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 MacOS (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 (MacOS 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 Interprocess Communication (IPC)
453 In general, don't directly access the system in code meant to be
454 portable. That means, no C<system>, C<exec>, C<fork>, C<pipe>,
455 C<``>, C<qx//>, C<open> with a C<|>, nor any of the other things
456 that makes being a perl hacker worth being.
458 Commands that launch external processes are generally supported on
459 most platforms (though many of them do not support any type of
460 forking). The problem with using them arises from what you invoke
461 them on. External tools are often named differently on different
462 platforms, may not be available in the same location, might accept
463 different arguments, can behave differently, and often present their
464 results in a platform-dependent way. Thus, you should seldom depend
465 on them to produce consistent results. (Then again, if you're calling
466 I<netstat -a>, you probably don't expect it to run on both Unix and CP/M.)
468 One especially common bit of Perl code is opening a pipe to B<sendmail>:
470 open(MAIL, '|/usr/lib/sendmail -t')
471 or die "cannot fork sendmail: $!";
473 This is fine for systems programming when sendmail is known to be
474 available. But it is not fine for many non-Unix systems, and even
475 some Unix systems that may not have sendmail installed. If a portable
476 solution is needed, see the various distributions on CPAN that deal
477 with it. Mail::Mailer and Mail::Send in the MailTools distribution are
478 commonly used, and provide several mailing methods, including mail,
479 sendmail, and direct SMTP (via Net::SMTP) if a mail transfer agent is
480 not available. Mail::Sendmail is a standalone module that provides
481 simple, platform-independent mailing.
483 The Unix System V IPC (C<msg*(), sem*(), shm*()>) is not available
484 even on all Unix platforms.
486 Do not use either the bare result of C<pack("N", 10, 20, 30, 40)> or
487 bare v-strings (such as C<v10.20.30.40>) to represent IPv4 addresses:
488 both forms just pack the four bytes into network order. That this
489 would be equal to the C language C<in_addr> struct (which is what the
490 socket code internally uses) is not guaranteed. To be portable use
491 the routines of the Socket extension, such as C<inet_aton()>,
492 C<inet_ntoa()>, and C<sockaddr_in()>.
494 The rule of thumb for portable code is: Do it all in portable Perl, or
495 use a module (that may internally implement it with platform-specific
496 code, but expose a common interface).
498 =head2 External Subroutines (XS)
500 XS code can usually be made to work with any platform, but dependent
501 libraries, header files, etc., might not be readily available or
502 portable, or the XS code itself might be platform-specific, just as Perl
503 code might be. If the libraries and headers are portable, then it is
504 normally reasonable to make sure the XS code is portable, too.
506 A different type of portability issue arises when writing XS code:
507 availability of a C compiler on the end-user's system. C brings
508 with it its own portability issues, and writing XS code will expose
509 you to some of those. Writing purely in Perl is an easier way to
512 =head2 Standard Modules
514 In general, the standard modules work across platforms. Notable
515 exceptions are the CPAN module (which currently makes connections to external
516 programs that may not be available), platform-specific modules (like
517 ExtUtils::MM_VMS), and DBM modules.
519 There is no one DBM module available on all platforms.
520 SDBM_File and the others are generally available on all Unix and DOSish
521 ports, but not in MacPerl, where only NBDM_File and DB_File are
524 The good news is that at least some DBM module should be available, and
525 AnyDBM_File will use whichever module it can find. Of course, then
526 the code needs to be fairly strict, dropping to the greatest common
527 factor (e.g., not exceeding 1K for each record), so that it will
528 work with any DBM module. See L<AnyDBM_File> for more details.
532 The system's notion of time of day and calendar date is controlled in
533 widely different ways. Don't assume the timezone is stored in C<$ENV{TZ}>,
534 and even if it is, don't assume that you can control the timezone through
537 Don't assume that the epoch starts at 00:00:00, January 1, 1970,
538 because that is OS- and implementation-specific. It is better to store a date
539 in an unambiguous representation. The ISO-8601 standard defines
540 "YYYY-MM-DD" as the date format. A text representation (like "1987-12-18")
541 can be easily converted into an OS-specific value using a module like
542 Date::Parse. An array of values, such as those returned by
543 C<localtime>, can be converted to an OS-specific representation using
546 When calculating specific times, such as for tests in time or date modules,
547 it may be appropriate to calculate an offset for the epoch.
550 $offset = Time::Local::timegm(0, 0, 0, 1, 0, 70);
552 The value for C<$offset> in Unix will be C<0>, but in Mac OS will be
553 some large number. C<$offset> can then be added to a Unix time value
554 to get what should be the proper value on any system.
556 =head2 Character sets and character encoding
558 Assume very little about character sets.
560 Assume nothing about numerical values (C<ord>, C<chr>) of characters.
561 Do not use explicit code point ranges (like \xHH-\xHH); use for
562 example symbolic character classes like C<[:print:]>.
564 Do not assume that the alphabetic characters are encoded contiguously
565 (in the numeric sense). There may be gaps.
567 Do not assume anything about the ordering of the characters.
568 The lowercase letters may come before or after the uppercase letters;
569 the lowercase and uppercase may be interlaced so that both `a' and `A'
570 come before `b'; the accented and other international characters may
571 be interlaced so that E<auml> comes before `b'.
573 =head2 Internationalisation
575 If you may assume POSIX (a rather large assumption), you may read
576 more about the POSIX locale system from L<perllocale>. The locale
577 system at least attempts to make things a little bit more portable,
578 or at least more convenient and native-friendly for non-English
579 users. The system affects character sets and encoding, and date
580 and time formatting--amongst other things.
582 =head2 System Resources
584 If your code is destined for systems with severely constrained (or
585 missing!) virtual memory systems then you want to be I<especially> mindful
586 of avoiding wasteful constructs such as:
588 # NOTE: this is no longer "bad" in perl5.005
589 for (0..10000000) {} # bad
590 for (my $x = 0; $x <= 10000000; ++$x) {} # good
592 @lines = <VERY_LARGE_FILE>; # bad
594 while (<FILE>) {$file .= $_} # sometimes bad
595 $file = join('', <FILE>); # better
597 The last two constructs may appear unintuitive to most people. The
598 first repeatedly grows a string, whereas the second allocates a
599 large chunk of memory in one go. On some systems, the second is
600 more efficient that the first.
604 Most multi-user platforms provide basic levels of security, usually
605 implemented at the filesystem level. Some, however, do
606 not-- unfortunately. Thus the notion of user id, or "home" directory,
607 or even the state of being logged-in, may be unrecognizable on many
608 platforms. If you write programs that are security-conscious, it
609 is usually best to know what type of system you will be running
610 under so that you can write code explicitly for that platform (or
613 Don't assume the UNIX filesystem access semantics: the operating
614 system or the filesystem may be using some ACL systems, which are
615 richer languages than the usual rwx. Even if the rwx exist,
616 their semantics might be different.
618 (From security viewpoint testing for permissions before attempting to
619 do something is silly anyway: if one tries this, there is potential
620 for race conditions-- someone or something might change the
621 permissions between the permissions check and the actual operation.
622 Just try the operation.)
624 Don't assume the UNIX user and group semantics: especially, don't
625 expect the C<< $< >> and C<< $> >> (or the C<$(> and C<$)>) to work
626 for switching identities (or memberships).
628 Don't assume set-uid and set-gid semantics. (And even if you do,
629 think twice: set-uid and set-gid are a known can of security worms.)
