| 1 |
.\" -*- mode: troff; coding: utf-8 -*- |
| 2 |
.\" Automatically generated by Pod::Man 5.0102 (Pod::Simple 3.45) |
| 3 |
.\" |
| 4 |
.\" Standard preamble: |
| 5 |
.\" ======================================================================== |
| 6 |
.de Sp \" Vertical space (when we can't use .PP) |
| 7 |
.if t .sp .5v |
| 8 |
.if n .sp |
| 9 |
.. |
| 10 |
.de Vb \" Begin verbatim text |
| 11 |
.ft CW |
| 12 |
.nf |
| 13 |
.ne \\$1 |
| 14 |
.. |
| 15 |
.de Ve \" End verbatim text |
| 16 |
.ft R |
| 17 |
.fi |
| 18 |
.. |
| 19 |
.\" \*(C` and \*(C' are quotes in nroff, nothing in troff, for use with C<>. |
| 20 |
.ie n \{\ |
| 21 |
. ds C` "" |
| 22 |
. ds C' "" |
| 23 |
'br\} |
| 24 |
.el\{\ |
| 25 |
. ds C` |
| 26 |
. ds C' |
| 27 |
'br\} |
| 28 |
.\" |
| 29 |
.\" Escape single quotes in literal strings from groff's Unicode transform. |
| 30 |
.ie \n(.g .ds Aq \(aq |
| 31 |
.el .ds Aq ' |
| 32 |
.\" |
| 33 |
.\" If the F register is >0, we'll generate index entries on stderr for |
| 34 |
.\" titles (.TH), headers (.SH), subsections (.SS), items (.Ip), and index |
| 35 |
.\" entries marked with X<> in POD. Of course, you'll have to process the |
| 36 |
.\" output yourself in some meaningful fashion. |
| 37 |
.\" |
| 38 |
.\" Avoid warning from groff about undefined register 'F'. |
| 39 |
.de IX |
| 40 |
.. |
| 41 |
.nr rF 0 |
| 42 |
.if \n(.g .if rF .nr rF 1 |
| 43 |
.if (\n(rF:(\n(.g==0)) \{\ |
| 44 |
. if \nF \{\ |
| 45 |
. de IX |
| 46 |
. tm Index:\\$1\t\\n%\t"\\$2" |
| 47 |
.. |
| 48 |
. if !\nF==2 \{\ |
| 49 |
. nr % 0 |
| 50 |
. nr F 2 |
| 51 |
. \} |
| 52 |
. \} |
| 53 |
.\} |
| 54 |
.rr rF |
| 55 |
.\" ======================================================================== |
| 56 |
.\" |
| 57 |
.IX Title "PERL581DELTA 1" |
| 58 |
.TH PERL581DELTA 1 2025-11-16 "perl v5.40.1" "Perl Programmers Reference Guide" |
| 59 |
.\" For nroff, turn off justification. Always turn off hyphenation; it makes |
| 60 |
.\" way too many mistakes in technical documents. |
| 61 |
.if n .ad l |
| 62 |
.nh |
| 63 |
.SH NAME |
| 64 |
perl581delta \- what is new for perl v5.8.1 |
| 65 |
.SH DESCRIPTION |
| 66 |
.IX Header "DESCRIPTION" |
| 67 |
This document describes differences between the 5.8.0 release and |
| 68 |
the 5.8.1 release. |
| 69 |
.PP |
| 70 |
If you are upgrading from an earlier release such as 5.6.1, first read |
| 71 |
the perl58delta, which describes differences between 5.6.0 and |
| 72 |
5.8.0. |
| 73 |
.PP |
| 74 |
In case you are wondering about 5.6.1, it was bug-fix-wise rather |
| 75 |
identical to the development release 5.7.1. Confused? This timeline |
| 76 |
hopefully helps a bit: it lists the new major releases, their maintenance |
| 77 |
releases, and the development releases. |
| 78 |
.PP |
| 79 |
.Vb 1 |
| 80 |
\& New Maintenance Development |
| 81 |
\& |
| 82 |
\& 5.6.0 2000\-Mar\-22 |
| 83 |
\& 5.7.0 2000\-Sep\-02 |
| 84 |
\& 5.6.1 2001\-Apr\-08 |
| 85 |
\& 5.7.1 2001\-Apr\-09 |
| 86 |
\& 5.7.2 2001\-Jul\-13 |
| 87 |
\& 5.7.3 2002\-Mar\-05 |
| 88 |
\& 5.8.0 2002\-Jul\-18 |
| 89 |
\& 5.8.1 2003\-Sep\-25 |
| 90 |
.Ve |
| 91 |
.SH "Incompatible Changes" |
| 92 |
.IX Header "Incompatible Changes" |
| 93 |
.SS "Hash Randomisation" |
| 94 |
.IX Subsection "Hash Randomisation" |
| 95 |
Mainly due to security reasons, the "random ordering" of hashes |
| 96 |
has been made even more random. Previously while the order of hash |
| 97 |
elements from \fBkeys()\fR, \fBvalues()\fR, and \fBeach()\fR was essentially random, |
| 98 |
it was still repeatable. Now, however, the order varies between |
| 99 |
different runs of Perl. |
| 100 |
.PP |
| 101 |
\&\fBPerl has never guaranteed any ordering of the hash keys\fR, and the |
| 102 |
ordering has already changed several times during the lifetime of |
| 103 |
Perl 5. Also, the ordering of hash keys has always been, and |
| 104 |
continues to be, affected by the insertion order. |
| 105 |
.PP |
| 106 |
The added randomness may affect applications. |
| 107 |
.PP |
| 108 |
One possible scenario is when output of an application has included |
| 109 |
hash data. For example, if you have used the Data::Dumper module to |
| 110 |
dump data into different files, and then compared the files to see |
| 111 |
whether the data has changed, now you will have false positives since |
| 112 |
the order in which hashes are dumped will vary. In general the cure |
| 113 |
is to sort the keys (or the values); in particular for Data::Dumper to |
| 114 |
use the \f(CW\*(C`Sortkeys\*(C'\fR option. If some particular order is really |
| 115 |
important, use tied hashes: for example the Tie::IxHash module |
| 116 |
which by default preserves the order in which the hash elements |
| 117 |
were added. |
| 118 |
.PP |
| 119 |
More subtle problem is reliance on the order of "global destruction". |
| 120 |
That is what happens at the end of execution: Perl destroys all data |
| 121 |
structures, including user data. If your destructors (the DESTROY |
| 122 |
subroutines) have assumed any particular ordering to the global |
| 123 |
destruction, there might be problems ahead. For example, in a |
| 124 |
destructor of one object you cannot assume that objects of any other |
| 125 |
class are still available, unless you hold a reference to them. |
| 126 |
If the environment variable PERL_DESTRUCT_LEVEL is set to a non-zero |
| 127 |
value, or if Perl is exiting a spawned thread, it will also destruct |
| 128 |
the ordinary references and the symbol tables that are no longer in use. |
| 129 |
You can't call a class method or an ordinary function on a class that |
| 130 |
has been collected that way. |
| 131 |
.PP |
| 132 |
The hash randomisation is certain to reveal hidden assumptions about |
| 133 |
some particular ordering of hash elements, and outright bugs: it |
| 134 |
revealed a few bugs in the Perl core and core modules. |
| 135 |
.PP |
| 136 |
To disable the hash randomisation in runtime, set the environment |
| 137 |
variable PERL_HASH_SEED to 0 (zero) before running Perl (for more |
| 138 |
information see "PERL_HASH_SEED" in perlrun), or to disable the feature |
| 139 |
completely in compile time, compile with \f(CW\*(C`\-DNO_HASH_SEED\*(C'\fR (see \fIINSTALL\fR). |
| 140 |
.PP |
| 141 |
See "Algorithmic Complexity Attacks" in perlsec for the original |
| 142 |
rationale behind this change. |
| 143 |
.SS "UTF\-8 On Filehandles No Longer Activated By Locale" |
| 144 |
.IX Subsection "UTF-8 On Filehandles No Longer Activated By Locale" |
| 145 |
In Perl 5.8.0 all filehandles, including the standard filehandles, |
| 146 |
were implicitly set to be in Unicode UTF\-8 if the locale settings |
| 147 |
indicated the use of UTF\-8. This feature caused too many problems, |
| 148 |
so the feature was turned off and redesigned: see "Core Enhancements". |
| 149 |
.SS "Single-number v\-strings are no longer v\-strings before ""=>""" |
| 150 |
.IX Subsection "Single-number v-strings are no longer v-strings before ""=>""" |
| 151 |
