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 \cU eq chr(21) eq CP-1047 21
196 LF eq \045 eq \x25 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;}
492 Don't assume that you can reach the public Internet.
494 Don't assume that there is only one way to get through firewalls
495 to the public Internet.
497 Don't assume that you can reach yourself or any node by the name
498 'localhost'. The same goes for '127.0.0.1'.
500 Don't assume that any particular port (service) will respond.
502 Don't assume that you can ping hosts and get replies.
504 =head2 Interprocess Communication (IPC)
506 In general, don't directly access the system in code meant to be
507 portable. That means, no C<system>, C<exec>, C<fork>, C<pipe>,
508 C<``>, C<qx//>, C<open> with a C<|>, nor any of the other things
509 that makes being a perl hacker worth being.
511 Commands that launch external processes are generally supported on
512 most platforms (though many of them do not support any type of
513 forking). The problem with using them arises from what you invoke
514 them on. External tools are often named differently on different
515 platforms, may not be available in the same location, might accept
516 different arguments, can behave differently, and often present their
517 results in a platform-dependent way. Thus, you should seldom depend
518 on them to produce consistent results. (Then again, if you're calling
519 I<netstat -a>, you probably don't expect it to run on both Unix and CP/M.)
521 One especially common bit of Perl code is opening a pipe to B<sendmail>:
523 open(MAIL, '|/usr/lib/sendmail -t')
524 or die "cannot fork sendmail: $!";
526 This is fine for systems programming when sendmail is known to be
527 available. But it is not fine for many non-Unix systems, and even
528 some Unix systems that may not have sendmail installed. If a portable
529 solution is needed, see the various distributions on CPAN that deal
530 with it. Mail::Mailer and Mail::Send in the MailTools distribution are
531 commonly used, and provide several mailing methods, including mail,
532 sendmail, and direct SMTP (via Net::SMTP) if a mail transfer agent is
533 not available. Mail::Sendmail is a standalone module that provides
534 simple, platform-independent mailing.
536 The Unix System V IPC (C<msg*(), sem*(), shm*()>) is not available
537 even on all Unix platforms.
539 Do not use either the bare result of C<pack("N", 10, 20, 30, 40)> or
540 bare v-strings (such as C<v10.20.30.40>) to represent IPv4 addresses:
541 both forms just pack the four bytes into network order. That this
542 would be equal to the C language C<in_addr> struct (which is what the
543 socket code internally uses) is not guaranteed. To be portable use
544 the routines of the Socket extension, such as C<inet_aton()>,
545 C<inet_ntoa()>, and C<sockaddr_in()>.
547 The rule of thumb for portable code is: Do it all in portable Perl, or
548 use a module (that may internally implement it with platform-specific
549 code, but expose a common interface).
551 =head2 External Subroutines (XS)
553 XS code can usually be made to work with any platform, but dependent
554 libraries, header files, etc., might not be readily available or
555 portable, or the XS code itself might be platform-specific, just as Perl
556 code might be. If the libraries and headers are portable, then it is
557 normally reasonable to make sure the XS code is portable, too.
559 A different type of portability issue arises when writing XS code:
560 availability of a C compiler on the end-user's system. C brings
561 with it its own portability issues, and writing XS code will expose
562 you to some of those. Writing purely in Perl is an easier way to
565 =head2 Standard Modules
567 In general, the standard modules work across platforms. Notable
568 exceptions are the CPAN module (which currently makes connections to external
569 programs that may not be available), platform-specific modules (like
570 ExtUtils::MM_VMS), and DBM modules.
572 There is no one DBM module available on all platforms.
573 SDBM_File and the others are generally available on all Unix and DOSish
574 ports, but not in MacPerl, where only NBDM_File and DB_File are
577 The good news is that at least some DBM module should be available, and
578 AnyDBM_File will use whichever module it can find. Of course, then
579 the code needs to be fairly strict, dropping to the greatest common
580 factor (e.g., not exceeding 1K for each record), so that it will
581 work with any DBM module. See L<AnyDBM_File> for more details.
585 The system's notion of time of day and calendar date is controlled in
586 widely different ways. Don't assume the timezone is stored in C<$ENV{TZ}>,
587 and even if it is, don't assume that you can control the timezone through
588 that variable. Don't assume anything about the three-letter timezone
589 abbreviations (for example that MST would be the Mountain Standard Time,
590 it's been known to stand for Moscow Standard Time). If you need to
591 use timezones, express them in some unambiguous format like the
592 exact number of minutes offset from UTC, or the POSIX timezone
595 Don't assume that the epoch starts at 00:00:00, January 1, 1970,
596 because that is OS- and implementation-specific. It is better to
597 store a date in an unambiguous representation. The ISO 8601 standard
598 defines YYYY-MM-DD as the date format, or YYYY-MM-DDTHH-MM-SS
599 (that's a literal "T" separating the date from the time).
600 Please do use the ISO 8601 instead of making us to guess what
601 date 02/03/04 might be. ISO 8601 even sorts nicely as-is.
602 A text representation (like "1987-12-18") can be easily converted
603 into an OS-specific value using a module like Date::Parse.
604 An array of values, such as those returned by C<localtime>, can be
605 converted to an OS-specific representation using Time::Local.
607 When calculating specific times, such as for tests in time or date modules,
608 it may be appropriate to calculate an offset for the epoch.
611 $offset = Time::Local::timegm(0, 0, 0, 1, 0, 70);
613 The value for C<$offset> in Unix will be C<0>, but in Mac OS will be
614 some large number. C<$offset> can then be added to a Unix time value
615 to get what should be the proper value on any system.
617 On Windows (at least), you shouldn't pass a negative value to C<gmtime> or
620 =head2 Character sets and character encoding
622 Assume very little about character sets.
624 Assume nothing about numerical values (C<ord>, C<chr>) of characters.
625 Do not use explicit code point ranges (like \xHH-\xHH); use for
626 example symbolic character classes like C<[:print:]>.
628 Do not assume that the alphabetic characters are encoded contiguously
629 (in the numeric sense). There may be gaps.
631 Do not assume anything about the ordering of the characters.
632 The lowercase letters may come before or after the uppercase letters;
633 the lowercase and uppercase may be interlaced so that both `a' and `A'
634 come before `b'; the accented and other international characters may
635 be interlaced so that E<auml> comes before `b'.
637 =head2 Internationalisation
639 If you may assume POSIX (a rather large assumption), you may read
640 more about the POSIX locale system from L<perllocale>. The locale
641 system at least attempts to make things a little bit more portable,
642 or at least more convenient and native-friendly for non-English
643 users. The system affects character sets and encoding, and date
644 and time formatting--amongst other things.
646 If you really want to be international, you should consider Unicode.
647 See L<perluniintro> and L<perlunicode> for more information.
649 If you want to use non-ASCII bytes (outside the bytes 0x00..0x7f) in
650 the "source code" of your code, to be portable you have to be explicit
651 about what bytes they are. Someone might for example be using your
652 code under a UTF-8 locale, in which case random native bytes might be
653 illegal ("Malformed UTF-8 ...") This means that for example embedding
654 ISO 8859-1 bytes beyond 0x7f into your strings might cause trouble
655 later. If the bytes are native 8-bit bytes, you can use the C<bytes>
656 pragma. If the bytes are in a string (regular expression being a
657 curious string), you can often also use the C<\xHH> notation instead
658 of embedding the bytes as-is. If they are in some particular legacy
659 encoding (ether single-byte or something more complicated), you can
660 use the C<encoding> pragma. (If you want to write your code in UTF-8,
661 you can use either the C<utf8> pragma, or the C<encoding> pragma.)
662 The C<bytes> and C<utf8> pragmata are available since Perl 5.6.0, and
663 the C<encoding> pragma since Perl 5.8.0.
