1PERL590DELTA(1) Perl Programmers Reference Guide PERL590DELTA(1)
2
3
4
6 perl590delta - what is new for perl v5.9.0
7
9 This document describes differences between the 5.8.0 release and the
10 5.9.0 release.
11
13 Hash Randomisation
14 Mainly due to security reasons, the "random ordering" of hashes has
15 been made even more random. Previously while the order of hash
16 elements from keys(), values(), and each() was essentially random, it
17 was still repeatable. Now, however, the order varies between different
18 runs of Perl.
19
20 Perl has never guaranteed any ordering of the hash keys, and the
21 ordering has already changed several times during the lifetime of Perl
22 5. Also, the ordering of hash keys has always been, and continues to
23 be, affected by the insertion order.
24
25 The added randomness may affect applications.
26
27 One possible scenario is when output of an application has included
28 hash data. For example, if you have used the Data::Dumper module to
29 dump data into different files, and then compared the files to see
30 whether the data has changed, now you will have false positives since
31 the order in which hashes are dumped will vary. In general the cure is
32 to sort the keys (or the values); in particular for Data::Dumper to use
33 the "Sortkeys" option. If some particular order is really important,
34 use tied hashes: for example the Tie::IxHash module which by default
35 preserves the order in which the hash elements were added.
36
37 More subtle problem is reliance on the order of "global destruction".
38 That is what happens at the end of execution: Perl destroys all data
39 structures, including user data. If your destructors (the DESTROY
40 subroutines) have assumed any particular ordering to the global
41 destruction, there might be problems ahead. For example, in a
42 destructor of one object you cannot assume that objects of any other
43 class are still available, unless you hold a reference to them. If the
44 environment variable PERL_DESTRUCT_LEVEL is set to a non-zero value, or
45 if Perl is exiting a spawned thread, it will also destruct the ordinary
46 references and the symbol tables that are no longer in use. You can't
47 call a class method or an ordinary function on a class that has been
48 collected that way.
49
50 The hash randomisation is certain to reveal hidden assumptions about
51 some particular ordering of hash elements, and outright bugs: it
52 revealed a few bugs in the Perl core and core modules.
53
54 To disable the hash randomisation in runtime, set the environment
55 variable PERL_HASH_SEED to 0 (zero) before running Perl (for more
56 information see "PERL_HASH_SEED" in perlrun), or to disable the feature
57 completely in compile time, compile with "-DNO_HASH_SEED" (see
58 INSTALL).
59
60 See "Algorithmic Complexity Attacks" in perlsec for the original
61 rationale behind this change.
62
63 UTF-8 On Filehandles No Longer Activated By Locale
64 In Perl 5.8.0 all filehandles, including the standard filehandles, were
65 implicitly set to be in Unicode UTF-8 if the locale settings indicated
66 the use of UTF-8. This feature caused too many problems, so the
67 feature was turned off and redesigned: see "Core Enhancements".
68
69 Single-number v-strings are no longer v-strings before "=>"
70 The version strings or v-strings (see "Version Strings" in perldata)
71 feature introduced in Perl 5.6.0 has been a source of some confusion--
72 especially when the user did not want to use it, but Perl thought it
73 knew better. Especially troublesome has been the feature that before a
74 "=>" a version string (a "v" followed by digits) has been interpreted
75 as a v-string instead of a string literal. In other words:
76
77 %h = ( v65 => 42 );
78
79 has meant since Perl 5.6.0
80
81 %h = ( 'A' => 42 );
82
83 (at least in platforms of ASCII progeny) Perl 5.8.1 restored the more
84 natural interpretation
85
86 %h = ( 'v65' => 42 );
87
88 The multi-number v-strings like v65.66 and 65.66.67 still continue to
89 be v-strings in Perl 5.8.
90
91 (Win32) The -C Switch Has Been Repurposed
92 The -C switch has changed in an incompatible way. The old semantics of
93 this switch only made sense in Win32 and only in the "use utf8"
94 universe in 5.6.x releases, and do not make sense for the Unicode
95 implementation in 5.8.0. Since this switch could not have been used by
96 anyone, it has been repurposed. The behavior that this switch enabled
97 in 5.6.x releases may be supported in a transparent, data-dependent
98 fashion in a future release.
