LDFLAGS to the linker command when linking a python extension. This lets
us pass the right flags to correctly find libraries required for certain
extensions, e.g. zlib.so.
Also explicitly comment out a few extensions that aren't being built and
should never be.
Changes are:
- The incompatibilities between Hugs and the Haskell Graphics Library
have been fixed, and binaries for the HGL are now available on the
Hugs download page.
- The missing standard libraries Directory, CPUTime, Time and Locale
have been added along with a complete implementation of Haskell98 IO.
- Hugs is now delivered with most of the hslibs libraries installed
in the lib/exts/ directory. The added modules cover the Edison,
Parsec, HaXml, QuickCheck, concurrent, monad, and html subdirectories
of hslibs.
- The :set option now refuses the user to set a module search path
that doesn't contain the Prelude. This is to protect users from
accidentally rendering their Hugs setups unusable, esp. so on
Windows machines where the options are persisted to the Registry.
- MacOS X is now one of the supported unix ports, with pre-built
binaries available on the download page.
- Experimental support is provided for hierarchical module names,
where a module name A.B.C is mapped onto the file path
A/B/C{.hs,.lhs} and appended to each of the path prefixes in
HUGSPATH until the name of a readable file is found.
Enable this package for all platforms.
Added GAWK_ENABLE_PORTALS to mk/bsd.pkg.defaults.mk to enable/disable gawk
handling file names that start with `/p/' as a 4.4 BSD type portal file.
Changes from 3.0.4 to 3.0.5:
- bug fix release only.
Changes from 3.0.5 to 3.0.6:
- bug fix release only.
Changes from 3.0.6 to 3.1.0:
- A new PROCINFO array provides info about the process. The non-I/O /dev/xxx
files are now obsolete, and their use always generates a warning.
- A new `mktime' builtin function was added for creating time stamps. The
`mktime' function written in awk was removed from the user's guide.
- New `--gen-po' option creates GNU gettext .po files for strings marked
with a leading underscore.
- Gawk now completely interprets special file names internally, ignoring the
existence of real /dev/stdin, /dev/stdout files, etc.
- The mmap code was removed. It was a worthwhile experiment that just
didn't work out.
- The BINMODE variable is new; on non-UNIX systems it affects how gawk
opens files for text vs. binary.
- Gawk no longer supports `next file' as two words.
- On systems that support it, gawk now sets the `close on exec' flag on all
files and pipes it opens. This makes sure that child processes run via
system() or pipes have plenty of file descriptors available.
- If `--posix' is in effect, newlines are not allowed after ?:.
- Weird OFMT/CONVFMT formats no longer cause fatal errors.
- Diagnostics about array parameters now include the parameter's name,
not just its number.
- It is now possible to open a two-way pipe via the `|&' operator.
See the discussion in the manual about putting `sort' into such a pipeline,
though. (NOTE! This is borrowed from ksh: it is not the same as
the same operator in csh!)
- The close() function now takes an optional second string argument
that allows closing one or the other end of the two-way pipe to
a co-process. This is needed to use `sort' in a co-process, see
the doc.
- If TCP/IP is available, special file names beginning with `/inet'
can be used with `|&' for IPC.
- With `--enable-portals' on the configure command line, gawk will also
treat file names that start with `/p/' as a 4.4 BSD type portal file,
i.e., a two-way pipe for `|&'.
- Unrecognized escapes, such as "\q" now always generate a warning.
- The LINT variable is new; it provides dynamic control over the --lint
option.
- Lint warnings can be made fatal by using --lint=fatal or `LINT = "fatal"'.
Use this if you're really serious about portable code.
- A number of lint warnings have been added. Most notably, gawk will
detect if a variable is used before assigned to. Warnings for
when a string that isn't a number gets converted to a number are
in the code but disabled; they seem to be too picky in practice.
Also, gawk will now warn about function parameter names that shadow
global variable names.
- It is now possible to dynamically add builtin functions on systems
that support dlopen. This facility is not (yet) as portable or well
integrated as it might be. *** WARNING *** THIS FEATURE WILL EVOLVE!
