PERL58DELTA(category21-suse.html) - phpMan

PERL58DELTA(1)         Perl Programmers Reference Guide         PERL58DELTA(1)

NAME
       perl58delta - what is new for perl v5.8.0
DESCRIPTION
       This document describes differences between the 5.6.0 release and the
       5.8.0 release.
       Many of the bug fixes in 5.8.0 were already seen in the 5.6.1
       maintenance release since the two releases were kept closely
       coordinated (while 5.8.0 was still called 5.7.something).
       Changes that were integrated into the 5.6.1 release are marked "[561]".
       Many of these changes have been further developed since 5.6.1 was
       released, those are marked "[561+]".
       You can see the list of changes in the 5.6.1 release (both from the
       5.005_03 release and the 5.6.0 release) by reading perl561delta.
Highlights In 5.8.0
       o   Better Unicode support
       o   New IO Implementation
       o   New Thread Implementation
       o   Better Numeric Accuracy
       o   Safe Signals
       o   Many New Modules
       o   More Extensive Regression Testing
Incompatible Changes
   Binary Incompatibility
       Perl 5.8 is not binary compatible with earlier releases of Perl.
       You have to recompile your XS modules.
       (Pure Perl modules should continue to work.)
       The major reason for the discontinuity is the new IO architecture
       called PerlIO.  PerlIO is the default configuration because without it
       many new features of Perl 5.8 cannot be used.  In other words: you just
       have to recompile your modules containing XS code, sorry about that.
       In future releases of Perl, non-PerlIO aware XS modules may become
       completely unsupported.  This shouldn't be too difficult for module
       authors, however: PerlIO has been designed as a drop-in replacement (at
       the source code level) for the stdio interface.
       Depending on your platform, there are also other reasons why we decided
       to break binary compatibility, please read on.
   64-bit platforms and malloc
       If your pointers are 64 bits wide, the Perl malloc is no longer being
       used because it does not work well with 8-byte pointers.  Also, usually
       the system mallocs on such platforms are much better optimized for such
       large memory models than the Perl malloc.  Some memory-hungry Perl
       applications like the PDL don't work well with Perl's malloc.  Finally,
       other applications than Perl (such as mod_perl) tend to prefer the
       system malloc.  Such platforms include Alpha and 64-bit HPPA, MIPS,
       PPC, and Sparc.
   AIX Dynaloading
       The AIX dynaloading now uses in AIX releases 4.3 and newer the native
       dlopen interface of AIX instead of the old emulated interface.  This
       change will probably break backward compatibility with compiled
       modules.  The change was made to make Perl more compliant with other
       applications like mod_perl which are using the AIX native interface.
   Attributes for "my" variables now handled at run-time
       The "my EXPR : ATTRS" syntax now applies variable attributes at run-
       time.  (Subroutine and "our" variables still get attributes applied at
       compile-time.)  See attributes for additional details.  In particular,
       however, this allows variable attributes to be useful for "tie"
       interfaces, which was a deficiency of earlier releases.  Note that the
       new semantics doesn't work with the Attribute::Handlers module (as of
       version 0.76).
   Socket Extension Dynamic in VMS
       The Socket extension is now dynamically loaded instead of being
       statically built in.  This may or may not be a problem with ancient
       TCP/IP stacks of VMS: we do not know since we weren't able to test Perl
       in such configurations.
   IEEE-format Floating Point Default on OpenVMS Alpha
       Perl now uses IEEE format (T_FLOAT) as the default internal floating
       point format on OpenVMS Alpha, potentially breaking binary
       compatibility with external libraries or existing data.  G_FLOAT is
       still available as a configuration option.  The default on VAX
       (D_FLOAT) has not changed.
   New Unicode Semantics (no more "use utf8", almost)
       Previously in Perl 5.6 to use Unicode one would say "use utf8" and then
       the operations (like string concatenation) were Unicode-aware in that
       lexical scope.
       This was found to be an inconvenient interface, and in Perl 5.8 the
       Unicode model has completely changed: now the "Unicodeness" is bound to
       the data itself, and for most of the time "use utf8" is not needed at
       all.  The only remaining use of "use utf8" is when the Perl script
       itself has been written in the UTF-8 encoding of Unicode.  (UTF-8 has
       not been made the default since there are many Perl scripts out there
       that are using various national eight-bit character sets, which would
       be illegal in UTF-8.)
       See perluniintro for the explanation of the current model, and utf8 for
       the current use of the utf8 pragma.
   New Unicode Properties
       Unicode scripts are now supported. Scripts are similar to (and superior
       to) Unicode blocks. The difference between scripts and blocks is that
       scripts are the glyphs used by a language or a group of languages,
       while the blocks are more artificial groupings of (mostly) 256
       characters based on the Unicode numbering.
       In general, scripts are more inclusive, but not universally so. For
       example, while the script "Latin" includes all the Latin characters and
       their various diacritic-adorned versions, it does not include the
       various punctuation or digits (since they are not solely "Latin").
       A number of other properties are now supported, including "\p{L&}",
       "\p{Any}" "\p{Assigned}", "\p{Unassigned}", "\p{Blank}" [561] and
       "\p{SpacePerl}" [561] (along with their "\P{...}" versions, of course).
       See perlunicode for details, and more additions.
       The "In" or "Is" prefix to names used with the "\p{...}" and "\P{...}"
       are now almost always optional. The only exception is that a "In"
       prefix is required to signify a Unicode block when a block name
       conflicts with a script name. For example, "\p{Tibetan}" refers to the
       script, while "\p{InTibetan}" refers to the block. When there is no
       name conflict, you can omit the "In" from the block name (e.g.
       "\p{BraillePatterns}"), but to be safe, it's probably best to always
       use the "In").
   REF(...) Instead Of SCALAR(...)
       A reference to a reference now stringifies as "REF(0x81485ec)" instead
       of "SCALAR(0x81485ec)" in order to be more consistent with the return
       value of ref().
   pack/unpack D/F recycled
       The undocumented pack/unpack template letters D/F have been recycled
       for better use: now they stand for long double (if supported by the
       platform) and NV (Perl internal floating point type).  (They used to be
       aliases for d/f, but you never knew that.)
   glob() now returns filenames in alphabetical order
       The list of filenames from glob() (or <...>) is now by default sorted
       alphabetically to be csh-compliant (which is what happened before in
       most Unix platforms).  (bsd_glob() does still sort platform natively,
       ASCII or EBCDIC, unless GLOB_ALPHASORT is specified.) [561]
   Deprecations
       o   The semantics of bless(REF, REF) were unclear and until someone
           proves it to make some sense, it is forbidden.
       o   The obsolete chat2 library that should never have been allowed to
           escape the laboratory has been decommissioned.
       o   Using chdir("") or chdir(undef) instead of explicit chdir() is
           doubtful.  A failure (think chdir(some_function()) can lead into
           unintended chdir() to the home directory, therefore this behaviour
           is deprecated.
       o   The builtin dump() function has probably outlived most of its
           usefulness.  The core-dumping functionality will remain in future
           available as an explicit call to "CORE::dump()", but in future
           releases the behaviour of an unqualified "dump()" call may change.
       o   The very dusty examples in the eg/ directory have been removed.
           Suggestions for new shiny examples welcome but the main issue is
           that the examples need to be documented, tested and (most
           importantly) maintained.
       o   The (bogus) escape sequences \8 and \9 now give an optional warning
           ("Unrecognized escape passed through").  There is no need to
           \-escape any "\w" character.
       o   The *glob{FILEHANDLE} is deprecated, use *glob{IO} instead.
       o   The "package;" syntax ("package" without an argument) has been
           deprecated.  Its semantics were never that clear and its
           implementation even less so.  If you have used that feature to
           disallow all but fully qualified variables, "use strict;" instead.
       o   The unimplemented POSIX regex features [[.cc.]] and [[=c=]] are
           still recognised but now cause fatal errors.  The previous
           behaviour of ignoring them by default and warning if requested was
           unacceptable since it, in a way, falsely promised that the features
           could be used.
       o   In future releases, non-PerlIO aware XS modules may become
           completely unsupported.  Since PerlIO is a drop-in replacement for
           stdio at the source code level, this shouldn't be that drastic a
           change.
       o   Previous versions of perl and some readings of some sections of
           Camel III implied that the ":raw" "discipline" was the inverse of
           ":crlf".  Turning off "clrfness" is no longer enough to make a
           stream truly binary. So the PerlIO ":raw" layer (or "discipline",
           to use the Camel book's older terminology) is now formally defined
           as being equivalent to binmode(FH) - which is in turn defined as
           doing whatever is necessary to pass each byte as-is without any
           translation.  In particular binmode(FH) - and hence ":raw" - will
           now turn off both CRLF and UTF-8 translation and remove other
           layers (e.g. :encoding()) which would modify byte stream.
       o   The current user-visible implementation of pseudo-hashes (the weird
           use of the first array element) is deprecated starting from Perl
           5.8.0 and will be removed in Perl 5.10.0, and the feature will be
           implemented differently.  Not only is the current interface rather
           ugly, but the current implementation slows down normal array and
           hash use quite noticeably. The "fields" pragma interface will
           remain available.  The restricted hashes interface is expected to
           be the replacement interface (see Hash::Util).  If your existing
           programs depends on the underlying implementation, consider using
           Class::PseudoHash from CPAN.
       o   The syntaxes "@a->[...]" and  "%h->{...}" have now been deprecated.
       o   After years of trying, suidperl is considered to be too complex to
           ever be considered truly secure.  The suidperl functionality is
           likely to be removed in a future release.
       o   The 5.005 threads model (module "Thread") is deprecated and
           expected to be removed in Perl 5.10.  Multithreaded code should be
           migrated to the new ithreads model (see threads, threads::shared
           and perlthrtut).
       o   The long deprecated uppercase aliases for the string comparison
           operators (EQ, NE, LT, LE, GE, GT) have now been removed.
       o   The tr///C and tr///U features have been removed and will not
           return; the interface was a mistake.  Sorry about that.  For
           similar functionality, see pack('U0', ...) and pack('C0', ...).
           [561]
       o   Earlier Perls treated "sub foo (@bar)" as equivalent to "sub foo
           (@)".  The prototypes are now checked better at compile-time for
           invalid syntax.  An optional warning is generated ("Illegal
           character in prototype...")  but this may be upgraded to a fatal
           error in a future release.
       o   The "exec LIST" and "system LIST" operations now produce warnings
           on tainted data and in some future release they will produce fatal
           errors.
       o   The existing behaviour when localising tied arrays and hashes is
           wrong, and will be changed in a future release, so do not rely on
           the existing behaviour. See "Localising Tied Arrays and Hashes Is
           Broken".
Core Enhancements
   Unicode Overhaul
       Unicode in general should be now much more usable than in Perl 5.6.0
       (or even in 5.6.1).  Unicode can be used in hash keys, Unicode in
       regular expressions should work now, Unicode in tr/// should work now,
       Unicode in I/O should work now.  See perluniintro for introduction and
       perlunicode for details.
       o   The Unicode Character Database coming with Perl has been upgraded
           to Unicode 3.2.0.  For more information, see
           http://www.unicode.org/ .  [561+] (5.6.1 has UCD 3.0.1.)
       o   For developers interested in enhancing Perl's Unicode capabilities:
           almost all the UCD files are included with the Perl distribution in
           the lib/unicore subdirectory.  The most notable omission, for space
           considerations, is the Unihan database.
       o   The properties \p{Blank} and \p{SpacePerl} have been added. "Blank"
           is like C isblank(), that is, it contains only "horizontal
           whitespace" (the space character is, the newline isn't), and the
           "SpacePerl" is the Unicode equivalent of "\s" (\p{Space} isn't,
           since that includes the vertical tabulator character, whereas "\s"
           doesn't.)
           See "New Unicode Properties" earlier in this document for
           additional information on changes with Unicode properties.
   PerlIO is Now The Default
       o   IO is now by default done via PerlIO rather than system's "stdio".
           PerlIO allows "layers" to be "pushed" onto a file handle to alter
           the handle's behaviour.  Layers can be specified at open time via
           3-arg form of open:
              open($fh,'>:crlf :utf8', $path) || ...
           or on already opened handles via extended "binmode":
              binmode($fh,':encoding(iso-8859-7)');
           The built-in layers are: unix (low level read/write), stdio (as in
           previous Perls), perlio (re-implementation of stdio buffering in a
           portable manner), crlf (does CRLF <=> "\n" translation as on Win32,
           but available on any platform).  A mmap layer may be available if
           platform supports it (mostly Unixes).
           Layers to be applied by default may be specified via the 'open'
           pragma.
           See "Installation and Configuration Improvements" for the effects
           of PerlIO on your architecture name.
       o   If your platform supports fork(), you can use the list form of
           "open" for pipes.  For example:
               open KID_PS, "-|", "ps", "aux" or die $!;
           forks the ps(1) command (without spawning a shell, as there are
           more than three arguments to open()), and reads its standard output
           via the "KID_PS" filehandle.  See perlipc.
       o   File handles can be marked as accepting Perl's internal encoding of
           Unicode (UTF-8 or UTF-EBCDIC depending on platform) by a pseudo
           layer ":utf8" :
              open($fh,">:utf8","Uni.txt");
           Note for EBCDIC users: the pseudo layer ":utf8" is erroneously
           named for you since it's not UTF-8 what you will be getting but
           instead UTF-EBCDIC.  See perlunicode, utf8, and
           http://www.unicode.org/unicode/reports/tr16/ for more information.
