3 perlport - Writing portable Perl
7 Perl runs on numerous operating systems. While most of them share
8 much in common, they also have their own unique features.
10 This document is meant to help you to find out what constitutes portable
11 Perl code. That way once you make a decision to write portably,
12 you know where the lines are drawn, and you can stay within them.
14 There is a tradeoff between taking full advantage of one particular
15 type of computer and taking advantage of a full range of them.
16 Naturally, as you broaden your range and become more diverse, the
17 common factors drop, and you are left with an increasingly smaller
18 area of common ground in which you can operate to accomplish a
19 particular task. Thus, when you begin attacking a problem, it is
20 important to consider under which part of the tradeoff curve you
21 want to operate. Specifically, you must decide whether it is
22 important that the task that you are coding have the full generality
23 of being portable, or whether to just get the job done right now.
24 This is the hardest choice to be made. The rest is easy, because
25 Perl provides many choices, whichever way you want to approach your
28 Looking at it another way, writing portable code is usually about
29 willfully limiting your available choices. Naturally, it takes
30 discipline and sacrifice to do that. The product of portability
31 and convenience may be a constant. You have been warned.
33 Be aware of two important points:
37 =item Not all Perl programs have to be portable
39 There is no reason you should not use Perl as a language to glue Unix
40 tools together, or to prototype a Macintosh application, or to manage the
41 Windows registry. If it makes no sense to aim for portability for one
42 reason or another in a given program, then don't bother.
44 =item Nearly all of Perl already I<is> portable
46 Don't be fooled into thinking that it is hard to create portable Perl
47 code. It isn't. Perl tries its level-best to bridge the gaps between
48 what's available on different platforms, and all the means available to
49 use those features. Thus almost all Perl code runs on any machine
50 without modification. But there are some significant issues in
51 writing portable code, and this document is entirely about those issues.
55 Here's the general rule: When you approach a task commonly done
56 using a whole range of platforms, think about writing portable
57 code. That way, you don't sacrifice much by way of the implementation
58 choices you can avail yourself of, and at the same time you can give
59 your users lots of platform choices. On the other hand, when you have to
60 take advantage of some unique feature of a particular platform, as is
61 often the case with systems programming (whether for Unix, Windows,
62 S<Mac OS>, VMS, etc.), consider writing platform-specific code.
64 When the code will run on only two or three operating systems, you
65 may need to consider only the differences of those particular systems.
66 The important thing is to decide where the code will run and to be
67 deliberate in your decision.
69 The material below is separated into three main sections: main issues of
70 portability (L<"ISSUES">, platform-specific issues (L<"PLATFORMS">, and
71 built-in perl functions that behave differently on various ports
72 (L<"FUNCTION IMPLEMENTATIONS">.
74 This information should not be considered complete; it includes possibly
75 transient information about idiosyncrasies of some of the ports, almost
76 all of which are in a state of constant evolution. Thus, this material
77 should be considered a perpetual work in progress
78 (C<< <IMG SRC="yellow_sign.gif" ALT="Under Construction"> >>).
84 In most operating systems, lines in files are terminated by newlines.
85 Just what is used as a newline may vary from OS to OS. Unix
86 traditionally uses C<\012>, one type of DOSish I/O uses C<\015\012>,
87 and S<Mac OS> uses C<\015>.
89 Perl uses C<\n> to represent the "logical" newline, where what is
90 logical may depend on the platform in use. In MacPerl, C<\n> always
91 means C<\015>. In DOSish perls, C<\n> usually means C<\012>, but
92 when accessing a file in "text" mode, STDIO translates it to (or
93 from) C<\015\012>, depending on whether you're reading or writing.
94 Unix does the same thing on ttys in canonical mode. C<\015\012>
95 is commonly referred to as CRLF.
97 A common cause of unportable programs is the misuse of chop() to trim
107 You can get away with this on Unix and Mac OS (they have a single
108 character end-of-line), but the same program will break under DOSish
109 perls because you're only chop()ing half the end-of-line. Instead,
110 chomp() should be used to trim newlines. The Dunce::Files module can
111 help audit your code for misuses of chop().
113 When dealing with binary files (or text files in binary mode) be sure
114 to explicitly set $/ to the appropriate value for your file format
115 before using chomp().
117 Because of the "text" mode translation, DOSish perls have limitations
118 in using C<seek> and C<tell> on a file accessed in "text" mode.
119 Stick to C<seek>-ing to locations you got from C<tell> (and no
120 others), and you are usually free to use C<seek> and C<tell> even
121 in "text" mode. Using C<seek> or C<tell> or other file operations
122 may be non-portable. If you use C<binmode> on a file, however, you
123 can usually C<seek> and C<tell> with arbitrary values in safety.
125 A common misconception in socket programming is that C<\n> eq C<\012>
126 everywhere. When using protocols such as common Internet protocols,
127 C<\012> and C<\015> are called for specifically, and the values of
128 the logical C<\n> and C<\r> (carriage return) are not reliable.
130 print SOCKET "Hi there, client!\r\n"; # WRONG
131 print SOCKET "Hi there, client!\015\012"; # RIGHT
133 However, using C<\015\012> (or C<\cM\cJ>, or C<\x0D\x0A>) can be tedious
134 and unsightly, as well as confusing to those maintaining the code. As
135 such, the Socket module supplies the Right Thing for those who want it.
137 use Socket qw(:DEFAULT :crlf);
138 print SOCKET "Hi there, client!$CRLF" # RIGHT
140 When reading from a socket, remember that the default input record
141 separator C<$/> is C<\n>, but robust socket code will recognize as
142 either C<\012> or C<\015\012> as end of line:
148 Because both CRLF and LF end in LF, the input record separator can
149 be set to LF and any CR stripped later. Better to write:
151 use Socket qw(:DEFAULT :crlf);
152 local($/) = LF; # not needed if $/ is already \012
155 s/$CR?$LF/\n/; # not sure if socket uses LF or CRLF, OK
156 # s/\015?\012/\n/; # same thing
159 This example is preferred over the previous one--even for Unix
160 platforms--because now any C<\015>'s (C<\cM>'s) are stripped out
161 (and there was much rejoicing).
163 Similarly, functions that return text data--such as a function that
164 fetches a web page--should sometimes translate newlines before
165 returning the data, if they've not yet been translated to the local
166 newline representation. A single line of code will often suffice:
168 $data =~ s/\015?\012/\n/g;
171 Some of this may be confusing. Here's a handy reference to the ASCII CR
172 and LF characters. You can print it out and stick it in your wallet.
174 LF eq \012 eq \x0A eq \cJ eq chr(10) eq ASCII 10
175 CR eq \015 eq \x0D eq \cM eq chr(13) eq ASCII 13
178 ---------------------------
181 \n * | LF | CRLF | CR |
182 \r * | CR | CR | LF |
183 ---------------------------
186 The Unix column assumes that you are not accessing a serial line
187 (like a tty) in canonical mode. If you are, then CR on input becomes
188 "\n", and "\n" on output becomes CRLF.
190 These are just the most common definitions of C<\n> and C<\r> in Perl.
191 There may well be others. For example, on an EBCDIC implementation such
192 as z/OS or OS/400 the above material is similar to "Unix" but the code
195 LF eq \025 eq \x15 eq chr(21) eq CP-1047 21
196 LF eq \045 eq \x25 eq \cU eq chr(37) eq CP-0037 37
197 CR eq \015 eq \x0D eq \cM eq chr(13) eq CP-1047 13
198 CR eq \015 eq \x0D eq \cM eq chr(13) eq CP-0037 13
201 ----------------------
206 ----------------------
209 =head2 Numbers endianness and Width
211 Different CPUs store integers and floating point numbers in different
212 orders (called I<endianness>) and widths (32-bit and 64-bit being the
213 most common today). This affects your programs when they attempt to transfer
214 numbers in binary format from one CPU architecture to another,
215 usually either "live" via network connection, or by storing the
216 numbers to secondary storage such as a disk file or tape.
218 Conflicting storage orders make utter mess out of the numbers. If a
219 little-endian host (Intel, VAX) stores 0x12345678 (305419896 in
220 decimal), a big-endian host (Motorola, Sparc, PA) reads it as
221 0x78563412 (2018915346 in decimal). Alpha and MIPS can be either:
222 Digital/Compaq used/uses them in little-endian mode; SGI/Cray uses
223 them in big-endian mode. To avoid this problem in network (socket)
224 connections use the C<pack> and C<unpack> formats C<n> and C<N>, the
225 "network" orders. These are guaranteed to be portable.
227 You can explore the endianness of your platform by unpacking a
228 data structure packed in native format such as:
230 print unpack("h*", pack("s2", 1, 2)), "\n";
231 # '10002000' on e.g. Intel x86 or Alpha 21064 in little-endian mode
232 # '00100020' on e.g. Motorola 68040
234 If you need to distinguish between endian architectures you could use
235 either of the variables set like so:
237 $is_big_endian = unpack("h*", pack("s", 1)) =~ /01/;
238 $is_little_endian = unpack("h*", pack("s", 1)) =~ /^1/;
240 Differing widths can cause truncation even between platforms of equal
241 endianness. The platform of shorter width loses the upper parts of the
242 number. There is no good solution for this problem except to avoid
243 transferring or storing raw binary numbers.
245 One can circumnavigate both these problems in two ways. Either
246 transfer and store numbers always in text format, instead of raw
247 binary, or else consider using modules like Data::Dumper (included in
248 the standard distribution as of Perl 5.005) and Storable (included as
249 of perl 5.8). Keeping all data as text significantly simplifies matters.
251 The v-strings are portable only up to v2147483647 (0x7FFFFFFF), that's
252 how far EBCDIC, or more precisely UTF-EBCDIC will go.