633 For those times when it is necessary to have platform-specific code,
634 consider keeping the platform-specific code in one place, making porting
635 to other platforms easier. Use the Config module and the special
636 variable C<$^O> to differentiate platforms, as described in
639 Be careful in the tests you supply with your module or programs.
640 Module code may be fully portable, but its tests might not be. This
641 often happens when tests spawn off other processes or call external
642 programs to aid in the testing, or when (as noted above) the tests
643 assume certain things about the filesystem and paths. Be careful
644 not to depend on a specific output style for errors, such as when
645 checking C<$!> after a system call. Some platforms expect a certain
646 output format, and perl on those platforms may have been adjusted
647 accordingly. Most specifically, don't anchor a regex when testing
652 Modules uploaded to CPAN are tested by a variety of volunteers on
653 different platforms. These CPAN testers are notified by mail of each
654 new upload, and reply to the list with PASS, FAIL, NA (not applicable to
655 this platform), or UNKNOWN (unknown), along with any relevant notations.
657 The purpose of the testing is twofold: one, to help developers fix any
658 problems in their code that crop up because of lack of testing on other
659 platforms; two, to provide users with information about whether
660 a given module works on a given platform.
664 =item Mailing list: cpan-testers@perl.org
666 =item Testing results: http://testers.cpan.org/
672 As of version 5.002, Perl is built with a C<$^O> variable that
673 indicates the operating system it was built on. This was implemented
674 to help speed up code that would otherwise have to C<use Config>
675 and use the value of C<$Config{osname}>. Of course, to get more
676 detailed information about the system, looking into C<%Config> is
677 certainly recommended.
679 C<%Config> cannot always be trusted, however, because it was built
680 at compile time. If perl was built in one place, then transferred
681 elsewhere, some values may be wrong. The values may even have been
682 edited after the fact.
686 Perl works on a bewildering variety of Unix and Unix-like platforms (see
687 e.g. most of the files in the F<hints/> directory in the source code kit).
688 On most of these systems, the value of C<$^O> (hence C<$Config{'osname'}>,
689 too) is determined either by lowercasing and stripping punctuation from the
690 first field of the string returned by typing C<uname -a> (or a similar command)
691 at the shell prompt or by testing the file system for the presence of
692 uniquely named files such as a kernel or header file. Here, for example,
693 are a few of the more popular Unix flavors:
695 uname $^O $Config{'archname'}
696 --------------------------------------------
698 BSD/OS bsdos i386-bsdos
700 dgux dgux AViiON-dgux
701 DYNIX/ptx dynixptx i386-dynixptx
702 FreeBSD freebsd freebsd-i386
703 Linux linux arm-linux
704 Linux linux i386-linux
705 Linux linux i586-linux
706 Linux linux ppc-linux
707 HP-UX hpux PA-RISC1.1
709 Mac OS X darwin darwin
710 MachTen PPC machten powerpc-machten
712 NeXT 4 next OPENSTEP-Mach
713 openbsd openbsd i386-openbsd
714 OSF1 dec_osf alpha-dec_osf
715 reliantunix-n svr4 RM400-svr4
716 SCO_SV sco_sv i386-sco_sv
717 SINIX-N svr4 RM400-svr4
718 sn4609 unicos CRAY_C90-unicos
719 sn6521 unicosmk t3e-unicosmk
720 sn9617 unicos CRAY_J90-unicos
721 SunOS solaris sun4-solaris
722 SunOS solaris i86pc-solaris
723 SunOS4 sunos sun4-sunos
725 Because the value of C<$Config{archname}> may depend on the
726 hardware architecture, it can vary more than the value of C<$^O>.
728 =head2 DOS and Derivatives
730 Perl has long been ported to Intel-style microcomputers running under
731 systems like PC-DOS, MS-DOS, OS/2, and most Windows platforms you can
732 bring yourself to mention (except for Windows CE, if you count that).
733 Users familiar with I<COMMAND.COM> or I<CMD.EXE> style shells should
734 be aware that each of these file specifications may have subtle
737 $filespec0 = "c:/foo/bar/file.txt";
738 $filespec1 = "c:\\foo\\bar\\file.txt";
739 $filespec2 = 'c:\foo\bar\file.txt';
740 $filespec3 = 'c:\\foo\\bar\\file.txt';
742 System calls accept either C</> or C<\> as the path separator.
743 However, many command-line utilities of DOS vintage treat C</> as
744 the option prefix, so may get confused by filenames containing C</>.
745 Aside from calling any external programs, C</> will work just fine,
746 and probably better, as it is more consistent with popular usage,
747 and avoids the problem of remembering what to backwhack and what
750 The DOS FAT filesystem can accommodate only "8.3" style filenames. Under
751 the "case-insensitive, but case-preserving" HPFS (OS/2) and NTFS (NT)
752 filesystems you may have to be careful about case returned with functions
753 like C<readdir> or used with functions like C<open> or C<opendir>.
755 DOS also treats several filenames as special, such as AUX, PRN,
756 NUL, CON, COM1, LPT1, LPT2, etc. Unfortunately, sometimes these
757 filenames won't even work if you include an explicit directory
758 prefix. It is best to avoid such filenames, if you want your code
759 to be portable to DOS and its derivatives. It's hard to know what
760 these all are, unfortunately.
762 Users of these operating systems may also wish to make use of
763 scripts such as I<pl2bat.bat> or I<pl2cmd> to
764 put wrappers around your scripts.
766 Newline (C<\n>) is translated as C<\015\012> by STDIO when reading from
767 and writing to files (see L<"Newlines">). C<binmode(FILEHANDLE)>
768 will keep C<\n> translated as C<\012> for that filehandle. Since it is a
769 no-op on other systems, C<binmode> should be used for cross-platform code
770 that deals with binary data. That's assuming you realize in advance
771 that your data is in binary. General-purpose programs should
772 often assume nothing about their data.
774 The C<$^O> variable and the C<$Config{archname}> values for various
775 DOSish perls are as follows:
777 OS $^O $Config{archname} ID Version
778 --------------------------------------------------------
782 Windows 3.1 ? ? 0 3 01
783 Windows 95 MSWin32 MSWin32-x86 1 4 00
784 Windows 98 MSWin32 MSWin32-x86 1 4 10
785 Windows ME MSWin32 MSWin32-x86 1 ?
786 Windows NT MSWin32 MSWin32-x86 2 4 xx
787 Windows NT MSWin32 MSWin32-ALPHA 2 4 xx
788 Windows NT MSWin32 MSWin32-ppc 2 4 xx
789 Windows 2000 MSWin32 MSWin32-x86 2 5 xx
790 Windows XP MSWin32 MSWin32-x86 2 ?
791 Windows CE MSWin32 ? 3
794 The various MSWin32 Perl's can distinguish the OS they are running on
795 via the value of the fifth element of the list returned from
796 Win32::GetOSVersion(). For example:
798 if ($^O eq 'MSWin32') {
799 my @os_version_info = Win32::GetOSVersion();
800 print +('3.1','95','NT')[$os_version_info[4]],"\n";
809 The djgpp environment for DOS, http://www.delorie.com/djgpp/
814 The EMX environment for DOS, OS/2, etc. emx@iaehv.nl,
815 http://www.leo.org/pub/comp/os/os2/leo/gnu/emx+gcc/index.html or
816 ftp://hobbes.nmsu.edu/pub/os2/dev/emx/ Also L<perlos2>.