           In future releases this naming may change.  See perluniintro for
           more information about UTF-8.
       o   If your environment variables (LC_ALL, LC_CTYPE, LANG) look like
           you want to use UTF-8 (any of the variables match "/utf-?8/i"),
           your STDIN, STDOUT, STDERR handles and the default open layer (see
           open) are marked as UTF-8.  (This feature, like other new features
           that combine Unicode and I/O, work only if you are using PerlIO,
           but that's the default.)
           Note that after this Perl really does assume that everything is
           UTF-8: for example if some input handle is not, Perl will probably
           very soon complain about the input data like this "Malformed UTF-8
           ..." since any old eight-bit data is not legal UTF-8.
           Note for code authors: if you want to enable your users to use
           UTF-8 as their default encoding  but in your code still have eight-
           bit I/O streams (such as images or zip files), you need to
           explicitly open() or binmode() with ":bytes" (see "open" in
           perlfunc and "binmode" in perlfunc), or you can just use
           "binmode(FH)" (nice for pre-5.8.0 backward compatibility).
       o   File handles can translate character encodings from/to Perl's
           internal Unicode form on read/write via the ":encoding()" layer.
       o   File handles can be opened to "in memory" files held in Perl
           scalars via:
              open($fh,'>', \$variable) || ...
       o   Anonymous temporary files are available without need to 'use
           FileHandle' or other module via
              open($fh,"+>", undef) || ...
           That is a literal undef, not an undefined value.
   ithreads
       The new interpreter threads ("ithreads" for short) implementation of
       multithreading, by Arthur Bergman, replaces the old "5.005 threads"
       implementation.  In the ithreads model any data sharing between threads
       must be explicit, as opposed to the model where data sharing was
       implicit.  See threads and threads::shared, and perlthrtut.
       As a part of the ithreads implementation Perl will also use any
       necessary and detectable reentrant libc interfaces.
   Restricted Hashes
       A restricted hash is restricted to a certain set of keys, no keys
       outside the set can be added.  Also individual keys can be restricted
       so that the key cannot be deleted and the value cannot be changed.  No
       new syntax is involved: the Hash::Util module is the interface.
   Safe Signals
       Perl used to be fragile in that signals arriving at inopportune moments
       could corrupt Perl's internal state.  Now Perl postpones handling of
       signals until it's safe (between opcodes).
       This change may have surprising side effects because signals no longer
       interrupt Perl instantly.  Perl will now first finish whatever it was
       doing, like finishing an internal operation (like sort()) or an
       external operation (like an I/O operation), and only then look at any
       arrived signals (and before starting the next operation).  No more
       corrupt internal state since the current operation is always finished
       first, but the signal may take more time to get heard.  Note that
       breaking out from potentially blocking operations should still work,
       though.
   Understanding of Numbers
       In general a lot of fixing has happened in the area of Perl's
       understanding of numbers, both integer and floating point.  Since in
       many systems the standard number parsing functions like "strtoul()" and
       "atof()" seem to have bugs, Perl tries to work around their
       deficiencies.  This results hopefully in more accurate numbers.
       Perl now tries internally to use integer values in numeric conversions
       and basic arithmetics (+ - * /) if the arguments are integers, and
       tries also to keep the results stored internally as integers.  This
       change leads to often slightly faster and always less lossy
       arithmetics. (Previously Perl always preferred floating point numbers
       in its math.)
   Arrays now always interpolate into double-quoted strings [561]
       In double-quoted strings, arrays now interpolate, no matter what.  The
       behavior in earlier versions of perl 5 was that arrays would
       interpolate into strings if the array had been mentioned before the
       string was compiled, and otherwise Perl would raise a fatal compile-
       time error.  In versions 5.000 through 5.003, the error was
               Literal @example now requires backslash
       In versions 5.004_01 through 5.6.0, the error was
               In string, @example now must be written as \@example
       The idea here was to get people into the habit of writing
       "fred\@example.com" when they wanted a literal "@" sign, just as they
       have always written "Give me back my \$5" when they wanted a literal
       "$" sign.
       Starting with 5.6.1, when Perl now sees an "@" sign in a double-quoted
       string, it always attempts to interpolate an array, regardless of
       whether or not the array has been used or declared already.  The fatal
       error has been downgraded to an optional warning:
               Possible unintended interpolation of @example in string
       This warns you that "fred AT example.com" is going to turn into "fred.com"
       if you don't backslash the "@".  See
       http://perl.plover.com/at-error.html for more details about the history
       here.
   Miscellaneous Changes
       o   AUTOLOAD is now lvaluable, meaning that you can add the :lvalue
           attribute to AUTOLOAD subroutines and you can assign to the
           AUTOLOAD return value.
       o   The $Config{byteorder} (and corresponding BYTEORDER in config.h)
           was previously wrong in platforms if sizeof(long) was 4, but
           sizeof(IV) was 8.  The byteorder was only sizeof(long) bytes long
           (1234 or 4321), but now it is correctly sizeof(IV) bytes long,
           (12345678 or 87654321).  (This problem didn't affect Windows
           platforms.)
           Also, $Config{byteorder} is now computed dynamically--this is more
           robust with "fat binaries" where an executable image contains
           binaries for more than one binary platform, and when cross-
           compiling.
       o   "perl -d:Module=arg,arg,arg" now works (previously one couldn't
           pass in multiple arguments.)
       o   "do" followed by a bareword now ensures that this bareword isn't a
           keyword (to avoid a bug where "do q(foo.pl)" tried to call a
           subroutine called "q").  This means that for example instead of "do
           format()" you must write "do &format()".
       o   The builtin dump() now gives an optional warning "dump() better
           written as CORE::dump()", meaning that by default "dump(...)" is
           resolved as the builtin dump() which dumps core and aborts, not as
           (possibly) user-defined "sub dump".  To call the latter, qualify
           the call as "&dump(...)".  (The whole dump() feature is to
           considered deprecated, and possibly removed/changed in future
           releases.)
       o   chomp() and chop() are now overridable.  Note, however, that their
           prototype (as given by "prototype("CORE::chomp")" is undefined,
           because it cannot be expressed and therefore one cannot really
           write replacements to override these builtins.
       o   END blocks are now run even if you exit/die in a BEGIN block.
           Internally, the execution of END blocks is now controlled by
           PL_exit_flags & PERL_EXIT_DESTRUCT_END. This enables the new
           behaviour for Perl embedders. This will default in 5.10. See
           perlembed.
       o   Formats now support zero-padded decimal fields.
       o   Although "you shouldn't do that", it was possible to write code
           that depends on Perl's hashed key order (Data::Dumper does this).
           The new algorithm "One-at-a-Time" produces a different hashed key
           order.  More details are in "Performance Enhancements".
       o   lstat(FILEHANDLE) now gives a warning because the operation makes
           no sense.  In future releases this may become a fatal error.
       o   Spurious syntax errors generated in certain situations, when glob()
           caused File::Glob to be loaded for the first time, have been fixed.
           [561]
       o   Lvalue subroutines can now return "undef" in list context.
           However, the lvalue subroutine feature still remains experimental.
           [561+]
       o   A lost warning "Can't declare ... dereference in my" has been
           restored (Perl had it earlier but it became lost in later
           releases.)
       o   A new special regular expression variable has been introduced: $^N,
           which contains the most-recently closed group (submatch).
       o   "no Module;" does not produce an error even if Module does not have
           an unimport() method.  This parallels the behavior of "use" vis-a-
           vis "import". [561]
       o   The numerical comparison operators return "undef" if either operand
           is a NaN.  Previously the behaviour was unspecified.
       o   "our" can now have an experimental optional attribute "unique" that
           affects how global variables are shared among multiple
           interpreters, see "our" in perlfunc.
       o   The following builtin functions are now overridable: each(),
           keys(), pop(), push(), shift(), splice(), unshift(). [561]
       o   "pack() / unpack()" can now group template letters with "()" and
           then apply repetition/count modifiers on the groups.
       o   "pack() / unpack()" can now process the Perl internal numeric
           types: IVs, UVs, NVs-- and also long doubles, if supported by the
           platform.  The template letters are "j", "J", "F", and "D".
       o   "pack('U0a*', ...)" can now be used to force a string to UTF-8.
       o   my __PACKAGE__ $obj now works. [561]
       o   POSIX::sleep() now returns the number of unslept seconds (as the
           POSIX standard says), as opposed to CORE::sleep() which returns the
           number of slept seconds.
       o   printf() and sprintf() now support parameter reordering using the
           "%\d+\$" and "*\d+\$" syntaxes.  For example
               printf "%2\$s %1\$s\n", "foo", "bar";
           will print "bar foo\n".  This feature helps in writing
           internationalised software, and in general when the order of the
           parameters can vary.
       o   The (\&) prototype now works properly. [561]
       o   prototype(\[$@%&]) is now available to implicitly create references
           (useful for example if you want to emulate the tie() interface).
       o   A new command-line option, "-t" is available.  It is the little
           brother of "-T": instead of dying on taint violations, lexical
           warnings are given.  This is only meant as a temporary debugging
           aid while securing the code of old legacy applications.  This is
           not a substitute for -T.
       o   In other taint news, the "exec LIST" and "system LIST" have now
           been considered too risky (think "exec @ARGV": it can start any
           program with any arguments), and now the said forms cause a warning
           under lexical warnings.  You should carefully launder the arguments
           to guarantee their validity.  In future releases of Perl the forms
           will become fatal errors so consider starting laundering now.
       o   Tied hash interfaces are now required to have the EXISTS and DELETE
           methods (either own or inherited).
       o   If tr/// is just counting characters, it doesn't attempt to modify
           its target.
       o   untie() will now call an UNTIE() hook if it exists.  See perltie
           for details. [561]
       o   "utime" in perlfunc now supports "utime undef, undef, @files" to
           change the file timestamps to the current time.
       o   The rules for allowing underscores (underbars) in numeric constants
           have been relaxed and simplified: now you can have an underscore
           simply between digits.
       o   Rather than relying on C's argv[0] (which may not contain a full
           pathname) where possible $^X is now set by asking the operating
           system.  (eg by reading /proc/self/exe on Linux, /proc/curproc/file
           on FreeBSD)
       o   A new variable, "${^TAINT}", indicates whether taint mode is
           enabled.
       o   You can now override the readline() builtin, and this overrides
           also the <FILEHANDLE> angle bracket operator.
       o   The command-line options -s and -F are now recognized on the
           shebang (#!) line.
       o   Use of the "/c" match modifier without an accompanying "/g"
           modifier elicits a new warning: "Use of /c modifier is meaningless
           without /g".
           Use of "/c" in substitutions, even with "/g", elicits "Use of /c
           modifier is meaningless in s///".
           Use of "/g" with "split" elicits "Use of /g modifier is meaningless
           in split".
       o   Support for the "CLONE" special subroutine had been added.  With
           ithreads, when a new thread is created, all Perl data is cloned,
           however non-Perl data cannot be cloned automatically.  In "CLONE"
           you can do whatever you need to do, like for example handle the
           cloning of non-Perl data, if necessary.  "CLONE" will be executed
           once for every package that has it defined or inherited.  It will
           be called in the context of the new thread, so all modifications
           are made in the new area.
           See perlmod
Modules and Pragmata
   New Modules and Pragmata
       o   "Attribute::Handlers", originally by Damian Conway and now
           maintained by Arthur Bergman, allows a class to define attribute
           handlers.
               package MyPack;
               use Attribute::Handlers;
               sub Wolf :ATTR(SCALAR) { print "howl!\n" }
               # later, in some package using or inheriting from MyPack...
               my MyPack $Fluffy : Wolf; # the attribute handler Wolf will be called
           Both variables and routines can have attribute handlers.  Handlers
           can be specific to type (SCALAR, ARRAY, HASH, or CODE), or specific
           to the exact compilation phase (BEGIN, CHECK, INIT, or END).  See
           Attribute::Handlers.
       o   "B::Concise", by Stephen McCamant, is a new compiler backend for
           walking the Perl syntax tree, printing concise info about ops.  The
           output is highly customisable.  See B::Concise. [561+]
       o   The new bignum, bigint, and bigrat pragmas, by Tels, implement
           transparent bignum support (using the Math::BigInt, Math::BigFloat,
           and Math::BigRat backends).
       o   "Class::ISA", by Sean Burke, is a module for reporting the search
           path for a class's ISA tree.  See Class::ISA.
       o   "Cwd" now has a split personality: if possible, an XS extension is
           used, (this will hopefully be faster, more secure, and more robust)
           but if not possible, the familiar Perl implementation is used.
       o   "Devel::PPPort", originally by Kenneth Albanowski and now
           maintained by Paul Marquess, has been added.  It is primarily used
           by "h2xs" to enhance portability of XS modules between different
           versions of Perl.  See Devel::PPPort.
       o   "Digest", frontend module for calculating digests (checksums), from
           Gisle Aas, has been added.  See Digest.
       o   "Digest::MD5" for calculating MD5 digests (checksums) as defined in
           RFC 1321, from Gisle Aas, has been added.  See Digest::MD5.
               use Digest::MD5 'md5_hex';
               $digest = md5_hex("Thirsty Camel");
               print $digest, "\n"; # 01d19d9d2045e005c3f1b80e8b164de1
           NOTE: the "MD5" backward compatibility module is deliberately not
           included since its further use is discouraged.
