaries for
1588 more than one binary platform.)
1592 Because PerlIO is now the default on most platforms, "-perlio" doesn't
1593 get appended to the $Config{archname} (also known as $^O) anymore.
1594 Instead, if you explicitly choose not to use perlio (Configure command
1595 line option -Uuseperlio), you will get "-stdio" appended.
1599 Another change related to the architecture name is that "-64all"
1600 (-Duse64bitall, or "maximally 64-bit") is appended only if your
1601 pointers are 64 bits wide. (To be exact, the use64bitall is ignored.)
1605 In AFS installations one can configure the root of the AFS to be
1606 somewhere else than the default F</afs> by using the Configure
1607 parameter C<-Dafsroot=/some/where/else>.
1611 APPLLIB_EXP, a less-know configuration-time definition, has been
1612 documented. It can be used to prepend site-specific directories
1613 to Perl's default search path (@INC), see INSTALL for information.
1617 The version of Berkeley DB used when the Perl (and, presumably, the
1618 DB_File extension) was built is now available as
1619 C<@Config{qw(db_version_major db_version_minor db_version_patch)}>
1620 from Perl and as C<DB_VERSION_MAJOR_CFG DB_VERSION_MINOR_CFG
1621 DB_VERSION_PATCH_CFG> from C.
1625 Building Berkeley DB3 for compatibility modes for DB, NDBM, and ODBM
1626 has been documented in INSTALL.
1630 If you have CPAN access (either network or a local copy such as a
1631 CD-ROM) you can during specify extra modules to Configure to build and
1632 install with Perl using the -Dextras=... option. See INSTALL for
1637 In addition to config.over a new override file, config.arch, is
1638 available. That is supposed to be used by hints file writers for
1639 architecture-wide changes (as opposed to config.over which is for
1644 If your file system supports symbolic links you can build Perl outside
1645 of the source directory by
1647 mkdir /tmp/perl/build/directory
1648 cd /tmp/perl/build/directory
1649 sh /path/to/perl/source/Configure -Dmksymlinks ...
1651 This will create in /tmp/perl/build/directory a tree of symbolic links
1652 pointing to files in /path/to/perl/source. The original files are left
1653 unaffected. After Configure has finished you can just say
1657 and Perl will be built and tested, all in /tmp/perl/build/directory.
1661 For Perl developers several new make targets for profiling
1662 and debugging have been added, see L<perlhack>.
1668 Use of the F<gprof> tool to profile Perl has been documented in
1669 L<perlhack>. There is a make target called "perl.gprof" for
1670 generating a gprofiled Perl executable.
1674 If you have GCC 3, there is a make target called "perl.gcov" for
1675 creating a gcoved Perl executable for coverage analysis. See
1680 If you are on IRIX or Tru64 platforms, new profiling/debugging options
1681 have been added, see L<perlhack> for more information about pixie and
1688 Guidelines of how to construct minimal Perl installations have
1689 been added to INSTALL.
1693 The Thread extension is now not built at all under ithreads
1694 (C<Configure -Duseithreads>) because it wouldn't work anyway (the
1695 Thread extension requires being Configured with C<-Duse5005threads>).
1697 But note that the Thread.pm interface is now shared by both
1702 The Gconvert macro ($Config{d_Gconvert}) used by perl for stringifying
1703 floating-point numbers is now more picky about using sprintf %.*g
1704 rules for the conversion. Some platforms that used to use gcvt may
1705 now resort to the slower sprintf.
1709 =head2 New Or Improved Platforms
1711 For the list of platforms known to support Perl,
1712 see L<perlport/"Supported Platforms">.
1718 AIX dynamic loading should be now better supported.
1722 AIX should now work better with gcc, threads, and 64-bitness. Also the
1723 long doubles support in AIX should be better now. See L<perlaix>.
1727 After a long pause, AmigaOS has been verified to be happy with Perl.
1731 AtheOS ( http://www.atheos.cx/ ) is a new platform.
1735 BeOS has been reclaimed.
1739 DG/UX platform now supports the 5.005-style threads. See L<perldgux>.
1743 DYNIX/ptx platform (a.k.a. dynixptx) is supported at or near osvers 4.5.2.
1747 EBCDIC platforms (z/OS, also known as OS/390, POSIX-BC, and VM/ESA)
1748 have been regained. Many test suite tests still fail and the
1749 co-existence of Unicode and EBCDIC isn't quite settled, but the
1750 situation is much better than with Perl 5.6. See L<perlos390>,
1751 L<perlbs2000> (for POSIX-BC), and L<perlvmesa> for more information.
1755 Building perl with -Duseithreads or -Duse5005threads now works under
1756 HP-UX 10.20 (previously it only worked under 10.30 or later). You will
1757 need a thread library package installed. See README.hpux.
