README.win32

If you read this file _as_is_, just ignore the funny characters you
see. It is written in the POD format (see pod/perlpod.pod) which is
specially designed to be readable as is.

=head1 NAME

perlwin32 - Perl under Win32

=head1 SYNOPSIS

These are instructions for building Perl under Windows NT (versions
3.51 or 4.0), using Visual C++ (versions 2.0 through 5.0) or Borland
C++ (version 5.02 or later).  Currently, this port is reported to build
under Windows95 using the 4DOS shell--the default shell that infests
Windows95 will not work (see below).  Note this caveat is only about
B<building> perl.  Once built, you should be able to B<use> it on
either Win32 platform (modulo the problems arising from the inferior
command shell).

=head1 DESCRIPTION

Before you start, you should glance through the README file
found in the top-level directory where the Perl distribution
was extracted.  Make sure you read and understand the terms under
which this software is being distributed.

Also make sure you read L<BUGS AND CAVEATS> below for the
known limitations of this port.

The INSTALL file in the perl top-level has much information that is
only relevant to people building Perl on Unix-like systems.  In
particular, you can safely ignore any information that talks about
"Configure".

You may also want to look at two other options for building
a perl that will work on Windows NT:  the README.cygwin32 and
README.os2 files, which each give a different set of rules to build
a Perl that will work on Win32 platforms.  Those two methods will
probably enable you to build a more Unix-compatible perl, but you
will also need to download and use various other build-time and
run-time support software described in those files.

This set of instructions is meant to describe a so-called "native"
port of Perl to Win32 platforms.  The resulting Perl requires no
additional software to run (other than what came with your operating
system).  Currently, this port is capable of using either the
Microsoft Visual C++ compiler, or the Borland C++ compiler.  The
ultimate goal is to support the other major compilers that can
generally be used to build Win32 applications.

This port currently supports MakeMaker (the set of modules that
is used to build extensions to perl).  Therefore, you should be
able to build and install most extensions found in the CPAN sites.
See L<Usage Hints> below for general hints about this.

=head2 Setting Up

=over 4

=item Command Shell

Use the default "cmd" shell that comes with NT.  Some versions of the
popular 4DOS/NT shell have incompatibilities that may cause you trouble.
If the build fails under that shell, try building again with the cmd
shell.  The Makefile also has known incompatibilites with the "command.com"
shell that comes with Windows95, so building under Windows95 should
be considered "unsupported".  However, there have been reports of successful
build attempts using 4DOS/NT version 3.00 under Windows95, using dmake, but
your mileage may vary.

The surest way to build it is on WindowsNT, using the cmd shell.

=item Borland C++

If you are using the Borland compiler, you will need dmake, a freely
available make that has very nice macro features and parallelability.
(The make that Borland supplies is seriously crippled, and will not
work for MakeMaker builds.)

A port of dmake for win32 platforms is available from:

    http://www-personal.umich.edu/~gsar/dmake-4.1-win32.zip

Fetch and install dmake somewhere on your path (follow the instructions
in the README.NOW file).

=item Microsoft Visual C++

The NMAKE that comes with Visual C++ will suffice for building.
If you did not choose to always initialize the Visual C++ compilation
environment variables when you installed Visual C++ on your system, you
will need to run the VCVARS32.BAT file usually found somewhere like
C:\MSDEV4.2\BIN.  This will set your build environment.

You can also use dmake to build using Visual C++, provided:
you set OSRELEASE to "microsft" (or whatever the directory name
under which the Visual C dmake configuration lives) in your environment,
and edit win32/config.vc to change "make=nmake" into "make=dmake".  The
latter step is only essential if you want to use dmake as your default
make for building extensions using MakeMaker.

=item Permissions

Depending on how you extracted the distribution, you have to make sure
some of the files are writable by you.  The easiest way to make sure of
this is to execute:

	attrib -R *.* /S

from the perl toplevel directory.  You don't I<have> to do this if you
used the right tools to extract the files in the standard distribution,
but it doesn't hurt to do so.

=back

=head2 Building

=over 4

=item *

Make sure you are in the "win32" subdirectory under the perl toplevel.
This directory contains a "Makefile" that will work with
versions of NMAKE that come with Visual C++ ver. 2.0 and above, and
a dmake "makefile.mk" that will work for both Borland and Visual C++
builds.  The defaults in the dmake makefile are setup to build using the
Borland compiler.

