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
192 such as z/OS (OS/390) or OS/400 (using the ILE, the PASE is ASCII-based)
193 the above material is similar to "Unix" but the code numbers change:
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 this limitation may also apply to changing filesystem metainformation
409 like file permissions or owners. Remember to C<close> files when you
410 are done with them. Don't C<unlink> or C<rename> an open file. Don't
411 C<tie> or C<open> a file already tied or opened; C<untie> or C<close>
414 Don't open the same file more than once at a time for writing, as some
415 operating systems put mandatory locks on such files.
417 Don't assume that write/modify permission on a directory gives the
418 right to add or delete files/directories in that directory. That is
419 filesystem specific: in some filesystems you need write/modify
420 permission also (or even just) in the file/directory itself. In some
421 filesystems (AFS, DFS) the permission to add/delete directory entries
422 is a completely separate permission.
424 Don't assume that a single C<unlink> completely gets rid of the file:
425 some filesystems (most notably the ones in VMS) have versioned
426 filesystems, and unlink() removes only the most recent one (it doesn't
427 remove all the versions because by default the native tools on those
428 platforms remove just the most recent version, too). The portable
429 idiom to remove all the versions of a file is
431 1 while unlink "file";
433 This will terminate if the file is undeleteable for some reason
434 (protected, not there, and so on).
436 Don't count on a specific environment variable existing in C<%ENV>.
437 Don't count on C<%ENV> entries being case-sensitive, or even
438 case-preserving. Don't try to clear %ENV by saying C<%ENV = ();>, or,
439 if you really have to, make it conditional on C<$^O ne 'VMS'> since in
440 VMS the C<%ENV> table is much more than a per-process key-value string
443 Don't count on signals or C<%SIG> for anything.
445 Don't count on filename globbing. Use C<opendir>, C<readdir>, and
448 Don't count on per-program environment variables, or per-program current
451 Don't count on specific values of C<$!>, neither numeric nor
452 especially the strings values-- users may switch their locales causing
453 error messages to be translated into their languages. If you can
454 trust a POSIXish environment, you can portably use the symbols defined
455 by the Errno module, like ENOENT. And don't trust on the values of C<$!>
456 at all except immediately after a failed system call.
458 =head2 Command names versus file pathnames
460 Don't assume that the name used to invoke a command or program with
461 C<system> or C<exec> can also be used to test for the existence of the
462 file that holds the executable code for that command or program.
463 First, many systems have "internal" commands that are built-in to the
464 shell or OS and while these commands can be invoked, there is no
465 corresponding file. Second, some operating systems (e.g., Cygwin,
466 DJGPP, OS/2, and VOS) have required suffixes for executable files;
467 these suffixes are generally permitted on the command name but are not
468 required. Thus, a command like "perl" might exist in a file named
469 "perl", "perl.exe", or "perl.pm", depending on the operating system.
470 The variable "_exe" in the Config module holds the executable suffix,
471 if any. Third, the VMS port carefully sets up $^X and
472 $Config{perlpath} so that no further processing is required. This is
473 just as well, because the matching regular expression used below would
474 then have to deal with a possible trailing version number in the VMS
477 To convert $^X to a file pathname, taking account of the requirements
478 of the various operating system possibilities, say:
482 {$thisperl .= $Config{_exe} unless $thisperl =~ m/$Config{_exe}$/i;}
484 To convert $Config{perlpath} to a file pathname, say:
486 $thisperl = $Config{perlpath};
488 {$thisperl .= $Config{_exe} unless $thisperl =~ m/$Config{_exe}$/i;}
490 =head2 Interprocess Communication (IPC)
492 In general, don't directly access the system in code meant to be
493 portable. That means, no C<system>, C<exec>, C<fork>, C<pipe>,
494 C<``>, C<qx//>, C<open> with a C<|>, nor any of the other things
495 that makes being a perl hacker worth being.
497 Commands that launch external processes are generally supported on
498 most platforms (though many of them do not support any type of
499 forking). The problem with using them arises from what you invoke
500 them on. External tools are often named differently on different
501 platforms, may not be available in the same location, might accept
502 different arguments, can behave differently, and often present their
503 results in a platform-dependent way. Thus, you should seldom depend
504 on them to produce consistent results. (Then again, if you're calling
505 I<netstat -a>, you probably don't expect it to run on both Unix and CP/M.)
507 One especially common bit of Perl code is opening a pipe to B<sendmail>:
509 open(MAIL, '|/usr/lib/sendmail -t')
510 or die "cannot fork sendmail: $!";
512 This is fine for systems programming when sendmail is known to be
513 available. But it is not fine for many non-Unix systems, and even
514 some Unix systems that may not have sendmail installed. If a portable
515 solution is needed, see the various distributions on CPAN that deal
516 with it. Mail::Mailer and Mail::Send in the MailTools distribution are
517 commonly used, and provide several mailing methods, including mail,
518 sendmail, and direct SMTP (via Net::SMTP) if a mail transfer agent is
519 not available. Mail::Sendmail is a standalone module that provides
520 simple, platform-independent mailing.
522 The Unix System V IPC (C<msg*(), sem*(), shm*()>) is not available
523 even on all Unix platforms.
525 Do not use either the bare result of C<pack("N", 10, 20, 30, 40)> or
526 bare v-strings (such as C<v10.20.30.40>) to represent IPv4 addresses:
527 both forms just pack the four bytes into network order. That this
528 would be equal to the C language C<in_addr> struct (which is what the
529 socket code internally uses) is not guaranteed. To be portable use
530 the routines of the Socket extension, such as C<inet_aton()>,
531 C<inet_ntoa()>, and C<sockaddr_in()>.
533 The rule of thumb for portable code is: Do it all in portable Perl, or
534 use a module (that may internally implement it with platform-specific
535 code, but expose a common interface).
537 =head2 External Subroutines (XS)
539 XS code can usually be made to work with any platform, but dependent
540 libraries, header files, etc., might not be readily available or
541 portable, or the XS code itself might be platform-specific, just as Perl
542 code might be. If the libraries and headers are portable, then it is
543 normally reasonable to make sure the XS code is portable, too.
545 A different type of portability issue arises when writing XS code:
546 availability of a C compiler on the end-user's system. C brings
547 with it its own portability issues, and writing XS code will expose
548 you to some of those. Writing purely in Perl is an easier way to
551 =head2 Standard Modules
553 In general, the standard modules work across platforms. Notable
554 exceptions are the CPAN module (which currently makes connections to external
555 programs that may not be available), platform-specific modules (like
556 ExtUtils::MM_VMS), and DBM modules.
558 There is no one DBM module available on all platforms.
559 SDBM_File and the others are generally available on all Unix and DOSish
560 ports, but not in MacPerl, where only NBDM_File and DB_File are
563 The good news is that at least some DBM module should be available, and
564 AnyDBM_File will use whichever module it can find. Of course, then
565 the code needs to be fairly strict, dropping to the greatest common
566 factor (e.g., not exceeding 1K for each record), so that it will
567 work with any DBM module. See L<AnyDBM_File> for more details.
571 The system's notion of time of day and calendar date is controlled in
572 widely different ways. Don't assume the timezone is stored in C<$ENV{TZ}>,
573 and even if it is, don't assume that you can control the timezone through
574 that variable. Don't assume anything about the three-letter timezone
575 abbreviations (for example that MST would be the Mountain Standard Time,
576 it's been known to stand for Moscow Standard Time). If you need to
577 use timezones, express them in some unambiguous format like the
578 exact number of minutes offset from UTC, or the POSIX timezone
581 Don't assume that the epoch starts at 00:00:00, January 1, 1970,
582 because that is OS- and implementation-specific. It is better to
583 store a date in an unambiguous representation. The ISO 8601 standard
584 defines YYYY-MM-DD as the date format, or YYYY-MM-DDTHH-MM-SS
585 (that's a literal "T" separating the date from the time).
586 Please do use the ISO 8601 instead of making us to guess what
587 date 02/03/04 might be. ISO 8601 even sorts nicely as-is.
