| 1 | =head1 NAME |
| 2 | |
| 3 | perl58delta - what is new for perl v5.8.0 |
| 4 | |
| 5 | =head1 DESCRIPTION |
| 6 | |
| 7 | This document describes differences between the 5.6.0 release and |
| 8 | the 5.8.0 release. |
| 9 | |
| 10 | Many of the bug fixes in 5.8.0 were already seen in the 5.6.1 |
| 11 | maintenance release since the two releases were kept closely |
| 12 | coordinated (while 5.8.0 was still called 5.7.something). |
| 13 | |
| 14 | Changes that were integrated into the 5.6.1 release are marked C<[561]>. |
| 15 | Many of these changes have been further developed since 5.6.1 was released, |
| 16 | those are marked C<[561+]>. |
| 17 | |
| 18 | You can see the list of changes in the 5.6.1 release (both from the |
| 19 | 5.005_03 release and the 5.6.0 release) by reading L<perl561delta>. |
| 20 | |
| 21 | =head1 Highlights In 5.8.0 |
| 22 | |
| 23 | =over 4 |
| 24 | |
| 25 | =item * |
| 26 | |
| 27 | Better Unicode support |
| 28 | |
| 29 | =item * |
| 30 | |
| 31 | New IO Implementation |
| 32 | |
| 33 | =item * |
| 34 | |
| 35 | New Thread Implementation |
| 36 | |
| 37 | =item * |
| 38 | |
| 39 | Better Numeric Accuracy |
| 40 | |
| 41 | =item * |
| 42 | |
| 43 | Safe Signals |
| 44 | |
| 45 | =item * |
| 46 | |
| 47 | Many New Modules |
| 48 | |
| 49 | =item * |
| 50 | |
| 51 | More Extensive Regression Testing |
| 52 | |
| 53 | =back |
| 54 | |
| 55 | =head1 Incompatible Changes |
| 56 | |
| 57 | =head2 Binary Incompatibility |
| 58 | |
| 59 | B<Perl 5.8 is not binary compatible with earlier releases of Perl.> |
| 60 | |
| 61 | B<You have to recompile your XS modules.> |
| 62 | |
| 63 | (Pure Perl modules should continue to work.) |
| 64 | |
| 65 | The major reason for the discontinuity is the new IO architecture |
| 66 | called PerlIO. PerlIO is the default configuration because without |
| 67 | it many new features of Perl 5.8 cannot be used. In other words: |
| 68 | you just have to recompile your modules containing XS code, sorry |
| 69 | about that. |
| 70 | |
| 71 | In future releases of Perl, non-PerlIO aware XS modules may become |
| 72 | completely unsupported. This shouldn't be too difficult for module |
| 73 | authors, however: PerlIO has been designed as a drop-in replacement |
| 74 | (at the source code level) for the stdio interface. |
| 75 | |
| 76 | Depending on your platform, there are also other reasons why |
| 77 | we decided to break binary compatibility, please read on. |
| 78 | |
| 79 | =head2 64-bit platforms and malloc |
| 80 | |
| 81 | If your pointers are 64 bits wide, the Perl malloc is no longer being |
| 82 | used because it does not work well with 8-byte pointers. Also, |
| 83 | usually the system mallocs on such platforms are much better optimized |
| 84 | for such large memory models than the Perl malloc. Some memory-hungry |
| 85 | Perl applications like the PDL don't work well with Perl's malloc. |
| 86 | Finally, other applications than Perl (such as mod_perl) tend to prefer |
| 87 | the system malloc. Such platforms include Alpha and 64-bit HPPA, |
| 88 | MIPS, PPC, and Sparc. |
| 89 | |
| 90 | =head2 AIX Dynaloading |
| 91 | |
| 92 | The AIX dynaloading now uses in AIX releases 4.3 and newer the native |
| 93 | dlopen interface of AIX instead of the old emulated interface. This |
| 94 | change will probably break backward compatibility with compiled |
| 95 | modules. The change was made to make Perl more compliant with other |
| 96 | applications like mod_perl which are using the AIX native interface. |
| 97 | |
| 98 | =head2 Attributes for C<my> variables now handled at run-time |
| 99 | |
| 100 | The C<my EXPR : ATTRS> syntax now applies variable attributes at |
| 101 | run-time. (Subroutine and C<our> variables still get attributes applied |
| 102 | at compile-time.) See L<attributes> for additional details. In particular, |
| 103 | however, this allows variable attributes to be useful for C<tie> interfaces, |
| 104 | which was a deficiency of earlier releases. Note that the new semantics |
| 105 | doesn't work with the Attribute::Handlers module (as of version 0.76). |
| 106 | |
| 107 | =head2 Socket Extension Dynamic in VMS |
| 108 | |
| 109 | The Socket extension is now dynamically loaded instead of being |
| 110 | statically built in. This may or may not be a problem with ancient |
| 111 | TCP/IP stacks of VMS: we do not know since we weren't able to test |
| 112 | Perl in such configurations. |
| 113 | |
| 114 | =head2 IEEE-format Floating Point Default on OpenVMS Alpha |
| 115 | |
| 116 | Perl now uses IEEE format (T_FLOAT) as the default internal floating |
| 117 | point format on OpenVMS Alpha, potentially breaking binary compatibility |
| 118 | with external libraries or existing data. G_FLOAT is still available as |
| 119 | a configuration option. The default on VAX (D_FLOAT) has not changed. |
| 120 | |
| 121 | =head2 New Unicode Semantics (no more C<use utf8>, almost) |
| 122 | |
| 123 | Previously in Perl 5.6 to use Unicode one would say "use utf8" and |
| 124 | then the operations (like string concatenation) were Unicode-aware |
| 125 | in that lexical scope. |
| 126 | |
| 127 | This was found to be an inconvenient interface, and in Perl 5.8 the |
| 128 | Unicode model has completely changed: now the "Unicodeness" is bound |
| 129 | to the data itself, and for most of the time "use utf8" is not needed |
| 130 | at all. The only remaining use of "use utf8" is when the Perl script |
| 131 | itself has been written in the UTF-8 encoding of Unicode. (UTF-8 has |
| 132 | not been made the default since there are many Perl scripts out there |
| 133 | that are using various national eight-bit character sets, which would |
| 134 | be illegal in UTF-8.) |
| 135 | |
| 136 | See L<perluniintro> for the explanation of the current model, |
| 137 | and L<utf8> for the current use of the utf8 pragma. |
| 138 | |
| 139 | =head2 New Unicode Properties |
| 140 | |
| 141 | Unicode I<scripts> are now supported. Scripts are similar to (and superior |
| 142 | to) Unicode I<blocks>. The difference between scripts and blocks is that |
| 143 | scripts are the glyphs used by a language or a group of languages, while |
| 144 | the blocks are more artificial groupings of (mostly) 256 characters based |
| 145 | on the Unicode numbering. |
| 146 | |
| 147 | In general, scripts are more inclusive, but not universally so. For |
| 148 | example, while the script C<Latin> includes all the Latin characters and |
| 149 | their various diacritic-adorned versions, it does not include the various |
| 150 | punctuation or digits (since they are not solely C<Latin>). |
| 151 | |
| 152 | A number of other properties are now supported, including C<\p{L&}>, |
| 153 | C<\p{Any}> C<\p{Assigned}>, C<\p{Unassigned}>, C<\p{Blank}> [561] and |
| 154 | C<\p{SpacePerl}> [561] (along with their C<\P{...}> versions, of course). |
| 155 | See L<perlunicode> for details, and more additions. |
| 156 | |
| 157 | The C<In> or C<Is> prefix to names used with the C<\p{...}> and C<\P{...}> |
| 158 | are now almost always optional. The only exception is that a C<In> prefix |
| 159 | is required to signify a Unicode block when a block name conflicts with a |
| 160 | script name. For example, C<\p{Tibetan}> refers to the script, while |
| 161 | C<\p{InTibetan}> refers to the block. When there is no name conflict, you |
| 162 | can omit the C<In> from the block name (e.g. C<\p{BraillePatterns}>), but |
| 163 | to be safe, it's probably best to always use the C<In>). |
| 164 | |
| 165 | =head2 REF(...) Instead Of SCALAR(...) |
| 166 | |
| 167 | A reference to a reference now stringifies as "REF(0x81485ec)" instead |
| 168 | of "SCALAR(0x81485ec)" in order to be more consistent with the return |
| 169 | value of ref(). |
| 170 | |
| 171 | =head2 pack/unpack D/F recycled |
| 172 | |
| 173 | The undocumented pack/unpack template letters D/F have been recycled |
| 174 | for better use: now they stand for long double (if supported by the |
| 175 | platform) and NV (Perl internal floating point type). (They used |
| 176 | to be aliases for d/f, but you never knew that.) |
| 177 | |
| 178 | =head2 glob() now returns filenames in alphabetical order |
| 179 | |
| 180 | The list of filenames from glob() (or <...>) is now by default sorted |
| 181 | alphabetically to be csh-compliant (which is what happened before |
| 182 | in most UNIX platforms). (bsd_glob() does still sort platform |
| 183 | natively, ASCII or EBCDIC, unless GLOB_ALPHASORT is specified.) [561] |
| 184 | |
| 185 | =head2 Deprecations |
| 186 | |
| 187 | =over 4 |
| 188 | |
| 189 | =item * |
| 190 | |
| 191 | The semantics of bless(REF, REF) were unclear and until someone proves |
| 192 | it to make some sense, it is forbidden. |
| 193 | |
| 194 | =item * |
| 195 | |
| 196 | The obsolete chat2 library that should never have been allowed |
| 197 | to escape the laboratory has been decommissioned. |
| 198 | |
| 199 | =item * |
| 200 | |
| 201 | Using chdir("") or chdir(undef) instead of explicit chdir() is |
| 202 | doubtful. A failure (think chdir(some_function()) can lead into |
| 203 | unintended chdir() to the home directory, therefore this behaviour |
| 204 | is deprecated. |
| 205 | |
| 206 | =item * |
| 207 | |
| 208 | The builtin dump() function has probably outlived most of its |
| 209 | usefulness. The core-dumping functionality will remain in future |
| 210 | available as an explicit call to C<CORE::dump()>, but in future |
| 211 | releases the behaviour of an unqualified C<dump()> call may change. |
| 212 | |
| 213 | =item * |
| 214 | |
| 215 | The very dusty examples in the eg/ directory have been removed. |
| 216 | Suggestions for new shiny examples welcome but the main issue is that |
| 217 | the examples need to be documented, tested and (most importantly) |
| 218 | maintained. |
| 219 | |
| 220 | =item * |
| 221 | |
| 222 | The (bogus) escape sequences \8 and \9 now give an optional warning |
| 223 | ("Unrecognized escape passed through"). There is no need to \-escape |
| 224 | any C<\w> character. |
| 225 | |
| 226 | =item * |
| 227 | |
| 228 | The *glob{FILEHANDLE} is deprecated, use *glob{IO} instead. |
| 229 | |
| 230 | =item * |
| 231 | |
| 232 | The C<package;> syntax (C<package> without an argument) has been |
| 233 | deprecated. Its semantics were never that clear and its |
| 234 | implementation even less so. If you have used that feature to |
| 235 | disallow all but fully qualified variables, C<use strict;> instead. |
| 236 | |
| 237 | =item * |
| 238 | |
| 239 | The unimplemented POSIX regex features [[.cc.]] and [[=c=]] are still |
| 240 | recognised but now cause fatal errors. The previous behaviour of |
| 241 | ignoring them by default and warning if requested was unacceptable |
| 242 | since it, in a way, falsely promised that the features could be used. |
| 243 | |
| 244 | =item * |
| 245 | |
| 246 | In future releases, non-PerlIO aware XS modules may become completely |
| 247 | unsupported. Since PerlIO is a drop-in replacement for stdio at the |
| 248 | source code level, this shouldn't be that drastic a change. |
| 249 | |
| 250 | =item * |
| 251 | |
| 252 | Previous versions of perl and some readings of some sections of Camel |
| 253 | III implied that the C<:raw> "discipline" was the inverse of C<:crlf>. |
| 254 | Turning off "clrfness" is no longer enough to make a stream truly |
| 255 | binary. So the PerlIO C<:raw> layer (or "discipline", to use the Camel |
| 256 | book's older terminology) is now formally defined as being equivalent |
| 257 | to binmode(FH) - which is in turn defined as doing whatever is |
| 258 | necessary to pass each byte as-is without any translation. In |
| 259 | particular binmode(FH) - and hence C<:raw> - will now turn off both |
| 260 | CRLF and UTF-8 translation and remove other layers (e.g. :encoding()) |
| 261 | which would modify byte stream. |
| 262 | |
| 263 | =item * |
| 264 | |
| 265 | The current user-visible implementation of pseudo-hashes (the weird |
| 266 | use of the first array element) is deprecated starting from Perl 5.8.0 |
| 267 | and will be removed in Perl 5.10.0, and the feature will be |
| 268 | implemented differently. Not only is the current interface rather |
| 269 | ugly, but the current implementation slows down normal array and hash |
| 270 | use quite noticeably. The C<fields> pragma interface will remain |
| 271 | available. The I<restricted hashes> interface is expected to |
| 272 | be the replacement interface (see L<Hash::Util>). If your existing |
| 273 | programs depends on the underlying implementation, consider using |
| 274 | L<Class::PseudoHash> from CPAN. |
| 275 | |
| 276 | =item * |
| 277 | |
| 278 | The syntaxes C<< @a->[...] >> and C<< %h->{...} >> have now been deprecated. |
| 279 | |
| 280 | =item * |
| 281 | |
| 282 | After years of trying, suidperl is considered to be too complex to |
| 283 | ever be considered truly secure. The suidperl functionality is likely |
| 284 | to be removed in a future release. |
| 285 | |
| 286 | =item * |
| 287 | |
| 288 | The 5.005 threads model (module C<Thread>) is deprecated and expected |
| 289 | to be removed in Perl 5.10. Multithreaded code should be migrated to |
| 290 | the new ithreads model (see L<threads>, L<threads::shared> and |
| 291 | L<perlthrtut>). |
| 292 | |
| 293 | =item * |
| 294 | |
| 295 | The long deprecated uppercase aliases for the string comparison |
| 296 | operators (EQ, NE, LT, LE, GE, GT) have now been removed. |
| 297 | |
| 298 | =item * |
| 299 | |
| 300 | The tr///C and tr///U features have been removed and will not return; |
| 301 | the interface was a mistake. Sorry about that. For similar |
| 302 | functionality, see pack('U0', ...) and pack('C0', ...). [561] |
| 303 | |
| 304 | =item * |
| 305 | |
| 306 | Earlier Perls treated "sub foo (@bar)" as equivalent to "sub foo (@)". |
| 307 | The prototypes are now checked better at compile-time for invalid |
| 308 | syntax. An optional warning is generated ("Illegal character in |
| 309 | prototype...") but this may be upgraded to a fatal error in a future |
| 310 | release. |
| 311 | |
| 312 | =item * |
| 313 | |
| 314 | The C<exec LIST> and C<system LIST> operations now produce warnings on |
| 315 | tainted data and in some future release they will produce fatal errors. |
| 316 | |
| 317 | =item * |
| 318 | |
| 319 | The existing behaviour when localising tied arrays and hashes is wrong, |
| 320 | and will be changed in a future release, so do not rely on the existing |
| 321 | behaviour. See L<"Localising Tied Arrays and Hashes Is Broken">. |
| 322 | |
| 323 | =back |
| 324 | |
| 325 | =head1 Core Enhancements |
| 326 | |
| 327 | =head2 Unicode Overhaul |
| 328 | |
| 329 | Unicode in general should be now much more usable than in Perl 5.6.0 |
| 330 | (or even in 5.6.1). Unicode can be used in hash keys, Unicode in |
| 331 | regular expressions should work now, Unicode in tr/// should work now, |
| 332 | Unicode in I/O should work now. See L<perluniintro> for introduction |
| 333 | and L<perlunicode> for details. |
| 334 | |
| 335 | =over 4 |
| 336 | |
| 337 | =item * |
| 338 | |
| 339 | The Unicode Character Database coming with Perl has been upgraded |
| 340 | to Unicode 3.2.0. For more information, see http://www.unicode.org/ . |
| 341 | [561+] (5.6.1 has UCD 3.0.1.) |
| 342 | |
| 343 | =item * |
| 344 | |
| 345 | For developers interested in enhancing Perl's Unicode capabilities: |
| 346 | almost all the UCD files are included with the Perl distribution in |
| 347 | the F<lib/unicore> subdirectory. The most notable omission, for space |
| 348 | considerations, is the Unihan database. |
| 349 | |
| 350 | =item * |
| 351 | |
| 352 | The properties \p{Blank} and \p{SpacePerl} have been added. "Blank" is like |
| 353 | C isblank(), that is, it contains only "horizontal whitespace" (the space |
| 354 | character is, the newline isn't), and the "SpacePerl" is the Unicode |
| 355 | equivalent of C<\s> (\p{Space} isn't, since that includes the vertical |
| 356 | tabulator character, whereas C<\s> doesn't.) |
| 357 | |
| 358 | See "New Unicode Properties" earlier in this document for additional |
| 359 | information on changes with Unicode properties. |
| 360 | |
| 361 | =back |
| 362 | |
| 363 | =head2 PerlIO is Now The Default |
| 364 | |
| 365 | =over 4 |
| 366 | |
| 367 | =item * |
| 368 | |
| 369 | IO is now by default done via PerlIO rather than system's "stdio". |
| 370 | PerlIO allows "layers" to be "pushed" onto a file handle to alter the |
| 371 | handle's behaviour. Layers can be specified at open time via 3-arg |
| 372 | form of open: |
| 373 | |
| 374 | open($fh,'>:crlf :utf8', $path) || ... |
| 375 | |
| 376 | or on already opened handles via extended C<binmode>: |
| 377 | |
| 378 | binmode($fh,':encoding(iso-8859-7)'); |
| 379 | |
| 380 | The built-in layers are: unix (low level read/write), stdio (as in |
| 381 | previous Perls), perlio (re-implementation of stdio buffering in a |
| 382 | portable manner), crlf (does CRLF <=> "\n" translation as on Win32, |
| 383 | but available on any platform). A mmap layer may be available if |
| 384 | platform supports it (mostly UNIXes). |
| 385 | |
| 386 | Layers to be applied by default may be specified via the 'open' pragma. |
| 387 | |
| 388 | See L</"Installation and Configuration Improvements"> for the effects |
| 389 | of PerlIO on your architecture name. |
| 390 | |
| 391 | =item * |
| 392 | |
| 393 | If your platform supports fork(), you can use the list form of C<open> |
| 394 | for pipes. For example: |
| 395 | |
| 396 | open KID_PS, "-|", "ps", "aux" or die $!; |
| 397 | |
| 398 | forks the ps(1) command (without spawning a shell, as there are more |
| 399 | than three arguments to open()), and reads its standard output via the |
| 400 | C<KID_PS> filehandle. See L<perlipc>. |
| 401 | |
| 402 | =item * |
| 403 | |
| 404 | File handles can be marked as accepting Perl's internal encoding of Unicode |
| 405 | (UTF-8 or UTF-EBCDIC depending on platform) by a pseudo layer ":utf8" : |
| 406 | |
| 407 | open($fh,">:utf8","Uni.txt"); |
| 408 | |
| 409 | Note for EBCDIC users: the pseudo layer ":utf8" is erroneously named |
| 410 | for you since it's not UTF-8 what you will be getting but instead |
| 411 | UTF-EBCDIC. See L<perlunicode>, L<utf8>, and |
| 412 | http://www.unicode.org/unicode/reports/tr16/ for more information. |
| 413 | In future releases this naming may change. See L<perluniintro> |
| 414 | for more information about UTF-8. |
| 415 | |
| 416 | =item * |
| 417 | |
| 418 | If your environment variables (LC_ALL, LC_CTYPE, LANG) look like you |
| 419 | want to use UTF-8 (any of the the variables match C</utf-?8/i>), your |
| 420 | STDIN, STDOUT, STDERR handles and the default open layer (see L<open>) |
| 421 | are marked as UTF-8. (This feature, like other new features that |
| 422 | combine Unicode and I/O, work only if you are using PerlIO, but that's |