1761 MacOS Classic (MacPerl has of course been available since
1762 perl 5.004 but now the source code bases of standard Perl
1763 and MacPerl have been synchronised)
1767 MacOS X (or Darwin) should now be able to build Perl even on HFS+
1768 filesystems. (The case-insensitivity confused the Perl build process.)
1772 NCR MP-RAS is now supported.
1776 All the NetBSD specific patches (except for the installation
1777 specific ones) have been merged back to the main distribution.
1781 NetWare from Novell is now supported. See L<perlnetware>.
1785 NonStop-UX is now supported.
1789 NEC SUPER-UX is now supported.
1793 All the OpenBSD specific patches (except for the installation
1794 specific ones) have been merged back to the main distribution.
1798 Perl has been tested with the GNU pth userlevel thread package
1799 ( http://www.gnu.org/software/pth/pth.html ) . All but one thread
1800 test worked, and that one failure was because of test results arriving
1801 in unexpected order.
1805 Amdahl UTS UNIX mainframe platform is now supported.
1809 WinCE is now supported. See L<perlce>.
1813 z/OS (formerly known as OS/390, formerly known as MVS OE) has now
1814 support for dynamic loading. This is not selected by default,
1815 however, you must specify -Dusedl in the arguments of Configure.
1819 =head1 Selected Bug Fixes
1821 Numerous memory leaks and uninitialized memory accesses have been
1822 hunted down. Most importantly anonymous subs used to leak quite
1829 The autouse pragma didn't work for Multi::Part::Function::Names.
1833 caller() could cause core dumps in certain situations. Carp was sometimes
1834 affected by this problem.
1838 chop(@list) in list context returned the characters chopped in
1839 reverse order. This has been reversed to be in the right order.
1843 Configure no longer includes the DBM libraries (dbm, gdbm, db, ndbm)
1844 when building the Perl binary. The only exception to this is SunOS 4.x,
1849 The behaviour of non-decimal but numeric string constants such as
1850 "0x23" was platform-dependent: in some platforms that was seen as 35,
1851 in some as 0, in some as a floating point number (don't ask). This
1852 was caused by Perl using the operating system libraries in a situation
1853 where the result of the string to number conversion is undefined: now
1854 Perl consistently handles such strings as zero in numeric contexts.
1858 The order of DESTROYs has been made more predictable.
1862 Several debugger fixes: exit code now reflects the script exit code,
1863 condition C<"0"> now treated correctly, the C<d> command now checks
1864 line number, the C<$.> no longer gets corrupted, all debugger output
1865 now goes correctly to the socket if RemotePort is set.
1869 Perl 5.6.0 could emit spurious warnings about redefinition of dl_error()
1870 when statically building extensions into perl. This has been corrected.
1874 L<dprofpp> -R didn't work.
1878 C<*foo{FORMAT}> now works.
1881 Infinity is now recognized as a number.
1885 UNIVERSAL::isa no longer caches methods incorrectly. (This broke
1886 the Tk extension with 5.6.0.)
1890 Lexicals I: lexicals outside an eval "" weren't resolved
1891 correctly inside a subroutine definition inside the eval "" if they
1892 were not already referenced in the top level of the eval""ed code.
1896 Lexicals II: lexicals leaked at file scope into subroutines that
1897 were declared before the lexicals.
1901 Lexical warnings now propagating correctly between scopes
1902 and into C<eval "...">.
1906 C<use warnings qw(FATAL all)> did not work as intended. This has been
1911 warnings::enabled() now reports the state of $^W correctly if the caller
1912 isn't using lexical warnings.
1916 Line renumbering with eval and C<#line> now works.
1920 Fixed numerous memory leaks, especially in eval "".
1924 mkdir() now ignores trailing slashes in the directory name,
1925 as mandated by POSIX.
1929 Some versions of glibc have a broken modfl(). This affects builds
1930 with C<-Duselongdouble>. This version of Perl detects this brokenness
1931 and has a workaround for it. The glibc release 2.2.2 is known to have
1932 fixed the modfl() bug.
1936 Modulus of unsigned numbers now works (4063328477 % 65535 used to
1937 return 27406, instead of 27047).
1941 Some "not a number" warnings introduced in 5.6.0 eliminated to be
1942 more compatible with 5.005. Infinity is now recognised as a number.
1946 Numeric conversions did not recognize changes in the string value
1947 properly in certain circumstances.
1951 Attributes (like :shared) didn't work with our().
1955 our() variables will not cause "will not stay shared" warnings.
1959 "our" variables of the same name declared in two sibling blocks
1960 resulted in bogus warnings about "redeclaration" of the variables.
