PKGREVISION++ for the previous change.

(rillig)

Fixed a segmentation fault when calling "socat -hhh".

(rillig)

See also: PLIST_SUBST

(rillig)

When updating a package, the "(perhaps the packing list is incorrectly
specified?)" warning isn't correct, so don't print this assumption at
all.

(rillig)

Fixed some of the gcc4 issues. The package doesn't built yet on
NetBSD-4.99.30.

(rillig)

The first of a series of patches for gcc4.

(rillig)

Some commands like "mktexlsr" and "texconfig" must not be called from
package Makefiles.

(rillig)

Fixed build with sunpro.

(rillig)

Got rid of _PKG_SILENT. Why use __file__ when a simple f does the same?

(rillig)

Replaced unreadable code with readable code by doing a little
substitution.

(rillig)

STEP_MSG may be a hidden command.

(rillig)

Replaced _PKG_SILENT and _PKG_DEBUG with RUN, after making sure that the
additional error checking won't break anything that isn't intended to
break.

(rillig)

Don't run pkg_create with the -v option. It prints a temporary package
name which isn't correct, and the "Value of SrcDir" that is used is not
important to any pkgsrc user. Instead, let the pkgsrc infrastructure
print the package name.

(rillig)

Explicitly mention that the PKGREVISION must be incremented when a
dependency of the package changes. I still have problems with some
packages that depend on GConf2, which doesn't exist anymore. The
dependency has silently changed to GConf, without increasing any
PKGREVISION. Same for gnome-vfs2 and gnome-vfs.

(rillig)

Since revision 1.50 of mk/subst.mk, the test for text files is no longer
unreliable, so re-enable it.

(rillig)

Made the wording more into an explanation.

(rillig)

Added some patches to fix the use of the sigsend() function, which is
not available on NetBSD.

(rillig)

Fixing permissions is done in the post-extract stage, since it belongs
there, not in post-patch.

There's no need to use xargs -0: Solaris doesn't know that option, POSIX
doesn't require it, and all the filenames are sane anyway.

(rillig)

The package x11/gdm embeds ${exec_prefix} into a macro containing a
PATH, in which it is not directly preceded by a quote character:

#define GDM_USER_PATH "/usr/bin:/bin:${exec_prefix}"

Catch this by complaining about "${" also when it appears after a colon.

Since this check is still disabled by default, it will not surprise anyone.

(rillig)

Make sure that the files found by check_prog are really regular files
and nothing else. This prevents bootstrap from exiting just because
there is a subdirectory named "awk" (or another tool) in one of the PATH
directories.

Fixes PR 37806.

(rillig)

Explained the evolution of mount(2) in NetBSD and how to deal with it.

(rillig)

Renamed some *.mk files to *.help, since they only contain comments.

(rillig)

Added a file where difficulties using specific C functions can be
documented. The first one is mremap(), since its prototype differs
between Linux and NetBSD.

(rillig)

Added a dependency to devel/getopt, since a2x requires it. Suggested by
Nick Goffee via private mail.

(rillig)

PKGREVISION++ for the invalid definition of shell functions.

(rillig)

Fixed the definition of shell functions, as reported in PR 37795.

See also:
http://www.opengroup.org/onlinepubs/000095399/utilities/xcu_chap02.html#tag_02_09_05
http://www.opengroup.org/onlinepubs/000095399/utilities/xcu_chap02.html#tag_02_09_04

(rillig)

If the package doesn't provide a SUBST_MESSAGE, generate a default one.
This way, file modifications cannot go unnoticed and at least appear in
the build logs. While here, fixed an ambiguity in the description of
SUBST_MESSAGE.

(rillig)

Don't rely on file(1) to distinguish text files from binary files. This
has proven too unreliable in the past. For example, some Makefile.in
files were classified as "Quake I or II world or extension", just
because they happen to start with the letters "PACK". This method was
also subject to subtle differences in the locale.