665 =head2 System Resources
667 If your code is destined for systems with severely constrained (or
668 missing!) virtual memory systems then you want to be I<especially> mindful
669 of avoiding wasteful constructs such as:
671 # NOTE: this is no longer "bad" in perl5.005
672 for (0..10000000) {} # bad
673 for (my $x = 0; $x <= 10000000; ++$x) {} # good
675 @lines = <VERY_LARGE_FILE>; # bad
677 while (<FILE>) {$file .= $_} # sometimes bad
678 $file = join('', <FILE>); # better
680 The last two constructs may appear unintuitive to most people. The
681 first repeatedly grows a string, whereas the second allocates a
682 large chunk of memory in one go. On some systems, the second is
683 more efficient that the first.
687 Most multi-user platforms provide basic levels of security, usually
688 implemented at the filesystem level. Some, however, do
689 not-- unfortunately. Thus the notion of user id, or "home" directory,
690 or even the state of being logged-in, may be unrecognizable on many
691 platforms. If you write programs that are security-conscious, it
692 is usually best to know what type of system you will be running
693 under so that you can write code explicitly for that platform (or
696 Don't assume the UNIX filesystem access semantics: the operating
697 system or the filesystem may be using some ACL systems, which are
698 richer languages than the usual rwx. Even if the rwx exist,
699 their semantics might be different.
701 (From security viewpoint testing for permissions before attempting to
702 do something is silly anyway: if one tries this, there is potential
703 for race conditions-- someone or something might change the
704 permissions between the permissions check and the actual operation.
705 Just try the operation.)
707 Don't assume the UNIX user and group semantics: especially, don't
708 expect the C<< $< >> and C<< $> >> (or the C<$(> and C<$)>) to work
709 for switching identities (or memberships).
711 Don't assume set-uid and set-gid semantics. (And even if you do,
712 think twice: set-uid and set-gid are a known can of security worms.)
716 For those times when it is necessary to have platform-specific code,
717 consider keeping the platform-specific code in one place, making porting
718 to other platforms easier. Use the Config module and the special
719 variable C<$^O> to differentiate platforms, as described in
722 Be careful in the tests you supply with your module or programs.
723 Module code may be fully portable, but its tests might not be. This
724 often happens when tests spawn off other processes or call external
725 programs to aid in the testing, or when (as noted above) the tests
726 assume certain things about the filesystem and paths. Be careful not
727 to depend on a specific output style for errors, such as when checking
728 C<$!> after a failed system call. Using C<$!> for anything else than
729 displaying it as output is doubtful (though see the Errno module for
730 testing reasonably portably for error value). Some platforms expect
731 a certain output format, and Perl on those platforms may have been
732 adjusted accordingly. Most specifically, don't anchor a regex when
733 testing an error value.
737 Modules uploaded to CPAN are tested by a variety of volunteers on
738 different platforms. These CPAN testers are notified by mail of each
739 new upload, and reply to the list with PASS, FAIL, NA (not applicable to
740 this platform), or UNKNOWN (unknown), along with any relevant notations.
742 The purpose of the testing is twofold: one, to help developers fix any
743 problems in their code that crop up because of lack of testing on other
744 platforms; two, to provide users with information about whether
745 a given module works on a given platform.
749 =item Mailing list: cpan-testers@perl.org
751 =item Testing results: http://testers.cpan.org/
757 As of version 5.002, Perl is built with a C<$^O> variable that
758 indicates the operating system it was built on. This was implemented
759 to help speed up code that would otherwise have to C<use Config>
760 and use the value of C<$Config{osname}>. Of course, to get more
761 detailed information about the system, looking into C<%Config> is
762 certainly recommended.
764 C<%Config> cannot always be trusted, however, because it was built
765 at compile time. If perl was built in one place, then transferred
766 elsewhere, some values may be wrong. The values may even have been
767 edited after the fact.
771 Perl works on a bewildering variety of Unix and Unix-like platforms (see
772 e.g. most of the files in the F<hints/> directory in the source code kit).
773 On most of these systems, the value of C<$^O> (hence C<$Config{'osname'}>,
774 too) is determined either by lowercasing and stripping punctuation from the
775 first field of the string returned by typing C<uname -a> (or a similar command)
776 at the shell prompt or by testing the file system for the presence of
777 uniquely named files such as a kernel or header file. Here, for example,
778 are a few of the more popular Unix flavors:
780 uname $^O $Config{'archname'}
781 --------------------------------------------
783 BSD/OS bsdos i386-bsdos
785 dgux dgux AViiON-dgux
786 DYNIX/ptx dynixptx i386-dynixptx
787 FreeBSD freebsd freebsd-i386
788 Linux linux arm-linux
789 Linux linux i386-linux
790 Linux linux i586-linux
791 Linux linux ppc-linux
792 HP-UX hpux PA-RISC1.1
794 Mac OS X darwin darwin
795 MachTen PPC machten powerpc-machten
797 NeXT 4 next OPENSTEP-Mach
798 openbsd openbsd i386-openbsd
799 OSF1 dec_osf alpha-dec_osf
800 reliantunix-n svr4 RM400-svr4
801 SCO_SV sco_sv i386-sco_sv
802 SINIX-N svr4 RM400-svr4
803 sn4609 unicos CRAY_C90-unicos
804 sn6521 unicosmk t3e-unicosmk
805 sn9617 unicos CRAY_J90-unicos
806 SunOS solaris sun4-solaris
807 SunOS solaris i86pc-solaris
808 SunOS4 sunos sun4-sunos
810 Because the value of C<$Config{archname}> may depend on the
811 hardware architecture, it can vary more than the value of C<$^O>.
813 =head2 DOS and Derivatives
815 Perl has long been ported to Intel-style microcomputers running under
816 systems like PC-DOS, MS-DOS, OS/2, and most Windows platforms you can
817 bring yourself to mention (except for Windows CE, if you count that).
818 Users familiar with I<COMMAND.COM> or I<CMD.EXE> style shells should
819 be aware that each of these file specifications may have subtle
822 $filespec0 = "c:/foo/bar/file.txt";
823 $filespec1 = "c:\\foo\\bar\\file.txt";
824 $filespec2 = 'c:\foo\bar\file.txt';
825 $filespec3 = 'c:\\foo\\bar\\file.txt';
827 System calls accept either C</> or C<\> as the path separator.
828 However, many command-line utilities of DOS vintage treat C</> as
829 the option prefix, so may get confused by filenames containing C</>.
830 Aside from calling any external programs, C</> will work just fine,
831 and probably better, as it is more consistent with popular usage,
832 and avoids the problem of remembering what to backwhack and what
835 The DOS FAT filesystem can accommodate only "8.3" style filenames. Under
836 the "case-insensitive, but case-preserving" HPFS (OS/2) and NTFS (NT)
837 filesystems you may have to be careful about case returned with functions
838 like C<readdir> or used with functions like C<open> or C<opendir>.
840 DOS also treats several filenames as special, such as AUX, PRN,
841 NUL, CON, COM1, LPT1, LPT2, etc. Unfortunately, sometimes these
842 filenames won't even work if you include an explicit directory
843 prefix. It is best to avoid such filenames, if you want your code
844 to be portable to DOS and its derivatives. It's hard to know what
845 these all are, unfortunately.
847 Users of these operating systems may also wish to make use of
848 scripts such as I<pl2bat.bat> or I<pl2cmd> to
849 put wrappers around your scripts.
851 Newline (C<\n>) is translated as C<\015\012> by STDIO when reading from
852 and writing to files (see L<"Newlines">). C<binmode(FILEHANDLE)>
853 will keep C<\n> translated as C<\012> for that filehandle. Since it is a
854 no-op on other systems, C<binmode> should be used for cross-platform code
855 that deals with binary data. That's assuming you realize in advance
856 that your data is in binary. General-purpose programs should
857 often assume nothing about their data.