254 =head2 Files and Filesystems
256 Most platforms these days structure files in a hierarchical fashion.
257 So, it is reasonably safe to assume that all platforms support the
258 notion of a "path" to uniquely identify a file on the system. How
259 that path is really written, though, differs considerably.
261 Although similar, file path specifications differ between Unix,
262 Windows, S<Mac OS>, OS/2, VMS, VOS, S<RISC OS>, and probably others.
263 Unix, for example, is one of the few OSes that has the elegant idea
264 of a single root directory.
266 DOS, OS/2, VMS, VOS, and Windows can work similarly to Unix with C</>
267 as path separator, or in their own idiosyncratic ways (such as having
268 several root directories and various "unrooted" device files such NIL:
271 S<Mac OS> uses C<:> as a path separator instead of C</>.
273 The filesystem may support neither hard links (C<link>) nor
274 symbolic links (C<symlink>, C<readlink>, C<lstat>).
276 The filesystem may support neither access timestamp nor change
277 timestamp (meaning that about the only portable timestamp is the
278 modification timestamp), or one second granularity of any timestamps
279 (e.g. the FAT filesystem limits the time granularity to two seconds).
281 The "inode change timestamp" (the C<-C> filetest) may really be the
282 "creation timestamp" (which it is not in UNIX).
284 VOS perl can emulate Unix filenames with C</> as path separator. The
285 native pathname characters greater-than, less-than, number-sign, and
286 percent-sign are always accepted.
288 S<RISC OS> perl can emulate Unix filenames with C</> as path
289 separator, or go native and use C<.> for path separator and C<:> to
290 signal filesystems and disk names.
292 Don't assume UNIX filesystem access semantics: that read, write,
293 and execute are all the permissions there are, and even if they exist,
294 that their semantics (for example what do r, w, and x mean on
295 a directory) are the UNIX ones. The various UNIX/POSIX compatibility
296 layers usually try to make interfaces like chmod() work, but sometimes
297 there simply is no good mapping.
299 If all this is intimidating, have no (well, maybe only a little)
300 fear. There are modules that can help. The File::Spec modules
301 provide methods to do the Right Thing on whatever platform happens
302 to be running the program.
304 use File::Spec::Functions;
305 chdir(updir()); # go up one directory
306 $file = catfile(curdir(), 'temp', 'file.txt');
307 # on Unix and Win32, './temp/file.txt'
308 # on Mac OS, ':temp:file.txt'
309 # on VMS, '[.temp]file.txt'
311 File::Spec is available in the standard distribution as of version
312 5.004_05. File::Spec::Functions is only in File::Spec 0.7 and later,
313 and some versions of perl come with version 0.6. If File::Spec
314 is not updated to 0.7 or later, you must use the object-oriented
315 interface from File::Spec (or upgrade File::Spec).
317 In general, production code should not have file paths hardcoded.
318 Making them user-supplied or read from a configuration file is
319 better, keeping in mind that file path syntax varies on different
322 This is especially noticeable in scripts like Makefiles and test suites,
323 which often assume C</> as a path separator for subdirectories.
325 Also of use is File::Basename from the standard distribution, which
326 splits a pathname into pieces (base filename, full path to directory,
329 Even when on a single platform (if you can call Unix a single platform),
330 remember not to count on the existence or the contents of particular
331 system-specific files or directories, like F</etc/passwd>,
332 F</etc/sendmail.conf>, F</etc/resolv.conf>, or even F</tmp/>. For
333 example, F</etc/passwd> may exist but not contain the encrypted
334 passwords, because the system is using some form of enhanced security.
335 Or it may not contain all the accounts, because the system is using NIS.
336 If code does need to rely on such a file, include a description of the
337 file and its format in the code's documentation, then make it easy for
338 the user to override the default location of the file.
340 Don't assume a text file will end with a newline. They should,
343 Do not have two files or directories of the same name with different
344 case, like F<test.pl> and F<Test.pl>, as many platforms have
345 case-insensitive (or at least case-forgiving) filenames. Also, try
346 not to have non-word characters (except for C<.>) in the names, and
347 keep them to the 8.3 convention, for maximum portability, onerous a
348 burden though this may appear.
350 Likewise, when using the AutoSplit module, try to keep your functions to
351 8.3 naming and case-insensitive conventions; or, at the least,
352 make it so the resulting files have a unique (case-insensitively)
355 Whitespace in filenames is tolerated on most systems, but not all,
356 and even on systems where it might be tolerated, some utilities
357 might become confused by such whitespace.
359 Many systems (DOS, VMS) cannot have more than one C<.> in their filenames.
361 Don't assume C<< > >> won't be the first character of a filename.
362 Always use C<< < >> explicitly to open a file for reading, or even
363 better, use the three-arg version of open, unless you want the user to
364 be able to specify a pipe open.
366 open(FILE, '<', $existing_file) or die $!;
368 If filenames might use strange characters, it is safest to open it
369 with C<sysopen> instead of C<open>. C<open> is magic and can
370 translate characters like C<< > >>, C<< < >>, and C<|>, which may
371 be the wrong thing to do. (Sometimes, though, it's the right thing.)
372 Three-arg open can also help protect against this translation in cases
373 where it is undesirable.
375 Don't use C<:> as a part of a filename since many systems use that for
376 their own semantics (Mac OS Classic for separating pathname components,
377 many networking schemes and utilities for separating the nodename and
378 the pathname, and so on). For the same reasons, avoid C<@>, C<;> and
381 Don't assume that in pathnames you can collapse two leading slashes
382 C<//> into one: some networking and clustering filesystems have special
383 semantics for that. Let the operating system to sort it out.
385 The I<portable filename characters> as defined by ANSI C are
387 a b c d e f g h i j k l m n o p q r t u v w x y z
388 A B C D E F G H I J K L M N O P Q R T U V W X Y Z
392 and the "-" shouldn't be the first character. If you want to be
393 hypercorrect, stay case-insensitive and within the 8.3 naming
394 convention (all the files and directories have to be unique within one
395 directory if their names are lowercased and truncated to eight
396 characters before the C<.>, if any, and to three characters after the
397 C<.>, if any). (And do not use C<.>s in directory names.)
399 =head2 System Interaction
401 Not all platforms provide a command line. These are usually platforms
402 that rely primarily on a Graphical User Interface (GUI) for user
403 interaction. A program requiring a command line interface might
404 not work everywhere. This is probably for the user of the program
405 to deal with, so don't stay up late worrying about it.
407 Some platforms can't delete or rename files held open by the system.
408 Remember to C<close> files when you are done with them. Don't
409 C<unlink> or C<rename> an open file. Don't C<tie> or C<open> a
410 file already tied or opened; C<untie> or C<close> it first.
412 Don't open the same file more than once at a time for writing, as some
413 operating systems put mandatory locks on such files.
415 Don't assume that write/modify permission on a directory gives the
416 right to add or delete files/directories in that directory. That is
417 filesystem specific: in some filesystems you need write/modify
418 permission also (or even just) in the file/directory itself. In some
419 filesystems (AFS, DFS) the permission to add/delete directory entries
420 is a completely separate permission.
422 Don't assume that a single C<unlink> completely gets rid of the file:
423 some filesystems (most notably the ones in VMS) have versioned
424 filesystems, and unlink() removes only the most recent one (it doesn't
425 remove all the versions because by default the native tools on those
426 platforms remove just the most recent version, too). The portable
427 idiom to remove all the versions of a file is
429 1 while unlink "file";
431 This will terminate if the file is undeleteable for some reason
432 (protected, not there, and so on).
434 Don't count on a specific environment variable existing in C<%ENV>.
435 Don't count on C<%ENV> entries being case-sensitive, or even
436 case-preserving. Don't try to clear %ENV by saying C<%ENV = ();>, or,
437 if you really have to, make it conditional on C<$^O ne 'VMS'> since in
438 VMS the C<%ENV> table is much more than a per-process key-value string
441 Don't count on signals or C<%SIG> for anything.
443 Don't count on filename globbing. Use C<opendir>, C<readdir>, and
446 Don't count on per-program environment variables, or per-program current
449 Don't count on specific values of C<$!>.
451 =head2 Command names versus file pathnames
453 Don't assume that the name used to invoke a command or program with
454 C<system> or C<exec> can also be used to test for the existence of the
455 file that holds the executable code for that command or program.
456 First, many systems have "internal" commands that are built-in to the
457 shell or OS and while these commands can be invoked, there is no
458 corresponding file. Second, some operating systems (e.g., Cygwin,
459 DJGPP, OS/2, and VOS) have required suffixes for executable files;
460 these suffixes are generally permitted on the command name but are not
461 required. Thus, a command like "perl" might exist in a file named
462 "perl", "perl.exe", or "perl.pm", depending on the operating system.
463 The variable "_exe" in the Config module holds the executable suffix,
464 if any. Third, the VMS port carefully sets up $^X and
465 $Config{perlpath} so that no further processing is required. This is
466 just as well, because the matching regular expression used below would
467 then have to deal with a possible trailing version number in the VMS
470 To convert $^X to a file pathname, taking account of the requirements
471 of the various operating system possibilities, say:
475 {$thisperl .= $Config{_exe} unless $thisperl =~ m/$Config{_exe}$/i;}
477 To convert $Config{perlpath} to a file pathname, say:
479 $thisperl = $Config{perlpath};
481 {$thisperl .= $Config{_exe} unless $thisperl =~ m/$Config{_exe}$/i;}
483 =head2 Interprocess Communication (IPC)
485 In general, don't directly access the system in code meant to be
486 portable. That means, no C<system>, C<exec>, C<fork>, C<pipe>,
487 C<``>, C<qx//>, C<open> with a C<|>, nor any of the other things
488 that makes being a perl hacker worth being.