| 423 | the default.) |
| 424 | |
| 425 | Note that after this Perl really does assume that everything is UTF-8: |
| 426 | for example if some input handle is not, Perl will probably very soon |
| 427 | complain about the input data like this "Malformed UTF-8 ..." since |
| 428 | any old eight-bit data is not legal UTF-8. |
| 429 | |
| 430 | Note for code authors: if you want to enable your users to use UTF-8 |
| 431 | as their default encoding but in your code still have eight-bit I/O streams |
| 432 | (such as images or zip files), you need to explicitly open() or binmode() |
| 433 | with C<:bytes> (see L<perlfunc/open> and L<perlfunc/binmode>), or you |
| 434 | can just use C<binmode(FH)> (nice for pre-5.8.0 backward compatibility). |
| 435 | |
| 436 | =item * |
| 437 | |
| 438 | File handles can translate character encodings from/to Perl's internal |
| 439 | Unicode form on read/write via the ":encoding()" layer. |
| 440 | |
| 441 | =item * |
| 442 | |
| 443 | File handles can be opened to "in memory" files held in Perl scalars via: |
| 444 | |
| 445 | open($fh,'>', \$variable) || ... |
| 446 | |
| 447 | =item * |
| 448 | |
| 449 | Anonymous temporary files are available without need to |
| 450 | 'use FileHandle' or other module via |
| 451 | |
| 452 | open($fh,"+>", undef) || ... |
| 453 | |
| 454 | That is a literal undef, not an undefined value. |
| 455 | |
| 456 | =back |
| 457 | |
| 458 | =head2 ithreads |
| 459 | |
| 460 | The new interpreter threads ("ithreads" for short) implementation of |
| 461 | multithreading, by Arthur Bergman, replaces the old "5.005 threads" |
| 462 | implementation. In the ithreads model any data sharing between |
| 463 | threads must be explicit, as opposed to the model where data sharing |
| 464 | was implicit. See L<threads> and L<threads::shared>, and |
| 465 | L<perlthrtut>. |
| 466 | |
| 467 | As a part of the ithreads implementation Perl will also use |
| 468 | any necessary and detectable reentrant libc interfaces. |
| 469 | |
| 470 | =head2 Restricted Hashes |
| 471 | |
| 472 | A restricted hash is restricted to a certain set of keys, no keys |
| 473 | outside the set can be added. Also individual keys can be restricted |
| 474 | so that the key cannot be deleted and the value cannot be changed. |
| 475 | No new syntax is involved: the Hash::Util module is the interface. |
| 476 | |
| 477 | =head2 Safe Signals |
| 478 | |
| 479 | Perl used to be fragile in that signals arriving at inopportune moments |
| 480 | could corrupt Perl's internal state. Now Perl postpones handling of |
| 481 | signals until it's safe (between opcodes). |
| 482 | |
| 483 | This change may have surprising side effects because signals no longer |
| 484 | interrupt Perl instantly. Perl will now first finish whatever it was |
| 485 | doing, like finishing an internal operation (like sort()) or an |
| 486 | external operation (like an I/O operation), and only then look at any |
| 487 | arrived signals (and before starting the next operation). No more corrupt |
| 488 | internal state since the current operation is always finished first, |
| 489 | but the signal may take more time to get heard. Note that breaking |
| 490 | out from potentially blocking operations should still work, though. |
| 491 | |
| 492 | =head2 Understanding of Numbers |
| 493 | |
| 494 | In general a lot of fixing has happened in the area of Perl's |
| 495 | understanding of numbers, both integer and floating point. Since in |
| 496 | many systems the standard number parsing functions like C<strtoul()> |
| 497 | and C<atof()> seem to have bugs, Perl tries to work around their |
| 498 | deficiencies. This results hopefully in more accurate numbers. |
| 499 | |
| 500 | Perl now tries internally to use integer values in numeric conversions |
| 501 | and basic arithmetics (+ - * /) if the arguments are integers, and |
| 502 | tries also to keep the results stored internally as integers. |
| 503 | This change leads to often slightly faster and always less lossy |
| 504 | arithmetics. (Previously Perl always preferred floating point numbers |
| 505 | in its math.) |
| 506 | |
| 507 | =head2 Arrays now always interpolate into double-quoted strings [561] |
| 508 | |
| 509 | In double-quoted strings, arrays now interpolate, no matter what. The |
| 510 | behavior in earlier versions of perl 5 was that arrays would interpolate |
| 511 | into strings if the array had been mentioned before the string was |
| 512 | compiled, and otherwise Perl would raise a fatal compile-time error. |
| 513 | In versions 5.000 through 5.003, the error was |
| 514 | |
| 515 | Literal @example now requires backslash |
| 516 | |
| 517 | In versions 5.004_01 through 5.6.0, the error was |
| 518 | |
| 519 | In string, @example now must be written as \@example |
| 520 | |
| 521 | The idea here was to get people into the habit of writing |
| 522 | C<"fred\@example.com"> when they wanted a literal C<@> sign, just as |
| 523 | they have always written C<"Give me back my \$5"> when they wanted a |
| 524 | literal C<$> sign. |
| 525 | |
| 526 | Starting with 5.6.1, when Perl now sees an C<@> sign in a |
| 527 | double-quoted string, it I<always> attempts to interpolate an array, |
| 528 | regardless of whether or not the array has been used or declared |
| 529 | already. The fatal error has been downgraded to an optional warning: |
| 530 | |
| 531 | Possible unintended interpolation of @example in string |
| 532 | |
| 533 | This warns you that C<"fred@example.com"> is going to turn into |
| 534 | C<fred.com> if you don't backslash the C<@>. |
| 535 | See http://www.plover.com/~mjd/perl/at-error.html for more details |
| 536 | about the history here. |
| 537 | |
| 538 | =head2 Miscellaneous Changes |
| 539 | |
| 540 | =over 4 |
| 541 | |
| 542 | =item * |
| 543 | |
| 544 | AUTOLOAD is now lvaluable, meaning that you can add the :lvalue attribute |
| 545 | to AUTOLOAD subroutines and you can assign to the AUTOLOAD return value. |
| 546 | |
| 547 | =item * |
| 548 | |
| 549 | The $Config{byteorder} (and corresponding BYTEORDER in config.h) was |
| 550 | previously wrong in platforms if sizeof(long) was 4, but sizeof(IV) |
| 551 | was 8. The byteorder was only sizeof(long) bytes long (1234 or 4321), |
| 552 | but now it is correctly sizeof(IV) bytes long, (12345678 or 87654321). |
| 553 | (This problem didn't affect Windows platforms.) |
| 554 | |
| 555 | Also, $Config{byteorder} is now computed dynamically--this is more |
| 556 | robust with "fat binaries" where an executable image contains binaries |
| 557 | for more than one binary platform, and when cross-compiling. |
| 558 | |
| 559 | =item * |
| 560 | |
| 561 | C<perl -d:Module=arg,arg,arg> now works (previously one couldn't pass |
| 562 | in multiple arguments.) |
| 563 | |
| 564 | =item * |
| 565 | |
| 566 | C<do> followed by a bareword now ensures that this bareword isn't |
| 567 | a keyword (to avoid a bug where C<do q(foo.pl)> tried to call a |
| 568 | subroutine called C<q>). This means that for example instead of |
| 569 | C<do format()> you must write C<do &format()>. |
| 570 | |
| 571 | =item * |
| 572 | |
| 573 | The builtin dump() now gives an optional warning |
| 574 | C<dump() better written as CORE::dump()>, |
| 575 | meaning that by default C<dump(...)> is resolved as the builtin |
| 576 | dump() which dumps core and aborts, not as (possibly) user-defined |
| 577 | C<sub dump>. To call the latter, qualify the call as C<&dump(...)>. |
| 578 | (The whole dump() feature is to considered deprecated, and possibly |
| 579 | removed/changed in future releases.) |
| 580 | |
| 581 | =item * |
| 582 | |
| 583 | chomp() and chop() are now overridable. Note, however, that their |
| 584 | prototype (as given by C<prototype("CORE::chomp")> is undefined, |
| 585 | because it cannot be expressed and therefore one cannot really write |
| 586 | replacements to override these builtins. |
| 587 | |
| 588 | =item * |
| 589 | |
| 590 | END blocks are now run even if you exit/die in a BEGIN block. |
| 591 | Internally, the execution of END blocks is now controlled by |
| 592 | PL_exit_flags & PERL_EXIT_DESTRUCT_END. This enables the new |
| 593 | behaviour for Perl embedders. This will default in 5.10. See |
| 594 | L<perlembed>. |
| 595 | |
| 596 | =item * |
| 597 | |
| 598 | Formats now support zero-padded decimal fields. |
| 599 | |
| 600 | =item * |
| 601 | |
| 602 | Although "you shouldn't do that", it was possible to write code that |
| 603 | depends on Perl's hashed key order (Data::Dumper does this). The new |
| 604 | algorithm "One-at-a-Time" produces a different hashed key order. |
| 605 | More details are in L</"Performance Enhancements">. |
| 606 | |
| 607 | =item * |
| 608 | |
| 609 | lstat(FILEHANDLE) now gives a warning because the operation makes no sense. |
| 610 | In future releases this may become a fatal error. |
| 611 | |
| 612 | =item * |
| 613 | |
| 614 | Spurious syntax errors generated in certain situations, when glob() |
| 615 | caused File::Glob to be loaded for the first time, have been fixed. [561] |
| 616 | |
| 617 | =item * |
| 618 | |
| 619 | Lvalue subroutines can now return C<undef> in list context. However, |
| 620 | the lvalue subroutine feature still remains experimental. [561+] |
| 621 | |
| 622 | =item * |
| 623 | |
| 624 | A lost warning "Can't declare ... dereference in my" has been |
| 625 | restored (Perl had it earlier but it became lost in later releases.) |
| 626 | |
| 627 | =item * |
| 628 | |
| 629 | A new special regular expression variable has been introduced: |
| 630 | C<$^N>, which contains the most-recently closed group (submatch). |
| 631 | |
| 632 | =item * |
| 633 | |
| 634 | C<no Module;> does not produce an error even if Module does not have an |
| 635 | unimport() method. This parallels the behavior of C<use> vis-a-vis |
| 636 | C<import>. [561] |
| 637 | |
| 638 | =item * |
| 639 | |
| 640 | The numerical comparison operators return C<undef> if either operand |
| 641 | is a NaN. Previously the behaviour was unspecified. |
| 642 | |
| 643 | =item * |
| 644 | |
| 645 | C<our> can now have an experimental optional attribute C<unique> that |
| 646 | affects how global variables are shared among multiple interpreters, |
| 647 | see L<perlfunc/our>. |
| 648 | |
| 649 | =item * |
| 650 | |
| 651 | The following builtin functions are now overridable: each(), keys(), |
| 652 | pop(), push(), shift(), splice(), unshift(). [561] |
| 653 | |
| 654 | =item * |
| 655 | |
| 656 | C<pack() / unpack()> can now group template letters with C<()> and then |
| 657 | apply repetition/count modifiers on the groups. |
| 658 | |
| 659 | =item * |
| 660 | |
| 661 | C<pack() / unpack()> can now process the Perl internal numeric types: |
| 662 | IVs, UVs, NVs-- and also long doubles, if supported by the platform. |
| 663 | The template letters are C<j>, C<J>, C<F>, and C<D>. |
| 664 | |
| 665 | =item * |
| 666 | |
| 667 | C<pack('U0a*', ...)> can now be used to force a string to UTF-8. |
| 668 | |
| 669 | =item * |
| 670 | |
| 671 | my __PACKAGE__ $obj now works. [561] |
| 672 | |
| 673 | =item * |
| 674 | |
| 675 | POSIX::sleep() now returns the number of I<unslept> seconds |
| 676 | (as the POSIX standard says), as opposed to CORE::sleep() which |
| 677 | returns the number of slept seconds. |
| 678 | |
| 679 | =item * |
| 680 | |
| 681 | printf() and sprintf() now support parameter reordering using the |
| 682 | C<%\d+\$> and C<*\d+\$> syntaxes. For example |
| 683 | |
| 684 | printf "%2\$s %1\$s\n", "foo", "bar"; |
| 685 | |
| 686 | will print "bar foo\n". This feature helps in writing |
| 687 | internationalised software, and in general when the order |
| 688 | of the parameters can vary. |
| 689 | |
| 690 | =item * |
| 691 | |
| 692 | The (\&) prototype now works properly. [561] |
| 693 | |
| 694 | =item * |
| 695 | |
| 696 | prototype(\[$@%&]) is now available to implicitly create references |
| 697 | (useful for example if you want to emulate the tie() interface). |
| 698 | |
| 699 | =item * |
| 700 | |
| 701 | A new command-line option, C<-t> is available. It is the |
| 702 | little brother of C<-T>: instead of dying on taint violations, |
| 703 | lexical warnings are given. B<This is only meant as a temporary |
| 704 | debugging aid while securing the code of old legacy applications. |
| 705 | This is not a substitute for -T.> |
| 706 | |
| 707 | =item * |
| 708 | |
| 709 | In other taint news, the C<exec LIST> and C<system LIST> have now been |
| 710 | considered too risky (think C<exec @ARGV>: it can start any program |
| 711 | with any arguments), and now the said forms cause a warning under |
| 712 | lexical warnings. You should carefully launder the arguments to |
| 713 | guarantee their validity. In future releases of Perl the forms will |
| 714 | become fatal errors so consider starting laundering now. |
| 715 | |
| 716 | =item * |
| 717 | |
| 718 | Tied hash interfaces are now required to have the EXISTS and DELETE |
| 719 | methods (either own or inherited). |
| 720 | |
| 721 | =item * |
| 722 | |
| 723 | If tr/// is just counting characters, it doesn't attempt to |
| 724 | modify its target. |
| 725 | |
| 726 | =item * |
| 727 | |
| 728 | untie() will now call an UNTIE() hook if it exists. See L<perltie> |
| 729 | for details. [561] |
| 730 | |
| 731 | =item * |
| 732 | |
| 733 | L<utime> now supports C<utime undef, undef, @files> to change the |
| 734 | file timestamps to the current time. |
| 735 | |
| 736 | =item * |
| 737 | |
| 738 | The rules for allowing underscores (underbars) in numeric constants |
| 739 | have been relaxed and simplified: now you can have an underscore |
| 740 | simply B<between digits>. |
| 741 | |
| 742 | =item * |
| 743 | |
| 744 | Rather than relying on C's argv[0] (which may not contain a full pathname) |
| 745 | where possible $^X is now set by asking the operating system. |
| 746 | (eg by reading F</proc/self/exe> on Linux, F</proc/curproc/file> on FreeBSD) |
| 747 | |
| 748 | =item * |
| 749 | |
| 750 | A new variable, C<${^TAINT}>, indicates whether taint mode is enabled. |
| 751 | |
| 752 | =item * |
| 753 | |
| 754 | You can now override the readline() builtin, and this overrides also |
| 755 | the <FILEHANDLE> angle bracket operator. |
| 756 | |
| 757 | =item * |
| 758 | |
| 759 | The command-line options -s and -F are now recognized on the shebang |
| 760 | (#!) line. |
| 761 | |
| 762 | =item * |
| 763 | |
| 764 | Use of the C</c> match modifier without an accompanying C</g> modifier |
| 765 | elicits a new warning: C<Use of /c modifier is meaningless without /g>. |
| 766 | |
| 767 | Use of C</c> in substitutions, even with C</g>, elicits |
| 768 | C<Use of /c modifier is meaningless in s///>. |
| 769 | |
| 770 | Use of C</g> with C<split> elicits C<Use of /g modifier is meaningless |
| 771 | in split>. |
| 772 | |
| 773 | =item * |
| 774 | |
| 775 | Support for the C<CLONE> special subroutine had been added. |
| 776 | With ithreads, when a new thread is created, all Perl data is cloned, |
| 777 | however non-Perl data cannot be cloned automatically. In C<CLONE> you |
| 778 | can do whatever you need to do, like for example handle the cloning of |
| 779 | non-Perl data, if necessary. C<CLONE> will be executed once for every |
| 780 | package that has it defined or inherited. It will be called in the |
| 781 | context of the new thread, so all modifications are made in the new area. |
| 782 | |
| 783 | See L<perlmod> |
| 784 | |
| 785 | =back |
| 786 | |
| 787 | =head1 Modules and Pragmata |
| 788 | |
| 789 | =head2 New Modules and Pragmata |
| 790 | |
| 791 | =over 4 |
| 792 | |
| 793 | =item * |
| 794 | |
| 795 | C<Attribute::Handlers>, originally by Damian Conway and now maintained |
| 796 | by Arthur Bergman, allows a class to define attribute handlers. |
| 797 | |
| 798 | package MyPack; |
| 799 | use Attribute::Handlers; |
| 800 | sub Wolf :ATTR(SCALAR) { print "howl!\n" } |
| 801 | |
| 802 | # later, in some package using or inheriting from MyPack... |
| 803 | |
| 804 | my MyPack $Fluffy : Wolf; # the attribute handler Wolf will be called |
| 805 | |
| 806 | Both variables and routines can have attribute handlers. Handlers can |
| 807 | be specific to type (SCALAR, ARRAY, HASH, or CODE), or specific to the |
| 808 | exact compilation phase (BEGIN, CHECK, INIT, or END). |
| 809 | See L<Attribute::Handlers>. |
| 810 | |
| 811 | =item * |
| 812 | |
| 813 | C<B::Concise>, by Stephen McCamant, is a new compiler backend for |
| 814 | walking the Perl syntax tree, printing concise info about ops. |
| 815 | The output is highly customisable. See L<B::Concise>. [561+] |
| 816 | |
| 817 | =item * |
| 818 | |
| 819 | The new bignum, bigint, and bigrat pragmas, by Tels, implement |
| 820 | transparent bignum support (using the Math::BigInt, Math::BigFloat, |
| 821 | and Math::BigRat backends). |
| 822 | |
| 823 | =item * |
| 824 | |
| 825 | C<Class::ISA>, by Sean Burke, is a module for reporting the search |
| 826 | path for a class's ISA tree. See L<Class::ISA>. |
| 827 | |
| 828 | =item * |
| 829 | |
| 830 | C<Cwd> now has a split personality: if possible, an XS extension is |
| 831 | used, (this will hopefully be faster, more secure, and more robust) |
| 832 | but if not possible, the familiar Perl implementation is used. |
| 833 | |
| 834 | =item * |
| 835 | |
| 836 | C<Devel::PPPort>, originally by Kenneth Albanowski and now |
| 837 | maintained by Paul Marquess, has been added. It is primarily used |
| 838 | by C<h2xs> to enhance portability of XS modules between different |
| 839 | versions of Perl. See L<Devel::PPPort>. |
| 840 | |
| 841 | =item * |
| 842 | |
| 843 | C<Digest>, frontend module for calculating digests (checksums), from |
| 844 | Gisle Aas, has been added. See L<Digest>. |
| 845 | |
| 846 | =item * |
| 847 | |
| 848 | C<Digest::MD5> for calculating MD5 digests (checksums) as defined in |
| 849 | RFC 1321, from Gisle Aas, has been added. See L<Digest::MD5>. |
| 850 | |
| 851 | use Digest::MD5 'md5_hex'; |
| 852 | |
| 853 | $digest = md5_hex("Thirsty Camel"); |
| 854 | |
| 855 | print $digest, "\n"; # 01d19d9d2045e005c3f1b80e8b164de1 |
| 856 | |
| 857 | NOTE: the C<MD5> backward compatibility module is deliberately not |
| 858 | included since its further use is discouraged. |
| 859 | |
| 860 | See also L<PerlIO::via::QuotedPrint>. |
| 861 | |
| 862 | =item * |
| 863 | |
| 864 | C<Encode>, originally by Nick Ing-Simmons and now maintained by Dan |
| 865 | Kogai, provides a mechanism to translate between different character |
| 866 | encodings. Support for Unicode, ISO-8859-1, and ASCII are compiled in |
| 867 | to the module. Several other encodings (like the rest of the |
| 868 | ISO-8859, CP*/Win*, Mac, KOI8-R, three variants EBCDIC, Chinese, |
| 869 | Japanese, and Korean encodings) are included and can be loaded at |
| 870 | runtime. (For space considerations, the largest Chinese encodings |