859 The C<$^O> variable and the C<$Config{archname}> values for various
860 DOSish perls are as follows:
862 OS $^O $Config{archname} ID Version
863 --------------------------------------------------------
867 Windows 3.1 ? ? 0 3 01
868 Windows 95 MSWin32 MSWin32-x86 1 4 00
869 Windows 98 MSWin32 MSWin32-x86 1 4 10
870 Windows ME MSWin32 MSWin32-x86 1 ?
871 Windows NT MSWin32 MSWin32-x86 2 4 xx
872 Windows NT MSWin32 MSWin32-ALPHA 2 4 xx
873 Windows NT MSWin32 MSWin32-ppc 2 4 xx
874 Windows 2000 MSWin32 MSWin32-x86 2 5 xx
875 Windows XP MSWin32 MSWin32-x86 2 ?
876 Windows CE MSWin32 ? 3
879 The various MSWin32 Perl's can distinguish the OS they are running on
880 via the value of the fifth element of the list returned from
881 Win32::GetOSVersion(). For example:
883 if ($^O eq 'MSWin32') {
884 my @os_version_info = Win32::GetOSVersion();
885 print +('3.1','95','NT')[$os_version_info[4]],"\n";
888 There are also Win32::IsWinNT() and Win32::IsWin95(), try C<perldoc Win32>,
889 and as of libwin32 0.19 (not part of the core Perl distribution)
890 Win32::GetOSName(). The very portable POSIX::uname() will work too:
892 c:\> perl -MPOSIX -we "print join '|', uname"
893 Windows NT|moonru|5.0|Build 2195 (Service Pack 2)|x86
901 The djgpp environment for DOS, http://www.delorie.com/djgpp/
906 The EMX environment for DOS, OS/2, etc. emx@iaehv.nl,
907 http://www.leo.org/pub/comp/os/os2/leo/gnu/emx+gcc/index.html or
908 ftp://hobbes.nmsu.edu/pub/os2/dev/emx/ Also L<perlos2>.
912 Build instructions for Win32 in L<perlwin32>, or under the Cygnus environment
917 The C<Win32::*> modules in L<Win32>.
921 The ActiveState Pages, http://www.activestate.com/
925 The Cygwin environment for Win32; F<README.cygwin> (installed
926 as L<perlcygwin>), http://www.cygwin.com/
930 The U/WIN environment for Win32,
931 http://www.research.att.com/sw/tools/uwin/
935 Build instructions for OS/2, L<perlos2>
941 Any module requiring XS compilation is right out for most people, because
942 MacPerl is built using non-free (and non-cheap!) compilers. Some XS
943 modules that can work with MacPerl are built and distributed in binary
946 Directories are specified as:
948 volume:folder:file for absolute pathnames
949 volume:folder: for absolute pathnames
950 :folder:file for relative pathnames
951 :folder: for relative pathnames
952 :file for relative pathnames
953 file for relative pathnames
955 Files are stored in the directory in alphabetical order. Filenames are
956 limited to 31 characters, and may include any character except for
957 null and C<:>, which is reserved as the path separator.
959 Instead of C<flock>, see C<FSpSetFLock> and C<FSpRstFLock> in the
960 Mac::Files module, or C<chmod(0444, ...)> and C<chmod(0666, ...)>.
962 In the MacPerl application, you can't run a program from the command line;
963 programs that expect C<@ARGV> to be populated can be edited with something
964 like the following, which brings up a dialog box asking for the command
968 @ARGV = split /\s+/, MacPerl::Ask('Arguments?');
971 A MacPerl script saved as a "droplet" will populate C<@ARGV> with the full
972 pathnames of the files dropped onto the script.
974 Mac users can run programs under a type of command line interface
975 under MPW (Macintosh Programmer's Workshop, a free development
976 environment from Apple). MacPerl was first introduced as an MPW
977 tool, and MPW can be used like a shell:
979 perl myscript.plx some arguments
981 ToolServer is another app from Apple that provides access to MPW tools
982 from MPW and the MacPerl app, which allows MacPerl programs to use
983 C<system>, backticks, and piped C<open>.
985 "S<Mac OS>" is the proper name for the operating system, but the value
986 in C<$^O> is "MacOS". To determine architecture, version, or whether
987 the application or MPW tool version is running, check:
989 $is_app = $MacPerl::Version =~ /App/;
990 $is_tool = $MacPerl::Version =~ /MPW/;
991 ($version) = $MacPerl::Version =~ /^(\S+)/;
992 $is_ppc = $MacPerl::Architecture eq 'MacPPC';
993 $is_68k = $MacPerl::Architecture eq 'Mac68K';
995 S<Mac OS X>, based on NeXT's OpenStep OS, runs MacPerl natively, under the
996 "Classic" environment. There is no "Carbon" version of MacPerl to run
997 under the primary Mac OS X environment. S<Mac OS X> and its Open Source
998 version, Darwin, both run Unix perl natively.
1006 MacPerl Development, http://dev.macperl.org/ .
1010 The MacPerl Pages, http://www.macperl.com/ .
1014 The MacPerl mailing lists, http://lists.perl.org/ .
1020 Perl on VMS is discussed in L<perlvms> in the perl distribution.
1021 Perl on VMS can accept either VMS- or Unix-style file
1022 specifications as in either of the following:
1024 $ perl -ne "print if /perl_setup/i" SYS$LOGIN:LOGIN.COM
1025 $ perl -ne "print if /perl_setup/i" /sys$login/login.com
1027 but not a mixture of both as in:
1029 $ perl -ne "print if /perl_setup/i" sys$login:/login.com
1030 Can't open sys$login:/login.com: file specification syntax error
1032 Interacting with Perl from the Digital Command Language (DCL) shell
1033 often requires a different set of quotation marks than Unix shells do.
1036 $ perl -e "print ""Hello, world.\n"""
1039 There are several ways to wrap your perl scripts in DCL F<.COM> files, if
1040 you are so inclined. For example:
1042 $ write sys$output "Hello from DCL!"
1044 $ then perl -x 'f$environment("PROCEDURE")
1045 $ else perl -x - 'p1 'p2 'p3 'p4 'p5 'p6 'p7 'p8
1046 $ deck/dollars="__END__"
1049 print "Hello from Perl!\n";
1054 Do take care with C<$ ASSIGN/nolog/user SYS$COMMAND: SYS$INPUT> if your
1055 perl-in-DCL script expects to do things like C<< $read = <STDIN>; >>.
1057 Filenames are in the format "name.extension;version". The maximum
1058 length for filenames is 39 characters, and the maximum length for
1059 extensions is also 39 characters. Version is a number from 1 to
1060 32767. Valid characters are C</[A-Z0-9$_-]/>.
1062 VMS's RMS filesystem is case-insensitive and does not preserve case.
1063 C<readdir> returns lowercased filenames, but specifying a file for
1064 opening remains case-insensitive. Files without extensions have a
1065 trailing period on them, so doing a C<readdir> with a file named F<A.;5>
1066 will return F<a.> (though that file could be opened with
1069 RMS had an eight level limit on directory depths from any rooted logical
1070 (allowing 16 levels overall) prior to VMS 7.2. Hence
1071 C<PERL_ROOT:[LIB.2.3.4.5.6.7.8]> is a valid directory specification but
1072 C<PERL_ROOT:[LIB.2.3.4.5.6.7.8.9]> is not. F<Makefile.PL> authors might
1073 have to take this into account, but at least they can refer to the former
1074 as C</PERL_ROOT/lib/2/3/4/5/6/7/8/>.