The new method counts the number of NUL bytes in the file. It does not
depend on the locale settings. The -c option of wc(1) counts bytes, not
characters, and tr(1), which may interpret multibyte sequences, is
protected by LC_ALL. It should also work with the historical
implementations of tr(1) that could not handle NUL bytes and discarded
them, since this is exactly the intention.

See also:
* pkg/2006/07/05/0000.html">http://mail-index.netbsd.org/tech-pkg/2006/07/05/0000.html
* PR 37793

(rillig)

Replaced _PKG_SILENT and _PKG_DEBUG with RUN.

(rillig)

Needs libz.

>From PR 37760.

(rillig)

+ ruby18-libxml

(rillig)

Imported textproc/ruby18-libxml-0.3.8.4 from PR 37759

(rillig)

Imported ruby18-libxml from PR 37759.

The Libxml-Ruby provides Ruby language bindings for the GNOME Libxml2
XML toolkit. Its primary advantage over REXML is performance - if speed
is your need, these are good libraries to consider.

Packaged by Naoto Morishima.

Status:

Vendor Tag: TNF
Release Tags: pkgsrc-base

(rillig)

sunpro needs some patches.

(rillig)

Needs bison and flex as tools, which obsoletes a BUILD_DEPENDS.

(rillig)

Added Solaris libraries. Removed patch-aa, since it is easier to just
override do-build.

(rillig)

+ Needs libtool.

(rillig)

Fixed build failures with sunpro.

(rillig)

Broken gettext detection.

(rillig)

Shortened the definition of a meta-package.

(rillig)

Sunpro fix.

(rillig)

Added a patch for sunpro.

(rillig)

Added a patch for sunpro.

(rillig)

Fixed ambiguous function call with sunpro.

(rillig)

Fixed typos.

(rillig)

Greg wants me to check for unused license files.

(rillig)

Check misspelled macro names, for example __NetBSD_Version instead of
__NetBSD_Version__. Inspired by emulators/wine/patches/patch-ac 1.37.

(rillig)

Explained the difference between an lvalue and an rvalue.

Documented the __STDC__ macro, especially Sun's implementation.

(rillig)

Made gdb build again on NetBSD-current. The cause was that <sys/vmparam.h>
uses kmutex_t, but doesn't include <sys/mutex.h>, where it is defined.

(rillig)

+ REPLACE_LOCALEDIR_PATTERNS

(rillig)

Some changes take a long time. This package had been broken since two
years. The fix was trivially copied from devel/coconut, which had the
same problem concerning an "invalid lvalue in increment".

(rillig)

C compilers also emit error messages that are worth being explain.

(rillig)

Without kernel sources, this package cannot be built.

(rillig)

Fixed invalid C++ code and some pkglint warnings.

(rillig)

+ BUILD_MAKE_CMD

(rillig)

Updated audio/xmms-wma to 1.0.5

(rillig)

Updated xmms-wma to 1.0.5.

Changes in 1.0.5
- Added opportunity to build plug-in for player BMP.
- Fix bug with blanks with use gnome-vfs in BMP.
- WARNING: BEEPMP-WMA conflict with BMP-WMA. Remove BMP-WMA or do not use BEEPMP-WMA.

(rillig)

Explained an error message from g++ that occurs quite often.

(rillig)

NetBSD defines KERN_ARND to a value different to 37.

(rillig)

Dragonfly is not the only platform that doesn't know about
_SC_NPROCESSORS_ONLN, NetBSD is another.

(rillig)

Since X11_TYPE cannot be "xorg", there's no point in checking for that
condition. What was really meant is X11_TYPE != "modular".

(rillig)

Deactivated the "NOTE: you can write "tool" instead of "${TOOL}", as
suggested by darcy@.

(rillig)

When a variable is compared to a string in an .if or .elif directive,
check that the string fits to the variable type.