| 871 | have been separated into their own CPAN module, Encode::HanExtra, |
| 872 | which Encode will use if available). See L<Encode>. |
| 873 | |
| 874 | Any encoding supported by Encode module is also available to the |
| 875 | ":encoding()" layer if PerlIO is used. |
| 876 | |
| 877 | =item * |
| 878 | |
| 879 | C<Hash::Util> is the interface to the new I<restricted hashes> |
| 880 | feature. (Implemented by Jeffrey Friedl, Nick Ing-Simmons, and |
| 881 | Michael Schwern.) See L<Hash::Util>. |
| 882 | |
| 883 | =item * |
| 884 | |
| 885 | C<I18N::Langinfo> can be used to query locale information. |
| 886 | See L<I18N::Langinfo>. |
| 887 | |
| 888 | =item * |
| 889 | |
| 890 | C<I18N::LangTags>, by Sean Burke, has functions for dealing with |
| 891 | RFC3066-style language tags. See L<I18N::LangTags>. |
| 892 | |
| 893 | =item * |
| 894 | |
| 895 | C<ExtUtils::Constant>, by Nicholas Clark, is a new tool for extension |
| 896 | writers for generating XS code to import C header constants. |
| 897 | See L<ExtUtils::Constant>. |
| 898 | |
| 899 | =item * |
| 900 | |
| 901 | C<Filter::Simple>, by Damian Conway, is an easy-to-use frontend to |
| 902 | Filter::Util::Call. See L<Filter::Simple>. |
| 903 | |
| 904 | # in MyFilter.pm: |
| 905 | |
| 906 | package MyFilter; |
| 907 | |
| 908 | use Filter::Simple sub { |
| 909 | while (my ($from, $to) = splice @_, 0, 2) { |
| 910 | s/$from/$to/g; |
| 911 | } |
| 912 | }; |
| 913 | |
| 914 | 1; |
| 915 | |
| 916 | # in user's code: |
| 917 | |
| 918 | use MyFilter qr/red/ => 'green'; |
| 919 | |
| 920 | print "red\n"; # this code is filtered, will print "green\n" |
| 921 | print "bored\n"; # this code is filtered, will print "bogreen\n" |
| 922 | |
| 923 | no MyFilter; |
| 924 | |
| 925 | print "red\n"; # this code is not filtered, will print "red\n" |
| 926 | |
| 927 | =item * |
| 928 | |
| 929 | C<File::Temp>, by Tim Jenness, allows one to create temporary files |
| 930 | and directories in an easy, portable, and secure way. See L<File::Temp>. |
| 931 | [561+] |
| 932 | |
| 933 | =item * |
| 934 | |
| 935 | C<Filter::Util::Call>, by Paul Marquess, provides you with the |
| 936 | framework to write I<source filters> in Perl. For most uses, the |
| 937 | frontend Filter::Simple is to be preferred. See L<Filter::Util::Call>. |
| 938 | |
| 939 | =item * |
| 940 | |
| 941 | C<if>, by Ilya Zakharevich, is a new pragma for conditional inclusion |
| 942 | of modules. |
| 943 | |
| 944 | =item * |
| 945 | |
| 946 | L<libnet>, by Graham Barr, is a collection of perl5 modules related |
| 947 | to network programming. See L<Net::FTP>, L<Net::NNTP>, L<Net::Ping> |
| 948 | (not part of libnet, but related), L<Net::POP3>, L<Net::SMTP>, |
| 949 | and L<Net::Time>. |
| 950 | |
| 951 | Perl installation leaves libnet unconfigured; use F<libnetcfg> |
| 952 | to configure it. |
| 953 | |
| 954 | =item * |
| 955 | |
| 956 | C<List::Util>, by Graham Barr, is a selection of general-utility |
| 957 | list subroutines, such as sum(), min(), first(), and shuffle(). |
| 958 | See L<List::Util>. |
| 959 | |
| 960 | =item * |
| 961 | |
| 962 | C<Locale::Constants>, C<Locale::Country>, C<Locale::Currency> |
| 963 | C<Locale::Language>, and L<Locale::Script>, by Neil Bowers, have |
| 964 | been added. They provide the codes for various locale standards, such |
| 965 | as "fr" for France, "usd" for US Dollar, and "ja" for Japanese. |
| 966 | |
| 967 | use Locale::Country; |
| 968 | |
| 969 | $country = code2country('jp'); # $country gets 'Japan' |
| 970 | $code = country2code('Norway'); # $code gets 'no' |
| 971 | |
| 972 | See L<Locale::Constants>, L<Locale::Country>, L<Locale::Currency>, |
| 973 | and L<Locale::Language>. |
| 974 | |
| 975 | =item * |
| 976 | |
| 977 | C<Locale::Maketext>, by Sean Burke, is a localization framework. See |
| 978 | L<Locale::Maketext>, and L<Locale::Maketext::TPJ13>. The latter is an |
| 979 | article about software localization, originally published in The Perl |
| 980 | Journal #13, and republished here with kind permission. |
| 981 | |
| 982 | =item * |
| 983 | |
| 984 | C<Math::BigRat> for big rational numbers, to accompany Math::BigInt and |
| 985 | Math::BigFloat, from Tels. See L<Math::BigRat>. |
| 986 | |
| 987 | =item * |
| 988 | |
| 989 | C<Memoize> can make your functions faster by trading space for time, |
| 990 | from Mark-Jason Dominus. See L<Memoize>. |
| 991 | |
| 992 | =item * |
| 993 | |
| 994 | C<MIME::Base64>, by Gisle Aas, allows you to encode data in base64, |
| 995 | as defined in RFC 2045 - I<MIME (Multipurpose Internet Mail |
| 996 | Extensions)>. |
| 997 | |
| 998 | use MIME::Base64; |
| 999 | |
| 1000 | $encoded = encode_base64('Aladdin:open sesame'); |
| 1001 | $decoded = decode_base64($encoded); |
| 1002 | |
| 1003 | print $encoded, "\n"; # "QWxhZGRpbjpvcGVuIHNlc2FtZQ==" |
| 1004 | |
| 1005 | See L<MIME::Base64>. |
| 1006 | |
| 1007 | =item * |
| 1008 | |
| 1009 | C<MIME::QuotedPrint>, by Gisle Aas, allows you to encode data |
| 1010 | in quoted-printable encoding, as defined in RFC 2045 - I<MIME |
| 1011 | (Multipurpose Internet Mail Extensions)>. |
| 1012 | |
| 1013 | use MIME::QuotedPrint; |
| 1014 | |
| 1015 | $encoded = encode_qp("\xDE\xAD\xBE\xEF"); |
| 1016 | $decoded = decode_qp($encoded); |
| 1017 | |
| 1018 | print $encoded, "\n"; # "=DE=AD=BE=EF\n" |
| 1019 | print $decoded, "\n"; # "\xDE\xAD\xBE\xEF\n" |
| 1020 | |
| 1021 | See also L<PerlIO::via::QuotedPrint>. |
| 1022 | |
| 1023 | =item * |
| 1024 | |
| 1025 | C<NEXT>, by Damian Conway, is a pseudo-class for method redispatch. |
| 1026 | See L<NEXT>. |
| 1027 | |
| 1028 | =item * |
| 1029 | |
| 1030 | C<open> is a new pragma for setting the default I/O layers |
| 1031 | for open(). |
| 1032 | |
| 1033 | =item * |
| 1034 | |
| 1035 | C<PerlIO::scalar>, by Nick Ing-Simmons, provides the implementation |
| 1036 | of IO to "in memory" Perl scalars as discussed above. It also serves |
| 1037 | as an example of a loadable PerlIO layer. Other future possibilities |
| 1038 | include PerlIO::Array and PerlIO::Code. See L<PerlIO::scalar>. |
| 1039 | |
| 1040 | =item * |
| 1041 | |
| 1042 | C<PerlIO::via>, by Nick Ing-Simmons, acts as a PerlIO layer and wraps |
| 1043 | PerlIO layer functionality provided by a class (typically implemented |
| 1044 | in Perl code). |
| 1045 | |
| 1046 | =item * |
| 1047 | |
| 1048 | C<PerlIO::via::QuotedPrint>, by Elizabeth Mattijsen, is an example |
| 1049 | of a C<PerlIO::via> class: |
| 1050 | |
| 1051 | use PerlIO::via::QuotedPrint; |
| 1052 | open($fh,">:via(QuotedPrint)",$path); |
| 1053 | |
| 1054 | This will automatically convert everything output to C<$fh> to |
| 1055 | Quoted-Printable. See L<PerlIO::via> and L<PerlIO::via::QuotedPrint>. |
| 1056 | |
| 1057 | =item * |
| 1058 | |
| 1059 | C<Pod::ParseLink>, by Russ Allbery, has been added, |
| 1060 | to parse LZ<><> links in pods as described in the new |
| 1061 | perlpodspec. |
| 1062 | |
| 1063 | =item * |
| 1064 | |
| 1065 | C<Pod::Text::Overstrike>, by Joe Smith, has been added. |
| 1066 | It converts POD data to formatted overstrike text. |
| 1067 | See L<Pod::Text::Overstrike>. [561+] |
| 1068 | |
| 1069 | =item * |
| 1070 | |
| 1071 | C<Scalar::Util> is a selection of general-utility scalar subroutines, |
| 1072 | such as blessed(), reftype(), and tainted(). See L<Scalar::Util>. |
| 1073 | |
| 1074 | =item * |
| 1075 | |
| 1076 | C<sort> is a new pragma for controlling the behaviour of sort(). |
| 1077 | |
| 1078 | =item * |
| 1079 | |
| 1080 | C<Storable> gives persistence to Perl data structures by allowing the |
| 1081 | storage and retrieval of Perl data to and from files in a fast and |
| 1082 | compact binary format. Because in effect Storable does serialisation |
| 1083 | of Perl data structures, with it you can also clone deep, hierarchical |
| 1084 | datastructures. Storable was originally created by Raphael Manfredi, |
| 1085 | but it is now maintained by Abhijit Menon-Sen. Storable has been |
| 1086 | enhanced to understand the two new hash features, Unicode keys and |
| 1087 | restricted hashes. See L<Storable>. |
| 1088 | |
| 1089 | =item * |
| 1090 | |
| 1091 | C<Switch>, by Damian Conway, has been added. Just by saying |
| 1092 | |
| 1093 | use Switch; |
| 1094 | |
| 1095 | you have C<switch> and C<case> available in Perl. |
| 1096 | |
| 1097 | use Switch; |
| 1098 | |
| 1099 | switch ($val) { |
| 1100 | |
| 1101 | case 1 { print "number 1" } |
| 1102 | case "a" { print "string a" } |
| 1103 | case [1..10,42] { print "number in list" } |
| 1104 | case (@array) { print "number in list" } |
| 1105 | case /\w+/ { print "pattern" } |
| 1106 | case qr/\w+/ { print "pattern" } |
| 1107 | case (%hash) { print "entry in hash" } |
| 1108 | case (\%hash) { print "entry in hash" } |
| 1109 | case (\&sub) { print "arg to subroutine" } |
| 1110 | else { print "previous case not true" } |
| 1111 | } |
| 1112 | |
| 1113 | See L<Switch>. |
| 1114 | |
| 1115 | =item * |
| 1116 | |
| 1117 | C<Test::More>, by Michael Schwern, is yet another framework for writing |
| 1118 | test scripts, more extensive than Test::Simple. See L<Test::More>. |
| 1119 | |
| 1120 | =item * |
| 1121 | |
| 1122 | C<Test::Simple>, by Michael Schwern, has basic utilities for writing |
| 1123 | tests. See L<Test::Simple>. |
| 1124 | |
| 1125 | =item * |
| 1126 | |
| 1127 | C<Text::Balanced>, by Damian Conway, has been added, for extracting |
| 1128 | delimited text sequences from strings. |
| 1129 | |
| 1130 | use Text::Balanced 'extract_delimited'; |
| 1131 | |
| 1132 | ($a, $b) = extract_delimited("'never say never', he never said", "'", ''); |
| 1133 | |
| 1134 | $a will be "'never say never'", $b will be ', he never said'. |
| 1135 | |
| 1136 | In addition to extract_delimited(), there are also extract_bracketed(), |
| 1137 | extract_quotelike(), extract_codeblock(), extract_variable(), |
| 1138 | extract_tagged(), extract_multiple(), gen_delimited_pat(), and |
| 1139 | gen_extract_tagged(). With these, you can implement rather advanced |
| 1140 | parsing algorithms. See L<Text::Balanced>. |
| 1141 | |
| 1142 | =item * |
| 1143 | |
| 1144 | C<threads>, by Arthur Bergman, is an interface to interpreter threads. |
| 1145 | Interpreter threads (ithreads) is the new thread model introduced in |
| 1146 | Perl 5.6 but only available as an internal interface for extension |
| 1147 | writers (and for Win32 Perl for C<fork()> emulation). See L<threads>, |
| 1148 | L<threads::shared>, and L<perlthrtut>. |
| 1149 | |
| 1150 | =item * |
| 1151 | |
| 1152 | C<threads::shared>, by Arthur Bergman, allows data sharing for |
| 1153 | interpreter threads. See L<threads::shared>. |
| 1154 | |
| 1155 | =item * |
| 1156 | |
| 1157 | C<Tie::File>, by Mark-Jason Dominus, associates a Perl array with the |
| 1158 | lines of a file. See L<Tie::File>. |
| 1159 | |
| 1160 | =item * |
| 1161 | |
| 1162 | C<Tie::Memoize>, by Ilya Zakharevich, provides on-demand loaded hashes. |
| 1163 | See L<Tie::Memoize>. |
| 1164 | |
| 1165 | =item * |
| 1166 | |
| 1167 | C<Tie::RefHash::Nestable>, by Edward Avis, allows storing hash |
| 1168 | references (unlike the standard Tie::RefHash) The module is contained |
| 1169 | within Tie::RefHash. See L<Tie::RefHash>. |
| 1170 | |
| 1171 | =item * |
| 1172 | |
| 1173 | C<Time::HiRes>, by Douglas E. Wegscheid, provides high resolution |
| 1174 | timing (ualarm, usleep, and gettimeofday). See L<Time::HiRes>. |
| 1175 | |
| 1176 | =item * |
| 1177 | |
| 1178 | C<Unicode::UCD> offers a querying interface to the Unicode Character |
| 1179 | Database. See L<Unicode::UCD>. |
| 1180 | |
| 1181 | =item * |
| 1182 | |
| 1183 | C<Unicode::Collate>, by SADAHIRO Tomoyuki, implements the UCA |
| 1184 | (Unicode Collation Algorithm) for sorting Unicode strings. |
| 1185 | See L<Unicode::Collate>. |
| 1186 | |
| 1187 | =item * |
| 1188 | |
| 1189 | C<Unicode::Normalize>, by SADAHIRO Tomoyuki, implements the various |
| 1190 | Unicode normalization forms. See L<Unicode::Normalize>. |
| 1191 | |
| 1192 | =item * |
| 1193 | |
| 1194 | C<XS::APItest>, by Tim Jenness, is a test extension that exercises XS |
| 1195 | APIs. Currently only C<printf()> is tested: how to output various |
| 1196 | basic data types from XS. |
| 1197 | |
| 1198 | =item * |
| 1199 | |
| 1200 | C<XS::Typemap>, by Tim Jenness, is a test extension that exercises |
| 1201 | XS typemaps. Nothing gets installed, but the code is worth studying |
| 1202 | for extension writers. |
| 1203 | |
| 1204 | =back |
| 1205 | |
| 1206 | =head2 Updated And Improved Modules and Pragmata |
| 1207 | |
| 1208 | =over 4 |
| 1209 | |
| 1210 | =item * |
| 1211 | |
| 1212 | The following independently supported modules have been updated to the |
| 1213 | newest versions from CPAN: CGI, CPAN, DB_File, File::Spec, File::Temp, |
| 1214 | Getopt::Long, Math::BigFloat, Math::BigInt, the podlators bundle |
| 1215 | (Pod::Man, Pod::Text), Pod::LaTeX [561+], Pod::Parser, Storable, |
| 1216 | Term::ANSIColor, Test, Text-Tabs+Wrap. |
| 1217 | |
| 1218 | =item * |
| 1219 | |
| 1220 | attributes::reftype() now works on tied arguments. |
| 1221 | |
| 1222 | =item * |
| 1223 | |
| 1224 | AutoLoader can now be disabled with C<no AutoLoader;>. |
| 1225 | |
| 1226 | =item * |
| 1227 | |
| 1228 | B::Deparse has been significantly enhanced by Robin Houston. It can |
| 1229 | now deparse almost all of the standard test suite (so that the tests |
| 1230 | still succeed). There is a make target "test.deparse" for trying this |
| 1231 | out. |
| 1232 | |
| 1233 | =item * |
| 1234 | |
| 1235 | Carp now has better interface documentation, and the @CARP_NOT |
| 1236 | interface has been added to get optional control over where errors |
| 1237 | are reported independently of @ISA, by Ben Tilly. |
| 1238 | |
| 1239 | =item * |
| 1240 | |
| 1241 | Class::Struct can now define the classes in compile time. |
| 1242 | |
| 1243 | =item * |
| 1244 | |
| 1245 | Class::Struct now assigns the array/hash element if the accessor |
| 1246 | is called with an array/hash element as the B<sole> argument. |
| 1247 | |
| 1248 | =item * |
| 1249 | |
| 1250 | The return value of Cwd::fastcwd() is now tainted. |
| 1251 | |
| 1252 | =item * |
| 1253 | |
| 1254 | Data::Dumper now has an option to sort hashes. |
| 1255 | |
| 1256 | =item * |
| 1257 | |
| 1258 | Data::Dumper now has an option to dump code references |
| 1259 | using B::Deparse. |
| 1260 | |
| 1261 | =item * |
| 1262 | |
| 1263 | DB_File now supports newer Berkeley DB versions, among |
| 1264 | other improvements. |
| 1265 | |
| 1266 | =item * |
| 1267 | |
| 1268 | Devel::Peek now has an interface for the Perl memory statistics |
| 1269 | (this works only if you are using perl's malloc, and if you have |
| 1270 | compiled with debugging). |
| 1271 | |
| 1272 | =item * |
| 1273 | |
| 1274 | The English module can now be used without the infamous performance |
| 1275 | hit by saying |
| 1276 | |
| 1277 | use English '-no_match_vars'; |
| 1278 | |
| 1279 | (Assuming, of course, that you don't need the troublesome variables |
| 1280 | C<$`>, C<$&>, or C<$'>.) Also, introduced C<@LAST_MATCH_START> and |
| 1281 | C<@LAST_MATCH_END> English aliases for C<@-> and C<@+>. |
| 1282 | |
| 1283 | =item * |
| 1284 | |
| 1285 | ExtUtils::MakeMaker has been significantly cleaned up and fixed. |
| 1286 | The enhanced version has also been backported to earlier releases |
| 1287 | of Perl and submitted to CPAN so that the earlier releases can |
| 1288 | enjoy the fixes. |
| 1289 | |
| 1290 | =item * |
| 1291 | |
| 1292 | The arguments of WriteMakefile() in Makefile.PL are now checked |
| 1293 | for sanity much more carefully than before. This may cause new |
| 1294 | warnings when modules are being installed. See L<ExtUtils::MakeMaker> |
| 1295 | for more details. |
| 1296 | |
| 1297 | =item * |
| 1298 | |
| 1299 | ExtUtils::MakeMaker now uses File::Spec internally, which hopefully |
| 1300 | leads to better portability. |
| 1301 | |
| 1302 | =item * |
| 1303 | |
| 1304 | Fcntl, Socket, and Sys::Syslog have been rewritten by Nicholas Clark |
| 1305 | to use the new-style constant dispatch section (see L<ExtUtils::Constant>). |
| 1306 | This means that they will be more robust and hopefully faster. |
| 1307 | |
| 1308 | =item * |
| 1309 | |
| 1310 | File::Find now chdir()s correctly when chasing symbolic links. [561] |
| 1311 | |
| 1312 | =item * |
| 1313 | |
| 1314 | File::Find now has pre- and post-processing callbacks. It also |
| 1315 | correctly changes directories when chasing symbolic links. Callbacks |
| 1316 | (naughtily) exiting with "next;" instead of "return;" now work. |
| 1317 | |
| 1318 | =item * |
| 1319 | |
| 1320 | File::Find is now (again) reentrant. It also has been made |
| 1321 | more portable. |
| 1322 | |
| 1323 | =item * |
| 1324 | |
| 1325 | The warnings issued by File::Find now belong to their own category. |
| 1326 | You can enable/disable them with C<use/no warnings 'File::Find';>. |
| 1327 | |
| 1328 | =item * |
| 1329 | |
| 1330 | File::Glob::glob() has been renamed to File::Glob::bsd_glob() |
| 1331 | because the name clashes with the builtin glob(). The older |
| 1332 | name is still available for compatibility, but is deprecated. [561] |
| 1333 | |
| 1334 | =item * |
| 1335 | |
| 1336 | File::Glob now supports C<GLOB_LIMIT> constant to limit the size of |
| 1337 | the returned list of filenames. |
| 1338 | |
| 1339 | =item * |
| 1340 | |
| 1341 | IPC::Open3 now allows the use of numeric file descriptors. |
| 1342 | |
| 1343 | =item * |
| 1344 | |
| 1345 | IO::Socket now has an atmark() method, which returns true if the socket |
| 1346 | is positioned at the out-of-band mark. The method is also exportable |
| 1347 | as a sockatmark() function. |
| 1348 | |
| 1349 | =item * |
| 1350 | |
| 1351 | IO::Socket::INET failed to open the specified port if the service name |
| 1352 | was not known. It now correctly uses the supplied port number as is. [561] |
| 1353 | |
| 1354 | =item * |
| 1355 | |
| 1356 | IO::Socket::INET has support for the ReusePort option (if your |
| 1357 | platform supports it). The Reuse option now has an alias, ReuseAddr. |
| 1358 | For clarity, you may want to prefer ReuseAddr. |
| 1359 | |
| 1360 | =item * |
| 1361 | |
| 1362 | IO::Socket::INET now supports a value of zero for C<LocalPort> |
| 1363 | (usually meaning that the operating system will make one up.) |
| 1364 | |
| 1365 | =item * |
| 1366 | |
| 1367 | 'use lib' now works identically to @INC. Removing directories |
| 1368 | with 'no lib' now works. |
| 1369 | |
| 1370 | =item * |
| 1371 | |
| 1372 | Math::BigFloat and Math::BigInt have undergone a full rewrite by Tels. |
| 1373 | They are now magnitudes faster, and they support various bignum |
| 1374 | libraries such as GMP and PARI as their backends. |
| 1375 | |
| 1376 | =item * |
| 1377 | |
| 1378 | Math::Complex handles inf, NaN etc., better. |
| 1379 | |
| 1380 | =item * |