The version strings or v\-strings (see "Version Strings" in perldata) |
| 152 |
feature introduced in Perl 5.6.0 has been a source of some confusion\-\- |
| 153 |
especially when the user did not want to use it, but Perl thought it |
| 154 |
knew better. Especially troublesome has been the feature that before |
| 155 |
a "=>" a version string (a "v" followed by digits) has been interpreted |
| 156 |
as a v\-string instead of a string literal. In other words: |
| 157 |
.PP |
| 158 |
.Vb 1 |
| 159 |
\& %h = ( v65 => 42 ); |
| 160 |
.Ve |
| 161 |
.PP |
| 162 |
has meant since Perl 5.6.0 |
| 163 |
.PP |
| 164 |
.Vb 1 |
| 165 |
\& %h = ( \*(AqA\*(Aq => 42 ); |
| 166 |
.Ve |
| 167 |
.PP |
| 168 |
(at least in platforms of ASCII progeny) Perl 5.8.1 restores the |
| 169 |
more natural interpretation |
| 170 |
.PP |
| 171 |
.Vb 1 |
| 172 |
\& %h = ( \*(Aqv65\*(Aq => 42 ); |
| 173 |
.Ve |
| 174 |
.PP |
| 175 |
The multi-number v\-strings like v65.66 and 65.66.67 still continue to |
| 176 |
be v\-strings in Perl 5.8. |
| 177 |
.SS "(Win32) The \-C Switch Has Been Repurposed" |
| 178 |
.IX Subsection "(Win32) The -C Switch Has Been Repurposed" |
| 179 |
The \-C switch has changed in an incompatible way. The old semantics |
| 180 |
of this switch only made sense in Win32 and only in the "use utf8" |
| 181 |
universe in 5.6.x releases, and do not make sense for the Unicode |
| 182 |
implementation in 5.8.0. Since this switch could not have been used |
| 183 |
by anyone, it has been repurposed. The behavior that this switch |
| 184 |
enabled in 5.6.x releases may be supported in a transparent, |
| 185 |
data-dependent fashion in a future release. |
| 186 |
.PP |
| 187 |
For the new life of this switch, see "UTF\-8 no longer default under |
| 188 |
UTF\-8 locales", and "\-C" in perlrun. |
| 189 |
.SS "(Win32) The /d Switch Of cmd.exe" |
| 190 |
.IX Subsection "(Win32) The /d Switch Of cmd.exe" |
| 191 |
Perl 5.8.1 uses the /d switch when running the cmd.exe shell |
| 192 |
internally for \fBsystem()\fR, backticks, and when opening pipes to external |
| 193 |
programs. The extra switch disables the execution of AutoRun commands |
| 194 |
from the registry, which is generally considered undesirable when |
| 195 |
running external programs. If you wish to retain compatibility with |
| 196 |
the older behavior, set PERL5SHELL in your environment to \f(CW\*(C`cmd /x/c\*(C'\fR. |
| 197 |
.SH "Core Enhancements" |
| 198 |
.IX Header "Core Enhancements" |
| 199 |
.SS "UTF\-8 no longer default under UTF\-8 locales" |
| 200 |
.IX Subsection "UTF-8 no longer default under UTF-8 locales" |
| 201 |
In Perl 5.8.0 many Unicode features were introduced. One of them |
| 202 |
was found to be of more nuisance than benefit: the automagic |
| 203 |
(and silent) "UTF\-8\-ification" of filehandles, including the |
| 204 |
standard filehandles, if the user's locale settings indicated |
| 205 |
use of UTF\-8. |
| 206 |
.PP |
| 207 |
For example, if you had \f(CW\*(C`en_US.UTF\-8\*(C'\fR as your locale, your STDIN and |
| 208 |
STDOUT were automatically "UTF\-8", in other words an implicit |
| 209 |
binmode(..., ":utf8") was made. This meant that trying to print, say, |
| 210 |
\&\fBchr\fR\|(0xff), ended up printing the bytes 0xc3 0xbf. Hardly what |
| 211 |
you had in mind unless you were aware of this feature of Perl 5.8.0. |
| 212 |
The problem is that the vast majority of people weren't: for example |
| 213 |
in RedHat releases 8 and 9 the \fBdefault\fR locale setting is UTF\-8, so |
| 214 |
all RedHat users got UTF\-8 filehandles, whether they wanted it or not. |
| 215 |
The pain was intensified by the Unicode implementation of Perl 5.8.0 |
| 216 |
(still) having nasty bugs, especially related to the use of s/// and |
| 217 |
tr///. (Bugs that have been fixed in 5.8.1) |
| 218 |
.PP |
| 219 |
Therefore a decision was made to backtrack the feature and change it |
| 220 |
from implicit silent default to explicit conscious option. The new |
| 221 |
Perl command line option \f(CW\*(C`\-C\*(C'\fR and its counterpart environment |
| 222 |
variable PERL_UNICODE can now be used to control how Perl and Unicode |
| 223 |
interact at interfaces like I/O and for example the command line |
| 224 |
arguments. See "\-C" in perlrun and "PERL_UNICODE" in perlrun for more |
| 225 |
information. |
| 226 |
.SS "Unsafe signals again available" |
| 227 |
.IX Subsection "Unsafe signals again available" |
| 228 |
In Perl 5.8.0 the so-called "safe signals" were introduced. This |
| 229 |
means that Perl no longer handles signals immediately but instead |
| 230 |
"between opcodes", when it is safe to do so. The earlier immediate |
| 231 |
handling easily could corrupt the internal state of Perl, resulting |
| 232 |
in mysterious crashes. |
| 233 |
.PP |
| 234 |
However, the new safer model has its problems too. Because now an |
| 235 |
opcode, a basic unit of Perl execution, is never interrupted but |
| 236 |
instead let to run to completion, certain operations that can take a |
| 237 |
long time now really do take a long time. For example, certain |
| 238 |
network operations have their own blocking and timeout mechanisms, and |
| 239 |
being able to interrupt them immediately would be nice. |
| 240 |
.PP |
| 241 |
Therefore perl 5.8.1 introduces a "backdoor" to restore the pre\-5.8.0 |
| 242 |
(pre\-5.7.3, really) signal behaviour. Just set the environment variable |
| 243 |
PERL_SIGNALS to \f(CW\*(C`unsafe\*(C'\fR, and the old immediate (and unsafe) |
| 244 |
signal handling behaviour returns. See "PERL_SIGNALS" in perlrun |
| 245 |
and "Deferred Signals (Safe Signals)" in perlipc. |
| 246 |
.PP |
| 247 |
In completely unrelated news, you can now use safe signals with |
| 248 |
POSIX::SigAction. See "POSIX::SigAction" in POSIX. |
| 249 |
.SS "Tied Arrays with Negative Array Indices" |
| 250 |
.IX Subsection "Tied Arrays with Negative Array Indices" |
| 251 |
Formerly, the indices passed to \f(CW\*(C`FETCH\*(C'\fR, \f(CW\*(C`STORE\*(C'\fR, \f(CW\*(C`EXISTS\*(C'\fR, and |
| 252 |
\&\f(CW\*(C`DELETE\*(C'\fR methods in tied array class were always non-negative. If |
| 253 |
the actual argument was negative, Perl would call FETCHSIZE implicitly |
| 254 |
and add the result to the index before passing the result to the tied |
| 255 |
array method. This behaviour is now optional. If the tied array class |
| 256 |
contains a package variable named \f(CW$NEGATIVE_INDICES\fR which is set to |
| 257 |
a true value, negative values will be passed to \f(CW\*(C`FETCH\*(C'\fR, \f(CW\*(C`STORE\*(C'\fR, |
| 258 |
\&\f(CW\*(C`EXISTS\*(C'\fR, and \f(CW\*(C`DELETE\*(C'\fR unchanged. |
| 259 |
.SS "local ${$x}" |
| 260 |
.IX Subsection "local ${$x}" |
| 261 |
The syntaxes |
| 262 |
.PP |
| 263 |
.Vb 3 |
| 264 |
\& local ${$x} |
| 265 |
\& local @{$x} |
| 266 |
\& local %{$x} |
| 267 |
.Ve |
| 268 |
.PP |
| 269 |
now do localise variables, given that the \f(CW$x\fR is a valid variable name. |
| 270 |
.SS "Unicode Character Database 4.0.0" |
| 271 |
.IX Subsection "Unicode Character Database 4.0.0" |