99
100 For the new life of this switch, see "UTF-8 no longer default under
101 UTF-8 locales", and "-C" in perlrun.
102
103 (Win32) The /d Switch Of cmd.exe
104 Since version 5.8.1, perl uses the /d switch when running the cmd.exe
105 shell internally for system(), backticks, and when opening pipes to
106 external programs. The extra switch disables the execution of AutoRun
107 commands from the registry, which is generally considered undesirable
108 when running external programs. If you wish to retain compatibility
109 with the older behavior, set PERL5SHELL in your environment to "cmd
110 /x/c".
111
112 The $* variable has been removed
113 $*, which was deprecated in favor of the "/s" and "/m" regexp
114 modifiers, has been removed.
115
117 Assertions
118 Perl 5.9.0 has experimental support for assertions. Note that the user
119 interface is not fully stabilized yet, and it may change until the
120 5.10.0 release. A new command-line switch, -A, is used to activate
121 assertions, which are declared with the "assertions" pragma. See
122 assertions.
123
124 Defined-or operators
125 A new operator "//" (defined-or) has been implemented. The following
126 statement:
127
128 $a // $b
129
130 is merely equivalent to
131
132 defined $a ? $a : $b
133
134 and
135
136 $c //= $d;
137
138 can be used instead of
139
140 $c = $d unless defined $c;
141
142 This operator has the same precedence and associativity as "||". It
143 has a low-precedence counterpart, "err", which has the same precedence
144 and associativity as "or". Special care has been taken to ensure that
145 those operators Do What You Mean while not breaking old code, but some
146 edge cases involving the empty regular expression may now parse
147 differently. See perlop for details.
148
149 UTF-8 no longer default under UTF-8 locales
150 In Perl 5.8.0 many Unicode features were introduced. One of them was
151 found to be of more nuisance than benefit: the automagic (and silent)
152 "UTF-8-ification" of filehandles, including the standard filehandles,
153 if the user's locale settings indicated use of UTF-8.
154
155 For example, if you had "en_US.UTF-8" as your locale, your STDIN and
156 STDOUT were automatically "UTF-8", in other words an implicit
157 binmode(..., ":utf8") was made. This meant that trying to print, say,
158 chr(0xff), ended up printing the bytes 0xc3 0xbf. Hardly what you had
159 in mind unless you were aware of this feature of Perl 5.8.0. The
160 problem is that the vast majority of people weren't: for example in
161 RedHat releases 8 and 9 the default locale setting is UTF-8, so all
162 RedHat users got UTF-8 filehandles, whether they wanted it or not. The
163 pain was intensified by the Unicode implementation of Perl 5.8.0
164 (still) having nasty bugs, especially related to the use of s/// and
165 tr///. (Bugs that have been fixed in 5.8.1)
166
167 Therefore a decision was made to backtrack the feature and change it
168 from implicit silent default to explicit conscious option. The new
169 Perl command line option "-C" and its counterpart environment variable
170 PERL_UNICODE can now be used to control how Perl and Unicode interact
171 at interfaces like I/O and for example the command line arguments. See
172 "-C" in perlrun and "PERL_UNICODE" in perlrun for more information.
173
174 Unsafe signals again available
175 In Perl 5.8.0 the so-called "safe signals" were introduced. This means
176 that Perl no longer handles signals immediately but instead "between
177 opcodes", when it is safe to do so. The earlier immediate handling
178 easily could corrupt the internal state of Perl, resulting in
179 mysterious crashes.
180
181 However, the new safer model has its problems too. Because now an
182 opcode, a basic unit of Perl execution, is never interrupted but
183 instead let to run to completion, certain operations that can take a
184 long time now really do take a long time. For example, certain network
185 operations have their own blocking and timeout mechanisms, and being
186 able to interrupt them immediately would be nice.
187
188 Therefore perl 5.8.1 introduced a "backdoor" to restore the pre-5.8.0
189 (pre-5.7.3, really) signal behaviour. Just set the environment
190 variable PERL_SIGNALS to "unsafe", and the old immediate (and unsafe)
191 signal handling behaviour returns. See "PERL_SIGNALS" in perlrun and
192 "Deferred Signals (Safe Signals)" in perlipc.