1961 The problem has been corrected.
1965 pack "Z" now correctly terminates the string with "\0".
1969 Fix password routines which in some shadow password platforms
1970 (e.g. HP-UX) caused getpwent() to return every other entry.
1974 The PERL5OPT environment variable (for passing command line arguments
1975 to Perl) didn't work for more than a single group of options.
1979 PERL5OPT with embedded spaces didn't work.
1983 printf() no longer resets the numeric locale to "C".
1987 C<qw(a\\b)> now parses correctly as C<'a\\b'>.
1991 pos() did not return the correct value within s///ge in earlier
1992 versions. This is now handled correctly.
1996 Printing quads (64-bit integers) with printf/sprintf now works
1997 without the q L ll prefixes (assuming you are on a quad-capable platform).
2001 Regular expressions on references and overloaded scalars now work.
2005 Right-hand side magic (GMAGIC) could in many cases such as string
2006 concatenation be invoked too many times.
2010 scalar() now forces scalar context even when used in void context.
2014 SOCKS support is now much more robust.
2018 sort() arguments are now compiled in the right wantarray context
2019 (they were accidentally using the context of the sort() itself).
2020 The comparison block is now run in scalar context, and the arguments
2021 to be sorted are always provided list context.
2025 Changed the POSIX character class C<[[:space:]]> to include the (very
2026 rarely used) vertical tab character. Added a new POSIX-ish character
2027 class C<[[:blank:]]> which stands for horizontal whitespace
2028 (currently, the space and the tab).
2032 The tainting behaviour of sprintf() has been rationalized. It does
2033 not taint the result of floating point formats anymore, making the
2034 behaviour consistent with that of string interpolation.
2038 Some cases of inconsistent taint propagation (such as within hash
2039 values) have been fixed.
2043 The RE engine found in Perl 5.6.0 accidentally pessimised certain kinds
2044 of simple pattern matches. These are now handled better.
2048 Regular expression debug output (whether through C<use re 'debug'>
2049 or via C<-Dr>) now looks better.
2053 Multi-line matches like C<"a\nxb\n" =~ /(?!\A)x/m> were flawed. The
2058 Use of $& could trigger a core dump under some situations. This
2063 The regular expression captured submatches ($1, $2, ...) are now
2064 more consistently unset if the match fails, instead of leaving false
2065 data lying around in them.
2069 readline() on files opened in "slurp" mode could return an extra "" at
2070 the end in certain situations. This has been corrected.
2074 Autovivification of symbolic references of special variables described
2075 in L<perlvar> (as in C<${$num}>) was accidentally disabled. This works
2080 Sys::Syslog ignored the C<LOG_AUTH> constant.
2084 All but the first argument of the IO syswrite() method are now optional.
2088 $AUTOLOAD, sort(), lock(), and spawning subprocesses
2089 in multiple threads simultaneously are now thread-safe.
2093 Tie::ARRAY SPLICE method was broken.
2097 Allow read-only string on left hand side of non-modifying tr///.
2101 If C<STDERR> is tied, warnings caused by C<warn> and C<die> now
2102 correctly pass to it.
2106 Several Unicode fixes.
2112 BOMs (byte order marks) in the beginning of Perl files
2113 (scripts, modules) should now be transparently skipped.
2114 UTF-16 (UCS-2) encoded Perl files should now be read correctly.
2118 The character tables have been updated to Unicode 3.2.0.
2122 Comparing with utf8 data does not magically upgrade non-utf8 data
2123 into utf8. (This was a problem for example if you were mixing data
2124 from I/O and Unicode data: your output might have got magically encoded
2129 Generating illegal Unicode code points like U+FFFE, or the UTF-16
2130 surrogates, now also generates an optional warning.
2134 C<IsAlnum>, C<IsAlpha>, and C<IsWord> now match titlecase.
2138 Concatenation with the C<.> operator or via variable interpolation,
2139 C<eq>, C<substr>, C<reverse>, C<quotemeta>, the C<x> operator,
2140 substitution with C<s///>, single-quoted UTF8, should now work.
2144 The C<tr///> operator now works. Note that the C<tr///CU>
2145 functionality has been removed (but see pack('U0', ...)).
2149 C<eval "v200"> now works.
2153 Perl 5.6.0 parsed m/\x{ab}/ incorrectly, leading to spurious warnings.
2154 This has been corrected.
2158 Zero entries were missing from the Unicode classes like C<IsDigit>.
2164 Large unsigned numbers (those above 2**31) could sometimes lose their
2165 unsignedness, causing bogus results in arithmetic operations.