| 272 |
The copy of the Unicode Character Database included in Perl 5.8 has |
| 273 |
been updated to 4.0.0 from 3.2.0. This means for example that the |
| 274 |
Unicode character properties are as in Unicode 4.0.0. |
| 275 |
.SS "Deprecation Warnings" |
| 276 |
.IX Subsection "Deprecation Warnings" |
| 277 |
There is one new feature deprecation. Perl 5.8.0 forgot to add |
| 278 |
some deprecation warnings, these warnings have now been added. |
| 279 |
Finally, a reminder of an impending feature removal. |
| 280 |
.PP |
| 281 |
\fI(Reminder) Pseudo-hashes are deprecated (really)\fR |
| 282 |
.IX Subsection "(Reminder) Pseudo-hashes are deprecated (really)" |
| 283 |
.PP |
| 284 |
Pseudo-hashes were deprecated in Perl 5.8.0 and will be removed in |
| 285 |
Perl 5.10.0, see perl58delta for details. Each attempt to access |
| 286 |
pseudo-hashes will trigger the warning \f(CW\*(C`Pseudo\-hashes are deprecated\*(C'\fR. |
| 287 |
If you really want to continue using pseudo-hashes but not to see the |
| 288 |
deprecation warnings, use: |
| 289 |
.PP |
| 290 |
.Vb 1 |
| 291 |
\& no warnings \*(Aqdeprecated\*(Aq; |
| 292 |
.Ve |
| 293 |
.PP |
| 294 |
Or you can continue to use the fields pragma, but please don't |
| 295 |
expect the data structures to be pseudohashes any more. |
| 296 |
.PP |
| 297 |
\fI(Reminder) 5.005\-style threads are deprecated (really)\fR |
| 298 |
.IX Subsection "(Reminder) 5.005-style threads are deprecated (really)" |
| 299 |
.PP |
| 300 |
5.005\-style threads (activated by \f(CW\*(C`use Thread;\*(C'\fR) were deprecated in |
| 301 |
Perl 5.8.0 and will be removed after Perl 5.8, see perl58delta for |
| 302 |
details. Each 5.005\-style thread creation will trigger the warning |
| 303 |
\&\f(CW\*(C`5.005 threads are deprecated\*(C'\fR. If you really want to continue |
| 304 |
using the 5.005 threads but not to see the deprecation warnings, use: |
| 305 |
.PP |
| 306 |
.Vb 1 |
| 307 |
\& no warnings \*(Aqdeprecated\*(Aq; |
| 308 |
.Ve |
| 309 |
.PP |
| 310 |
\fI(Reminder) The $* variable is deprecated (really)\fR |
| 311 |
.IX Subsection "(Reminder) The $* variable is deprecated (really)" |
| 312 |
.PP |
| 313 |
The \f(CW$*\fR variable controlling multi-line matching has been deprecated |
| 314 |
and will be removed after 5.8. The variable has been deprecated for a |
| 315 |
long time, and a deprecation warning \f(CW\*(C`Use of $* is deprecated\*(C'\fR is given, |
| 316 |
now the variable will just finally be removed. The functionality has |
| 317 |
been supplanted by the \f(CW\*(C`/s\*(C'\fR and \f(CW\*(C`/m\*(C'\fR modifiers on pattern matching. |
| 318 |
If you really want to continue using the \f(CW$*\fR\-variable but not to see |
| 319 |
the deprecation warnings, use: |
| 320 |
.PP |
| 321 |
.Vb 1 |
| 322 |
\& no warnings \*(Aqdeprecated\*(Aq; |
| 323 |
.Ve |
| 324 |
.SS "Miscellaneous Enhancements" |
| 325 |
.IX Subsection "Miscellaneous Enhancements" |
| 326 |
\&\f(CW\*(C`map\*(C'\fR in void context is no longer expensive. \f(CW\*(C`map\*(C'\fR is now context |
| 327 |
aware, and will not construct a list if called in void context. |
| 328 |
.PP |
| 329 |
If a socket gets closed by the server while printing to it, the client |
| 330 |
now gets a SIGPIPE. While this new feature was not planned, it fell |
| 331 |
naturally out of PerlIO changes, and is to be considered an accidental |
| 332 |
feature. |
| 333 |
.PP |
| 334 |
PerlIO::get_layers(FH) returns the names of the PerlIO layers |
| 335 |
active on a filehandle. |
| 336 |
.PP |
| 337 |
PerlIO::via layers can now have an optional UTF8 method to |
| 338 |
indicate whether the layer wants to "auto\-:utf8" the stream. |
| 339 |
.PP |
| 340 |
\&\fButf8::is_utf8()\fR has been added as a quick way to test whether |
| 341 |
a scalar is encoded internally in UTF\-8 (Unicode). |
| 342 |
.SH "Modules and Pragmata" |
| 343 |
.IX Header "Modules and Pragmata" |
| 344 |
.SS "Updated Modules And Pragmata" |
| 345 |
.IX Subsection "Updated Modules And Pragmata" |
| 346 |
The following modules and pragmata have been updated since Perl 5.8.0: |
| 347 |
.IP base 4 |
| 348 |
.IX Item "base" |
| 349 |
.PD 0 |
| 350 |
.IP B::Bytecode 4 |
| 351 |
.IX Item "B::Bytecode" |
| 352 |
.PD |
| 353 |
In much better shape than it used to be. Still far from perfect, but |
| 354 |
maybe worth a try. |
| 355 |
.IP B::Concise 4 |
| 356 |
.IX Item "B::Concise" |
| 357 |
.PD 0 |
| 358 |
.IP B::Deparse 4 |
| 359 |
.IX Item "B::Deparse" |
| 360 |
.IP Benchmark 4 |
| 361 |
.IX Item "Benchmark" |
| 362 |
.PD |
| 363 |
An optional feature, \f(CW\*(C`:hireswallclock\*(C'\fR, now allows for high |
| 364 |
resolution wall clock times (uses Time::HiRes). |
| 365 |
.IP ByteLoader 4 |
| 366 |
.IX Item "ByteLoader" |
| 367 |
See B::Bytecode. |
| 368 |
.IP bytes 4 |
| 369 |
.IX Item "bytes" |
| 370 |
Now has bytes::substr. |
| 371 |
.IP CGI 4 |
| 372 |
.IX Item "CGI" |
| 373 |
.PD 0 |
| 374 |
.IP charnames 4 |
| 375 |
.IX Item "charnames" |
| 376 |
.PD |
| 377 |
One can now have custom character name aliases. |
| 378 |
.IP CPAN 4 |
| 379 |
.IX Item "CPAN" |
| 380 |
There is now a simple command line frontend to the CPAN.pm |
| 381 |
module called \fIcpan\fR. |
| 382 |
.IP Data::Dumper 4 |
| 383 |
.IX Item "Data::Dumper" |
| 384 |
A new option, Pair, allows choosing the separator between hash keys |
| 385 |
and values. |
| 386 |
.IP DB_File 4 |
| 387 |
.IX Item "DB_File" |
| 388 |
.PD 0 |
| 389 |
.IP Devel::PPPort 4 |
| 390 |
.IX Item "Devel::PPPort" |
| 391 |
.IP Digest::MD5 4 |
| 392 |
.IX Item "Digest::MD5" |
| 393 |
.IP Encode 4 |
| 394 |
.IX Item "Encode" |
| 395 |
.PD |
| 396 |
Significant updates on the encoding pragma functionality |
| 397 |
(tr/// and the DATA filehandle, formats). |
| 398 |
.Sp |
| 399 |
If a filehandle has been marked as to have an encoding, unmappable |
| 400 |
characters are detected already during input, not later (when the |
| 401 |
corrupted data is being used). |
| 402 |
.Sp |
| 403 |
The ISO 8859\-6 conversion table has been corrected (the 0x30..0x39 |
| 404 |
erroneously mapped to U+0660..U+0669, instead of U+0030..U+0039). The |
| 405 |
GSM 03.38 conversion did not handle escape sequences correctly. The |
| 406 |
UTF\-7 encoding has been added (making Encode feature-complete with |
| 407 |
Unicode::String). |
| 408 |
.IP fields 4 |
| 409 |
.IX Item "fields" |
| 410 |
.PD 0 |
| 411 |
.IP libnet 4 |
| 412 |
.IX Item "libnet" |
| 413 |
.IP Math::BigInt 4 |
| 414 |
.IX Item "Math::BigInt" |
| 415 |
.PD |
| 416 |
A lot of bugs have been fixed since v1.60, the version included in Perl |
| 417 |
v5.8.0. Especially noteworthy are the bug in Calc that caused div and mod to |
| 418 |
fail for some large values, and the fixes to the handling of bad inputs. |
| 419 |
.Sp |
| 420 |
Some new features were added, e.g. the \fBbroot()\fR method, you can now pass |
| 421 |
parameters to \fBconfig()\fR to change some settings at runtime, and it is now |
| 422 |
possible to trap the creation of NaN and infinity. |