193
194 In completely unrelated news, you can now use safe signals with
195 POSIX::SigAction. See "POSIX::SigAction" in POSIX.
196
197 Tied Arrays with Negative Array Indices
198 Formerly, the indices passed to "FETCH", "STORE", "EXISTS", and
199 "DELETE" methods in tied array class were always non-negative. If the
200 actual argument was negative, Perl would call FETCHSIZE implicitly and
201 add the result to the index before passing the result to the tied array
202 method. This behaviour is now optional. If the tied array class
203 contains a package variable named $NEGATIVE_INDICES which is set to a
204 true value, negative values will be passed to "FETCH", "STORE",
205 "EXISTS", and "DELETE" unchanged.
206
207 local ${$x}
208 The syntaxes
209
210 local ${$x}
211 local @{$x}
212 local %{$x}
213
214 now do localise variables, given that the $x is a valid variable name.
215
216 Unicode Character Database 4.0.0
217 The copy of the Unicode Character Database included in Perl 5.8 has
218 been updated to 4.0.0 from 3.2.0. This means for example that the
219 Unicode character properties are as in Unicode 4.0.0.
220
221 Miscellaneous Enhancements
222 "unpack()" now defaults to unpacking the $_.
223
224 "map" in void context is no longer expensive. "map" is now context
225 aware, and will not construct a list if called in void context.
226
227 If a socket gets closed by the server while printing to it, the client
228 now gets a SIGPIPE. While this new feature was not planned, it fell
229 naturally out of PerlIO changes, and is to be considered an accidental
230 feature.
231
232 PerlIO::get_layers(FH) returns the names of the PerlIO layers active on
233 a filehandle.
234
235 PerlIO::via layers can now have an optional UTF8 method to indicate
236 whether the layer wants to "auto-:utf8" the stream.
237
238 utf8::is_utf8() has been added as a quick way to test whether a scalar
239 is encoded internally in UTF-8 (Unicode).
240
242 Updated Modules And Pragmata
243 The following modules and pragmata have been updated since Perl 5.8.0:
244
245 base
246 B::Bytecode
247 In much better shape than it used to be. Still far from perfect,
248 but maybe worth a try.
249
250 B::Concise
251 B::Deparse
252 Benchmark
253 An optional feature, ":hireswallclock", now allows for high
254 resolution wall clock times (uses Time::HiRes).
255
256 ByteLoader
257 See B::Bytecode.
258
259 bytes
260 Now has bytes::substr.
261
262 CGI
263 charnames
264 One can now have custom character name aliases.
265
266 CPAN
267 There is now a simple command line frontend to the CPAN.pm module
268 called cpan.
269
270 Data::Dumper
271 A new option, Pair, allows choosing the separator between hash keys
272 and values.
273
274 DB_File
275 Devel::PPPort
276 Digest::MD5
277 Encode
278 Significant updates on the encoding pragma functionality (tr/// and
279 the DATA filehandle, formats).
280
281 If a filehandle has been marked as to have an encoding, unmappable
282 characters are detected already during input, not later (when the
283 corrupted data is being used).
284
285 The ISO 8859-6 conversion table has been corrected (the 0x30..0x39
286 erroneously mapped to U+0660..U+0669, instead of U+0030..U+0039).
287 The GSM 03.38 conversion did not handle escape sequences correctly.
288 The UTF-7 encoding has been added (making Encode feature-complete
289 with Unicode::String).
290
291 fields
292 libnet
293 Math::BigInt
294 A lot of bugs have been fixed since v1.60, the version included in
295 Perl v5.8.0. Especially noteworthy are the bug in Calc that caused
296 div and mod to fail for some large values, and the fixes to the
297 handling of bad inputs.
298
299 Some new features were added, e.g. the broot() method, you can now
300 pass parameters to config() to change some settings at runtime, and
301 it is now possible to trap the creation of NaN and infinity.
302
303 As usual, some optimizations took place and made the math overall a
304 tad faster. In some cases, quite a lot faster, actually. Especially
305 alternative libraries like Math::BigInt::GMP benefit from this. In
306 addition, a lot of the quite clunky routines like fsqrt() and
307 flog() are now much much faster.
308
309 MIME::Base64
310 NEXT
311 Diamond inheritance now works.
312
313 Net::Ping
314 PerlIO::scalar
315 Reading from non-string scalars (like the special variables, see
316 perlvar) now works.
317
318 podlators
319 Pod::LaTeX
320 PodParsers
321 Pod::Perldoc
322 Complete rewrite. As a side-effect, no longer refuses to startup
323 when run by root.
324
325 Scalar::Util
326 New utilities: refaddr, isvstring, looks_like_number,
327 set_prototype.