588 A text representation (like "1987-12-18") can be easily converted
589 into an OS-specific value using a module like Date::Parse.
590 An array of values, such as those returned by C<localtime>, can be
591 converted to an OS-specific representation using Time::Local.
593 When calculating specific times, such as for tests in time or date modules,
594 it may be appropriate to calculate an offset for the epoch.
597 $offset = Time::Local::timegm(0, 0, 0, 1, 0, 70);
599 The value for C<$offset> in Unix will be C<0>, but in Mac OS will be
600 some large number. C<$offset> can then be added to a Unix time value
601 to get what should be the proper value on any system.
603 =head2 Character sets and character encoding
605 Assume very little about character sets.
607 Assume nothing about numerical values (C<ord>, C<chr>) of characters.
608 Do not use explicit code point ranges (like \xHH-\xHH); use for
609 example symbolic character classes like C<[:print:]>.
611 Do not assume that the alphabetic characters are encoded contiguously
612 (in the numeric sense). There may be gaps.
614 Do not assume anything about the ordering of the characters.
615 The lowercase letters may come before or after the uppercase letters;
616 the lowercase and uppercase may be interlaced so that both `a' and `A'
617 come before `b'; the accented and other international characters may
618 be interlaced so that E<auml> comes before `b'.
620 =head2 Internationalisation
622 If you may assume POSIX (a rather large assumption), you may read
623 more about the POSIX locale system from L<perllocale>. The locale
624 system at least attempts to make things a little bit more portable,
625 or at least more convenient and native-friendly for non-English
626 users. The system affects character sets and encoding, and date
627 and time formatting--amongst other things.
629 If you really want to be international, you should consider Unicode.
630 See L<perluniintro> and L<perlunicode> for more information.
632 =head2 System Resources
634 If your code is destined for systems with severely constrained (or
635 missing!) virtual memory systems then you want to be I<especially> mindful
636 of avoiding wasteful constructs such as:
638 # NOTE: this is no longer "bad" in perl5.005
639 for (0..10000000) {} # bad
640 for (my $x = 0; $x <= 10000000; ++$x) {} # good
642 @lines = <VERY_LARGE_FILE>; # bad
644 while (<FILE>) {$file .= $_} # sometimes bad
645 $file = join('', <FILE>); # better
647 The last two constructs may appear unintuitive to most people. The
648 first repeatedly grows a string, whereas the second allocates a
649 large chunk of memory in one go. On some systems, the second is
650 more efficient that the first.
654 Most multi-user platforms provide basic levels of security, usually
655 implemented at the filesystem level. Some, however, do
656 not-- unfortunately. Thus the notion of user id, or "home" directory,
657 or even the state of being logged-in, may be unrecognizable on many
658 platforms. If you write programs that are security-conscious, it
659 is usually best to know what type of system you will be running
660 under so that you can write code explicitly for that platform (or
663 Don't assume the UNIX filesystem access semantics: the operating
664 system or the filesystem may be using some ACL systems, which are
665 richer languages than the usual rwx. Even if the rwx exist,
666 their semantics might be different.
668 (From security viewpoint testing for permissions before attempting to
669 do something is silly anyway: if one tries this, there is potential
670 for race conditions-- someone or something might change the
671 permissions between the permissions check and the actual operation.
672 Just try the operation.)
674 Don't assume the UNIX user and group semantics: especially, don't
675 expect the C<< $< >> and C<< $> >> (or the C<$(> and C<$)>) to work
676 for switching identities (or memberships).
678 Don't assume set-uid and set-gid semantics. (And even if you do,
679 think twice: set-uid and set-gid are a known can of security worms.)
683 For those times when it is necessary to have platform-specific code,
684 consider keeping the platform-specific code in one place, making porting
685 to other platforms easier. Use the Config module and the special
686 variable C<$^O> to differentiate platforms, as described in
689 Be careful in the tests you supply with your module or programs.
690 Module code may be fully portable, but its tests might not be. This
691 often happens when tests spawn off other processes or call external
692 programs to aid in the testing, or when (as noted above) the tests
693 assume certain things about the filesystem and paths. Be careful not
694 to depend on a specific output style for errors, such as when checking
695 C<$!> after a failed system call. Using C<$!> for anything else than
696 displaying it as output is doubtful (though see the Errno module for
697 testing reasonably portably for error value). Some platforms expect
698 a certain output format, and Perl on those platforms may have been
699 adjusted accordingly. Most specifically, don't anchor a regex when
700 testing an error value.
704 Modules uploaded to CPAN are tested by a variety of volunteers on
705 different platforms. These CPAN testers are notified by mail of each
706 new upload, and reply to the list with PASS, FAIL, NA (not applicable to
707 this platform), or UNKNOWN (unknown), along with any relevant notations.
709 The purpose of the testing is twofold: one, to help developers fix any
710 problems in their code that crop up because of lack of testing on other
711 platforms; two, to provide users with information about whether
712 a given module works on a given platform.
716 =item Mailing list: cpan-testers@perl.org
718 =item Testing results: http://testers.cpan.org/
724 As of version 5.002, Perl is built with a C<$^O> variable that
725 indicates the operating system it was built on. This was implemented
726 to help speed up code that would otherwise have to C<use Config>
727 and use the value of C<$Config{osname}>. Of course, to get more
728 detailed information about the system, looking into C<%Config> is
729 certainly recommended.
731 C<%Config> cannot always be trusted, however, because it was built
732 at compile time. If perl was built in one place, then transferred
733 elsewhere, some values may be wrong. The values may even have been
734 edited after the fact.
738 Perl works on a bewildering variety of Unix and Unix-like platforms (see
739 e.g. most of the files in the F<hints/> directory in the source code kit).
740 On most of these systems, the value of C<$^O> (hence C<$Config{'osname'}>,
741 too) is determined either by lowercasing and stripping punctuation from the
742 first field of the string returned by typing C<uname -a> (or a similar command)
743 at the shell prompt or by testing the file system for the presence of
744 uniquely named files such as a kernel or header file. Here, for example,
745 are a few of the more popular Unix flavors:
747 uname $^O $Config{'archname'}
748 --------------------------------------------
750 BSD/OS bsdos i386-bsdos
752 dgux dgux AViiON-dgux
753 DYNIX/ptx dynixptx i386-dynixptx
754 FreeBSD freebsd freebsd-i386
755 Linux linux arm-linux
756 Linux linux i386-linux
757 Linux linux i586-linux
758 Linux linux ppc-linux
759 HP-UX hpux PA-RISC1.1
761 Mac OS X darwin darwin
762 MachTen PPC machten powerpc-machten
764 NeXT 4 next OPENSTEP-Mach
765 openbsd openbsd i386-openbsd
766 OSF1 dec_osf alpha-dec_osf
767 reliantunix-n svr4 RM400-svr4
768 SCO_SV sco_sv i386-sco_sv
769 SINIX-N svr4 RM400-svr4
770 sn4609 unicos CRAY_C90-unicos
771 sn6521 unicosmk t3e-unicosmk
772 sn9617 unicos CRAY_J90-unicos
773 SunOS solaris sun4-solaris
774 SunOS solaris i86pc-solaris
775 SunOS4 sunos sun4-sunos
777 Because the value of C<$Config{archname}> may depend on the
778 hardware architecture, it can vary more than the value of C<$^O>.
780 =head2 DOS and Derivatives
782 Perl has long been ported to Intel-style microcomputers running under
783 systems like PC-DOS, MS-DOS, OS/2, and most Windows platforms you can
784 bring yourself to mention (except for Windows CE, if you count that).
785 Users familiar with I<COMMAND.COM> or I<CMD.EXE> style shells should
786 be aware that each of these file specifications may have subtle
789 $filespec0 = "c:/foo/bar/file.txt";
790 $filespec1 = "c:\\foo\\bar\\file.txt";
791 $filespec2 = 'c:\foo\bar\file.txt';
792 $filespec3 = 'c:\\foo\\bar\\file.txt';
794 System calls accept either C</> or C<\> as the path separator.
795 However, many command-line utilities of DOS vintage treat C</> as
796 the option prefix, so may get confused by filenames containing C</>.