| 423 |
.Sp |
| 424 |
As usual, some optimizations took place and made the math overall a tad |
| 425 |
faster. In some cases, quite a lot faster, actually. Especially alternative |
| 426 |
libraries like Math::BigInt::GMP benefit from this. In addition, a lot of the |
| 427 |
quite clunky routines like \fBfsqrt()\fR and \fBflog()\fR are now much much faster. |
| 428 |
.IP MIME::Base64 4 |
| 429 |
.IX Item "MIME::Base64" |
| 430 |
.PD 0 |
| 431 |
.IP NEXT 4 |
| 432 |
.IX Item "NEXT" |
| 433 |
.PD |
| 434 |
Diamond inheritance now works. |
| 435 |
.IP Net::Ping 4 |
| 436 |
.IX Item "Net::Ping" |
| 437 |
.PD 0 |
| 438 |
.IP PerlIO::scalar 4 |
| 439 |
.IX Item "PerlIO::scalar" |
| 440 |
.PD |
| 441 |
Reading from non-string scalars (like the special variables, see |
| 442 |
perlvar) now works. |
| 443 |
.IP podlators 4 |
| 444 |
.IX Item "podlators" |
| 445 |
.PD 0 |
| 446 |
.IP Pod::LaTeX 4 |
| 447 |
.IX Item "Pod::LaTeX" |
| 448 |
.IP PodParsers 4 |
| 449 |
.IX Item "PodParsers" |
| 450 |
.IP Pod::Perldoc 4 |
| 451 |
.IX Item "Pod::Perldoc" |
| 452 |
.PD |
| 453 |
Complete rewrite. As a side-effect, no longer refuses to startup when |
| 454 |
run by root. |
| 455 |
.IP Scalar::Util 4 |
| 456 |
.IX Item "Scalar::Util" |
| 457 |
New utilities: refaddr, isvstring, looks_like_number, set_prototype. |
| 458 |
.IP Storable 4 |
| 459 |
.IX Item "Storable" |
| 460 |
Can now store code references (via B::Deparse, so not foolproof). |
| 461 |
.IP strict 4 |
| 462 |
.IX Item "strict" |
| 463 |
Earlier versions of the strict pragma did not check the parameters |
| 464 |
implicitly passed to its "import" (use) and "unimport" (no) routine. |
| 465 |
This caused the false idiom such as: |
| 466 |
.Sp |
| 467 |
.Vb 2 |
| 468 |
\& use strict qw(@ISA); |
| 469 |
\& @ISA = qw(Foo); |
| 470 |
.Ve |
| 471 |
.Sp |
| 472 |
This however (probably) raised the false expectation that the strict |
| 473 |
refs, vars and subs were being enforced (and that \f(CW@ISA\fR was somehow |
| 474 |
"declared"). But the strict refs, vars, and subs are \fBnot\fR enforced |
| 475 |
when using this false idiom. |
| 476 |
.Sp |
| 477 |
Starting from Perl 5.8.1, the above \fBwill\fR cause an error to be |
| 478 |
raised. This may cause programs which used to execute seemingly |
| 479 |
correctly without warnings and errors to fail when run under 5.8.1. |
| 480 |
This happens because |
| 481 |
.Sp |
| 482 |
.Vb 1 |
| 483 |
\& use strict qw(@ISA); |
| 484 |
.Ve |
| 485 |
.Sp |
| 486 |
will now fail with the error: |
| 487 |
.Sp |
| 488 |
.Vb 1 |
| 489 |
\& Unknown \*(Aqstrict\*(Aq tag(s) \*(Aq@ISA\*(Aq |
| 490 |
.Ve |
| 491 |
.Sp |
| 492 |
The remedy to this problem is to replace this code with the correct idiom: |
| 493 |
.Sp |
| 494 |
.Vb 3 |
| 495 |
\& use strict; |
| 496 |
\& use vars qw(@ISA); |
| 497 |
\& @ISA = qw(Foo); |
| 498 |
.Ve |
| 499 |
.IP Term::ANSIcolor 4 |
| 500 |
.IX Item "Term::ANSIcolor" |
| 501 |
.PD 0 |
| 502 |
.IP Test::Harness 4 |
| 503 |
.IX Item "Test::Harness" |
| 504 |
.PD |
| 505 |
Now much more picky about extra or missing output from test scripts. |
| 506 |
.IP Test::More 4 |
| 507 |
.IX Item "Test::More" |
| 508 |
.PD 0 |
| 509 |
.IP Test::Simple 4 |
| 510 |
.IX Item "Test::Simple" |
| 511 |
.IP Text::Balanced 4 |
| 512 |
.IX Item "Text::Balanced" |
| 513 |
.IP Time::HiRes 4 |
| 514 |
.IX Item "Time::HiRes" |
| 515 |
.PD |
| 516 |
Use of \fBnanosleep()\fR, if available, allows mixing subsecond sleeps with |
| 517 |
alarms. |
| 518 |
.IP threads 4 |
| 519 |
.IX Item "threads" |
| 520 |
Several fixes, for example for \fBjoin()\fR problems and memory |
| 521 |
leaks. In some platforms (like Linux) that use glibc the minimum memory |
| 522 |
footprint of one ithread has been reduced by several hundred kilobytes. |
| 523 |
.IP threads::shared 4 |
| 524 |
.IX Item "threads::shared" |
| 525 |
Many memory leaks have been fixed. |
| 526 |
.IP Unicode::Collate 4 |
| 527 |
.IX Item "Unicode::Collate" |
| 528 |
.PD 0 |
| 529 |
.IP Unicode::Normalize 4 |
| 530 |
.IX Item "Unicode::Normalize" |
| 531 |
.IP Win32::GetFolderPath 4 |
| 532 |
.IX Item "Win32::GetFolderPath" |
| 533 |
.IP Win32::GetOSVersion 4 |
| 534 |
.IX Item "Win32::GetOSVersion" |
| 535 |
.PD |
| 536 |
Now returns extra information. |
| 537 |
.SH "Utility Changes" |
| 538 |
.IX Header "Utility Changes" |
| 539 |
The \f(CW\*(C`h2xs\*(C'\fR utility now produces a more modern layout: |
| 540 |
\&\fIFoo\-Bar/lib/Foo/Bar.pm\fR instead of \fIFoo/Bar/Bar.pm\fR. |
| 541 |
Also, the boilerplate test is now called \fIt/Foo\-Bar.t\fR |
| 542 |
instead of \fIt/1.t\fR. |
| 543 |
.PP |
| 544 |
The Perl debugger (\fIlib/perl5db.pl\fR) has now been extensively |
| 545 |
documented and bugs found while documenting have been fixed. |
| 546 |
.PP |
| 547 |
\&\f(CW\*(C`perldoc\*(C'\fR has been rewritten from scratch to be more robust and |
| 548 |
feature rich. |
| 549 |
.PP |
| 550 |
\&\f(CW\*(C`perlcc \-B\*(C'\fR works now at least somewhat better, while \f(CW\*(C`perlcc \-c\*(C'\fR |
| 551 |
is rather more broken. (The Perl compiler suite as a whole continues |
| 552 |
to be experimental.) |
| 553 |
.SH "New Documentation" |
| 554 |
.IX Header "New Documentation" |
| 555 |
perl573delta has been added to list the differences between the |
| 556 |
(now quite obsolete) development releases 5.7.2 and 5.7.3. |
| 557 |
.PP |
| 558 |
perl58delta has been added: it is the perldelta of 5.8.0, detailing |
| 559 |
the differences between 5.6.0 and 5.8.0. |
| 560 |
.PP |
| 561 |
perlartistic has been added: it is the Artistic License in pod format, |
| 562 |
making it easier for modules to refer to it. |
| 563 |
.PP |
| 564 |
perlcheat has been added: it is a Perl cheat sheet. |
| 565 |
.PP |
| 566 |
perlgpl has been added: it is the GNU General Public License in pod |
| 567 |
format, making it easier for modules to refer to it. |
| 568 |
.PP |
| 569 |
perlmacosx has been added to tell about the installation and use |
| 570 |
of Perl in Mac OS X. |
| 571 |
.PP |
| 572 |
perlos400 has been added to tell about the installation and use |
| 573 |
of Perl in OS/400 PASE. |
| 574 |
.PP |
| 575 |
perlreref has been added: it is a regular expressions quick reference. |
| 576 |
.SH "Installation and Configuration Improvements" |
| 577 |
.IX Header "Installation and Configuration Improvements" |
| 578 |
The Unix standard Perl location, \fI/usr/bin/perl\fR, is no longer |
| 579 |
overwritten by default if it exists. This change was very prudent |
| 580 |
because so many Unix vendors already provide a \fI/usr/bin/perl\fR, |
| 581 |
but simultaneously many system utilities may depend on that |
| 582 |
exact version of Perl, so better not to overwrite it. |
| 583 |
.PP |
| 584 |
One can now specify installation directories for site and vendor man |
| 585 |
and HTML pages, and site and vendor scripts. See \fIINSTALL\fR. |
| 586 |
.PP |
| 587 |
One can now specify a destination directory for Perl installation |
| 588 |
by specifying the DESTDIR variable for \f(CW\*(C`make install\*(C'\fR. (This feature |