328
329 Storable
330 Can now store code references (via B::Deparse, so not foolproof).
331
332 strict
333 Earlier versions of the strict pragma did not check the parameters
334 implicitly passed to its "import" (use) and "unimport" (no)
335 routine. This caused the false idiom such as:
336
337 use strict qw(@ISA);
338 @ISA = qw(Foo);
339
340 This however (probably) raised the false expectation that the
341 strict refs, vars and subs were being enforced (and that @ISA was
342 somehow "declared"). But the strict refs, vars, and subs are not
343 enforced when using this false idiom.
344
345 Starting from Perl 5.8.1, the above will cause an error to be
346 raised. This may cause programs which used to execute seemingly
347 correctly without warnings and errors to fail when run under 5.8.1.
348 This happens because
349
350 use strict qw(@ISA);
351
352 will now fail with the error:
353
354 Unknown 'strict' tag(s) '@ISA'
355
356 The remedy to this problem is to replace this code with the correct
357 idiom:
358
359 use strict;
360 use vars qw(@ISA);
361 @ISA = qw(Foo);
362
363 Term::ANSIcolor
364 Test::Harness
365 Now much more picky about extra or missing output from test
366 scripts.
367
368 Test::More
369 Test::Simple
370 Text::Balanced
371 Time::HiRes
372 Use of nanosleep(), if available, allows mixing subsecond sleeps
373 with alarms.
374
375 threads
376 Several fixes, for example for join() problems and memory leaks.
377 In some platforms (like Linux) that use glibc the minimum memory
378 footprint of one ithread has been reduced by several hundred
379 kilobytes.
380
381 threads::shared
382 Many memory leaks have been fixed.
383
384 Unicode::Collate
385 Unicode::Normalize
386 Win32::GetFolderPath
387 Win32::GetOSVersion
388 Now returns extra information.
389
391 The "h2xs" utility now produces a more modern layout:
392 Foo-Bar/lib/Foo/Bar.pm instead of Foo/Bar/Bar.pm. Also, the
393 boilerplate test is now called t/Foo-Bar.t instead of t/1.t.
394
395 The Perl debugger (lib/perl5db.pl) has now been extensively documented
396 and bugs found while documenting have been fixed.
397
398 "perldoc" has been rewritten from scratch to be more robust and feature
399 rich.
400
401 "perlcc -B" works now at least somewhat better, while "perlcc -c" is
402 rather more broken. (The Perl compiler suite as a whole continues to
403 be experimental.)
404
406 perl573delta has been added to list the differences between the (now
407 quite obsolete) development releases 5.7.2 and 5.7.3.
408
409 perl58delta and perl581delta have been added: these are the perldeltas
410 of 5.8.0 and 5.8.1, detailing the differences respectively between
411 5.6.0 and 5.8.0, and between 5.8.0 and 5.8.1.
412
413 perlartistic has been added: it is the Artistic License in pod format,
414 making it easier for modules to refer to it.
415
416 perlcheat has been added: it is a Perl cheat sheet.
417
418 perlgpl has been added: it is the GNU General Public License in pod
419 format, making it easier for modules to refer to it.
420
421 perlmacosx has been added to tell about the installation and use of
422 Perl in Mac OS X.
423
424 perlos400 has been added to tell about the installation and use of Perl
425 in OS/400 PASE.
426
427 perlreref has been added: it is a regular expressions quick reference.
428
430 The Unix standard Perl location, /usr/bin/perl, is no longer
431 overwritten by default if it exists. This change was very prudent
432 because so many Unix vendors already provide a /usr/bin/perl, but
433 simultaneously many system utilities may depend on that exact version
434 of Perl, so better not to overwrite it.
435
436 One can now specify installation directories for site and vendor man
437 and HTML pages, and site and vendor scripts. See INSTALL.
438
439 One can now specify a destination directory for Perl installation by
440 specifying the DESTDIR variable for "make install". (This feature is
441 slightly different from the previous "Configure -Dinstallprefix=...".)
442 See INSTALL.
443
444 gcc versions 3.x introduced a new warning that caused a lot of noise
445 during Perl compilation: "gcc -Ialreadyknowndirectory (warning:
446 changing search order)". This warning has now been avoided by
447 Configure weeding out such directories before the compilation.
448
449 One can now build subsets of Perl core modules by using the Configure
450 flags "-Dnoextensions=..." and "-Donlyextensions=...", see INSTALL.