2169 =head2 Platform Specific Changes and Fixes
2177 Perl now works on post-4.0 BSD/OSes.
2183 Setting C<$0> now works (as much as possible; see L<perlvar> for details).
2189 Numerous updates; currently synchronised with Cygwin 1.3.10.
2193 Previously DYNIX/ptx had problems in its Configure probe for non-blocking I/O.
2199 EPOC update after Perl 5.6.0. See README.epoc.
2205 Perl now works on post-3.0 FreeBSDs.
2211 README.hpux updated; C<Configure -Duse64bitall> now works.
2217 Numerous compilation flag and hint enhancements; accidental mixing
2218 of 32-bit and 64-bit libraries (a doomed attempt) made much harder.
2228 Long doubles should now work (see INSTALL).
2232 Linux previously had problems related to sockaddrlen when using
2233 accept(), revcfrom() (in Perl: recv()), getpeername(), and getsockname().
2241 Compilation of the standard Perl distribution in MacOS Classic should
2242 now work if you have the Metrowerks development environment and
2243 the missing Mac-specific toolkit bits. Contact the macperl mailing
2250 MPE/iX update after Perl 5.6.0. See README.mpeix.
2256 Perl now works on NetBSD/sparc.
2262 Now works with usethreads (see INSTALL).
2268 64-bitness using the Sun Workshop compiler now works.
2272 Tru64 (aka Digital UNIX, aka DEC OSF/1)
2274 The operating system version letter now recorded in $Config{osvers}.
2275 Allow compiling with gcc (previously explicitly forbidden). Compiling
2276 with gcc still not recommended because buggy code results, even with
2283 Fixed various alignment problems that lead into core dumps either
2284 during build or later; no longer dies on math errors at runtime;
2285 now using full quad integers (64 bits), previously was using
2286 only 46 bit integers for speed.
2292 chdir() now works better despite a CRT bug; now works with MULTIPLICITY
2293 (see INSTALL); now works with Perl's malloc.
2295 The tainting of C<%ENV> elements via C<keys> or C<values> was previously
2296 unimplemented. It now works as documented.
2298 The C<waitpid> emulation has been improved. The worst bug (now fixed)
2299 was that a pid of -1 would cause a wildcard search of all processes on
2300 the system. The most significant enhancement is that we can now
2301 usually get the completion status of a terminated process.
2303 POSIX-style signals are now emulated much better on VMS versions prior
2306 The C<system> function and backticks operator have improved
2307 functionality and better error handling.
2309 File access tests now use current process privileges rather than the
2310 user's default privileges, which could sometimes result in a mismatch
2311 between reported access and actual access.
2321 accept() no longer leaks memory.
2325 Borland C++ v5.5 is now a supported compiler that can build Perl.
2326 However, the generated binaries continue to be incompatible with those
2327 generated by the other supported compilers (GCC and Visual C++).
2331 Better chdir() return value for a non-existent directory.
2335 Duping socket handles with open(F, ">&MYSOCK") now works under Windows 9x.
2339 New %ENV entries now propagate to subprocesses.
2343 Current directory entries in %ENV are now correctly propagated to child
2348 $ENV{LIB} now used to search for libs under Visual C.
2352 fork() emulation has been improved in various ways, but still continues
2353 to be experimental. See L<perlfork> for known bugs and caveats.
2357 A failed (pseudo)fork now returns undef and sets errno to EAGAIN.
2361 Win32::GetCwd() correctly returns C:\ instead of C: when at the drive root.
2362 Other bugs in chdir() and Cwd::cwd() have also been fixed.
2366 HTML files will be installed in c:\perl\html instead of c:\perl\lib\pod\html
2370 The makefiles now provide a single switch to bulk-enable all the features
2371 enabled in ActiveState ActivePerl (a popular Win32 binary distribution).
2375 Allow REG_EXPAND_SZ keys in the registry.
2379 Can now send() from all threads, not just the first one.
2383 Fake signal handling reenabled, bugs and all.
2387 %SIG has been enabled under USE_ITHREADS, but its use is completely
2388 unsupported under all configurations.
2392 Less stack reserved per thread so that more threads can run
2393 concurrently. (Still 16M per thread.)
2397 C<File::Spec->tmpdir()> now prefers C:/temp over /tmp
2398 (works better when perl is running as service).
2402 Better UNC path handling under ithreads.
2406 wait(), waitpid() and backticks now return the correct exit status under
2411 winsock handle leak fixed.
2415 The Perl parser has been stress tested using both random input and
2416 Markov chain input and the few found crashes and lockups have been
2423 =head1 New or Changed Diagnostics
2429 The lexical warnings category "deprecated" is no longer a sub-category
2430 of the "syntax" category. It is now a top-level category in its own
2435 All regular expression compilation error messages are now hopefully
2436 easier to understand both because the error message now comes before
2437 the failed regex and because the point of failure is now clearly
2438 marked by a C<E<lt>-- HERE> marker.