820 Build instructions for Win32 in L<perlwin32>, or under the Cygnus environment
825 The C<Win32::*> modules in L<Win32>.
829 The ActiveState Pages, http://www.activestate.com/
833 The Cygwin environment for Win32; F<README.cygwin> (installed
834 as L<perlcygwin>), http://www.cygwin.com/
838 The U/WIN environment for Win32,
839 http://www.research.att.com/sw/tools/uwin/
843 Build instructions for OS/2, L<perlos2>
849 Any module requiring XS compilation is right out for most people, because
850 MacPerl is built using non-free (and non-cheap!) compilers. Some XS
851 modules that can work with MacPerl are built and distributed in binary
854 Directories are specified as:
856 volume:folder:file for absolute pathnames
857 volume:folder: for absolute pathnames
858 :folder:file for relative pathnames
859 :folder: for relative pathnames
860 :file for relative pathnames
861 file for relative pathnames
863 Files are stored in the directory in alphabetical order. Filenames are
864 limited to 31 characters, and may include any character except for
865 null and C<:>, which is reserved as the path separator.
867 Instead of C<flock>, see C<FSpSetFLock> and C<FSpRstFLock> in the
868 Mac::Files module, or C<chmod(0444, ...)> and C<chmod(0666, ...)>.
870 In the MacPerl application, you can't run a program from the command line;
871 programs that expect C<@ARGV> to be populated can be edited with something
872 like the following, which brings up a dialog box asking for the command
876 @ARGV = split /\s+/, MacPerl::Ask('Arguments?');
879 A MacPerl script saved as a "droplet" will populate C<@ARGV> with the full
880 pathnames of the files dropped onto the script.
882 Mac users can run programs under a type of command line interface
883 under MPW (Macintosh Programmer's Workshop, a free development
884 environment from Apple). MacPerl was first introduced as an MPW
885 tool, and MPW can be used like a shell:
887 perl myscript.plx some arguments
889 ToolServer is another app from Apple that provides access to MPW tools
890 from MPW and the MacPerl app, which allows MacPerl programs to use
891 C<system>, backticks, and piped C<open>.
893 "S<Mac OS>" is the proper name for the operating system, but the value
894 in C<$^O> is "MacOS". To determine architecture, version, or whether
895 the application or MPW tool version is running, check:
897 $is_app = $MacPerl::Version =~ /App/;
898 $is_tool = $MacPerl::Version =~ /MPW/;
899 ($version) = $MacPerl::Version =~ /^(\S+)/;
900 $is_ppc = $MacPerl::Architecture eq 'MacPPC';
901 $is_68k = $MacPerl::Architecture eq 'Mac68K';
903 S<Mac OS X>, based on NeXT's OpenStep OS, runs MacPerl natively, under the
904 "Classic" environment. There is no "Carbon" version of MacPerl to run
905 under the primary Mac OS X environment. S<Mac OS X> and its Open Source
906 version, Darwin, both run Unix perl natively.
914 MacPerl Development, http://dev.macperl.org/ .
918 The MacPerl Pages, http://www.macperl.com/ .
922 The MacPerl mailing lists, http://lists.perl.org/ .
928 Perl on VMS is discussed in L<perlvms> in the perl distribution.
929 Perl on VMS can accept either VMS- or Unix-style file
930 specifications as in either of the following:
932 $ perl -ne "print if /perl_setup/i" SYS$LOGIN:LOGIN.COM
933 $ perl -ne "print if /perl_setup/i" /sys$login/login.com
935 but not a mixture of both as in:
937 $ perl -ne "print if /perl_setup/i" sys$login:/login.com
938 Can't open sys$login:/login.com: file specification syntax error
940 Interacting with Perl from the Digital Command Language (DCL) shell
941 often requires a different set of quotation marks than Unix shells do.
944 $ perl -e "print ""Hello, world.\n"""
947 There are several ways to wrap your perl scripts in DCL F<.COM> files, if
948 you are so inclined. For example:
950 $ write sys$output "Hello from DCL!"
952 $ then perl -x 'f$environment("PROCEDURE")
953 $ else perl -x - 'p1 'p2 'p3 'p4 'p5 'p6 'p7 'p8
954 $ deck/dollars="__END__"
957 print "Hello from Perl!\n";
962 Do take care with C<$ ASSIGN/nolog/user SYS$COMMAND: SYS$INPUT> if your
963 perl-in-DCL script expects to do things like C<< $read = <STDIN>; >>.
965 Filenames are in the format "name.extension;version". The maximum
966 length for filenames is 39 characters, and the maximum length for
967 extensions is also 39 characters. Version is a number from 1 to
968 32767. Valid characters are C</[A-Z0-9$_-]/>.
970 VMS's RMS filesystem is case-insensitive and does not preserve case.
971 C<readdir> returns lowercased filenames, but specifying a file for
972 opening remains case-insensitive. Files without extensions have a
973 trailing period on them, so doing a C<readdir> with a file named F<A.;5>
974 will return F<a.> (though that file could be opened with
977 RMS had an eight level limit on directory depths from any rooted logical
978 (allowing 16 levels overall) prior to VMS 7.2. Hence
979 C<PERL_ROOT:[LIB.2.3.4.5.6.7.8]> is a valid directory specification but
980 C<PERL_ROOT:[LIB.2.3.4.5.6.7.8.9]> is not. F<Makefile.PL> authors might
981 have to take this into account, but at least they can refer to the former
982 as C</PERL_ROOT/lib/2/3/4/5/6/7/8/>.
984 The VMS::Filespec module, which gets installed as part of the build
985 process on VMS, is a pure Perl module that can easily be installed on
986 non-VMS platforms and can be helpful for conversions to and from RMS
989 What C<\n> represents depends on the type of file opened. It usually
990 represents C<\012> but it could also be C<\015>, C<\012>, C<\015\012>,
991 C<\000>, C<\040>, or nothing depending on the file organiztion and
992 record format. The VMS::Stdio module provides access to the
993 special fopen() requirements of files with unusual attributes on VMS.
995 TCP/IP stacks are optional on VMS, so socket routines might not be
996 implemented. UDP sockets may not be supported.
998 The value of C<$^O> on OpenVMS is "VMS". To determine the architecture
999 that you are running on without resorting to loading all of C<%Config>
1000 you can examine the content of the C<@INC> array like so:
1002 if (grep(/VMS_AXP/, @INC)) {
1003 print "I'm on Alpha!\n";
1005 } elsif (grep(/VMS_VAX/, @INC)) {
1006 print "I'm on VAX!\n";
1009 print "I'm not so sure about where $^O is...\n";
1012 On VMS, perl determines the UTC offset from the C<SYS$TIMEZONE_DIFFERENTIAL>
1013 logical name. Although the VMS epoch began at 17-NOV-1858 00:00:00.00,
1014 calls to C<localtime> are adjusted to count offsets from
1015 01-JAN-1970 00:00:00.00, just like Unix.
1023 F<README.vms> (installed as L<README_vms>), L<perlvms>
1027 vmsperl list, majordomo@perl.org
1029 (Put the words C<subscribe vmsperl> in message body.)
1033 vmsperl on the web, http://www.sidhe.org/vmsperl/index.html
1039 Perl on VOS is discussed in F<README.vos> in the perl distribution
1040 (installed as L<perlvos>). Perl on VOS can accept either VOS- or
1041 Unix-style file specifications as in either of the following:
1043 C<< $ perl -ne "print if /perl_setup/i" >system>notices >>
1044 C<< $ perl -ne "print if /perl_setup/i" /system/notices >>
1046 or even a mixture of both as in:
1048 C<< $ perl -ne "print if /perl_setup/i" >system/notices >>
1050 Even though VOS allows the slash character to appear in object
1051 names, because the VOS port of Perl interprets it as a pathname
1052 delimiting character, VOS files, directories, or links whose names
1053 contain a slash character cannot be processed. Such files must be
1054 renamed before they can be processed by Perl. Note that VOS limits
1055 file names to 32 or fewer characters.