| 1381 | |
| 1382 | Net::Ping has been considerably enhanced by Rob Brown: multihoming is |
| 1383 | now supported, Win32 functionality is better, there is now time |
| 1384 | measuring functionality (optionally high-resolution using |
| 1385 | Time::HiRes), and there is now "external" protocol which uses |
| 1386 | Net::Ping::External module which runs your external ping utility and |
| 1387 | parses the output. A version of Net::Ping::External is available in |
| 1388 | CPAN. |
| 1389 | |
| 1390 | Note that some of the Net::Ping tests are disabled when running |
| 1391 | under the Perl distribution since one cannot assume one or more |
| 1392 | of the following: enabled echo port at localhost, full Internet |
| 1393 | connectivity, or sympathetic firewalls. You can set the environment |
| 1394 | variable PERL_TEST_Net_Ping to "1" (one) before running the Perl test |
| 1395 | suite to enable all the Net::Ping tests. |
| 1396 | |
| 1397 | =item * |
| 1398 | |
| 1399 | POSIX::sigaction() is now much more flexible and robust. |
| 1400 | You can now install coderef handlers, 'DEFAULT', and 'IGNORE' |
| 1401 | handlers, installing new handlers was not atomic. |
| 1402 | |
| 1403 | =item * |
| 1404 | |
| 1405 | In Safe, C<%INC> is now localised in a Safe compartment so that |
| 1406 | use/require work. |
| 1407 | |
| 1408 | =item * |
| 1409 | |
| 1410 | In SDBM_File on dosish platforms, some keys went missing because of |
| 1411 | lack of support for files with "holes". A workaround for the problem |
| 1412 | has been added. |
| 1413 | |
| 1414 | =item * |
| 1415 | |
| 1416 | In Search::Dict one can now have a pre-processing hook for the |
| 1417 | lines being searched. |
| 1418 | |
| 1419 | =item * |
| 1420 | |
| 1421 | The Shell module now has an OO interface. |
| 1422 | |
| 1423 | =item * |
| 1424 | |
| 1425 | In Sys::Syslog there is now a failover mechanism that will go |
| 1426 | through alternative connection mechanisms until the message |
| 1427 | is successfully logged. |
| 1428 | |
| 1429 | =item * |
| 1430 | |
| 1431 | The Test module has been significantly enhanced. |
| 1432 | |
| 1433 | =item * |
| 1434 | |
| 1435 | Time::Local::timelocal() does not handle fractional seconds anymore. |
| 1436 | The rationale is that neither does localtime(), and timelocal() and |
| 1437 | localtime() are supposed to be inverses of each other. |
| 1438 | |
| 1439 | =item * |
| 1440 | |
| 1441 | The vars pragma now supports declaring fully qualified variables. |
| 1442 | (Something that C<our()> does not and will not support.) |
| 1443 | |
| 1444 | =item * |
| 1445 | |
| 1446 | The C<utf8::> name space (as in the pragma) provides various |
| 1447 | Perl-callable functions to provide low level access to Perl's |
| 1448 | internal Unicode representation. At the moment only length() |
| 1449 | has been implemented. |
| 1450 | |
| 1451 | =back |
| 1452 | |
| 1453 | =head1 Utility Changes |
| 1454 | |
| 1455 | =over 4 |
| 1456 | |
| 1457 | =item * |
| 1458 | |
| 1459 | Emacs perl mode (emacs/cperl-mode.el) has been updated to version |
| 1460 | 4.31. |
| 1461 | |
| 1462 | =item * |
| 1463 | |
| 1464 | F<emacs/e2ctags.pl> is now much faster. |
| 1465 | |
| 1466 | =item * |
| 1467 | |
| 1468 | C<enc2xs> is a tool for people adding their own encodings to the |
| 1469 | Encode module. |
| 1470 | |
| 1471 | =item * |
| 1472 | |
| 1473 | C<h2ph> now supports C trigraphs. |
| 1474 | |
| 1475 | =item * |
| 1476 | |
| 1477 | C<h2xs> now produces a template README. |
| 1478 | |
| 1479 | =item * |
| 1480 | |
| 1481 | C<h2xs> now uses C<Devel::PPPort> for better portability between |
| 1482 | different versions of Perl. |
| 1483 | |
| 1484 | =item * |
| 1485 | |
| 1486 | C<h2xs> uses the new L<ExtUtils::Constant|ExtUtils::Constant> module |
| 1487 | which will affect newly created extensions that define constants. |
| 1488 | Since the new code is more correct (if you have two constants where the |
| 1489 | first one is a prefix of the second one, the first constant B<never> |
| 1490 | got defined), less lossy (it uses integers for integer constant, |
| 1491 | as opposed to the old code that used floating point numbers even for |
| 1492 | integer constants), and slightly faster, you might want to consider |
| 1493 | regenerating your extension code (the new scheme makes regenerating |
| 1494 | easy). L<h2xs> now also supports C trigraphs. |
| 1495 | |
| 1496 | =item * |
| 1497 | |
| 1498 | C<libnetcfg> has been added to configure libnet. |
| 1499 | |
| 1500 | =item * |
| 1501 | |
| 1502 | C<perlbug> is now much more robust. It also sends the bug report to |
| 1503 | perl.org, not perl.com. |
| 1504 | |
| 1505 | =item * |
| 1506 | |
| 1507 | C<perlcc> has been rewritten and its user interface (that is, |
| 1508 | command line) is much more like that of the UNIX C compiler, cc. |
| 1509 | (The perlbc tools has been removed. Use C<perlcc -B> instead.) |
| 1510 | B<Note that perlcc is still considered very experimental and |
| 1511 | unsupported.> [561] |
| 1512 | |
| 1513 | =item * |
| 1514 | |
| 1515 | C<perlivp> is a new Installation Verification Procedure utility |
| 1516 | for running any time after installing Perl. |
| 1517 | |
| 1518 | =item * |
| 1519 | |
| 1520 | C<piconv> is an implementation of the character conversion utility |
| 1521 | C<iconv>, demonstrating the new Encode module. |
| 1522 | |
| 1523 | =item * |
| 1524 | |
| 1525 | C<pod2html> now allows specifying a cache directory. |
| 1526 | |
| 1527 | =item * |
| 1528 | |
| 1529 | C<pod2html> now produces XHTML 1.0. |
| 1530 | |
| 1531 | =item * |
| 1532 | |
| 1533 | C<pod2html> now understands POD written using different line endings |
| 1534 | (PC-like CRLF versus UNIX-like LF versus MacClassic-like CR). |
| 1535 | |
| 1536 | =item * |
| 1537 | |
| 1538 | C<s2p> has been completely rewritten in Perl. (It is in fact a full |
| 1539 | implementation of sed in Perl: you can use the sed functionality by |
| 1540 | using the C<psed> utility.) |
| 1541 | |
| 1542 | =item * |
| 1543 | |
| 1544 | C<xsubpp> now understands POD documentation embedded in the *.xs |
| 1545 | files. [561] |
| 1546 | |
| 1547 | =item * |
| 1548 | |
| 1549 | C<xsubpp> now supports the OUT keyword. |
| 1550 | |
| 1551 | =back |
| 1552 | |
| 1553 | =head1 New Documentation |
| 1554 | |
| 1555 | =over 4 |
| 1556 | |
| 1557 | =item * |
| 1558 | |
| 1559 | perl56delta details the changes between the 5.005 release and the |
| 1560 | 5.6.0 release. |
| 1561 | |
| 1562 | =item * |
| 1563 | |
| 1564 | perlclib documents the internal replacements for standard C library |
| 1565 | functions. (Interesting only for extension writers and Perl core |
| 1566 | hackers.) [561+] |
| 1567 | |
| 1568 | =item * |
| 1569 | |
| 1570 | perldebtut is a Perl debugging tutorial. [561+] |
| 1571 | |
| 1572 | =item * |
| 1573 | |
| 1574 | perlebcdic contains considerations for running Perl on EBCDIC |
| 1575 | platforms. [561+] |
| 1576 | |
| 1577 | =item * |
| 1578 | |
| 1579 | perlintro is a gentle introduction to Perl. |
| 1580 | |
| 1581 | =item * |
| 1582 | |
| 1583 | perliol documents the internals of PerlIO with layers. |
| 1584 | |
| 1585 | =item * |
| 1586 | |
| 1587 | perlmodstyle is a style guide for writing modules. |
| 1588 | |
| 1589 | =item * |
| 1590 | |
| 1591 | perlnewmod tells about writing and submitting a new module. [561+] |
| 1592 | |
| 1593 | =item * |
| 1594 | |
| 1595 | perlpacktut is a pack() tutorial. |
| 1596 | |
| 1597 | =item * |
| 1598 | |
| 1599 | perlpod has been rewritten to be clearer and to record the best |
| 1600 | practices gathered over the years. |
| 1601 | |
| 1602 | =item * |
| 1603 | |
| 1604 | perlpodspec is a more formal specification of the pod format, |
| 1605 | mainly of interest for writers of pod applications, not to |
| 1606 | people writing in pod. |
| 1607 | |
| 1608 | =item * |
| 1609 | |
| 1610 | perlretut is a regular expression tutorial. [561+] |
| 1611 | |
| 1612 | =item * |
| 1613 | |
| 1614 | perlrequick is a regular expressions quick-start guide. |
| 1615 | Yes, much quicker than perlretut. [561] |
| 1616 | |
| 1617 | =item * |
| 1618 | |
| 1619 | perltodo has been updated. |
| 1620 | |
| 1621 | =item * |
| 1622 | |
| 1623 | perltootc has been renamed as perltooc (to not to conflict |
| 1624 | with perltoot in filesystems restricted to "8.3" names). |
| 1625 | |
| 1626 | =item * |
| 1627 | |
| 1628 | perluniintro is an introduction to using Unicode in Perl. |
| 1629 | (perlunicode is more of a detailed reference and background |
| 1630 | information) |
| 1631 | |
| 1632 | =item * |
| 1633 | |
| 1634 | perlutil explains the command line utilities packaged with the Perl |
| 1635 | distribution. [561+] |
| 1636 | |
| 1637 | =back |
| 1638 | |
| 1639 | The following platform-specific documents are available before |
| 1640 | the installation as README.I<platform>, and after the installation |
| 1641 | as perlI<platform>: |
| 1642 | |
| 1643 | perlaix perlamiga perlapollo perlbeos perlbs2000 |
| 1644 | perlce perlcygwin perldgux perldos perlepoc perlfreebsd perlhpux |
| 1645 | perlhurd perlirix perlmachten perlmacos perlmint perlmpeix |
| 1646 | perlnetware perlos2 perlos390 perlplan9 perlqnx perlsolaris |
| 1647 | perltru64 perluts perlvmesa perlvms perlvos perlwin32 |
| 1648 | |
| 1649 | These documents usually detail one or more of the following subjects: |
| 1650 | configuring, building, testing, installing, and sometimes also using |
| 1651 | Perl on the said platform. |
| 1652 | |
| 1653 | Eastern Asian Perl users are now welcomed in their own languages: |
| 1654 | README.jp (Japanese), README.ko (Korean), README.cn (simplified |
| 1655 | Chinese) and README.tw (traditional Chinese), which are written in |
| 1656 | normal pod but encoded in EUC-JP, EUC-KR, EUC-CN and Big5. These |
| 1657 | will get installed as |
| 1658 | |
| 1659 | perljp perlko perlcn perltw |
| 1660 | |
| 1661 | =over 4 |
| 1662 | |
| 1663 | =item * |
| 1664 | |
| 1665 | The documentation for the POSIX-BC platform is called "BS2000", to avoid |
| 1666 | confusion with the Perl POSIX module. |
| 1667 | |
| 1668 | =item * |
| 1669 | |
| 1670 | The documentation for the WinCE platform is called perlce (README.ce |
| 1671 | in the source code kit), to avoid confusion with the perlwin32 |
| 1672 | documentation on 8.3-restricted filesystems. |
| 1673 | |
| 1674 | =back |
| 1675 | |
| 1676 | =head1 Performance Enhancements |
| 1677 | |
| 1678 | =over 4 |
| 1679 | |
| 1680 | =item * |
| 1681 | |
| 1682 | map() could get pathologically slow when the result list it generates |
| 1683 | is larger than the source list. The performance has been improved for |
| 1684 | common scenarios. [561] |
| 1685 | |
| 1686 | =item * |
| 1687 | |
| 1688 | sort() is also fully reentrant, in the sense that the sort function |
| 1689 | can itself call sort(). This did not work reliably in previous |
| 1690 | releases. [561] |
| 1691 | |
| 1692 | =item * |
| 1693 | |
| 1694 | sort() has been changed to use primarily mergesort internally as |
| 1695 | opposed to the earlier quicksort. For very small lists this may |
| 1696 | result in slightly slower sorting times, but in general the speedup |
| 1697 | should be at least 20%. Additional bonuses are that the worst case |
| 1698 | behaviour of sort() is now better (in computer science terms it now |
| 1699 | runs in time O(N log N), as opposed to quicksort's Theta(N**2) |
| 1700 | worst-case run time behaviour), and that sort() is now stable |
| 1701 | (meaning that elements with identical keys will stay ordered as they |
| 1702 | were before the sort). See the C<sort> pragma for information. |
| 1703 | |
| 1704 | The story in more detail: suppose you want to serve yourself a little |
| 1705 | slice of Pi. |
| 1706 | |
| 1707 | @digits = ( 3,1,4,1,5,9 ); |
| 1708 | |
| 1709 | A numerical sort of the digits will yield (1,1,3,4,5,9), as expected. |
| 1710 | Which C<1> comes first is hard to know, since one C<1> looks pretty |
| 1711 | much like any other. You can regard this as totally trivial, |
| 1712 | or somewhat profound. However, if you just want to sort the even |
| 1713 | digits ahead of the odd ones, then what will |
| 1714 | |
| 1715 | sort { ($a % 2) <=> ($b % 2) } @digits; |
| 1716 | |
| 1717 | yield? The only even digit, C<4>, will come first. But how about |
| 1718 | the odd numbers, which all compare equal? With the quicksort algorithm |
| 1719 | used to implement Perl 5.6 and earlier, the order of ties is left up |
| 1720 | to the sort. So, as you add more and more digits of Pi, the order |
| 1721 | in which the sorted even and odd digits appear will change. |
| 1722 | and, for sufficiently large slices of Pi, the quicksort algorithm |
| 1723 | in Perl 5.8 won't return the same results even if reinvoked with the |
| 1724 | same input. The justification for this rests with quicksort's |
| 1725 | worst case behavior. If you run |
| 1726 | |
| 1727 | sort { $a <=> $b } ( 1 .. $N , 1 .. $N ); |
| 1728 | |
| 1729 | (something you might approximate if you wanted to merge two sorted |
| 1730 | arrays using sort), doubling $N doesn't just double the quicksort time, |
| 1731 | it I<quadruples> it. Quicksort has a worst case run time that can |
| 1732 | grow like N**2, so-called I<quadratic> behaviour, and it can happen |
| 1733 | on patterns that may well arise in normal use. You won't notice this |
| 1734 | for small arrays, but you I<will> notice it with larger arrays, |
| 1735 | and you may not live long enough for the sort to complete on arrays |
| 1736 | of a million elements. So the 5.8 quicksort scrambles large arrays |
| 1737 | before sorting them, as a statistical defence against quadratic behaviour. |
| 1738 | But that means if you sort the same large array twice, ties may be |
| 1739 | broken in different ways. |
| 1740 | |
| 1741 | Because of the unpredictability of tie-breaking order, and the quadratic |
| 1742 | worst-case behaviour, quicksort was I<almost> replaced completely with |
| 1743 | a stable mergesort. I<Stable> means that ties are broken to preserve |
| 1744 | the original order of appearance in the input array. So |
| 1745 | |
| 1746 | sort { ($a % 2) <=> ($b % 2) } (3,1,4,1,5,9); |
| 1747 | |
| 1748 | will yield (4,3,1,1,5,9), guaranteed. The even and odd numbers |
| 1749 | appear in the output in the same order they appeared in the input. |
| 1750 | Mergesort has worst case O(N log N) behaviour, the best value |
| 1751 | attainable. And, ironically, this mergesort does particularly |
| 1752 | well where quicksort goes quadratic: mergesort sorts (1..$N, 1..$N) |
| 1753 | in O(N) time. But quicksort was rescued at the last moment because |
| 1754 | it is faster than mergesort on certain inputs and platforms. |
| 1755 | For example, if you really I<don't> care about the order of even |
| 1756 | and odd digits, quicksort will run in O(N) time; it's very good |
| 1757 | at sorting many repetitions of a small number of distinct elements. |
| 1758 | The quicksort divide and conquer strategy works well on platforms |
| 1759 | with relatively small, very fast, caches. Eventually, the problem gets |
| 1760 | whittled down to one that fits in the cache, from which point it |
| 1761 | benefits from the increased memory speed. |
| 1762 | |
| 1763 | Quicksort was rescued by implementing a sort pragma to control aspects |
| 1764 | of the sort. The B<stable> subpragma forces stable behaviour, |
| 1765 | regardless of algorithm. The B<_quicksort> and B<_mergesort> |
| 1766 | subpragmas are heavy-handed ways to select the underlying implementation. |
| 1767 | The leading C<_> is a reminder that these subpragmas may not survive |
| 1768 | beyond 5.8. More appropriate mechanisms for selecting the implementation |
| 1769 | exist, but they wouldn't have arrived in time to save quicksort. |
| 1770 | |
| 1771 | =item * |
| 1772 | |
| 1773 | Hashes now use Bob Jenkins "One-at-a-Time" hashing key algorithm |
| 1774 | ( http://burtleburtle.net/bob/hash/doobs.html ). This algorithm is |
| 1775 | reasonably fast while producing a much better spread of values than |
| 1776 | the old hashing algorithm (originally by Chris Torek, later tweaked by |
| 1777 | Ilya Zakharevich). Hash values output from the algorithm on a hash of |
| 1778 | all 3-char printable ASCII keys comes much closer to passing the |
| 1779 | DIEHARD random number generation tests. According to perlbench, this |
| 1780 | change has not affected the overall speed of Perl. |
| 1781 | |
| 1782 | =item * |
| 1783 | |
| 1784 | unshift() should now be noticeably faster. |
| 1785 | |
| 1786 | =back |
| 1787 | |
| 1788 | =head1 Installation and Configuration Improvements |
| 1789 | |
| 1790 | =head2 Generic Improvements |
| 1791 | |
| 1792 | =over 4 |
| 1793 | |
| 1794 | =item * |
| 1795 | |
| 1796 | INSTALL now explains how you can configure Perl to use 64-bit |
| 1797 | integers even on non-64-bit platforms. |
| 1798 | |
| 1799 | =item * |
| 1800 | |
| 1801 | Policy.sh policy change: if you are reusing a Policy.sh file |
| 1802 | (see INSTALL) and you use Configure -Dprefix=/foo/bar and in the old |
| 1803 | Policy $prefix eq $siteprefix and $prefix eq $vendorprefix, all of |
| 1804 | them will now be changed to the new prefix, /foo/bar. (Previously |
| 1805 | only $prefix changed.) If you do not like this new behaviour, |
| 1806 | specify prefix, siteprefix, and vendorprefix explicitly. |
| 1807 | |
| 1808 | =item * |
| 1809 | |
| 1810 | A new optional location for Perl libraries, otherlibdirs, is available. |
| 1811 | It can be used for example for vendor add-ons without disturbing Perl's |
| 1812 | own library directories. |
| 1813 | |
| 1814 | =item * |
| 1815 | |
| 1816 | In many platforms, the vendor-supplied 'cc' is too stripped-down to |
| 1817 | build Perl (basically, 'cc' doesn't do ANSI C). If this seems |
| 1818 | to be the case and 'cc' does not seem to be the GNU C compiler |
| 1819 | 'gcc', an automatic attempt is made to find and use 'gcc' instead. |
| 1820 | |
| 1821 | =item * |
| 1822 | |
| 1823 | gcc needs to closely track the operating system release to avoid |
| 1824 | build problems. If Configure finds that gcc was built for a different |
| 1825 | operating system release than is running, it now gives a clearly visible |
| 1826 | warning that there may be trouble ahead. |
| 1827 | |
| 1828 | =item * |
| 1829 | |
| 1830 | Since Perl 5.8 is not binary-compatible with previous releases |
| 1831 | of Perl, Configure no longer suggests including the 5.005 |
| 1832 | modules in @INC. |
| 1833 | |
| 1834 | =item * |
| 1835 | |
| 1836 | Configure C<-S> can now run non-interactively. [561] |
| 1837 | |
| 1838 | =item * |
| 1839 | |
| 1840 | Configure support for pdp11-style memory models has been removed due |
| 1841 | to obsolescence. [561] |
| 1842 | |
| 1843 | =item * |
| 1844 | |
| 1845 | configure.gnu now works with options with whitespace in them. |
| 1846 | |
| 1847 | =item * |