797 Aside from calling any external programs, C</> will work just fine,
798 and probably better, as it is more consistent with popular usage,
799 and avoids the problem of remembering what to backwhack and what
802 The DOS FAT filesystem can accommodate only "8.3" style filenames. Under
803 the "case-insensitive, but case-preserving" HPFS (OS/2) and NTFS (NT)
804 filesystems you may have to be careful about case returned with functions
805 like C<readdir> or used with functions like C<open> or C<opendir>.
807 DOS also treats several filenames as special, such as AUX, PRN,
808 NUL, CON, COM1, LPT1, LPT2, etc. Unfortunately, sometimes these
809 filenames won't even work if you include an explicit directory
810 prefix. It is best to avoid such filenames, if you want your code
811 to be portable to DOS and its derivatives. It's hard to know what
812 these all are, unfortunately.
814 Users of these operating systems may also wish to make use of
815 scripts such as I<pl2bat.bat> or I<pl2cmd> to
816 put wrappers around your scripts.
818 Newline (C<\n>) is translated as C<\015\012> by STDIO when reading from
819 and writing to files (see L<"Newlines">). C<binmode(FILEHANDLE)>
820 will keep C<\n> translated as C<\012> for that filehandle. Since it is a
821 no-op on other systems, C<binmode> should be used for cross-platform code
822 that deals with binary data. That's assuming you realize in advance
823 that your data is in binary. General-purpose programs should
824 often assume nothing about their data.
826 The C<$^O> variable and the C<$Config{archname}> values for various
827 DOSish perls are as follows:
829 OS $^O $Config{archname} ID Version
830 --------------------------------------------------------
834 Windows 3.1 ? ? 0 3 01
835 Windows 95 MSWin32 MSWin32-x86 1 4 00
836 Windows 98 MSWin32 MSWin32-x86 1 4 10
837 Windows ME MSWin32 MSWin32-x86 1 ?
838 Windows NT MSWin32 MSWin32-x86 2 4 xx
839 Windows NT MSWin32 MSWin32-ALPHA 2 4 xx
840 Windows NT MSWin32 MSWin32-ppc 2 4 xx
841 Windows 2000 MSWin32 MSWin32-x86 2 5 xx
842 Windows XP MSWin32 MSWin32-x86 2 ?
843 Windows CE MSWin32 ? 3
846 The various MSWin32 Perl's can distinguish the OS they are running on
847 via the value of the fifth element of the list returned from
848 Win32::GetOSVersion(). For example:
850 if ($^O eq 'MSWin32') {
851 my @os_version_info = Win32::GetOSVersion();
852 print +('3.1','95','NT')[$os_version_info[4]],"\n";
855 There are also Win32::IsWinNT() and Win32::IsWin95(), try C<perldoc Win32>,
856 and as of libwin32 0.19 (not part of the core Perl distribution)
857 Win32::GetOSName(). The very portable POSIX::uname() will work too:
859 c:\> perl -MPOSIX -we "print join '|', uname"
860 Windows NT|moonru|5.0|Build 2195 (Service Pack 2)|x86
868 The djgpp environment for DOS, http://www.delorie.com/djgpp/
873 The EMX environment for DOS, OS/2, etc. emx@iaehv.nl,
874 http://www.leo.org/pub/comp/os/os2/leo/gnu/emx+gcc/index.html or
875 ftp://hobbes.nmsu.edu/pub/os2/dev/emx/ Also L<perlos2>.
879 Build instructions for Win32 in L<perlwin32>, or under the Cygnus environment
884 The C<Win32::*> modules in L<Win32>.
888 The ActiveState Pages, http://www.activestate.com/
892 The Cygwin environment for Win32; F<README.cygwin> (installed
893 as L<perlcygwin>), http://www.cygwin.com/
897 The U/WIN environment for Win32,
898 http://www.research.att.com/sw/tools/uwin/
902 Build instructions for OS/2, L<perlos2>
908 Any module requiring XS compilation is right out for most people, because
909 MacPerl is built using non-free (and non-cheap!) compilers. Some XS
910 modules that can work with MacPerl are built and distributed in binary
913 Directories are specified as:
915 volume:folder:file for absolute pathnames
916 volume:folder: for absolute pathnames
917 :folder:file for relative pathnames
918 :folder: for relative pathnames
919 :file for relative pathnames
920 file for relative pathnames
922 Files are stored in the directory in alphabetical order. Filenames are
923 limited to 31 characters, and may include any character except for
924 null and C<:>, which is reserved as the path separator.
926 Instead of C<flock>, see C<FSpSetFLock> and C<FSpRstFLock> in the
927 Mac::Files module, or C<chmod(0444, ...)> and C<chmod(0666, ...)>.
929 In the MacPerl application, you can't run a program from the command line;
930 programs that expect C<@ARGV> to be populated can be edited with something
931 like the following, which brings up a dialog box asking for the command
935 @ARGV = split /\s+/, MacPerl::Ask('Arguments?');
938 A MacPerl script saved as a "droplet" will populate C<@ARGV> with the full
939 pathnames of the files dropped onto the script.
941 Mac users can run programs under a type of command line interface
942 under MPW (Macintosh Programmer's Workshop, a free development
943 environment from Apple). MacPerl was first introduced as an MPW
944 tool, and MPW can be used like a shell:
946 perl myscript.plx some arguments
948 ToolServer is another app from Apple that provides access to MPW tools
949 from MPW and the MacPerl app, which allows MacPerl programs to use
950 C<system>, backticks, and piped C<open>.
952 "S<Mac OS>" is the proper name for the operating system, but the value
953 in C<$^O> is "MacOS". To determine architecture, version, or whether
954 the application or MPW tool version is running, check:
956 $is_app = $MacPerl::Version =~ /App/;
957 $is_tool = $MacPerl::Version =~ /MPW/;
958 ($version) = $MacPerl::Version =~ /^(\S+)/;
959 $is_ppc = $MacPerl::Architecture eq 'MacPPC';
960 $is_68k = $MacPerl::Architecture eq 'Mac68K';
962 S<Mac OS X>, based on NeXT's OpenStep OS, runs MacPerl natively, under the
963 "Classic" environment. There is no "Carbon" version of MacPerl to run
964 under the primary Mac OS X environment. S<Mac OS X> and its Open Source
965 version, Darwin, both run Unix perl natively.
973 MacPerl Development, http://dev.macperl.org/ .
977 The MacPerl Pages, http://www.macperl.com/ .
981 The MacPerl mailing lists, http://lists.perl.org/ .
987 Perl on VMS is discussed in L<perlvms> in the perl distribution.
988 Perl on VMS can accept either VMS- or Unix-style file
989 specifications as in either of the following:
991 $ perl -ne "print if /perl_setup/i" SYS$LOGIN:LOGIN.COM
992 $ perl -ne "print if /perl_setup/i" /sys$login/login.com
994 but not a mixture of both as in:
996 $ perl -ne "print if /perl_setup/i" sys$login:/login.com
997 Can't open sys$login:/login.com: file specification syntax error
999 Interacting with Perl from the Digital Command Language (DCL) shell
1000 often requires a different set of quotation marks than Unix shells do.
1003 $ perl -e "print ""Hello, world.\n"""
1006 There are several ways to wrap your perl scripts in DCL F<.COM> files, if
1007 you are so inclined. For example:
1009 $ write sys$output "Hello from DCL!"
1011 $ then perl -x 'f$environment("PROCEDURE")
1012 $ else perl -x - 'p1 'p2 'p3 'p4 'p5 'p6 'p7 'p8
1013 $ deck/dollars="__END__"
1016 print "Hello from Perl!\n";
1021 Do take care with C<$ ASSIGN/nolog/user SYS$COMMAND: SYS$INPUT> if your
1022 perl-in-DCL script expects to do things like C<< $read = <STDIN>; >>.