- Profiling has been added! A separate version of gawk, named pgawk, is
built and generates a run-time execution profile. The --profile option
can be used to change the default output file. In regular gawk, this
option pretty-prints the parse tree.
- Gawk has been internationalized, using GNU gettext. Translations for
future distributions are most welcome.
- New asort() function for sorting arrays. See the doc for details.
- The match function takes an optional array third argument to hold
the text matched by parenthesized sub-expressions.
- The bit op functions and octal and hex source code constants are on by
default, no longer a configure-time option. Recognition of non-decimal
data is now enabled at runtime with --non-decimal-data command line option.
- Internationalization features available at the awk level: new TEXTDOMAIN
variable and bindtextdomain() and dcgettext() functions. printf formats
may contain the "%2$3.5d" kind of notation for use in translations. See
the texinfo manual for details.
- The return value from close() has been rationalized. Most notably,
closing something that wasn't open returns -1 but remains non-fatal.
- The array effeciency change from 3.0.5 was reverted; the semantics were
not right. Additionally, index values of previously stored elements
can no longer change dynamically.
- The new option --dump-variables dumps a list of all global variables and
their final types and values to a file you give, or to `awkvars.out'.
- Gawk now uses a recent version of random.c courtesy of the FreeBSD
project.
- The gawk source code now uses ANSI C function definitions (new style),
with ansi2knr to translate code for old compilers.
- `for (iggy in foo)' loops should be more robust now in the face of
adding/deleting elements in the middle; they loop over just the elements
that are present in the array when the loop starts.
initial import of onyx 3.0.2 package. from DESCR:
Onyx is an embeddable stack-based threaded interpreted language. This package
contains both a stand alone interpreter and a library that can be used to embed
Onyx in an application. Extensive documentation is included.
WWW: http://www.canonware.com/
which only differs in setting the entry point. Let this use the default
linker script for alpha but force the entry point so this will work on older
NetBSD installs for alpha.
Not very many overall changes. Main ones include
1. Support for powerpc, arm32 and vax
2. Makefile.gcc can now be included by anything which depends on gcc versions.
If the version installed isn't 2.95.3 it'll add itself as a BUILD_DEPENDS.
(XXX: any of the makefile's in pkgsrc should be checked and change to use
this)
3. Remove special PLIST.NetBSD-sparc as it's no longer needed
4. Change post-extract loop to pick up any arch files from FILESDIR without
having to hardcode all the archs
5. Remove arch restrictions as this should work on any arch supported by the
main source tree as of 03/28/02
6. Add PKGREVISION as this clearly isn't stock 2.95.3 (it doesn't change
gcc --version so version checks won't care).
as suggested by Johnny Lam
-don't try to build the "locale" module - it doesn't work anyway
-pull up a fix from Python CVS to get more consistent math exceptions
(eg pow(1e-200,2) vs. math.pow(1e-200,2))
This will be part of 2.2.1.
${MACHINE_GNU_PLATFORM} moved ${RUBY_SITEARCHLIBDIR} on some ports,
and also bump the dependency and PKGREVISION in ruby-gtk, which is
the only package that seems to use that.
from its --host argument, which is exactly ${MACHINE_GNU_PLATFORM}. Use that
for the PLIST instead of ${MACHINE_ARCH}--${LOWER_OPSYS}, which may lack the
"elf" suffix on certain platforms.
This code was assuming it could copy double's in/out of char *'s with just
casts and normal copies. This blows up on anything which enforces alignments.
Change the generic case for the ATOM to just have a special double field.
For the serialization routines memcpy the double in and provide a union to
memcpy it out to that also contains a single double value. This ensures
alignment is correct and it won't SIGBUS anymore.
Bump pkg to nb1
Provide a trampoline implementation that doesn't presume r11 is free (it's not
in dynamically linked programs and the ELF ABI says as much but somehow linux
managed to avoid this it seems). Use r13 for the time being since while gcc
will allocate it, it's the last "local" one allocated so none of the clisp
code hits that (nothing needed 17+ local registers in use ever). Update the
vacall implementation to match the trampoline calls but otherwise for
any assembly use the linux code.