           See also PerlIO::via::QuotedPrint.
       o   "Encode", originally by Nick Ing-Simmons and now maintained by Dan
           Kogai, provides a mechanism to translate between different
           character encodings.  Support for Unicode, ISO-8859-1, and ASCII
           are compiled in to the module.  Several other encodings (like the
           rest of the ISO-8859, CP*/Win*, Mac, KOI8-R, three variants EBCDIC,
           Chinese, Japanese, and Korean encodings) are included and can be
           loaded at runtime.  (For space considerations, the largest Chinese
           encodings have been separated into their own CPAN module,
           Encode::HanExtra, which Encode will use if available).  See Encode.
           Any encoding supported by Encode module is also available to the
           ":encoding()" layer if PerlIO is used.
       o   "Hash::Util" is the interface to the new restricted hashes feature.
           (Implemented by Jeffrey Friedl, Nick Ing-Simmons, and Michael
           Schwern.)  See Hash::Util.
       o   "I18N::Langinfo" can be used to query locale information.  See
           I18N::Langinfo.
       o   "I18N::LangTags", by Sean Burke, has functions for dealing with
           RFC3066-style language tags.  See I18N::LangTags.
       o   "ExtUtils::Constant", by Nicholas Clark, is a new tool for
           extension writers for generating XS code to import C header
           constants.  See ExtUtils::Constant.
       o   "Filter::Simple", by Damian Conway, is an easy-to-use frontend to
           Filter::Util::Call.  See Filter::Simple.
               # in MyFilter.pm:
               package MyFilter;
               use Filter::Simple sub {
                   while (my ($from, $to) = splice @_, 0, 2) {
                           s/$from/$to/g;
                   }
               };
               1;
               # in user's code:
               use MyFilter qr/red/ => 'green';
               print "red\n";   # this code is filtered, will print "green\n"
               print "bored\n"; # this code is filtered, will print "bogreen\n"
               no MyFilter;
               print "red\n";   # this code is not filtered, will print "red\n"
       o   "File::Temp", by Tim Jenness, allows one to create temporary files
           and directories in an easy, portable, and secure way.  See
           File::Temp.  [561+]
       o   "Filter::Util::Call", by Paul Marquess, provides you with the
           framework to write source filters in Perl.  For most uses, the
           frontend Filter::Simple is to be preferred.  See
           Filter::Util::Call.
       o   "if", by Ilya Zakharevich, is a new pragma for conditional
           inclusion of modules.
       o   libnet, by Graham Barr, is a collection of perl5 modules related to
           network programming.  See Net::FTP, Net::NNTP, Net::Ping (not part
           of libnet, but related), Net::POP3, Net::SMTP, and Net::Time.
           Perl installation leaves libnet unconfigured; use libnetcfg to
           configure it.
       o   "List::Util", by Graham Barr, is a selection of general-utility
           list subroutines, such as sum(), min(), first(), and shuffle().
           See List::Util.
       o   "Locale::Constants", "Locale::Country", "Locale::Currency"
           "Locale::Language", and Locale::Script, by Neil Bowers, have been
           added.  They provide the codes for various locale standards, such
           as "fr" for France, "usd" for US Dollar, and "ja" for Japanese.
               use Locale::Country;
               $country = code2country('jp');               # $country gets 'Japan'
               $code    = country2code('Norway');           # $code gets 'no'
           See Locale::Constants, Locale::Country, Locale::Currency, and
           Locale::Language.
       o   "Locale::Maketext", by Sean Burke, is a localization framework.
           See Locale::Maketext, and Locale::Maketext::TPJ13.  The latter is
           an article about software localization, originally published in The
           Perl Journal #13, and republished here with kind permission.
       o   "Math::BigRat" for big rational numbers, to accompany Math::BigInt
           and Math::BigFloat, from Tels.  See Math::BigRat.
       o   "Memoize" can make your functions faster by trading space for time,
           from Mark-Jason Dominus.  See Memoize.
       o   "MIME::Base64", by Gisle Aas, allows you to encode data in base64,
           as defined in RFC 2045 - MIME (Multipurpose Internet Mail
           Extensions).
               use MIME::Base64;
               $encoded = encode_base64('Aladdin:open sesame');
               $decoded = decode_base64($encoded);
               print $encoded, "\n"; # "QWxhZGRpbjpvcGVuIHNlc2FtZQ=="
           See MIME::Base64.
       o   "MIME::QuotedPrint", by Gisle Aas, allows you to encode data in
           quoted-printable encoding, as defined in RFC 2045 - MIME
           (Multipurpose Internet Mail Extensions).
               use MIME::QuotedPrint;
               $encoded = encode_qp("\xDE\xAD\xBE\xEF");
               $decoded = decode_qp($encoded);
               print $encoded, "\n"; # "=DE=AD=BE=EF\n"
               print $decoded, "\n"; # "\xDE\xAD\xBE\xEF\n"
           See also PerlIO::via::QuotedPrint.
       o   "NEXT", by Damian Conway, is a pseudo-class for method redispatch.
           See NEXT.
       o   "open" is a new pragma for setting the default I/O layers for
           open().
       o   "PerlIO::scalar", by Nick Ing-Simmons, provides the implementation
           of IO to "in memory" Perl scalars as discussed above.  It also
           serves as an example of a loadable PerlIO layer.  Other future
           possibilities include PerlIO::Array and PerlIO::Code.  See
           PerlIO::scalar.
       o   "PerlIO::via", by Nick Ing-Simmons, acts as a PerlIO layer and
           wraps PerlIO layer functionality provided by a class (typically
           implemented in Perl code).
       o   "PerlIO::via::QuotedPrint", by Elizabeth Mattijsen, is an example
           of a "PerlIO::via" class:
               use PerlIO::via::QuotedPrint;
               open($fh,">:via(QuotedPrint)",$path);
           This will automatically convert everything output to $fh to Quoted-
           Printable.  See PerlIO::via and PerlIO::via::QuotedPrint.
       o   "Pod::ParseLink", by Russ Allbery, has been added, to parse L<>
           links in pods as described in the new perlpodspec.
       o   "Pod::Text::Overstrike", by Joe Smith, has been added.  It converts
           POD data to formatted overstrike text.  See Pod::Text::Overstrike.
           [561+]
       o   "Scalar::Util" is a selection of general-utility scalar
           subroutines, such as blessed(), reftype(), and tainted().  See
           Scalar::Util.
       o   "sort" is a new pragma for controlling the behaviour of sort().
       o   "Storable" gives persistence to Perl data structures by allowing
           the storage and retrieval of Perl data to and from files in a fast
           and compact binary format.  Because in effect Storable does
           serialisation of Perl data structures, with it you can also clone
           deep, hierarchical datastructures.  Storable was originally created
           by Raphael Manfredi, but it is now maintained by Abhijit Menon-Sen.
           Storable has been enhanced to understand the two new hash features,
           Unicode keys and restricted hashes.  See Storable.
       o   "Switch", by Damian Conway, has been added.  Just by saying
               use Switch;
           you have "switch" and "case" available in Perl.
               use Switch;
               switch ($val) {
                           case 1          { print "number 1" }
                           case "a"        { print "string a" }
                           case [1..10,42] { print "number in list" }
                           case (@array)   { print "number in list" }
                           case /\w+/      { print "pattern" }
                           case qr/\w+/    { print "pattern" }
                           case (%hash)    { print "entry in hash" }
                           case (\%hash)   { print "entry in hash" }
                           case (\&sub)    { print "arg to subroutine" }
                           else            { print "previous case not true" }
               }
           See Switch.
       o   "Test::More", by Michael Schwern, is yet another framework for
           writing test scripts, more extensive than Test::Simple.  See
           Test::More.
       o   "Test::Simple", by Michael Schwern, has basic utilities for writing
           tests.   See Test::Simple.
       o   "Text::Balanced", by Damian Conway, has been added, for extracting
           delimited text sequences from strings.
               use Text::Balanced 'extract_delimited';
               ($a, $b) = extract_delimited("'never say never', he never said", "'", '');
           $a will be "'never say never'", $b will be ', he never said'.
           In addition to extract_delimited(), there are also
           extract_bracketed(), extract_quotelike(), extract_codeblock(),
           extract_variable(), extract_tagged(), extract_multiple(),
           gen_delimited_pat(), and gen_extract_tagged().  With these, you can
           implement rather advanced parsing algorithms.  See Text::Balanced.
       o   "threads", by Arthur Bergman, is an interface to interpreter
           threads.  Interpreter threads (ithreads) is the new thread model
           introduced in Perl 5.6 but only available as an internal interface
           for extension writers (and for Win32 Perl for "fork()" emulation).
           See threads, threads::shared, and perlthrtut.
       o   "threads::shared", by Arthur Bergman, allows data sharing for
           interpreter threads.  See threads::shared.
       o   "Tie::File", by Mark-Jason Dominus, associates a Perl array with
           the lines of a file.  See Tie::File.
       o   "Tie::Memoize", by Ilya Zakharevich, provides on-demand loaded
           hashes.  See Tie::Memoize.
       o   "Tie::RefHash::Nestable", by Edward Avis, allows storing hash
           references (unlike the standard Tie::RefHash)  The module is
           contained within Tie::RefHash.  See Tie::RefHash.
       o   "Time::HiRes", by Douglas E. Wegscheid, provides high resolution
           timing (ualarm, usleep, and gettimeofday).  See Time::HiRes.
       o   "Unicode::UCD" offers a querying interface to the Unicode Character
           Database.  See Unicode::UCD.
       o   "Unicode::Collate", by SADAHIRO Tomoyuki, implements the UCA
           (Unicode Collation Algorithm) for sorting Unicode strings.  See
           Unicode::Collate.
       o   "Unicode::Normalize", by SADAHIRO Tomoyuki, implements the various
           Unicode normalization forms.  See Unicode::Normalize.
       o   "XS::APItest", by Tim Jenness, is a test extension that exercises
           XS APIs.  Currently only "printf()" is tested: how to output
           various basic data types from XS.
       o   "XS::Typemap", by Tim Jenness, is a test extension that exercises
           XS typemaps.  Nothing gets installed, but the code is worth
           studying for extension writers.
   Updated And Improved Modules and Pragmata
       o   The following independently supported modules have been updated to
           the newest versions from CPAN: CGI, CPAN, DB_File, File::Spec,
           File::Temp, Getopt::Long, Math::BigFloat, Math::BigInt, the
           podlators bundle (Pod::Man, Pod::Text), Pod::LaTeX [561+],
           Pod::Parser, Storable, Term::ANSIColor, Test, Text-Tabs+Wrap.
       o   attributes::reftype() now works on tied arguments.
       o   AutoLoader can now be disabled with "no AutoLoader;".
       o   B::Deparse has been significantly enhanced by Robin Houston.  It
           can now deparse almost all of the standard test suite (so that the
           tests still succeed).  There is a make target "test.deparse" for
           trying this out.
       o   Carp now has better interface documentation, and the @CARP_NOT
           interface has been added to get optional control over where errors
           are reported independently of @ISA, by Ben Tilly.
       o   Class::Struct can now define the classes in compile time.
       o   Class::Struct now assigns the array/hash element if the accessor is
           called with an array/hash element as the sole argument.
       o   The return value of Cwd::fastcwd() is now tainted.
       o   Data::Dumper now has an option to sort hashes.
       o   Data::Dumper now has an option to dump code references using
           B::Deparse.
       o   DB_File now supports newer Berkeley DB versions, among other
           improvements.
       o   Devel::Peek now has an interface for the Perl memory statistics
           (this works only if you are using perl's malloc, and if you have
           compiled with debugging).
       o   The English module can now be used without the infamous performance
           hit by saying
                   use English '-no_match_vars';
           (Assuming, of course, that you don't need the troublesome variables
           "$`", $&, or "$'".)  Also, introduced @LAST_MATCH_START and
           @LAST_MATCH_END English aliases for "@-" and "@+".
       o   ExtUtils::MakeMaker has been significantly cleaned up and fixed.
           The enhanced version has also been backported to earlier releases
           of Perl and submitted to CPAN so that the earlier releases can
           enjoy the fixes.
       o   The arguments of WriteMakefile() in Makefile.PL are now checked for
           sanity much more carefully than before.  This may cause new
           warnings when modules are being installed.  See ExtUtils::MakeMaker
           for more details.
       o   ExtUtils::MakeMaker now uses File::Spec internally, which hopefully
           leads to better portability.
       o   Fcntl, Socket, and Sys::Syslog have been rewritten by Nicholas
           Clark to use the new-style constant dispatch section (see
           ExtUtils::Constant).  This means that they will be more robust and
           hopefully faster.
       o   File::Find now chdir()s correctly when chasing symbolic links.
           [561]
       o   File::Find now has pre- and post-processing callbacks.  It also
           correctly changes directories when chasing symbolic links.
           Callbacks (naughtily) exiting with "next;" instead of "return;" now
           work.
       o   File::Find is now (again) reentrant.  It also has been made more
           portable.
       o   The warnings issued by File::Find now belong to their own category.