| 589 |
is slightly different from the previous \f(CW\*(C`Configure \-Dinstallprefix=...\*(C'\fR.) |
| 590 |
See \fIINSTALL\fR. |
| 591 |
.PP |
| 592 |
gcc versions 3.x introduced a new warning that caused a lot of noise |
| 593 |
during Perl compilation: \f(CW\*(C`gcc \-Ialreadyknowndirectory (warning: |
| 594 |
changing search order)\*(C'\fR. This warning has now been avoided by |
| 595 |
Configure weeding out such directories before the compilation. |
| 596 |
.PP |
| 597 |
One can now build subsets of Perl core modules by using the |
| 598 |
Configure flags \f(CW\*(C`\-Dnoextensions=...\*(C'\fR and \f(CW\*(C`\-Donlyextensions=...\*(C'\fR, |
| 599 |
see \fIINSTALL\fR. |
| 600 |
.SS "Platform-specific enhancements" |
| 601 |
.IX Subsection "Platform-specific enhancements" |
| 602 |
In Cygwin Perl can now be built with threads (\f(CW\*(C`Configure \-Duseithreads\*(C'\fR). |
| 603 |
This works with both Cygwin 1.3.22 and Cygwin 1.5.3. |
| 604 |
.PP |
| 605 |
In newer FreeBSD releases Perl 5.8.0 compilation failed because of |
| 606 |
trying to use \fImalloc.h\fR, which in FreeBSD is just a dummy file, and |
| 607 |
a fatal error to even try to use. Now \fImalloc.h\fR is not used. |
| 608 |
.PP |
| 609 |
Perl is now known to build also in Hitachi HI-UXMPP. |
| 610 |
.PP |
| 611 |
Perl is now known to build again in LynxOS. |
| 612 |
.PP |
| 613 |
Mac OS X now installs with Perl version number embedded in |
| 614 |
installation directory names for easier upgrading of user-compiled |
| 615 |
Perl, and the installation directories in general are more standard. |
| 616 |
In other words, the default installation no longer breaks the |
| 617 |
Apple-provided Perl. On the other hand, with \f(CW\*(C`Configure \-Dprefix=/usr\*(C'\fR |
| 618 |
you can now really replace the Apple-supplied Perl (\fBplease be careful\fR). |
| 619 |
.PP |
| 620 |
Mac OS X now builds Perl statically by default. This change was done |
| 621 |
mainly for faster startup times. The Apple-provided Perl is still |
| 622 |
dynamically linked and shared, and you can enable the sharedness for |
| 623 |
your own Perl builds by \f(CW\*(C`Configure \-Duseshrplib\*(C'\fR. |
| 624 |
.PP |
| 625 |
Perl has been ported to IBM's OS/400 PASE environment. The best way |
| 626 |
to build a Perl for PASE is to use an AIX host as a cross-compilation |
| 627 |
environment. See README.os400. |
| 628 |
.PP |
| 629 |
Yet another cross-compilation option has been added: now Perl builds |
| 630 |
on OpenZaurus, a Linux distribution based on Mandrake + Embedix for |
| 631 |
the Sharp Zaurus PDA. See the Cross/README file. |
| 632 |
.PP |
| 633 |
Tru64 when using gcc 3 drops the optimisation for \fItoke.c\fR to \f(CW\*(C`\-O2\*(C'\fR |
| 634 |
because of gigantic memory use with the default \f(CW\*(C`\-O3\*(C'\fR. |
| 635 |
.PP |
| 636 |
Tru64 can now build Perl with the newer Berkeley DBs. |
| 637 |
.PP |
| 638 |
Building Perl on WinCE has been much enhanced, see \fIREADME.ce\fR |
| 639 |
and \fIREADME.perlce\fR. |
| 640 |
.SH "Selected Bug Fixes" |
| 641 |
.IX Header "Selected Bug Fixes" |
| 642 |
.SS "Closures, eval and lexicals" |
| 643 |
.IX Subsection "Closures, eval and lexicals" |
| 644 |
There have been many fixes in the area of anonymous subs, lexicals and |
| 645 |
closures. Although this means that Perl is now more "correct", it is |
| 646 |
possible that some existing code will break that happens to rely on |
| 647 |
the faulty behaviour. In practice this is unlikely unless your code |
| 648 |
contains a very complex nesting of anonymous subs, evals and lexicals. |
| 649 |
.SS "Generic fixes" |
| 650 |
.IX Subsection "Generic fixes" |
| 651 |
If an input filehandle is marked \f(CW\*(C`:utf8\*(C'\fR and Perl sees illegal UTF\-8 |
| 652 |
coming in when doing \f(CW\*(C`<FH>\*(C'\fR, if warnings are enabled a warning is |
| 653 |
immediately given \- instead of being silent about it and Perl being |
| 654 |
unhappy about the broken data later. (The \f(CW:encoding(utf8)\fR layer |
| 655 |
also works the same way.) |
| 656 |
.PP |
| 657 |
binmode(SOCKET, ":utf8") only worked on the input side, not on the |
| 658 |
output side of the socket. Now it works both ways. |
| 659 |
.PP |
| 660 |
For threaded Perls certain system database functions like \fBgetpwent()\fR |
| 661 |
and \fBgetgrent()\fR now grow their result buffer dynamically, instead of |
| 662 |
failing. This means that at sites with lots of users and groups the |
| 663 |
functions no longer fail by returning only partial results. |
| 664 |
.PP |
| 665 |
Perl 5.8.0 had accidentally broken the capability for users |
| 666 |
to define their own uppercase<\->lowercase Unicode mappings |
| 667 |
(as advertised by the Camel). This feature has been fixed and |
| 668 |
is also documented better. |
| 669 |
.PP |
| 670 |
In 5.8.0 this |
| 671 |
.PP |
| 672 |
.Vb 1 |
| 673 |
\& $some_unicode .= <FH>; |
| 674 |
.Ve |
| 675 |
.PP |
| 676 |
didn't work correctly but instead corrupted the data. This has now |
| 677 |
been fixed. |
| 678 |
.PP |
| 679 |
Tied methods like FETCH etc. may now safely access tied values, i.e. |
| 680 |
resulting in a recursive call to FETCH etc. Remember to break the |
| 681 |
recursion, though. |
| 682 |
.PP |
| 683 |
At startup Perl blocks the SIGFPE signal away since there isn't much |
| 684 |
Perl can do about it. Previously this blocking was in effect also for |
| 685 |
programs executed from within Perl. Now Perl restores the original |
| 686 |
SIGFPE handling routine, whatever it was, before running external |
| 687 |
programs. |
| 688 |
.PP |
| 689 |
Linenumbers in Perl scripts may now be greater than 65536, or 2**16. |
| 690 |
(Perl scripts have always been able to be larger than that, it's just |
| 691 |
that the linenumber for reported errors and warnings have "wrapped |
| 692 |
around".) While scripts that large usually indicate a need to rethink |
| 693 |
your code a bit, such Perl scripts do exist, for example as results |
| 694 |
from generated code. Now linenumbers can go all the way to |
| 695 |
4294967296, or 2**32. |
| 696 |
.SS "Platform-specific fixes" |
| 697 |
.IX Subsection "Platform-specific fixes" |
| 698 |
Linux |
| 699 |
.IP \(bu 4 |
| 700 |
Setting \f(CW$0\fR works again (with certain limitations that |
| 701 |
Perl cannot do much about: see "$0" in perlvar) |
| 702 |
.PP |
| 703 |
HP-UX |
| 704 |
.IP \(bu 4 |
| 705 |
Setting \f(CW$0\fR now works. |
| 706 |
.PP |
| 707 |
VMS |
| 708 |
.IP \(bu 4 |
| 709 |
Configuration now tests for the presence of \f(CWpoll()\fR, and IO::Poll |
| 710 |
now uses the vendor-supplied function if detected. |
| 711 |
.IP \(bu 4 |
| 712 |
A rare access violation at Perl start-up could occur if the Perl image was |
| 713 |
installed with privileges or if there was an identifier with the |
| 714 |
subsystem attribute set in the process's rightslist. Either of these |
| 715 |
circumstances triggered tainting code that contained a pointer bug. |
| 716 |