1076 The VMS::Filespec module, which gets installed as part of the build
1077 process on VMS, is a pure Perl module that can easily be installed on
1078 non-VMS platforms and can be helpful for conversions to and from RMS
1081 What C<\n> represents depends on the type of file opened. It usually
1082 represents C<\012> but it could also be C<\015>, C<\012>, C<\015\012>,
1083 C<\000>, C<\040>, or nothing depending on the file organiztion and
1084 record format. The VMS::Stdio module provides access to the
1085 special fopen() requirements of files with unusual attributes on VMS.
1087 TCP/IP stacks are optional on VMS, so socket routines might not be
1088 implemented. UDP sockets may not be supported.
1090 The value of C<$^O> on OpenVMS is "VMS". To determine the architecture
1091 that you are running on without resorting to loading all of C<%Config>
1092 you can examine the content of the C<@INC> array like so:
1094 if (grep(/VMS_AXP/, @INC)) {
1095 print "I'm on Alpha!\n";
1097 } elsif (grep(/VMS_VAX/, @INC)) {
1098 print "I'm on VAX!\n";
1101 print "I'm not so sure about where $^O is...\n";
1104 On VMS, perl determines the UTC offset from the C<SYS$TIMEZONE_DIFFERENTIAL>
1105 logical name. Although the VMS epoch began at 17-NOV-1858 00:00:00.00,
1106 calls to C<localtime> are adjusted to count offsets from
1107 01-JAN-1970 00:00:00.00, just like Unix.
1115 F<README.vms> (installed as L<README_vms>), L<perlvms>
1119 vmsperl list, majordomo@perl.org
1121 (Put the words C<subscribe vmsperl> in message body.)
1125 vmsperl on the web, http://www.sidhe.org/vmsperl/index.html
1131 Perl on VOS is discussed in F<README.vos> in the perl distribution
1132 (installed as L<perlvos>). Perl on VOS can accept either VOS- or
1133 Unix-style file specifications as in either of the following:
1135 C<< $ perl -ne "print if /perl_setup/i" >system>notices >>
1136 C<< $ perl -ne "print if /perl_setup/i" /system/notices >>
1138 or even a mixture of both as in:
1140 C<< $ perl -ne "print if /perl_setup/i" >system/notices >>
1142 Even though VOS allows the slash character to appear in object
1143 names, because the VOS port of Perl interprets it as a pathname
1144 delimiting character, VOS files, directories, or links whose names
1145 contain a slash character cannot be processed. Such files must be
1146 renamed before they can be processed by Perl. Note that VOS limits
1147 file names to 32 or fewer characters.
1149 The value of C<$^O> on VOS is "VOS". To determine the architecture that
1150 you are running on without resorting to loading all of C<%Config> you
1151 can examine the content of the @INC array like so:
1154 print "I'm on a Stratus box!\n";
1156 print "I'm not on a Stratus box!\n";
1166 F<README.vos> (installed as L<perlvos>)
1170 The VOS mailing list.
1172 There is no specific mailing list for Perl on VOS. You can post
1173 comments to the comp.sys.stratus newsgroup, or subscribe to the general
1174 Stratus mailing list. Send a letter with "subscribe Info-Stratus" in
1175 the message body to majordomo@list.stratagy.com.
1179 VOS Perl on the web at http://ftp.stratus.com/pub/vos/posix/posix.html
1183 =head2 EBCDIC Platforms
1185 Recent versions of Perl have been ported to platforms such as OS/400 on
1186 AS/400 minicomputers as well as OS/390, VM/ESA, and BS2000 for S/390
1187 Mainframes. Such computers use EBCDIC character sets internally (usually
1188 Character Code Set ID 0037 for OS/400 and either 1047 or POSIX-BC for S/390
1189 systems). On the mainframe perl currently works under the "Unix system
1190 services for OS/390" (formerly known as OpenEdition), VM/ESA OpenEdition, or
1191 the BS200 POSIX-BC system (BS2000 is supported in perl 5.6 and greater).
1192 See L<perlos390> for details. Note that for OS/400 there is also a port of
1193 Perl 5.8.1/5.9.0 or later to the PASE which is ASCII-based (as opposed to
1194 ILE which is EBCDIC-based), see L<perlos400>.
1196 As of R2.5 of USS for OS/390 and Version 2.3 of VM/ESA these Unix
1197 sub-systems do not support the C<#!> shebang trick for script invocation.
1198 Hence, on OS/390 and VM/ESA perl scripts can be executed with a header
1199 similar to the following simple script:
1202 eval 'exec /usr/local/bin/perl -S $0 ${1+"$@"}'
1204 #!/usr/local/bin/perl # just a comment really
1206 print "Hello from perl!\n";
1208 OS/390 will support the C<#!> shebang trick in release 2.8 and beyond.
1209 Calls to C<system> and backticks can use POSIX shell syntax on all
1212 On the AS/400, if PERL5 is in your library list, you may need
1213 to wrap your perl scripts in a CL procedure to invoke them like so:
1216 CALL PGM(PERL5/PERL) PARM('/QOpenSys/hello.pl')
1219 This will invoke the perl script F<hello.pl> in the root of the
1220 QOpenSys file system. On the AS/400 calls to C<system> or backticks
1223 On these platforms, bear in mind that the EBCDIC character set may have
1224 an effect on what happens with some perl functions (such as C<chr>,
1225 C<pack>, C<print>, C<printf>, C<ord>, C<sort>, C<sprintf>, C<unpack>), as
1226 well as bit-fiddling with ASCII constants using operators like C<^>, C<&>
1227 and C<|>, not to mention dealing with socket interfaces to ASCII computers
1228 (see L<"Newlines">).
1230 Fortunately, most web servers for the mainframe will correctly
1231 translate the C<\n> in the following statement to its ASCII equivalent
1232 (C<\r> is the same under both Unix and OS/390 & VM/ESA):
1234 print "Content-type: text/html\r\n\r\n";
1236 The values of C<$^O> on some of these platforms includes:
1238 uname $^O $Config{'archname'}
1239 --------------------------------------------
1242 POSIX-BC posix-bc BS2000-posix-bc
1245 Some simple tricks for determining if you are running on an EBCDIC
1246 platform could include any of the following (perhaps all):
1248 if ("\t" eq "\05") { print "EBCDIC may be spoken here!\n"; }
1250 if (ord('A') == 193) { print "EBCDIC may be spoken here!\n"; }
1252 if (chr(169) eq 'z') { print "EBCDIC may be spoken here!\n"; }
1254 One thing you may not want to rely on is the EBCDIC encoding
1255 of punctuation characters since these may differ from code page to code
1256 page (and once your module or script is rumoured to work with EBCDIC,
1257 folks will want it to work with all EBCDIC character sets).
1267 L<perlos390>, F<README.os390>, F<perlbs2000>, F<README.vmesa>,
1272 The perl-mvs@perl.org list is for discussion of porting issues as well as
1273 general usage issues for all EBCDIC Perls. Send a message body of
1274 "subscribe perl-mvs" to majordomo@perl.org.
1278 AS/400 Perl information at
1279 http://as400.rochester.ibm.com/
1280 as well as on CPAN in the F<ports/> directory.
1284 =head2 Acorn RISC OS
1286 Because Acorns use ASCII with newlines (C<\n>) in text files as C<\012> like
1287 Unix, and because Unix filename emulation is turned on by default,
1288 most simple scripts will probably work "out of the box". The native
1289 filesystem is modular, and individual filesystems are free to be
1290 case-sensitive or insensitive, and are usually case-preserving. Some
1291 native filesystems have name length limits, which file and directory
1292 names are silently truncated to fit. Scripts should be aware that the
1293 standard filesystem currently has a name length limit of B<10>
1294 characters, with up to 77 items in a directory, but other filesystems
1295 may not impose such limitations.