1024 Filenames are in the format "name.extension;version". The maximum
1025 length for filenames is 39 characters, and the maximum length for
1026 extensions is also 39 characters. Version is a number from 1 to
1027 32767. Valid characters are C</[A-Z0-9$_-]/>.
1029 VMS's RMS filesystem is case-insensitive and does not preserve case.
1030 C<readdir> returns lowercased filenames, but specifying a file for
1031 opening remains case-insensitive. Files without extensions have a
1032 trailing period on them, so doing a C<readdir> with a file named F<A.;5>
1033 will return F<a.> (though that file could be opened with
1036 RMS had an eight level limit on directory depths from any rooted logical
1037 (allowing 16 levels overall) prior to VMS 7.2. Hence
1038 C<PERL_ROOT:[LIB.2.3.4.5.6.7.8]> is a valid directory specification but
1039 C<PERL_ROOT:[LIB.2.3.4.5.6.7.8.9]> is not. F<Makefile.PL> authors might
1040 have to take this into account, but at least they can refer to the former
1041 as C</PERL_ROOT/lib/2/3/4/5/6/7/8/>.
1043 The VMS::Filespec module, which gets installed as part of the build
1044 process on VMS, is a pure Perl module that can easily be installed on
1045 non-VMS platforms and can be helpful for conversions to and from RMS
1048 What C<\n> represents depends on the type of file opened. It usually
1049 represents C<\012> but it could also be C<\015>, C<\012>, C<\015\012>,
1050 C<\000>, C<\040>, or nothing depending on the file organiztion and
1051 record format. The VMS::Stdio module provides access to the
1052 special fopen() requirements of files with unusual attributes on VMS.
1054 TCP/IP stacks are optional on VMS, so socket routines might not be
1055 implemented. UDP sockets may not be supported.
1057 The value of C<$^O> on OpenVMS is "VMS". To determine the architecture
1058 that you are running on without resorting to loading all of C<%Config>
1059 you can examine the content of the C<@INC> array like so:
1061 if (grep(/VMS_AXP/, @INC)) {
1062 print "I'm on Alpha!\n";
1064 } elsif (grep(/VMS_VAX/, @INC)) {
1065 print "I'm on VAX!\n";
1068 print "I'm not so sure about where $^O is...\n";
1071 On VMS, perl determines the UTC offset from the C<SYS$TIMEZONE_DIFFERENTIAL>
1072 logical name. Although the VMS epoch began at 17-NOV-1858 00:00:00.00,
1073 calls to C<localtime> are adjusted to count offsets from
1074 01-JAN-1970 00:00:00.00, just like Unix.
1082 F<README.vms> (installed as L<README_vms>), L<perlvms>
1086 vmsperl list, majordomo@perl.org
1088 (Put the words C<subscribe vmsperl> in message body.)
1092 vmsperl on the web, http://www.sidhe.org/vmsperl/index.html
1098 Perl on VOS is discussed in F<README.vos> in the perl distribution
1099 (installed as L<perlvos>). Perl on VOS can accept either VOS- or
1100 Unix-style file specifications as in either of the following:
1102 C<< $ perl -ne "print if /perl_setup/i" >system>notices >>
1103 C<< $ perl -ne "print if /perl_setup/i" /system/notices >>
1105 or even a mixture of both as in:
1107 C<< $ perl -ne "print if /perl_setup/i" >system/notices >>
1109 Even though VOS allows the slash character to appear in object
1110 names, because the VOS port of Perl interprets it as a pathname
1111 delimiting character, VOS files, directories, or links whose names
1112 contain a slash character cannot be processed. Such files must be
1113 renamed before they can be processed by Perl. Note that VOS limits
1114 file names to 32 or fewer characters.
1116 The value of C<$^O> on VOS is "VOS". To determine the architecture that
1117 you are running on without resorting to loading all of C<%Config> you
1118 can examine the content of the @INC array like so:
1121 print "I'm on a Stratus box!\n";
1123 print "I'm not on a Stratus box!\n";
1133 F<README.vos> (installed as L<perlvos>)
1137 The VOS mailing list.
1139 There is no specific mailing list for Perl on VOS. You can post
1140 comments to the comp.sys.stratus newsgroup, or subscribe to the general
1141 Stratus mailing list. Send a letter with "subscribe Info-Stratus" in
1142 the message body to majordomo@list.stratagy.com.
1146 VOS Perl on the web at http://ftp.stratus.com/pub/vos/posix/posix.html
1150 =head2 EBCDIC Platforms
1152 Recent versions of Perl have been ported to platforms such as OS/400 on
1153 AS/400 minicomputers as well as OS/390, VM/ESA, and BS2000 for S/390
1154 Mainframes. Such computers use EBCDIC character sets internally (usually
1155 Character Code Set ID 0037 for OS/400 and either 1047 or POSIX-BC for S/390
1156 systems). On the mainframe perl currently works under the "Unix system
1157 services for OS/390" (formerly known as OpenEdition), VM/ESA OpenEdition, or
1158 the BS200 POSIX-BC system (BS2000 is supported in perl 5.6 and greater).
1159 See L<perlos390> for details. Note that for OS/400 there is also a port of
1160 Perl 5.8.1/5.9.0 or later to the PASE which is ASCII-based (as opposed to
1161 ILE which is EBCDIC-based), see L<perlos400>.
1163 As of R2.5 of USS for OS/390 and Version 2.3 of VM/ESA these Unix
1164 sub-systems do not support the C<#!> shebang trick for script invocation.
1165 Hence, on OS/390 and VM/ESA perl scripts can be executed with a header
1166 similar to the following simple script:
1169 eval 'exec /usr/local/bin/perl -S $0 ${1+"$@"}'
1171 #!/usr/local/bin/perl # just a comment really
1173 print "Hello from perl!\n";
1175 OS/390 will support the C<#!> shebang trick in release 2.8 and beyond.
1176 Calls to C<system> and backticks can use POSIX shell syntax on all
1179 On the AS/400, if PERL5 is in your library list, you may need
1180 to wrap your perl scripts in a CL procedure to invoke them like so:
1183 CALL PGM(PERL5/PERL) PARM('/QOpenSys/hello.pl')
1186 This will invoke the perl script F<hello.pl> in the root of the
1187 QOpenSys file system. On the AS/400 calls to C<system> or backticks
1190 On these platforms, bear in mind that the EBCDIC character set may have
1191 an effect on what happens with some perl functions (such as C<chr>,
1192 C<pack>, C<print>, C<printf>, C<ord>, C<sort>, C<sprintf>, C<unpack>), as
1193 well as bit-fiddling with ASCII constants using operators like C<^>, C<&>
1194 and C<|>, not to mention dealing with socket interfaces to ASCII computers
1195 (see L<"Newlines">).
1197 Fortunately, most web servers for the mainframe will correctly
1198 translate the C<\n> in the following statement to its ASCII equivalent
1199 (C<\r> is the same under both Unix and OS/390 & VM/ESA):
1201 print "Content-type: text/html\r\n\r\n";
1203 The values of C<$^O> on some of these platforms includes:
1205 uname $^O $Config{'archname'}
1206 --------------------------------------------
1209 POSIX-BC posix-bc BS2000-posix-bc
1212 Some simple tricks for determining if you are running on an EBCDIC
1213 platform could include any of the following (perhaps all):
1215 if ("\t" eq "\05") { print "EBCDIC may be spoken here!\n"; }
1217 if (ord('A') == 193) { print "EBCDIC may be spoken here!\n"; }
1219 if (chr(169) eq 'z') { print "EBCDIC may be spoken here!\n"; }
1221 One thing you may not want to rely on is the EBCDIC encoding
1222 of punctuation characters since these may differ from code page to code
1223 page (and once your module or script is rumoured to work with EBCDIC,
1224 folks will want it to work with all EBCDIC character sets).