1057 Perl on VOS can be built using two different compilers and two different
1058 versions of the POSIX runtime. The recommended method for building full
1059 Perl is with the GNU C compiler and the generally-available version of
1060 VOS POSIX support. See F<README.vos> (installed as L<perlvos>) for
1061 restrictions that apply when Perl is built using the VOS Standard C
1062 compiler or the alpha version of VOS POSIX support.
1064 The value of C<$^O> on VOS is "VOS". To determine the architecture that
1065 you are running on without resorting to loading all of C<%Config> you
1066 can examine the content of the @INC array like so:
1069 print "I'm on a Stratus box!\n";
1071 print "I'm not on a Stratus box!\n";
1075 if (grep(/860/, @INC)) {
1076 print "This box is a Stratus XA/R!\n";
1078 } elsif (grep(/7100/, @INC)) {
1079 print "This box is a Stratus HP 7100 or 8xxx!\n";
1081 } elsif (grep(/8000/, @INC)) {
1082 print "This box is a Stratus HP 8xxx!\n";
1085 print "This box is a Stratus 68K!\n";
1094 F<README.vos> (installed as L<perlvos>)
1098 The VOS mailing list.
1100 There is no specific mailing list for Perl on VOS. You can post
1101 comments to the comp.sys.stratus newsgroup, or subscribe to the general
1102 Stratus mailing list. Send a letter with "subscribe Info-Stratus" in
1103 the message body to majordomo@list.stratagy.com.
1107 VOS Perl on the web at http://ftp.stratus.com/pub/vos/posix/posix.html
1111 =head2 EBCDIC Platforms
1113 Recent versions of Perl have been ported to platforms such as OS/400 on
1114 AS/400 minicomputers as well as OS/390, VM/ESA, and BS2000 for S/390
1115 Mainframes. Such computers use EBCDIC character sets internally (usually
1116 Character Code Set ID 0037 for OS/400 and either 1047 or POSIX-BC for S/390
1117 systems). On the mainframe perl currently works under the "Unix system
1118 services for OS/390" (formerly known as OpenEdition), VM/ESA OpenEdition, or
1119 the BS200 POSIX-BC system (BS2000 is supported in perl 5.6 and greater).
1120 See L<perlos390> for details.
1122 As of R2.5 of USS for OS/390 and Version 2.3 of VM/ESA these Unix
1123 sub-systems do not support the C<#!> shebang trick for script invocation.
1124 Hence, on OS/390 and VM/ESA perl scripts can be executed with a header
1125 similar to the following simple script:
1128 eval 'exec /usr/local/bin/perl -S $0 ${1+"$@"}'
1130 #!/usr/local/bin/perl # just a comment really
1132 print "Hello from perl!\n";
1134 OS/390 will support the C<#!> shebang trick in release 2.8 and beyond.
1135 Calls to C<system> and backticks can use POSIX shell syntax on all
1138 On the AS/400, if PERL5 is in your library list, you may need
1139 to wrap your perl scripts in a CL procedure to invoke them like so:
1142 CALL PGM(PERL5/PERL) PARM('/QOpenSys/hello.pl')
1145 This will invoke the perl script F<hello.pl> in the root of the
1146 QOpenSys file system. On the AS/400 calls to C<system> or backticks
1149 On these platforms, bear in mind that the EBCDIC character set may have
1150 an effect on what happens with some perl functions (such as C<chr>,
1151 C<pack>, C<print>, C<printf>, C<ord>, C<sort>, C<sprintf>, C<unpack>), as
1152 well as bit-fiddling with ASCII constants using operators like C<^>, C<&>
1153 and C<|>, not to mention dealing with socket interfaces to ASCII computers
1154 (see L<"Newlines">).
1156 Fortunately, most web servers for the mainframe will correctly
1157 translate the C<\n> in the following statement to its ASCII equivalent
1158 (C<\r> is the same under both Unix and OS/390 & VM/ESA):
1160 print "Content-type: text/html\r\n\r\n";
1162 The values of C<$^O> on some of these platforms includes:
1164 uname $^O $Config{'archname'}
1165 --------------------------------------------
1168 POSIX-BC posix-bc BS2000-posix-bc
1171 Some simple tricks for determining if you are running on an EBCDIC
1172 platform could include any of the following (perhaps all):
1174 if ("\t" eq "\05") { print "EBCDIC may be spoken here!\n"; }
1176 if (ord('A') == 193) { print "EBCDIC may be spoken here!\n"; }
1178 if (chr(169) eq 'z') { print "EBCDIC may be spoken here!\n"; }
1180 One thing you may not want to rely on is the EBCDIC encoding
1181 of punctuation characters since these may differ from code page to code
1182 page (and once your module or script is rumoured to work with EBCDIC,
1183 folks will want it to work with all EBCDIC character sets).
1193 L<perlos390>, F<README.os390>, F<perlbs2000>, F<README.vmesa>,
1198 The perl-mvs@perl.org list is for discussion of porting issues as well as
1199 general usage issues for all EBCDIC Perls. Send a message body of
1200 "subscribe perl-mvs" to majordomo@perl.org.
1204 AS/400 Perl information at
1205 http://as400.rochester.ibm.com/
1206 as well as on CPAN in the F<ports/> directory.
1210 =head2 Acorn RISC OS
1212 Because Acorns use ASCII with newlines (C<\n>) in text files as C<\012> like
1213 Unix, and because Unix filename emulation is turned on by default,
1214 most simple scripts will probably work "out of the box". The native
1215 filesystem is modular, and individual filesystems are free to be
1216 case-sensitive or insensitive, and are usually case-preserving. Some
1217 native filesystems have name length limits, which file and directory
1218 names are silently truncated to fit. Scripts should be aware that the
1219 standard filesystem currently has a name length limit of B<10>
1220 characters, with up to 77 items in a directory, but other filesystems
1221 may not impose such limitations.
1223 Native filenames are of the form
1225 Filesystem#Special_Field::DiskName.$.Directory.Directory.File
1229 Special_Field is not usually present, but may contain . and $ .
1230 Filesystem =~ m|[A-Za-z0-9_]|
1231 DsicName =~ m|[A-Za-z0-9_/]|
1232 $ represents the root directory
1233 . is the path separator
1234 @ is the current directory (per filesystem but machine global)
1235 ^ is the parent directory
1236 Directory and File =~ m|[^\0- "\.\$\%\&:\@\\^\|\177]+|
1238 The default filename translation is roughly C<tr|/.|./|;>
1240 Note that C<"ADFS::HardDisk.$.File" ne 'ADFS::HardDisk.$.File'> and that
1241 the second stage of C<$> interpolation in regular expressions will fall
1242 foul of the C<$.> if scripts are not careful.
1244 Logical paths specified by system variables containing comma-separated
1245 search lists are also allowed; hence C<System:Modules> is a valid
1246 filename, and the filesystem will prefix C<Modules> with each section of
1247 C<System$Path> until a name is made that points to an object on disk.
1248 Writing to a new file C<System:Modules> would be allowed only if
1249 C<System$Path> contains a single item list. The filesystem will also
1250 expand system variables in filenames if enclosed in angle brackets, so
1251 C<< <System$Dir>.Modules >> would look for the file
1252 S<C<$ENV{'System$Dir'} . 'Modules'>>. The obvious implication of this is
1253 that B<fully qualified filenames can start with C<< <> >>> and should
1254 be protected when C<open> is used for input.