490 Commands that launch external processes are generally supported on
491 most platforms (though many of them do not support any type of
492 forking). The problem with using them arises from what you invoke
493 them on. External tools are often named differently on different
494 platforms, may not be available in the same location, might accept
495 different arguments, can behave differently, and often present their
496 results in a platform-dependent way. Thus, you should seldom depend
497 on them to produce consistent results. (Then again, if you're calling
498 I<netstat -a>, you probably don't expect it to run on both Unix and CP/M.)
500 One especially common bit of Perl code is opening a pipe to B<sendmail>:
502 open(MAIL, '|/usr/lib/sendmail -t')
503 or die "cannot fork sendmail: $!";
505 This is fine for systems programming when sendmail is known to be
506 available. But it is not fine for many non-Unix systems, and even
507 some Unix systems that may not have sendmail installed. If a portable
508 solution is needed, see the various distributions on CPAN that deal
509 with it. Mail::Mailer and Mail::Send in the MailTools distribution are
510 commonly used, and provide several mailing methods, including mail,
511 sendmail, and direct SMTP (via Net::SMTP) if a mail transfer agent is
512 not available. Mail::Sendmail is a standalone module that provides
513 simple, platform-independent mailing.
515 The Unix System V IPC (C<msg*(), sem*(), shm*()>) is not available
516 even on all Unix platforms.
518 Do not use either the bare result of C<pack("N", 10, 20, 30, 40)> or
519 bare v-strings (such as C<v10.20.30.40>) to represent IPv4 addresses:
520 both forms just pack the four bytes into network order. That this
521 would be equal to the C language C<in_addr> struct (which is what the
522 socket code internally uses) is not guaranteed. To be portable use
523 the routines of the Socket extension, such as C<inet_aton()>,
524 C<inet_ntoa()>, and C<sockaddr_in()>.
526 The rule of thumb for portable code is: Do it all in portable Perl, or
527 use a module (that may internally implement it with platform-specific
528 code, but expose a common interface).
530 =head2 External Subroutines (XS)
532 XS code can usually be made to work with any platform, but dependent
533 libraries, header files, etc., might not be readily available or
534 portable, or the XS code itself might be platform-specific, just as Perl
535 code might be. If the libraries and headers are portable, then it is
536 normally reasonable to make sure the XS code is portable, too.
538 A different type of portability issue arises when writing XS code:
539 availability of a C compiler on the end-user's system. C brings
540 with it its own portability issues, and writing XS code will expose
541 you to some of those. Writing purely in Perl is an easier way to
544 =head2 Standard Modules
546 In general, the standard modules work across platforms. Notable
547 exceptions are the CPAN module (which currently makes connections to external
548 programs that may not be available), platform-specific modules (like
549 ExtUtils::MM_VMS), and DBM modules.
551 There is no one DBM module available on all platforms.
552 SDBM_File and the others are generally available on all Unix and DOSish
553 ports, but not in MacPerl, where only NBDM_File and DB_File are
556 The good news is that at least some DBM module should be available, and
557 AnyDBM_File will use whichever module it can find. Of course, then
558 the code needs to be fairly strict, dropping to the greatest common
559 factor (e.g., not exceeding 1K for each record), so that it will
560 work with any DBM module. See L<AnyDBM_File> for more details.
564 The system's notion of time of day and calendar date is controlled in
565 widely different ways. Don't assume the timezone is stored in C<$ENV{TZ}>,
566 and even if it is, don't assume that you can control the timezone through
569 Don't assume that the epoch starts at 00:00:00, January 1, 1970,
570 because that is OS- and implementation-specific. It is better to store a date
571 in an unambiguous representation. The ISO-8601 standard defines
572 "YYYY-MM-DD" as the date format. A text representation (like "1987-12-18")
573 can be easily converted into an OS-specific value using a module like
574 Date::Parse. An array of values, such as those returned by
575 C<localtime>, can be converted to an OS-specific representation using
578 When calculating specific times, such as for tests in time or date modules,
579 it may be appropriate to calculate an offset for the epoch.
582 $offset = Time::Local::timegm(0, 0, 0, 1, 0, 70);
584 The value for C<$offset> in Unix will be C<0>, but in Mac OS will be
585 some large number. C<$offset> can then be added to a Unix time value
586 to get what should be the proper value on any system.
588 =head2 Character sets and character encoding
590 Assume very little about character sets.
592 Assume nothing about numerical values (C<ord>, C<chr>) of characters.
593 Do not use explicit code point ranges (like \xHH-\xHH); use for
594 example symbolic character classes like C<[:print:]>.
596 Do not assume that the alphabetic characters are encoded contiguously
597 (in the numeric sense). There may be gaps.
599 Do not assume anything about the ordering of the characters.
600 The lowercase letters may come before or after the uppercase letters;
601 the lowercase and uppercase may be interlaced so that both `a' and `A'
602 come before `b'; the accented and other international characters may
603 be interlaced so that E<auml> comes before `b'.
605 =head2 Internationalisation
607 If you may assume POSIX (a rather large assumption), you may read
608 more about the POSIX locale system from L<perllocale>. The locale
609 system at least attempts to make things a little bit more portable,
610 or at least more convenient and native-friendly for non-English
611 users. The system affects character sets and encoding, and date
612 and time formatting--amongst other things.
614 =head2 System Resources
616 If your code is destined for systems with severely constrained (or
617 missing!) virtual memory systems then you want to be I<especially> mindful
618 of avoiding wasteful constructs such as:
620 # NOTE: this is no longer "bad" in perl5.005
621 for (0..10000000) {} # bad
622 for (my $x = 0; $x <= 10000000; ++$x) {} # good
624 @lines = <VERY_LARGE_FILE>; # bad
626 while (<FILE>) {$file .= $_} # sometimes bad
627 $file = join('', <FILE>); # better
629 The last two constructs may appear unintuitive to most people. The
630 first repeatedly grows a string, whereas the second allocates a
631 large chunk of memory in one go. On some systems, the second is
632 more efficient that the first.
636 Most multi-user platforms provide basic levels of security, usually
637 implemented at the filesystem level. Some, however, do
638 not-- unfortunately. Thus the notion of user id, or "home" directory,
639 or even the state of being logged-in, may be unrecognizable on many
640 platforms. If you write programs that are security-conscious, it
641 is usually best to know what type of system you will be running
642 under so that you can write code explicitly for that platform (or
645 Don't assume the UNIX filesystem access semantics: the operating
646 system or the filesystem may be using some ACL systems, which are
647 richer languages than the usual rwx. Even if the rwx exist,
648 their semantics might be different.
650 (From security viewpoint testing for permissions before attempting to
651 do something is silly anyway: if one tries this, there is potential
652 for race conditions-- someone or something might change the
653 permissions between the permissions check and the actual operation.
654 Just try the operation.)
656 Don't assume the UNIX user and group semantics: especially, don't
657 expect the C<< $< >> and C<< $> >> (or the C<$(> and C<$)>) to work
658 for switching identities (or memberships).
660 Don't assume set-uid and set-gid semantics. (And even if you do,
661 think twice: set-uid and set-gid are a known can of security worms.)
665 For those times when it is necessary to have platform-specific code,
666 consider keeping the platform-specific code in one place, making porting
667 to other platforms easier. Use the Config module and the special
668 variable C<$^O> to differentiate platforms, as described in
671 Be careful in the tests you supply with your module or programs.
672 Module code may be fully portable, but its tests might not be. This
673 often happens when tests spawn off other processes or call external
674 programs to aid in the testing, or when (as noted above) the tests
675 assume certain things about the filesystem and paths. Be careful
676 not to depend on a specific output style for errors, such as when
677 checking C<$!> after a system call. Some platforms expect a certain
678 output format, and perl on those platforms may have been adjusted
679 accordingly. Most specifically, don't anchor a regex when testing
684 Modules uploaded to CPAN are tested by a variety of volunteers on
685 different platforms. These CPAN testers are notified by mail of each
686 new upload, and reply to the list with PASS, FAIL, NA (not applicable to
687 this platform), or UNKNOWN (unknown), along with any relevant notations.
689 The purpose of the testing is twofold: one, to help developers fix any
690 problems in their code that crop up because of lack of testing on other
691 platforms; two, to provide users with information about whether
692 a given module works on a given platform.
696 =item Mailing list: cpan-testers@perl.org
698 =item Testing results: http://testers.cpan.org/
704 As of version 5.002, Perl is built with a C<$^O> variable that
705 indicates the operating system it was built on. This was implemented
706 to help speed up code that would otherwise have to C<use Config>
707 and use the value of C<$Config{osname}>. Of course, to get more
708 detailed information about the system, looking into C<%Config> is
709 certainly recommended.
711 C<%Config> cannot always be trusted, however, because it was built
712 at compile time. If perl was built in one place, then transferred
713 elsewhere, some values may be wrong. The values may even have been
714 edited after the fact.
718 Perl works on a bewildering variety of Unix and Unix-like platforms (see
719 e.g. most of the files in the F<hints/> directory in the source code kit).