451
452 Platform-specific enhancements
453 In Cygwin Perl can now be built with threads ("Configure
454 -Duseithreads"). This works with both Cygwin 1.3.22 and Cygwin 1.5.3.
455
456 In newer FreeBSD releases Perl 5.8.0 compilation failed because of
457 trying to use malloc.h, which in FreeBSD is just a dummy file, and a
458 fatal error to even try to use. Now malloc.h is not used.
459
460 Perl is now known to build also in Hitachi HI-UXMPP.
461
462 Perl is now known to build again in LynxOS.
463
464 Mac OS X now installs with Perl version number embedded in installation
465 directory names for easier upgrading of user-compiled Perl, and the
466 installation directories in general are more standard. In other words,
467 the default installation no longer breaks the Apple-provided Perl. On
468 the other hand, with "Configure -Dprefix=/usr" you can now really
469 replace the Apple-supplied Perl (please be careful).
470
471 Mac OS X now builds Perl statically by default. This change was done
472 mainly for faster startup times. The Apple-provided Perl is still
473 dynamically linked and shared, and you can enable the sharedness for
474 your own Perl builds by "Configure -Duseshrplib".
475
476 Perl has been ported to IBM's OS/400 PASE environment. The best way to
477 build a Perl for PASE is to use an AIX host as a cross-compilation
478 environment. See README.os400.
479
480 Yet another cross-compilation option has been added: now Perl builds on
481 OpenZaurus, an Linux distribution based on Mandrake + Embedix for the
482 Sharp Zaurus PDA. See the Cross/README file.
483
484 Tru64 when using gcc 3 drops the optimisation for toke.c to "-O2"
485 because of gigantic memory use with the default "-O3".
486
487 Tru64 can now build Perl with the newer Berkeley DBs.
488
489 Building Perl on WinCE has been much enhanced, see README.ce and
490 README.perlce.
491
493 Closures, eval and lexicals
494 There have been many fixes in the area of anonymous subs, lexicals and
495 closures. Although this means that Perl is now more "correct", it is
496 possible that some existing code will break that happens to rely on the
497 faulty behaviour. In practice this is unlikely unless your code
498 contains a very complex nesting of anonymous subs, evals and lexicals.
499
500 Generic fixes
501 If an input filehandle is marked ":utf8" and Perl sees illegal UTF-8
502 coming in when doing "<FH>", if warnings are enabled a warning is
503 immediately given - instead of being silent about it and Perl being
504 unhappy about the broken data later. (The ":encoding(utf8)" layer also
505 works the same way.)
506
507 binmode(SOCKET, ":utf8") only worked on the input side, not on the
508 output side of the socket. Now it works both ways.
509
510 For threaded Perls certain system database functions like getpwent()
511 and getgrent() now grow their result buffer dynamically, instead of
512 failing. This means that at sites with lots of users and groups the
513 functions no longer fail by returning only partial results.
514
515 Perl 5.8.0 had accidentally broken the capability for users to define
516 their own uppercase<->lowercase Unicode mappings (as advertised by the
517 Camel). This feature has been fixed and is also documented better.
518
519 In 5.8.0 this
520
521 $some_unicode .= <FH>;
522
523 didn't work correctly but instead corrupted the data. This has now
524 been fixed.
525
526 Tied methods like FETCH etc. may now safely access tied values, i.e.
527 resulting in a recursive call to FETCH etc. Remember to break the
528 recursion, though.
529
530 At startup Perl blocks the SIGFPE signal away since there isn't much
531 Perl can do about it. Previously this blocking was in effect also for
532 programs executed from within Perl. Now Perl restores the original
533 SIGFPE handling routine, whatever it was, before running external
534 programs.
535
536 Linenumbers in Perl scripts may now be greater than 65536, or 2**16.
537 (Perl scripts have always been able to be larger than that, it's just
538 that the linenumber for reported errors and warnings have "wrapped
539 around".) While scripts that large usually indicate a need to rethink
540 your code a bit, such Perl scripts do exist, for example as results
541 from generated code. Now linenumbers can go all the way to 4294967296,
542 or 2**32.
543
544 Platform-specific fixes
545 Linux
546
547 · Setting $0 works again (with certain limitations that Perl cannot
548 do much about: see "$0" in perlvar)
549
550 HP-UX
551
552 · Setting $0 now works.
553
554 VMS
555
556 · Configuration now tests for the presence of "poll()", and IO::Poll
557 now uses the vendor-supplied function if detected.