=item *

Edit the Makefile (or makefile.mk, if using dmake) and change the values
of INST_DRV and INST_TOP.  If you are using Visual C++ ver. 2.0, uncomment
the line that sets "CCTYPE=MSVC20".  You can also enable various build
flags.

If you have either the source or a library that contains des_fcrypt(),
enable the appropriate option in the makefile.  des_fcrypt() is not
bundled with the distribution due to US Government restrictions
on the export of cryptographic software.  Nevertheless, this routine
is part of the "libdes" library (written by Ed Young) which is widely
available worldwide, usually along with SSLeay (for example:
"ftp://fractal.mta.ca/pub/crypto/SSLeay/DES/").  Set CRYPT_SRC to the
name of the file that implements des_fcrypt().  Alternatively, if
you have built a library that contains des_fcrypt(), you can set
CRYPT_LIB to point to the library name.

Perl will also build without des_fcrypt(), but the crypt() builtin will
fail at run time.

You will also have to make sure CCHOME points to wherever you installed
your compiler.

=item *

Type "nmake" (or "dmake" if you are using that make).

This should build everything.  Specifically, it will create perl.exe,
perl.dll, and perlglob.exe at the perl toplevel, and various other
extension dll's under the lib\auto directory.  If the build fails for
any reason, make sure you have done the previous steps correctly.

The build process may produce "harmless" compiler warnings (more or
less copiously, depending on how picky your compiler gets).  The
maintainers are aware of these warnings, thankyouverymuch. :)

When building using Visual C++, a perl95.exe will also get built.  This
executable is only needed on Windows95, and should be used instead of
perl.exe, and then only if you want sockets to work properly on Windows95.
This is necessitated by a bug in the Microsoft C Runtime that cannot be
worked around in the "normal" perl.exe.  perl95.exe gets built with its
own private copy of the C Runtime that is not accessible to extensions
(which see the DLL version of the CRT).  Be aware, therefore, that this
perl95.exe will have esoteric problems with extensions like perl/Tk that
themselves use the C Runtime heavily, or want to free() pointers
malloc()-ed by perl.

You can avoid the perl95.exe problems completely if you use Borland
C++ for building perl (perl95.exe is not needed and will not be built
in that case).

=back

=head2 Testing

Type "nmake test" (or "dmake test").  This will run most of the tests from
the testsuite (many tests will be skipped, and but no test should fail).

If some tests do fail, it may be because you are using a different command
shell than the native "cmd.exe".

If you used the Borland compiler, you may see a failure in op/taint.t
arising from the inability to find the Borland Runtime DLLs on the system
default path.  You will need to copy the DLLs reported by the messages
from where Borland chose to install it, into the Windows system directory
(usually somewhere like C:\WINNT\SYSTEM32), and rerun the test.

Please report any other failures as described under L<BUGS AND CAVEATS>.

=head2 Installation

Type "nmake install" (or "dmake install").  This will put the newly
built perl and the libraries under whatever C<INST_TOP> points to in the
Makefile.  It will also install the pod documentation under
C<$INST_TOP\lib\pod> and HTML versions of the same under
C<$INST_TOP\lib\pod\html>.  To use the Perl you just installed,
set your PATH environment variable to C<$INST_TOP\bin>.

=head2 Usage Hints

=over 4

=item Environment Variables

The installation paths that you set during the build get compiled
into perl, so you don't have to do anything additional to start
using that perl (except add its location to your PATH variable).

If you put extensions in unusual places, you can set PERL5LIB
to a list of paths separated by semicolons where you want perl
to look for libraries.  Look for descriptions of other environment
variables you can set in L<perlrun>.

You can also control the shell that perl uses to run system() and
backtick commands via PERL5SHELL.  See L<perlrun>.

Sometime in the future, some of the configuration information
for perl will be moved into the Windows registry.