| 1848 | |
| 1849 | installperl now outputs everything to STDERR. |
| 1850 | |
| 1851 | =item * |
| 1852 | |
| 1853 | Because PerlIO is now the default on most platforms, "-perlio" doesn't |
| 1854 | get appended to the $Config{archname} (also known as $^O) anymore. |
| 1855 | Instead, if you explicitly choose not to use perlio (Configure command |
| 1856 | line option -Uuseperlio), you will get "-stdio" appended. |
| 1857 | |
| 1858 | =item * |
| 1859 | |
| 1860 | Another change related to the architecture name is that "-64all" |
| 1861 | (-Duse64bitall, or "maximally 64-bit") is appended only if your |
| 1862 | pointers are 64 bits wide. (To be exact, the use64bitall is ignored.) |
| 1863 | |
| 1864 | =item * |
| 1865 | |
| 1866 | In AFS installations, one can configure the root of the AFS to be |
| 1867 | somewhere else than the default F</afs> by using the Configure |
| 1868 | parameter C<-Dafsroot=/some/where/else>. |
| 1869 | |
| 1870 | =item * |
| 1871 | |
| 1872 | APPLLIB_EXP, a lesser-known configuration-time definition, has been |
| 1873 | documented. It can be used to prepend site-specific directories |
| 1874 | to Perl's default search path (@INC); see INSTALL for information. |
| 1875 | |
| 1876 | =item * |
| 1877 | |
| 1878 | The version of Berkeley DB used when the Perl (and, presumably, the |
| 1879 | DB_File extension) was built is now available as |
| 1880 | C<@Config{qw(db_version_major db_version_minor db_version_patch)}> |
| 1881 | from Perl and as C<DB_VERSION_MAJOR_CFG DB_VERSION_MINOR_CFG |
| 1882 | DB_VERSION_PATCH_CFG> from C. |
| 1883 | |
| 1884 | =item * |
| 1885 | |
| 1886 | Building Berkeley DB3 for compatibility modes for DB, NDBM, and ODBM |
| 1887 | has been documented in INSTALL. |
| 1888 | |
| 1889 | =item * |
| 1890 | |
| 1891 | If you have CPAN access (either network or a local copy such as a |
| 1892 | CD-ROM) you can during specify extra modules to Configure to build and |
| 1893 | install with Perl using the -Dextras=... option. See INSTALL for |
| 1894 | more details. |
| 1895 | |
| 1896 | =item * |
| 1897 | |
| 1898 | In addition to config.over, a new override file, config.arch, is |
| 1899 | available. This file is supposed to be used by hints file writers |
| 1900 | for architecture-wide changes (as opposed to config.over which is |
| 1901 | for site-wide changes). |
| 1902 | |
| 1903 | =item * |
| 1904 | |
| 1905 | If your file system supports symbolic links, you can build Perl outside |
| 1906 | of the source directory by |
| 1907 | |
| 1908 | mkdir perl/build/directory |
| 1909 | cd perl/build/directory |
| 1910 | sh /path/to/perl/source/Configure -Dmksymlinks ... |
| 1911 | |
| 1912 | This will create in perl/build/directory a tree of symbolic links |
| 1913 | pointing to files in /path/to/perl/source. The original files are left |
| 1914 | unaffected. After Configure has finished, you can just say |
| 1915 | |
| 1916 | make all test |
| 1917 | |
| 1918 | and Perl will be built and tested, all in perl/build/directory. |
| 1919 | [561] |
| 1920 | |
| 1921 | =item * |
| 1922 | |
| 1923 | For Perl developers, several new make targets for profiling |
| 1924 | and debugging have been added; see L<perlhack>. |
| 1925 | |
| 1926 | =over 8 |
| 1927 | |
| 1928 | =item * |
| 1929 | |
| 1930 | Use of the F<gprof> tool to profile Perl has been documented in |
| 1931 | L<perlhack>. There is a make target called "perl.gprof" for |
| 1932 | generating a gprofiled Perl executable. |
| 1933 | |
| 1934 | =item * |
| 1935 | |
| 1936 | If you have GCC 3, there is a make target called "perl.gcov" for |
| 1937 | creating a gcoved Perl executable for coverage analysis. See |
| 1938 | L<perlhack>. |
| 1939 | |
| 1940 | =item * |
| 1941 | |
| 1942 | If you are on IRIX or Tru64 platforms, new profiling/debugging options |
| 1943 | have been added; see L<perlhack> for more information about pixie and |
| 1944 | Third Degree. |
| 1945 | |
| 1946 | =back |
| 1947 | |
| 1948 | =item * |
| 1949 | |
| 1950 | Guidelines of how to construct minimal Perl installations have |
| 1951 | been added to INSTALL. |
| 1952 | |
| 1953 | =item * |
| 1954 | |
| 1955 | The Thread extension is now not built at all under ithreads |
| 1956 | (C<Configure -Duseithreads>) because it wouldn't work anyway (the |
| 1957 | Thread extension requires being Configured with C<-Duse5005threads>). |
| 1958 | |
| 1959 | B<Note that the 5.005 threads are unsupported and deprecated: if you |
| 1960 | have code written for the old threads you should migrate it to the |
| 1961 | new ithreads model.> |
| 1962 | |
| 1963 | =item * |
| 1964 | |
| 1965 | The Gconvert macro ($Config{d_Gconvert}) used by perl for stringifying |
| 1966 | floating-point numbers is now more picky about using sprintf %.*g |
| 1967 | rules for the conversion. Some platforms that used to use gcvt may |
| 1968 | now resort to the slower sprintf. |
| 1969 | |
| 1970 | =item * |
| 1971 | |
| 1972 | The obsolete method of making a special (e.g., debugging) flavor |
| 1973 | of perl by saying |
| 1974 | |
| 1975 | make LIBPERL=libperld.a |
| 1976 | |
| 1977 | has been removed. Use -DDEBUGGING instead. |
| 1978 | |
| 1979 | =back |
| 1980 | |
| 1981 | =head2 New Or Improved Platforms |
| 1982 | |
| 1983 | For the list of platforms known to support Perl, |
| 1984 | see L<perlport/"Supported Platforms">. |
| 1985 | |
| 1986 | =over 4 |
| 1987 | |
| 1988 | =item * |
| 1989 | |
| 1990 | AIX dynamic loading should be now better supported. |
| 1991 | |
| 1992 | =item * |
| 1993 | |
| 1994 | AIX should now work better with gcc, threads, and 64-bitness. Also the |
| 1995 | long doubles support in AIX should be better now. See L<perlaix>. |
| 1996 | |
| 1997 | =item * |
| 1998 | |
| 1999 | AtheOS ( http://www.atheos.cx/ ) is a new platform. |
| 2000 | |
| 2001 | =item * |
| 2002 | |
| 2003 | BeOS has been reclaimed. |
| 2004 | |
| 2005 | =item * |
| 2006 | |
| 2007 | The DG/UX platform now supports 5.005-style threads. |
| 2008 | See L<perldgux>. |
| 2009 | |
| 2010 | =item * |
| 2011 | |
| 2012 | The DYNIX/ptx platform (also known as dynixptx) is supported at or |
| 2013 | near osvers 4.5.2. |
| 2014 | |
| 2015 | =item * |
| 2016 | |
| 2017 | EBCDIC platforms (z/OS (also known as OS/390), POSIX-BC, and VM/ESA) |
| 2018 | have been regained. Many test suite tests still fail and the |
| 2019 | co-existence of Unicode and EBCDIC isn't quite settled, but the |
| 2020 | situation is much better than with Perl 5.6. See L<perlos390>, |
| 2021 | L<perlbs2000> (for POSIX-BC), and L<perlvmesa> for more information. |
| 2022 | |
| 2023 | =item * |
| 2024 | |
| 2025 | Building perl with -Duseithreads or -Duse5005threads now works under |
| 2026 | HP-UX 10.20 (previously it only worked under 10.30 or later). You will |
| 2027 | need a thread library package installed. See README.hpux. [561] |
| 2028 | |
| 2029 | =item * |
| 2030 | |
| 2031 | Mac OS Classic is now supported in the mainstream source package |
| 2032 | (MacPerl has of course been available since perl 5.004 but now the |
| 2033 | source code bases of standard Perl and MacPerl have been synchronised) |
| 2034 | [561] |
| 2035 | |
| 2036 | =item * |
| 2037 | |
| 2038 | Mac OS X (or Darwin) should now be able to build Perl even on HFS+ |
| 2039 | filesystems. (The case-insensitivity used to confuse the Perl build |
| 2040 | process.) |
| 2041 | |
| 2042 | =item * |
| 2043 | |
| 2044 | NCR MP-RAS is now supported. [561] |
| 2045 | |
| 2046 | =item * |
| 2047 | |
| 2048 | All the NetBSD specific patches (except for the installation |
| 2049 | specific ones) have been merged back to the main distribution. |
| 2050 | |
| 2051 | =item * |
| 2052 | |
| 2053 | NetWare from Novell is now supported. See L<perlnetware>. |
| 2054 | |
| 2055 | =item * |
| 2056 | |
| 2057 | NonStop-UX is now supported. [561] |
| 2058 | |
| 2059 | =item * |
| 2060 | |
| 2061 | NEC SUPER-UX is now supported. |
| 2062 | |
| 2063 | =item * |
| 2064 | |
| 2065 | All the OpenBSD specific patches (except for the installation |
| 2066 | specific ones) have been merged back to the main distribution. |
| 2067 | |
| 2068 | =item * |
| 2069 | |
| 2070 | Perl has been tested with the GNU pth userlevel thread package |
| 2071 | ( http://www.gnu.org/software/pth/pth.html ). All thread tests |
| 2072 | of Perl now work, but not without adding some yield()s to the tests, |
| 2073 | so while pth (and other userlevel thread implementations) can be |
| 2074 | considered to be "working" with Perl ithreads, keep in mind the |
| 2075 | possible non-preemptability of the underlying thread implementation. |
| 2076 | |
| 2077 | =item * |
| 2078 | |
| 2079 | Stratus VOS is now supported using Perl's native build method |
| 2080 | (Configure). This is the recommended method to build Perl on |
| 2081 | VOS. The older methods, which build miniperl, are still |
| 2082 | available. See L<perlvos>. [561+] |
| 2083 | |
| 2084 | =item * |
| 2085 | |
| 2086 | The Amdahl UTS UNIX mainframe platform is now supported. [561] |
| 2087 | |
| 2088 | =item * |
| 2089 | |
| 2090 | WinCE is now supported. See L<perlce>. |
| 2091 | |
| 2092 | =item * |
| 2093 | |
| 2094 | z/OS (formerly known as OS/390, formerly known as MVS OE) now has |
| 2095 | support for dynamic loading. This is not selected by default, |
| 2096 | however, you must specify -Dusedl in the arguments of Configure. [561] |
| 2097 | |
| 2098 | =back |
| 2099 | |
| 2100 | =head1 Selected Bug Fixes |
| 2101 | |
| 2102 | Numerous memory leaks and uninitialized memory accesses have been |
| 2103 | hunted down. Most importantly, anonymous subs used to leak quite |
| 2104 | a bit. [561] |
| 2105 | |
| 2106 | =over 4 |
| 2107 | |
| 2108 | =item * |
| 2109 | |
| 2110 | The autouse pragma didn't work for Multi::Part::Function::Names. |
| 2111 | |
| 2112 | =item * |
| 2113 | |
| 2114 | caller() could cause core dumps in certain situations. Carp was |
| 2115 | sometimes affected by this problem. In particular, caller() now |
| 2116 | returns a subroutine name of C<(unknown)> for subroutines that have |
| 2117 | been removed from the symbol table. |
| 2118 | |
| 2119 | =item * |
| 2120 | |
| 2121 | chop(@list) in list context returned the characters chopped in |
| 2122 | reverse order. This has been reversed to be in the right order. [561] |
| 2123 | |
| 2124 | =item * |
| 2125 | |
| 2126 | Configure no longer includes the DBM libraries (dbm, gdbm, db, ndbm) |
| 2127 | when building the Perl binary. The only exception to this is SunOS 4.x, |
| 2128 | which needs them. [561] |
| 2129 | |
| 2130 | =item * |
| 2131 | |
| 2132 | The behaviour of non-decimal but numeric string constants such as |
| 2133 | "0x23" was platform-dependent: in some platforms that was seen as 35, |
| 2134 | in some as 0, in some as a floating point number (don't ask). This |
| 2135 | was caused by Perl's using the operating system libraries in a situation |
| 2136 | where the result of the string to number conversion is undefined: now |
| 2137 | Perl consistently handles such strings as zero in numeric contexts. |
| 2138 | |
| 2139 | =item * |
| 2140 | |
| 2141 | Several debugger fixes: exit code now reflects the script exit code, |
| 2142 | condition C<"0"> now treated correctly, the C<d> command now checks |
| 2143 | line number, C<$.> no longer gets corrupted, and all debugger output |
| 2144 | now goes correctly to the socket if RemotePort is set. [561] |
| 2145 | |
| 2146 | =item * |
| 2147 | |
| 2148 | The debugger (perl5db.pl) has been modified to present a more |
| 2149 | consistent commands interface, via (CommandSet=580). perl5db.t was |
| 2150 | also added to test the changes, and as a placeholder for further tests. |
| 2151 | |
| 2152 | See L<perldebug>. |
| 2153 | |
| 2154 | =item * |
| 2155 | |
| 2156 | The debugger has a new C<dumpDepth> option to control the maximum |
| 2157 | depth to which nested structures are dumped. The C<x> command has |
| 2158 | been extended so that C<x N EXPR> dumps out the value of I<EXPR> to a |
| 2159 | depth of at most I<N> levels. |
| 2160 | |
| 2161 | =item * |
| 2162 | |
| 2163 | The debugger can now show lexical variables if you have the CPAN |
| 2164 | module PadWalker installed. |
| 2165 | |
| 2166 | =item * |
| 2167 | |
| 2168 | The order of DESTROYs has been made more predictable. |
| 2169 | |
| 2170 | =item * |
| 2171 | |
| 2172 | Perl 5.6.0 could emit spurious warnings about redefinition of |
| 2173 | dl_error() when statically building extensions into perl. |
| 2174 | This has been corrected. [561] |
| 2175 | |
| 2176 | =item * |
| 2177 | |
| 2178 | L<dprofpp> -R didn't work. |
| 2179 | |
| 2180 | =item * |
| 2181 | |
| 2182 | C<*foo{FORMAT}> now works. |
| 2183 | |
| 2184 | =item * |
| 2185 | |
| 2186 | Infinity is now recognized as a number. |
| 2187 | |
| 2188 | =item * |
| 2189 | |
| 2190 | UNIVERSAL::isa no longer caches methods incorrectly. (This broke |
| 2191 | the Tk extension with 5.6.0.) [561] |
| 2192 | |
| 2193 | =item * |
| 2194 | |
| 2195 | Lexicals I: lexicals outside an eval "" weren't resolved |
| 2196 | correctly inside a subroutine definition inside the eval "" if they |
| 2197 | were not already referenced in the top level of the eval""ed code. |
| 2198 | |
| 2199 | =item * |
| 2200 | |
| 2201 | Lexicals II: lexicals leaked at file scope into subroutines that |
| 2202 | were declared before the lexicals. |
| 2203 | |
| 2204 | =item * |
| 2205 | |
| 2206 | Lexical warnings now propagating correctly between scopes |
| 2207 | and into C<eval "...">. |
| 2208 | |
| 2209 | =item * |
| 2210 | |
| 2211 | C<use warnings qw(FATAL all)> did not work as intended. This has been |
| 2212 | corrected. [561] |
| 2213 | |
| 2214 | =item * |
| 2215 | |
| 2216 | warnings::enabled() now reports the state of $^W correctly if the caller |
| 2217 | isn't using lexical warnings. [561] |
| 2218 | |
| 2219 | =item * |
| 2220 | |
| 2221 | Line renumbering with eval and C<#line> now works. [561] |
| 2222 | |
| 2223 | =item * |
| 2224 | |
| 2225 | Fixed numerous memory leaks, especially in eval "". |
| 2226 | |
| 2227 | =item * |
| 2228 | |
| 2229 | Localised tied variables no longer leak memory |
| 2230 | |
| 2231 | use Tie::Hash; |
| 2232 | tie my %tied_hash => 'Tie::StdHash'; |
| 2233 | |
| 2234 | ... |
| 2235 | |
| 2236 | # Used to leak memory every time local() was called; |
| 2237 | # in a loop, this added up. |
| 2238 | local($tied_hash{Foo}) = 1; |
| 2239 | |
| 2240 | =item * |
| 2241 | |
| 2242 | Localised hash elements (and %ENV) are correctly unlocalised to not |
| 2243 | exist, if they didn't before they were localised. |
| 2244 | |
| 2245 | |
| 2246 | use Tie::Hash; |
| 2247 | tie my %tied_hash => 'Tie::StdHash'; |
| 2248 | |
| 2249 | ... |
| 2250 | |
| 2251 | # Nothing has set the FOO element so far |
| 2252 | |
| 2253 | { local $tied_hash{FOO} = 'Bar' } |
| 2254 | |
| 2255 | # This used to print, but not now. |
| 2256 | print "exists!\n" if exists $tied_hash{FOO}; |
| 2257 | |
| 2258 | As a side effect of this fix, tied hash interfaces B<must> define |
| 2259 | the EXISTS and DELETE methods. |
| 2260 | |
| 2261 | =item * |
| 2262 | |
| 2263 | mkdir() now ignores trailing slashes in the directory name, |
| 2264 | as mandated by POSIX. |
| 2265 | |
| 2266 | =item * |
| 2267 | |
| 2268 | Some versions of glibc have a broken modfl(). This affects builds |
| 2269 | with C<-Duselongdouble>. This version of Perl detects this brokenness |
| 2270 | and has a workaround for it. The glibc release 2.2.2 is known to have |
| 2271 | fixed the modfl() bug. |
| 2272 | |
| 2273 | =item * |
| 2274 | |
| 2275 | Modulus of unsigned numbers now works (4063328477 % 65535 used to |
| 2276 | return 27406, instead of 27047). [561] |
| 2277 | |
| 2278 | =item * |
| 2279 | |
| 2280 | Some "not a number" warnings introduced in 5.6.0 eliminated to be |
| 2281 | more compatible with 5.005. Infinity is now recognised as a number. [561] |
| 2282 | |
| 2283 | =item * |
| 2284 | |
| 2285 | Numeric conversions did not recognize changes in the string value |
| 2286 | properly in certain circumstances. [561] |
| 2287 | |
| 2288 | =item * |
| 2289 | |
| 2290 | Attributes (such as :shared) didn't work with our(). |
| 2291 | |
| 2292 | =item * |
| 2293 | |
| 2294 | our() variables will not cause bogus "Variable will not stay shared" |
| 2295 | warnings. [561] |
| 2296 | |
| 2297 | =item * |
| 2298 | |
| 2299 | "our" variables of the same name declared in two sibling blocks |
| 2300 | resulted in bogus warnings about "redeclaration" of the variables. |
| 2301 | The problem has been corrected. [561] |
| 2302 | |
| 2303 | =item * |
| 2304 | |
| 2305 | pack "Z" now correctly terminates the string with "\0". |
| 2306 | |
| 2307 | =item * |
| 2308 | |
| 2309 | Fix password routines which in some shadow password platforms |
| 2310 | (e.g. HP-UX) caused getpwent() to return every other entry. |
| 2311 | |
| 2312 | =item * |
| 2313 | |
| 2314 | The PERL5OPT environment variable (for passing command line arguments |
| 2315 | to Perl) didn't work for more than a single group of options. [561] |
| 2316 | |
| 2317 | =item * |
| 2318 | |
| 2319 | PERL5OPT with embedded spaces didn't work. |
| 2320 | |
| 2321 | =item * |
| 2322 | |
| 2323 | printf() no longer resets the numeric locale to "C". |
| 2324 | |
| 2325 | =item * |
| 2326 | |
| 2327 | C<qw(a\\b)> now parses correctly as C<'a\\b'>: that is, as three |
| 2328 | characters, not four. [561] |
| 2329 | |
| 2330 | =item * |
| 2331 | |
| 2332 | pos() did not return the correct value within s///ge in earlier |
| 2333 | versions. This is now handled correctly. [561] |
| 2334 | |
| 2335 | =item * |
| 2336 | |
| 2337 | Printing quads (64-bit integers) with printf/sprintf now works |
| 2338 | without the q L ll prefixes (assuming you are on a quad-capable platform). |
| 2339 | |
| 2340 | =item * |
| 2341 | |
| 2342 | Regular expressions on references and overloaded scalars now work. [561+] |
| 2343 | |
| 2344 | =item * |
| 2345 | |
| 2346 | Right-hand side magic (GMAGIC) could in many cases such as string |
| 2347 | concatenation be invoked too many times. |
| 2348 | |
| 2349 | =item * |
| 2350 | |
| 2351 | scalar() now forces scalar context even when used in void context. |
| 2352 | |
| 2353 | =item * |
| 2354 | |
| 2355 | SOCKS support is now much more robust. |
| 2356 | |
| 2357 | =item * |
| 2358 | |
| 2359 | sort() arguments are now compiled in the right wantarray context |
| 2360 | (they were accidentally using the context of the sort() itself). |
| 2361 | The comparison block is now run in scalar context, and the arguments |
| 2362 | to be sorted are always provided list context. [561] |
| 2363 | |
| 2364 | =item * |
| 2365 | |
| 2366 | Changed the POSIX character class C<[[:space:]]> to include the (very |
| 2367 | rarely used) vertical tab character. Added a new POSIX-ish character |
| 2368 | class C<[[:blank:]]> which stands for horizontal whitespace |
| 2369 | (currently, the space and the tab). |