Currently, the generated warnings are mainly for X11_TYPE, for which
"xorg" is not a valid value. (hi joerg!)

(rillig)

pkgsrc doesn't have "xorg" anymore.

(rillig)

Updated devel/kscope to 1.6.0

(rillig)

Update kscope to 1.6.0.

Changes since 1.5.0:
* Improved: Updated the documentation
* Improved: Implemented multiple-call nodes in the call-graph
* Improved: Workspace toolbar buttons re-arranged
* Improved: Hitting Enter in the "Open Project" dialogue's list of
  recent projects opens the selected project
* Bug fix: Handle file extensions correctly in the file list
* Bug fix: Avoid hanging status bar messages when a project is closed
* Bug fix: Prevent infinite loops in the file finder in the presence of
  recursive symbolic links
* Bug fix: Child windows should not disappear when dialogues are invoked
* Bug fix: Added padding to the call graph, to avoid chopped-off drawing
  of nodes close to the borders
* Bug fix: Honour the preferred font when drawing the graph
* Bug fix: Delete temporary dot files used for drawing graphs
* Bug fix: Save call-tree files when a project is closed
* Bug fix: Clean-up a project's directory name after it is created

Maintainer update by Martijn van Buul via PR 37701.

(rillig)

Improved an explanation for an error message, since the previous version
didn't really explain anything.

(rillig)

Explicitly record the PKGNAME via the @name command. That way, it is
possible to create the package file using a temporary file first, and if
everything has succeeded, to rename it to the real name. This time, I
tested it creating various binary packages and installing them
afterwards, so I'm pretty sure it works now.

(rillig)

Improved the wording of the "Makefiles must not be included" error message.

(rillig)

Removed options.mk, since it is not included anymore.

(rillig)

Patch files start with the CVS Id.

(rillig)

Wrote ${.ALLSRC} instead of $> to make the file more readable for those
who don't see this variable daily. This also prevents a pkglint internal
error.

(rillig)

ERROR: patches/patch-ab:45: This code must not be included in patches.

(rillig)

ERROR: patches/patch-qa:29: This code must not be included in patches.

(rillig)

ERROR: patches/patch-ab:71: This code must not be included in patches.

(rillig)

Fixed name mismatch.

(rillig)

Fixed category name in BUILDLINK_PKGSRCDIR.

(hi wiz!)

(rillig)

Fixed tools definition, as reported by pkglint.

(rillig)

ERROR: patches/patch-ac:35: This code must not be included in patches.

(hi joerg!)

(rillig)

Patch files have to start with a CVS Id.

(rillig)

Fixed naming mismatches reported by pkglint.

(rillig)

Fixed a few pkglint warnings.

(rillig)

Sent the patch upstream.

(rillig)

Updated chat/amsn to 0.97

(rillig)

Updated amsn to 0.97.

No ChangeLog available. :(

Fixes PR 37509.

(rillig)

Upper-case keywords may also end with a colon. This is used in
mk/check/check-portability.mk.

(rillig)

Nuked 45 patches that were all doing the same mechanical transformation.
We have the SUBST framework for a reason.

(hi joerg! And don't tell me it's too unreliable, I won't believe it.)

(rillig)

Don't give a warning for incomplete RCS tags in patch files, as they
probably will not do any harm.

(rillig)

Reactivated the code to warn about missing comments in patches. It seems
that the time has come that we can no longer find out what a patch does,
unless we document it. The CVS history is not of great help in many
cases, patches get moved in the repository and so on.

To not scare casual pkglint users, these warnings are only shown if
pkglint is run with the -Wstyle or -Wall option.

(rillig)

Rewrote the help parser. You can get a list of all help topics by
running "bmake help topic=:index" now. This change is in preparation of
importing this help into the pkgsrc guide. There are still too many
false positives to be useful.

(rillig)

The @imake-man macro should not yet be used, since it is subject to
further changes.