=item File Globbing

By default, perl spawns an external program to do file globbing.
The install process installs both a perlglob.exe and a perlglob.bat
that perl can use for this purpose.  Note that with the default
installation, perlglob.exe will be found by the system before
perlglob.bat.

perlglob.exe relies on the argv expansion done by the C Runtime of
the particular compiler you used, and therefore behaves very
differently depending on the Runtime used to build it.  To preserve
compatiblity, perlglob.bat (a perl script that can be used portably)
is installed.  Besides being portable, perlglob.bat also offers
enhanced globbing functionality.

If you want perl to use perlglob.bat instead of perlglob.exe, just
delete perlglob.exe from the install location (or move it somewhere
perl cannot find).  Using File::DosGlob.pm (which implements the core
functionality of perlglob.bat) to override the internal CORE::glob()
works about 10 times faster than spawing perlglob.exe, and you should
take this approach when writing new modules.  See File::DosGlob for
details.

=item Using perl from the command line

If you are accustomed to using perl from various command-line
shells found in UNIX environments, you will be less than pleased
with what Windows NT offers by way of a command shell.

The crucial thing to understand about the "cmd" shell (which is
the default on Windows NT) is that it does not do any wildcard
expansions of command-line arguments (so wildcards need not be
quoted).  It also provides only rudimentary quoting.  The only
(useful) quote character is the double quote (").  It can be used to
protect spaces in arguments and other special characters.  The
Windows NT documentation has almost no description of how the
quoting rules are implemented, but here are some general observations
based on experiments:  The shell breaks arguments at spaces and
passes them to programs in argc/argv.  Doublequotes can be used
to prevent arguments with spaces in them from being split up.
You can put a double quote in an argument by escaping it with
a backslash and enclosing the whole argument within double quotes.
The backslash and the pair of double quotes surrounding the
argument will be stripped by the shell.

The file redirection characters "<", ">", and "|" cannot be quoted
by double quotes (there are probably more such).  Single quotes
will protect those three file redirection characters, but the
single quotes don't get stripped by the shell (just to make this
type of quoting completely useless).  The caret "^" has also
been observed to behave as a quoting character (and doesn't get
stripped by the shell also).

Here are some examples of usage of the "cmd" shell:

This prints two doublequotes:

    perl -e "print '\"\"' "

This does the same:

    perl -e "print \"\\\"\\\"\" "

This prints "bar" and writes "foo" to the file "blurch":

    perl -e "print 'foo'; print STDERR 'bar'" > blurch

This prints "foo" ("bar" disappears into nowhereland):

    perl -e "print 'foo'; print STDERR 'bar'" 2> nul

This prints "bar" and writes "foo" into the file "blurch":

    perl -e "print 'foo'; print STDERR 'bar'" 1> blurch

This pipes "foo" to the "less" pager and prints "bar" on the console:

    perl -e "print 'foo'; print STDERR 'bar'" | less

This pipes "foo\nbar\n" to the less pager:

    perl -le "print 'foo'; print STDERR 'bar'" 2>&1 | less

This pipes "foo" to the pager and writes "bar" in the file "blurch":

    perl -e "print 'foo'; print STDERR 'bar'" 2> blurch | less


Discovering the usefulness of the "command.com" shell on Windows95
is left as an exercise to the reader :)

=item Building Extensions

The Comprehensive Perl Archive Network (CPAN) offers a wealth
of extensions, some of which require a C compiler to build.
Look in http://www.perl.com/ for more information on CPAN.

Most extensions (whether they require a C compiler or not) can
be built, tested and installed with the standard mantra:

    perl Makefile.PL
    $MAKE
    $MAKE test
    $MAKE install

where $MAKE stands for NMAKE or DMAKE.  Some extensions may not
provide a testsuite (so "$MAKE test" may not do anything, or fail),
but most serious ones do.

If a module implements XSUBs, you will need one of the supported
C compilers.  You must make sure you have set up the environment for
the compiler for command-line compilation.

If a module does not build for some reason, look carefully for
why it failed, and report problems to the module author.  If
it looks like the extension building support is at fault, report
that with full details of how the build failed using the perlbug
utility.

=item Win32 Specific Extensions

A number of extensions specific to the Win32 platform are available
from CPAN.  You may find that many of these extensions are meant to
be used under the Activeware port of Perl, which used to be the only
native port for the Win32 platform.  Since the Activeware port does not
have adequate support for Perl's extension building tools, these
extensions typically do not support those tools either, and therefore
cannot be built using the generic steps shown in the previous section.