| 2370 | |
| 2371 | =item * |
| 2372 | |
| 2373 | The tainting behaviour of sprintf() has been rationalized. It does |
| 2374 | not taint the result of floating point formats anymore, making the |
| 2375 | behaviour consistent with that of string interpolation. [561] |
| 2376 | |
| 2377 | =item * |
| 2378 | |
| 2379 | Some cases of inconsistent taint propagation (such as within hash |
| 2380 | values) have been fixed. |
| 2381 | |
| 2382 | =item * |
| 2383 | |
| 2384 | The RE engine found in Perl 5.6.0 accidentally pessimised certain kinds |
| 2385 | of simple pattern matches. These are now handled better. [561] |
| 2386 | |
| 2387 | =item * |
| 2388 | |
| 2389 | Regular expression debug output (whether through C<use re 'debug'> |
| 2390 | or via C<-Dr>) now looks better. [561] |
| 2391 | |
| 2392 | =item * |
| 2393 | |
| 2394 | Multi-line matches like C<"a\nxb\n" =~ /(?!\A)x/m> were flawed. The |
| 2395 | bug has been fixed. [561] |
| 2396 | |
| 2397 | =item * |
| 2398 | |
| 2399 | Use of $& could trigger a core dump under some situations. This |
| 2400 | is now avoided. [561] |
| 2401 | |
| 2402 | =item * |
| 2403 | |
| 2404 | The regular expression captured submatches ($1, $2, ...) are now |
| 2405 | more consistently unset if the match fails, instead of leaving false |
| 2406 | data lying around in them. [561] |
| 2407 | |
| 2408 | =item * |
| 2409 | |
| 2410 | readline() on files opened in "slurp" mode could return an extra |
| 2411 | "" (blank line) at the end in certain situations. This has been |
| 2412 | corrected. [561] |
| 2413 | |
| 2414 | =item * |
| 2415 | |
| 2416 | Autovivification of symbolic references of special variables described |
| 2417 | in L<perlvar> (as in C<${$num}>) was accidentally disabled. This works |
| 2418 | again now. [561] |
| 2419 | |
| 2420 | =item * |
| 2421 | |
| 2422 | Sys::Syslog ignored the C<LOG_AUTH> constant. |
| 2423 | |
| 2424 | =item * |
| 2425 | |
| 2426 | $AUTOLOAD, sort(), lock(), and spawning subprocesses |
| 2427 | in multiple threads simultaneously are now thread-safe. |
| 2428 | |
| 2429 | =item * |
| 2430 | |
| 2431 | Tie::Array's SPLICE method was broken. |
| 2432 | |
| 2433 | =item * |
| 2434 | |
| 2435 | Allow a read-only string on the left-hand side of a non-modifying tr///. |
| 2436 | |
| 2437 | =item * |
| 2438 | |
| 2439 | If C<STDERR> is tied, warnings caused by C<warn> and C<die> now |
| 2440 | correctly pass to it. |
| 2441 | |
| 2442 | =item * |
| 2443 | |
| 2444 | Several Unicode fixes. |
| 2445 | |
| 2446 | =over 8 |
| 2447 | |
| 2448 | =item * |
| 2449 | |
| 2450 | BOMs (byte order marks) at the beginning of Perl files |
| 2451 | (scripts, modules) should now be transparently skipped. |
| 2452 | UTF-16 and UCS-2 encoded Perl files should now be read correctly. |
| 2453 | |
| 2454 | =item * |
| 2455 | |
| 2456 | The character tables have been updated to Unicode 3.2.0. |
| 2457 | |
| 2458 | =item * |
| 2459 | |
| 2460 | Comparing with utf8 data does not magically upgrade non-utf8 data |
| 2461 | into utf8. (This was a problem for example if you were mixing data |
| 2462 | from I/O and Unicode data: your output might have got magically encoded |
| 2463 | as UTF-8.) |
| 2464 | |
| 2465 | =item * |
| 2466 | |
| 2467 | Generating illegal Unicode code points such as U+FFFE, or the UTF-16 |
| 2468 | surrogates, now also generates an optional warning. |
| 2469 | |
| 2470 | =item * |
| 2471 | |
| 2472 | C<IsAlnum>, C<IsAlpha>, and C<IsWord> now match titlecase. |
| 2473 | |
| 2474 | =item * |
| 2475 | |
| 2476 | Concatenation with the C<.> operator or via variable interpolation, |
| 2477 | C<eq>, C<substr>, C<reverse>, C<quotemeta>, the C<x> operator, |
| 2478 | substitution with C<s///>, single-quoted UTF-8, should now work. |
| 2479 | |
| 2480 | =item * |
| 2481 | |
| 2482 | The C<tr///> operator now works. Note that the C<tr///CU> |
| 2483 | functionality has been removed (but see pack('U0', ...)). |
| 2484 | |
| 2485 | =item * |
| 2486 | |
| 2487 | C<eval "v200"> now works. |
| 2488 | |
| 2489 | =item * |
| 2490 | |
| 2491 | Perl 5.6.0 parsed m/\x{ab}/ incorrectly, leading to spurious warnings. |
| 2492 | This has been corrected. [561] |
| 2493 | |
| 2494 | =item * |
| 2495 | |
| 2496 | Zero entries were missing from the Unicode classes such as C<IsDigit>. |
| 2497 | |
| 2498 | =back |
| 2499 | |
| 2500 | =item * |
| 2501 | |
| 2502 | Large unsigned numbers (those above 2**31) could sometimes lose their |
| 2503 | unsignedness, causing bogus results in arithmetic operations. [561] |
| 2504 | |
| 2505 | =item * |
| 2506 | |
| 2507 | The Perl parser has been stress tested using both random input and |
| 2508 | Markov chain input and the few found crashes and lockups have been |
| 2509 | fixed. |
| 2510 | |
| 2511 | =back |
| 2512 | |
| 2513 | =head2 Platform Specific Changes and Fixes |
| 2514 | |
| 2515 | =over 4 |
| 2516 | |
| 2517 | =item * |
| 2518 | |
| 2519 | BSDI 4.* |
| 2520 | |
| 2521 | Perl now works on post-4.0 BSD/OSes. |
| 2522 | |
| 2523 | =item * |
| 2524 | |
| 2525 | All BSDs |
| 2526 | |
| 2527 | Setting C<$0> now works (as much as possible; see L<perlvar> for details). |
| 2528 | |
| 2529 | =item * |
| 2530 | |
| 2531 | Cygwin |
| 2532 | |
| 2533 | Numerous updates; currently synchronised with Cygwin 1.3.10. |
| 2534 | |
| 2535 | =item * |
| 2536 | |
| 2537 | Previously DYNIX/ptx had problems in its Configure probe for non-blocking I/O. |
| 2538 | |
| 2539 | =item * |
| 2540 | |
| 2541 | EPOC |
| 2542 | |
| 2543 | EPOC now better supported. See README.epoc. [561] |
| 2544 | |
| 2545 | =item * |
| 2546 | |
| 2547 | FreeBSD 3.* |
| 2548 | |
| 2549 | Perl now works on post-3.0 FreeBSDs. |
| 2550 | |
| 2551 | =item * |
| 2552 | |
| 2553 | HP-UX |
| 2554 | |
| 2555 | README.hpux updated; C<Configure -Duse64bitall> now works; |
| 2556 | now uses HP-UX malloc instead of Perl malloc. |
| 2557 | |
| 2558 | =item * |
| 2559 | |
| 2560 | IRIX |
| 2561 | |
| 2562 | Numerous compilation flag and hint enhancements; accidental mixing |
| 2563 | of 32-bit and 64-bit libraries (a doomed attempt) made much harder. |
| 2564 | |
| 2565 | =item * |
| 2566 | |
| 2567 | Linux |
| 2568 | |
| 2569 | =over 8 |
| 2570 | |
| 2571 | =item * |
| 2572 | |
| 2573 | Long doubles should now work (see INSTALL). [561] |
| 2574 | |
| 2575 | =item * |
| 2576 | |
| 2577 | Linux previously had problems related to sockaddrlen when using |
| 2578 | accept(), recvfrom() (in Perl: recv()), getpeername(), and |
| 2579 | getsockname(). |
| 2580 | |
| 2581 | =back |
| 2582 | |
| 2583 | =item * |
| 2584 | |
| 2585 | Mac OS Classic |
| 2586 | |
| 2587 | Compilation of the standard Perl distribution in Mac OS Classic should |
| 2588 | now work if you have the Metrowerks development environment and the |
| 2589 | missing Mac-specific toolkit bits. Contact the macperl mailing list |
| 2590 | for details. |
| 2591 | |
| 2592 | =item * |
| 2593 | |
| 2594 | MPE/iX |
| 2595 | |
| 2596 | MPE/iX update after Perl 5.6.0. See README.mpeix. [561] |
| 2597 | |
| 2598 | =item * |
| 2599 | |
| 2600 | NetBSD/threads: try installing the GNU pth (should be in the |
| 2601 | packages collection, or http://www.gnu.org/software/pth/), |
| 2602 | and Configure with -Duseithreads. |
| 2603 | |
| 2604 | =item * |
| 2605 | |
| 2606 | NetBSD/sparc |
| 2607 | |
| 2608 | Perl now works on NetBSD/sparc. |
| 2609 | |
| 2610 | =item * |
| 2611 | |
| 2612 | OS/2 |
| 2613 | |
| 2614 | Now works with usethreads (see INSTALL). [561] |
| 2615 | |
| 2616 | =item * |
| 2617 | |
| 2618 | Solaris |
| 2619 | |
| 2620 | 64-bitness using the Sun Workshop compiler now works. |
| 2621 | |
| 2622 | =item * |
| 2623 | |
| 2624 | Stratus VOS |
| 2625 | |
| 2626 | The native build method requires at least VOS Release 14.5.0 |
| 2627 | and GNU C++/GNU Tools 2.0.1 or later. The Perl pack function |
| 2628 | now maps overflowed values to +infinity and underflowed values |
| 2629 | to -infinity. |
| 2630 | |
| 2631 | =item * |
| 2632 | |
| 2633 | Tru64 (aka Digital UNIX, aka DEC OSF/1) |
| 2634 | |
| 2635 | The operating system version letter now recorded in $Config{osvers}. |
| 2636 | Allow compiling with gcc (previously explicitly forbidden). Compiling |
| 2637 | with gcc still not recommended because buggy code results, even with |
| 2638 | gcc 2.95.2. |
| 2639 | |
| 2640 | =item * |
| 2641 | |
| 2642 | Unicos |
| 2643 | |
| 2644 | Fixed various alignment problems that lead into core dumps either |
| 2645 | during build or later; no longer dies on math errors at runtime; |
| 2646 | now using full quad integers (64 bits), previously was using |
| 2647 | only 46 bit integers for speed. |
| 2648 | |
| 2649 | =item * |
| 2650 | |
| 2651 | VMS |
| 2652 | |
| 2653 | See L</"Socket Extension Dynamic in VMS"> and L</"IEEE-format Floating Point |
| 2654 | Default on OpenVMS Alpha"> for important changes not otherwise listed here. |
| 2655 | |
| 2656 | chdir() now works better despite a CRT bug; now works with MULTIPLICITY |
| 2657 | (see INSTALL); now works with Perl's malloc. |
| 2658 | |
| 2659 | The tainting of C<%ENV> elements via C<keys> or C<values> was previously |
| 2660 | unimplemented. It now works as documented. |
| 2661 | |
| 2662 | The C<waitpid> emulation has been improved. The worst bug (now fixed) |
| 2663 | was that a pid of -1 would cause a wildcard search of all processes on |
| 2664 | the system. |
| 2665 | |
| 2666 | POSIX-style signals are now emulated much better on VMS versions prior |
| 2667 | to 7.0. |
| 2668 | |
| 2669 | The C<system> function and backticks operator have improved |
| 2670 | functionality and better error handling. [561] |
| 2671 | |
| 2672 | File access tests now use current process privileges rather than the |
| 2673 | user's default privileges, which could sometimes result in a mismatch |
| 2674 | between reported access and actual access. This improvement is only |
| 2675 | available on VMS v6.0 and later. |
| 2676 | |
| 2677 | There is a new C<kill> implementation based on C<sys$sigprc> that allows |
| 2678 | older VMS systems (pre-7.0) to use C<kill> to send signals rather than |
| 2679 | simply force exit. This implementation also allows later systems to |
| 2680 | call C<kill> from within a signal handler. |
| 2681 | |
| 2682 | Iterative logical name translations are now limited to 10 iterations in |
| 2683 | imitation of SHOW LOGICAL and other OpenVMS facilities. |
| 2684 | |
| 2685 | =item * |
| 2686 | |
| 2687 | Windows |
| 2688 | |
| 2689 | =over 8 |
| 2690 | |
| 2691 | =item * |
| 2692 | |
| 2693 | Signal handling now works better than it used to. It is now implemented |
| 2694 | using a Windows message loop, and is therefore less prone to random |
| 2695 | crashes. |
| 2696 | |
| 2697 | =item * |
| 2698 | |
| 2699 | fork() emulation is now more robust, but still continues to have a few |
| 2700 | esoteric bugs and caveats. See L<perlfork> for details. [561+] |
| 2701 | |
| 2702 | =item * |
| 2703 | |
| 2704 | A failed (pseudo)fork now returns undef and sets errno to EAGAIN. [561] |
| 2705 | |
| 2706 | =item * |
| 2707 | |
| 2708 | The following modules now work on Windows: |
| 2709 | |
| 2710 | ExtUtils::Embed [561] |
| 2711 | IO::Pipe |
| 2712 | IO::Poll |
| 2713 | Net::Ping |
| 2714 | |
| 2715 | =item * |
| 2716 | |
| 2717 | IO::File::new_tmpfile() is no longer limited to 32767 invocations |
| 2718 | per-process. |
| 2719 | |
| 2720 | =item * |
| 2721 | |
| 2722 | Better chdir() return value for a non-existent directory. |
| 2723 | |
| 2724 | =item * |
| 2725 | |
| 2726 | Compiling perl using the 64-bit Platform SDK tools is now supported. |
| 2727 | |
| 2728 | =item * |
| 2729 | |
| 2730 | The Win32::SetChildShowWindow() builtin can be used to control the |
| 2731 | visibility of windows created by child processes. See L<Win32> for |
| 2732 | details. |
| 2733 | |
| 2734 | =item * |
| 2735 | |
| 2736 | Non-blocking waits for child processes (or pseudo-processes) are |
| 2737 | supported via C<waitpid($pid, &POSIX::WNOHANG)>. |
| 2738 | |
| 2739 | =item * |
| 2740 | |
| 2741 | The behavior of system() with multiple arguments has been rationalized. |
| 2742 | Each unquoted argument will be automatically quoted to protect whitespace, |
| 2743 | and any existing whitespace in the arguments will be preserved. This |
| 2744 | improves the portability of system(@args) by avoiding the need for |
| 2745 | Windows C<cmd> shell specific quoting in perl programs. |
| 2746 | |
| 2747 | Note that this means that some scripts that may have relied on earlier |
| 2748 | buggy behavior may no longer work correctly. For example, |
| 2749 | C<system("nmake /nologo", @args)> will now attempt to run the file |
| 2750 | C<nmake /nologo> and will fail when such a file isn't found. |
| 2751 | On the other hand, perl will now execute code such as |
| 2752 | C<system("c:/Program Files/MyApp/foo.exe", @args)> correctly. |
| 2753 | |
| 2754 | =item * |
| 2755 | |
| 2756 | The perl header files no longer suppress common warnings from the |
| 2757 | Microsoft Visual C++ compiler. This means that additional warnings may |
| 2758 | now show up when compiling XS code. |
| 2759 | |
| 2760 | =item * |
| 2761 | |
| 2762 | Borland C++ v5.5 is now a supported compiler that can build Perl. |
| 2763 | However, the generated binaries continue to be incompatible with those |
| 2764 | generated by the other supported compilers (GCC and Visual C++). [561] |
| 2765 | |
| 2766 | =item * |
| 2767 | |
| 2768 | Duping socket handles with open(F, ">&MYSOCK") now works under Windows 9x. |
| 2769 | [561] |
| 2770 | |
| 2771 | =item * |
| 2772 | |
| 2773 | Current directory entries in %ENV are now correctly propagated to child |
| 2774 | processes. [561] |
| 2775 | |
| 2776 | =item * |
| 2777 | |
| 2778 | New %ENV entries now propagate to subprocesses. [561] |
| 2779 | |
| 2780 | =item * |
| 2781 | |
| 2782 | Win32::GetCwd() correctly returns C:\ instead of C: when at the drive root. |
| 2783 | Other bugs in chdir() and Cwd::cwd() have also been fixed. [561] |
| 2784 | |
| 2785 | =item * |
| 2786 | |
| 2787 | The makefiles now default to the features enabled in ActiveState ActivePerl |
| 2788 | (a popular Win32 binary distribution). [561] |
| 2789 | |
| 2790 | =item * |
| 2791 | |
| 2792 | HTML files will now be installed in c:\perl\html instead of |
| 2793 | c:\perl\lib\pod\html |
| 2794 | |
| 2795 | =item * |
| 2796 | |
| 2797 | REG_EXPAND_SZ keys are now allowed in registry settings used by perl. [561] |
| 2798 | |
| 2799 | =item * |
| 2800 | |
| 2801 | Can now send() from all threads, not just the first one. [561] |
| 2802 | |
| 2803 | =item * |
| 2804 | |
| 2805 | ExtUtils::MakeMaker now uses $ENV{LIB} to search for libraries. [561] |
| 2806 | |
| 2807 | =item * |
| 2808 | |
| 2809 | Less stack reserved per thread so that more threads can run |
| 2810 | concurrently. (Still 16M per thread.) [561] |
| 2811 | |
| 2812 | =item * |
| 2813 | |
| 2814 | C<< File::Spec->tmpdir() >> now prefers C:/temp over /tmp |
| 2815 | (works better when perl is running as service). |
| 2816 | |
| 2817 | =item * |
| 2818 | |
| 2819 | Better UNC path handling under ithreads. [561] |
| 2820 | |
| 2821 | =item * |
| 2822 | |
| 2823 | wait(), waitpid(), and backticks now return the correct exit status |
| 2824 | under Windows 9x. [561] |
| 2825 | |
| 2826 | =item * |
| 2827 | |
| 2828 | A socket handle leak in accept() has been fixed. [561] |
| 2829 | |
| 2830 | =back |
| 2831 | |
| 2832 | =back |
| 2833 | |
| 2834 | =head1 New or Changed Diagnostics |
| 2835 | |
| 2836 | Please see L<perldiag> for more details. |
| 2837 | |
| 2838 | =over 4 |
| 2839 | |
| 2840 | =item * |
| 2841 | |
| 2842 | Ambiguous range in the transliteration operator (like a-z-9) now |
| 2843 | gives a warning. |
| 2844 | |
| 2845 | =item * |
| 2846 | |
| 2847 | chdir("") and chdir(undef) now give a deprecation warning because they |
| 2848 | cause a possible unintentional chdir to the home directory. |
| 2849 | Say chdir() if you really mean that. |
| 2850 | |
| 2851 | =item * |
| 2852 | |
| 2853 | Two new debugging options have been added: if you have compiled your |
| 2854 | Perl with debugging, you can use the -DT [561] and -DR options to trace |
| 2855 | tokenising and to add reference counts to displaying variables, |
| 2856 | respectively. |
| 2857 | |
| 2858 | =item * |
| 2859 | |
| 2860 | The lexical warnings category "deprecated" is no longer a sub-category |
| 2861 | of the "syntax" category. It is now a top-level category in its own |
| 2862 | right. |
| 2863 | |
| 2864 | =item * |
| 2865 | |
| 2866 | Unadorned dump() will now give a warning suggesting to |
| 2867 | use explicit CORE::dump() if that's what really is meant. |
| 2868 | |
| 2869 | =item * |
| 2870 | |
| 2871 | The "Unrecognized escape" warning has been extended to include C<\8>, |
| 2872 | C<\9>, and C<\_>. There is no need to escape any of the C<\w> characters. |
| 2873 | |
| 2874 | =item * |
| 2875 | |
| 2876 | All regular expression compilation error messages are now hopefully |
| 2877 | easier to understand both because the error message now comes before |
| 2878 | the failed regex and because the point of failure is now clearly |
| 2879 | marked by a C<E<lt>-- HERE> marker. |
| 2880 | |
| 2881 | =item * |
| 2882 | |
| 2883 | Various I/O (and socket) functions like binmode(), close(), and so |
| 2884 | forth now more consistently warn if they are used illogically either |
| 2885 | on a yet unopened or on an already closed filehandle (or socket). |
| 2886 | |
| 2887 | =item * |
| 2888 | |
| 2889 | Using lstat() on a filehandle now gives a warning. (It's a non-sensical |
| 2890 | thing to do.) |
| 2891 | |
| 2892 | =item * |
| 2893 | |
| 2894 | The C<-M> and C<-m> options now warn if you didn't supply the module name. |
| 2895 | |
| 2896 | =item * |
| 2897 | |
| 2898 | If you in C<use> specify a required minimum version, modules matching |
| 2899 | the name and but not defining a $VERSION will cause a fatal failure. |
| 2900 | |
| 2901 | =item * |
| 2902 | |
| 2903 | Using negative offset for vec() in lvalue context is now a warnable offense. |
| 2904 | |
| 2905 | =item * |
| 2906 | |
| 2907 | Odd number of arguments to overload::constant now elicits a warning. |
| 2908 | |
| 2909 | =item * |
| 2910 | |
| 2911 | Odd number of elements in anonymous hash now elicits a warning. |
| 2912 | |
| 2913 | =item * |
| 2914 | |
| 2915 | The various "opened only for", "on closed", "never opened" warnings |
| 2916 | drop the C<main::> prefix for filehandles in the C<main> package, |
| 2917 | for example C<STDIN> instead of C<main::STDIN>. |
| 2918 | |
| 2919 | =item * |
| 2920 | |
| 2921 | Subroutine prototypes are now checked more carefully, you may |
| 2922 | get warnings for example if you have used non-prototype characters. |
| 2923 | |
| 2924 | =item * |
| 2925 | |
| 2926 | If an attempt to use a (non-blessed) reference as an array index |
| 2927 | is made, a warning is given. |