558
559 · A rare access violation at Perl start-up could occur if the Perl
560 image was installed with privileges or if there was an identifier
561 with the subsystem attribute set in the process's rightslist.
562 Either of these circumstances triggered tainting code that
563 contained a pointer bug. The faulty pointer arithmetic has been
564 fixed.
565
566 · The length limit on values (not keys) in the %ENV hash has been
567 raised from 255 bytes to 32640 bytes (except when the
568 PERL_ENV_TABLES setting overrides the default use of logical names
569 for %ENV). If it is necessary to access these long values from
570 outside Perl, be aware that they are implemented using search list
571 logical names that store the value in pieces, each 255-byte piece
572 (up to 128 of them) being an element in the search list. When doing
573 a lookup in %ENV from within Perl, the elements are combined into a
574 single value. The existing VMS-specific ability to access
575 individual elements of a search list logical name via the
576 $ENV{'foo;N'} syntax (where N is the search list index) is
577 unimpaired.
578
579 · The piping implementation now uses local rather than global DCL
580 symbols for inter-process communication.
581
582 · File::Find could become confused when navigating to a relative
583 directory whose name collided with a logical name. This problem
584 has been corrected by adding directory syntax to relative path
585 names, thus preventing logical name translation.
586
587 Win32
588
589 · A memory leak in the fork() emulation has been fixed.
590
591 · The return value of the ioctl() built-in function was accidentally
592 broken in 5.8.0. This has been corrected.
593
594 · The internal message loop executed by perl during blocking
595 operations sometimes interfered with messages that were external to
596 Perl. This often resulted in blocking operations terminating
597 prematurely or returning incorrect results, when Perl was executing
598 under environments that could generate Windows messages. This has
599 been corrected.
600
601 · Pipes and sockets are now automatically in binary mode.
602
603 · The four-argument form of select() did not preserve $! (errno)
604 properly when there were errors in the underlying call. This is
605 now fixed.
606
607 · The "CR CR LF" problem of has been fixed, binmode(FH, ":crlf") is
608 now effectively a no-op.
609
611 All the warnings related to pack() and unpack() were made more
612 informative and consistent.
613
614 Changed "A thread exited while %d threads were running"
615 The old version
616
617 A thread exited while %d other threads were still running
618
619 was misleading because the "other" included also the thread giving the
620 warning.
621
622 Removed "Attempt to clear a restricted hash"
623 It is not illegal to clear a restricted hash, so the warning was
624 removed.
625
626 New "Illegal declaration of anonymous subroutine"
627 You must specify the block of code for "sub".
628
629 Changed "Invalid range "%s" in transliteration operator"
630 The old version
631
632 Invalid [] range "%s" in transliteration operator
633
634 was simply wrong because there are no "[] ranges" in tr///.
635
636 New "Missing control char name in \c"
637 Self-explanatory.
638
639 New "Newline in left-justified string for %s"
640 The padding spaces would appear after the newline, which is probably
641 not what you had in mind.
642
643 New "Possible precedence problem on bitwise %c operator"
644 If you think this
645
646 $x & $y == 0
647
648 tests whether the bitwise AND of $x and $y is zero, you will like this
649 warning.
650
651 New "read() on %s filehandle %s"
652 You cannot read() (or sysread()) from a closed or unopened filehandle.
653
654 New "Tied variable freed while still in use"
655 Something pulled the plug on a live tied variable, Perl plays safe by
656 bailing out.
657
658 New "To%s: illegal mapping '%s'"
659 An illegal user-defined Unicode casemapping was specified.
660
661 New "Use of freed value in iteration"
662 Something modified the values being iterated over. This is not good.
663
665 These news matter to you only if you either write XS code or like to
666 know about or hack Perl internals (using Devel::Peek or any of the
667 "B::" modules counts), or like to run Perl with the "-D" option.
668
669 The embedding examples of perlembed have been reviewed to be up to date
670 and consistent: for example, the correct use of PERL_SYS_INIT3() and
671 PERL_SYS_TERM().
672
673 Extensive reworking of the pad code (the code responsible for lexical
674 variables) has been conducted by Dave Mitchell.
675
676 Extensive work on the v-strings by John Peacock.
677
678 UTF-8 length and position cache: to speed up the handling of Unicode
679 (UTF-8) scalars, a cache was introduced. Potential problems exist if
680 an extension bypasses the official APIs and directly modifies the PV of
681 an SV: the UTF-8 cache does not get cleared as it should.