1234 L<perlos390>, F<README.os390>, F<perlbs2000>, F<README.vmesa>,
1239 The perl-mvs@perl.org list is for discussion of porting issues as well as
1240 general usage issues for all EBCDIC Perls. Send a message body of
1241 "subscribe perl-mvs" to majordomo@perl.org.
1245 AS/400 Perl information at
1246 http://as400.rochester.ibm.com/
1247 as well as on CPAN in the F<ports/> directory.
1251 =head2 Acorn RISC OS
1253 Because Acorns use ASCII with newlines (C<\n>) in text files as C<\012> like
1254 Unix, and because Unix filename emulation is turned on by default,
1255 most simple scripts will probably work "out of the box". The native
1256 filesystem is modular, and individual filesystems are free to be
1257 case-sensitive or insensitive, and are usually case-preserving. Some
1258 native filesystems have name length limits, which file and directory
1259 names are silently truncated to fit. Scripts should be aware that the
1260 standard filesystem currently has a name length limit of B<10>
1261 characters, with up to 77 items in a directory, but other filesystems
1262 may not impose such limitations.
1264 Native filenames are of the form
1266 Filesystem#Special_Field::DiskName.$.Directory.Directory.File
1270 Special_Field is not usually present, but may contain . and $ .
1271 Filesystem =~ m|[A-Za-z0-9_]|
1272 DsicName =~ m|[A-Za-z0-9_/]|
1273 $ represents the root directory
1274 . is the path separator
1275 @ is the current directory (per filesystem but machine global)
1276 ^ is the parent directory
1277 Directory and File =~ m|[^\0- "\.\$\%\&:\@\\^\|\177]+|
1279 The default filename translation is roughly C<tr|/.|./|;>
1281 Note that C<"ADFS::HardDisk.$.File" ne 'ADFS::HardDisk.$.File'> and that
1282 the second stage of C<$> interpolation in regular expressions will fall
1283 foul of the C<$.> if scripts are not careful.
1285 Logical paths specified by system variables containing comma-separated
1286 search lists are also allowed; hence C<System:Modules> is a valid
1287 filename, and the filesystem will prefix C<Modules> with each section of
1288 C<System$Path> until a name is made that points to an object on disk.
1289 Writing to a new file C<System:Modules> would be allowed only if
1290 C<System$Path> contains a single item list. The filesystem will also
1291 expand system variables in filenames if enclosed in angle brackets, so
1292 C<< <System$Dir>.Modules >> would look for the file
1293 S<C<$ENV{'System$Dir'} . 'Modules'>>. The obvious implication of this is
1294 that B<fully qualified filenames can start with C<< <> >>> and should
1295 be protected when C<open> is used for input.
1297 Because C<.> was in use as a directory separator and filenames could not
1298 be assumed to be unique after 10 characters, Acorn implemented the C
1299 compiler to strip the trailing C<.c> C<.h> C<.s> and C<.o> suffix from
1300 filenames specified in source code and store the respective files in
1301 subdirectories named after the suffix. Hence files are translated:
1304 C:foo.h C:h.foo (logical path variable)
1305 sys/os.h sys.h.os (C compiler groks Unix-speak)
1306 10charname.c c.10charname
1307 10charname.o o.10charname
1308 11charname_.c c.11charname (assuming filesystem truncates at 10)
1310 The Unix emulation library's translation of filenames to native assumes
1311 that this sort of translation is required, and it allows a user-defined list
1312 of known suffixes that it will transpose in this fashion. This may
1313 seem transparent, but consider that with these rules C<foo/bar/baz.h>
1314 and C<foo/bar/h/baz> both map to C<foo.bar.h.baz>, and that C<readdir> and
1315 C<glob> cannot and do not attempt to emulate the reverse mapping. Other
1316 C<.>'s in filenames are translated to C</>.
1318 As implied above, the environment accessed through C<%ENV> is global, and
1319 the convention is that program specific environment variables are of the
1320 form C<Program$Name>. Each filesystem maintains a current directory,
1321 and the current filesystem's current directory is the B<global> current
1322 directory. Consequently, sociable programs don't change the current
1323 directory but rely on full pathnames, and programs (and Makefiles) cannot
1324 assume that they can spawn a child process which can change the current
1325 directory without affecting its parent (and everyone else for that
1328 Because native operating system filehandles are global and are currently
1329 allocated down from 255, with 0 being a reserved value, the Unix emulation
1330 library emulates Unix filehandles. Consequently, you can't rely on
1331 passing C<STDIN>, C<STDOUT>, or C<STDERR> to your children.
1333 The desire of users to express filenames of the form
1334 C<< <Foo$Dir>.Bar >> on the command line unquoted causes problems,
1335 too: C<``> command output capture has to perform a guessing game. It
1336 assumes that a string C<< <[^<>]+\$[^<>]> >> is a
1337 reference to an environment variable, whereas anything else involving
1338 C<< < >> or C<< > >> is redirection, and generally manages to be 99%
1339 right. Of course, the problem remains that scripts cannot rely on any
1340 Unix tools being available, or that any tools found have Unix-like command
1343 Extensions and XS are, in theory, buildable by anyone using free
1344 tools. In practice, many don't, as users of the Acorn platform are
1345 used to binary distributions. MakeMaker does run, but no available
1346 make currently copes with MakeMaker's makefiles; even if and when
1347 this should be fixed, the lack of a Unix-like shell will cause
1348 problems with makefile rules, especially lines of the form C<cd
1349 sdbm && make all>, and anything using quoting.
1351 "S<RISC OS>" is the proper name for the operating system, but the value
1352 in C<$^O> is "riscos" (because we don't like shouting).
1356 Perl has been ported to many platforms that do not fit into any of
1357 the categories listed above. Some, such as AmigaOS, Atari MiNT,
1358 BeOS, HP MPE/iX, QNX, Plan 9, and VOS, have been well-integrated
1359 into the standard Perl source code kit. You may need to see the
1360 F<ports/> directory on CPAN for information, and possibly binaries,
1361 for the likes of: aos, Atari ST, lynxos, riscos, Novell Netware,
1362 Tandem Guardian, I<etc.> (Yes, we know that some of these OSes may
1363 fall under the Unix category, but we are not a standards body.)
1365 Some approximate operating system names and their C<$^O> values
1366 in the "OTHER" category include:
1368 OS $^O $Config{'archname'}
1369 ------------------------------------------
1370 Amiga DOS amigaos m68k-amigos
1372 MPE/iX mpeix PA-RISC1.1
1380 Amiga, F<README.amiga> (installed as L<perlamiga>).
1384 Atari, F<README.mint> and Guido Flohr's web page
1385 http://stud.uni-sb.de/~gufl0000/
1389 Be OS, F<README.beos>
1393 HP 300 MPE/iX, F<README.mpeix> and Mark Bixby's web page
1394 http://www.bixby.org/mark/perlix.html
1398 A free perl5-based PERL.NLM for Novell Netware is available in
1399 precompiled binary and source code form from http://www.novell.com/
1400 as well as from CPAN.
1404 S<Plan 9>, F<README.plan9>
1408 =head1 FUNCTION IMPLEMENTATIONS
1410 Listed below are functions that are either completely unimplemented
1411 or else have been implemented differently on various platforms.
1412 Following each description will be, in parentheses, a list of
1413 platforms that the description applies to.
1415 The list may well be incomplete, or even wrong in some places. When
1416 in doubt, consult the platform-specific README files in the Perl
1417 source distribution, and any other documentation resources accompanying
1420 Be aware, moreover, that even among Unix-ish systems there are variations.
1422 For many functions, you can also query C<%Config>, exported by
1423 default from the Config module. For example, to check whether the
1424 platform has the C<lstat> call, check C<$Config{d_lstat}>. See
1425 L<Config> for a full description of available variables.
1427 =head2 Alphabetical Listing of Perl Functions
1437 C<-r>, C<-w>, and C<-x> have a limited meaning only; directories
1438 and applications are executable, and there are no uid/gid
1439 considerations. C<-o> is not supported. (S<Mac OS>)
1441 C<-r>, C<-w>, C<-x>, and C<-o> tell whether the file is accessible,
1442 which may not reflect UIC-based file protections. (VMS)
1444 C<-s> returns the size of the data fork, not the total size of data fork
1445 plus resource fork. (S<Mac OS>).