(rillig)

Instead of just recording whether a topic is relevant or not, collect
all applicable keywords (like variable names, make targets or explicit
keywords). This will help printing an index of all keywords.

(rillig)

Fixed gcc4 build failures.

(rillig)

+ AUTO_MKDIRS

(rillig)

+ AUTO_MKDIRS

(rillig)

It's a META_PACKAGE.

(rillig)

In the check for variable ordering in package Makefiles, MASTER_SITES is
not required to appear, since META_PACKAGEs don't need to define it.

(rillig)

It's a META_PACKAGE.

(rillig)

It's a META_PACKAGE.

(rillig)

Removed the special-case handling of PKG_SYSCONFDIR for NetBSD. Now the
configuration files are installed in the usual pkgsrc place, not in
/etc. PKGREVISION++

Ok'ed by jlam@.

(rillig)

Added code to analyze the conditions in .if and .elif directives. This
will soon be extended to checking for spelling mistakes in PKG_OPTIONS
handling.

(rillig)

Whenever a Makefile.common is included by another file, that file should
be mentioned in the Makefile.common.

(rillig)

Updated pkgtools/pkglint to 4.82

(rillig)

Updated pkglint to 4.82.

Changes since 4.81:
- Moved lintpkgsrc to its own package.
- Less internal errors when parsing shell code.
- White-space checks in PLIST.

(rillig)

Defining a meta-package is done via the META_PACKAGE variable.

In Makefile.common, the MAINTAINER should only be set with the ?= operator.

(rillig)

Adjusted the guide to the separation of pkglint and lintpkgsrc.

(rillig)

Replaced all references to pkglint with lintpkgsrc, which has its own
package since a few days.

(rillig)

If a command line argument is used, it should be mentioned by exactly
this name (pkg-name) and not some related description, to make searching
easy.

(rillig)

Reverted the change that tried to make binary packages more sane because
it had severe consequences: pkg_create gets lots of information from the
filename into which the package is written. The extension decides what
compression to apply, and the basename gets recorded as the @name. This
part needs more work.

Noticed by stoned@.

(rillig)

Made the .include directives simpler, since the directory of the
including file is always the first in the search path.

(rillig)

Prepared pkglint for the upcoming cleanup of the .include directives in
mk/*.

(rillig)

Fixed pkglint warnings.

(rillig)

When creating the binary package, first create a temporary file, and if
everything went well, rename it to the real name. That way, it is less
likely that broken binary packages are created. It is a common
assumption that binary package files, if they exist, are usable.

An example for a broken binary package is security/sudo-1.6.9p10, in
which sbin/visudo wasn't readable when creating the package as an
unprivileged user.

(rillig)

Install the binaries readable for the owner, so that a package can be
created in unprivileged pkgsrc mode. PKGREVISION++

(rillig)

When calling open(2), don't try to get an error message from libarchive,
since it won't know anything about it. (hi joerg!)

(rillig)

+ lintpkgsrc

(rillig)

Imported pkgtools/lintpkgsrc-4.82

(rillig)

Imported lintpkgsrc.

lintpkgsrc checks various things on the complete pkgsrc tree.

Status:

Vendor Tag: TNF
Release Tags: pkgsrc-base

(rillig)

Fixed a pkglint warning, which revealed a naming conflict. Fixed that
too.

(rillig)

Added a way to determine the PKGPATH when checking a package.

(rillig)

regen

(rillig)

Patches that replace the == argument to test(1) with a single = don't
need to be commented upon at all. Fixes PR 37167.

(rillig)

I don't have access to IRIX anymore.

(rillig)

If one of the NO_*_ON_* variables contains something other that
${RESTRICTED}, print the variable name in the warning message.

While here, added an explanation for the warning.

(rillig)

All characters in PLIST files are relevant, so make trailing white-space
an error. Who would ever want to install files whose name ends in
white-space?

(rillig)