720 On most of these systems, the value of C<$^O> (hence C<$Config{'osname'}>,
721 too) is determined either by lowercasing and stripping punctuation from the
722 first field of the string returned by typing C<uname -a> (or a similar command)
723 at the shell prompt or by testing the file system for the presence of
724 uniquely named files such as a kernel or header file. Here, for example,
725 are a few of the more popular Unix flavors:
727 uname $^O $Config{'archname'}
728 --------------------------------------------
730 BSD/OS bsdos i386-bsdos
732 dgux dgux AViiON-dgux
733 DYNIX/ptx dynixptx i386-dynixptx
734 FreeBSD freebsd freebsd-i386
735 Linux linux arm-linux
736 Linux linux i386-linux
737 Linux linux i586-linux
738 Linux linux ppc-linux
739 HP-UX hpux PA-RISC1.1
741 Mac OS X darwin darwin
742 MachTen PPC machten powerpc-machten
744 NeXT 4 next OPENSTEP-Mach
745 openbsd openbsd i386-openbsd
746 OSF1 dec_osf alpha-dec_osf
747 reliantunix-n svr4 RM400-svr4
748 SCO_SV sco_sv i386-sco_sv
749 SINIX-N svr4 RM400-svr4
750 sn4609 unicos CRAY_C90-unicos
751 sn6521 unicosmk t3e-unicosmk
752 sn9617 unicos CRAY_J90-unicos
753 SunOS solaris sun4-solaris
754 SunOS solaris i86pc-solaris
755 SunOS4 sunos sun4-sunos
757 Because the value of C<$Config{archname}> may depend on the
758 hardware architecture, it can vary more than the value of C<$^O>.
760 =head2 DOS and Derivatives
762 Perl has long been ported to Intel-style microcomputers running under
763 systems like PC-DOS, MS-DOS, OS/2, and most Windows platforms you can
764 bring yourself to mention (except for Windows CE, if you count that).
765 Users familiar with I<COMMAND.COM> or I<CMD.EXE> style shells should
766 be aware that each of these file specifications may have subtle
769 $filespec0 = "c:/foo/bar/file.txt";
770 $filespec1 = "c:\\foo\\bar\\file.txt";
771 $filespec2 = 'c:\foo\bar\file.txt';
772 $filespec3 = 'c:\\foo\\bar\\file.txt';
774 System calls accept either C</> or C<\> as the path separator.
775 However, many command-line utilities of DOS vintage treat C</> as
776 the option prefix, so may get confused by filenames containing C</>.
777 Aside from calling any external programs, C</> will work just fine,
778 and probably better, as it is more consistent with popular usage,
779 and avoids the problem of remembering what to backwhack and what
782 The DOS FAT filesystem can accommodate only "8.3" style filenames. Under
783 the "case-insensitive, but case-preserving" HPFS (OS/2) and NTFS (NT)
784 filesystems you may have to be careful about case returned with functions
785 like C<readdir> or used with functions like C<open> or C<opendir>.
787 DOS also treats several filenames as special, such as AUX, PRN,
788 NUL, CON, COM1, LPT1, LPT2, etc. Unfortunately, sometimes these
789 filenames won't even work if you include an explicit directory
790 prefix. It is best to avoid such filenames, if you want your code
791 to be portable to DOS and its derivatives. It's hard to know what
792 these all are, unfortunately.
794 Users of these operating systems may also wish to make use of
795 scripts such as I<pl2bat.bat> or I<pl2cmd> to
796 put wrappers around your scripts.
798 Newline (C<\n>) is translated as C<\015\012> by STDIO when reading from
799 and writing to files (see L<"Newlines">). C<binmode(FILEHANDLE)>
800 will keep C<\n> translated as C<\012> for that filehandle. Since it is a
801 no-op on other systems, C<binmode> should be used for cross-platform code
802 that deals with binary data. That's assuming you realize in advance
803 that your data is in binary. General-purpose programs should
804 often assume nothing about their data.
806 The C<$^O> variable and the C<$Config{archname}> values for various
807 DOSish perls are as follows:
809 OS $^O $Config{archname} ID Version
810 --------------------------------------------------------
814 Windows 3.1 ? ? 0 3 01
815 Windows 95 MSWin32 MSWin32-x86 1 4 00
816 Windows 98 MSWin32 MSWin32-x86 1 4 10
817 Windows ME MSWin32 MSWin32-x86 1 ?
818 Windows NT MSWin32 MSWin32-x86 2 4 xx
819 Windows NT MSWin32 MSWin32-ALPHA 2 4 xx
820 Windows NT MSWin32 MSWin32-ppc 2 4 xx
821 Windows 2000 MSWin32 MSWin32-x86 2 5 xx
822 Windows XP MSWin32 MSWin32-x86 2 ?
823 Windows CE MSWin32 ? 3
826 The various MSWin32 Perl's can distinguish the OS they are running on
827 via the value of the fifth element of the list returned from
828 Win32::GetOSVersion(). For example:
830 if ($^O eq 'MSWin32') {
831 my @os_version_info = Win32::GetOSVersion();
832 print +('3.1','95','NT')[$os_version_info[4]],"\n";
835 There are also Win32::IsWinNT() and Win32::IsWin95(),
836 try C<perldoc Win32>.
844 The djgpp environment for DOS, http://www.delorie.com/djgpp/
849 The EMX environment for DOS, OS/2, etc. emx@iaehv.nl,
850 http://www.leo.org/pub/comp/os/os2/leo/gnu/emx+gcc/index.html or
851 ftp://hobbes.nmsu.edu/pub/os2/dev/emx/ Also L<perlos2>.
855 Build instructions for Win32 in L<perlwin32>, or under the Cygnus environment
860 The C<Win32::*> modules in L<Win32>.
864 The ActiveState Pages, http://www.activestate.com/
868 The Cygwin environment for Win32; F<README.cygwin> (installed
869 as L<perlcygwin>), http://www.cygwin.com/
873 The U/WIN environment for Win32,
874 http://www.research.att.com/sw/tools/uwin/
878 Build instructions for OS/2, L<perlos2>
884 Any module requiring XS compilation is right out for most people, because
885 MacPerl is built using non-free (and non-cheap!) compilers. Some XS
886 modules that can work with MacPerl are built and distributed in binary
889 Directories are specified as:
891 volume:folder:file for absolute pathnames
892 volume:folder: for absolute pathnames
893 :folder:file for relative pathnames
894 :folder: for relative pathnames
895 :file for relative pathnames
896 file for relative pathnames
898 Files are stored in the directory in alphabetical order. Filenames are
899 limited to 31 characters, and may include any character except for
900 null and C<:>, which is reserved as the path separator.
902 Instead of C<flock>, see C<FSpSetFLock> and C<FSpRstFLock> in the
903 Mac::Files module, or C<chmod(0444, ...)> and C<chmod(0666, ...)>.
905 In the MacPerl application, you can't run a program from the command line;
906 programs that expect C<@ARGV> to be populated can be edited with something
907 like the following, which brings up a dialog box asking for the command
911 @ARGV = split /\s+/, MacPerl::Ask('Arguments?');
914 A MacPerl script saved as a "droplet" will populate C<@ARGV> with the full
915 pathnames of the files dropped onto the script.
917 Mac users can run programs under a type of command line interface
918 under MPW (Macintosh Programmer's Workshop, a free development
919 environment from Apple). MacPerl was first introduced as an MPW
920 tool, and MPW can be used like a shell:
922 perl myscript.plx some arguments
924 ToolServer is another app from Apple that provides access to MPW tools
925 from MPW and the MacPerl app, which allows MacPerl programs to use
926 C<system>, backticks, and piped C<open>.
928 "S<Mac OS>" is the proper name for the operating system, but the value
929 in C<$^O> is "MacOS". To determine architecture, version, or whether
930 the application or MPW tool version is running, check:
932 $is_app = $MacPerl::Version =~ /App/;
933 $is_tool = $MacPerl::Version =~ /MPW/;
934 ($version) = $MacPerl::Version =~ /^(\S+)/;
935 $is_ppc = $MacPerl::Architecture eq 'MacPPC';
936 $is_68k = $MacPerl::Architecture eq 'Mac68K';
938 S<Mac OS X>, based on NeXT's OpenStep OS, runs MacPerl natively, under the
939 "Classic" environment. There is no "Carbon" version of MacPerl to run
940 under the primary Mac OS X environment. S<Mac OS X> and its Open Source
941 version, Darwin, both run Unix perl natively.
949 MacPerl Development, http://dev.macperl.org/ .
953 The MacPerl Pages, http://www.macperl.com/ .
957 The MacPerl mailing lists, http://lists.perl.org/ .
963 Perl on VMS is discussed in L<perlvms> in the perl distribution.
964 Perl on VMS can accept either VMS- or Unix-style file
965 specifications as in either of the following:
967 $ perl -ne "print if /perl_setup/i" SYS$LOGIN:LOGIN.COM
968 $ perl -ne "print if /perl_setup/i" /sys$login/login.com
970 but not a mixture of both as in:
972 $ perl -ne "print if /perl_setup/i" sys$login:/login.com
973 Can't open sys$login:/login.com: file specification syntax error
975 Interacting with Perl from the Digital Command Language (DCL) shell
976 often requires a different set of quotation marks than Unix shells do.
979 $ perl -e "print ""Hello, world.\n"""
982 There are several ways to wrap your perl scripts in DCL F<.COM> files, if
983 you are so inclined. For example:
985 $ write sys$output "Hello from DCL!"
987 $ then perl -x 'f$environment("PROCEDURE")
988 $ else perl -x - 'p1 'p2 'p3 'p4 'p5 'p6 'p7 'p8
989 $ deck/dollars="__END__"
992 print "Hello from Perl!\n";
997 Do take care with C<$ ASSIGN/nolog/user SYS$COMMAND: SYS$INPUT> if your
998 perl-in-DCL script expects to do things like C<< $read = <STDIN>; >>.
1000 Filenames are in the format "name.extension;version". The maximum
1001 length for filenames is 39 characters, and the maximum length for
1002 extensions is also 39 characters. Version is a number from 1 to
1003 32767. Valid characters are C</[A-Z0-9$_-]/>.