Make these changes specific to netbsd, update the test cases so everything
works.
This now passes all the clisp tests when done/installed.
all dependencies on packages depending on "png" which contain shared
libraries, all for the (imminent) update to the "png" package.
[List courtesy of John Darrow, courtesy of "bulk-build".]
and buildlinked by me.
Changes:
Now distributed under the GPL.
All .fas files generated by previous CLISP versions are invalid
and must be recompiled.
Hostname resolution is now optional in EXT:SOCKET-STREAM-PEER and
EXT:SOCKET-STREAM-LOCAL. EXT:SOCKET-STATUS now accepts SOCKET-SERVERs too,
and the direction of the checks can be specified. Added install.bat for
win32 installation. Fixed handling of circular structs and pointers to
functions in the FFI. Fixed binary I/O for streams with element type longer
than one byte, but not a whole number of bytes.
Summary of changes:
- removal of USE_GTEXINFO
- addition of mk/texinfo.mk
- inclusion of this file in package Makefiles requiring it
- `install-info' substituted by `${INSTALL_INFO}' in PLISTs
- tuning of mk/bsd.pkg.mk:
removal of USE_GTEXINFO
INSTALL_INFO added to PLIST_SUBST
`${INSTALL_INFO}' replace `install-info' in target rules
print-PLIST target now generate `${INSTALL_INFO}' instead of `install-info'
- a couple of new patch files added for a handful of packages
- setting of the TEXINFO_OVERRIDE "switch" in packages Makefiles requiring it
- devel/cssc marked requiring texinfo 4.0
- a couple of packages Makefiles were tuned with respect of INFO_FILES and
makeinfo command usage
See -newly added by this commit- section 10.24 of Packages.txt for
further information.
The primary objective of this release was bug fixes, not new features.
Highlights of the release follow, see the ChangeLog for more detail and
the code for even finer detail.
Build
- --enable-debug now works and provides correct CXXFLAGS
- firend class issues brought into compliance with lang spec
- VC++ enablement of __int64 for long long
- tolerate non-ISO conformant encoding names when searching
- correct namespace generation in java.g (so the parser won't
regress every time we update the grammer again)
- include jikesapi.h in the distribution (Note that this is
at best a very early alpha preview)
- prevent bad code generation by VC++ in IEEEfloat
- finished removal of dead EBCDIC code - to be replaced by
use of JikesAPI class latter
Parser
- remove some parenthesized expressions not allowed by JLS
- synchronized(null){} isn't valid, don't allow it
- prevent core dump when local class method omits return type
- inner classes and static members error handling cleanup
- fix assertion `this_type -> HeaderProcessed()' failures
- don't allow invalid combinations of abstract, native and strictfp
- don't allow qualified explicit this constructor
- make sure string constants are compiled inline
- fix assertion from NULL in SemanticError::Report
- be more compliant with JLS 14.4.2 and scope of duplication of token
- fix regression in 1.14 when ?: mixed primitive and reference types
Emitter
- reverse the logic of large branches to use goto_w and jsr_w
for branch offsets requiring more than two bytes
- never set ACC_STATIC bit on local classes
- make sure private methods and anonymous classes are always
final, and anonymous classes are never static.