682
683 APIs obsoleted in Perl 5.8.0, like sv_2pv, sv_catpvn, sv_catsv,
684 sv_setsv, are again available.
685
686 Certain Perl core C APIs like cxinc and regatom are no longer available
687 at all to code outside the Perl core of the Perl core extensions. This
688 is intentional. They never should have been available with the shorter
689 names, and if you application depends on them, you should (be ashamed
690 and) contact perl5-porters to discuss what are the proper APIs.
691
692 Certain Perl core C APIs like "Perl_list" are no longer available
693 without their "Perl_" prefix. If your XS module stops working because
694 some functions cannot be found, in many cases a simple fix is to add
695 the "Perl_" prefix to the function and the thread context "aTHX_" as
696 the first argument of the function call. This is also how it should
697 always have been done: letting the Perl_-less forms to leak from the
698 core was an accident. For cleaner embedding you can also force this
699 for all APIs by defining at compile time the cpp define
700 PERL_NO_SHORT_NAMES.
701
702 Perl_save_bool() has been added.
703
704 Regexp objects (those created with "qr") now have S-magic rather than
705 R-magic. This fixed regexps of the form /...(??{...;$x})/ to no longer
706 ignore changes made to $x. The S-magic avoids dropping the caching
707 optimization and making (??{...}) constructs obscenely slow (and
708 consequently useless). See also "Magic Variables" in perlguts.
709 Regexp::Copy was affected by this change.
710
711 The Perl internal debugging macros DEBUG() and DEB() have been renamed
712 to PERL_DEBUG() and PERL_DEB() to avoid namespace conflicts.
713
714 "-DL" removed (the leaktest had been broken and unsupported for years,
715 use alternative debugging mallocs or tools like valgrind and Purify).
716
717 Verbose modifier "v" added for "-DXv" and "-Dsv", see perlrun.
718
720 In Perl 5.8.0 there were about 69000 separate tests in about 700 test
721 files, in Perl 5.9.0 there are about 77000 separate tests in about 780
722 test files. The exact numbers depend on the Perl configuration and on
723 the operating system platform.
724
726 The hash randomisation mentioned in "Incompatible Changes" is
727 definitely problematic: it will wake dormant bugs and shake out bad
728 assumptions.
729
730 Many of the rarer platforms that worked 100% or pretty close to it with
731 perl 5.8.0 have been left a little bit untended since their maintainers
732 have been otherwise busy lately, and therefore there will be more
733 failures on those platforms. Such platforms include Mac OS Classic,
734 IBM z/OS (and other EBCDIC platforms), and NetWare. The most common
735 Perl platforms (Unix and Unix-like, Microsoft platforms, and VMS) have
736 large enough testing and expert population that they are doing well.
737
738 Tied hashes in scalar context
739 Tied hashes do not currently return anything useful in scalar context,
740 for example when used as boolean tests:
741
742 if (%tied_hash) { ... }
743
744 The current nonsensical behaviour is always to return false, regardless
745 of whether the hash is empty or has elements.
746
747 The root cause is that there is no interface for the implementors of
748 tied hashes to implement the behaviour of a hash in scalar context.
749
750 Net::Ping 450_service and 510_ping_udp failures
751 The subtests 9 and 18 of lib/Net/Ping/t/450_service.t, and the subtest
752 2 of lib/Net/Ping/t/510_ping_udp.t might fail if you have an unusual
753 networking setup. For example in the latter case the test is trying to
754 send a UDP ping to the IP address 127.0.0.1.
755
756 B::C
757 The C-generating compiler backend B::C (the frontend being "perlcc -c")
758 is even more broken than it used to be because of the extensive lexical
759 variable changes. (The good news is that B::Bytecode and ByteLoader
760 are better than they used to be.)
761
763 EBCDIC Platforms
764 IBM z/OS and other EBCDIC platforms continue to be problematic
765 regarding Unicode support. Many Unicode tests are skipped when they
766 really should be fixed.
767
768 Cygwin 1.5 problems
769 In Cygwin 1.5 the io/tell and op/sysio tests have failures for some yet
770 unknown reason. In 1.5.5 the threads tests stress_cv, stress_re, and
771 stress_string are failing unless the environment variable PERLIO is set
772 to "perlio" (which makes also the io/tell failure go away).