1297 Native filenames are of the form
1299 Filesystem#Special_Field::DiskName.$.Directory.Directory.File
1303 Special_Field is not usually present, but may contain . and $ .
1304 Filesystem =~ m|[A-Za-z0-9_]|
1305 DsicName =~ m|[A-Za-z0-9_/]|
1306 $ represents the root directory
1307 . is the path separator
1308 @ is the current directory (per filesystem but machine global)
1309 ^ is the parent directory
1310 Directory and File =~ m|[^\0- "\.\$\%\&:\@\\^\|\177]+|
1312 The default filename translation is roughly C<tr|/.|./|;>
1314 Note that C<"ADFS::HardDisk.$.File" ne 'ADFS::HardDisk.$.File'> and that
1315 the second stage of C<$> interpolation in regular expressions will fall
1316 foul of the C<$.> if scripts are not careful.
1318 Logical paths specified by system variables containing comma-separated
1319 search lists are also allowed; hence C<System:Modules> is a valid
1320 filename, and the filesystem will prefix C<Modules> with each section of
1321 C<System$Path> until a name is made that points to an object on disk.
1322 Writing to a new file C<System:Modules> would be allowed only if
1323 C<System$Path> contains a single item list. The filesystem will also
1324 expand system variables in filenames if enclosed in angle brackets, so
1325 C<< <System$Dir>.Modules >> would look for the file
1326 S<C<$ENV{'System$Dir'} . 'Modules'>>. The obvious implication of this is
1327 that B<fully qualified filenames can start with C<< <> >>> and should
1328 be protected when C<open> is used for input.
1330 Because C<.> was in use as a directory separator and filenames could not
1331 be assumed to be unique after 10 characters, Acorn implemented the C
1332 compiler to strip the trailing C<.c> C<.h> C<.s> and C<.o> suffix from
1333 filenames specified in source code and store the respective files in
1334 subdirectories named after the suffix. Hence files are translated:
1337 C:foo.h C:h.foo (logical path variable)
1338 sys/os.h sys.h.os (C compiler groks Unix-speak)
1339 10charname.c c.10charname
1340 10charname.o o.10charname
1341 11charname_.c c.11charname (assuming filesystem truncates at 10)
1343 The Unix emulation library's translation of filenames to native assumes
1344 that this sort of translation is required, and it allows a user-defined list
1345 of known suffixes that it will transpose in this fashion. This may
1346 seem transparent, but consider that with these rules C<foo/bar/baz.h>
1347 and C<foo/bar/h/baz> both map to C<foo.bar.h.baz>, and that C<readdir> and
1348 C<glob> cannot and do not attempt to emulate the reverse mapping. Other
1349 C<.>'s in filenames are translated to C</>.
1351 As implied above, the environment accessed through C<%ENV> is global, and
1352 the convention is that program specific environment variables are of the
1353 form C<Program$Name>. Each filesystem maintains a current directory,
1354 and the current filesystem's current directory is the B<global> current
1355 directory. Consequently, sociable programs don't change the current
1356 directory but rely on full pathnames, and programs (and Makefiles) cannot
1357 assume that they can spawn a child process which can change the current
1358 directory without affecting its parent (and everyone else for that
1361 Because native operating system filehandles are global and are currently
1362 allocated down from 255, with 0 being a reserved value, the Unix emulation
1363 library emulates Unix filehandles. Consequently, you can't rely on
1364 passing C<STDIN>, C<STDOUT>, or C<STDERR> to your children.
1366 The desire of users to express filenames of the form
1367 C<< <Foo$Dir>.Bar >> on the command line unquoted causes problems,
1368 too: C<``> command output capture has to perform a guessing game. It
1369 assumes that a string C<< <[^<>]+\$[^<>]> >> is a
1370 reference to an environment variable, whereas anything else involving
1371 C<< < >> or C<< > >> is redirection, and generally manages to be 99%
1372 right. Of course, the problem remains that scripts cannot rely on any
1373 Unix tools being available, or that any tools found have Unix-like command
1376 Extensions and XS are, in theory, buildable by anyone using free
1377 tools. In practice, many don't, as users of the Acorn platform are
1378 used to binary distributions. MakeMaker does run, but no available
1379 make currently copes with MakeMaker's makefiles; even if and when
1380 this should be fixed, the lack of a Unix-like shell will cause
1381 problems with makefile rules, especially lines of the form C<cd
1382 sdbm && make all>, and anything using quoting.
1384 "S<RISC OS>" is the proper name for the operating system, but the value
1385 in C<$^O> is "riscos" (because we don't like shouting).
1389 Perl has been ported to many platforms that do not fit into any of
1390 the categories listed above. Some, such as AmigaOS, Atari MiNT,
1391 BeOS, HP MPE/iX, QNX, Plan 9, and VOS, have been well-integrated
1392 into the standard Perl source code kit. You may need to see the
1393 F<ports/> directory on CPAN for information, and possibly binaries,
1394 for the likes of: aos, Atari ST, lynxos, riscos, Novell Netware,
1395 Tandem Guardian, I<etc.> (Yes, we know that some of these OSes may
1396 fall under the Unix category, but we are not a standards body.)
1398 Some approximate operating system names and their C<$^O> values
1399 in the "OTHER" category include:
1401 OS $^O $Config{'archname'}
1402 ------------------------------------------
1403 Amiga DOS amigaos m68k-amigos
1405 MPE/iX mpeix PA-RISC1.1
1413 Amiga, F<README.amiga> (installed as L<perlamiga>).
1417 Atari, F<README.mint> and Guido Flohr's web page
1418 http://stud.uni-sb.de/~gufl0000/
1422 Be OS, F<README.beos>
1426 HP 300 MPE/iX, F<README.mpeix> and Mark Bixby's web page
1427 http://www.bixby.org/mark/perlix.html
1431 A free perl5-based PERL.NLM for Novell Netware is available in
1432 precompiled binary and source code form from http://www.novell.com/
1433 as well as from CPAN.
1437 S<Plan 9>, F<README.plan9>
1441 =head1 FUNCTION IMPLEMENTATIONS
1443 Listed below are functions that are either completely unimplemented
1444 or else have been implemented differently on various platforms.
1445 Following each description will be, in parentheses, a list of
1446 platforms that the description applies to.
1448 The list may well be incomplete, or even wrong in some places. When
1449 in doubt, consult the platform-specific README files in the Perl
1450 source distribution, and any other documentation resources accompanying
1453 Be aware, moreover, that even among Unix-ish systems there are variations.
1455 For many functions, you can also query C<%Config>, exported by
1456 default from the Config module. For example, to check whether the
1457 platform has the C<lstat> call, check C<$Config{d_lstat}>. See
1458 L<Config> for a full description of available variables.
1460 =head2 Alphabetical Listing of Perl Functions
1470 C<-r>, C<-w>, and C<-x> have a limited meaning only; directories
1471 and applications are executable, and there are no uid/gid
1472 considerations. C<-o> is not supported. (S<Mac OS>)
1474 C<-r>, C<-w>, C<-x>, and C<-o> tell whether the file is accessible,
1475 which may not reflect UIC-based file protections. (VMS)
1477 C<-s> returns the size of the data fork, not the total size of data fork
1478 plus resource fork. (S<Mac OS>).
1480 C<-s> by name on an open file will return the space reserved on disk,
1481 rather than the current extent. C<-s> on an open filehandle returns the
1482 current size. (S<RISC OS>)
1484 C<-R>, C<-W>, C<-X>, C<-O> are indistinguishable from C<-r>, C<-w>,
1485 C<-x>, C<-o>. (S<Mac OS>, Win32, VMS, S<RISC OS>)
1487 C<-b>, C<-c>, C<-k>, C<-g>, C<-p>, C<-u>, C<-A> are not implemented.
1490 C<-g>, C<-k>, C<-l>, C<-p>, C<-u>, C<-A> are not particularly meaningful.
1491 (Win32, VMS, S<RISC OS>)
1493 C<-d> is true if passed a device spec without an explicit directory.
1496 C<-T> and C<-B> are implemented, but might misclassify Mac text files
1497 with foreign characters; this is the case will all platforms, but may
1498 affect S<Mac OS> often. (S<Mac OS>)
1500 C<-x> (or C<-X>) determine if a file ends in one of the executable
1501 suffixes. C<-S> is meaningless. (Win32)
1503 C<-x> (or C<-X>) determine if a file has an executable file type.
1510 Not implemented. (Win32)
1512 =item binmode FILEHANDLE
1514 Meaningless. (S<Mac OS>, S<RISC OS>)
1516 Reopens file and restores pointer; if function fails, underlying
1517 filehandle may be closed, or pointer may be in a different position.