1256 Because C<.> was in use as a directory separator and filenames could not
1257 be assumed to be unique after 10 characters, Acorn implemented the C
1258 compiler to strip the trailing C<.c> C<.h> C<.s> and C<.o> suffix from
1259 filenames specified in source code and store the respective files in
1260 subdirectories named after the suffix. Hence files are translated:
1263 C:foo.h C:h.foo (logical path variable)
1264 sys/os.h sys.h.os (C compiler groks Unix-speak)
1265 10charname.c c.10charname
1266 10charname.o o.10charname
1267 11charname_.c c.11charname (assuming filesystem truncates at 10)
1269 The Unix emulation library's translation of filenames to native assumes
1270 that this sort of translation is required, and it allows a user-defined list
1271 of known suffixes that it will transpose in this fashion. This may
1272 seem transparent, but consider that with these rules C<foo/bar/baz.h>
1273 and C<foo/bar/h/baz> both map to C<foo.bar.h.baz>, and that C<readdir> and
1274 C<glob> cannot and do not attempt to emulate the reverse mapping. Other
1275 C<.>'s in filenames are translated to C</>.
1277 As implied above, the environment accessed through C<%ENV> is global, and
1278 the convention is that program specific environment variables are of the
1279 form C<Program$Name>. Each filesystem maintains a current directory,
1280 and the current filesystem's current directory is the B<global> current
1281 directory. Consequently, sociable programs don't change the current
1282 directory but rely on full pathnames, and programs (and Makefiles) cannot
1283 assume that they can spawn a child process which can change the current
1284 directory without affecting its parent (and everyone else for that
1287 Because native operating system filehandles are global and are currently
1288 allocated down from 255, with 0 being a reserved value, the Unix emulation
1289 library emulates Unix filehandles. Consequently, you can't rely on
1290 passing C<STDIN>, C<STDOUT>, or C<STDERR> to your children.
1292 The desire of users to express filenames of the form
1293 C<< <Foo$Dir>.Bar >> on the command line unquoted causes problems,
1294 too: C<``> command output capture has to perform a guessing game. It
1295 assumes that a string C<< <[^<>]+\$[^<>]> >> is a
1296 reference to an environment variable, whereas anything else involving
1297 C<< < >> or C<< > >> is redirection, and generally manages to be 99%
1298 right. Of course, the problem remains that scripts cannot rely on any
1299 Unix tools being available, or that any tools found have Unix-like command
1302 Extensions and XS are, in theory, buildable by anyone using free
1303 tools. In practice, many don't, as users of the Acorn platform are
1304 used to binary distributions. MakeMaker does run, but no available
1305 make currently copes with MakeMaker's makefiles; even if and when
1306 this should be fixed, the lack of a Unix-like shell will cause
1307 problems with makefile rules, especially lines of the form C<cd
1308 sdbm && make all>, and anything using quoting.
1310 "S<RISC OS>" is the proper name for the operating system, but the value
1311 in C<$^O> is "riscos" (because we don't like shouting).
1315 Perl has been ported to many platforms that do not fit into any of
1316 the categories listed above. Some, such as AmigaOS, Atari MiNT,
1317 BeOS, HP MPE/iX, QNX, Plan 9, and VOS, have been well-integrated
1318 into the standard Perl source code kit. You may need to see the
1319 F<ports/> directory on CPAN for information, and possibly binaries,
1320 for the likes of: aos, Atari ST, lynxos, riscos, Novell Netware,
1321 Tandem Guardian, I<etc.> (Yes, we know that some of these OSes may
1322 fall under the Unix category, but we are not a standards body.)
1324 Some approximate operating system names and their C<$^O> values
1325 in the "OTHER" category include:
1327 OS $^O $Config{'archname'}
1328 ------------------------------------------
1329 Amiga DOS amigaos m68k-amigos
1330 MPE/iX mpeix PA-RISC1.1
1338 Amiga, F<README.amiga> (installed as L<perlamiga>).
1342 Atari, F<README.mint> and Guido Flohr's web page
1343 http://stud.uni-sb.de/~gufl0000/
1347 Be OS, F<README.beos>
1351 HP 300 MPE/iX, F<README.mpeix> and Mark Bixby's web page
1352 http://www.bixby.org/mark/perlix.html
1356 A free perl5-based PERL.NLM for Novell Netware is available in
1357 precompiled binary and source code form from http://www.novell.com/
1358 as well as from CPAN.
1362 Plan 9, F<README.plan9>
1366 =head1 FUNCTION IMPLEMENTATIONS
1368 Listed below are functions that are either completely unimplemented
1369 or else have been implemented differently on various platforms.
1370 Following each description will be, in parentheses, a list of
1371 platforms that the description applies to.
1373 The list may well be incomplete, or even wrong in some places. When
1374 in doubt, consult the platform-specific README files in the Perl
1375 source distribution, and any other documentation resources accompanying
1378 Be aware, moreover, that even among Unix-ish systems there are variations.
1380 For many functions, you can also query C<%Config>, exported by
1381 default from the Config module. For example, to check whether the
1382 platform has the C<lstat> call, check C<$Config{d_lstat}>. See
1383 L<Config> for a full description of available variables.
1385 =head2 Alphabetical Listing of Perl Functions
1395 C<-r>, C<-w>, and C<-x> have a limited meaning only; directories
1396 and applications are executable, and there are no uid/gid
1397 considerations. C<-o> is not supported. (S<Mac OS>)
1399 C<-r>, C<-w>, C<-x>, and C<-o> tell whether the file is accessible,
1400 which may not reflect UIC-based file protections. (VMS)
1402 C<-s> returns the size of the data fork, not the total size of data fork
1403 plus resource fork. (S<Mac OS>).
1405 C<-s> by name on an open file will return the space reserved on disk,
1406 rather than the current extent. C<-s> on an open filehandle returns the
1407 current size. (S<RISC OS>)
1409 C<-R>, C<-W>, C<-X>, C<-O> are indistinguishable from C<-r>, C<-w>,
1410 C<-x>, C<-o>. (S<Mac OS>, Win32, VMS, S<RISC OS>)
1412 C<-b>, C<-c>, C<-k>, C<-g>, C<-p>, C<-u>, C<-A> are not implemented.
1415 C<-g>, C<-k>, C<-l>, C<-p>, C<-u>, C<-A> are not particularly meaningful.
1416 (Win32, VMS, S<RISC OS>)
1418 C<-d> is true if passed a device spec without an explicit directory.
1421 C<-T> and C<-B> are implemented, but might misclassify Mac text files
1422 with foreign characters; this is the case will all platforms, but may
1423 affect S<Mac OS> often. (S<Mac OS>)
1425 C<-x> (or C<-X>) determine if a file ends in one of the executable
1426 suffixes. C<-S> is meaningless. (Win32)
1428 C<-x> (or C<-X>) determine if a file has an executable file type.
1435 Not implemented. (Win32)
1437 =item binmode FILEHANDLE
1439 Meaningless. (S<Mac OS>, S<RISC OS>)
1441 Reopens file and restores pointer; if function fails, underlying
1442 filehandle may be closed, or pointer may be in a different position.