1447 C<-s> by name on an open file will return the space reserved on disk,
1448 rather than the current extent. C<-s> on an open filehandle returns the
1449 current size. (S<RISC OS>)
1451 C<-R>, C<-W>, C<-X>, C<-O> are indistinguishable from C<-r>, C<-w>,
1452 C<-x>, C<-o>. (S<Mac OS>, Win32, VMS, S<RISC OS>)
1454 C<-b>, C<-c>, C<-k>, C<-g>, C<-p>, C<-u>, C<-A> are not implemented.
1457 C<-g>, C<-k>, C<-l>, C<-p>, C<-u>, C<-A> are not particularly meaningful.
1458 (Win32, VMS, S<RISC OS>)
1460 C<-d> is true if passed a device spec without an explicit directory.
1463 C<-T> and C<-B> are implemented, but might misclassify Mac text files
1464 with foreign characters; this is the case will all platforms, but may
1465 affect S<Mac OS> often. (S<Mac OS>)
1467 C<-x> (or C<-X>) determine if a file ends in one of the executable
1468 suffixes. C<-S> is meaningless. (Win32)
1470 C<-x> (or C<-X>) determine if a file has an executable file type.
1477 Not implemented. (Win32)
1479 =item binmode FILEHANDLE
1481 Meaningless. (S<Mac OS>, S<RISC OS>)
1483 Reopens file and restores pointer; if function fails, underlying
1484 filehandle may be closed, or pointer may be in a different position.
1487 The value returned by C<tell> may be affected after the call, and
1488 the filehandle may be flushed. (Win32)
1492 Only limited meaning. Disabling/enabling write permission is mapped to
1493 locking/unlocking the file. (S<Mac OS>)
1495 Only good for changing "owner" read-write access, "group", and "other"
1496 bits are meaningless. (Win32)
1498 Only good for changing "owner" and "other" read-write access. (S<RISC OS>)
1500 Access permissions are mapped onto VOS access-control list changes. (VOS)
1502 The actual permissions set depend on the value of the C<CYGWIN>
1503 in the SYSTEM environment settings. (Cygwin)
1507 Not implemented. (S<Mac OS>, Win32, S<Plan 9>, S<RISC OS>)
1509 Does nothing, but won't fail. (Win32)
1511 A little funky, because VOS's notion of ownership is a little funky (VOS).
1513 =item chroot FILENAME
1517 Not implemented. (S<Mac OS>, Win32, VMS, S<Plan 9>, S<RISC OS>, VOS, VM/ESA)
1519 =item crypt PLAINTEXT,SALT
1521 May not be available if library or source was not provided when building
1526 Not implemented. (VMS, S<Plan 9>, VOS)
1528 =item dbmopen HASH,DBNAME,MODE
1530 Not implemented. (VMS, S<Plan 9>, VOS)
1534 Not useful. (S<Mac OS>, S<RISC OS>)
1536 Not implemented. (Win32)
1538 Invokes VMS debugger. (VMS)
1542 Not implemented. (S<Mac OS>)
1544 Implemented via Spawn. (VM/ESA)
1546 Does not automatically flush output handles on some platforms.
1547 (SunOS, Solaris, HP-UX)
1553 Emulates UNIX exit() (which considers C<exit 1> to indicate an error) by
1554 mapping the C<1> to SS$_ABORT (C<44>). This behavior may be overridden
1555 with the pragma C<use vmsish 'exit'>. As with the CRTL's exit()
1556 function, C<exit 0> is also mapped to an exit status of SS$_NORMAL
1557 (C<1>); this mapping cannot be overridden. Any other argument to exit()
1558 is used directly as Perl's exit status. (VMS)
1560 =item fcntl FILEHANDLE,FUNCTION,SCALAR
1562 Not implemented. (Win32, VMS)
1564 =item flock FILEHANDLE,OPERATION
1566 Not implemented (S<Mac OS>, VMS, S<RISC OS>, VOS).
1568 Available only on Windows NT (not on Windows 95). (Win32)
1572 Not implemented. (S<Mac OS>, AmigaOS, S<RISC OS>, VM/ESA, VMS)
1574 Emulated using multiple interpreters. See L<perlfork>. (Win32)
1576 Does not automatically flush output handles on some platforms.
1577 (SunOS, Solaris, HP-UX)
1581 Not implemented. (S<Mac OS>, S<RISC OS>)
1585 Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>)
1589 Not implemented. (S<Mac OS>, Win32, S<RISC OS>)
1591 =item getpriority WHICH,WHO
1593 Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>, VOS, VM/ESA)
1597 Not implemented. (S<Mac OS>, Win32)
1599 Not useful. (S<RISC OS>)
1603 Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>)
1605 =item getnetbyname NAME
1607 Not implemented. (S<Mac OS>, Win32, S<Plan 9>)
1611 Not implemented. (S<Mac OS>, Win32)
1613 Not useful. (S<RISC OS>)
1617 Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>)
1619 =item getnetbyaddr ADDR,ADDRTYPE
1621 Not implemented. (S<Mac OS>, Win32, S<Plan 9>)
1623 =item getprotobynumber NUMBER
1625 Not implemented. (S<Mac OS>)
1627 =item getservbyport PORT,PROTO
1629 Not implemented. (S<Mac OS>)
1633 Not implemented. (S<Mac OS>, Win32, VM/ESA)
1637 Not implemented. (S<Mac OS>, Win32, VMS, VM/ESA)
1641 Not implemented. (S<Mac OS>, Win32)
1645 Not implemented. (S<Mac OS>, Win32, S<Plan 9>)
1649 Not implemented. (S<Mac OS>, Win32, S<Plan 9>)
1653 Not implemented. (Win32, S<Plan 9>)
1655 =item sethostent STAYOPEN
1657 Not implemented. (S<Mac OS>, Win32, S<Plan 9>, S<RISC OS>)
1659 =item setnetent STAYOPEN
1661 Not implemented. (S<Mac OS>, Win32, S<Plan 9>, S<RISC OS>)
1663 =item setprotoent STAYOPEN
1665 Not implemented. (S<Mac OS>, Win32, S<Plan 9>, S<RISC OS>)
1667 =item setservent STAYOPEN
1669 Not implemented. (S<Plan 9>, Win32, S<RISC OS>)
1673 Not implemented. (S<Mac OS>, MPE/iX, VM/ESA, Win32)
1677 Not implemented. (S<Mac OS>, MPE/iX, S<RISC OS>, VM/ESA, VMS, Win32)
1681 Not implemented. (S<Mac OS>, Win32)
1685 Not implemented. (S<Mac OS>, Win32, S<Plan 9>)
1689 Not implemented. (S<Mac OS>, Win32, S<Plan 9>)
1693 Not implemented. (S<Plan 9>, Win32)
1695 =item getsockopt SOCKET,LEVEL,OPTNAME
1697 Not implemented. (S<Plan 9>)
1703 This operator is implemented via the File::Glob extension on most
1704 platforms. See L<File::Glob> for portability information.
1706 =item ioctl FILEHANDLE,FUNCTION,SCALAR
1708 Not implemented. (VMS)
1710 Available only for socket handles, and it does what the ioctlsocket() call
1711 in the Winsock API does. (Win32)
1713 Available only for socket handles. (S<RISC OS>)
1715 =item kill SIGNAL, LIST
1717 C<kill(0, LIST)> is implemented for the sake of taint checking;
1718 use with other signals is unimplemented. (S<Mac OS>)
1720 Not implemented, hence not useful for taint checking. (S<RISC OS>)
1722 C<kill()> doesn't have the semantics of C<raise()>, i.e. it doesn't send
1723 a signal to the identified process like it does on Unix platforms.