1520 The value returned by C<tell> may be affected after the call, and
1521 the filehandle may be flushed. (Win32)
1525 Only limited meaning. Disabling/enabling write permission is mapped to
1526 locking/unlocking the file. (S<Mac OS>)
1528 Only good for changing "owner" read-write access, "group", and "other"
1529 bits are meaningless. (Win32)
1531 Only good for changing "owner" and "other" read-write access. (S<RISC OS>)
1533 Access permissions are mapped onto VOS access-control list changes. (VOS)
1535 The actual permissions set depend on the value of the C<CYGWIN>
1536 in the SYSTEM environment settings. (Cygwin)
1540 Not implemented. (S<Mac OS>, Win32, S<Plan 9>, S<RISC OS>)
1542 Does nothing, but won't fail. (Win32)
1544 A little funky, because VOS's notion of ownership is a little funky (VOS).
1546 =item chroot FILENAME
1550 Not implemented. (S<Mac OS>, Win32, VMS, S<Plan 9>, S<RISC OS>, VOS, VM/ESA)
1552 =item crypt PLAINTEXT,SALT
1554 May not be available if library or source was not provided when building
1559 Not implemented. (VMS, S<Plan 9>, VOS)
1561 =item dbmopen HASH,DBNAME,MODE
1563 Not implemented. (VMS, S<Plan 9>, VOS)
1567 Not useful. (S<Mac OS>, S<RISC OS>)
1569 Not implemented. (Win32)
1571 Invokes VMS debugger. (VMS)
1575 Not implemented. (S<Mac OS>)
1577 Implemented via Spawn. (VM/ESA)
1579 Does not automatically flush output handles on some platforms.
1580 (SunOS, Solaris, HP-UX)
1586 Emulates UNIX exit() (which considers C<exit 1> to indicate an error) by
1587 mapping the C<1> to SS$_ABORT (C<44>). This behavior may be overridden
1588 with the pragma C<use vmsish 'exit'>. As with the CRTL's exit()
1589 function, C<exit 0> is also mapped to an exit status of SS$_NORMAL
1590 (C<1>); this mapping cannot be overridden. Any other argument to exit()
1591 is used directly as Perl's exit status. (VMS)
1593 =item fcntl FILEHANDLE,FUNCTION,SCALAR
1595 Not implemented. (Win32, VMS)
1597 =item flock FILEHANDLE,OPERATION
1599 Not implemented (S<Mac OS>, VMS, S<RISC OS>, VOS).
1601 Available only on Windows NT (not on Windows 95). (Win32)
1605 Not implemented. (S<Mac OS>, AmigaOS, S<RISC OS>, VM/ESA, VMS)
1607 Emulated using multiple interpreters. See L<perlfork>. (Win32)
1609 Does not automatically flush output handles on some platforms.
1610 (SunOS, Solaris, HP-UX)
1614 Not implemented. (S<Mac OS>, S<RISC OS>)
1618 Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>)
1622 Not implemented. (S<Mac OS>, Win32, S<RISC OS>)
1624 =item getpriority WHICH,WHO
1626 Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>, VOS, VM/ESA)
1630 Not implemented. (S<Mac OS>, Win32)
1632 Not useful. (S<RISC OS>)
1636 Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>)
1638 =item getnetbyname NAME
1640 Not implemented. (S<Mac OS>, Win32, S<Plan 9>)
1644 Not implemented. (S<Mac OS>, Win32)
1646 Not useful. (S<RISC OS>)
1650 Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>)
1652 =item getnetbyaddr ADDR,ADDRTYPE
1654 Not implemented. (S<Mac OS>, Win32, S<Plan 9>)
1656 =item getprotobynumber NUMBER
1658 Not implemented. (S<Mac OS>)
1660 =item getservbyport PORT,PROTO
1662 Not implemented. (S<Mac OS>)
1666 Not implemented. (S<Mac OS>, Win32, VM/ESA)
1670 Not implemented. (S<Mac OS>, Win32, VMS, VM/ESA)
1674 C<gethostbyname('localhost')> does not work everywhere: you may have
1675 to use C<gethostbyname('127.0.0.1')>. (S<Mac OS>, S<Irix 5>)
1679 Not implemented. (S<Mac OS>, Win32)
1683 Not implemented. (S<Mac OS>, Win32, S<Plan 9>)
1687 Not implemented. (S<Mac OS>, Win32, S<Plan 9>)
1691 Not implemented. (Win32, S<Plan 9>)
1693 =item sethostent STAYOPEN
1695 Not implemented. (S<Mac OS>, Win32, S<Plan 9>, S<RISC OS>)
1697 =item setnetent STAYOPEN
1699 Not implemented. (S<Mac OS>, Win32, S<Plan 9>, S<RISC OS>)
1701 =item setprotoent STAYOPEN
1703 Not implemented. (S<Mac OS>, Win32, S<Plan 9>, S<RISC OS>)
1705 =item setservent STAYOPEN
1707 Not implemented. (S<Plan 9>, Win32, S<RISC OS>)
1711 Not implemented. (S<Mac OS>, MPE/iX, VM/ESA, Win32)
1715 Not implemented. (S<Mac OS>, MPE/iX, S<RISC OS>, VM/ESA, VMS, Win32)
1719 Not implemented. (S<Mac OS>, Win32)
1723 Not implemented. (S<Mac OS>, Win32, S<Plan 9>)
1727 Not implemented. (S<Mac OS>, Win32, S<Plan 9>)
1731 Not implemented. (S<Plan 9>, Win32)
1733 =item getsockopt SOCKET,LEVEL,OPTNAME
1735 Not implemented. (S<Plan 9>)
1741 This operator is implemented via the File::Glob extension on most
1742 platforms. See L<File::Glob> for portability information.
1744 =item ioctl FILEHANDLE,FUNCTION,SCALAR
1746 Not implemented. (VMS)
1748 Available only for socket handles, and it does what the ioctlsocket() call
1749 in the Winsock API does. (Win32)
1751 Available only for socket handles. (S<RISC OS>)
1753 =item kill SIGNAL, LIST
1755 C<kill(0, LIST)> is implemented for the sake of taint checking;
1756 use with other signals is unimplemented. (S<Mac OS>)
1758 Not implemented, hence not useful for taint checking. (S<RISC OS>)
1760 C<kill()> doesn't have the semantics of C<raise()>, i.e. it doesn't send
1761 a signal to the identified process like it does on Unix platforms.