1445 The value returned by C<tell> may be affected after the call, and
1446 the filehandle may be flushed. (Win32)
1450 Only limited meaning. Disabling/enabling write permission is mapped to
1451 locking/unlocking the file. (S<Mac OS>)
1453 Only good for changing "owner" read-write access, "group", and "other"
1454 bits are meaningless. (Win32)
1456 Only good for changing "owner" and "other" read-write access. (S<RISC OS>)
1458 Access permissions are mapped onto VOS access-control list changes. (VOS)
1460 The actual permissions set depend on the value of the C<CYGWIN>
1461 in the SYSTEM environment settings. (Cygwin)
1465 Not implemented. (S<Mac OS>, Win32, Plan9, S<RISC OS>, VOS)
1467 Does nothing, but won't fail. (Win32)
1469 =item chroot FILENAME
1473 Not implemented. (S<Mac OS>, Win32, VMS, Plan9, S<RISC OS>, VOS, VM/ESA)
1475 =item crypt PLAINTEXT,SALT
1477 May not be available if library or source was not provided when building
1480 Not implemented. (VOS)
1484 Not implemented. (VMS, Plan9, VOS)
1486 =item dbmopen HASH,DBNAME,MODE
1488 Not implemented. (VMS, Plan9, VOS)
1492 Not useful. (S<Mac OS>, S<RISC OS>)
1494 Not implemented. (Win32)
1496 Invokes VMS debugger. (VMS)
1500 Not implemented. (S<Mac OS>)
1502 Implemented via Spawn. (VM/ESA)
1504 Does not automatically flush output handles on some platforms.
1505 (SunOS, Solaris, HP-UX)
1511 Emulates UNIX exit() (which considers C<exit 1> to indicate an error) by
1512 mapping the C<1> to SS$_ABORT (C<44>). This behavior may be overridden
1513 with the pragma C<use vmsish 'exit'>. As with the CRTL's exit()
1514 function, C<exit 0> is also mapped to an exit status of SS$_NORMAL
1515 (C<1>); this mapping cannot be overridden. Any other argument to exit()
1516 is used directly as Perl's exit status. (VMS)
1518 =item fcntl FILEHANDLE,FUNCTION,SCALAR
1520 Not implemented. (Win32, VMS)
1522 =item flock FILEHANDLE,OPERATION
1524 Not implemented (S<Mac OS>, VMS, S<RISC OS>, VOS).
1526 Available only on Windows NT (not on Windows 95). (Win32)
1530 Not implemented. (S<Mac OS>, AmigaOS, S<RISC OS>, VOS, VM/ESA)
1532 Emulated using multiple interpreters. See L<perlfork>. (Win32)
1534 Does not automatically flush output handles on some platforms.
1535 (SunOS, Solaris, HP-UX)
1539 Not implemented. (S<Mac OS>, S<RISC OS>)
1543 Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>, VOS)
1547 Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>)
1549 =item getpriority WHICH,WHO
1551 Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>, VOS, VM/ESA)
1555 Not implemented. (S<Mac OS>, Win32)
1557 Not useful. (S<RISC OS>)
1561 Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>)
1563 =item getnetbyname NAME
1565 Not implemented. (S<Mac OS>, Win32, Plan9)
1569 Not implemented. (S<Mac OS>, Win32)
1571 Not useful. (S<RISC OS>)
1575 Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>)
1577 =item getnetbyaddr ADDR,ADDRTYPE
1579 Not implemented. (S<Mac OS>, Win32, Plan9)
1581 =item getprotobynumber NUMBER
1583 Not implemented. (S<Mac OS>)
1585 =item getservbyport PORT,PROTO
1587 Not implemented. (S<Mac OS>)
1591 Not implemented. (S<Mac OS>, Win32, VM/ESA)
1595 Not implemented. (S<Mac OS>, Win32, VMS, VM/ESA)
1599 Not implemented. (S<Mac OS>, Win32)
1603 Not implemented. (S<Mac OS>, Win32, Plan9)
1607 Not implemented. (S<Mac OS>, Win32, Plan9)
1611 Not implemented. (Win32, Plan9)
1613 =item sethostent STAYOPEN
1615 Not implemented. (S<Mac OS>, Win32, Plan9, S<RISC OS>)
1617 =item setnetent STAYOPEN
1619 Not implemented. (S<Mac OS>, Win32, Plan9, S<RISC OS>)
1621 =item setprotoent STAYOPEN
1623 Not implemented. (S<Mac OS>, Win32, Plan9, S<RISC OS>)
1625 =item setservent STAYOPEN
1627 Not implemented. (Plan9, Win32, S<RISC OS>)
1631 Not implemented. (S<Mac OS>, MPE/iX, VM/ESA, Win32)
1635 Not implemented. (S<Mac OS>, MPE/iX, S<RISC OS>, VM/ESA, VMS, Win32)
1639 Not implemented. (S<Mac OS>, Win32)
1643 Not implemented. (S<Mac OS>, Win32, Plan9)
1647 Not implemented. (S<Mac OS>, Win32, Plan9)
1651 Not implemented. (Plan9, Win32)
1653 =item getsockopt SOCKET,LEVEL,OPTNAME
1655 Not implemented. (Plan9)
1661 This operator is implemented via the File::Glob extension on most
1662 platforms. See L<File::Glob> for portability information.
1664 =item ioctl FILEHANDLE,FUNCTION,SCALAR
1666 Not implemented. (VMS)
1668 Available only for socket handles, and it does what the ioctlsocket() call
1669 in the Winsock API does. (Win32)
1671 Available only for socket handles. (S<RISC OS>)
1673 =item kill SIGNAL, LIST
1675 C<kill(0, LIST)> is implemented for the sake of taint checking;
1676 use with other signals is unimplemented. (S<Mac OS>)
1678 Not implemented, hence not useful for taint checking. (S<RISC OS>)
1680 C<kill()> doesn't have the semantics of C<raise()>, i.e. it doesn't send
1681 a signal to the identified process like it does on Unix platforms.
1682 Instead C<kill($sig, $pid)> terminates the process identified by $pid,
1683 and makes it exit immediately with exit status $sig. As in Unix, if
1684 $sig is 0 and the specified process exists, it returns true without
1685 actually terminating it. (Win32)
1687 =item link OLDFILE,NEWFILE
1689 Not implemented. (S<Mac OS>, MPE/iX, VMS, S<RISC OS>)
1691 Link count not updated because hard links are not quite that hard
1692 (They are sort of half-way between hard and soft links). (AmigaOS)