1724 Instead C<kill($sig, $pid)> terminates the process identified by $pid,
1725 and makes it exit immediately with exit status $sig. As in Unix, if
1726 $sig is 0 and the specified process exists, it returns true without
1727 actually terminating it. (Win32)
1729 =item link OLDFILE,NEWFILE
1731 Not implemented. (S<Mac OS>, MPE/iX, VMS, S<RISC OS>)
1733 Link count not updated because hard links are not quite that hard
1734 (They are sort of half-way between hard and soft links). (AmigaOS)
1736 Hard links are implemented on Win32 (Windows NT and Windows 2000)
1739 =item lstat FILEHANDLE
1745 Not implemented. (VMS, S<RISC OS>)
1747 Return values (especially for device and inode) may be bogus. (Win32)
1749 =item msgctl ID,CMD,ARG
1751 =item msgget KEY,FLAGS
1753 =item msgsnd ID,MSG,FLAGS
1755 =item msgrcv ID,VAR,SIZE,TYPE,FLAGS
1757 Not implemented. (S<Mac OS>, Win32, VMS, S<Plan 9>, S<RISC OS>, VOS)
1759 =item open FILEHANDLE,EXPR
1761 =item open FILEHANDLE
1763 The C<|> variants are supported only if ToolServer is installed.
1766 open to C<|-> and C<-|> are unsupported. (S<Mac OS>, Win32, S<RISC OS>)
1768 Opening a process does not automatically flush output handles on some
1769 platforms. (SunOS, Solaris, HP-UX)
1771 =item pipe READHANDLE,WRITEHANDLE
1773 Very limited functionality. (MiNT)
1779 Not implemented. (Win32, VMS, S<RISC OS>)
1781 =item select RBITS,WBITS,EBITS,TIMEOUT
1783 Only implemented on sockets. (Win32, VMS)
1785 Only reliable on sockets. (S<RISC OS>)
1787 Note that the C<select FILEHANDLE> form is generally portable.
1789 =item semctl ID,SEMNUM,CMD,ARG
1791 =item semget KEY,NSEMS,FLAGS
1793 =item semop KEY,OPSTRING
1795 Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>, VOS)
1799 Not implemented. (S<Mac OS>, MPE/iX, VMS, Win32, S<RISC OS>, VOS)
1801 =item setpgrp PID,PGRP
1803 Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>, VOS)
1805 =item setpriority WHICH,WHO,PRIORITY
1807 Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>, VOS)
1811 Not implemented. (S<Mac OS>, MPE/iX, Win32, S<RISC OS>, VOS)
1813 =item setsockopt SOCKET,LEVEL,OPTNAME,OPTVAL
1815 Not implemented. (S<Plan 9>)
1817 =item shmctl ID,CMD,ARG
1819 =item shmget KEY,SIZE,FLAGS
1821 =item shmread ID,VAR,POS,SIZE
1823 =item shmwrite ID,STRING,POS,SIZE
1825 Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>, VOS)
1827 =item sockatmark SOCKET
1829 A relatively recent addition to socket functions, may not
1830 be implemented even in UNIX platforms.
1832 =item socketpair SOCKET1,SOCKET2,DOMAIN,TYPE,PROTOCOL
1834 Not implemented. (Win32, VMS, S<RISC OS>, VOS, VM/ESA)
1836 =item stat FILEHANDLE
1842 Platforms that do not have rdev, blksize, or blocks will return these
1843 as '', so numeric comparison or manipulation of these fields may cause
1844 'not numeric' warnings.
1846 mtime and atime are the same thing, and ctime is creation time instead of
1847 inode change time. (S<Mac OS>).
1849 ctime not supported on UFS (S<Mac OS X>).
1851 ctime is creation time instead of inode change time (Win32).
1853 device and inode are not meaningful. (Win32)
1855 device and inode are not necessarily reliable. (VMS)
1857 mtime, atime and ctime all return the last modification time. Device and
1858 inode are not necessarily reliable. (S<RISC OS>)
1860 dev, rdev, blksize, and blocks are not available. inode is not
1861 meaningful and will differ between stat calls on the same file. (os2)
1863 some versions of cygwin when doing a stat("foo") and if not finding it
1864 may then attempt to stat("foo.exe") (Cygwin)
1866 =item symlink OLDFILE,NEWFILE
1868 Not implemented. (Win32, VMS, S<RISC OS>)
1872 Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>, VOS, VM/ESA)
1874 =item sysopen FILEHANDLE,FILENAME,MODE,PERMS
1876 The traditional "0", "1", and "2" MODEs are implemented with different
1877 numeric values on some systems. The flags exported by C<Fcntl>
1878 (O_RDONLY, O_WRONLY, O_RDWR) should work everywhere though. (S<Mac
1879 OS>, OS/390, VM/ESA)
1883 In general, do not assume the UNIX/POSIX semantics that you can shift
1884 C<$?> right by eight to get the exit value, or that C<$? & 127>
1885 would give you the number of the signal that terminated the program,
1886 or that C<$? & 128> would test true if the program was terminated by a
1887 coredump. Instead, use the POSIX W*() interfaces: for example, use
1888 WIFEXITED($?) and WEXITVALUE($?) to test for a normal exit and the exit
1889 value, WIFSIGNALED($?) and WTERMSIG($?) for a signal exit and the
1890 signal. Core dumping is not a portable concept, so there's no portable
1891 way to test for that.
1893 Only implemented if ToolServer is installed. (S<Mac OS>)
1895 As an optimization, may not call the command shell specified in
1896 C<$ENV{PERL5SHELL}>. C<system(1, @args)> spawns an external
1897 process and immediately returns its process designator, without
1898 waiting for it to terminate. Return value may be used subsequently
1899 in C<wait> or C<waitpid>. Failure to spawn() a subprocess is indicated
1900 by setting $? to "255 << 8". C<$?> is set in a way compatible with
1901 Unix (i.e. the exitstatus of the subprocess is obtained by "$? >> 8",
1902 as described in the documentation). (Win32)
1904 There is no shell to process metacharacters, and the native standard is
1905 to pass a command line terminated by "\n" "\r" or "\0" to the spawned
1906 program. Redirection such as C<< > foo >> is performed (if at all) by
1907 the run time library of the spawned program. C<system> I<list> will call
1908 the Unix emulation library's C<exec> emulation, which attempts to provide
1909 emulation of the stdin, stdout, stderr in force in the parent, providing
1910 the child program uses a compatible version of the emulation library.
1911 I<scalar> will call the native command line direct and no such emulation
1912 of a child Unix program will exists. Mileage B<will> vary. (S<RISC OS>)
1914 Far from being POSIX compliant. Because there may be no underlying
1915 /bin/sh tries to work around the problem by forking and execing the
1916 first token in its argument string. Handles basic redirection
1917 ("<" or ">") on its own behalf. (MiNT)
1919 Does not automatically flush output handles on some platforms.
1920 (SunOS, Solaris, HP-UX)
1922 The return value is POSIX-like (shifted up by 8 bits), which only allows
1923 room for a made-up value derived from the severity bits of the native
1924 32-bit condition code (unless overridden by C<use vmsish 'status'>).