1005 VMS's RMS filesystem is case-insensitive and does not preserve case.
1006 C<readdir> returns lowercased filenames, but specifying a file for
1007 opening remains case-insensitive. Files without extensions have a
1008 trailing period on them, so doing a C<readdir> with a file named F<A.;5>
1009 will return F<a.> (though that file could be opened with
1012 RMS had an eight level limit on directory depths from any rooted logical
1013 (allowing 16 levels overall) prior to VMS 7.2. Hence
1014 C<PERL_ROOT:[LIB.2.3.4.5.6.7.8]> is a valid directory specification but
1015 C<PERL_ROOT:[LIB.2.3.4.5.6.7.8.9]> is not. F<Makefile.PL> authors might
1016 have to take this into account, but at least they can refer to the former
1017 as C</PERL_ROOT/lib/2/3/4/5/6/7/8/>.
1019 The VMS::Filespec module, which gets installed as part of the build
1020 process on VMS, is a pure Perl module that can easily be installed on
1021 non-VMS platforms and can be helpful for conversions to and from RMS
1024 What C<\n> represents depends on the type of file opened. It usually
1025 represents C<\012> but it could also be C<\015>, C<\012>, C<\015\012>,
1026 C<\000>, C<\040>, or nothing depending on the file organiztion and
1027 record format. The VMS::Stdio module provides access to the
1028 special fopen() requirements of files with unusual attributes on VMS.
1030 TCP/IP stacks are optional on VMS, so socket routines might not be
1031 implemented. UDP sockets may not be supported.
1033 The value of C<$^O> on OpenVMS is "VMS". To determine the architecture
1034 that you are running on without resorting to loading all of C<%Config>
1035 you can examine the content of the C<@INC> array like so:
1037 if (grep(/VMS_AXP/, @INC)) {
1038 print "I'm on Alpha!\n";
1040 } elsif (grep(/VMS_VAX/, @INC)) {
1041 print "I'm on VAX!\n";
1044 print "I'm not so sure about where $^O is...\n";
1047 On VMS, perl determines the UTC offset from the C<SYS$TIMEZONE_DIFFERENTIAL>
1048 logical name. Although the VMS epoch began at 17-NOV-1858 00:00:00.00,
1049 calls to C<localtime> are adjusted to count offsets from
1050 01-JAN-1970 00:00:00.00, just like Unix.
1058 F<README.vms> (installed as L<README_vms>), L<perlvms>
1062 vmsperl list, majordomo@perl.org
1064 (Put the words C<subscribe vmsperl> in message body.)
1068 vmsperl on the web, http://www.sidhe.org/vmsperl/index.html
1074 Perl on VOS is discussed in F<README.vos> in the perl distribution
1075 (installed as L<perlvos>). Perl on VOS can accept either VOS- or
1076 Unix-style file specifications as in either of the following:
1078 C<< $ perl -ne "print if /perl_setup/i" >system>notices >>
1079 C<< $ perl -ne "print if /perl_setup/i" /system/notices >>
1081 or even a mixture of both as in:
1083 C<< $ perl -ne "print if /perl_setup/i" >system/notices >>
1085 Even though VOS allows the slash character to appear in object
1086 names, because the VOS port of Perl interprets it as a pathname
1087 delimiting character, VOS files, directories, or links whose names
1088 contain a slash character cannot be processed. Such files must be
1089 renamed before they can be processed by Perl. Note that VOS limits
1090 file names to 32 or fewer characters.
1092 Perl on VOS can be built using two different compilers and two different
1093 versions of the POSIX runtime. The recommended method for building full
1094 Perl is with the GNU C compiler and the generally-available version of
1095 VOS POSIX support. See F<README.vos> (installed as L<perlvos>) for
1096 restrictions that apply when Perl is built using the VOS Standard C
1097 compiler or the alpha version of VOS POSIX support.
1099 The value of C<$^O> on VOS is "VOS". To determine the architecture that
1100 you are running on without resorting to loading all of C<%Config> you
1101 can examine the content of the @INC array like so:
1104 print "I'm on a Stratus box!\n";
1106 print "I'm not on a Stratus box!\n";
1110 if (grep(/860/, @INC)) {
1111 print "This box is a Stratus XA/R!\n";
1113 } elsif (grep(/7100/, @INC)) {
1114 print "This box is a Stratus HP 7100 or 8xxx!\n";
1116 } elsif (grep(/8000/, @INC)) {
1117 print "This box is a Stratus HP 8xxx!\n";
1120 print "This box is a Stratus 68K!\n";
1129 F<README.vos> (installed as L<perlvos>)
1133 The VOS mailing list.
1135 There is no specific mailing list for Perl on VOS. You can post
1136 comments to the comp.sys.stratus newsgroup, or subscribe to the general
1137 Stratus mailing list. Send a letter with "subscribe Info-Stratus" in
1138 the message body to majordomo@list.stratagy.com.
1142 VOS Perl on the web at http://ftp.stratus.com/pub/vos/posix/posix.html
1146 =head2 EBCDIC Platforms
1148 Recent versions of Perl have been ported to platforms such as OS/400 on
1149 AS/400 minicomputers as well as OS/390, VM/ESA, and BS2000 for S/390
1150 Mainframes. Such computers use EBCDIC character sets internally (usually
1151 Character Code Set ID 0037 for OS/400 and either 1047 or POSIX-BC for S/390
1152 systems). On the mainframe perl currently works under the "Unix system
1153 services for OS/390" (formerly known as OpenEdition), VM/ESA OpenEdition, or
1154 the BS200 POSIX-BC system (BS2000 is supported in perl 5.6 and greater).
1155 See L<perlos390> for details.
1157 As of R2.5 of USS for OS/390 and Version 2.3 of VM/ESA these Unix
1158 sub-systems do not support the C<#!> shebang trick for script invocation.
1159 Hence, on OS/390 and VM/ESA perl scripts can be executed with a header
1160 similar to the following simple script:
1163 eval 'exec /usr/local/bin/perl -S $0 ${1+"$@"}'
1165 #!/usr/local/bin/perl # just a comment really
1167 print "Hello from perl!\n";
1169 OS/390 will support the C<#!> shebang trick in release 2.8 and beyond.
1170 Calls to C<system> and backticks can use POSIX shell syntax on all
1173 On the AS/400, if PERL5 is in your library list, you may need
1174 to wrap your perl scripts in a CL procedure to invoke them like so:
1177 CALL PGM(PERL5/PERL) PARM('/QOpenSys/hello.pl')
1180 This will invoke the perl script F<hello.pl> in the root of the
1181 QOpenSys file system. On the AS/400 calls to C<system> or backticks
1184 On these platforms, bear in mind that the EBCDIC character set may have
1185 an effect on what happens with some perl functions (such as C<chr>,
1186 C<pack>, C<print>, C<printf>, C<ord>, C<sort>, C<sprintf>, C<unpack>), as
1187 well as bit-fiddling with ASCII constants using operators like C<^>, C<&>
1188 and C<|>, not to mention dealing with socket interfaces to ASCII computers
1189 (see L<"Newlines">).
1191 Fortunately, most web servers for the mainframe will correctly
1192 translate the C<\n> in the following statement to its ASCII equivalent
1193 (C<\r> is the same under both Unix and OS/390 & VM/ESA):
1195 print "Content-type: text/html\r\n\r\n";
1197 The values of C<$^O> on some of these platforms includes:
1199 uname $^O $Config{'archname'}
1200 --------------------------------------------
1203 POSIX-BC posix-bc BS2000-posix-bc
1206 Some simple tricks for determining if you are running on an EBCDIC
1207 platform could include any of the following (perhaps all):
1209 if ("\t" eq "\05") { print "EBCDIC may be spoken here!\n"; }
1211 if (ord('A') == 193) { print "EBCDIC may be spoken here!\n"; }
1213 if (chr(169) eq 'z') { print "EBCDIC may be spoken here!\n"; }
1215 One thing you may not want to rely on is the EBCDIC encoding
1216 of punctuation characters since these may differ from code page to code
1217 page (and once your module or script is rumoured to work with EBCDIC,
1218 folks will want it to work with all EBCDIC character sets).
1228 L<perlos390>, F<README.os390>, F<perlbs2000>, F<README.vmesa>,
1233 The perl-mvs@perl.org list is for discussion of porting issues as well as
1234 general usage issues for all EBCDIC Perls. Send a message body of
1235 "subscribe perl-mvs" to majordomo@perl.org.
1239 AS/400 Perl information at
1240 http://as400.rochester.ibm.com/
1241 as well as on CPAN in the F<ports/> directory.
1245 =head2 Acorn RISC OS
1247 Because Acorns use ASCII with newlines (C<\n>) in text files as C<\012> like
1248 Unix, and because Unix filename emulation is turned on by default,
1249 most simple scripts will probably work "out of the box". The native
1250 filesystem is modular, and individual filesystems are free to be
1251 case-sensitive or insensitive, and are usually case-preserving. Some
1252 native filesystems have name length limits, which file and directory
1253 names are silently truncated to fit. Scripts should be aware that the
1254 standard filesystem currently has a name length limit of B<10>
1255 characters, with up to 77 items in a directory, but other filesystems
1256 may not impose such limitations.
1258 Native filenames are of the form
1260 Filesystem#Special_Field::DiskName.$.Directory.Directory.File
1264 Special_Field is not usually present, but may contain . and $ .
1265 Filesystem =~ m|[A-Za-z0-9_]|
1266 DsicName =~ m|[A-Za-z0-9_/]|
1267 $ represents the root directory
1268 . is the path separator
1269 @ is the current directory (per filesystem but machine global)
1270 ^ is the parent directory
1271 Directory and File =~ m|[^\0- "\.\$\%\&:\@\\^\|\177]+|
1273 The default filename translation is roughly C<tr|/.|./|;>
1275 Note that C<"ADFS::HardDisk.$.File" ne 'ADFS::HardDisk.$.File'> and that
1276 the second stage of C<$> interpolation in regular expressions will fall
1277 foul of the C<$.> if scripts are not careful.