1694 Hard links are implemented on Win32 (Windows NT and Windows 2000)
1697 =item lstat FILEHANDLE
1703 Not implemented. (VMS, S<RISC OS>)
1705 Return values (especially for device and inode) may be bogus. (Win32)
1707 =item msgctl ID,CMD,ARG
1709 =item msgget KEY,FLAGS
1711 =item msgsnd ID,MSG,FLAGS
1713 =item msgrcv ID,VAR,SIZE,TYPE,FLAGS
1715 Not implemented. (S<Mac OS>, Win32, VMS, Plan9, S<RISC OS>, VOS)
1717 =item open FILEHANDLE,EXPR
1719 =item open FILEHANDLE
1721 The C<|> variants are supported only if ToolServer is installed.
1724 open to C<|-> and C<-|> are unsupported. (S<Mac OS>, Win32, S<RISC OS>)
1726 Opening a process does not automatically flush output handles on some
1727 platforms. (SunOS, Solaris, HP-UX)
1729 =item pipe READHANDLE,WRITEHANDLE
1731 Very limited functionality. (MiNT)
1737 Not implemented. (Win32, VMS, S<RISC OS>)
1739 =item select RBITS,WBITS,EBITS,TIMEOUT
1741 Only implemented on sockets. (Win32, VMS)
1743 Only reliable on sockets. (S<RISC OS>)
1745 Note that the C<select FILEHANDLE> form is generally portable.
1747 =item semctl ID,SEMNUM,CMD,ARG
1749 =item semget KEY,NSEMS,FLAGS
1751 =item semop KEY,OPSTRING
1753 Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>, VOS)
1757 Not implemented. (S<Mac OS>, MPE/iX, VMS, Win32, VMS, S<RISC OS>)
1759 =item setpgrp PID,PGRP
1761 Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>, VOS)
1763 =item setpriority WHICH,WHO,PRIORITY
1765 Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>, VOS)
1769 Not implemented. (S<Mac OS>, MPE/iX, Win32, S<RISC OS>)
1771 =item setsockopt SOCKET,LEVEL,OPTNAME,OPTVAL
1773 Not implemented. (Plan9)
1775 =item shmctl ID,CMD,ARG
1777 =item shmget KEY,SIZE,FLAGS
1779 =item shmread ID,VAR,POS,SIZE
1781 =item shmwrite ID,STRING,POS,SIZE
1783 Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>, VOS)
1785 =item sockatmark SOCKET
1787 A relatively recent addition to socket functions, may not
1788 be implemented even in UNIX platforms.
1790 =item socketpair SOCKET1,SOCKET2,DOMAIN,TYPE,PROTOCOL
1792 Not implemented. (Win32, VMS, S<RISC OS>, VOS, VM/ESA)
1794 =item stat FILEHANDLE
1800 Platforms that do not have rdev, blksize, or blocks will return these
1801 as '', so numeric comparison or manipulation of these fields may cause
1802 'not numeric' warnings.
1804 mtime and atime are the same thing, and ctime is creation time instead of
1805 inode change time. (S<Mac OS>).
1807 ctime not supported on UFS (S<Mac OS X>).
1809 ctime is creation time instead of inode change time (Win32).
1811 device and inode are not meaningful. (Win32)
1813 device and inode are not necessarily reliable. (VMS)
1815 mtime, atime and ctime all return the last modification time. Device and
1816 inode are not necessarily reliable. (S<RISC OS>)
1818 dev, rdev, blksize, and blocks are not available. inode is not
1819 meaningful and will differ between stat calls on the same file. (os2)
1821 some versions of cygwin when doing a stat("foo") and if not finding it
1822 may then attempt to stat("foo.exe") (Cygwin)
1824 =item symlink OLDFILE,NEWFILE
1826 Not implemented. (Win32, VMS, S<RISC OS>)
1830 Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>, VOS, VM/ESA)
1832 =item sysopen FILEHANDLE,FILENAME,MODE,PERMS
1834 The traditional "0", "1", and "2" MODEs are implemented with different
1835 numeric values on some systems. The flags exported by C<Fcntl>
1836 (O_RDONLY, O_WRONLY, O_RDWR) should work everywhere though. (S<Mac
1837 OS>, OS/390, VM/ESA)
1841 In general, do not assume the UNIX/POSIX semantics that you can shift
1842 C<$?> right by eight to get the exit value, or that C<$? & 127>
1843 would give you the number of the signal that terminated the program,
1844 or that C<$? & 128> would test true if the program was terminated by a
1845 coredump. Instead, use the POSIX W*() interfaces: for example, use
1846 WIFEXITED($?) and WEXITVALUE($?) to test for a normal exit and the exit
1847 value, WIFSIGNALED($?) and WTERMSIG($?) for a signal exit and the
1848 signal. Core dumping is not a portable concept, so there's no portable
1849 way to test for that.
1851 Only implemented if ToolServer is installed. (S<Mac OS>)
1853 As an optimization, may not call the command shell specified in
1854 C<$ENV{PERL5SHELL}>. C<system(1, @args)> spawns an external
1855 process and immediately returns its process designator, without
1856 waiting for it to terminate. Return value may be used subsequently
1857 in C<wait> or C<waitpid>. Failure to spawn() a subprocess is indicated
1858 by setting $? to "255 << 8". C<$?> is set in a way compatible with
1859 Unix (i.e. the exitstatus of the subprocess is obtained by "$? >> 8",
1860 as described in the documentation). (Win32)
1862 There is no shell to process metacharacters, and the native standard is
1863 to pass a command line terminated by "\n" "\r" or "\0" to the spawned
1864 program. Redirection such as C<< > foo >> is performed (if at all) by
1865 the run time library of the spawned program. C<system> I<list> will call
1866 the Unix emulation library's C<exec> emulation, which attempts to provide
1867 emulation of the stdin, stdout, stderr in force in the parent, providing
1868 the child program uses a compatible version of the emulation library.
1869 I<scalar> will call the native command line direct and no such emulation
1870 of a child Unix program will exists. Mileage B<will> vary. (S<RISC OS>)
1872 Far from being POSIX compliant. Because there may be no underlying
1873 /bin/sh tries to work around the problem by forking and execing the
1874 first token in its argument string. Handles basic redirection
1875 ("<" or ">") on its own behalf. (MiNT)
1877 Does not automatically flush output handles on some platforms.
1878 (SunOS, Solaris, HP-UX)
1880 The return value is POSIX-like (shifted up by 8 bits), which only allows
1881 room for a made-up value derived from the severity bits of the native
1882 32-bit condition code (unless overridden by C<use vmsish 'status'>).
1883 For more details see L<perlvms/$?>. (VMS)
1887 Only the first entry returned is nonzero. (S<Mac OS>)
1889 "cumulative" times will be bogus. On anything other than Windows NT
1890 or Windows 2000, "system" time will be bogus, and "user" time is
1891 actually the time returned by the clock() function in the C runtime
1894 Not useful. (S<RISC OS>)
1896 =item truncate FILEHANDLE,LENGTH
1898 =item truncate EXPR,LENGTH
1900 Not implemented. (Older versions of VMS)
1902 Truncation to zero-length only. (VOS)
1904 If a FILEHANDLE is supplied, it must be writable and opened in append
1905 mode (i.e., use C<<< open(FH, '>>filename') >>>
1906 or C<sysopen(FH,...,O_APPEND|O_RDWR)>. If a filename is supplied, it
1907 should not be held open elsewhere. (Win32)
1913 Returns undef where unavailable, as of version 5.005.
1915 C<umask> works but the correct permissions are set only when the file
1916 is finally closed. (AmigaOS)
1920 Only the modification time is updated. (S<BeOS>, S<Mac OS>, VMS, S<RISC OS>)
1922 May not behave as expected. Behavior depends on the C runtime
1923 library's implementation of utime(), and the filesystem being
1924 used. The FAT filesystem typically does not support an "access
1925 time" field, and it may limit timestamps to a granularity of
1926 two seconds. (Win32)
1930 =item waitpid PID,FLAGS
1932 Not implemented. (S<Mac OS>, VOS)
1934 Can only be applied to process handles returned for processes spawned
1935 using C<system(1, ...)> or pseudo processes created with C<fork()>. (Win32)
1937 Not useful. (S<RISC OS>)
1945 =item v1.48, 02 February 2001
1947 Various updates from perl5-porters over the past year, supported
1948 platforms update from Jarkko Hietaniemi.
1950 =item v1.47, 22 March 2000
1952 Various cleanups from Tom Christiansen, including migration of
1953 long platform listings from L<perl>.