1925 For more details see L<perlvms/$?>. (VMS)
1929 Only the first entry returned is nonzero. (S<Mac OS>)
1931 "cumulative" times will be bogus. On anything other than Windows NT
1932 or Windows 2000, "system" time will be bogus, and "user" time is
1933 actually the time returned by the clock() function in the C runtime
1936 Not useful. (S<RISC OS>)
1938 =item truncate FILEHANDLE,LENGTH
1940 =item truncate EXPR,LENGTH
1942 Not implemented. (Older versions of VMS)
1944 Truncation to same-or-shorter lengths only. (VOS)
1946 If a FILEHANDLE is supplied, it must be writable and opened in append
1947 mode (i.e., use C<<< open(FH, '>>filename') >>>
1948 or C<sysopen(FH,...,O_APPEND|O_RDWR)>. If a filename is supplied, it
1949 should not be held open elsewhere. (Win32)
1955 Returns undef where unavailable, as of version 5.005.
1957 C<umask> works but the correct permissions are set only when the file
1958 is finally closed. (AmigaOS)
1962 Only the modification time is updated. (S<BeOS>, S<Mac OS>, VMS, S<RISC OS>)
1964 May not behave as expected. Behavior depends on the C runtime
1965 library's implementation of utime(), and the filesystem being
1966 used. The FAT filesystem typically does not support an "access
1967 time" field, and it may limit timestamps to a granularity of
1968 two seconds. (Win32)
1972 =item waitpid PID,FLAGS
1974 Not implemented. (S<Mac OS>)
1976 Can only be applied to process handles returned for processes spawned
1977 using C<system(1, ...)> or pseudo processes created with C<fork()>. (Win32)
1979 Not useful. (S<RISC OS>)
1987 =item v1.49, 12 August 2002
1989 Updates for VOS from Paul Green.
1991 =item v1.48, 02 February 2001
1993 Various updates from perl5-porters over the past year, supported
1994 platforms update from Jarkko Hietaniemi.
1996 =item v1.47, 22 March 2000
1998 Various cleanups from Tom Christiansen, including migration of
1999 long platform listings from L<perl>.
2001 =item v1.46, 12 February 2000
2003 Updates for VOS and MPE/iX. (Peter Prymmer) Other small changes.
2005 =item v1.45, 20 December 1999
2007 Small changes from 5.005_63 distribution, more changes to EBCDIC info.
2009 =item v1.44, 19 July 1999
2011 A bunch of updates from Peter Prymmer for C<$^O> values,
2012 endianness, File::Spec, VMS, BS2000, OS/400.
2014 =item v1.43, 24 May 1999
2016 Added a lot of cleaning up from Tom Christiansen.
2018 =item v1.42, 22 May 1999
2020 Added notes about tests, sprintf/printf, and epoch offsets.
2022 =item v1.41, 19 May 1999
2024 Lots more little changes to formatting and content.
2026 Added a bunch of C<$^O> and related values
2027 for various platforms; fixed mail and web addresses, and added
2028 and changed miscellaneous notes. (Peter Prymmer)
2030 =item v1.40, 11 April 1999
2032 Miscellaneous changes.
2034 =item v1.39, 11 February 1999
2036 Changes from Jarkko and EMX URL fixes Michael Schwern. Additional
2037 note about newlines added.
2039 =item v1.38, 31 December 1998
2041 More changes from Jarkko.
2043 =item v1.37, 19 December 1998
2045 More minor changes. Merge two separate version 1.35 documents.
2047 =item v1.36, 9 September 1998
2049 Updated for Stratus VOS. Also known as version 1.35.
2051 =item v1.35, 13 August 1998
2053 Integrate more minor changes, plus addition of new sections under
2054 L<"ISSUES">: L<"Numbers endianness and Width">,
2055 L<"Character sets and character encoding">,
2056 L<"Internationalisation">.
2058 =item v1.33, 06 August 1998
2060 Integrate more minor changes.
2062 =item v1.32, 05 August 1998
2064 Integrate more minor changes.
2066 =item v1.30, 03 August 1998
2068 Major update for RISC OS, other minor changes.
2070 =item v1.23, 10 July 1998
2072 First public release with perl5.005.
2076 =head1 Supported Platforms
2078 As of July 2002 (the Perl release 5.8.0), the following platforms are
2079 able to build Perl from the standard source code distribution
2080 available at http://www.cpan.org/src/index.html
2103 OS/400 (using the PASE) (since Perl 5.8.1/5.9.0)
2110 Tru64 UNIX (DEC OSF/1, Digital UNIX)
2115 Win95/98/ME/2K/XP 2)
2120 1) in DOS mode either the DOS or OS/2 ports can be used
2121 2) compilers: Borland, MinGW (GCC), VC6
2123 The following platforms worked with the previous releases (5.6 and
2124 5.7), but we did not manage either to fix or to test these in time
2125 for the 5.8.0 release. There is a very good chance that many of these
2126 will work fine with the 5.8.0.
2139 Known to be broken for 5.8.0 (but 5.6.1 and 5.7.2 can be used):
2143 The following platforms have been known to build Perl from source in
2144 the past (5.005_03 and earlier), but we haven't been able to verify
2145 their status for the current release, either because the
2146 hardware/software platforms are rare or because we don't have an
2147 active champion on these platforms--or both. They used to work,
2148 though, so go ahead and try compiling them, and let perlbug@perl.org
2182 The following platforms have their own source code distributions and
2183 binaries available via http://www.cpan.org/ports/
2187 OS/400 (ILE) 5.005_02
2188 Tandem Guardian 5.004
2190 The following platforms have only binaries available via
2191 http://www.cpan.org/ports/index.html :
2195 Acorn RISCOS 5.005_02
2199 Although we do suggest that you always build your own Perl from
2200 the source code, both for maximal configurability and for security,
2201 in case you are in a hurry you can check
2202 http://www.cpan.org/ports/index.html for binary distributions.
2206 L<perlaix>, L<perlamiga>, L<perlapollo>, L<perlbeos>, L<perlbs2000>,
2207 L<perlce>, L<perlcygwin>, L<perldgux>, L<perldos>, L<perlepoc>,
2208 L<perlebcdic>, L<perlfreebsd>, L<perlhurd>, L<perlhpux>, L<perlirix>,
2209 L<perlmachten>, L<perlmacos>, L<perlmint>, L<perlmpeix>,
2210 L<perlnetware>, L<perlos2>, L<perlos390>, L<perlos400>,
2211 L<perlplan9>, L<perlqnx>, L<perlsolaris>, L<perltru64>,
2212 L<perlunicode>, L<perlvmesa>, L<perlvms>, L<perlvos>,
2213 L<perlwin32>, and L<Win32>.
2215 =head1 AUTHORS / CONTRIBUTORS
2217 Abigail <abigail@foad.org>,
2218 Charles Bailey <bailey@newman.upenn.edu>,
2219 Graham Barr <gbarr@pobox.com>,
2220 Tom Christiansen <tchrist@perl.com>,
2221 Nicholas Clark <nick@ccl4.org>,
2222 Thomas Dorner <Thomas.Dorner@start.de>,
2223 Andy Dougherty <doughera@lafayette.edu>,
2224 Dominic Dunlop <domo@computer.org>,
2225 Neale Ferguson <neale@vma.tabnsw.com.au>,
2226 David J. Fiander <davidf@mks.com>,
2227 Paul Green <Paul.Green@stratus.com>,
2228 M.J.T. Guy <mjtg@cam.ac.uk>,
2229 Jarkko Hietaniemi <jhi@iki.fi>,
2230 Luther Huffman <lutherh@stratcom.com>,
2231 Nick Ing-Simmons <nick@ing-simmons.net>,
2232 Andreas J. KE<ouml>nig <a.koenig@mind.de>,
2233 Markus Laker <mlaker@contax.co.uk>,
2234 Andrew M. Langmead <aml@world.std.com>,
2235 Larry Moore <ljmoore@freespace.net>,
2236 Paul Moore <Paul.Moore@uk.origin-it.com>,
2237 Chris Nandor <pudge@pobox.com>,
2238 Matthias Neeracher <neeracher@mac.com>,
2239 Philip Newton <pne@cpan.org>,
2240 Gary Ng <71564.1743@CompuServe.COM>,
2241 Tom Phoenix <rootbeer@teleport.com>,
2242 AndrE<eacute> Pirard <A.Pirard@ulg.ac.be>,
2243 Peter Prymmer <pvhp@forte.com>,
2244 Hugo van der Sanden <hv@crypt0.demon.co.uk>,
2245 Gurusamy Sarathy <gsar@activestate.com>,
2246 Paul J. Schinder <schinder@pobox.com>,
2247 Michael G Schwern <schwern@pobox.com>,
2248 Dan Sugalski <dan@sidhe.org>,
2249 Nathan Torkington <gnat@frii.com>.