2442 The various "opened only for", "on closed", "never opened" warnings
2443 drop the C<main::> prefix for filehandles in the C<main> package,
2444 for example C<STDIN> instead of C<main::STDIN>.
2448 The "Unrecognized escape" warning has been extended to include C<\8>,
2449 C<\9>, and C<\_>. There is no need to escape any of the C<\w> characters.
2453 Two new debugging options have been added: if you have compiled your
2454 Perl with debugging, you can use the -DT and -DR options to trace
2455 tokenising and to add reference counts to displaying variables,
2460 The debugger (perl5db.pl) has been modified to present a more
2461 consistent commands interface, via (CommandSet=580). perl5db.t was
2462 also added to test the changes, and as a placeholder for further tests.
2468 The debugger has a new C<dumpDepth> option to control the maximum
2469 depth to which nested structures are dumped. The C<x> command has
2470 been extended so that C<x N EXPR> dumps out the value of I<EXPR> to a
2471 depth of at most I<N> levels.
2475 The debugger can now show lexical variables if you have the CPAN
2476 module PadWalker installed.
2480 If an attempt to use a (non-blessed) reference as an array index
2481 is made, a warning is given.
2485 C<push @a;> and C<unshift @a;> (with no values to push or unshift)
2486 now give a warning. This may be a problem for generated and evaled
2491 If you try to L<perlfunc/pack> a number less than 0 or larger than 255
2492 using the C<"C"> format you will get an optional warning. Similarly
2493 for the C<"c"> format and a number less than -128 or more than 127.
2497 Certain regex modifiers such as C<(?o)> make sense only if applied to
2498 the entire regex. You will get an optional warning if you try to do
2503 Using arrays or hashes as references (e.g. C<< %foo->{bar} >>
2504 has been deprecated for a while. Now you will get an optional warning.
2508 Using C<sort> in scalar context now issues an optional warning.
2509 This didn't do anything useful, as the sort was not performed.
2513 =head1 Changed Internals
2519 perlapi.pod (a companion to perlguts) now attempts to document the
2524 You can now build a really minimal perl called microperl.
2525 Building microperl does not require even running Configure;
2526 C<make -f Makefile.micro> should be enough. Beware: microperl makes
2527 many assumptions, some of which may be too bold; the resulting
2528 executable may crash or otherwise misbehave in wondrous ways.
2529 For careful hackers only.
2533 Added rsignal(), whichsig(), do_join(), op_clear, op_null,
2534 ptr_table_clear(), ptr_table_free(), sv_setref_uv(), and several UTF-8
2535 interfaces to the publicised API. For the full list of the available
2536 APIs see L<perlapi>.
2540 Made possible to propagate customised exceptions via croak()ing.
2544 Now xsubs can have attributes just like subs. (Well, at least the
2545 built-in attributes.)
2549 dTHR and djSP have been obsoleted; the former removed (because it's
2550 a no-op) and the latter replaced with dSP.
2554 PERL_OBJECT has been completely removed.
2558 The MAGIC constants (e.g. C<'P'>) have been macrofied
2559 (e.g. C<PERL_MAGIC_TIED>) for better source code readability
2560 and maintainability.
2564 The regex compiler now maintains a structure that identifies nodes in
2565 the compiled bytecode with the corresponding syntactic features of the
2566 original regex expression. The information is attached to the new
2567 C<offsets> member of the C<struct regexp>. See L<perldebguts> for more
2568 complete information.
2572 The C code has been made much more C<gcc -Wall> clean. Some warning
2573 messages still remain in some platforms, so if you are compiling with
2574 gcc you may see some warnings about dubious practices. The warnings
2575 are being worked on.
2579 F<perly.c>, F<sv.c>, and F<sv.h> have now been extensively commented.
2583 Documentation on how to use the Perl source repository has been added
2584 to F<Porting/repository.pod>.
2588 There are now several profiling make targets.
2592 =head1 Security Vulnerability Closed
2594 (This change was already made in 5.7.0 but bears repeating here.)
2596 A potential security vulnerability in the optional suidperl component
2597 of Perl was identified in August 2000. suidperl is neither built nor
2598 installed by default. As of November 2001 the only known vulnerable
2599 platform is Linux, most likely all Linux distributions. CERT and
2600 various vendors and distributors have been alerted about the vulnerability.
2601 See http://www.cpan.org/src/5.0/sperl-2000-08-05/sperl-2000-08-05.txt
2602 for more information.