1955 =item v1.46, 12 February 2000
1957 Updates for VOS and MPE/iX. (Peter Prymmer) Other small changes.
1959 =item v1.45, 20 December 1999
1961 Small changes from 5.005_63 distribution, more changes to EBCDIC info.
1963 =item v1.44, 19 July 1999
1965 A bunch of updates from Peter Prymmer for C<$^O> values,
1966 endianness, File::Spec, VMS, BS2000, OS/400.
1968 =item v1.43, 24 May 1999
1970 Added a lot of cleaning up from Tom Christiansen.
1972 =item v1.42, 22 May 1999
1974 Added notes about tests, sprintf/printf, and epoch offsets.
1976 =item v1.41, 19 May 1999
1978 Lots more little changes to formatting and content.
1980 Added a bunch of C<$^O> and related values
1981 for various platforms; fixed mail and web addresses, and added
1982 and changed miscellaneous notes. (Peter Prymmer)
1984 =item v1.40, 11 April 1999
1986 Miscellaneous changes.
1988 =item v1.39, 11 February 1999
1990 Changes from Jarkko and EMX URL fixes Michael Schwern. Additional
1991 note about newlines added.
1993 =item v1.38, 31 December 1998
1995 More changes from Jarkko.
1997 =item v1.37, 19 December 1998
1999 More minor changes. Merge two separate version 1.35 documents.
2001 =item v1.36, 9 September 1998
2003 Updated for Stratus VOS. Also known as version 1.35.
2005 =item v1.35, 13 August 1998
2007 Integrate more minor changes, plus addition of new sections under
2008 L<"ISSUES">: L<"Numbers endianness and Width">,
2009 L<"Character sets and character encoding">,
2010 L<"Internationalisation">.
2012 =item v1.33, 06 August 1998
2014 Integrate more minor changes.
2016 =item v1.32, 05 August 1998
2018 Integrate more minor changes.
2020 =item v1.30, 03 August 1998
2022 Major update for RISC OS, other minor changes.
2024 =item v1.23, 10 July 1998
2026 First public release with perl5.005.
2030 =head1 Supported Platforms
2032 As of early 2001 (the Perl releases 5.6.1 and 5.7.1), the following
2033 platforms are able to build Perl from the standard source code
2034 distribution available at http://www.cpan.org/src/index.html
2057 Tru64 UNIX (DEC OSF/1, Digital UNIX)
2063 1) in DOS mode either the DOS or OS/2 ports can be used
2064 2) Mac OS Classic (pre-X) is almost 5.6.1-ready; building from
2065 the source does work with 5.6.1, but additional MacOS specific
2066 source code is needed for a complete build. See the web
2067 site http://dev.macperl.org/ for more information.
2068 3) compilers: Borland, Cygwin, Mingw32 EGCS/GCC, VC++
2070 The following platforms worked for the previous releases (5.6.0 and 5.7.0),
2071 but we did not manage to test these in time for the 5.7.1 release.
2072 There is a very good chance that these will work fine with the 5.7.1.
2090 The following platform worked for the 5.005_03 major release but not
2091 for 5.6.0. Standardization on UTF-8 as the internal string
2092 representation in 5.6.0 and 5.6.1 introduced incompatibilities in this
2093 EBCDIC platform. While Perl 5.7.1 will build on this platform some
2094 regression tests may fail and the C<use utf8;> pragma typically
2095 introduces text handling errors.
2099 1) previously known as MVS, about to become z/OS.
2101 Strongly related to the OS/390 platform by also being EBCDIC-based
2102 mainframe platforms are the following platforms:
2107 These are also expected to work, albeit with no UTF-8 support, under 5.6.1
2108 for the same reasons as OS/390. Contact the mailing list perl-mvs@perl.org
2111 The following platforms have been known to build Perl from source in
2112 the past (5.005_03 and earlier), but we haven't been able to verify
2113 their status for the current release, either because the
2114 hardware/software platforms are rare or because we don't have an
2115 active champion on these platforms--or both. They used to work,
2116 though, so go ahead and try compiling them, and let perlbug@perl.org
2155 Support for the following platform is planned for a future Perl release:
2159 The following platforms have their own source code distributions and
2160 binaries available via http://www.cpan.org/ports/
2166 Tandem Guardian 5.004
2168 The following platforms have only binaries available via
2169 http://www.cpan.org/ports/index.html :
2173 Acorn RISCOS 5.005_02
2177 Although we do suggest that you always build your own Perl from
2178 the source code, both for maximal configurability and for security,
2179 in case you are in a hurry you can check
2180 http://www.cpan.org/ports/index.html for binary distributions.
2184 L<perlaix>, L<perlapollo>, L<perlamiga>, L<perlbeos>, L<perlbs200>,
2185 L<perlce>, L<perlcygwin>, L<perldgux>, L<perldos>, L<perlepoc>, L<perlebcdic>,
2186 L<perlhurd>, L<perlhpux>, L<perlmachten>, L<perlmacos>, L<perlmint>,
2187 L<perlmpeix>, L<perlnetware>, L<perlos2>, L<perlos390>, L<perlplan9>,
2188 L<perlqnx>, L<perlsolaris>, L<perltru64>, L<perlunicode>,
2189 L<perlvmesa>, L<perlvms>, L<perlvos>, L<perlwin32>, and L<Win32>.
2191 =head1 AUTHORS / CONTRIBUTORS
2193 Abigail <abigail@foad.org>,
2194 Charles Bailey <bailey@newman.upenn.edu>,
2195 Graham Barr <gbarr@pobox.com>,
2196 Tom Christiansen <tchrist@perl.com>,
2197 Nicholas Clark <nick@ccl4.org>,
2198 Thomas Dorner <Thomas.Dorner@start.de>,
2199 Andy Dougherty <doughera@lafayette.edu>,
2200 Dominic Dunlop <domo@computer.org>,
2201 Neale Ferguson <neale@vma.tabnsw.com.au>,
2202 David J. Fiander <davidf@mks.com>,
2203 Paul Green <Paul_Green@stratus.com>,
2204 M.J.T. Guy <mjtg@cam.ac.uk>,
2205 Jarkko Hietaniemi <jhi@iki.fi>,
2206 Luther Huffman <lutherh@stratcom.com>,
2207 Nick Ing-Simmons <nick@ing-simmons.net>,
2208 Andreas J. KE<ouml>nig <a.koenig@mind.de>,
2209 Markus Laker <mlaker@contax.co.uk>,
2210 Andrew M. Langmead <aml@world.std.com>,
2211 Larry Moore <ljmoore@freespace.net>,
2212 Paul Moore <Paul.Moore@uk.origin-it.com>,
2213 Chris Nandor <pudge@pobox.com>,
2214 Matthias Neeracher <neeracher@mac.com>,
2215 Philip Newton <pne@cpan.org>,
2216 Gary Ng <71564.1743@CompuServe.COM>,
2217 Tom Phoenix <rootbeer@teleport.com>,
2218 AndrE<eacute> Pirard <A.Pirard@ulg.ac.be>,
2219 Peter Prymmer <pvhp@forte.com>,
2220 Hugo van der Sanden <hv@crypt0.demon.co.uk>,
2221 Gurusamy Sarathy <gsar@activestate.com>,
2222 Paul J. Schinder <schinder@pobox.com>,
2223 Michael G Schwern <schwern@pobox.com>,
2224 Dan Sugalski <dan@sidhe.org>,
2225 Nathan Torkington <gnat@frii.com>.