| 2928 | |
| 2929 | =item * |
| 2930 | |
| 2931 | C<push @a;> and C<unshift @a;> (with no values to push or unshift) |
| 2932 | now give a warning. This may be a problem for generated and evaled |
| 2933 | code. |
| 2934 | |
| 2935 | =item * |
| 2936 | |
| 2937 | If you try to L<perlfunc/pack> a number less than 0 or larger than 255 |
| 2938 | using the C<"C"> format you will get an optional warning. Similarly |
| 2939 | for the C<"c"> format and a number less than -128 or more than 127. |
| 2940 | |
| 2941 | =item * |
| 2942 | |
| 2943 | pack C<P> format now demands an explicit size. |
| 2944 | |
| 2945 | =item * |
| 2946 | |
| 2947 | unpack C<w> now warns of unterminated compressed integers. |
| 2948 | |
| 2949 | =item * |
| 2950 | |
| 2951 | Warnings relating to the use of PerlIO have been added. |
| 2952 | |
| 2953 | =item * |
| 2954 | |
| 2955 | Certain regex modifiers such as C<(?o)> make sense only if applied to |
| 2956 | the entire regex. You will get an optional warning if you try to do |
| 2957 | otherwise. |
| 2958 | |
| 2959 | =item * |
| 2960 | |
| 2961 | Variable length lookbehind has not yet been implemented, trying to |
| 2962 | use it will tell that. |
| 2963 | |
| 2964 | =item * |
| 2965 | |
| 2966 | Using arrays or hashes as references (e.g. C<< %foo->{bar} >> |
| 2967 | has been deprecated for a while. Now you will get an optional warning. |
| 2968 | |
| 2969 | =item * |
| 2970 | |
| 2971 | Warnings relating to the use of the new restricted hashes feature |
| 2972 | have been added. |
| 2973 | |
| 2974 | =item * |
| 2975 | |
| 2976 | Self-ties of arrays and hashes are not supported and fatal errors |
| 2977 | will happen even at an attempt to do so. |
| 2978 | |
| 2979 | =item * |
| 2980 | |
| 2981 | Using C<sort> in scalar context now issues an optional warning. |
| 2982 | This didn't do anything useful, as the sort was not performed. |
| 2983 | |
| 2984 | =item * |
| 2985 | |
| 2986 | Using the /g modifier in split() is meaningless and will cause a warning. |
| 2987 | |
| 2988 | =item * |
| 2989 | |
| 2990 | Using splice() past the end of an array now causes a warning. |
| 2991 | |
| 2992 | =item * |
| 2993 | |
| 2994 | Malformed Unicode encodings (UTF-8 and UTF-16) cause a lot of warnings, |
| 2995 | as does trying to use UTF-16 surrogates (which are unimplemented). |
| 2996 | |
| 2997 | =item * |
| 2998 | |
| 2999 | Trying to use Unicode characters on an I/O stream without marking the |
| 3000 | stream's encoding (using open() or binmode()) will cause "Wide character" |
| 3001 | warnings. |
| 3002 | |
| 3003 | =item * |
| 3004 | |
| 3005 | Use of v-strings in use/require causes a (backward) portability warning. |
| 3006 | |
| 3007 | =item * |
| 3008 | |
| 3009 | Warnings relating to the use interpreter threads and their shared data |
| 3010 | have been added. |
| 3011 | |
| 3012 | =back |
| 3013 | |
| 3014 | =head1 Changed Internals |
| 3015 | |
| 3016 | =over 4 |
| 3017 | |
| 3018 | =item * |
| 3019 | |
| 3020 | PerlIO is now the default. |
| 3021 | |
| 3022 | =item * |
| 3023 | |
| 3024 | perlapi.pod (a companion to perlguts) now attempts to document the |
| 3025 | internal API. |
| 3026 | |
| 3027 | =item * |
| 3028 | |
| 3029 | You can now build a really minimal perl called microperl. |
| 3030 | Building microperl does not require even running Configure; |
| 3031 | C<make -f Makefile.micro> should be enough. Beware: microperl makes |
| 3032 | many assumptions, some of which may be too bold; the resulting |
| 3033 | executable may crash or otherwise misbehave in wondrous ways. |
| 3034 | For careful hackers only. |
| 3035 | |
| 3036 | =item * |
| 3037 | |
| 3038 | Added rsignal(), whichsig(), do_join(), op_clear, op_null, |
| 3039 | ptr_table_clear(), ptr_table_free(), sv_setref_uv(), and several UTF-8 |
| 3040 | interfaces to the publicised API. For the full list of the available |
| 3041 | APIs see L<perlapi>. |
| 3042 | |
| 3043 | =item * |
| 3044 | |
| 3045 | Made possible to propagate customised exceptions via croak()ing. |
| 3046 | |
| 3047 | =item * |
| 3048 | |
| 3049 | Now xsubs can have attributes just like subs. (Well, at least the |
| 3050 | built-in attributes.) |
| 3051 | |
| 3052 | =item * |
| 3053 | |
| 3054 | dTHR and djSP have been obsoleted; the former removed (because it's |
| 3055 | a no-op) and the latter replaced with dSP. |
| 3056 | |
| 3057 | =item * |
| 3058 | |
| 3059 | PERL_OBJECT has been completely removed. |
| 3060 | |
| 3061 | =item * |
| 3062 | |
| 3063 | The MAGIC constants (e.g. C<'P'>) have been macrofied |
| 3064 | (e.g. C<PERL_MAGIC_TIED>) for better source code readability |
| 3065 | and maintainability. |
| 3066 | |
| 3067 | =item * |
| 3068 | |
| 3069 | The regex compiler now maintains a structure that identifies nodes in |
| 3070 | the compiled bytecode with the corresponding syntactic features of the |
| 3071 | original regex expression. The information is attached to the new |
| 3072 | C<offsets> member of the C<struct regexp>. See L<perldebguts> for more |
| 3073 | complete information. |
| 3074 | |
| 3075 | =item * |
| 3076 | |
| 3077 | The C code has been made much more C<gcc -Wall> clean. Some warning |
| 3078 | messages still remain in some platforms, so if you are compiling with |
| 3079 | gcc you may see some warnings about dubious practices. The warnings |
| 3080 | are being worked on. |
| 3081 | |
| 3082 | =item * |
| 3083 | |
| 3084 | F<perly.c>, F<sv.c>, and F<sv.h> have now been extensively commented. |
| 3085 | |
| 3086 | =item * |
| 3087 | |
| 3088 | Documentation on how to use the Perl source repository has been added |
| 3089 | to F<Porting/repository.pod>. |
| 3090 | |
| 3091 | =item * |
| 3092 | |
| 3093 | There are now several profiling make targets. |
| 3094 | |
| 3095 | =back |
| 3096 | |
| 3097 | =head1 Security Vulnerability Closed [561] |
| 3098 | |
| 3099 | (This change was already made in 5.7.0 but bears repeating here.) |
| 3100 | (5.7.0 came out before 5.6.1: the development branch 5.7 released |
| 3101 | earlier than the maintenance branch 5.6) |
| 3102 | |
| 3103 | A potential security vulnerability in the optional suidperl component |
| 3104 | of Perl was identified in August 2000. suidperl is neither built nor |
| 3105 | installed by default. As of November 2001 the only known vulnerable |
| 3106 | platform is Linux, most likely all Linux distributions. CERT and |
| 3107 | various vendors and distributors have been alerted about the vulnerability. |
| 3108 | See http://www.cpan.org/src/5.0/sperl-2000-08-05/sperl-2000-08-05.txt |
| 3109 | for more information. |
| 3110 | |
| 3111 | The problem was caused by Perl trying to report a suspected security |
| 3112 | exploit attempt using an external program, /bin/mail. On Linux |
| 3113 | platforms the /bin/mail program had an undocumented feature which |
| 3114 | when combined with suidperl gave access to a root shell, resulting in |
| 3115 | a serious compromise instead of reporting the exploit attempt. If you |
| 3116 | don't have /bin/mail, or if you have 'safe setuid scripts', or if |
| 3117 | suidperl is not installed, you are safe. |
| 3118 | |
| 3119 | The exploit attempt reporting feature has been completely removed from |
| 3120 | Perl 5.8.0 (and the maintenance release 5.6.1, and it was removed also |
| 3121 | from all the Perl 5.7 releases), so that particular vulnerability |
| 3122 | isn't there anymore. However, further security vulnerabilities are, |
| 3123 | unfortunately, always possible. The suidperl functionality is most |
| 3124 | probably going to be removed in Perl 5.10. In any case, suidperl |
| 3125 | should only be used by security experts who know exactly what they are |
| 3126 | doing and why they are using suidperl instead of some other solution |
| 3127 | such as sudo ( see http://www.courtesan.com/sudo/ ). |
| 3128 | |
| 3129 | =head1 New Tests |
| 3130 | |
| 3131 | Several new tests have been added, especially for the F<lib> and |
| 3132 | F<ext> subsections. There are now about 69 000 individual tests |
| 3133 | (spread over about 700 test scripts), in the regression suite (5.6.1 |
| 3134 | has about 11 700 tests, in 258 test scripts) The exact numbers depend |
| 3135 | on the platform and Perl configuration used. Many of the new tests |
| 3136 | are of course introduced by the new modules, but still in general Perl |
| 3137 | is now more thoroughly tested. |
| 3138 | |
| 3139 | Because of the large number of tests, running the regression suite |
| 3140 | will take considerably longer time than it used to: expect the suite |
| 3141 | to take up to 4-5 times longer to run than in perl 5.6. On a really |
| 3142 | fast machine you can hope to finish the suite in about 6-8 minutes |
| 3143 | (wallclock time). |
| 3144 | |
| 3145 | The tests are now reported in a different order than in earlier Perls. |
| 3146 | (This happens because the test scripts from under t/lib have been moved |
| 3147 | to be closer to the library/extension they are testing.) |
| 3148 | |
| 3149 | =head1 Known Problems |
| 3150 | |
| 3151 | =head2 The Compiler Suite Is Still Very Experimental |
| 3152 | |
| 3153 | The compiler suite is slowly getting better but it continues to be |
| 3154 | highly experimental. Use in production environments is discouraged. |
| 3155 | |
| 3156 | =head2 Localising Tied Arrays and Hashes Is Broken |
| 3157 | |
| 3158 | local %tied_array; |
| 3159 | |
| 3160 | doesn't work as one would expect: the old value is restored |
| 3161 | incorrectly. This will be changed in a future release, but we don't |
| 3162 | know yet what the new semantics will exactly be. In any case, the |
| 3163 | change will break existing code that relies on the current |
| 3164 | (ill-defined) semantics, so just avoid doing this in general. |
| 3165 | |
| 3166 | =head2 Building Extensions Can Fail Because Of Largefiles |
| 3167 | |
| 3168 | Some extensions like mod_perl are known to have issues with |
| 3169 | `largefiles', a change brought by Perl 5.6.0 in which file offsets |
| 3170 | default to 64 bits wide, where supported. Modules may fail to compile |
| 3171 | at all, or they may compile and work incorrectly. Currently, there |
| 3172 | is no good solution for the problem, but Configure now provides |
| 3173 | appropriate non-largefile ccflags, ldflags, libswanted, and libs |
| 3174 | in the %Config hash (e.g., $Config{ccflags_nolargefiles}) so the |
| 3175 | extensions that are having problems can try configuring themselves |
| 3176 | without the largefileness. This is admittedly not a clean solution, |
| 3177 | and the solution may not even work at all. One potential failure is |
| 3178 | whether one can (or, if one can, whether it's a good idea to) link |
| 3179 | together at all binaries with different ideas about file offsets; |
| 3180 | all this is platform-dependent. |
| 3181 | |
| 3182 | =head2 Modifying $_ Inside for(..) |
| 3183 | |
| 3184 | for (1..5) { $_++ } |
| 3185 | |
| 3186 | works without complaint. It shouldn't. (You should be able to |
| 3187 | modify only lvalue elements inside the loops.) You can see the |
| 3188 | correct behaviour by replacing the 1..5 with 1, 2, 3, 4, 5. |
| 3189 | |
| 3190 | =head2 mod_perl 1.26 Doesn't Build With Threaded Perl |
| 3191 | |
| 3192 | Use mod_perl 1.27 or higher. |
| 3193 | |
| 3194 | =head2 lib/ftmp-security tests warn 'system possibly insecure' |
| 3195 | |
| 3196 | Don't panic. Read the 'make test' section of INSTALL instead. |
| 3197 | |
| 3198 | =head2 libwww-perl (LWP) fails base/date #51 |
| 3199 | |
| 3200 | Use libwww-perl 5.65 or later. |
| 3201 | |
| 3202 | =head2 PDL failing some tests |
| 3203 | |
| 3204 | Use PDL 2.3.4 or later. |
| 3205 | |
| 3206 | =head2 Perl_get_sv |
| 3207 | |
| 3208 | You may get errors like 'Undefined symbol "Perl_get_sv"' or "can't |
| 3209 | resolve symbol 'Perl_get_sv'", or the symbol may be "Perl_sv_2pv". |
| 3210 | This probably means that you are trying to use an older shared Perl |
| 3211 | library (or extensions linked with such) with Perl 5.8.0 executable. |
| 3212 | Perl used to have such a subroutine, but that is no more the case. |
| 3213 | Check your shared library path, and any shared Perl libraries in those |
| 3214 | directories. |
| 3215 | |
| 3216 | Sometimes this problem may also indicate a partial Perl 5.8.0 |
| 3217 | installation, see L</"Mac OS X dyld undefined symbols"> for an |
| 3218 | example and how to deal with it. |
| 3219 | |
| 3220 | =head2 Self-tying Problems |
| 3221 | |
| 3222 | Self-tying of arrays and hashes is broken in rather deep and |
| 3223 | hard-to-fix ways. As a stop-gap measure to avoid people from getting |
| 3224 | frustrated at the mysterious results (core dumps, most often), it is |
| 3225 | forbidden for now (you will get a fatal error even from an attempt). |
| 3226 | |
| 3227 | A change to self-tying of globs has caused them to be recursively |
| 3228 | referenced (see: L<perlobj/"Two-Phased Garbage Collection">). You |
| 3229 | will now need an explicit untie to destroy a self-tied glob. This |
| 3230 | behaviour may be fixed at a later date. |
| 3231 | |
| 3232 | Self-tying of scalars and IO thingies works. |
| 3233 | |
| 3234 | =head2 ext/threads/t/libc |
| 3235 | |
| 3236 | If this test fails, it indicates that your libc (C library) is not |
| 3237 | threadsafe. This particular test stress tests the localtime() call to |
| 3238 | find out whether it is threadsafe. See L<perlthrtut> for more information. |
| 3239 | |
| 3240 | =head2 Failure of Thread (5.005-style) tests |
| 3241 | |
| 3242 | B<Note that support for 5.005-style threading is deprecated, |
| 3243 | experimental and practically unsupported. In 5.10, it is expected |
| 3244 | to be removed. You should migrate your code to ithreads.> |
| 3245 | |
| 3246 | The following tests are known to fail due to fundamental problems in |
| 3247 | the 5.005 threading implementation. These are not new failures--Perl |
| 3248 | 5.005_0x has the same bugs, but didn't have these tests. |
| 3249 | |
| 3250 | ../ext/B/t/xref.t 255 65280 14 12 85.71% 3-14 |
| 3251 | ../ext/List/Util/t/first.t 255 65280 7 4 57.14% 2 5-7 |
| 3252 | ../lib/English.t 2 512 54 2 3.70% 2-3 |
| 3253 | ../lib/FileCache.t 5 1 20.00% 5 |
| 3254 | ../lib/Filter/Simple/t/data.t 6 3 50.00% 1-3 |
| 3255 | ../lib/Filter/Simple/t/filter_only. 9 3 33.33% 1-2 5 |
| 3256 | ../lib/Math/BigInt/t/bare_mbf.t 1627 4 0.25% 8 11 1626-1627 |
| 3257 | ../lib/Math/BigInt/t/bigfltpm.t 1629 4 0.25% 10 13 1628- |
| 3258 | 1629 |
| 3259 | ../lib/Math/BigInt/t/sub_mbf.t 1633 4 0.24% 8 11 1632-1633 |
| 3260 | ../lib/Math/BigInt/t/with_sub.t 1628 4 0.25% 9 12 1627-1628 |
| 3261 | ../lib/Tie/File/t/31_autodefer.t 255 65280 65 32 49.23% 34-65 |
| 3262 | ../lib/autouse.t 10 1 10.00% 4 |
| 3263 | op/flip.t 15 1 6.67% 15 |
| 3264 | |
| 3265 | These failures are unlikely to get fixed as 5.005-style threads |
| 3266 | are considered fundamentally broken. (Basically what happens is that |
| 3267 | competing threads can corrupt shared global state, one good example |
| 3268 | being regular expression engine's state.) |
| 3269 | |
| 3270 | =head2 Timing problems |
| 3271 | |
| 3272 | The following tests may fail intermittently because of timing |
| 3273 | problems, for example if the system is heavily loaded. |
| 3274 | |
| 3275 | t/op/alarm.t |
| 3276 | ext/Time/HiRes/HiRes.t |
| 3277 | lib/Benchmark.t |
| 3278 | lib/Memoize/t/expmod_t.t |
| 3279 | lib/Memoize/t/speed.t |
| 3280 | |
| 3281 | In case of failure please try running them manually, for example |
| 3282 | |
| 3283 | ./perl -Ilib ext/Time/HiRes/HiRes.t |
| 3284 | |
| 3285 | =head2 Tied/Magical Array/Hash Elements Do Not Autovivify |
| 3286 | |
| 3287 | For normal arrays C<$foo = \$bar[1]> will assign C<undef> to |
| 3288 | C<$bar[1]> (assuming that it didn't exist before), but for |
| 3289 | tied/magical arrays and hashes such autovivification does not happen |
| 3290 | because there is currently no way to catch the reference creation. |
| 3291 | The same problem affects slicing over non-existent indices/keys of |
| 3292 | a tied/magical array/hash. |
| 3293 | |
| 3294 | =head2 Unicode in package/class and subroutine names does not work |
| 3295 | |
| 3296 | One can have Unicode in identifier names, but not in package/class or |
| 3297 | subroutine names. While some limited functionality towards this does |
| 3298 | exist as of Perl 5.8.0, that is more accidental than designed; use of |
| 3299 | Unicode for the said purposes is unsupported. |
| 3300 | |
| 3301 | One reason of this unfinishedness is its (currently) inherent |
| 3302 | unportability: since both package names and subroutine names may |
| 3303 | need to be mapped to file and directory names, the Unicode capability |
| 3304 | of the filesystem becomes important-- and there unfortunately aren't |
| 3305 | portable answers. |
| 3306 | |
| 3307 | =head1 Platform Specific Problems |
| 3308 | |
| 3309 | =head2 AIX |
| 3310 | |
| 3311 | =over 4 |
| 3312 | |
| 3313 | =item * |
| 3314 | |
| 3315 | If using the AIX native make command, instead of just "make" issue |
| 3316 | "make all". In some setups the former has been known to spuriously |
| 3317 | also try to run "make install". Alternatively, you may want to use |
| 3318 | GNU make. |
| 3319 | |
| 3320 | =item * |
| 3321 | |
| 3322 | In AIX 4.2, Perl extensions that use C++ functions that use statics |
| 3323 | may have problems in that the statics are not getting initialized. |
| 3324 | In newer AIX releases, this has been solved by linking Perl with |
| 3325 | the libC_r library, but unfortunately in AIX 4.2 the said library |
| 3326 | has an obscure bug where the various functions related to time |
| 3327 | (such as time() and gettimeofday()) return broken values, and |
| 3328 | therefore in AIX 4.2 Perl is not linked against libC_r. |
| 3329 | |
| 3330 | =item * |
| 3331 | |
| 3332 | vac 5.0.0.0 May Produce Buggy Code For Perl |
| 3333 | |
| 3334 | The AIX C compiler vac version 5.0.0.0 may produce buggy code, |
| 3335 | resulting in a few random tests failing when run as part of "make |
| 3336 | test", but when the failing tests are run by hand, they succeed. |
| 3337 | We suggest upgrading to at least vac version 5.0.1.0, that has been |
| 3338 | known to compile Perl correctly. "lslpp -L|grep vac.C" will tell |
| 3339 | you the vac version. See README.aix. |
| 3340 | |
| 3341 | =item * |
| 3342 | |
| 3343 | If building threaded Perl, you may get compilation warning from pp_sys.c: |
| 3344 | |
| 3345 | "pp_sys.c", line 4651.39: 1506-280 (W) Function argument assignment between types "unsigned char*" and "const void*" is not allowed. |
| 3346 | |
| 3347 | This is harmless; it is caused by the getnetbyaddr() and getnetbyaddr_r() |
| 3348 | having slightly different types for their first argument. |
| 3349 | |
| 3350 | =back |
| 3351 | |
| 3352 | =head2 Alpha systems with old gccs fail several tests |
| 3353 | |