2604 The problem was caused by Perl trying to report a suspected security
2605 exploit attempt using an external program, /bin/mail. On Linux
2606 platforms the /bin/mail program had an undocumented feature which
2607 when combined with suidperl gave access to a root shell, resulting in
2608 a serious compromise instead of reporting the exploit attempt. If you
2609 don't have /bin/mail, or if you have 'safe setuid scripts', or if
2610 suidperl is not installed, you are safe.
2612 The exploit attempt reporting feature has been completely removed from
2613 Perl 5.8.0 (and the maintenance release 5.6.1, and it was removed also
2614 from all the Perl 5.7 releases), so that particular vulnerability
2615 isn't there anymore. However, further security vulnerabilities are,
2616 unfortunately, always possible. The suidperl functionality is most
2617 probably going to be removed in Perl 5.10. In any case, suidperl
2618 should only be used by security experts who know exactly what they are
2619 doing and why they are using suidperl instead of some other solution
2620 such as sudo ( see http://www.courtesan.com/sudo/ ).
2624 Several new tests have been added, especially for the F<lib>
2625 subsection. There are now about 56 000 individual tests (spread over
2626 about 620 test scripts), in the regression suite (5.6.1 has about
2627 11700 tests, in 258 test scripts) Many of the new tests are introduced
2628 by the new modules, but still in general Perl is now more thoroughly
2631 Because of the large number of tests, running the regression suite
2632 will take considerably longer time than it used to: expect the suite
2633 to take up to 4-5 times longer to run than in perl 5.6. In a really
2634 fast machine you can hope to finish the suite in about 6-8 minutes
2637 The tests are now reported in a different order than in earlier Perls.
2638 (This happens because the test scripts from under t/lib have been moved
2639 to be closer to the library/extension they are testing.)
2641 =head1 Known Problems
2649 In AIX 4.2 Perl extensions that use C++ functions that use statics
2650 may have problems in that the statics are not getting initialized.
2651 In newer AIX releases this has been solved by linking Perl with
2652 the libC_r library, but unfortunately in AIX 4.2 the said library
2653 has an obscure bug where the various functions related to time
2654 (such as time() and gettimeofday()) return broken values, and
2655 therefore in AIX 4.2 Perl is not linked against the libC_r.
2659 vac 5.0.0.0 May Produce Buggy Code For Perl
2661 The AIX C compiler vac version 5.0.0.0 may produce buggy code,
2662 resulting in few random tests failing, but when the failing tests
2663 are run by hand, they succeed. We suggest upgrading to at least
2664 vac version 5.0.1.0, that has been known to compile Perl correctly.
2665 "lslpp -L|grep vac.C" will tell you the vac version. See README.aix.
2669 If building threaded Perl, you may get compilation warning from pp_sys.c:
2671 "pp_sys.c", line 4651.39: 1506-280 (W) Function argument assignment between types "unsigned char*" and "const void*" is not allowed.
2673 This is harmless; it is caused by the getnetbyaddr() and getnetbyaddr_r()
2674 having slightly different types for their first argument.
2678 =head2 Amiga Perl Invoking Mystery
2680 One cannot call Perl using the C<volume:> syntax, that is, C<perl -v>
2681 works, but for example C<bin:perl -v> doesn't. The exact reason isn't
2682 known but the current suspect is the F<ixemul> library.
2684 =head2 FreeBSD Failing locale Test 117 For ISO8859-15 Locales
2686 The ISO8859-15 locales may fail the locale test 117 in FreeBSD.
2687 This is caused by the characters \xFF (y with diaeresis) and \xBE
2688 (Y with diaeresis) not behaving correctly when being matched
2691 =head2 mod_perl 1.26 Doesn't Build With Threaded Perl
2693 Use mod_perl 1.27 or higher.
2695 =head2 lib/ftmp-security tests warn 'system possibly insecure'
2697 Don't panic. Read INSTALL 'make test' section instead.
2699 =head2 HP-UX lib/posix Subtest 9 Fails When LP64-Configured
2701 If perl is configured with -Duse64bitall, the successful result of the
2702 subtest 10 of lib/posix may arrive before the successful result of the
2703 subtest 9, which confuses the test harness so much that it thinks the
2706 =head2 Linux With Sfio Fails op/misc Test 48
2712 Please remember to set your environment variable LC_ALL to "C"
2713 (setenv LC_ALL C) before running "make test" to avoid a lot of
2714 warnings about the broken locales of Mac OS X.
2716 The following tests are known to fail:
2718 Failed Test Stat Wstat Total Fail Failed List of Failed
2719 -------------------------------------------------------------------------
2720 ../ext/DB_File/t/db-btree.t 0 11 ?? ?? % ??