           You can enable/disable them with "use/no warnings 'File::Find';".
       o   File::Glob::glob() has been renamed to File::Glob::bsd_glob()
           because the name clashes with the builtin glob().  The older name
           is still available for compatibility, but is deprecated. [561]
       o   File::Glob now supports "GLOB_LIMIT" constant to limit the size of
           the returned list of filenames.
       o   IPC::Open3 now allows the use of numeric file descriptors.
       o   IO::Socket now has an atmark() method, which returns true if the
           socket is positioned at the out-of-band mark.  The method is also
           exportable as a sockatmark() function.
       o   IO::Socket::INET failed to open the specified port if the service
           name was not known.  It now correctly uses the supplied port number
           as is. [561]
       o   IO::Socket::INET has support for the ReusePort option (if your
           platform supports it).  The Reuse option now has an alias,
           ReuseAddr.  For clarity, you may want to prefer ReuseAddr.
       o   IO::Socket::INET now supports a value of zero for "LocalPort"
           (usually meaning that the operating system will make one up.)
       o   'use lib' now works identically to @INC.  Removing directories with
           'no lib' now works.
       o   Math::BigFloat and Math::BigInt have undergone a full rewrite by
           Tels.  They are now magnitudes faster, and they support various
           bignum libraries such as GMP and PARI as their backends.
       o   Math::Complex handles inf, NaN etc., better.
       o   Net::Ping has been considerably enhanced by Rob Brown: multihoming
           is now supported, Win32 functionality is better, there is now time
           measuring functionality (optionally high-resolution using
           Time::HiRes), and there is now "external" protocol which uses
           Net::Ping::External module which runs your external ping utility
           and parses the output.  A version of Net::Ping::External is
           available in CPAN.
           Note that some of the Net::Ping tests are disabled when running
           under the Perl distribution since one cannot assume one or more of
           the following: enabled echo port at localhost, full Internet
           connectivity, or sympathetic firewalls.  You can set the
           environment variable PERL_TEST_Net_Ping to "1" (one) before running
           the Perl test suite to enable all the Net::Ping tests.
       o   POSIX::sigaction() is now much more flexible and robust.  You can
           now install coderef handlers, 'DEFAULT', and 'IGNORE' handlers,
           installing new handlers was not atomic.
       o   In Safe, %INC is now localised in a Safe compartment so that
           use/require work.
       o   In SDBM_File on dosish platforms, some keys went missing because of
           lack of support for files with "holes".  A workaround for the
           problem has been added.
       o   In Search::Dict one can now have a pre-processing hook for the
           lines being searched.
       o   The Shell module now has an OO interface.
       o   In Sys::Syslog there is now a failover mechanism that will go
           through alternative connection mechanisms until the message is
           successfully logged.
       o   The Test module has been significantly enhanced.
       o   Time::Local::timelocal() does not handle fractional seconds
           anymore.  The rationale is that neither does localtime(), and
           timelocal() and localtime() are supposed to be inverses of each
           other.
       o   The vars pragma now supports declaring fully qualified variables.
           (Something that "our()" does not and will not support.)
       o   The "utf8::" name space (as in the pragma) provides various Perl-
           callable functions to provide low level access to Perl's internal
           Unicode representation.  At the moment only length() has been
           implemented.
Utility Changes
       o   Emacs perl mode (emacs/cperl-mode.el) has been updated to version
           4.31.
       o   emacs/e2ctags.pl is now much faster.
       o   "enc2xs" is a tool for people adding their own encodings to the
           Encode module.
       o   "h2ph" now supports C trigraphs.
       o   "h2xs" now produces a template README.
       o   "h2xs" now uses "Devel::PPPort" for better portability between
           different versions of Perl.
       o   "h2xs" uses the new ExtUtils::Constant module which will affect
           newly created extensions that define constants.  Since the new code
           is more correct (if you have two constants where the first one is a
           prefix of the second one, the first constant never got defined),
           less lossy (it uses integers for integer constant, as opposed to
           the old code that used floating point numbers even for integer
           constants), and slightly faster, you might want to consider
           regenerating your extension code (the new scheme makes regenerating
           easy).  h2xs now also supports C trigraphs.
       o   "libnetcfg" has been added to configure libnet.
       o   "perlbug" is now much more robust.  It also sends the bug report to
           perl.org, not perl.com.
       o   "perlcc" has been rewritten and its user interface (that is,
           command line) is much more like that of the Unix C compiler, cc.
           (The perlbc tools has been removed.  Use "perlcc -B" instead.)
           Note that perlcc is still considered very experimental and
           unsupported. [561]
       o   "perlivp" is a new Installation Verification Procedure utility for
           running any time after installing Perl.
       o   "piconv" is an implementation of the character conversion utility
           "iconv", demonstrating the new Encode module.
       o   "pod2html" now allows specifying a cache directory.
       o   "pod2html" now produces XHTML 1.0.
       o   "pod2html" now understands POD written using different line endings
           (PC-like CRLF versus Unix-like LF versus MacClassic-like CR).
       o   "s2p" has been completely rewritten in Perl.  (It is in fact a full
           implementation of sed in Perl: you can use the sed functionality by
           using the "psed" utility.)
       o   "xsubpp" now understands POD documentation embedded in the *.xs
           files. [561]
       o   "xsubpp" now supports the OUT keyword.
New Documentation
       o   perl56delta details the changes between the 5.005 release and the
           5.6.0 release.
       o   perlclib documents the internal replacements for standard C library
           functions.  (Interesting only for extension writers and Perl core
           hackers.) [561+]
       o   perldebtut is a Perl debugging tutorial. [561+]
       o   perlebcdic contains considerations for running Perl on EBCDIC
           platforms. [561+]
       o   perlintro is a gentle introduction to Perl.
       o   perliol documents the internals of PerlIO with layers.
       o   perlmodstyle is a style guide for writing modules.
       o   perlnewmod tells about writing and submitting a new module. [561+]
       o   perlpacktut is a pack() tutorial.
       o   perlpod has been rewritten to be clearer and to record the best
           practices gathered over the years.
       o   perlpodspec is a more formal specification of the pod format,
           mainly of interest for writers of pod applications, not to people
           writing in pod.
       o   perlretut is a regular expression tutorial. [561+]
       o   perlrequick is a regular expressions quick-start guide.  Yes, much
           quicker than perlretut. [561]
       o   perltodo has been updated.
       o   perltootc has been renamed as perltooc (to not to conflict with
           perltoot in filesystems restricted to "8.3" names).
       o   perluniintro is an introduction to using Unicode in Perl.
           (perlunicode is more of a detailed reference and background
           information)
       o   perlutil explains the command line utilities packaged with the Perl
           distribution. [561+]
       The following platform-specific documents are available before the
       installation as README.platform, and after the installation as
       perlplatform:
           perlaix perlamiga perlapollo perlbeos perlbs2000
           perlce perlcygwin perldgux perldos perlepoc perlfreebsd perlhpux
           perlhurd perlirix perlmachten perlmacos perlmint perlmpeix
           perlnetware perlos2 perlos390 perlplan9 perlqnx perlsolaris
           perltru64 perluts perlvmesa perlvms perlvos perlwin32
       These documents usually detail one or more of the following subjects:
       configuring, building, testing, installing, and sometimes also using
       Perl on the said platform.
       Eastern Asian Perl users are now welcomed in their own languages:
       README.jp (Japanese), README.ko (Korean), README.cn (simplified
       Chinese) and README.tw (traditional Chinese), which are written in
       normal pod but encoded in EUC-JP, EUC-KR, EUC-CN and Big5.  These will
       get installed as
          perljp perlko perlcn perltw
       o   The documentation for the POSIX-BC platform is called "BS2000", to
           avoid confusion with the Perl POSIX module.
       o   The documentation for the WinCE platform is called perlce
           (README.ce in the source code kit), to avoid confusion with the
           perlwin32 documentation on 8.3-restricted filesystems.
Performance Enhancements
       o   map() could get pathologically slow when the result list it
           generates is larger than the source list.  The performance has been
           improved for common scenarios. [561]
       o   sort() is also fully reentrant, in the sense that the sort function
           can itself call sort().  This did not work reliably in previous
           releases. [561]
       o   sort() has been changed to use primarily mergesort internally as
           opposed to the earlier quicksort.  For very small lists this may
           result in slightly slower sorting times, but in general the speedup
           should be at least 20%.  Additional bonuses are that the worst case
           behaviour of sort() is now better (in computer science terms it now
           runs in time O(N log N), as opposed to quicksort's Theta(N**2)
           worst-case run time behaviour), and that sort() is now stable
           (meaning that elements with identical keys will stay ordered as
           they were before the sort).  See the "sort" pragma for information.
           The story in more detail: suppose you want to serve yourself a
           little slice of Pi.
               @digits = ( 3,1,4,1,5,9 );
           A numerical sort of the digits will yield (1,1,3,4,5,9), as
           expected.  Which 1 comes first is hard to know, since one 1 looks
           pretty much like any other.  You can regard this as totally
           trivial, or somewhat profound.  However, if you just want to sort
           the even digits ahead of the odd ones, then what will
               sort { ($a % 2) <=> ($b % 2) } @digits;
           yield?  The only even digit, 4, will come first.  But how about the
           odd numbers, which all compare equal?  With the quicksort algorithm
           used to implement Perl 5.6 and earlier, the order of ties is left
           up to the sort.  So, as you add more and more digits of Pi, the
           order in which the sorted even and odd digits appear will change.
           and, for sufficiently large slices of Pi, the quicksort algorithm
           in Perl 5.8 won't return the same results even if reinvoked with
           the same input.  The justification for this rests with quicksort's
           worst case behavior.  If you run
              sort { $a <=> $b } ( 1 .. $N , 1 .. $N );
           (something you might approximate if you wanted to merge two sorted
           arrays using sort), doubling $N doesn't just double the quicksort
           time, it quadruples it.  Quicksort has a worst case run time that
           can grow like N**2, so-called quadratic behaviour, and it can
           happen on patterns that may well arise in normal use.  You won't
           notice this for small arrays, but you will notice it with larger
           arrays, and you may not live long enough for the sort to complete
           on arrays of a million elements.  So the 5.8 quicksort scrambles
           large arrays before sorting them, as a statistical defence against
           quadratic behaviour.  But that means if you sort the same large
           array twice, ties may be broken in different ways.
           Because of the unpredictability of tie-breaking order, and the
           quadratic worst-case behaviour, quicksort was almost replaced
           completely with a stable mergesort.  Stable means that ties are
           broken to preserve the original order of appearance in the input
           array.  So
               sort { ($a % 2) <=> ($b % 2) } (3,1,4,1,5,9);
           will yield (4,3,1,1,5,9), guaranteed.  The even and odd numbers
           appear in the output in the same order they appeared in the input.
           Mergesort has worst case O(N log N) behaviour, the best value
           attainable.  And, ironically, this mergesort does particularly well
           where quicksort goes quadratic:  mergesort sorts (1..$N, 1..$N) in
           O(N) time.  But quicksort was rescued at the last moment because it
           is faster than mergesort on certain inputs and platforms.  For
           example, if you really don't care about the order of even and odd
           digits, quicksort will run in O(N) time; it's very good at sorting
           many repetitions of a small number of distinct elements.  The
           quicksort divide and conquer strategy works well on platforms with
           relatively small, very fast, caches.  Eventually, the problem gets
           whittled down to one that fits in the cache, from which point it
           benefits from the increased memory speed.
           Quicksort was rescued by implementing a sort pragma to control
           aspects of the sort.  The stable subpragma forces stable behaviour,
           regardless of algorithm.  The _quicksort and _mergesort subpragmas
           are heavy-handed ways to select the underlying implementation.  The
           leading "_" is a reminder that these subpragmas may not survive
           beyond 5.8.  More appropriate mechanisms for selecting the
           implementation exist, but they wouldn't have arrived in time to
           save quicksort.
       o   Hashes now use Bob Jenkins "One-at-a-Time" hashing key algorithm (
           http://burtleburtle.net/bob/hash/doobs.html ).  This algorithm is
           reasonably fast while producing a much better spread of values than
           the old hashing algorithm (originally by Chris Torek, later tweaked
           by Ilya Zakharevich).  Hash values output from the algorithm on a
           hash of all 3-char printable ASCII keys comes much closer to
           passing the DIEHARD random number generation tests.  According to
           perlbench, this change has not affected the overall speed of Perl.
       o   unshift() should now be noticeably faster.
Installation and Configuration Improvements
   Generic Improvements
       o   INSTALL now explains how you can configure Perl to use 64-bit
           integers even on non-64-bit platforms.
       o   Policy.sh policy change: if you are reusing a Policy.sh file (see
           INSTALL) and you use Configure -Dprefix=/foo/bar and in the old
           Policy $prefix eq $siteprefix and $prefix eq $vendorprefix, all of
           them will now be changed to the new prefix, /foo/bar.  (Previously
           only $prefix changed.)  If you do not like this new behaviour,
           specify prefix, siteprefix, and vendorprefix explicitly.
       o   A new optional location for Perl libraries, otherlibdirs, is
           available.  It can be used for example for vendor add-ons without
           disturbing Perl's own library directories.
       o   In many platforms, the vendor-supplied 'cc' is too stripped-down to
           build Perl (basically, 'cc' doesn't do ANSI C).  If this seems to
           be the case and 'cc' does not seem to be the GNU C compiler 'gcc',
           an automatic attempt is made to find and use 'gcc' instead.
       o   gcc needs to closely track the operating system release to avoid
           build problems. If Configure finds that gcc was built for a
           different operating system release than is running, it now gives a
           clearly visible warning that there may be trouble ahead.
       o   Since Perl 5.8 is not binary-compatible with previous releases of
           Perl, Configure no longer suggests including the 5.005 modules in
           @INC.
       o   Configure "-S" can now run non-interactively. [561]
       o   Configure support for pdp11-style memory models has been removed
           due to obsolescence. [561]
       o   configure.gnu now works with options with whitespace in them.
       o   installperl now outputs everything to STDERR.
       o   Because PerlIO is now the default on most platforms, "-perlio"
           doesn't get appended to the $Config{archname} (also known as $^O)
           anymore.  Instead, if you explicitly choose not to use perlio
           (Configure command line option -Uuseperlio), you will get "-stdio"
           appended.
       o   Another change related to the architecture name is that "-64all"
           (-Duse64bitall, or "maximally 64-bit") is appended only if your
           pointers are 64 bits wide.  (To be exact, the use64bitall is
           ignored.)
       o   In AFS installations, one can configure the root of the AFS to be
           somewhere else than the default /afs by using the Configure
           parameter "-Dafsroot=/some/where/else".
       o   APPLLIB_EXP, a lesser-known configuration-time definition, has been
           documented.  It can be used to prepend site-specific directories to
           Perl's default search path (@INC); see INSTALL for information.
       o   The version of Berkeley DB used when the Perl (and, presumably, the
           DB_File extension) was built is now available as
           @Config{qw(db_version_major db_version_minor db_version_patch)}
           from Perl and as "DB_VERSION_MAJOR_CFG DB_VERSION_MINOR_CFG
           DB_VERSION_PATCH_CFG" from C.
       o   Building Berkeley DB3 for compatibility modes for DB, NDBM, and
           ODBM has been documented in INSTALL.
       o   If you have CPAN access (either network or a local copy such as a
           CD-ROM) you can during specify extra modules to Configure to build
           and install with Perl using the -Dextras=...  option.  See INSTALL
           for more details.
       o   In addition to config.over, a new override file, config.arch, is
           available.  This file is supposed to be used by hints file writers
           for architecture-wide changes (as opposed to config.over which is
           for site-wide changes).
       o   If your file system supports symbolic links, you can build Perl
           outside of the source directory by
                   mkdir perl/build/directory
                   cd perl/build/directory
                   sh /path/to/perl/source/Configure -Dmksymlinks ...