1762 Instead C<kill($sig, $pid)> terminates the process identified by $pid,
1763 and makes it exit immediately with exit status $sig. As in Unix, if
1764 $sig is 0 and the specified process exists, it returns true without
1765 actually terminating it. (Win32)
1767 =item link OLDFILE,NEWFILE
1769 Not implemented. (S<Mac OS>, MPE/iX, VMS, S<RISC OS>)
1771 Link count not updated because hard links are not quite that hard
1772 (They are sort of half-way between hard and soft links). (AmigaOS)
1774 Hard links are implemented on Win32 (Windows NT and Windows 2000)
1777 =item lstat FILEHANDLE
1783 Not implemented. (VMS, S<RISC OS>)
1785 Return values (especially for device and inode) may be bogus. (Win32)
1787 =item msgctl ID,CMD,ARG
1789 =item msgget KEY,FLAGS
1791 =item msgsnd ID,MSG,FLAGS
1793 =item msgrcv ID,VAR,SIZE,TYPE,FLAGS
1795 Not implemented. (S<Mac OS>, Win32, VMS, S<Plan 9>, S<RISC OS>, VOS)
1797 =item open FILEHANDLE,EXPR
1799 =item open FILEHANDLE
1801 The C<|> variants are supported only if ToolServer is installed.
1804 open to C<|-> and C<-|> are unsupported. (S<Mac OS>, Win32, S<RISC OS>)
1806 Opening a process does not automatically flush output handles on some
1807 platforms. (SunOS, Solaris, HP-UX)
1809 =item pipe READHANDLE,WRITEHANDLE
1811 Very limited functionality. (MiNT)
1817 Not implemented. (Win32, VMS, S<RISC OS>)
1819 =item select RBITS,WBITS,EBITS,TIMEOUT
1821 Only implemented on sockets. (Win32, VMS)
1823 Only reliable on sockets. (S<RISC OS>)
1825 Note that the C<select FILEHANDLE> form is generally portable.
1827 =item semctl ID,SEMNUM,CMD,ARG
1829 =item semget KEY,NSEMS,FLAGS
1831 =item semop KEY,OPSTRING
1833 Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>, VOS)
1837 Not implemented. (S<Mac OS>, MPE/iX, VMS, Win32, S<RISC OS>, VOS)
1839 =item setpgrp PID,PGRP
1841 Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>, VOS)
1843 =item setpriority WHICH,WHO,PRIORITY
1845 Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>, VOS)
1849 Not implemented. (S<Mac OS>, MPE/iX, Win32, S<RISC OS>, VOS)
1851 =item setsockopt SOCKET,LEVEL,OPTNAME,OPTVAL
1853 Not implemented. (S<Plan 9>)
1855 =item shmctl ID,CMD,ARG
1857 =item shmget KEY,SIZE,FLAGS
1859 =item shmread ID,VAR,POS,SIZE
1861 =item shmwrite ID,STRING,POS,SIZE
1863 Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>, VOS)
1865 =item sockatmark SOCKET
1867 A relatively recent addition to socket functions, may not
1868 be implemented even in UNIX platforms.
1870 =item socketpair SOCKET1,SOCKET2,DOMAIN,TYPE,PROTOCOL
1872 Not implemented. (Win32, VMS, S<RISC OS>, VOS, VM/ESA)
1874 =item stat FILEHANDLE
1880 Platforms that do not have rdev, blksize, or blocks will return these
1881 as '', so numeric comparison or manipulation of these fields may cause
1882 'not numeric' warnings.
1884 mtime and atime are the same thing, and ctime is creation time instead of
1885 inode change time. (S<Mac OS>).
1887 ctime not supported on UFS (S<Mac OS X>).
1889 ctime is creation time instead of inode change time (Win32).
1891 device and inode are not meaningful. (Win32)
1893 device and inode are not necessarily reliable. (VMS)
1895 mtime, atime and ctime all return the last modification time. Device and
1896 inode are not necessarily reliable. (S<RISC OS>)
1898 dev, rdev, blksize, and blocks are not available. inode is not
1899 meaningful and will differ between stat calls on the same file. (os2)
1901 some versions of cygwin when doing a stat("foo") and if not finding it
1902 may then attempt to stat("foo.exe") (Cygwin)
1904 =item symlink OLDFILE,NEWFILE
1906 Not implemented. (Win32, VMS, S<RISC OS>)
1910 Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>, VOS, VM/ESA)
1912 =item sysopen FILEHANDLE,FILENAME,MODE,PERMS
1914 The traditional "0", "1", and "2" MODEs are implemented with different
1915 numeric values on some systems. The flags exported by C<Fcntl>
1916 (O_RDONLY, O_WRONLY, O_RDWR) should work everywhere though. (S<Mac
1917 OS>, OS/390, VM/ESA)
1921 In general, do not assume the UNIX/POSIX semantics that you can shift
1922 C<$?> right by eight to get the exit value, or that C<$? & 127>
1923 would give you the number of the signal that terminated the program,
1924 or that C<$? & 128> would test true if the program was terminated by a
1925 coredump. Instead, use the POSIX W*() interfaces: for example, use
1926 WIFEXITED($?) and WEXITVALUE($?) to test for a normal exit and the exit
1927 value, WIFSIGNALED($?) and WTERMSIG($?) for a signal exit and the
1928 signal. Core dumping is not a portable concept, so there's no portable
1929 way to test for that.
1931 Only implemented if ToolServer is installed. (S<Mac OS>)
1933 As an optimization, may not call the command shell specified in
1934 C<$ENV{PERL5SHELL}>. C<system(1, @args)> spawns an external
1935 process and immediately returns its process designator, without
1936 waiting for it to terminate. Return value may be used subsequently
1937 in C<wait> or C<waitpid>. Failure to spawn() a subprocess is indicated
1938 by setting $? to "255 << 8". C<$?> is set in a way compatible with
1939 Unix (i.e. the exitstatus of the subprocess is obtained by "$? >> 8",
1940 as described in the documentation). (Win32)
1942 There is no shell to process metacharacters, and the native standard is
1943 to pass a command line terminated by "\n" "\r" or "\0" to the spawned
1944 program. Redirection such as C<< > foo >> is performed (if at all) by
1945 the run time library of the spawned program. C<system> I<list> will call
1946 the Unix emulation library's C<exec> emulation, which attempts to provide
1947 emulation of the stdin, stdout, stderr in force in the parent, providing
1948 the child program uses a compatible version of the emulation library.
1949 I<scalar> will call the native command line direct and no such emulation
1950 of a child Unix program will exists. Mileage B<will> vary. (S<RISC OS>)
1952 Far from being POSIX compliant. Because there may be no underlying
1953 /bin/sh tries to work around the problem by forking and execing the
1954 first token in its argument string. Handles basic redirection
1955 ("<" or ">") on its own behalf. (MiNT)
1957 Does not automatically flush output handles on some platforms.
1958 (SunOS, Solaris, HP-UX)
1960 The return value is POSIX-like (shifted up by 8 bits), which only allows
1961 room for a made-up value derived from the severity bits of the native
1962 32-bit condition code (unless overridden by C<use vmsish 'status'>).
1963 For more details see L<perlvms/$?>. (VMS)
1967 Only the first entry returned is nonzero. (S<Mac OS>)
1969 "cumulative" times will be bogus. On anything other than Windows NT
1970 or Windows 2000, "system" time will be bogus, and "user" time is
1971 actually the time returned by the clock() function in the C runtime
1974 Not useful. (S<RISC OS>)
1976 =item truncate FILEHANDLE,LENGTH
1978 =item truncate EXPR,LENGTH
1980 Not implemented. (Older versions of VMS)
1982 Truncation to same-or-shorter lengths only. (VOS)
1984 If a FILEHANDLE is supplied, it must be writable and opened in append
1985 mode (i.e., use C<<< open(FH, '>>filename') >>>
1986 or C<sysopen(FH,...,O_APPEND|O_RDWR)>. If a filename is supplied, it
1987 should not be held open elsewhere. (Win32)
1993 Returns undef where unavailable, as of version 5.005.
1995 C<umask> works but the correct permissions are set only when the file
1996 is finally closed. (AmigaOS)
2000 Only the modification time is updated. (S<BeOS>, S<Mac OS>, VMS, S<RISC OS>)
2002 May not behave as expected. Behavior depends on the C runtime
2003 library's implementation of utime(), and the filesystem being
2004 used. The FAT filesystem typically does not support an "access
2005 time" field, and it may limit timestamps to a granularity of
2006 two seconds. (Win32)
2010 =item waitpid PID,FLAGS
2012 Not implemented. (S<Mac OS>)
2014 Can only be applied to process handles returned for processes spawned
2015 using C<system(1, ...)> or pseudo processes created with C<fork()>. (Win32)
2017 Not useful. (S<RISC OS>)
2025 =item v1.49, 12 August 2002
2027 Updates for VOS from Paul Green.
2029 =item v1.48, 02 February 2001
2031 Various updates from perl5-porters over the past year, supported
2032 platforms update from Jarkko Hietaniemi.