- second round of JPDA pampering; believed to make it happy now
- prevent LocalVariableTable_attribute::AddLocalVariable assertion
- fix VerifyErrors caused by reuse of local vars in try/catch and
synchronized blocks
User Interface
- correct classpath handling issues on cygwin
- treat @files as one argument per line, regardless of whitespace
embedding friendly, as suggested in PR pkg/14520 by Jarkko Torppa
-having to touch Setup.in anyway, comment out the non-64-bit-clean
modules regularily and remove the Makefile magic which led to the
same effect
on any platform (we can make a separate pkg for it later)
-pull in a patch which used to be in devel/py-readline here because
it is Python version specific
-bump PKGREVISION
changing the default module/library search path to have site_perl come
before the standard directories. In other words, the previous search path
on an i386 was:
/usr/pkg/lib/perl5/5.6.1/i386-netbsd
/usr/pkg/lib/perl5/5.6.1
/usr/pkg/lib/perl5/site_perl/5.6.1/i386-netbsd
/usr/pkg/lib/perl5/site_perl/5.6.1
/usr/pkg/lib/perl5/site_perl
but it is now:
/usr/pkg/lib/perl5/site_perl/5.6.1/i386-netbsd
/usr/pkg/lib/perl5/site_perl/5.6.1
/usr/pkg/lib/perl5/site_perl
/usr/pkg/lib/perl5/5.6.1/i386-netbsd
/usr/pkg/lib/perl5/5.6.1
The rationale for this is that when we install a module that is newer than
one in the standard library, the new module goes into the site_perl
directory as it's an add-on module. However, we can't use the newer module
without modifying either the scripts of the perl environment to find the
newer module explicitly because of the order of the library search path:
the site_perl directories come after the standard directories. The normal
solution is to directly replace the module in the standard library with
the newer module. However, this isn't really on option when installing
via pkgsrc because the older module files are owned by the perl package.
By placing the the site_perl directories before the standard directories,
newer modules that we install via pkgsrc are simply found before the older
ones in the standard library.
pure bugfix release
-install the "pydoc" script as "pydoc2.1" and make sure the corresponding
python version is used
-disable the audioop, imageop and rgbimg extensions - these don't work on
64-bit platforms anyway, I haven't seen a place where they are used yet,
and in case a need shows up, we better make an own extension pkg from
it which is marked FOR_PLATFORM_ONLY
o improve SortedCollection performance and ensures that the Directory
class>>#image method returns the *current* rather than the default
image path.
o fix a few bugs that were apparent when trying to extend base classes.
o bugfixes in the I/O subsystem, missed a few I/O available events if
many happened in a row.
o The #(a b) syntax for symbols inside Arrays has been obsoleted,
since 2.0 will parse it according to the ANSI standard.
o Calls to the virtual machine from plugins, and objects that are passed
as OOPs in call-outs, put OOPs in the incubator rather than in the
registry; call-outs are wrapped in incSavePointer/incRestorePointer.
o Command line parsing uses getopt and thus behaves exactly like other
programs (previously there were some discrepancies)
o Errors are signaled if a file specified on the command line is not found.
o Events can be passed to the Smalltalk image via an ObjectMemory class.
o Fixed bug in evalExpr and typeNameToOOP (gave a parse error).
o Removed the `make optimize' mess.
o Supported two additional ways to pass objects from Smalltalk to C:
#selfSmalltalk and #variadicSmalltalk, which are similar to respectively
#self and #variadic but pass raw object pointers to the C function
instead of attempting automatic conversions.
o #bindWith:... methods now accept other objects than Strings as parameters.
o Complete hierarchy of exceptions, with more meaningful error message
and possibility of more fine-grained exception handling.
o FileStream calls are not blocking and can preempt the current Process.
o FileStream handling has been rewritten; the buffering is now done
by Smalltalk code rather than implied in stdio. Unbuffered file
descriptor access (which used to be provided by UnixStream, defined
by the TCP package) is provided by FileStream's parent, FileDescriptor.
o Many methods in SystemDictionary were moved to ObjectMemory (a new class);
the old ones are now deprecated.
o SortedCollection's #includes:, #indexOf:, and #occurrencesOf: can
check for objects that could not be inserted in the collection
(e.g. an Integer in a collection of Strings). Fixed bugs in the same
methods related to sort blocks for which sort-block equality
(a <= b and b <= a) does not imply equality.
o Support for init blocks will be removed in a future version, as it
was replaced by the much more powerful ObjectMemory class.
o The SystemDictionary>>#enableGC: method does not exist any more, since
it only caused harm (the correct way to obtain its effect is to use the
incubator, since what we want is to unregister a batch of many objects
at the same time).
o The TCP library does not poll the socket for I/O, but relies on the
system's preemptive I/O facilities. As a result, the polling period
methods in Socket have disappeared.
o Usual round of bug fixes
New goodies:
o MD5 checksums
o Perl regular expressions
o Support for localization, internationalization and multiple character sets
Makefile and PLIST files, using PLIST_SUBST accordingly.