           This will create in perl/build/directory a tree of symbolic links
           pointing to files in /path/to/perl/source.  The original files are
           left unaffected.  After Configure has finished, you can just say
                   make all test
           and Perl will be built and tested, all in perl/build/directory.
           [561]
       o   For Perl developers, several new make targets for profiling and
           debugging have been added; see perlhack.
           o       Use of the gprof tool to profile Perl has been documented
                   in perlhack.  There is a make target called "perl.gprof"
                   for generating a gprofiled Perl executable.
           o       If you have GCC 3, there is a make target called
                   "perl.gcov" for creating a gcoved Perl executable for
                   coverage analysis.  See perlhack.
           o       If you are on IRIX or Tru64 platforms, new
                   profiling/debugging options have been added; see perlhack
                   for more information about pixie and Third Degree.
       o   Guidelines of how to construct minimal Perl installations have been
           added to INSTALL.
       o   The Thread extension is now not built at all under ithreads
           ("Configure -Duseithreads") because it wouldn't work anyway (the
           Thread extension requires being Configured with
           "-Duse5005threads").
           Note that the 5.005 threads are unsupported and deprecated: if you
           have code written for the old threads you should migrate it to the
           new ithreads model.
       o   The Gconvert macro ($Config{d_Gconvert}) used by perl for
           stringifying floating-point numbers is now more picky about using
           sprintf %.*g rules for the conversion.  Some platforms that used to
           use gcvt may now resort to the slower sprintf.
       o   The obsolete method of making a special (e.g., debugging) flavor of
           perl by saying
                   make LIBPERL=libperld.a
           has been removed. Use -DDEBUGGING instead.
   New Or Improved Platforms
       For the list of platforms known to support Perl, see "Supported
       Platforms" in perlport.
       o   AIX dynamic loading should be now better supported.
       o   AIX should now work better with gcc, threads, and 64-bitness.  Also
           the long doubles support in AIX should be better now.  See perlaix.
       o   AtheOS ( http://www.atheos.cx/ ) is a new platform.
       o   BeOS has been reclaimed.
       o   The DG/UX platform now supports 5.005-style threads.  See perldgux.
       o   The DYNIX/ptx platform (also known as dynixptx) is supported at or
           near osvers 4.5.2.
       o   EBCDIC platforms (z/OS (also known as OS/390), POSIX-BC, and
           VM/ESA) have been regained.  Many test suite tests still fail and
           the co-existence of Unicode and EBCDIC isn't quite settled, but the
           situation is much better than with Perl 5.6.  See perlos390,
           perlbs2000 (for POSIX-BC), and perlvmesa for more information.
       o   Building perl with -Duseithreads or -Duse5005threads now works
           under HP-UX 10.20 (previously it only worked under 10.30 or later).
           You will need a thread library package installed. See README.hpux.
           [561]
       o   Mac OS Classic is now supported in the mainstream source package
           (MacPerl has of course been available since perl 5.004 but now the
           source code bases of standard Perl and MacPerl have been
           synchronised) [561]
       o   Mac OS X (or Darwin) should now be able to build Perl even on HFS+
           filesystems.  (The case-insensitivity used to confuse the Perl
           build process.)
       o   NCR MP-RAS is now supported. [561]
       o   All the NetBSD specific patches (except for the installation
           specific ones) have been merged back to the main distribution.
       o   NetWare from Novell is now supported.  See perlnetware.
       o   NonStop-UX is now supported. [561]
       o   NEC SUPER-UX is now supported.
       o   All the OpenBSD specific patches (except for the installation
           specific ones) have been merged back to the main distribution.
       o   Perl has been tested with the GNU pth userlevel thread package (
           http://www.gnu.org/software/pth/pth.html ).  All thread tests of
           Perl now work, but not without adding some yield()s to the tests,
           so while pth (and other userlevel thread implementations) can be
           considered to be "working" with Perl ithreads, keep in mind the
           possible non-preemptability of the underlying thread
           implementation.
       o   Stratus VOS is now supported using Perl's native build method
           (Configure).  This is the recommended method to build Perl on VOS.
           The older methods, which build miniperl, are still available.  See
           perlvos. [561+]
       o   The Amdahl UTS Unix mainframe platform is now supported. [561]
       o   WinCE is now supported.  See perlce.
       o   z/OS (formerly known as OS/390, formerly known as MVS OE) now has
           support for dynamic loading.  This is not selected by default,
           however, you must specify -Dusedl in the arguments of Configure.
           [561]
Selected Bug Fixes
       Numerous memory leaks and uninitialized memory accesses have been
       hunted down.  Most importantly, anonymous subs used to leak quite a
       bit. [561]
       o   The autouse pragma didn't work for Multi::Part::Function::Names.
       o   caller() could cause core dumps in certain situations.  Carp was
           sometimes affected by this problem.  In particular, caller() now
           returns a subroutine name of "(unknown)" for subroutines that have
           been removed from the symbol table.
       o   chop(@list) in list context returned the characters chopped in
           reverse order.  This has been reversed to be in the right order.
           [561]
       o   Configure no longer includes the DBM libraries (dbm, gdbm, db,
           ndbm) when building the Perl binary.  The only exception to this is
           SunOS 4.x, which needs them. [561]
       o   The behaviour of non-decimal but numeric string constants such as
           "0x23" was platform-dependent: in some platforms that was seen as
           35, in some as 0, in some as a floating point number (don't ask).
           This was caused by Perl's using the operating system libraries in a
           situation where the result of the string to number conversion is
           undefined: now Perl consistently handles such strings as zero in
           numeric contexts.
       o   Several debugger fixes: exit code now reflects the script exit
           code, condition "0" now treated correctly, the "d" command now
           checks line number, $. no longer gets corrupted, and all debugger
           output now goes correctly to the socket if RemotePort is set. [561]
       o   The debugger (perl5db.pl) has been modified to present a more
           consistent commands interface, via (CommandSet=580).  perl5db.t was
           also added to test the changes, and as a placeholder for further
           tests.
           See perldebug.
       o   The debugger has a new "dumpDepth" option to control the maximum
           depth to which nested structures are dumped.  The "x" command has
           been extended so that "x N EXPR" dumps out the value of EXPR to a
           depth of at most N levels.
       o   The debugger can now show lexical variables if you have the CPAN
           module PadWalker installed.
       o   The order of DESTROYs has been made more predictable.
       o   Perl 5.6.0 could emit spurious warnings about redefinition of
           dl_error() when statically building extensions into perl.  This has
           been corrected. [561]
       o   dprofpp -R didn't work.
       o   *foo{FORMAT} now works.
       o   Infinity is now recognized as a number.
       o   UNIVERSAL::isa no longer caches methods incorrectly.  (This broke
           the Tk extension with 5.6.0.) [561]
       o   Lexicals I: lexicals outside an eval "" weren't resolved correctly
           inside a subroutine definition inside the eval "" if they were not
           already referenced in the top level of the eval""ed code.
       o   Lexicals II: lexicals leaked at file scope into subroutines that
           were declared before the lexicals.
       o   Lexical warnings now propagating correctly between scopes and into
           "eval "..."".
       o   "use warnings qw(FATAL all)" did not work as intended.  This has
           been corrected. [561]
       o   warnings::enabled() now reports the state of $^W correctly if the
           caller isn't using lexical warnings. [561]
       o   Line renumbering with eval and "#line" now works. [561]
       o   Fixed numerous memory leaks, especially in eval "".
       o   Localised tied variables no longer leak memory
               use Tie::Hash;
               tie my %tied_hash => 'Tie::StdHash';
               ...
               # Used to leak memory every time local() was called;
               # in a loop, this added up.
               local($tied_hash{Foo}) = 1;
       o   Localised hash elements (and %ENV) are correctly unlocalised to not
           exist, if they didn't before they were localised.
               use Tie::Hash;
               tie my %tied_hash => 'Tie::StdHash';
               ...
               # Nothing has set the FOO element so far
               { local $tied_hash{FOO} = 'Bar' }
               # This used to print, but not now.
               print "exists!\n" if exists $tied_hash{FOO};
           As a side effect of this fix, tied hash interfaces must define the
           EXISTS and DELETE methods.
       o   mkdir() now ignores trailing slashes in the directory name, as
           mandated by POSIX.
       o   Some versions of glibc have a broken modfl().  This affects builds
           with "-Duselongdouble".  This version of Perl detects this
           brokenness and has a workaround for it.  The glibc release 2.2.2 is
           known to have fixed the modfl() bug.
       o   Modulus of unsigned numbers now works (4063328477 % 65535 used to
           return 27406, instead of 27047). [561]
       o   Some "not a number" warnings introduced in 5.6.0 eliminated to be
           more compatible with 5.005.  Infinity is now recognised as a
           number. [561]
       o   Numeric conversions did not recognize changes in the string value
           properly in certain circumstances. [561]
       o   Attributes (such as :shared) didn't work with our().
       o   our() variables will not cause bogus "Variable will not stay
           shared" warnings. [561]
       o   "our" variables of the same name declared in two sibling blocks
           resulted in bogus warnings about "redeclaration" of the variables.
           The problem has been corrected. [561]
       o   pack "Z" now correctly terminates the string with "\0".
       o   Fix password routines which in some shadow password platforms (e.g.
           HP-UX) caused getpwent() to return every other entry.
       o   The PERL5OPT environment variable (for passing command line
           arguments to Perl) didn't work for more than a single group of
           options. [561]
       o   PERL5OPT with embedded spaces didn't work.
       o   printf() no longer resets the numeric locale to "C".
       o   "qw(a\\b)" now parses correctly as 'a\\b': that is, as three
           characters, not four. [561]
       o   pos() did not return the correct value within s///ge in earlier
           versions.  This is now handled correctly. [561]
       o   Printing quads (64-bit integers) with printf/sprintf now works
           without the q L ll prefixes (assuming you are on a quad-capable
           platform).
       o   Regular expressions on references and overloaded scalars now work.
           [561+]
       o   Right-hand side magic (GMAGIC) could in many cases such as string
           concatenation be invoked too many times.
       o   scalar() now forces scalar context even when used in void context.
       o   SOCKS support is now much more robust.
       o   sort() arguments are now compiled in the right wantarray context
           (they were accidentally using the context of the sort() itself).
           The comparison block is now run in scalar context, and the
           arguments to be sorted are always provided list context. [561]
       o   Changed the POSIX character class "[[:space:]]" to include the
           (very rarely used) vertical tab character.  Added a new POSIX-ish
           character class "[[:blank:]]" which stands for horizontal
           whitespace (currently, the space and the tab).
       o   The tainting behaviour of sprintf() has been rationalized.  It does
           not taint the result of floating point formats anymore, making the
           behaviour consistent with that of string interpolation. [561]
       o   Some cases of inconsistent taint propagation (such as within hash
           values) have been fixed.
       o   The RE engine found in Perl 5.6.0 accidentally pessimised certain
           kinds of simple pattern matches.  These are now handled better.
           [561]
       o   Regular expression debug output (whether through "use re 'debug'"
           or via "-Dr") now looks better. [561]
       o   Multi-line matches like ""a\nxb\n" =~ /(?!\A)x/m" were flawed.  The
           bug has been fixed. [561]
       o   Use of $& could trigger a core dump under some situations.  This is
           now avoided. [561]
       o   The regular expression captured submatches ($1, $2, ...) are now
           more consistently unset if the match fails, instead of leaving
           false data lying around in them. [561]
       o   readline() on files opened in "slurp" mode could return an extra ""
           (blank line) at the end in certain situations.  This has been
           corrected. [561]
       o   Autovivification of symbolic references of special variables
           described in perlvar (as in "${$num}") was accidentally disabled.