The faulty pointer arithmetic has been fixed. |
| 717 |
.IP \(bu 4 |
| 718 |
The length limit on values (not keys) in the \f(CW%ENV\fR hash has been raised |
| 719 |
from 255 bytes to 32640 bytes (except when the PERL_ENV_TABLES setting |
| 720 |
overrides the default use of logical names for \f(CW%ENV\fR). If it is |
| 721 |
necessary to access these long values from outside Perl, be aware that |
| 722 |
they are implemented using search list logical names that store the |
| 723 |
value in pieces, each 255\-byte piece (up to 128 of them) being an |
| 724 |
element in the search list. When doing a lookup in \f(CW%ENV\fR from within |
| 725 |
Perl, the elements are combined into a single value. The existing |
| 726 |
VMS-specific ability to access individual elements of a search list |
| 727 |
logical name via the \f(CW$ENV\fR{'foo;N'} syntax (where N is the search list |
| 728 |
index) is unimpaired. |
| 729 |
.IP \(bu 4 |
| 730 |
The piping implementation now uses local rather than global DCL |
| 731 |
symbols for inter-process communication. |
| 732 |
.IP \(bu 4 |
| 733 |
File::Find could become confused when navigating to a relative |
| 734 |
directory whose name collided with a logical name. This problem has |
| 735 |
been corrected by adding directory syntax to relative path names, thus |
| 736 |
preventing logical name translation. |
| 737 |
.PP |
| 738 |
Win32 |
| 739 |
.IP \(bu 4 |
| 740 |
A memory leak in the \fBfork()\fR emulation has been fixed. |
| 741 |
.IP \(bu 4 |
| 742 |
The return value of the \fBioctl()\fR built-in function was accidentally |
| 743 |
broken in 5.8.0. This has been corrected. |
| 744 |
.IP \(bu 4 |
| 745 |
The internal message loop executed by perl during blocking operations |
| 746 |
sometimes interfered with messages that were external to Perl. |
| 747 |
This often resulted in blocking operations terminating prematurely or |
| 748 |
returning incorrect results, when Perl was executing under environments |
| 749 |
that could generate Windows messages. This has been corrected. |
| 750 |
.IP \(bu 4 |
| 751 |
Pipes and sockets are now automatically in binary mode. |
| 752 |
.IP \(bu 4 |
| 753 |
The four-argument form of \fBselect()\fR did not preserve $! (errno) properly |
| 754 |
when there were errors in the underlying call. This is now fixed. |
| 755 |
.IP \(bu 4 |
| 756 |
The "CR CR LF" problem of has been fixed, binmode(FH, ":crlf") |
| 757 |
is now effectively a no-op. |
| 758 |
.SH "New or Changed Diagnostics" |
| 759 |
.IX Header "New or Changed Diagnostics" |
| 760 |
All the warnings related to \fBpack()\fR and \fBunpack()\fR were made more |
| 761 |
informative and consistent. |
| 762 |
.ie n .SS "Changed ""A thread exited while %d threads were running""" |
| 763 |
.el .SS "Changed ""A thread exited while \f(CW%d\fP threads were running""" |
| 764 |
.IX Subsection "Changed ""A thread exited while %d threads were running""" |
| 765 |
The old version |
| 766 |
.PP |
| 767 |
.Vb 1 |
| 768 |
\& A thread exited while %d other threads were still running |
| 769 |
.Ve |
| 770 |
.PP |
| 771 |
was misleading because the "other" included also the thread giving |
| 772 |
the warning. |
| 773 |
.SS "Removed ""Attempt to clear a restricted hash""" |
| 774 |
.IX Subsection "Removed ""Attempt to clear a restricted hash""" |
| 775 |
It is not illegal to clear a restricted hash, so the warning |
| 776 |
was removed. |
| 777 |
.SS "New ""Illegal declaration of anonymous subroutine""" |
| 778 |
.IX Subsection "New ""Illegal declaration of anonymous subroutine""" |
| 779 |
You must specify the block of code for \f(CW\*(C`sub\*(C'\fR. |
| 780 |
.SS "Changed ""Invalid range ""%s"" in transliteration operator""" |
| 781 |
.IX Subsection "Changed ""Invalid range ""%s"" in transliteration operator""" |
| 782 |
The old version |
| 783 |
.PP |
| 784 |
.Vb 1 |
| 785 |
\& Invalid [] range "%s" in transliteration operator |
| 786 |
.Ve |
| 787 |
.PP |
| 788 |
was simply wrong because there are no "[] ranges" in tr///. |
| 789 |
.SS "New ""Missing control char name in \ec""" |
| 790 |
.IX Subsection "New ""Missing control char name in c""" |
| 791 |
Self-explanatory. |
| 792 |
.ie n .SS "New ""Newline in left-justified string for %s""" |
| 793 |
.el .SS "New ""Newline in left-justified string for \f(CW%s\fP""" |
| 794 |
.IX Subsection "New ""Newline in left-justified string for %s""" |
| 795 |
The padding spaces would appear after the newline, which is |
| 796 |
probably not what you had in mind. |
| 797 |
.ie n .SS "New ""Possible precedence problem on bitwise %c operator""" |
| 798 |
.el .SS "New ""Possible precedence problem on bitwise \f(CW%c\fP operator""" |
| 799 |
.IX Subsection "New ""Possible precedence problem on bitwise %c operator""" |
| 800 |
If you think this |
| 801 |
.PP |
| 802 |
.Vb 1 |
| 803 |
\& $x & $y == 0 |
| 804 |
.Ve |
| 805 |
.PP |
| 806 |
tests whether the bitwise AND of \f(CW$x\fR and \f(CW$y\fR is zero, |
| 807 |
you will like this warning. |
| 808 |
.SS "New ""Pseudo-hashes are deprecated""" |
| 809 |
.IX Subsection "New ""Pseudo-hashes are deprecated""" |
| 810 |
This warning should have been already in 5.8.0, since they are. |
| 811 |
.ie n .SS "New ""\fBread()\fP on %s filehandle %s""" |
| 812 |
.el .SS "New ""\fBread()\fP on \f(CW%s\fP filehandle \f(CW%s\fP""" |
| 813 |
.IX Subsection "New ""read() on %s filehandle %s""" |
| 814 |
You cannot \fBread()\fR (or \fBsysread()\fR) from a closed or unopened filehandle. |
| 815 |
.SS "New ""5.005 threads are deprecated""" |
| 816 |
.IX Subsection "New ""5.005 threads are deprecated""" |
| 817 |
This warning should have been already in 5.8.0, since they are. |
| 818 |
.SS "New ""Tied variable freed while still in use""" |
| 819 |
.IX Subsection "New ""Tied variable freed while still in use""" |
| 820 |
Something pulled the plug on a live tied variable, Perl plays |
| 821 |
safe by bailing out. |
| 822 |
.SS "New ""To%s: illegal mapping '%s'""" |
| 823 |
.IX Subsection "New ""To%s: illegal mapping '%s'""" |
| 824 |
An illegal user-defined Unicode casemapping was specified. |
| 825 |
.SS "New ""Use of freed value in iteration""" |
| 826 |
.IX Subsection "New ""Use of freed value in iteration""" |
| 827 |
Something modified the values being iterated over. This is not good. |
| 828 |
.SH "Changed Internals" |
| 829 |
.IX Header "Changed Internals" |
| 830 |
These news matter to you only if you either write XS code or like to |
| 831 |
know about or hack Perl internals (using Devel::Peek or any of the |
| 832 |
\&\f(CW\*(C`B::\*(C'\fR modules counts), or like to run Perl with the \f(CW\*(C`\-D\*(C'\fR option. |
| 833 |
.PP |
| 834 |
The embedding examples of perlembed have been reviewed to be |
| 835 |
up to date and consistent: for example, the correct use of |
| 836 |
\&\fBPERL_SYS_INIT3()\fR and \fBPERL_SYS_TERM()\fR. |
| 837 |
.PP |
| 838 |
Extensive reworking of the pad code (the code responsible |
| 839 |
for lexical variables) has been conducted by Dave Mitchell. |
| 840 |
.PP |
| 841 |