1279 Logical paths specified by system variables containing comma-separated
1280 search lists are also allowed; hence C<System:Modules> is a valid
1281 filename, and the filesystem will prefix C<Modules> with each section of
1282 C<System$Path> until a name is made that points to an object on disk.
1283 Writing to a new file C<System:Modules> would be allowed only if
1284 C<System$Path> contains a single item list. The filesystem will also
1285 expand system variables in filenames if enclosed in angle brackets, so
1286 C<< <System$Dir>.Modules >> would look for the file
1287 S<C<$ENV{'System$Dir'} . 'Modules'>>. The obvious implication of this is
1288 that B<fully qualified filenames can start with C<< <> >>> and should
1289 be protected when C<open> is used for input.
1291 Because C<.> was in use as a directory separator and filenames could not
1292 be assumed to be unique after 10 characters, Acorn implemented the C
1293 compiler to strip the trailing C<.c> C<.h> C<.s> and C<.o> suffix from
1294 filenames specified in source code and store the respective files in
1295 subdirectories named after the suffix. Hence files are translated:
1298 C:foo.h C:h.foo (logical path variable)
1299 sys/os.h sys.h.os (C compiler groks Unix-speak)
1300 10charname.c c.10charname
1301 10charname.o o.10charname
1302 11charname_.c c.11charname (assuming filesystem truncates at 10)
1304 The Unix emulation library's translation of filenames to native assumes
1305 that this sort of translation is required, and it allows a user-defined list
1306 of known suffixes that it will transpose in this fashion. This may
1307 seem transparent, but consider that with these rules C<foo/bar/baz.h>
1308 and C<foo/bar/h/baz> both map to C<foo.bar.h.baz>, and that C<readdir> and
1309 C<glob> cannot and do not attempt to emulate the reverse mapping. Other
1310 C<.>'s in filenames are translated to C</>.
1312 As implied above, the environment accessed through C<%ENV> is global, and
1313 the convention is that program specific environment variables are of the
1314 form C<Program$Name>. Each filesystem maintains a current directory,
1315 and the current filesystem's current directory is the B<global> current
1316 directory. Consequently, sociable programs don't change the current
1317 directory but rely on full pathnames, and programs (and Makefiles) cannot
1318 assume that they can spawn a child process which can change the current
1319 directory without affecting its parent (and everyone else for that
1322 Because native operating system filehandles are global and are currently
1323 allocated down from 255, with 0 being a reserved value, the Unix emulation
1324 library emulates Unix filehandles. Consequently, you can't rely on
1325 passing C<STDIN>, C<STDOUT>, or C<STDERR> to your children.
1327 The desire of users to express filenames of the form
1328 C<< <Foo$Dir>.Bar >> on the command line unquoted causes problems,
1329 too: C<``> command output capture has to perform a guessing game. It
1330 assumes that a string C<< <[^<>]+\$[^<>]> >> is a
1331 reference to an environment variable, whereas anything else involving
1332 C<< < >> or C<< > >> is redirection, and generally manages to be 99%
1333 right. Of course, the problem remains that scripts cannot rely on any
1334 Unix tools being available, or that any tools found have Unix-like command
1337 Extensions and XS are, in theory, buildable by anyone using free
1338 tools. In practice, many don't, as users of the Acorn platform are
1339 used to binary distributions. MakeMaker does run, but no available
1340 make currently copes with MakeMaker's makefiles; even if and when
1341 this should be fixed, the lack of a Unix-like shell will cause
1342 problems with makefile rules, especially lines of the form C<cd
1343 sdbm && make all>, and anything using quoting.
1345 "S<RISC OS>" is the proper name for the operating system, but the value
1346 in C<$^O> is "riscos" (because we don't like shouting).
1350 Perl has been ported to many platforms that do not fit into any of
1351 the categories listed above. Some, such as AmigaOS, Atari MiNT,
1352 BeOS, HP MPE/iX, QNX, Plan 9, and VOS, have been well-integrated
1353 into the standard Perl source code kit. You may need to see the
1354 F<ports/> directory on CPAN for information, and possibly binaries,
1355 for the likes of: aos, Atari ST, lynxos, riscos, Novell Netware,
1356 Tandem Guardian, I<etc.> (Yes, we know that some of these OSes may
1357 fall under the Unix category, but we are not a standards body.)
1359 Some approximate operating system names and their C<$^O> values
1360 in the "OTHER" category include:
1362 OS $^O $Config{'archname'}
1363 ------------------------------------------
1364 Amiga DOS amigaos m68k-amigos
1366 MPE/iX mpeix PA-RISC1.1
1374 Amiga, F<README.amiga> (installed as L<perlamiga>).
1378 Atari, F<README.mint> and Guido Flohr's web page
1379 http://stud.uni-sb.de/~gufl0000/
1383 Be OS, F<README.beos>
1387 HP 300 MPE/iX, F<README.mpeix> and Mark Bixby's web page
1388 http://www.bixby.org/mark/perlix.html
1392 A free perl5-based PERL.NLM for Novell Netware is available in
1393 precompiled binary and source code form from http://www.novell.com/
1394 as well as from CPAN.
1398 Plan 9, F<README.plan9>
1402 =head1 FUNCTION IMPLEMENTATIONS
1404 Listed below are functions that are either completely unimplemented
1405 or else have been implemented differently on various platforms.
1406 Following each description will be, in parentheses, a list of
1407 platforms that the description applies to.
1409 The list may well be incomplete, or even wrong in some places. When
1410 in doubt, consult the platform-specific README files in the Perl
1411 source distribution, and any other documentation resources accompanying
1414 Be aware, moreover, that even among Unix-ish systems there are variations.
1416 For many functions, you can also query C<%Config>, exported by
1417 default from the Config module. For example, to check whether the
1418 platform has the C<lstat> call, check C<$Config{d_lstat}>. See
1419 L<Config> for a full description of available variables.
1421 =head2 Alphabetical Listing of Perl Functions
1431 C<-r>, C<-w>, and C<-x> have a limited meaning only; directories
1432 and applications are executable, and there are no uid/gid
1433 considerations. C<-o> is not supported. (S<Mac OS>)
1435 C<-r>, C<-w>, C<-x>, and C<-o> tell whether the file is accessible,
1436 which may not reflect UIC-based file protections. (VMS)
1438 C<-s> returns the size of the data fork, not the total size of data fork
1439 plus resource fork. (S<Mac OS>).
1441 C<-s> by name on an open file will return the space reserved on disk,
1442 rather than the current extent. C<-s> on an open filehandle returns the
1443 current size. (S<RISC OS>)
1445 C<-R>, C<-W>, C<-X>, C<-O> are indistinguishable from C<-r>, C<-w>,
1446 C<-x>, C<-o>. (S<Mac OS>, Win32, VMS, S<RISC OS>)
1448 C<-b>, C<-c>, C<-k>, C<-g>, C<-p>, C<-u>, C<-A> are not implemented.
1451 C<-g>, C<-k>, C<-l>, C<-p>, C<-u>, C<-A> are not particularly meaningful.
1452 (Win32, VMS, S<RISC OS>)
1454 C<-d> is true if passed a device spec without an explicit directory.
1457 C<-T> and C<-B> are implemented, but might misclassify Mac text files
1458 with foreign characters; this is the case will all platforms, but may
1459 affect S<Mac OS> often. (S<Mac OS>)
1461 C<-x> (or C<-X>) determine if a file ends in one of the executable
1462 suffixes. C<-S> is meaningless. (Win32)
1464 C<-x> (or C<-X>) determine if a file has an executable file type.
1471 Not implemented. (Win32)
1473 =item binmode FILEHANDLE
1475 Meaningless. (S<Mac OS>, S<RISC OS>)
1477 Reopens file and restores pointer; if function fails, underlying
1478 filehandle may be closed, or pointer may be in a different position.
1481 The value returned by C<tell> may be affected after the call, and
1482 the filehandle may be flushed. (Win32)
1486 Only limited meaning. Disabling/enabling write permission is mapped to
1487 locking/unlocking the file. (S<Mac OS>)
1489 Only good for changing "owner" read-write access, "group", and "other"
1490 bits are meaningless. (Win32)
1492 Only good for changing "owner" and "other" read-write access. (S<RISC OS>)
1494 Access permissions are mapped onto VOS access-control list changes. (VOS)
1496 The actual permissions set depend on the value of the C<CYGWIN>
1497 in the SYSTEM environment settings. (Cygwin)
1501 Not implemented. (S<Mac OS>, Win32, Plan9, S<RISC OS>, VOS)
1503 Does nothing, but won't fail. (Win32)
1505 =item chroot FILENAME
1509 Not implemented. (S<Mac OS>, Win32, VMS, Plan9, S<RISC OS>, VOS, VM/ESA)
1511 =item crypt PLAINTEXT,SALT
1513 May not be available if library or source was not provided when building
1516 Not implemented. (VOS)
1520 Not implemented. (VMS, Plan9, VOS)
1522 =item dbmopen HASH,DBNAME,MODE
1524 Not implemented. (VMS, Plan9, VOS)
1528 Not useful. (S<Mac OS>, S<RISC OS>)
1530 Not implemented. (Win32)
1532 Invokes VMS debugger. (VMS)
1536 Not implemented. (S<Mac OS>)
1538 Implemented via Spawn. (VM/ESA)
1540 Does not automatically flush output handles on some platforms.
1541 (SunOS, Solaris, HP-UX)
1547 Emulates UNIX exit() (which considers C<exit 1> to indicate an error) by
1548 mapping the C<1> to SS$_ABORT (C<44>). This behavior may be overridden
1549 with the pragma C<use vmsish 'exit'>. As with the CRTL's exit()
1550 function, C<exit 0> is also mapped to an exit status of SS$_NORMAL
1551 (C<1>); this mapping cannot be overridden. Any other argument to exit()
1552 is used directly as Perl's exit status. (VMS)