2034 =item v1.47, 22 March 2000
2036 Various cleanups from Tom Christiansen, including migration of
2037 long platform listings from L<perl>.
2039 =item v1.46, 12 February 2000
2041 Updates for VOS and MPE/iX. (Peter Prymmer) Other small changes.
2043 =item v1.45, 20 December 1999
2045 Small changes from 5.005_63 distribution, more changes to EBCDIC info.
2047 =item v1.44, 19 July 1999
2049 A bunch of updates from Peter Prymmer for C<$^O> values,
2050 endianness, File::Spec, VMS, BS2000, OS/400.
2052 =item v1.43, 24 May 1999
2054 Added a lot of cleaning up from Tom Christiansen.
2056 =item v1.42, 22 May 1999
2058 Added notes about tests, sprintf/printf, and epoch offsets.
2060 =item v1.41, 19 May 1999
2062 Lots more little changes to formatting and content.
2064 Added a bunch of C<$^O> and related values
2065 for various platforms; fixed mail and web addresses, and added
2066 and changed miscellaneous notes. (Peter Prymmer)
2068 =item v1.40, 11 April 1999
2070 Miscellaneous changes.
2072 =item v1.39, 11 February 1999
2074 Changes from Jarkko and EMX URL fixes Michael Schwern. Additional
2075 note about newlines added.
2077 =item v1.38, 31 December 1998
2079 More changes from Jarkko.
2081 =item v1.37, 19 December 1998
2083 More minor changes. Merge two separate version 1.35 documents.
2085 =item v1.36, 9 September 1998
2087 Updated for Stratus VOS. Also known as version 1.35.
2089 =item v1.35, 13 August 1998
2091 Integrate more minor changes, plus addition of new sections under
2092 L<"ISSUES">: L<"Numbers endianness and Width">,
2093 L<"Character sets and character encoding">,
2094 L<"Internationalisation">.
2096 =item v1.33, 06 August 1998
2098 Integrate more minor changes.
2100 =item v1.32, 05 August 1998
2102 Integrate more minor changes.
2104 =item v1.30, 03 August 1998
2106 Major update for RISC OS, other minor changes.
2108 =item v1.23, 10 July 1998
2110 First public release with perl5.005.
2114 =head1 Supported Platforms
2116 As of July 2002 (the Perl release 5.8.0), the following platforms are
2117 able to build Perl from the standard source code distribution
2118 available at http://www.cpan.org/src/index.html
2129 HI-UXMPP (Hitachi) (5.8.0 worked but we didn't know it)
2139 ReliantUNIX (formerly SINIX)
2141 OpenVMS (formerly VMS)
2142 Open UNIX (Unixware) (since Perl 5.8.1/5.9.0)
2144 OS/400 (using the PASE) (since Perl 5.8.1/5.9.0)
2146 POSIX-BC (formerly BS2000)
2151 Tru64 UNIX (formerly DEC OSF/1, Digital UNIX)
2156 Win95/98/ME/2K/XP 2)
2158 z/OS (formerly OS/390)
2161 1) in DOS mode either the DOS or OS/2 ports can be used
2162 2) compilers: Borland, MinGW (GCC), VC6
2164 The following platforms worked with the previous releases (5.6 and
2165 5.7), but we did not manage either to fix or to test these in time
2166 for the 5.8.0 release. There is a very good chance that many of these
2167 will work fine with the 5.8.0.
2180 Known to be broken for 5.8.0 (but 5.6.1 and 5.7.2 can be used):
2184 The following platforms have been known to build Perl from source in
2185 the past (5.005_03 and earlier), but we haven't been able to verify
2186 their status for the current release, either because the
2187 hardware/software platforms are rare or because we don't have an
2188 active champion on these platforms--or both. They used to work,
2189 though, so go ahead and try compiling them, and let perlbug@perl.org
2223 The following platforms have their own source code distributions and
2224 binaries available via http://www.cpan.org/ports/
2228 OS/400 (ILE) 5.005_02
2229 Tandem Guardian 5.004
2231 The following platforms have only binaries available via
2232 http://www.cpan.org/ports/index.html :
2236 Acorn RISCOS 5.005_02
2240 Although we do suggest that you always build your own Perl from
2241 the source code, both for maximal configurability and for security,
2242 in case you are in a hurry you can check
2243 http://www.cpan.org/ports/index.html for binary distributions.
2247 L<perlaix>, L<perlamiga>, L<perlapollo>, L<perlbeos>, L<perlbs2000>,
2248 L<perlce>, L<perlcygwin>, L<perldgux>, L<perldos>, L<perlepoc>,
2249 L<perlebcdic>, L<perlfreebsd>, L<perlhurd>, L<perlhpux>, L<perlirix>,
2250 L<perlmachten>, L<perlmacos>, L<perlmint>, L<perlmpeix>,
2251 L<perlnetware>, L<perlos2>, L<perlos390>, L<perlos400>,
2252 L<perlplan9>, L<perlqnx>, L<perlsolaris>, L<perltru64>,
2253 L<perlunicode>, L<perlvmesa>, L<perlvms>, L<perlvos>,
2254 L<perlwin32>, and L<Win32>.
2256 =head1 AUTHORS / CONTRIBUTORS
2258 Abigail <abigail@foad.org>,
2259 Charles Bailey <bailey@newman.upenn.edu>,
2260 Graham Barr <gbarr@pobox.com>,
2261 Tom Christiansen <tchrist@perl.com>,
2262 Nicholas Clark <nick@ccl4.org>,
2263 Thomas Dorner <Thomas.Dorner@start.de>,
2264 Andy Dougherty <doughera@lafayette.edu>,
2265 Dominic Dunlop <domo@computer.org>,
2266 Neale Ferguson <neale@vma.tabnsw.com.au>,
2267 David J. Fiander <davidf@mks.com>,
2268 Paul Green <Paul.Green@stratus.com>,
2269 M.J.T. Guy <mjtg@cam.ac.uk>,
2270 Jarkko Hietaniemi <jhi@iki.fi>,
2271 Luther Huffman <lutherh@stratcom.com>,
2272 Nick Ing-Simmons <nick@ing-simmons.net>,
2273 Andreas J. KE<ouml>nig <a.koenig@mind.de>,
2274 Markus Laker <mlaker@contax.co.uk>,
2275 Andrew M. Langmead <aml@world.std.com>,
2276 Larry Moore <ljmoore@freespace.net>,
2277 Paul Moore <Paul.Moore@uk.origin-it.com>,
2278 Chris Nandor <pudge@pobox.com>,
2279 Matthias Neeracher <neeracher@mac.com>,
2280 Philip Newton <pne@cpan.org>,
2281 Gary Ng <71564.1743@CompuServe.COM>,
2282 Tom Phoenix <rootbeer@teleport.com>,
2283 AndrE<eacute> Pirard <A.Pirard@ulg.ac.be>,
2284 Peter Prymmer <pvhp@forte.com>,
2285 Hugo van der Sanden <hv@crypt0.demon.co.uk>,
2286 Gurusamy Sarathy <gsar@activestate.com>,
2287 Paul J. Schinder <schinder@pobox.com>,
2288 Michael G Schwern <schwern@pobox.com>,
2289 Dan Sugalski <dan@sidhe.org>,
2290 Nathan Torkington <gnat@frii.com>.