To ensure smooth transitioning of existing code that uses the
Activeware port, there is a bundle of Win32 extensions that contains
all of the Activeware extensions and most other Win32 extensions from
CPAN in source form, along with many added bugfixes, and with MakeMaker
support.  This bundle is available at:

   http://www.perl.com/CPAN/authors/id/GSAR/libwin32-0.09.tar.gz

See the README in that distribution for building and installation
instructions.  Look for later versions that may be available at the
same location.

It is expected that authors of Win32 specific extensions will begin
distributing their work in MakeMaker compatible form subsequent to
the 5.004 release of perl, at which point the need for a dedicated
bundle such as the above should diminish.

=item Running Perl Scripts

Perl scripts on UNIX use the "#!" (a.k.a "shebang") line to
indicate to the OS that it should execute the file using perl.
Win32 has no comparable means to indicate arbitrary files are
executables.

Instead, all available methods to execute plain text files on
Win32 rely on the file "extension".  There are three methods
to use this to execute perl scripts:

=over 8

=item 1

There is a facility called "file extension associations" that will
work in Windows NT 4.0.  This can be manipulated via the two
commands "assoc" and "ftype" that come standard with Windows NT
4.0.  Type "ftype /?" for a complete example of how to set this
up for perl scripts (Say what?  You thought Windows NT wasn't
perl-ready? :).

=item 2

Since file associations don't work everywhere, and there are
reportedly bugs with file associations where it does work, the
old method of wrapping the perl script to make it look like a
regular batch file to the OS, may be used.  The install process
makes available the "pl2bat.bat" script which can be used to wrap
perl scripts into batch files.  For example:

	pl2bat foo.pl

will create the file "FOO.BAT".  Note "pl2bat" strips any
.pl suffix and adds a .bat suffix to the generated file.

If you use the 4DOS/NT or similar command shell, note that
"pl2bat" uses the "%*" variable in the generated batch file to
refer to all the command line arguments, so you may need to make
sure that construct works in batch files.  As of this writing,
4DOS/NT users will need a "ParameterChar = *" statement in their
4NT.INI file, or will need to execute "setdos /p*" in the 4DOS/NT
startup file to enable this to work.

=item 3

Using "pl2bat" has a few problems:  the file name gets changed,
so scripts that rely on C<$0> to find what they must do may not
run properly; running "pl2bat" replicates the contents of the
original script, and so this process can be maintenance intensive
if the originals get updated often.  A different approach that
avoids both problems is possible.

A script called "runperl.bat" is available that can be copied
to any filename (along with the .bat suffix).  For example,
if you call it "foo.bat", it will run the file "foo" when it is
executed.  Since you can run batch files on Win32 platforms simply
by typing the name (without the extension), this effectively
runs the file "foo", when you type either "foo" or "foo.bat".
With this method, "foo.bat" can even be in a different location
than the file "foo", as long as "foo" is available somewhere on
the PATH.  If your scripts are on a filesystem that allows symbolic
links, you can even avoid copying "runperl.bat".

Here's a diversion:  copy "runperl.bat" to "runperl", and type
"runperl".  Explain the observed behavior, or lack thereof. :)
Hint: .gnidnats llits er'uoy fi ,"lrepnur" eteled :tniH

=back

=item Miscellaneous Things

A full set of HTML documentation is installed, so you should be
able to use it if you have a web browser installed on your
system.

C<perldoc> is also a useful tool for browsing information contained
in the documentation, especially in conjunction with a pager
like C<less> (recent versions of which have Win32 support).  You may
have to set the PAGER environment variable to use a specific pager.
"perldoc -f foo" will print information about the perl operator
"foo".

If you find bugs in perl, you can run C<perlbug> to create a
bug report (you may have to send it manually if C<perlbug> cannot
find a mailer on your system).

=back

=head1 BUGS AND CAVEATS

This port should be considered beta quality software at the present
time because some details are still in flux and there may be
changes in any of these areas: build process, installation structure,
supported utilities/modules, and supported perl functionality.
In particular, functionality specific to the Win32 environment may
ultimately be supported as either core modules or extensions.  The
beta status implies, among other things, that you should be prepared
to recompile extensions when binary incompatibilites arise due to
changes in the internal structure of the code.