1554 =item fcntl FILEHANDLE,FUNCTION,SCALAR
1556 Not implemented. (Win32, VMS)
1558 =item flock FILEHANDLE,OPERATION
1560 Not implemented (S<Mac OS>, VMS, S<RISC OS>, VOS).
1562 Available only on Windows NT (not on Windows 95). (Win32)
1566 Not implemented. (S<Mac OS>, AmigaOS, S<RISC OS>, VOS, VM/ESA)
1568 Emulated using multiple interpreters. See L<perlfork>. (Win32)
1570 Does not automatically flush output handles on some platforms.
1571 (SunOS, Solaris, HP-UX)
1575 Not implemented. (S<Mac OS>, S<RISC OS>)
1579 Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>, VOS)
1583 Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>)
1585 =item getpriority WHICH,WHO
1587 Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>, VOS, VM/ESA)
1591 Not implemented. (S<Mac OS>, Win32)
1593 Not useful. (S<RISC OS>)
1597 Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>)
1599 =item getnetbyname NAME
1601 Not implemented. (S<Mac OS>, Win32, Plan9)
1605 Not implemented. (S<Mac OS>, Win32)
1607 Not useful. (S<RISC OS>)
1611 Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>)
1613 =item getnetbyaddr ADDR,ADDRTYPE
1615 Not implemented. (S<Mac OS>, Win32, Plan9)
1617 =item getprotobynumber NUMBER
1619 Not implemented. (S<Mac OS>)
1621 =item getservbyport PORT,PROTO
1623 Not implemented. (S<Mac OS>)
1627 Not implemented. (S<Mac OS>, Win32, VM/ESA)
1631 Not implemented. (S<Mac OS>, Win32, VMS, VM/ESA)
1635 Not implemented. (S<Mac OS>, Win32)
1639 Not implemented. (S<Mac OS>, Win32, Plan9)
1643 Not implemented. (S<Mac OS>, Win32, Plan9)
1647 Not implemented. (Win32, Plan9)
1649 =item sethostent STAYOPEN
1651 Not implemented. (S<Mac OS>, Win32, Plan9, S<RISC OS>)
1653 =item setnetent STAYOPEN
1655 Not implemented. (S<Mac OS>, Win32, Plan9, S<RISC OS>)
1657 =item setprotoent STAYOPEN
1659 Not implemented. (S<Mac OS>, Win32, Plan9, S<RISC OS>)
1661 =item setservent STAYOPEN
1663 Not implemented. (Plan9, Win32, S<RISC OS>)
1667 Not implemented. (S<Mac OS>, MPE/iX, VM/ESA, Win32)
1671 Not implemented. (S<Mac OS>, MPE/iX, S<RISC OS>, VM/ESA, VMS, Win32)
1675 Not implemented. (S<Mac OS>, Win32)
1679 Not implemented. (S<Mac OS>, Win32, Plan9)
1683 Not implemented. (S<Mac OS>, Win32, Plan9)
1687 Not implemented. (Plan9, Win32)
1689 =item getsockopt SOCKET,LEVEL,OPTNAME
1691 Not implemented. (Plan9)
1697 This operator is implemented via the File::Glob extension on most
1698 platforms. See L<File::Glob> for portability information.
1700 =item ioctl FILEHANDLE,FUNCTION,SCALAR
1702 Not implemented. (VMS)
1704 Available only for socket handles, and it does what the ioctlsocket() call
1705 in the Winsock API does. (Win32)
1707 Available only for socket handles. (S<RISC OS>)
1709 =item kill SIGNAL, LIST
1711 C<kill(0, LIST)> is implemented for the sake of taint checking;
1712 use with other signals is unimplemented. (S<Mac OS>)
1714 Not implemented, hence not useful for taint checking. (S<RISC OS>)
1716 C<kill()> doesn't have the semantics of C<raise()>, i.e. it doesn't send
1717 a signal to the identified process like it does on Unix platforms.
1718 Instead C<kill($sig, $pid)> terminates the process identified by $pid,
1719 and makes it exit immediately with exit status $sig. As in Unix, if
1720 $sig is 0 and the specified process exists, it returns true without
1721 actually terminating it. (Win32)
1723 =item link OLDFILE,NEWFILE
1725 Not implemented. (S<Mac OS>, MPE/iX, VMS, S<RISC OS>)
1727 Link count not updated because hard links are not quite that hard
1728 (They are sort of half-way between hard and soft links). (AmigaOS)
1730 Hard links are implemented on Win32 (Windows NT and Windows 2000)
1733 =item lstat FILEHANDLE
1739 Not implemented. (VMS, S<RISC OS>)
1741 Return values (especially for device and inode) may be bogus. (Win32)
1743 =item msgctl ID,CMD,ARG
1745 =item msgget KEY,FLAGS
1747 =item msgsnd ID,MSG,FLAGS
1749 =item msgrcv ID,VAR,SIZE,TYPE,FLAGS
1751 Not implemented. (S<Mac OS>, Win32, VMS, Plan9, S<RISC OS>, VOS)
1753 =item open FILEHANDLE,EXPR
1755 =item open FILEHANDLE
1757 The C<|> variants are supported only if ToolServer is installed.
1760 open to C<|-> and C<-|> are unsupported. (S<Mac OS>, Win32, S<RISC OS>)
1762 Opening a process does not automatically flush output handles on some
1763 platforms. (SunOS, Solaris, HP-UX)
1765 =item pipe READHANDLE,WRITEHANDLE
1767 Very limited functionality. (MiNT)
1773 Not implemented. (Win32, VMS, S<RISC OS>)
1775 =item select RBITS,WBITS,EBITS,TIMEOUT
1777 Only implemented on sockets. (Win32, VMS)
1779 Only reliable on sockets. (S<RISC OS>)
1781 Note that the C<select FILEHANDLE> form is generally portable.
1783 =item semctl ID,SEMNUM,CMD,ARG
1785 =item semget KEY,NSEMS,FLAGS
1787 =item semop KEY,OPSTRING
1789 Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>, VOS)
1793 Not implemented. (S<Mac OS>, MPE/iX, VMS, Win32, VMS, S<RISC OS>)
1795 =item setpgrp PID,PGRP
1797 Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>, VOS)
1799 =item setpriority WHICH,WHO,PRIORITY
1801 Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>, VOS)
1805 Not implemented. (S<Mac OS>, MPE/iX, Win32, S<RISC OS>)
1807 =item setsockopt SOCKET,LEVEL,OPTNAME,OPTVAL
1809 Not implemented. (Plan9)
1811 =item shmctl ID,CMD,ARG
1813 =item shmget KEY,SIZE,FLAGS
1815 =item shmread ID,VAR,POS,SIZE
1817 =item shmwrite ID,STRING,POS,SIZE
1819 Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>, VOS)
1821 =item sockatmark SOCKET
1823 A relatively recent addition to socket functions, may not
1824 be implemented even in UNIX platforms.
1826 =item socketpair SOCKET1,SOCKET2,DOMAIN,TYPE,PROTOCOL
1828 Not implemented. (Win32, VMS, S<RISC OS>, VOS, VM/ESA)
1830 =item stat FILEHANDLE
1836 Platforms that do not have rdev, blksize, or blocks will return these
1837 as '', so numeric comparison or manipulation of these fields may cause
1838 'not numeric' warnings.
1840 mtime and atime are the same thing, and ctime is creation time instead of
1841 inode change time. (S<Mac OS>).
1843 ctime not supported on UFS (S<Mac OS X>).
1845 ctime is creation time instead of inode change time (Win32).
1847 device and inode are not meaningful. (Win32)
1849 device and inode are not necessarily reliable. (VMS)
1851 mtime, atime and ctime all return the last modification time. Device and
1852 inode are not necessarily reliable. (S<RISC OS>)
1854 dev, rdev, blksize, and blocks are not available. inode is not
1855 meaningful and will differ between stat calls on the same file. (os2)
1857 some versions of cygwin when doing a stat("foo") and if not finding it
1858 may then attempt to stat("foo.exe") (Cygwin)
1860 =item symlink OLDFILE,NEWFILE
1862 Not implemented. (Win32, VMS, S<RISC OS>)
1866 Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>, VOS, VM/ESA)
1868 =item sysopen FILEHANDLE,FILENAME,MODE,PERMS
1870 The traditional "0", "1", and "2" MODEs are implemented with different
1871 numeric values on some systems. The flags exported by C<Fcntl>
1872 (O_RDONLY, O_WRONLY, O_RDWR) should work everywhere though. (S<Mac
1873 OS>, OS/390, VM/ESA)
1877 In general, do not assume the UNIX/POSIX semantics that you can shift
1878 C<$?> right by eight to get the exit value, or that C<$? & 127>
1879 would give you the number of the signal that terminated the program,
1880 or that C<$? & 128> would test true if the program was terminated by a
1881 coredump. Instead, use the POSIX W*() interfaces: for example, use
1882 WIFEXITED($?) and WEXITVALUE($?) to test for a normal exit and the exit
1883 value, WIFSIGNALED($?) and WTERMSIG($?) for a signal exit and the
1884 signal. Core dumping is not a portable concept, so there's no portable
1885 way to test for that.