           This works again now. [561]
       o   Sys::Syslog ignored the "LOG_AUTH" constant.
       o   $AUTOLOAD, sort(), lock(), and spawning subprocesses in multiple
           threads simultaneously are now thread-safe.
       o   Tie::Array's SPLICE method was broken.
       o   Allow a read-only string on the left-hand side of a non-modifying
           tr///.
       o   If "STDERR" is tied, warnings caused by "warn" and "die" now
           correctly pass to it.
       o   Several Unicode fixes.
           o       BOMs (byte order marks) at the beginning of Perl files
                   (scripts, modules) should now be transparently skipped.
                   UTF-16 and UCS-2 encoded Perl files should now be read
                   correctly.
           o       The character tables have been updated to Unicode 3.2.0.
           o       Comparing with utf8 data does not magically upgrade
                   non-utf8 data into utf8.  (This was a problem for example
                   if you were mixing data from I/O and Unicode data: your
                   output might have got magically encoded as UTF-8.)
           o       Generating illegal Unicode code points such as U+FFFE, or
                   the UTF-16 surrogates, now also generates an optional
                   warning.
           o       "IsAlnum", "IsAlpha", and "IsWord" now match titlecase.
           o       Concatenation with the "." operator or via variable
                   interpolation, "eq", "substr", "reverse", "quotemeta", the
                   "x" operator, substitution with "s///", single-quoted
                   UTF-8, should now work.
           o       The "tr///" operator now works.  Note that the "tr///CU"
                   functionality has been removed (but see pack('U0', ...)).
           o       "eval "v200"" now works.
           o       Perl 5.6.0 parsed m/\x{ab}/ incorrectly, leading to
                   spurious warnings.  This has been corrected. [561]
           o       Zero entries were missing from the Unicode classes such as
                   "IsDigit".
       o   Large unsigned numbers (those above 2**31) could sometimes lose
           their unsignedness, causing bogus results in arithmetic operations.
           [561]
       o   The Perl parser has been stress tested using both random input and
           Markov chain input and the few found crashes and lockups have been
           fixed.
   Platform Specific Changes and Fixes
       o   BSDI 4.*
           Perl now works on post-4.0 BSD/OSes.
       o   All BSDs
           Setting $0 now works (as much as possible; see perlvar for
           details).
       o   Cygwin
           Numerous updates; currently synchronised with Cygwin 1.3.10.
       o   Previously DYNIX/ptx had problems in its Configure probe for non-
           blocking I/O.
       o   EPOC
           EPOC now better supported.  See README.epoc. [561]
       o   FreeBSD 3.*
           Perl now works on post-3.0 FreeBSDs.
       o   HP-UX
           README.hpux updated; "Configure -Duse64bitall" now works; now uses
           HP-UX malloc instead of Perl malloc.
       o   IRIX
           Numerous compilation flag and hint enhancements; accidental mixing
           of 32-bit and 64-bit libraries (a doomed attempt) made much harder.
       o   Linux
           o       Long doubles should now work (see INSTALL). [561]
           o       Linux previously had problems related to sockaddrlen when
                   using accept(), recvfrom() (in Perl: recv()),
                   getpeername(), and getsockname().
       o   Mac OS Classic
           Compilation of the standard Perl distribution in Mac OS Classic
           should now work if you have the Metrowerks development environment
           and the missing Mac-specific toolkit bits.  Contact the macperl
           mailing list for details.
       o   MPE/iX
           MPE/iX update after Perl 5.6.0.  See README.mpeix. [561]
       o   NetBSD/threads: try installing the GNU pth (should be in the
           packages collection, or http://www.gnu.org/software/pth/), and
           Configure with -Duseithreads.
       o   NetBSD/sparc
           Perl now works on NetBSD/sparc.
       o   OS/2
           Now works with usethreads (see INSTALL). [561]
       o   Solaris
           64-bitness using the Sun Workshop compiler now works.
       o   Stratus VOS
           The native build method requires at least VOS Release 14.5.0 and
           GNU C++/GNU Tools 2.0.1 or later.  The Perl pack function now maps
           overflowed values to +infinity and underflowed values to -infinity.
       o   Tru64 (aka Digital UNIX, aka DEC OSF/1)
           The operating system version letter now recorded in
           $Config{osvers}.  Allow compiling with gcc (previously explicitly
           forbidden).  Compiling with gcc still not recommended because buggy
           code results, even with gcc 2.95.2.
       o   Unicos
           Fixed various alignment problems that lead into core dumps either
           during build or later; no longer dies on math errors at runtime;
           now using full quad integers (64 bits), previously was using only
           46 bit integers for speed.
       o   VMS
           See "Socket Extension Dynamic in VMS" and "IEEE-format Floating
           Point Default on OpenVMS Alpha" for important changes not otherwise
           listed here.
           chdir() now works better despite a CRT bug; now works with
           MULTIPLICITY (see INSTALL); now works with Perl's malloc.
           The tainting of %ENV elements via "keys" or "values" was previously
           unimplemented.  It now works as documented.
           The "waitpid" emulation has been improved.  The worst bug (now
           fixed) was that a pid of -1 would cause a wildcard search of all
           processes on the system.
           POSIX-style signals are now emulated much better on VMS versions
           prior to 7.0.
           The "system" function and backticks operator have improved
           functionality and better error handling. [561]
           File access tests now use current process privileges rather than
           the user's default privileges, which could sometimes result in a
           mismatch between reported access and actual access.  This
           improvement is only available on VMS v6.0 and later.
           There is a new "kill" implementation based on "sys$sigprc" that
           allows older VMS systems (pre-7.0) to use "kill" to send signals
           rather than simply force exit.  This implementation also allows
           later systems to call "kill" from within a signal handler.
           Iterative logical name translations are now limited to 10
           iterations in imitation of SHOW LOGICAL and other OpenVMS
           facilities.
       o   Windows
           o       Signal handling now works better than it used to.  It is
                   now implemented using a Windows message loop, and is
                   therefore less prone to random crashes.
           o       fork() emulation is now more robust, but still continues to
                   have a few esoteric bugs and caveats.  See perlfork for
                   details. [561+]
           o       A failed (pseudo)fork now returns undef and sets errno to
                   EAGAIN. [561]
           o       The following modules now work on Windows:
                       ExtUtils::Embed         [561]
                       IO::Pipe
                       IO::Poll
                       Net::Ping
           o       IO::File::new_tmpfile() is no longer limited to 32767
                   invocations per-process.
           o       Better chdir() return value for a non-existent directory.
           o       Compiling perl using the 64-bit Platform SDK tools is now
                   supported.
           o       The Win32::SetChildShowWindow() builtin can be used to
                   control the visibility of windows created by child
                   processes.  See Win32 for details.
           o       Non-blocking waits for child processes (or pseudo-
                   processes) are supported via "waitpid($pid,
                   &POSIX::WNOHANG)".
           o       The behavior of system() with multiple arguments has been
                   rationalized.  Each unquoted argument will be automatically
                   quoted to protect whitespace, and any existing whitespace
                   in the arguments will be preserved.  This improves the
                   portability of system(@args) by avoiding the need for
                   Windows "cmd" shell specific quoting in perl programs.
                   Note that this means that some scripts that may have relied
                   on earlier buggy behavior may no longer work correctly.
                   For example, "system("nmake /nologo", @args)" will now
                   attempt to run the file "nmake /nologo" and will fail when
                   such a file isn't found.  On the other hand, perl will now
                   execute code such as "system("c:/Program
                   Files/MyApp/foo.exe", @args)" correctly.
           o       The perl header files no longer suppress common warnings
                   from the Microsoft Visual C++ compiler.  This means that
                   additional warnings may now show up when compiling XS code.
           o       Borland C++ v5.5 is now a supported compiler that can build
                   Perl.  However, the generated binaries continue to be
                   incompatible with those generated by the other supported
                   compilers (GCC and Visual C++). [561]
           o       Duping socket handles with open(F, ">&MYSOCK") now works
                   under Windows 9x.  [561]
           o       Current directory entries in %ENV are now correctly
                   propagated to child processes. [561]
           o       New %ENV entries now propagate to subprocesses. [561]
           o       Win32::GetCwd() correctly returns C:\ instead of C: when at
                   the drive root.  Other bugs in chdir() and Cwd::cwd() have
                   also been fixed. [561]
           o       The makefiles now default to the features enabled in
                   ActiveState ActivePerl (a popular Win32 binary
                   distribution). [561]
           o       HTML files will now be installed in c:\perl\html instead of
                   c:\perl\lib\pod\html
           o       REG_EXPAND_SZ keys are now allowed in registry settings
                   used by perl. [561]
           o       Can now send() from all threads, not just the first one.
                   [561]
           o       ExtUtils::MakeMaker now uses $ENV{LIB} to search for
                   libraries. [561]
           o       Less stack reserved per thread so that more threads can run
                   concurrently. (Still 16M per thread.) [561]
           o       "File::Spec->tmpdir()" now prefers C:/temp over /tmp (works
                   better when perl is running as service).
           o       Better UNC path handling under ithreads. [561]
           o       wait(), waitpid(), and backticks now return the correct
                   exit status under Windows 9x. [561]
           o       A socket handle leak in accept() has been fixed. [561]
New or Changed Diagnostics
       Please see perldiag for more details.
       o   Ambiguous range in the transliteration operator (like a-z-9) now
           gives a warning.
       o   chdir("") and chdir(undef) now give a deprecation warning because
           they cause a possible unintentional chdir to the home directory.
           Say chdir() if you really mean that.
       o   Two new debugging options have been added: if you have compiled
           your Perl with debugging, you can use the -DT [561] and -DR options
           to trace tokenising and to add reference counts to displaying
           variables, respectively.
       o   The lexical warnings category "deprecated" is no longer a sub-
           category of the "syntax" category. It is now a top-level category
           in its own right.
       o   Unadorned dump() will now give a warning suggesting to use explicit
           CORE::dump() if that's what really is meant.
       o   The "Unrecognized escape" warning has been extended to include
           "\8", "\9", and "\_".  There is no need to escape any of the "\w"
           characters.
       o   All regular expression compilation error messages are now hopefully
           easier to understand both because the error message now comes
           before the failed regex and because the point of failure is now
           clearly marked by a "<-- HERE" marker.
       o   Various I/O (and socket) functions like binmode(), close(), and so
           forth now more consistently warn if they are used illogically
           either on a yet unopened or on an already closed filehandle (or
           socket).
       o   Using lstat() on a filehandle now gives a warning.  (It's a non-
           sensical thing to do.)
       o   The "-M" and "-m" options now warn if you didn't supply the module
           name.
       o   If you in "use" specify a required minimum version, modules
           matching the name and but not defining a $VERSION will cause a
           fatal failure.
       o   Using negative offset for vec() in lvalue context is now a warnable
           offense.
       o   Odd number of arguments to overload::constant now elicits a
           warning.
       o   Odd number of elements in anonymous hash now elicits a warning.
       o   The various "opened only for", "on closed", "never opened" warnings
           drop the "main::" prefix for filehandles in the "main" package, for
           example "STDIN" instead of "main::STDIN".
       o   Subroutine prototypes are now checked more carefully, you may get
           warnings for example if you have used non-prototype characters.
       o   If an attempt to use a (non-blessed) reference as an array index is
           made, a warning is given.
       o   "push @a;" and "unshift @a;" (with no values to push or unshift)
           now give a warning.  This may be a problem for generated and
           eval'ed code.
       o   If you try to "pack" in perlfunc a number less than 0 or larger
           than 255 using the "C" format you will get an optional warning.
           Similarly for the "c" format and a number less than -128 or more
           than 127.
       o   pack "P" format now demands an explicit size.
       o   unpack "w" now warns of unterminated compressed integers.
       o   Warnings relating to the use of PerlIO have been added.
       o   Certain regex modifiers such as "(?o)" make sense only if applied
           to the entire regex.  You will get an optional warning if you try
           to do otherwise.
       o   Variable length lookbehind has not yet been implemented, trying to
           use it will tell that.
       o   Using arrays or hashes as references (e.g. "%foo->{bar}" has been
           deprecated for a while.  Now you will get an optional warning.
       o   Warnings relating to the use of the new restricted hashes feature
           have been added.
       o   Self-ties of arrays and hashes are not supported and fatal errors
           will happen even at an attempt to do so.
       o   Using "sort" in scalar context now issues an optional warning.
           This didn't do anything useful, as the sort was not performed.
       o   Using the /g modifier in split() is meaningless and will cause a
           warning.
       o   Using splice() past the end of an array now causes a warning.
       o   Malformed Unicode encodings (UTF-8 and UTF-16) cause a lot of
           warnings, as does trying to use UTF-16 surrogates (which are
           unimplemented).
       o   Trying to use Unicode characters on an I/O stream without marking
           the stream's encoding (using open() or binmode()) will cause "Wide
           character" warnings.
       o   Use of v-strings in use/require causes a (backward) portability
           warning.
       o   Warnings relating to the use interpreter threads and their shared
           data have been added.
Changed Internals
       o   PerlIO is now the default.
       o   perlapi.pod (a companion to perlguts) now attempts to document the
           internal API.
       o   You can now build a really minimal perl called microperl.  Building
           microperl does not require even running Configure; "make -f
           Makefile.micro" should be enough.  Beware: microperl makes many
           assumptions, some of which may be too bold; the resulting
           executable may crash or otherwise misbehave in wondrous ways.  For
           careful hackers only.
       o   Added rsignal(), whichsig(), do_join(), op_clear, op_null,
           ptr_table_clear(), ptr_table_free(), sv_setref_uv(), and several
           UTF-8 interfaces to the publicised API.  For the full list of the
           available APIs see perlapi.
       o   Made possible to propagate customised exceptions via croak()ing.
       o   Now xsubs can have attributes just like subs.  (Well, at least the
           built-in attributes.)
       o   dTHR and djSP have been obsoleted; the former removed (because it's
           a no-op) and the latter replaced with dSP.
       o   PERL_OBJECT has been completely removed.
       o   The MAGIC constants (e.g. 'P') have been macrofied (e.g.