2721 ../ext/DB_File/t/db-recno.t 149 3 2.01% 61 63 65
2722 ../ext/POSIX/t/posix.t 31 1 3.23% 10
2724 If you are building on a UFS partition, you will also probably see
2725 t/op/stat.t subtest #9 fail. This is caused by Darwin's UFS not
2726 supporting inode change time.
2728 =head2 op/sprintf tests 91, 129, and 130
2730 The op/sprintf tests 91, 129, and 130 are known to fail on some platforms.
2731 Examples include any platform using sfio, and Compaq/Tandem's NonStop-UX.
2733 The test 91 is known to fail at QNX6 (nto), because C<sprintf '%e',0>
2734 incorrectly produces C<0.000000e+0> instead of C<0.000000e+00>.
2736 For the tests 129 and 130 the failing platforms do not comply with
2737 the ANSI C Standard, line 19ff on page 134 of ANSI X3.159 1989 to
2738 be exact. (They produce something other than "1" and "-1" when
2739 formatting 0.6 and -0.6 using the printf format "%.0f", most often
2740 they produce "0" and "-0".)
2742 =head2 Failure of Thread (5.005-style) tests
2744 B<Note that support for 5.005-style threading remains experimental
2745 and practically unsupported.>
2747 The following tests are known to fail due to fundamental problems in
2748 the 5.005 threading implementation. These are not new failures--Perl
2749 5.005_0x has the same bugs, but didn't have these tests.
2751 ../ext/List/Util/t/first.t 255 65280 7 4 57.14% 2 5-7
2752 ../lib/English.t 2 512 54 2 3.70% 2-3
2753 ../lib/Filter/Simple/t/data.t 6 3 50.00% 1-3
2754 ../lib/Filter/Simple/t/filter_onl 9 3 33.33% 1-2 5
2755 ../lib/autouse.t 10 1 10.00% 4
2756 op/flip.t 15 1 6.67% 15
2758 These failures are unlikely to get fixed as the 5.005-style
2759 threads are considered fundamentally broken.
2763 Failed Test Stat Wstat Total Fail Failed List of Failed
2764 -------------------------------------------------------------------------
2765 ../ext/Socket/socketpair.t 1 256 45 1 2.22% 12
2766 ../lib/Math/Trig.t 26 1 3.85% 25
2767 ../lib/warnings.t 460 1 0.22% 425
2768 io/fs.t 36 1 2.78% 31
2769 op/numconvert.t 1440 13 0.90% 208 509-510
2770 657-658 665-666 829-830 989-990 1149-1150
2772 =head2 UNICOS and UNICOS/mk
2774 The io/fs test #31 is failing because in UNICOS and UNICOS/mk the Perl
2775 truncate() cannot be used to grow the size of filehandles, only to
2776 reduce the size. The workaround is to truncate files instead of
2785 During Configure the test
2787 Guessing which symbols your C compiler and preprocessor define...
2789 will probably fail with error messages like
2791 CC-20 cc: ERROR File = try.c, Line = 3
2792 The identifier "bad" is undefined.
2794 bad switch yylook 79bad switch yylook 79bad switch yylook 79bad switch yylook 79#ifdef A29K
2797 CC-65 cc: ERROR File = try.c, Line = 3
2798 A semicolon is expected at this point.
2800 This is caused by a bug in awk utility of UNICOS/mk. You can ignore
2801 the error, but it does cause a slight problem: you cannot fully
2802 benefit from the h2ph utility (see L<h2ph>) that can be used to
2803 convert C headers to Perl libraries, mainly used to be able to access
2804 from Perl the constants defined using C preprocessor, cpp. Because of
2805 the above error parts of the converted headers will be invisible.
2806 Luckily, these days the need for h2ph is rare.
2810 If building Perl with the interpreter threads (ithreads), the
2811 getgrent(), getgrnam(), and getgrgid() functions cannot return the
2812 list of the group members due to a bug in the multithreaded support of
2813 UNICOS/mk. What this means that in list context the functions will
2814 return only three values, not four.
2820 There are a few known test failures, see L<perluts>.
2824 There should be no reported test failures with a default configuration,
2825 though there are a number of tests marked TODO that point to areas
2826 needing further debugging and/or porting work.
2830 In multi-CPU boxes there are some problems with the I/O buffering:
2831 some output may appear twice.
2833 =head2 XML::Parser not working
2835 Use XML::Parser 2.31 or later.
2837 =head2 Localising a Tied Variable Leaks Memory
2840 tie my %tie_hash => 'Tie::StdHash';
2844 local($tie_hash{Foo}) = 1; # leaks
2846 Code like the above is known to leak memory every time the local()
2849 =head2 z/OS (OS/390)
2851 z/OS has rather many test failures but the situation is actually
2852 better than it was in 5.6.0, it's just that so many new modules and
2853 tests have been added.