Extensive work on the v\-strings by John Peacock. |
| 842 |
.PP |
| 843 |
UTF\-8 length and position cache: to speed up the handling of Unicode |
| 844 |
(UTF\-8) scalars, a cache was introduced. Potential problems exist if |
| 845 |
an extension bypasses the official APIs and directly modifies the PV |
| 846 |
of an SV: the UTF\-8 cache does not get cleared as it should. |
| 847 |
.PP |
| 848 |
APIs obsoleted in Perl 5.8.0, like sv_2pv, sv_catpvn, sv_catsv, |
| 849 |
sv_setsv, are again available. |
| 850 |
.PP |
| 851 |
Certain Perl core C APIs like cxinc and regatom are no longer |
| 852 |
available at all to code outside the Perl core of the Perl core |
| 853 |
extensions. This is intentional. They never should have been |
| 854 |
available with the shorter names, and if you application depends on |
| 855 |
them, you should (be ashamed and) contact perl5\-porters to discuss |
| 856 |
what are the proper APIs. |
| 857 |
.PP |
| 858 |
Certain Perl core C APIs like \f(CW\*(C`Perl_list\*(C'\fR are no longer available |
| 859 |
without their \f(CW\*(C`Perl_\*(C'\fR prefix. If your XS module stops working |
| 860 |
because some functions cannot be found, in many cases a simple fix is |
| 861 |
to add the \f(CW\*(C`Perl_\*(C'\fR prefix to the function and the thread context |
| 862 |
\&\f(CW\*(C`aTHX_\*(C'\fR as the first argument of the function call. This is also how |
| 863 |
it should always have been done: letting the Perl_\-less forms to leak |
| 864 |
from the core was an accident. For cleaner embedding you can also |
| 865 |
force this for all APIs by defining at compile time the cpp define |
| 866 |
PERL_NO_SHORT_NAMES. |
| 867 |
.PP |
| 868 |
\&\fBPerl_save_bool()\fR has been added. |
| 869 |
.PP |
| 870 |
Regexp objects (those created with \f(CW\*(C`qr\*(C'\fR) now have S\-magic rather than |
| 871 |
R\-magic. This fixed regexps of the form /...(??{...;$x})/ to no |
| 872 |
longer ignore changes made to \f(CW$x\fR. The S\-magic avoids dropping |
| 873 |
the caching optimization and making (??{...}) constructs obscenely |
| 874 |
slow (and consequently useless). See also "Magic Variables" in perlguts. |
| 875 |
Regexp::Copy was affected by this change. |
| 876 |
.PP |
| 877 |
The Perl internal debugging macros \fBDEBUG()\fR and \fBDEB()\fR have been renamed |
| 878 |
to \fBPERL_DEBUG()\fR and \fBPERL_DEB()\fR to avoid namespace conflicts. |
| 879 |
.PP |
| 880 |
\&\f(CW\*(C`\-DL\*(C'\fR removed (the leaktest had been broken and unsupported for years, |
| 881 |
use alternative debugging mallocs or tools like valgrind and Purify). |
| 882 |
.PP |
| 883 |
Verbose modifier \f(CW\*(C`v\*(C'\fR added for \f(CW\*(C`\-DXv\*(C'\fR and \f(CW\*(C`\-Dsv\*(C'\fR, see perlrun. |
| 884 |
.SH "New Tests" |
| 885 |
.IX Header "New Tests" |
| 886 |
In Perl 5.8.0 there were about 69000 separate tests in about 700 test files, |
| 887 |
in Perl 5.8.1 there are about 77000 separate tests in about 780 test files. |
| 888 |
The exact numbers depend on the Perl configuration and on the operating |
| 889 |
system platform. |
| 890 |
.SH "Known Problems" |
| 891 |
.IX Header "Known Problems" |
| 892 |
The hash randomisation mentioned in "Incompatible Changes" is definitely |
| 893 |
problematic: it will wake dormant bugs and shake out bad assumptions. |
| 894 |
.PP |
| 895 |
If you want to use mod_perl 2.x with Perl 5.8.1, you will need |
| 896 |
mod_perl\-1.99_10 or higher. Earlier versions of mod_perl 2.x |
| 897 |
do not work with the randomised hashes. (mod_perl 1.x works fine.) |
| 898 |
You will also need Apache::Test 1.04 or higher. |
| 899 |
.PP |
| 900 |
Many of the rarer platforms that worked 100% or pretty close to it |
| 901 |
with perl 5.8.0 have been left a little bit untended since their |
| 902 |
maintainers have been otherwise busy lately, and therefore there will |
| 903 |
be more failures on those platforms. Such platforms include Mac OS |
| 904 |
Classic, IBM z/OS (and other EBCDIC platforms), and NetWare. The most |
| 905 |
common Perl platforms (Unix and Unix-like, Microsoft platforms, and |
| 906 |
VMS) have large enough testing and expert population that they are |
| 907 |
doing well. |
| 908 |
.SS "Tied hashes in scalar context" |
| 909 |
.IX Subsection "Tied hashes in scalar context" |
| 910 |
Tied hashes do not currently return anything useful in scalar context, |
| 911 |
for example when used as boolean tests: |
| 912 |
.PP |
| 913 |
.Vb 1 |
| 914 |
\& if (%tied_hash) { ... } |
| 915 |
.Ve |
| 916 |
.PP |
| 917 |
The current nonsensical behaviour is always to return false, |
| 918 |
regardless of whether the hash is empty or has elements. |
| 919 |
.PP |
| 920 |
The root cause is that there is no interface for the implementors of |
| 921 |
tied hashes to implement the behaviour of a hash in scalar context. |
| 922 |
.SS "Net::Ping 450_service and 510_ping_udp failures" |
| 923 |
.IX Subsection "Net::Ping 450_service and 510_ping_udp failures" |
| 924 |
The subtests 9 and 18 of lib/Net/Ping/t/450_service.t, and the |
| 925 |
subtest 2 of lib/Net/Ping/t/510_ping_udp.t might fail if you have |
| 926 |
an unusual networking setup. For example in the latter case the |
| 927 |
test is trying to send a UDP ping to the IP address 127.0.0.1. |
| 928 |
.SS B::C |
| 929 |
.IX Subsection "B::C" |
| 930 |
The C\-generating compiler backend B::C (the frontend being |
| 931 |
\&\f(CW\*(C`perlcc \-c\*(C'\fR) is even more broken than it used to be because of |
| 932 |
the extensive lexical variable changes. (The good news is that |
| 933 |
B::Bytecode and ByteLoader are better than they used to be.) |
| 934 |
.SH "Platform Specific Problems" |
| 935 |
.IX Header "Platform Specific Problems" |
| 936 |
.SS "EBCDIC Platforms" |
| 937 |
.IX Subsection "EBCDIC Platforms" |
| 938 |
IBM z/OS and other EBCDIC platforms continue to be problematic |
| 939 |
regarding Unicode support. Many Unicode tests are skipped when |
| 940 |
they really should be fixed. |
| 941 |
.SS "Cygwin 1.5 problems" |
| 942 |
.IX Subsection "Cygwin 1.5 problems" |
| 943 |
In Cygwin 1.5 the \fIio/tell\fR and \fIop/sysio\fR tests have failures for |
| 944 |
some yet unknown reason. In 1.5.5 the threads tests stress_cv, |
| 945 |
stress_re, and stress_string are failing unless the environment |
| 946 |
variable PERLIO is set to "perlio" (which makes also the io/tell |
| 947 |
failure go away). |
| 948 |
.PP |
| 949 |
Perl 5.8.1 does build and work well with Cygwin 1.3: with (uname \-a) |
| 950 |
\&\f(CW\*(C`CYGWIN_NT\-5.0 ... 1.3.22(0.78/3/2) 2003\-03\-18 09:20 i686 ...\*(C'\fR |
| 951 |
a 100% "make test" was achieved with \f(CW\*(C`Configure \-des \-Duseithreads\*(C'\fR. |
| 952 |
.SS "HP-UX: HP cc warnings about sendfile and sendpath" |
| 953 |
.IX Subsection "HP-UX: HP cc warnings about sendfile and sendpath" |
| 954 |
With certain HP C compiler releases (e.g. B.11.11.02) you will |
| 955 |
get many warnings like this (lines wrapped for easier reading): |
| 956 |
.PP |
| 957 |
.Vb 6 |
| 958 |
\& cc: "/usr/include/sys/socket.h", line 504: warning 562: |