           "PERL_MAGIC_TIED") for better source code readability and
           maintainability.
       o   The regex compiler now maintains a structure that identifies nodes
           in the compiled bytecode with the corresponding syntactic features
           of the original regex expression.  The information is attached to
           the new "offsets" member of the "struct regexp". See perldebguts
           for more complete information.
       o   The C code has been made much more "gcc -Wall" clean.  Some warning
           messages still remain in some platforms, so if you are compiling
           with gcc you may see some warnings about dubious practices.  The
           warnings are being worked on.
       o   perly.c, sv.c, and sv.h have now been extensively commented.
       o   Documentation on how to use the Perl source repository has been
           added to Porting/repository.pod.
       o   There are now several profiling make targets.
Security Vulnerability Closed [561]
       (This change was already made in 5.7.0 but bears repeating here.)
       (5.7.0 came out before 5.6.1: the development branch 5.7 released
       earlier than the maintenance branch 5.6)
       A potential security vulnerability in the optional suidperl component
       of Perl was identified in August 2000.  suidperl is neither built nor
       installed by default.  As of November 2001 the only known vulnerable
       platform is Linux, most likely all Linux distributions.  CERT and
       various vendors and distributors have been alerted about the
       vulnerability.  See
       http://www.cpan.org/src/5.0/sperl-2000-08-05/sperl-2000-08-05.txt for
       more information.
       The problem was caused by Perl trying to report a suspected security
       exploit attempt using an external program, /bin/mail.  On Linux
       platforms the /bin/mail program had an undocumented feature which when
       combined with suidperl gave access to a root shell, resulting in a
       serious compromise instead of reporting the exploit attempt.  If you
       don't have /bin/mail, or if you have 'safe setuid scripts', or if
       suidperl is not installed, you are safe.
       The exploit attempt reporting feature has been completely removed from
       Perl 5.8.0 (and the maintenance release 5.6.1, and it was removed also
       from all the Perl 5.7 releases), so that particular vulnerability isn't
       there anymore.  However, further security vulnerabilities are,
       unfortunately, always possible.  The suidperl functionality is most
       probably going to be removed in Perl 5.10.  In any case, suidperl
       should only be used by security experts who know exactly what they are
       doing and why they are using suidperl instead of some other solution
       such as sudo ( see http://www.courtesan.com/sudo/ ).
New Tests
       Several new tests have been added, especially for the lib and ext
       subsections.  There are now about 69 000 individual tests (spread over
       about 700 test scripts), in the regression suite (5.6.1 has about 11
       700 tests, in 258 test scripts)  The exact numbers depend on the
       platform and Perl configuration used.  Many of the new tests are of
       course introduced by the new modules, but still in general Perl is now
       more thoroughly tested.
       Because of the large number of tests, running the regression suite will
       take considerably longer time than it used to: expect the suite to take
       up to 4-5 times longer to run than in perl 5.6.  On a really fast
       machine you can hope to finish the suite in about 6-8 minutes
       (wallclock time).
       The tests are now reported in a different order than in earlier Perls.
       (This happens because the test scripts from under t/lib have been moved
       to be closer to the library/extension they are testing.)
Known Problems
   The Compiler Suite Is Still Very Experimental
       The compiler suite is slowly getting better but it continues to be
       highly experimental.  Use in production environments is discouraged.
   Localising Tied Arrays and Hashes Is Broken
           local %tied_array;
       doesn't work as one would expect: the old value is restored
       incorrectly.  This will be changed in a future release, but we don't
       know yet what the new semantics will exactly be.  In any case, the
       change will break existing code that relies on the current (ill-
       defined) semantics, so just avoid doing this in general.
   Building Extensions Can Fail Because Of Largefiles
       Some extensions like mod_perl are known to have issues with
       `largefiles', a change brought by Perl 5.6.0 in which file offsets
       default to 64 bits wide, where supported.  Modules may fail to compile
       at all, or they may compile and work incorrectly.  Currently, there is
       no good solution for the problem, but Configure now provides
       appropriate non-largefile ccflags, ldflags, libswanted, and libs in the
       %Config hash (e.g., $Config{ccflags_nolargefiles}) so the extensions
       that are having problems can try configuring themselves without the
       largefileness.  This is admittedly not a clean solution, and the
       solution may not even work at all.  One potential failure is whether
       one can (or, if one can, whether it's a good idea to) link together at
       all binaries with different ideas about file offsets; all this is
       platform-dependent.
   Modifying $_ Inside for(..)
          for (1..5) { $_++ }
       works without complaint.  It shouldn't.  (You should be able to modify
       only lvalue elements inside the loops.)  You can see the correct
       behaviour by replacing the 1..5 with 1, 2, 3, 4, 5.
   mod_perl 1.26 Doesn't Build With Threaded Perl
       Use mod_perl 1.27 or higher.
   lib/ftmp-security tests warn 'system possibly insecure'
       Don't panic.  Read the 'make test' section of INSTALL instead.
   libwww-perl (LWP) fails base/date #51
       Use libwww-perl 5.65 or later.
   PDL failing some tests
       Use PDL 2.3.4 or later.
   Perl_get_sv
       You may get errors like 'Undefined symbol "Perl_get_sv"' or "can't
       resolve symbol 'Perl_get_sv'", or the symbol may be "Perl_sv_2pv".
       This probably means that you are trying to use an older shared Perl
       library (or extensions linked with such) with Perl 5.8.0 executable.
       Perl used to have such a subroutine, but that is no more the case.
       Check your shared library path, and any shared Perl libraries in those
       directories.
       Sometimes this problem may also indicate a partial Perl 5.8.0
       installation, see "Mac OS X dyld undefined symbols" for an example and
       how to deal with it.
   Self-tying Problems
       Self-tying of arrays and hashes is broken in rather deep and hard-to-
       fix ways.  As a stop-gap measure to avoid people from getting
       frustrated at the mysterious results (core dumps, most often), it is
       forbidden for now (you will get a fatal error even from an attempt).
       A change to self-tying of globs has caused them to be recursively
       referenced (see: "Two-Phased Garbage Collection" in perlobj).  You will
       now need an explicit untie to destroy a self-tied glob.  This behaviour
       may be fixed at a later date.
       Self-tying of scalars and IO thingies works.
   ext/threads/t/libc
       If this test fails, it indicates that your libc (C library) is not
       threadsafe.  This particular test stress tests the localtime() call to
       find out whether it is threadsafe.  See perlthrtut for more
       information.
   Failure of Thread (5.005-style) tests
       Note that support for 5.005-style threading is deprecated, experimental
       and practically unsupported.  In 5.10, it is expected to be removed.
       You should migrate your code to ithreads.
       The following tests are known to fail due to fundamental problems in
       the 5.005 threading implementation. These are not new failures--Perl
       5.005_0x has the same bugs, but didn't have these tests.
        ../ext/B/t/xref.t                    255 65280    14   12  85.71%  3-14
        ../ext/List/Util/t/first.t           255 65280     7    4  57.14%  2 5-7
        ../lib/English.t                       2   512    54    2   3.70%  2-3
        ../lib/FileCache.t                                 5    1  20.00%  5
        ../lib/Filter/Simple/t/data.t                      6    3  50.00%  1-3
        ../lib/Filter/Simple/t/filter_only.                9    3  33.33%  1-2 5
        ../lib/Math/BigInt/t/bare_mbf.t                 1627    4   0.25%  8 11 1626-1627
        ../lib/Math/BigInt/t/bigfltpm.t                 1629    4   0.25%  10 13 1628-
                                                                           1629
        ../lib/Math/BigInt/t/sub_mbf.t                  1633    4   0.24%  8 11 1632-1633
        ../lib/Math/BigInt/t/with_sub.t                 1628    4   0.25%  9 12 1627-1628
        ../lib/Tie/File/t/31_autodefer.t     255 65280    65   32  49.23%  34-65
        ../lib/autouse.t                                  10    1  10.00%  4
        op/flip.t                                         15    1   6.67%  15
       These failures are unlikely to get fixed as 5.005-style threads are
       considered fundamentally broken.  (Basically what happens is that
       competing threads can corrupt shared global state, one good example
       being regular expression engine's state.)
   Timing problems
       The following tests may fail intermittently because of timing problems,
       for example if the system is heavily loaded.
           t/op/alarm.t
           ext/Time/HiRes/HiRes.t
           lib/Benchmark.t
           lib/Memoize/t/expmod_t.t
           lib/Memoize/t/speed.t
       In case of failure please try running them manually, for example
           ./perl -Ilib ext/Time/HiRes/HiRes.t
   Tied/Magical Array/Hash Elements Do Not Autovivify
       For normal arrays "$foo = \$bar[1]" will assign "undef" to $bar[1]
       (assuming that it didn't exist before), but for tied/magical arrays and
       hashes such autovivification does not happen because there is currently
       no way to catch the reference creation.  The same problem affects
       slicing over non-existent indices/keys of a tied/magical array/hash.
   Unicode in package/class and subroutine names does not work
       One can have Unicode in identifier names, but not in package/class or
       subroutine names.  While some limited functionality towards this does
       exist as of Perl 5.8.0, that is more accidental than designed; use of
       Unicode for the said purposes is unsupported.
       One reason of this unfinishedness is its (currently) inherent
       unportability: since both package names and subroutine names may need
       to be mapped to file and directory names, the Unicode capability of the
       filesystem becomes important-- and there unfortunately aren't portable
       answers.
Platform Specific Problems
   AIX
       o   If using the AIX native make command, instead of just "make" issue
           "make all".  In some setups the former has been known to spuriously
           also try to run "make install".  Alternatively, you may want to use
           GNU make.
       o   In AIX 4.2, Perl extensions that use C++ functions that use statics
           may have problems in that the statics are not getting initialized.
           In newer AIX releases, this has been solved by linking Perl with
           the libC_r library, but unfortunately in AIX 4.2 the said library
           has an obscure bug where the various functions related to time
           (such as time() and gettimeofday()) return broken values, and
           therefore in AIX 4.2 Perl is not linked against libC_r.
       o   vac 5.0.0.0 May Produce Buggy Code For Perl
           The AIX C compiler vac version 5.0.0.0 may produce buggy code,
           resulting in a few random tests failing when run as part of "make
           test", but when the failing tests are run by hand, they succeed.
           We suggest upgrading to at least vac version 5.0.1.0, that has been
           known to compile Perl correctly.  "lslpp -L|grep vac.C" will tell
           you the vac version.  See README.aix.
       o   If building threaded Perl, you may get compilation warning from
           pp_sys.c:
             "pp_sys.c", line 4651.39: 1506-280 (W) Function argument assignment between types "unsigned char*" and "const void*" is not allowed.
           This is harmless; it is caused by the getnetbyaddr() and
           getnetbyaddr_r() having slightly different types for their first
           argument.
   Alpha systems with old gccs fail several tests
       If you see op/pack, op/pat, op/regexp, or ext/Storable tests failing in
       a Linux/alpha or *BSD/Alpha, it's probably time to upgrade your gcc.
       gccs prior to 2.95.3 are definitely not good enough, and gcc 3.1 may be
       even better.  (RedHat Linux/alpha with gcc 3.1 reported no problems, as
       did Linux 2.4.18 with gcc 2.95.4.)  (In Tru64, it is preferable to use
       the bundled C compiler.)
   AmigaOS
       Perl 5.8.0 doesn't build in AmigaOS.  It broke at some point during the
       ithreads work and we could not find Amiga experts to unbreak the
       problems.  Perl 5.6.1 still works for AmigaOS (as does the 5.7.2
       development release).
   BeOS
       The following tests fail on 5.8.0 Perl in BeOS Personal 5.03:
        t/op/lfs............................FAILED at test 17
        t/op/magic..........................FAILED at test 24
        ext/Fcntl/t/syslfs..................FAILED at test 17
        ext/File/Glob/t/basic...............FAILED at test 3
        ext/POSIX/t/sigaction...............FAILED at test 13
        ext/POSIX/t/waitpid.................FAILED at test 1
       See perlbeos (README.beos) for more details.
   Cygwin "unable to remap"
       For example when building the Tk extension for Cygwin, you may get an
       error message saying "unable to remap".  This is known problem with
       Cygwin, and a workaround is detailed in here:
       http://sources.redhat.com/ml/cygwin/2001-12/msg00894.html
   Cygwin ndbm tests fail on FAT
       One can build but not install (or test the build of) the NDBM_File on
       FAT filesystems.  Installation (or build) on NTFS works fine.  If one
       attempts the test on a FAT install (or build) the following failures
       are expected:
        ../ext/NDBM_File/ndbm.t       13  3328    71   59  83.10%  1-2 4 16-71
        ../ext/ODBM_File/odbm.t      255 65280    ??   ??       %  ??
        ../lib/AnyDBM_File.t           2   512    12    2  16.67%  1 4
        ../lib/Memoize/t/errors.t      0   139    11    5  45.45%  7-11
        ../lib/Memoize/t/tie_ndbm.t   13  3328     4    4 100.00%  1-4
        run/fresh_perl.t                          97    1   1.03%  91
       NDBM_File fails and ODBM_File just coredumps.