An effort has been made to ensure that the DLLs produced by the two
supported compilers are compatible with each other (despite the
best efforts of the compiler vendors).  Extension binaries produced
by one compiler should also coexist with a perl binary built by
a different compiler.  In order to accomplish this, PERL.DLL provides
a layer of runtime code that uses the C Runtime that perl was compiled
with.  Extensions which include "perl.h" will transparently access
the functions in this layer, thereby ensuring that both perl and
extensions use the same runtime functions.

If you have had prior exposure to Perl on Unix platforms, you will notice
this port exhibits behavior different from what is documented.  Most of the
differences fall under one of these categories.  We do not consider
any of them to be serious limitations (especially when compared to the
limited nature of some of the Win32 OSes themselves :)

=over 8

=item *

C<stat()> and C<lstat()> functions may not behave as documented.  They
may return values that bear no resemblance to those reported on Unix
platforms, and some fields (like the the one for inode) may be completely
bogus.

=item *

The following functions are currently unavailable: C<fork()>,
C<dump()>, C<chown()>, C<link()>, C<symlink()>, C<chroot()>,
C<setpgrp()> and related security functions, C<setpriority()>,
C<getpriority()>, C<syscall()>, C<fcntl()>, C<getpw*()>,
C<msg*()>, C<shm*()>, C<sem*()>, C<alarm()>, C<socketpair()>,
C<*netent()>, C<*protoent()>, C<*servent()>, C<*hostent()>,
C<getnetby*()>.
This list is possibly incomplete.

=item *

Various C<socket()> related calls are supported, but they may not
behave as on Unix platforms.

=item *

The four-argument C<select()> call is only supported on sockets.

=item *

The C<ioctl()> call is only supported on sockets (where it provides the
functionality of ioctlsocket() in the Winsock API).

=item *

Failure to spawn() a subprocess is indicated by setting $? to "255 << 8".
C<$?> is set in a way compatible with Unix (i.e. the exitstatus of the
subprocess is obtained by "$? >> 8", as described in the documentation).

=item *

You can expect problems building modules available on CPAN if you
build perl itself with -DUSE_THREADS.  These problems should be resolved
as we get closer to 5.005.

=item *

C<utime()>, C<times()> and process-related functions may not
behave as described in the documentation, and some of the
returned values or effects may be bogus.

=item *

Signal handling may not behave as on Unix platforms (where it
doesn't exactly "behave", either :).  For instance, calling C<die()>
or C<exit()> from signal handlers will cause an exception, since most
implementations of C<signal()> on Win32 are severely crippled.
Thus, signals may work only for simple things like setting a flag
variable in the handler.  Using signals under this port should
currently be considered unsupported.

=item *

File globbing may not behave as on Unix platforms.  In particular,
if you don't use perlglob.bat for globbing, it will understand
wildcards only in the filename component (and not in the pathname).
In other words, something like "print <*/*.pl>" will not print all the
perl scripts in all the subdirectories one level under the current one
(like it does on UNIX platforms).  perlglob.exe is also dependent on
the particular implementation of wildcard expansion in the vendor
libraries used to build it (which varies wildly at the present time).
Using perlglob.bat (or File::DosGlob) avoids these limitations, but
still only provides DOS semantics (read "warts") for globbing.

=back

Please send detailed descriptions of any problems and solutions that 
you may find to <F<perlbug@perl.com>>, along with the output produced
by C<perl -V>.

=head1 AUTHORS

=over 4

Gary Ng E<lt>71564.1743@CompuServe.COME<gt>

Gurusamy Sarathy E<lt>gsar@umich.eduE<gt>

Nick Ing-Simmons E<lt>nick@ni-s.u-net.comE<gt>

=back

This document is maintained by Gurusamy Sarathy.

=head1 SEE ALSO

L<perl>

=head1 HISTORY

This port was originally contributed by Gary Ng around 5.003_24,
and borrowed from the Hip Communications port that was available
at the time.

Nick Ing-Simmons and Gurusamy Sarathy have made numerous and
sundry hacks since then.

Borland support was added in 5.004_01 (Gurusamy Sarathy).

Last updated: 3 January 1998

=cut