| 959 |
\& Redeclaration of "sendfile" with a different storage class specifier: |
| 960 |
\& "sendfile" will have internal linkage. |
| 961 |
\& cc: "/usr/include/sys/socket.h", line 505: warning 562: |
| 962 |
\& Redeclaration of "sendpath" with a different storage class specifier: |
| 963 |
\& "sendpath" will have internal linkage. |
| 964 |
.Ve |
| 965 |
.PP |
| 966 |
The warnings show up both during the build of Perl and during certain |
| 967 |
lib/ExtUtils tests that invoke the C compiler. The warning, however, |
| 968 |
is not serious and can be ignored. |
| 969 |
.SS "IRIX: t/uni/tr_7jis.t falsely failing" |
| 970 |
.IX Subsection "IRIX: t/uni/tr_7jis.t falsely failing" |
| 971 |
The test t/uni/tr_7jis.t is known to report failure under 'make test' |
| 972 |
or the test harness with certain releases of IRIX (at least IRIX 6.5 |
| 973 |
and MIPSpro Compilers Version 7.3.1.1m), but if run manually the test |
| 974 |
fully passes. |
| 975 |
.SS "Mac OS X: no usemymalloc" |
| 976 |
.IX Subsection "Mac OS X: no usemymalloc" |
| 977 |
The Perl malloc (\f(CW\*(C`\-Dusemymalloc\*(C'\fR) does not work at all in Mac OS X. |
| 978 |
This is not that serious, though, since the native malloc works just |
| 979 |
fine. |
| 980 |
.SS "Tru64: No threaded builds with GNU cc (gcc)" |
| 981 |
.IX Subsection "Tru64: No threaded builds with GNU cc (gcc)" |
| 982 |
In the latest Tru64 releases (e.g. v5.1B or later) gcc cannot be used |
| 983 |
to compile a threaded Perl (\-Duseithreads) because the system |
| 984 |
\&\f(CW\*(C`<pthread.h>\*(C'\fR file doesn't know about gcc. |
| 985 |
.SS "Win32: sysopen, sysread, syswrite" |
| 986 |
.IX Subsection "Win32: sysopen, sysread, syswrite" |
| 987 |
As of the 5.8.0 release, \fBsysopen()\fR/\fBsysread()\fR/\fBsyswrite()\fR do not behave |
| 988 |
like they used to in 5.6.1 and earlier with respect to "text" mode. |
| 989 |
These built-ins now always operate in "binary" mode (even if \fBsysopen()\fR |
| 990 |
was passed the O_TEXT flag, or if \fBbinmode()\fR was used on the file |
| 991 |
handle). Note that this issue should only make a difference for disk |
| 992 |
files, as sockets and pipes have always been in "binary" mode in the |
| 993 |
Windows port. As this behavior is currently considered a bug, |
| 994 |
compatible behavior may be re-introduced in a future release. Until |
| 995 |
then, the use of \fBsysopen()\fR, \fBsysread()\fR and \fBsyswrite()\fR is not supported |
| 996 |
for "text" mode operations. |
| 997 |
.SH "Future Directions" |
| 998 |
.IX Header "Future Directions" |
| 999 |
The following things \fBmight\fR happen in future. The first publicly |
| 1000 |
available releases having these characteristics will be the developer |
| 1001 |
releases Perl 5.9.x, culminating in the Perl 5.10.0 release. These |
| 1002 |
are our best guesses at the moment: we reserve the right to rethink. |
| 1003 |
.IP \(bu 4 |
| 1004 |
PerlIO will become The Default. Currently (in Perl 5.8.x) the stdio |
| 1005 |
library is still used if Perl thinks it can use certain tricks to |
| 1006 |
make stdio go \fBreally\fR fast. For future releases our goal is to |
| 1007 |
make PerlIO go even faster. |
| 1008 |
.IP \(bu 4 |
| 1009 |
A new feature called \fIassertions\fR will be available. This means that |
| 1010 |
one can have code called assertions sprinkled in the code: usually |
| 1011 |
they are optimised away, but they can be enabled with the \f(CW\*(C`\-A\*(C'\fR option. |
| 1012 |
.IP \(bu 4 |
| 1013 |
A new operator \f(CW\*(C`//\*(C'\fR (defined-or) will be available. This means that |
| 1014 |
one will be able to say |
| 1015 |
.Sp |
| 1016 |
.Vb 1 |
| 1017 |
\& $a // $b |
| 1018 |
.Ve |
| 1019 |
.Sp |
| 1020 |
instead of |
| 1021 |
.Sp |
| 1022 |
.Vb 1 |
| 1023 |
\& defined $a ? $a : $b |
| 1024 |
.Ve |
| 1025 |
.Sp |
| 1026 |
and |
| 1027 |
.Sp |
| 1028 |
.Vb 1 |
| 1029 |
\& $c //= $d; |
| 1030 |
.Ve |
| 1031 |
.Sp |
| 1032 |
instead of |
| 1033 |
.Sp |
| 1034 |
.Vb 1 |
| 1035 |
\& $c = $d unless defined $c; |
| 1036 |
.Ve |
| 1037 |
.Sp |
| 1038 |
The operator will have the same precedence and associativity as \f(CW\*(C`||\*(C'\fR. |
| 1039 |
A source code patch against the Perl 5.8.1 sources will be available |
| 1040 |
in CPAN as \fIauthors/id/H/HM/HMBRAND/dor\-5.8.1.diff\fR. |
| 1041 |
.IP \(bu 4 |
| 1042 |
\&\f(CWunpack()\fR will default to unpacking the \f(CW$_\fR. |
| 1043 |
.IP \(bu 4 |
| 1044 |
Various Copy-On-Write techniques will be investigated in hopes |
| 1045 |
of speeding up Perl. |
| 1046 |
.IP \(bu 4 |
| 1047 |
CPANPLUS, Inline, and Module::Build will become core modules. |
| 1048 |
.IP \(bu 4 |
| 1049 |
The ability to write true lexically scoped pragmas will be introduced. |
| 1050 |
.IP \(bu 4 |
| 1051 |
Work will continue on the bytecompiler and byteloader. |
| 1052 |
.IP \(bu 4 |
| 1053 |
v\-strings as they currently exist are scheduled to be deprecated. The |
| 1054 |
v\-less form (1.2.3) will become a "version object" when used with \f(CW\*(C`use\*(C'\fR, |
| 1055 |
\&\f(CW\*(C`require\*(C'\fR, and \f(CW$VERSION\fR. $^V will also be a "version object" so the |
| 1056 |
printf("%vd",...) construct will no longer be needed. The v\-ful version |
| 1057 |
(v1.2.3) will become obsolete. The equivalence of strings and v\-strings (e.g. |
| 1058 |
that currently 5.8.0 is equal to "\e5\e8\e0") will go away. \fBThere may be no |
| 1059 |
deprecation warning for v\-strings\fR, though: it is quite hard to detect when |
| 1060 |
v\-strings are being used safely, and when they are not. |
| 1061 |
.IP \(bu 4 |
| 1062 |
5.005 Threads Will Be Removed |
| 1063 |
.IP \(bu 4 |
| 1064 |
The \f(CW$*\fR Variable Will Be Removed |
| 1065 |
(it was deprecated a long time ago) |
| 1066 |
.IP \(bu 4 |
| 1067 |
Pseudohashes Will Be Removed |
| 1068 |
.SH "Reporting Bugs" |
| 1069 |
.IX Header "Reporting Bugs" |
| 1070 |
If you find what you think is a bug, you might check the articles |
| 1071 |
recently posted to the comp.lang.perl.misc newsgroup and the perl |
| 1072 |
bug database at http://bugs.perl.org/ . There may also be |
| 1073 |
information at http://www.perl.com/ , the Perl Home Page. |
| 1074 |
.PP |
| 1075 |
If you believe you have an unreported bug, please run the \fBperlbug\fR |
| 1076 |
program included with your release. Be sure to trim your bug down |
| 1077 |
to a tiny but sufficient test case. Your bug report, along with the |
| 1078 |
output of \f(CW\*(C`perl \-V\*(C'\fR, will be sent off to [email protected] to be |
| 1079 |
analysed by the Perl porting team. You can browse and search |
| 1080 |
the Perl 5 bugs at http://bugs.perl.org/ |
| 1081 |
.SH "SEE ALSO" |
| 1082 |
.IX Header "SEE ALSO" |
| 1083 |
The \fIChanges\fR file for exhaustive details on what changed. |
| 1084 |
.PP |
| 1085 |
The \fIINSTALL\fR file for how to build Perl. |
| 1086 |
.PP |
| 1087 |
The \fIREADME\fR file for general stuff. |
| 1088 |
.PP |
| 1089 |
The \fIArtistic\fR and \fICopying\fR files for copyright information. |