Also use an ALT_ARCH definition, with PLIST_SUBST, to make the two PLISTS
more generic, and set up ALT_ARCH accordingly.
NetBSD Packages Collection.
Dylan is an object-oriented, dynamic, infix, garbage-collected
programming language with support for multiple inheritence, multiple
dispatch (an advanced form of polymorphism), typed and untyped
variables, closures and exceptions. Dylan also supports pattern-based
hygenic macros. These allow you to define new control constructs and
fully integrate them with the language.
Dylan (theoretically) combines the performance of C or C++ with the
rapid development of Perl and the expressiveness of LISP. It looks
similar to C or Pascal, so experienced programmers can learn to write
simple programs quickly.
Gwydion Dylan provides two implementations of the Dylan programming
language: Mindy and d2c. Mindy is bytecode compiler and interpreter,
and d2c is a Dylan-to-C compiler. Mindy compiles programs quickly,
but the resulting executables run slowly. On the other hand, d2c
compiles programs slowly, but they run quickly.
Gwydion Dylan was originally written by the Gwydion Group at CMU as
part of a research project studying advanced hypercode development
environments. It is now maintained by a group of volunteers.
the NetBSD Packages Collection.
Dylan is an object-oriented, dynamic, infix, garbage-collected
programming language with support for multiple inheritence, multiple
dispatch (an advanced form of polymorphism), typed and untyped
variables, closures and exceptions. Dylan also supports pattern-based
hygenic macros. These allow you to define new control constructs and
fully integrate them with the language.
Dylan (theoretically) combines the performance of C or C++ with the
rapid development of Perl and the expressiveness of LISP. It looks
similar to C or Pascal, so experienced programmers can learn to write
simple programs quickly.
Gwydion Dylan provides two implementations of the Dylan programming
language: Mindy and d2c. Mindy is bytecode compiler and interpreter,
and d2c is a Dylan-to-C compiler. Mindy compiles programs quickly,
but the resulting executables run slowly. On the other hand, d2c
compiles programs slowly, but they run quickly. This package contains
mindy.
Gwydion Dylan was originally written by the Gwydion Group at CMU as
part of a research project studying advanced hypercode development
environments. It is now maintained by a group of volunteers.
For just now, the only supported platform is NetBSD/i386, but they
are fairly easy to add, so all contributions gratefully received.
Regen the patches files (one file per patch).
Fix for NetBSD ELF platform: files/stab-elf.c and config/untested/elf-netbsd-cc
(by way of patches/patch-af) adjusted.
Thanks to Nick for the helping hand and to Alistair for the previous work.
This should close PR 13527.
Make X11 support conditional by way if the ELK_USE_X11 variable.
I tried to enable Motif support: only Motif 1.X should work but
even with the lesstif12 package it does not - it does compile though.
So Motif support if commented out but leaved in case someone want to fix
ELK.
version 2.95.2. This package includes all changes to "gcc" from the
new toolchain in NetBSD-current. The only tested (and enabled) platform
is "NetBSD-*-i386" so far.
see ${LOCALBASE}/share/doc/ruby/NEWS file.
plus two fixes noted on ruby-dev@ruby-lang.org.
[ruby-dev:15549] lib/irb/locale.rb
[ruby-dev:15551] lib/weakref.rb
And some ruby pakcage own changes:
- Handle proper RUBY_ARCH variable on current.
- REPLACE_RUBY is processed by sed(1) not ruby.
- Proper handle with RUBY_EXTCONF.
- USE_RUBY_SETUP dosen't implicit define USE_RUBY_AMSTD now.
Besides more bug fixes and several enhancements some of the
NetBSD relevant changes (taken from "changes" in the
distribution):
2001-04-03 (doc fixes) numerous doc corrections and clarifications.
Update of READMEs.