| 3354 | If you see op/pack, op/pat, op/regexp, or ext/Storable tests failing |
| 3355 | in a Linux/alpha or *BSD/Alpha, it's probably time to upgrade your gcc. |
| 3356 | gccs prior to 2.95.3 are definitely not good enough, and gcc 3.1 may |
| 3357 | be even better. (RedHat Linux/alpha with gcc 3.1 reported no problems, |
| 3358 | as did Linux 2.4.18 with gcc 2.95.4.) (In Tru64, it is preferable to |
| 3359 | use the bundled C compiler.) |
| 3360 | |
| 3361 | =head2 AmigaOS |
| 3362 | |
| 3363 | Perl 5.8.0 doesn't build in AmigaOS. It broke at some point during |
| 3364 | the ithreads work and we could not find Amiga experts to unbreak the |
| 3365 | problems. Perl 5.6.1 still works for AmigaOS (as does the the 5.7.2 |
| 3366 | development release). |
| 3367 | |
| 3368 | =head2 BeOS |
| 3369 | |
| 3370 | The following tests fail on 5.8.0 Perl in BeOS Personal 5.03: |
| 3371 | |
| 3372 | t/op/lfs............................FAILED at test 17 |
| 3373 | t/op/magic..........................FAILED at test 24 |
| 3374 | ext/Fcntl/t/syslfs..................FAILED at test 17 |
| 3375 | ext/File/Glob/t/basic...............FAILED at test 3 |
| 3376 | ext/POSIX/t/sigaction...............FAILED at test 13 |
| 3377 | ext/POSIX/t/waitpid.................FAILED at test 1 |
| 3378 | |
| 3379 | See L<perlbeos> (README.beos) for more details. |
| 3380 | |
| 3381 | =head2 Cygwin "unable to remap" |
| 3382 | |
| 3383 | For example when building the Tk extension for Cygwin, |
| 3384 | you may get an error message saying "unable to remap". |
| 3385 | This is known problem with Cygwin, and a workaround is |
| 3386 | detailed in here: http://sources.redhat.com/ml/cygwin/2001-12/msg00894.html |
| 3387 | |
| 3388 | =head2 Cygwin ndbm tests fail on FAT |
| 3389 | |
| 3390 | One can build but not install (or test the build of) the NDBM_File |
| 3391 | on FAT filesystems. Installation (or build) on NTFS works fine. |
| 3392 | If one attempts the test on a FAT install (or build) the following |
| 3393 | failures are expected: |
| 3394 | |
| 3395 | ../ext/NDBM_File/ndbm.t 13 3328 71 59 83.10% 1-2 4 16-71 |
| 3396 | ../ext/ODBM_File/odbm.t 255 65280 ?? ?? % ?? |
| 3397 | ../lib/AnyDBM_File.t 2 512 12 2 16.67% 1 4 |
| 3398 | ../lib/Memoize/t/errors.t 0 139 11 5 45.45% 7-11 |
| 3399 | ../lib/Memoize/t/tie_ndbm.t 13 3328 4 4 100.00% 1-4 |
| 3400 | run/fresh_perl.t 97 1 1.03% 91 |
| 3401 | |
| 3402 | NDBM_File fails and ODBM_File just coredumps. |
| 3403 | |
| 3404 | If you intend to run only on FAT (or if using AnyDBM_File on FAT), |
| 3405 | run Configure with the -Ui_ndbm and -Ui_dbm options to prevent |
| 3406 | NDBM_File and ODBM_File being built. |
| 3407 | |
| 3408 | =head2 DJGPP Failures |
| 3409 | |
| 3410 | t/op/stat............................FAILED at test 29 |
| 3411 | lib/File/Find/t/find.................FAILED at test 1 |
| 3412 | lib/File/Find/t/taint................FAILED at test 1 |
| 3413 | lib/h2xs.............................FAILED at test 15 |
| 3414 | lib/Pod/t/eol........................FAILED at test 1 |
| 3415 | lib/Test/Harness/t/strap-analyze.....FAILED at test 8 |
| 3416 | lib/Test/Harness/t/test-harness......FAILED at test 23 |
| 3417 | lib/Test/Simple/t/exit...............FAILED at test 1 |
| 3418 | |
| 3419 | The above failures are known as of 5.8.0 with native builds with long |
| 3420 | filenames, but there are a few more if running under dosemu because of |
| 3421 | limitations (and maybe bugs) of dosemu: |
| 3422 | |
| 3423 | t/comp/cpp...........................FAILED at test 3 |
| 3424 | t/op/inccode.........................(crash) |
| 3425 | |
| 3426 | and a few lib/ExtUtils tests, and several hundred Encode/t/Aliases.t |
| 3427 | failures that work fine with long filenames. So you really might |
| 3428 | prefer native builds and long filenames. |
| 3429 | |
| 3430 | =head2 FreeBSD built with ithreads coredumps reading large directories |
| 3431 | |
| 3432 | This is a known bug in FreeBSD 4.5's readdir_r(), it has been fixed in |
| 3433 | FreeBSD 4.6 (see L<perlfreebsd> (README.freebsd)). |
| 3434 | |
| 3435 | =head2 FreeBSD Failing locale Test 117 For ISO 8859-15 Locales |
| 3436 | |
| 3437 | The ISO 8859-15 locales may fail the locale test 117 in FreeBSD. |
| 3438 | This is caused by the characters \xFF (y with diaeresis) and \xBE |
| 3439 | (Y with diaeresis) not behaving correctly when being matched |
| 3440 | case-insensitively. Apparently this problem has been fixed in |
| 3441 | the latest FreeBSD releases. |
| 3442 | ( http://www.freebsd.org/cgi/query-pr.cgi?pr=34308 ) |
| 3443 | |
| 3444 | =head2 IRIX fails ext/List/Util/t/shuffle.t or Digest::MD5 |
| 3445 | |
| 3446 | IRIX with MIPSpro 7.3.1.2m or 7.3.1.3m compiler may fail the List::Util |
| 3447 | test ext/List/Util/t/shuffle.t by dumping core. This seems to be |
| 3448 | a compiler error since if compiled with gcc no core dump ensues, and |
| 3449 | no failures have been seen on the said test on any other platform. |
| 3450 | |
| 3451 | Similarly, building the Digest::MD5 extension has been |
| 3452 | known to fail with "*** Termination code 139 (bu21)". |
| 3453 | |
| 3454 | The cure is to drop optimization level (Configure -Doptimize=-O2). |
| 3455 | |
| 3456 | =head2 HP-UX lib/posix Subtest 9 Fails When LP64-Configured |
| 3457 | |
| 3458 | If perl is configured with -Duse64bitall, the successful result of the |
| 3459 | subtest 10 of lib/posix may arrive before the successful result of the |
| 3460 | subtest 9, which confuses the test harness so much that it thinks the |
| 3461 | subtest 9 failed. |
| 3462 | |
| 3463 | =head2 Linux with glibc 2.2.5 fails t/op/int subtest #6 with -Duse64bitint |
| 3464 | |
| 3465 | This is a known bug in the glibc 2.2.5 with long long integers. |
| 3466 | ( http://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=65612 ) |
| 3467 | |
| 3468 | =head2 Linux With Sfio Fails op/misc Test 48 |
| 3469 | |
| 3470 | No known fix. |
| 3471 | |
| 3472 | =head2 Mac OS X |
| 3473 | |
| 3474 | Please remember to set your environment variable LC_ALL to "C" |
| 3475 | (setenv LC_ALL C) before running "make test" to avoid a lot of |
| 3476 | warnings about the broken locales of Mac OS X. |
| 3477 | |
| 3478 | The following tests are known to fail in Mac OS X 10.1.5 because of |
| 3479 | buggy (old) implementations of Berkeley DB included in Mac OS X: |
| 3480 | |
| 3481 | Failed Test Stat Wstat Total Fail Failed List of Failed |
| 3482 | ------------------------------------------------------------------------- |
| 3483 | ../ext/DB_File/t/db-btree.t 0 11 ?? ?? % ?? |
| 3484 | ../ext/DB_File/t/db-recno.t 149 3 2.01% 61 63 65 |
| 3485 | |
| 3486 | If you are building on a UFS partition, you will also probably see |
| 3487 | t/op/stat.t subtest #9 fail. This is caused by Darwin's UFS not |
| 3488 | supporting inode change time. |
| 3489 | |
| 3490 | Also the ext/POSIX/t/posix.t subtest #10 fails but it is skipped for |
| 3491 | now because the failure is Apple's fault, not Perl's (blocked signals |
| 3492 | are lost). |
| 3493 | |
| 3494 | If you Configure with ithreads, ext/threads/t/libc.t will fail. Again, |
| 3495 | this is not Perl's fault-- the libc of Mac OS X is not threadsafe |
| 3496 | (in this particular test, the localtime() call is found to be |
| 3497 | threadunsafe.) |
| 3498 | |
| 3499 | =head2 Mac OS X dyld undefined symbols |
| 3500 | |
| 3501 | If after installing Perl 5.8.0 you are getting warnings about missing |
| 3502 | symbols, for example |
| 3503 | |
| 3504 | dyld: perl Undefined symbols |
| 3505 | _perl_sv_2pv |
| 3506 | _perl_get_sv |
| 3507 | |
| 3508 | you probably have an old pre-Perl-5.8.0 installation (or parts of one) |
| 3509 | in /Library/Perl (the undefined symbols used to exist in pre-5.8.0 Perls). |
| 3510 | It seems that for some reason "make install" doesn't always completely |
| 3511 | overwrite the files in /Library/Perl. You can move the old Perl |
| 3512 | shared library out of the way like this: |
| 3513 | |
| 3514 | cd /Library/Perl/darwin/CORE |
| 3515 | mv libperl.dylib libperlold.dylib |
| 3516 | |
| 3517 | and then reissue "make install". Note that the above of course is |
| 3518 | extremely disruptive for anything using the /usr/local/bin/perl. |
| 3519 | If that doesn't help, you may have to try removing all the .bundle |
| 3520 | files from beneath /Library/Perl, and again "make install"-ing. |
| 3521 | |
| 3522 | =head2 OS/2 Test Failures |
| 3523 | |
| 3524 | The following tests are known to fail on OS/2 (for clarity |
| 3525 | only the failures are shown, not the full error messages): |
| 3526 | |
| 3527 | ../lib/ExtUtils/t/Mkbootstrap.t 1 256 18 1 5.56% 8 |
| 3528 | ../lib/ExtUtils/t/Packlist.t 1 256 34 1 2.94% 17 |
| 3529 | ../lib/ExtUtils/t/basic.t 1 256 17 1 5.88% 14 |
| 3530 | lib/os2_process.t 2 512 227 2 0.88% 174 209 |
| 3531 | lib/os2_process_kid.t 227 2 0.88% 174 209 |
| 3532 | lib/rx_cmprt.t 255 65280 18 3 16.67% 16-18 |
| 3533 | |
| 3534 | =head2 op/sprintf tests 91, 129, and 130 |
| 3535 | |
| 3536 | The op/sprintf tests 91, 129, and 130 are known to fail on some platforms. |
| 3537 | Examples include any platform using sfio, and Compaq/Tandem's NonStop-UX. |
| 3538 | |
| 3539 | Test 91 is known to fail on QNX6 (nto), because C<sprintf '%e',0> |
| 3540 | incorrectly produces C<0.000000e+0> instead of C<0.000000e+00>. |
| 3541 | |
| 3542 | For tests 129 and 130, the failing platforms do not comply with |
| 3543 | the ANSI C Standard: lines 19ff on page 134 of ANSI X3.159 1989, to |
| 3544 | be exact. (They produce something other than "1" and "-1" when |
| 3545 | formatting 0.6 and -0.6 using the printf format "%.0f"; most often, |
| 3546 | they produce "0" and "-0".) |
| 3547 | |
| 3548 | =head2 SCO |
| 3549 | |
| 3550 | The socketpair tests are known to be unhappy in SCO 3.2v5.0.4: |
| 3551 | |
| 3552 | ext/Socket/socketpair.t...............FAILED tests 15-45 |
| 3553 | |
| 3554 | =head2 Solaris 2.5 |
| 3555 | |
| 3556 | In case you are still using Solaris 2.5 (aka SunOS 5.5), you may |
| 3557 | experience failures (the test core dumping) in lib/locale.t. |
| 3558 | The suggested cure is to upgrade your Solaris. |
| 3559 | |
| 3560 | =head2 Solaris x86 Fails Tests With -Duse64bitint |
| 3561 | |
| 3562 | The following tests are known to fail in Solaris x86 with Perl |
| 3563 | configured to use 64 bit integers: |
| 3564 | |
| 3565 | ext/Data/Dumper/t/dumper.............FAILED at test 268 |
| 3566 | ext/Devel/Peek/Peek..................FAILED at test 7 |
| 3567 | |
| 3568 | =head2 SUPER-UX (NEC SX) |
| 3569 | |
| 3570 | The following tests are known to fail on SUPER-UX: |
| 3571 | |
| 3572 | op/64bitint...........................FAILED tests 29-30, 32-33, 35-36 |
| 3573 | op/arith..............................FAILED tests 128-130 |
| 3574 | op/pack...............................FAILED tests 25-5625 |
| 3575 | op/pow................................ |
| 3576 | op/taint..............................# msgsnd failed |
| 3577 | ../ext/IO/lib/IO/t/io_poll............FAILED tests 3-4 |
| 3578 | ../ext/IPC/SysV/ipcsysv...............FAILED tests 2, 5-6 |
| 3579 | ../ext/IPC/SysV/t/msg.................FAILED tests 2, 4-6 |
| 3580 | ../ext/Socket/socketpair..............FAILED tests 12 |
| 3581 | ../lib/IPC/SysV.......................FAILED tests 2, 5-6 |
| 3582 | ../lib/warnings.......................FAILED tests 115-116, 118-119 |
| 3583 | |
| 3584 | The op/pack failure ("Cannot compress negative numbers at op/pack.t line 126") |
| 3585 | is serious but as of yet unsolved. It points at some problems with the |
| 3586 | signedness handling of the C compiler, as do the 64bitint, arith, and pow |
| 3587 | failures. Most of the rest point at problems with SysV IPC. |
| 3588 | |
| 3589 | =head2 Term::ReadKey not working on Win32 |
| 3590 | |
| 3591 | Use Term::ReadKey 2.20 or later. |
| 3592 | |
| 3593 | =head2 UNICOS/mk |
| 3594 | |
| 3595 | =over 4 |
| 3596 | |
| 3597 | =item * |
| 3598 | |
| 3599 | During Configure, the test |
| 3600 | |
| 3601 | Guessing which symbols your C compiler and preprocessor define... |
| 3602 | |
| 3603 | will probably fail with error messages like |
| 3604 | |
| 3605 | CC-20 cc: ERROR File = try.c, Line = 3 |
| 3606 | The identifier "bad" is undefined. |
| 3607 | |
| 3608 | bad switch yylook 79bad switch yylook 79bad switch yylook 79bad switch yylook 79#ifdef A29K |
| 3609 | ^ |
| 3610 | |
| 3611 | CC-65 cc: ERROR File = try.c, Line = 3 |
| 3612 | A semicolon is expected at this point. |
| 3613 | |
| 3614 | This is caused by a bug in the awk utility of UNICOS/mk. You can ignore |
| 3615 | the error, but it does cause a slight problem: you cannot fully |
| 3616 | benefit from the h2ph utility (see L<h2ph>) that can be used to |
| 3617 | convert C headers to Perl libraries, mainly used to be able to access |
| 3618 | from Perl the constants defined using C preprocessor, cpp. Because of |
| 3619 | the above error, parts of the converted headers will be invisible. |
| 3620 | Luckily, these days the need for h2ph is rare. |
| 3621 | |
| 3622 | =item * |
| 3623 | |
| 3624 | If building Perl with interpreter threads (ithreads), the |
| 3625 | getgrent(), getgrnam(), and getgrgid() functions cannot return the |
| 3626 | list of the group members due to a bug in the multithreaded support of |
| 3627 | UNICOS/mk. What this means is that in list context the functions will |
| 3628 | return only three values, not four. |
| 3629 | |
| 3630 | =back |
| 3631 | |
| 3632 | =head2 UTS |
| 3633 | |
| 3634 | There are a few known test failures, see L<perluts> (README.uts). |
| 3635 | |
| 3636 | =head2 VOS (Stratus) |
| 3637 | |
| 3638 | When Perl is built using the native build process on VOS Release |
| 3639 | 14.5.0 and GNU C++/GNU Tools 2.0.1, all attempted tests either |
| 3640 | pass or result in TODO (ignored) failures. |
| 3641 | |
| 3642 | =head2 VMS |
| 3643 | |
| 3644 | There should be no reported test failures with a default configuration, |
| 3645 | though there are a number of tests marked TODO that point to areas |
| 3646 | needing further debugging and/or porting work. |
| 3647 | |
| 3648 | =head2 Win32 |
| 3649 | |
| 3650 | In multi-CPU boxes, there are some problems with the I/O buffering: |
| 3651 | some output may appear twice. |
| 3652 | |
| 3653 | =head2 XML::Parser not working |
| 3654 | |
| 3655 | Use XML::Parser 2.31 or later. |
| 3656 | |
| 3657 | =head2 z/OS (OS/390) |
| 3658 | |
| 3659 | z/OS has rather many test failures but the situation is actually much |
| 3660 | better than it was in 5.6.0; it's just that so many new modules and |
| 3661 | tests have been added. |
| 3662 | |
| 3663 | Failed Test Stat Wstat Total Fail Failed List of Failed |
| 3664 | --------------------------------------------------------------------------- |
| 3665 | ../ext/Data/Dumper/t/dumper.t 357 8 2.24% 311 314 325 327 |
| 3666 | 331 333 337 339 |
| 3667 | ../ext/IO/lib/IO/t/io_unix.t 5 4 80.00% 2-5 |
| 3668 | ../ext/Storable/t/downgrade.t 12 3072 169 12 7.10% 14-15 46-47 78-79 |
| 3669 | 110-111 150 161 |
| 3670 | ../lib/ExtUtils/t/Constant.t 121 30976 48 48 100.00% 1-48 |
| 3671 | ../lib/ExtUtils/t/Embed.t 9 9 100.00% 1-9 |
| 3672 | op/pat.t 922 7 0.76% 665 776 785 832- |
| 3673 | 834 845 |
| 3674 | op/sprintf.t 224 3 1.34% 98 100 136 |
| 3675 | op/tr.t 97 5 5.15% 63 71-74 |
| 3676 | uni/fold.t 780 6 0.77% 61 169 196 661 |
| 3677 | 710-711 |
| 3678 | |
| 3679 | The failures in dumper.t and downgrade.t are problems in the tests, |
| 3680 | those in io_unix and sprintf are problems in the USS (UDP sockets and |
| 3681 | printf formats). The pat, tr, and fold failures are genuine Perl |
| 3682 | problems caused by EBCDIC (and in the pat and fold cases, combining |
| 3683 | that with Unicode). The Constant and Embed are probably problems in |
| 3684 | the tests (since they test Perl's ability to build extensions, and |
| 3685 | that seems to be working reasonably well.) |
| 3686 | |
| 3687 | =head2 Unicode Support on EBCDIC Still Spotty |
| 3688 | |
| 3689 | Though mostly working, Unicode support still has problem spots on |
| 3690 | EBCDIC platforms. One such known spot are the C<\p{}> and C<\P{}> |
| 3691 | regular expression constructs for code points less than 256: the |
| 3692 | C<pP> are testing for Unicode code points, not knowing about EBCDIC. |
| 3693 | |
| 3694 | =head2 Seen In Perl 5.7 But Gone Now |
| 3695 | |
| 3696 | C<Time::Piece> (previously known as C<Time::Object>) was removed |
| 3697 | because it was felt that it didn't have enough value in it to be a |
| 3698 | core module. It is still a useful module, though, and is available |
| 3699 | from the CPAN. |
| 3700 | |
| 3701 | Perl 5.8 unfortunately does not build anymore on AmigaOS; this broke |
| 3702 | accidentally at some point. Since there are not that many Amiga |
| 3703 | developers available, we could not get this fixed and tested in time |
| 3704 | for 5.8.0. Perl 5.6.1 still works for AmigaOS (as does the the 5.7.2 |
| 3705 | development release). |
| 3706 | |
| 3707 | The C<PerlIO::Scalar> and C<PerlIO::Via> (capitalised) were renamed as |
| 3708 | C<PerlIO::scalar> and C<PerlIO::via> (all lowercase) just before 5.8.0. |
| 3709 | The main rationale was to have all core PerlIO layers to have all |
| 3710 | lowercase names. The "plugins" are named as usual, for example |
| 3711 | C<PerlIO::via::QuotedPrint>. |
| 3712 | |
| 3713 | The C<threads::shared::queue> and C<threads::shared::semaphore> were |
| 3714 | renamed as C<Thread::Queue> and C<Thread::Semaphore> just before 5.8.0. |
| 3715 | The main rationale was to have thread modules to obey normal naming, |
| 3716 | C<Thread::> (the C<threads> and C<threads::shared> themselves are |
| 3717 | more pragma-like, they affect compile-time, so they stay lowercase). |
| 3718 | |
| 3719 | =head1 Reporting Bugs |
| 3720 | |
| 3721 | If you find what you think is a bug, you might check the articles |
| 3722 | recently posted to the comp.lang.perl.misc newsgroup and the perl |
| 3723 | bug database at http://bugs.perl.org/ . There may also be |
| 3724 | information at http://www.perl.com/ , the Perl Home Page. |
| 3725 | |
| 3726 | If you believe you have an unreported bug, please run the B<perlbug> |
| 3727 | program included with your release. Be sure to trim your bug down |
| 3728 | to a tiny but sufficient test case. Your bug report, along with the |
| 3729 | output of C<perl -V>, will be sent off to perlbug@perl.org to be |
| 3730 | analysed by the Perl porting team. |
| 3731 | |
| 3732 | =head1 SEE ALSO |
| 3733 | |
| 3734 | The F<Changes> file for exhaustive details on what changed. |
| 3735 | |
| 3736 | The F<INSTALL> file for how to build Perl. |
| 3737 | |
| 3738 | The F<README> file for general stuff. |
| 3739 | |
| 3740 | The F<Artistic> and F<Copying> files for copyright information. |
| 3741 | |
| 3742 | =head1 HISTORY |
| 3743 | |
| 3744 | Written by Jarkko Hietaniemi <F<jhi@iki.fi>>. |
| 3745 | |
| 3746 | =cut |