       If you intend to run only on FAT (or if using AnyDBM_File on FAT), run
       Configure with the -Ui_ndbm and -Ui_dbm options to prevent NDBM_File
       and ODBM_File being built.
   DJGPP Failures
        t/op/stat............................FAILED at test 29
        lib/File/Find/t/find.................FAILED at test 1
        lib/File/Find/t/taint................FAILED at test 1
        lib/h2xs.............................FAILED at test 15
        lib/Pod/t/eol........................FAILED at test 1
        lib/Test/Harness/t/strap-analyze.....FAILED at test 8
        lib/Test/Harness/t/test-harness......FAILED at test 23
        lib/Test/Simple/t/exit...............FAILED at test 1
       The above failures are known as of 5.8.0 with native builds with long
       filenames, but there are a few more if running under dosemu because of
       limitations (and maybe bugs) of dosemu:
        t/comp/cpp...........................FAILED at test 3
        t/op/inccode.........................(crash)
       and a few lib/ExtUtils tests, and several hundred Encode/t/Aliases.t
       failures that work fine with long filenames.  So you really might
       prefer native builds and long filenames.
   FreeBSD built with ithreads coredumps reading large directories
       This is a known bug in FreeBSD 4.5's readdir_r(), it has been fixed in
       FreeBSD 4.6 (see perlfreebsd (README.freebsd)).
   FreeBSD Failing locale Test 117 For ISO 8859-15 Locales
       The ISO 8859-15 locales may fail the locale test 117 in FreeBSD.  This
       is caused by the characters \xFF (y with diaeresis) and \xBE (Y with
       diaeresis) not behaving correctly when being matched case-
       insensitively.  Apparently this problem has been fixed in the latest
       FreeBSD releases.  ( http://www.freebsd.org/cgi/query-pr.cgi?pr=34308 )
   IRIX fails ext/List/Util/t/shuffle.t or Digest::MD5
       IRIX with MIPSpro 7.3.1.2m or 7.3.1.3m compiler may fail the List::Util
       test ext/List/Util/t/shuffle.t by dumping core.  This seems to be a
       compiler error since if compiled with gcc no core dump ensues, and no
       failures have been seen on the said test on any other platform.
       Similarly, building the Digest::MD5 extension has been known to fail
       with "*** Termination code 139 (bu21)".
       The cure is to drop optimization level (Configure -Doptimize=-O2).
   HP-UX lib/posix Subtest 9 Fails When LP64-Configured
       If perl is configured with -Duse64bitall, the successful result of the
       subtest 10 of lib/posix may arrive before the successful result of the
       subtest 9, which confuses the test harness so much that it thinks the
       subtest 9 failed.
   Linux with glibc 2.2.5 fails t/op/int subtest #6 with -Duse64bitint
       This is a known bug in the glibc 2.2.5 with long long integers.  (
       http://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=65612 )
   Linux With Sfio Fails op/misc Test 48
       No known fix.
   Mac OS X
       Please remember to set your environment variable LC_ALL to "C" (setenv
       LC_ALL C) before running "make test" to avoid a lot of warnings about
       the broken locales of Mac OS X.
       The following tests are known to fail in Mac OS X 10.1.5 because of
       buggy (old) implementations of Berkeley DB included in Mac OS X:
        Failed Test                 Stat Wstat Total Fail  Failed  List of Failed
        -------------------------------------------------------------------------
        ../ext/DB_File/t/db-btree.t    0    11    ??   ??       %  ??
        ../ext/DB_File/t/db-recno.t              149    3   2.01%  61 63 65
       If you are building on a UFS partition, you will also probably see
       t/op/stat.t subtest #9 fail.  This is caused by Darwin's UFS not
       supporting inode change time.
       Also the ext/POSIX/t/posix.t subtest #10 fails but it is skipped for
       now because the failure is Apple's fault, not Perl's (blocked signals
       are lost).
       If you Configure with ithreads, ext/threads/t/libc.t will fail. Again,
       this is not Perl's fault-- the libc of Mac OS X is not threadsafe (in
       this particular test, the localtime() call is found to be
       threadunsafe.)
   Mac OS X dyld undefined symbols
       If after installing Perl 5.8.0 you are getting warnings about missing
       symbols, for example
           dyld: perl Undefined symbols
           _perl_sv_2pv
           _perl_get_sv
       you probably have an old pre-Perl-5.8.0 installation (or parts of one)
       in /Library/Perl (the undefined symbols used to exist in pre-5.8.0
       Perls).  It seems that for some reason "make install" doesn't always
       completely overwrite the files in /Library/Perl.  You can move the old
       Perl shared library out of the way like this:
           cd /Library/Perl/darwin/CORE
           mv libperl.dylib libperlold.dylib
       and then reissue "make install".  Note that the above of course is
       extremely disruptive for anything using the /usr/local/bin/perl.  If
       that doesn't help, you may have to try removing all the .bundle files
       from beneath /Library/Perl, and again "make install"-ing.
   OS/2 Test Failures
       The following tests are known to fail on OS/2 (for clarity only the
       failures are shown, not the full error messages):
        ../lib/ExtUtils/t/Mkbootstrap.t    1   256    18    1   5.56%  8
        ../lib/ExtUtils/t/Packlist.t       1   256    34    1   2.94%  17
        ../lib/ExtUtils/t/basic.t          1   256    17    1   5.88%  14
        lib/os2_process.t                  2   512   227    2   0.88%  174 209
        lib/os2_process_kid.t                        227    2   0.88%  174 209
        lib/rx_cmprt.t                   255 65280    18    3  16.67%  16-18
   op/sprintf tests 91, 129, and 130
       The op/sprintf tests 91, 129, and 130 are known to fail on some
       platforms.  Examples include any platform using sfio, and
       Compaq/Tandem's NonStop-UX.
       Test 91 is known to fail on QNX6 (nto), because "sprintf '%e',0"
       incorrectly produces 0.000000e+0 instead of 0.000000e+00.
       For tests 129 and 130, the failing platforms do not comply with the
       ANSI C Standard: lines 19ff on page 134 of ANSI X3.159 1989, to be
       exact.  (They produce something other than "1" and "-1" when formatting
       0.6 and -0.6 using the printf format "%.0f"; most often, they produce
       "0" and "-0".)
   SCO
       The socketpair tests are known to be unhappy in SCO 3.2v5.0.4:
        ext/Socket/socketpair.t...............FAILED tests 15-45
   Solaris 2.5
       In case you are still using Solaris 2.5 (aka SunOS 5.5), you may
       experience failures (the test core dumping) in lib/locale.t.  The
       suggested cure is to upgrade your Solaris.
   Solaris x86 Fails Tests With -Duse64bitint
       The following tests are known to fail in Solaris x86 with Perl
       configured to use 64 bit integers:
        ext/Data/Dumper/t/dumper.............FAILED at test 268
        ext/Devel/Peek/Peek..................FAILED at test 7
   SUPER-UX (NEC SX)
       The following tests are known to fail on SUPER-UX:
        op/64bitint...........................FAILED tests 29-30, 32-33, 35-36
        op/arith..............................FAILED tests 128-130
        op/pack...............................FAILED tests 25-5625
        op/pow................................
        op/taint..............................# msgsnd failed
        ../ext/IO/lib/IO/t/io_poll............FAILED tests 3-4
        ../ext/IPC/SysV/ipcsysv...............FAILED tests 2, 5-6
        ../ext/IPC/SysV/t/msg.................FAILED tests 2, 4-6
        ../ext/Socket/socketpair..............FAILED tests 12
        ../lib/IPC/SysV.......................FAILED tests 2, 5-6
        ../lib/warnings.......................FAILED tests 115-116, 118-119
       The op/pack failure ("Cannot compress negative numbers at op/pack.t
       line 126") is serious but as of yet unsolved.  It points at some
       problems with the signedness handling of the C compiler, as do the
       64bitint, arith, and pow failures.  Most of the rest point at problems
       with SysV IPC.
   Term::ReadKey not working on Win32
       Use Term::ReadKey 2.20 or later.
   UNICOS/mk
       o   During Configure, the test
               Guessing which symbols your C compiler and preprocessor define...
           will probably fail with error messages like
               CC-20 cc: ERROR File = try.c, Line = 3
                 The identifier "bad" is undefined.
                 bad switch yylook 79bad switch yylook 79bad switch yylook 79bad switch yylook 79#ifdef A29K
                 ^
               CC-65 cc: ERROR File = try.c, Line = 3
                 A semicolon is expected at this point.
           This is caused by a bug in the awk utility of UNICOS/mk.  You can
           ignore the error, but it does cause a slight problem: you cannot
           fully benefit from the h2ph utility (see h2ph) that can be used to
           convert C headers to Perl libraries, mainly used to be able to
           access from Perl the constants defined using C preprocessor, cpp.
           Because of the above error, parts of the converted headers will be
           invisible.  Luckily, these days the need for h2ph is rare.
       o   If building Perl with interpreter threads (ithreads), the
           getgrent(), getgrnam(), and getgrgid() functions cannot return the
           list of the group members due to a bug in the multithreaded support
           of UNICOS/mk.  What this means is that in list context the
           functions will return only three values, not four.
   UTS
       There are a few known test failures, see perluts (README.uts).
   VOS (Stratus)
       When Perl is built using the native build process on VOS Release 14.5.0
       and GNU C++/GNU Tools 2.0.1, all attempted tests either pass or result
       in TODO (ignored) failures.
   VMS
       There should be no reported test failures with a default configuration,
       though there are a number of tests marked TODO that point to areas
       needing further debugging and/or porting work.
   Win32
       In multi-CPU boxes, there are some problems with the I/O buffering:
       some output may appear twice.
   XML::Parser not working
       Use XML::Parser 2.31 or later.
   z/OS (OS/390)
       z/OS has rather many test failures but the situation is actually much
       better than it was in 5.6.0; it's just that so many new modules and
       tests have been added.
        Failed Test                   Stat Wstat Total Fail  Failed  List of Failed
        ---------------------------------------------------------------------------
        ../ext/Data/Dumper/t/dumper.t              357    8   2.24%  311 314 325 327
                                                                     331 333 337 339
        ../ext/IO/lib/IO/t/io_unix.t                 5    4  80.00%  2-5
        ../ext/Storable/t/downgrade.t   12  3072   169   12   7.10%  14-15 46-47 78-79
                                                                     110-111 150 161
        ../lib/ExtUtils/t/Constant.t   121 30976    48   48 100.00%  1-48
        ../lib/ExtUtils/t/Embed.t                    9    9 100.00%  1-9
        op/pat.t                                   922    7   0.76%  665 776 785 832-
                                                                     834 845
        op/sprintf.t                               224    3   1.34%  98 100 136
        op/tr.t                                     97    5   5.15%  63 71-74
        uni/fold.t                                 780    6   0.77%  61 169 196 661
                                                                     710-711
       The failures in dumper.t and downgrade.t are problems in the tests,
       those in io_unix and sprintf are problems in the USS (UDP sockets and
       printf formats).  The pat, tr, and fold failures are genuine Perl
       problems caused by EBCDIC (and in the pat and fold cases, combining
       that with Unicode).  The Constant and Embed are probably problems in
       the tests (since they test Perl's ability to build extensions, and that
       seems to be working reasonably well.)
   Unicode Support on EBCDIC Still Spotty
       Though mostly working, Unicode support still has problem spots on
       EBCDIC platforms.  One such known spot are the "\p{}" and "\P{}"
       regular expression constructs for code points less than 256: the "pP"
       are testing for Unicode code points, not knowing about EBCDIC.
   Seen In Perl 5.7 But Gone Now
       "Time::Piece" (previously known as "Time::Object") was removed because
       it was felt that it didn't have enough value in it to be a core module.
       It is still a useful module, though, and is available from the CPAN.
       Perl 5.8 unfortunately does not build anymore on AmigaOS; this broke
       accidentally at some point.  Since there are not that many Amiga
       developers available, we could not get this fixed and tested in time
       for 5.8.0.  Perl 5.6.1 still works for AmigaOS (as does the 5.7.2
       development release).
       The "PerlIO::Scalar" and "PerlIO::Via" (capitalised) were renamed as
       "PerlIO::scalar" and "PerlIO::via" (all lowercase) just before 5.8.0.
       The main rationale was to have all core PerlIO layers to have all
       lowercase names.  The "plugins" are named as usual, for example
       "PerlIO::via::QuotedPrint".
       The "threads::shared::queue" and "threads::shared::semaphore" were
       renamed as "Thread::Queue" and "Thread::Semaphore" just before 5.8.0.
       The main rationale was to have thread modules to obey normal naming,
       "Thread::" (the "threads" and "threads::shared" themselves are more
       pragma-like, they affect compile-time, so they stay lowercase).
Reporting Bugs
       If you find what you think is a bug, you might check the articles
       recently posted to the comp.lang.perl.misc newsgroup and the perl bug
       database at http://bugs.perl.org/ .  There may also be information at
       http://www.perl.com/ , the Perl Home Page.
       If you believe you have an unreported bug, please run the perlbug
       program included with your release.  Be sure to trim your bug down to a
       tiny but sufficient test case.  Your bug report, along with the output
       of "perl -V", will be sent off to perlbug AT perl.org to be analysed by
       the Perl porting team.
SEE ALSO
       The Changes file for exhaustive details on what changed.
       The INSTALL file for how to build Perl.
       The README file for general stuff.
       The Artistic and Copying files for copyright information.
HISTORY
       Written by Jarkko Hietaniemi <jhi AT iki.fi>.

perl v5.16.3                      2013-03-04                    PERL58DELTA(1)