2001-03-29 (bug fix) prevent potential race condition and security leak in
tmp filename creation on Unix. (max)
2001-03-13 (bug fix) Correctly possible memory corruption in string map {}
$str (fellows)
2001-01-30 (bug fix) Fixed possible hangs in fcopy. (porter)
2000-11-23 (mem leak) fixed potential memory leak in error case of lsort
(fellows)
2000-11-01 (mem leak) Corrected excessive mem use of info exists on a
non-existent array element (hobbs)
2000-09-27 (bug fix) fixed a bug introduced by a partial fix in 8.3.2 that
didn't set nonBlocking correctly when resetting the flags for the write
side (mem leak) Correct mem leak in channels when statePtr was released
(hobbs)
2001-07-18 (bug fix) corrected memory overwrite error when buffer size
of a channel is changed after channel use has already begun (kupries, porter)
2001-08-07 (bug fix) corrected bytecode stack management during [break]
(see test foreach-5.5) (sofer, tallneil, jstrot)
2001-08-08 (new features) updated packages msgcat 1.1.1, opt 0.4.3,
tcltest 1.0.1, dependencies checked (porter)
2001-08-24 (bug fix) [auto_import] now matches patterns like
[namespace import], not like [string match] (porter)
**** POTENTIAL INCOMPATABILITY ****
2001-08-27 (new feature) added Tcl_SetMainLoop() to enable loading Tk as a
true package (hobbs)
2001-09-06 (new feature) http 2.4: honor the Content-encoding and charset
parameters; add -binary switch for forcing the issue (hobbs, saoukhi, orwell)
2001-09-10 (bug fix) protect against alias loops (hobbs)
2001-09-19 (bug fix) [format] and [scan] corrected for 64-bit machines (rmax)
2001-09-26 (bug fix) corrected potential deadlock in channels that do not
provide a BlockModeProc (kupries, kogorman)
2001-10-11 (bug fix) corrected cleanup of self-referential bytecodes at
interpreter deletion (sofer, rbrunner)
references of the pkglint package.
_PKGSRCDIR is an internal definition in bsd.pkg.mk, and a few packages
which would like to refer to other packages in the build tree. It should
not be set by users, but neither should it stop a user from building a
package if it is defined, so make it obvious that this is the case.
provide a way to avoid building a statically linked perl on platforms
where it doesn't matter. Currently, by empirical evidence in pkg/14871,
this includes mipsel and probably mipseb. Other platforms can add
themselves if/when they discover it doesn't matter for them either.
Closes pkg/14871 by John Hawkinson <jhawk@mit.edu>.
we don't accidentally add options that the linker doesn't understand, such
as "-Wl,-R*". This should fix pkg/14907 by John Klos john@sixgirls.org
where the a.out linker ld doesn't understand extra flags passed in from
LDFLAGS from the pkgsrc environment when building a perl package.
from the distfile version number. G/C the version number stuff from
perl5/Makefile.common, preserving only PERL5_DIST_VERS as it's still used
by libperl.
pkgsrc. Instead, a new variable PKGREVISION is invented that can get
bumped independent of DISTNAME and PKGNAME.
Example #1:
DISTNAME= foo-X.Y
PKGREVISION= Z
=> PKGNAME= foo-X.YnbZ
Example #2:
DISTNAME= barthing-X.Y
PKGNAME= bar-X.Y
PKGREVISION= Z
=> PKGNAME= bar=X.YnbZ (!)
On subsequent changes, only PKGREVISION needs to be bumped, no more risk
of getting DISTNAME changed accidentally.
buildlink.mk file can replace the use of USE_PERL5: it defines all of the
same perl variables as bsd.pkg.mk, adds a dependency on perl>=${PERL5_REQD}
and conditionally includes bsd.perl.mk. It also creates a buildlinkified
Config.pm that makes the local include/library search path include
${BUILDLINK_DIR} instead of hard-coding ${LOCALBASE}. There's more to be
done to make this more useful in strongly buildlinkifying a package, but
it's useful now as-is in weakly buildlinkifying a package.
Pr 13866 by Brian Gregor.
New Features Include:
- A Foreign Function Interface closely modeled after the one
provided by GHC.
- Built-in, Hood-like debugging support.
- A new syntax for recursive monad bindings.
- A new GUI under Windows that doesn't consume all CPU time.