773
774 Perl 5.8.1 does build and work well with Cygwin 1.3: with (uname -a)
775 "CYGWIN_NT-5.0 ... 1.3.22(0.78/3/2) 2003-03-18 09:20 i686 ..." a 100%
776 "make test" was achieved with "Configure -des -Duseithreads".
777
778 HP-UX: HP cc warnings about sendfile and sendpath
779 With certain HP C compiler releases (e.g. B.11.11.02) you will get many
780 warnings like this (lines wrapped for easier reading):
781
782 cc: "/usr/include/sys/socket.h", line 504: warning 562:
783 Redeclaration of "sendfile" with a different storage class specifier:
784 "sendfile" will have internal linkage.
785 cc: "/usr/include/sys/socket.h", line 505: warning 562:
786 Redeclaration of "sendpath" with a different storage class specifier:
787 "sendpath" will have internal linkage.
788
789 The warnings show up both during the build of Perl and during certain
790 lib/ExtUtils tests that invoke the C compiler. The warning, however,
791 is not serious and can be ignored.
792
793 IRIX: t/uni/tr_7jis.t falsely failing
794 The test t/uni/tr_7jis.t is known to report failure under 'make test'
795 or the test harness with certain releases of IRIX (at least IRIX 6.5
796 and MIPSpro Compilers Version 7.3.1.1m), but if run manually the test
797 fully passes.
798
799 Mac OS X: no usemymalloc
800 The Perl malloc ("-Dusemymalloc") does not work at all in Mac OS X.
801 This is not that serious, though, since the native malloc works just
802 fine.
803
804 Tru64: No threaded builds with GNU cc (gcc)
805 In the latest Tru64 releases (e.g. v5.1B or later) gcc cannot be used
806 to compile a threaded Perl (-Duseithreads) because the system
807 "<pthread.h>" file doesn't know about gcc.
808
809 Win32: sysopen, sysread, syswrite
810 As of the 5.8.0 release, sysopen()/sysread()/syswrite() do not behave
811 like they used to in 5.6.1 and earlier with respect to "text" mode.
812 These built-ins now always operate in "binary" mode (even if sysopen()
813 was passed the O_TEXT flag, or if binmode() was used on the file
814 handle). Note that this issue should only make a difference for disk
815 files, as sockets and pipes have always been in "binary" mode in the
816 Windows port. As this behavior is currently considered a bug,
817 compatible behavior may be re-introduced in a future release. Until
818 then, the use of sysopen(), sysread() and syswrite() is not supported
819 for "text" mode operations.
820
822 Here are some things that are planned for perl 5.10.0 :
823
824 · Various Copy-On-Write techniques will be investigated in hopes of
825 speeding up Perl.
826
827 · CPANPLUS, Inline, and Module::Build will become core modules.
828
829 · The ability to write true lexically scoped pragmas will be
830 introduced, perhaps via a "pragma" pragma.
831
832 · Work will continue on the bytecompiler and byteloader.
833
834 · v-strings as they currently exist are scheduled to be deprecated.
835 The v-less form (1.2.3) will become a "version object" when used
836 with "use", "require", and $VERSION. $^V will also be a "version
837 object" so the printf("%vd",...) construct will no longer be
838 needed. The v-ful version (v1.2.3) will become obsolete. The
839 equivalence of strings and v-strings (e.g. that currently 5.8.0 is
840 equal to "\5\8\0") will go away. There may be no deprecation
841 warning for v-strings, though: it is quite hard to detect when
842 v-strings are being used safely, and when they are not.
843
845 If you find what you think is a bug, you might check the articles
846 recently posted to the comp.lang.perl.misc newsgroup and the perl bug
847 database at http://bugs.perl.org/. There may also be information at
848 http://www.perl.com/, the Perl Home Page.
849
850 If you believe you have an unreported bug, please run the perlbug
851 program included with your release. Be sure to trim your bug down to a
852 tiny but sufficient test case. Your bug report, along with the output
853 of "perl -V", will be sent off to perlbug@perl.org to be analysed by
854 the Perl porting team. You can browse and search the Perl 5 bugs at
855 http://bugs.perl.org/.
856
858 The Changes file for exhaustive details on what changed.
859
860 The INSTALL file for how to build Perl.
861
862 The README file for general stuff.
863
864 The Artistic and Copying files for copyright information.
865
866
867
868perl v5.12.4 2011-06-01 PERL590DELTA(1)