1887 Only implemented if ToolServer is installed. (S<Mac OS>)
1889 As an optimization, may not call the command shell specified in
1890 C<$ENV{PERL5SHELL}>. C<system(1, @args)> spawns an external
1891 process and immediately returns its process designator, without
1892 waiting for it to terminate. Return value may be used subsequently
1893 in C<wait> or C<waitpid>. Failure to spawn() a subprocess is indicated
1894 by setting $? to "255 << 8". C<$?> is set in a way compatible with
1895 Unix (i.e. the exitstatus of the subprocess is obtained by "$? >> 8",
1896 as described in the documentation). (Win32)
1898 There is no shell to process metacharacters, and the native standard is
1899 to pass a command line terminated by "\n" "\r" or "\0" to the spawned
1900 program. Redirection such as C<< > foo >> is performed (if at all) by
1901 the run time library of the spawned program. C<system> I<list> will call
1902 the Unix emulation library's C<exec> emulation, which attempts to provide
1903 emulation of the stdin, stdout, stderr in force in the parent, providing
1904 the child program uses a compatible version of the emulation library.
1905 I<scalar> will call the native command line direct and no such emulation
1906 of a child Unix program will exists. Mileage B<will> vary. (S<RISC OS>)
1908 Far from being POSIX compliant. Because there may be no underlying
1909 /bin/sh tries to work around the problem by forking and execing the
1910 first token in its argument string. Handles basic redirection
1911 ("<" or ">") on its own behalf. (MiNT)
1913 Does not automatically flush output handles on some platforms.
1914 (SunOS, Solaris, HP-UX)
1916 The return value is POSIX-like (shifted up by 8 bits), which only allows
1917 room for a made-up value derived from the severity bits of the native
1918 32-bit condition code (unless overridden by C<use vmsish 'status'>).
1919 For more details see L<perlvms/$?>. (VMS)
1923 Only the first entry returned is nonzero. (S<Mac OS>)
1925 "cumulative" times will be bogus. On anything other than Windows NT
1926 or Windows 2000, "system" time will be bogus, and "user" time is
1927 actually the time returned by the clock() function in the C runtime
1930 Not useful. (S<RISC OS>)
1932 =item truncate FILEHANDLE,LENGTH
1934 =item truncate EXPR,LENGTH
1936 Not implemented. (Older versions of VMS)
1938 Truncation to zero-length only. (VOS)
1940 If a FILEHANDLE is supplied, it must be writable and opened in append
1941 mode (i.e., use C<<< open(FH, '>>filename') >>>
1942 or C<sysopen(FH,...,O_APPEND|O_RDWR)>. If a filename is supplied, it
1943 should not be held open elsewhere. (Win32)
1949 Returns undef where unavailable, as of version 5.005.
1951 C<umask> works but the correct permissions are set only when the file
1952 is finally closed. (AmigaOS)
1956 Only the modification time is updated. (S<BeOS>, S<Mac OS>, VMS, S<RISC OS>)
1958 May not behave as expected. Behavior depends on the C runtime
1959 library's implementation of utime(), and the filesystem being
1960 used. The FAT filesystem typically does not support an "access
1961 time" field, and it may limit timestamps to a granularity of
1962 two seconds. (Win32)
1966 =item waitpid PID,FLAGS
1968 Not implemented. (S<Mac OS>, VOS)
1970 Can only be applied to process handles returned for processes spawned
1971 using C<system(1, ...)> or pseudo processes created with C<fork()>. (Win32)
1973 Not useful. (S<RISC OS>)
1981 =item v1.48, 02 February 2001
1983 Various updates from perl5-porters over the past year, supported
1984 platforms update from Jarkko Hietaniemi.
1986 =item v1.47, 22 March 2000
1988 Various cleanups from Tom Christiansen, including migration of
1989 long platform listings from L<perl>.
1991 =item v1.46, 12 February 2000
1993 Updates for VOS and MPE/iX. (Peter Prymmer) Other small changes.
1995 =item v1.45, 20 December 1999
1997 Small changes from 5.005_63 distribution, more changes to EBCDIC info.
1999 =item v1.44, 19 July 1999
2001 A bunch of updates from Peter Prymmer for C<$^O> values,
2002 endianness, File::Spec, VMS, BS2000, OS/400.
2004 =item v1.43, 24 May 1999
2006 Added a lot of cleaning up from Tom Christiansen.
2008 =item v1.42, 22 May 1999
2010 Added notes about tests, sprintf/printf, and epoch offsets.
2012 =item v1.41, 19 May 1999
2014 Lots more little changes to formatting and content.
2016 Added a bunch of C<$^O> and related values
2017 for various platforms; fixed mail and web addresses, and added
2018 and changed miscellaneous notes. (Peter Prymmer)
2020 =item v1.40, 11 April 1999
2022 Miscellaneous changes.
2024 =item v1.39, 11 February 1999
2026 Changes from Jarkko and EMX URL fixes Michael Schwern. Additional
2027 note about newlines added.
2029 =item v1.38, 31 December 1998
2031 More changes from Jarkko.
2033 =item v1.37, 19 December 1998
2035 More minor changes. Merge two separate version 1.35 documents.
2037 =item v1.36, 9 September 1998
2039 Updated for Stratus VOS. Also known as version 1.35.
2041 =item v1.35, 13 August 1998
2043 Integrate more minor changes, plus addition of new sections under
2044 L<"ISSUES">: L<"Numbers endianness and Width">,
2045 L<"Character sets and character encoding">,
2046 L<"Internationalisation">.
2048 =item v1.33, 06 August 1998
2050 Integrate more minor changes.
2052 =item v1.32, 05 August 1998
2054 Integrate more minor changes.
2056 =item v1.30, 03 August 1998
2058 Major update for RISC OS, other minor changes.
2060 =item v1.23, 10 July 1998
2062 First public release with perl5.005.
2066 =head1 Supported Platforms
2068 As of June 2002 (the Perl release 5.8.0), the following platforms are
2069 able to build Perl from the standard source code distribution
2070 available at http://www.cpan.org/src/index.html
2097 Tru64 UNIX (DEC OSF/1, Digital UNIX)
2102 Win95/98/ME/2K/XP 2)
2107 1) in DOS mode either the DOS or OS/2 ports can be used
2108 2) compilers: Borland, Cygwin (GCC), MinGW (GCC), VC6
2110 The following platforms worked with the previous releases (5.6 and
2111 5.7), but we did not manage either to fix or to test these in time
2112 for the 5.8.0 release. There is a very good chance that many of these
2113 will work fine with the 5.8.0. The only one known for certain to be
2114 broken for 5.8.0 is the AmigaOS.
2128 The following platforms have been known to build Perl from source in
2129 the past (5.005_03 and earlier), but we haven't been able to verify
2130 their status for the current release, either because the
2131 hardware/software platforms are rare or because we don't have an
2132 active champion on these platforms--or both. They used to work,
2133 though, so go ahead and try compiling them, and let perlbug@perl.org
2170 The following platforms have their own source code distributions and
2171 binaries available via http://www.cpan.org/ports/
2176 Tandem Guardian 5.004
2178 The following platforms have only binaries available via
2179 http://www.cpan.org/ports/index.html :
2183 Acorn RISCOS 5.005_02
2187 Although we do suggest that you always build your own Perl from
2188 the source code, both for maximal configurability and for security,
2189 in case you are in a hurry you can check
2190 http://www.cpan.org/ports/index.html for binary distributions.
2194 L<perlaix>, L<perlamiga>, L<perlapollo>, L<perlbeos>, L<perlbs2000>,
2195 L<perlce>, L<perlcygwin>, L<perldgux>, L<perldos>, L<perlepoc>, L<perlebcdic>,
2196 L<perlhurd>, L<perlhpux>, L<perlmachten>, L<perlmacos>, L<perlmint>,
2197 L<perlmpeix>, L<perlnetware>, L<perlos2>, L<perlos390>, L<perlplan9>,
2198 L<perlqnx>, L<perlsolaris>, L<perltru64>, L<perlunicode>,
2199 L<perlvmesa>, L<perlvms>, L<perlvos>, L<perlwin32>, and L<Win32>.
2201 =head1 AUTHORS / CONTRIBUTORS
2203 Abigail <abigail@foad.org>,
2204 Charles Bailey <bailey@newman.upenn.edu>,
2205 Graham Barr <gbarr@pobox.com>,
2206 Tom Christiansen <tchrist@perl.com>,
2207 Nicholas Clark <nick@ccl4.org>,
2208 Thomas Dorner <Thomas.Dorner@start.de>,
2209 Andy Dougherty <doughera@lafayette.edu>,
2210 Dominic Dunlop <domo@computer.org>,
2211 Neale Ferguson <neale@vma.tabnsw.com.au>,
2212 David J. Fiander <davidf@mks.com>,
2213 Paul Green <Paul_Green@stratus.com>,
2214 M.J.T. Guy <mjtg@cam.ac.uk>,
2215 Jarkko Hietaniemi <jhi@iki.fi>,
2216 Luther Huffman <lutherh@stratcom.com>,
2217 Nick Ing-Simmons <nick@ing-simmons.net>,
2218 Andreas J. KE<ouml>nig <a.koenig@mind.de>,
2219 Markus Laker <mlaker@contax.co.uk>,
2220 Andrew M. Langmead <aml@world.std.com>,
2221 Larry Moore <ljmoore@freespace.net>,
2222 Paul Moore <Paul.Moore@uk.origin-it.com>,
2223 Chris Nandor <pudge@pobox.com>,
2224 Matthias Neeracher <neeracher@mac.com>,
2225 Philip Newton <pne@cpan.org>,
2226 Gary Ng <71564.1743@CompuServe.COM>,
2227 Tom Phoenix <rootbeer@teleport.com>,
2228 AndrE<eacute> Pirard <A.Pirard@ulg.ac.be>,
2229 Peter Prymmer <pvhp@forte.com>,
2230 Hugo van der Sanden <hv@crypt0.demon.co.uk>,
2231 Gurusamy Sarathy <gsar@activestate.com>,
2232 Paul J. Schinder <schinder@pobox.com>,
2233 Michael G Schwern <schwern@pobox.com>,
2234 Dan Sugalski <dan@sidhe.org>,
2235 Nathan Torkington <gnat@frii.com>.