Changes from 5.32.1 (the last pkgsrc version):
10/1/01 5.33.0 <mark@doradosoftware.com> found that expect's diagnostics
didn't include the "no" after testing for a full buffer.
Hemang Lavana <hlavana@cisco.com> noted that "debug" (Dbg_On)
calls didn't always force the debugger into step mode.
Martin Kammerhofer <dada@sbox.tugraz.at> noted that the man
page neglected to document interpreter -eof.
Chris Clare <clarec@nortelnetworks.com> provided fix for
multiple decl in C lib.
Sheng Wang <wangs@sh.bel.alcatel.be> found interact's
can-match code had broken. It was missing the special hook
that Henry had added just for this purpose. How strange.
Dieter Fiebelkorn <dieter@fiebelkorn.net> requested addition
to config.guess for Power*Macintosh:Darwin for MacOSX.
Aside - to download latest config.guess:
cvs -d :pserver:anoncvs@subversions.gnu.org:/cvs checkout \
config
Added pipeline example to unbuffer man page.
8/4/00 5.32.2 Allen J. Newton <anewton@alturia.fleet.org> provided code for
generating passwords with special characters in mkpasswd.
Brent Welch <welch@ajubasolutions.com> changed the fix1line
install script so that "autoexpect" and other scripts that
get installed into the platform-independent bin directory
generically invoke "expect" from the users PATH instead
of hardwiring the platform-specific expect pathname.
Changes:
* New module `rep.util.md5', has two functions for generating MD5
message digests (of files or strings)
* Changes to the `rep.io.sockets' function:
In the `socket-server' function the HOST and/or PORT arguments may
be false, meaning to listen on all addresses and to choose a
random unused port.
New functions `socket-peer-address' and `socket-peer-port', these
always returns the details of the far end of the connetion.
`socket-address' and `socket-port' have been changed to always
return the details of the local connection point.
* New function in `rep.system' module, `crypt'. A wrapper for the
system's `crypt' function (if it has one)
* New function in `rep.threads' mdoule, `make-suspended-thread'
* New module `rep.net.rpc', provides a text-stream based RPC
mechanism for Lisp programs. Similar in some ways to untyped CORBA.
(This is still in the experimental stage - its interface may
change in forthcoming releases)
* New functions in `rep.data' module, `list->vector' and
`vector->list'
* New macro `define-special-form'. A combination of `defvar' and
`setq' - it always makes the variable special and it always sets
it to the given value
* New module `rep.test.framework' implementing `assert', `check' and
`test' macros. This provides a framework for implementing unit
tests in Lisp modules (such that running the interpreter with the
`--check' option will run all tests that have been set up to be
autoloaded
The automatic truncation in gensolpkg doesn't work for packages which
have the same package name for the first 5-6 chars.
e.g. amanda-server and amanda-client would be named amanda and amanda.
Now, we add a SVR4_PKGNAME and use amacl for amanda-client and amase for
amanda-server.
All svr4 packages also have a vendor tag, so we have to reserve some chars
for this tag, which is normaly 3 or 4 chars. Thats why we can only use 6
or 5 chars for SVR4_PKGNAME. I used 5 for all the packages, to give the
vendor tag enough room.
All p5-* packages and a few other packages have now a SVR4_PKGNAME.
directly against -lperl when built. Combined with the previous update of
perl to add ${LOCALBASE}/lib to the rpath when creating shared
libraries/modules, these two changes make using mod_perl.so (ap-perl) more
painless. All perl shared modules now contain interlibrary dependencies to
the shared libraries they need. Instead of needing (at least on ELF):
LoadFile !/usr/lib/libm.so
LoadFile lib/perl5/5.6.1/i386-netbsd/CORE/libperl.so
...
# Any other libraries needed by perl shared modules need to listed
# with LoadFile here.
...
LoadModule perl_module lib/httpd/mod_perl.so
AddModule mod_perl.c
you'd need only the last two lines as the mod_perl authors intended.
I've tested this patch for many weeks now, successfully loading and using
the p5-Apache-ASP module as my test bed, and I haven't noticed any problems
with normal perl usage.
Also comment the Makefile slightly better.