2855 Failed Test Stat Wstat Total Fail Failed List of Failed
2856 -------------------------------------------------------------------------------
2857 ../ext/Data/Dumper/t/dumper.t 339 8 2.36% 311 314 325 327
2859 ../ext/IO/lib/IO/t/io_unix.t 5 4 80.00% 2-5
2860 ../ext/Storable/t/utf8hash.t 10 2560 148 10 6.76% 1 5 72 76 143-148
2861 ../lib/utf8.t 94 13 13.83% 27 30-31 43 46 73
2864 ../lib/ExtUtils/t/Constant.t 121 30976 48 48 100.00% 1-48
2865 ../lib/ExtUtils/t/Embed.t 9 9 100.00% 1-9
2866 op/pat.t 900 9 1.00% 242-243 665 776
2868 op/sprintf.t 224 3 1.34% 98 100 136
2869 op/tr.t 97 5 5.15% 63 71-74
2870 run/fresh_perl.t 94 3 3.19% 92-94
2871 uni/fold.t 780 6 0.77% 61 169 196 661
2874 =head2 Localising Tied Arrays and Hashes Is Broken
2878 doesn't work as one would expect: the old value is restored
2881 =head2 Self-tying of Arrays and Hashes Is Forbidden
2883 Self-tying of arrays and hashes is broken in rather deep and
2884 hard-to-fix ways. As a stop-gap measure to avoid people from getting
2885 frustrated at the mysterious results (core dumps, most often) it is
2886 for now forbidden (you will get a fatal error even from an attempt).
2888 =head2 Building Extensions Can Fail Because Of Largefiles
2890 Some extensions like mod_perl are known to have issues with
2891 `largefiles', a change brought by Perl 5.6.0 in which file offsets
2892 default to 64 bits wide, where supported. Modules may fail to compile
2893 at all or compile and work incorrectly. Currently there is no good
2894 solution for the problem, but Configure now provides appropriate
2895 non-largefile ccflags, ldflags, libswanted, and libs in the %Config
2896 hash (e.g., $Config{ccflags_nolargefiles}) so the extensions that are
2897 having problems can try configuring themselves without the
2898 largefileness. This is admittedly not a clean solution, and the
2899 solution may not even work at all. One potential failure is whether
2900 one can (or, if one can, whether it's a good idea) link together at
2901 all binaries with different ideas about file offsets, all this is
2904 =head2 Unicode Support on EBCDIC Still Spotty
2906 Though mostly working, Unicode support still has problem spots on
2907 EBCDIC platforms. One such known spot are the C<\p{}> and C<\P{}>
2908 regular expression constructs for code points less than 256: the
2909 pP are testing for Unicode code points, not knowing about EBCDIC.
2911 =head2 The Compiler Suite Is Still Experimental
2913 The compiler suite is slowly getting better but it continues to be
2914 highly experimental. Use in production environments is discouraged.
2916 =head2 The Long Double Support Is Still Experimental
2918 The ability to configure Perl's numbers to use "long doubles",
2919 floating point numbers of hopefully better accuracy, is still
2920 experimental. The implementations of long doubles are not yet
2921 widespread and the existing implementations are not quite mature
2922 or standardised, therefore trying to support them is a rare
2923 and moving target. The gain of more precision may also be offset
2924 by slowdown in computations (more bits to move around, and the
2925 operations are more likely to be executed by less optimised
2928 =head2 Seen In Perl 5.7 But Gone Now
2930 C<Time::Piece> (previously known as C<Time::Object>) was removed
2931 because it was felt that it didn't have enough value in it to be a
2932 core module. It is still a useful module, though, and is available
2935 =head1 Reporting Bugs
2937 If you find what you think is a bug, you might check the articles
2938 recently posted to the comp.lang.perl.misc newsgroup and the perl
2939 bug database at http://bugs.perl.org/ There may also be
2940 information at http://www.perl.com/ , the Perl Home Page.
2942 If you believe you have an unreported bug, please run the B<perlbug>
2943 program included with your release. Be sure to trim your bug down
2944 to a tiny but sufficient test case. Your bug report, along with the
2945 output of C<perl -V>, will be sent off to perlbug@perl.org to be
2946 analysed by the Perl porting team.
2950 The F<Changes> file for exhaustive details on what changed.
2952 The F<INSTALL> file for how to build Perl.
2954 The F<README> file for general stuff.
2956 The F<Artistic> and F<Copying> files for copyright information.
2960 Written by Jarkko Hietaniemi <F<jhi@iki.fi>>.