Use EXIT_FAILURE, rather than hardcoding a value of '1' in err(3) and
errx(3).
Plug a memory leak by modifying the signature of
_pkgdb_getPKGDB_FILE() to take a buffer and its size, and fill in that
buffer and return it.
Add an enumerated type which is used in pkgdb_open() to specify
read-only or read-write mode.
Modify the return value of pkgdb_open() to have a more logical boolean
value.
- replace occurrences of "make" with @MAKE@
- add @MAKE@ to the list of variables to get sod (I never learned
properly how to conjugate "sed")
- warn about NO_PATCH, since it hinders more than it helps
- remove NO_PATCH from Makefile to quell new pkglint warning
punt and invoke "make show-vars VARNAMES=PKGNAME" and get make to tell
us.
This addresses the issue of:
Cannot extract digest-UNDEFINED version (/usr/pkgsrc/pkgtools/digest/Makefile)
Cannot extract pkg_install-UNDEFINED version (/usr/pkgsrc/pkgtools/pkg_install/Makefile)
Unknown package: 'digest' version 20010807
Unknown package: 'pkg_install' version 20021123
Patch reviewed by the wiz.
Changes:
- Support XFree86 4.2.99.3, so we support the upcoming 4.3
- There is no need to link bitmap files, which was introduced while
switching to bulidlink2
- Add the following new files:
include/X11/Xcursor/Xcursor.h
include/X11/extensions/bigreqstr.h
include/X11/extensions/extutil.h
include/X11/extensions/mitmiscstr.h
include/X11/extensions/multibufst.h
include/X11/extensions/shapestr.h
include/X11/extensions/shmstr.h
include/X11/extensions/syncstr.h
include/X11/extensions/xcmiscstr.h
include/X11/extensions/xteststr.h
lib/llibXcursor.a
lib/libXcursor.so
lib/libXcursor.so.1
lib/libXcursor.so.1.0
lib/libXrandr.so.2
lib/libXrandr.so.2.0
lib/libXrender.so.1.2
lib/libXv.so
lib/libXv.so.1
lib/libXv.so.1.0
(1) Graph recoloring, so that as you delete and rebuild bunches of
packages, you can see how far along you are (I use this a *lot*).
(2) Subgraph grouping (based on work done by David Maxwell) which
clusters groups of packages with common prefixes in their names inside
a box.
(3) The explicit dependency on the graphviz pkg has been removed so
that pkgdepgraph can be run on "client machines" that only wish to
generate graph data, not to view it (ie, machines that have packages
installed but which do not wish to incur the overhead of X or the
other stuff that graphviz requires). The fact that you should install
graphviz *somewhere* so that you can view the data is noted by the new
INSTALL script.
Some useless comments have also been removed.
- Added an EXTRAMK variable to allow the addition of any kind of stuff
into the generated mk.conf.
- While creating the template, quote ${LOCALBASE} so it does not get
replaced. Reported by Douglas Brebner in PR pkg/19509.
databases/db/buildlink2.mk.
On platforms that lack Berkeley DB, pkg_install will build and
install from bootstrap-pkgsrc, but will be missing some important
functionality. On such systems, once bootstrapped, install this
package to get full functionality. (We'll take care of this
automatically in the future.)
Platforms that are reported to lack Berkeley DB in the base system:
Solaris, some Linuxes. There are probably others.
Suggested by "Jeremy C. Reed" <reed@reedmedia.net> on tech-pkg@.
This stops current loop:
- PKGTOOLS_REQD bump
- update pkg_install
- ZOULARIS_VERSION is too small, newer pkg_install required
- update pkg_install
- ZOULARIS_VERSION is too small, newer pkg_install required
...
into ${FILESDIR}. (Using autoconf greatly simplifies maintenance
of this package across multiple platforms.) Copy the sources into
${WRKSRC} during "do-extract". We automagically determine VERSION
by inspecting ${FILESDIR}/configure.
USE_BUILDLINK2=no because this is a pkgsrc bootstrap tool.
and the "do-extract" target copies the sources into ${WRKSRC}. We
automagically determine VERSION by inspecting ${FILESDIR}/lib/version.h.
Thanks to autoconf, all the patches and almost all of the cruftulent
${OPSYS}-specific stuff in the Makefile are gone.
othersrc/bootstrap-pkgsrc/pkg_install into ${FILESDIR}.
Using autoconf greatly simplifies maintenance of this package across
multiple platforms. Placing the extracted source under ${FILESDIR}
will eventually let us use the same source for bootstrapping as
well, though we're not ready for this yet.
- Added ability to tune LOCALBASE and PKG_SYSCONFBASE.
- Deprecated PTHREAD_TYPE variable, not useful any more.
- Multiple error checking (directory presence) added.
- Several bugfixes (specially in removepkgs target).
- Replaced "echo -n" with "printf" (more portable).
- Replaced "! -z" with "-n" when doing tests.
- Changed my email address to the NetBSD one.
inter-dependencies of your installed packages. The "dot" language
is interpreted by the graphviz package to make graphs. There are
several uses for such information.
(1) A graphical representation of information is always good to
look at.
(2) The output itself can be sorted and filtered to provide a list
of packages to delete in order that they can be rebuilt (to replace
out-of-date components).
(3) You can visually estimate the work involved in (or impact of)
removing a given component in order to replace it.
Work around a deficiency in the 1.6 pkg_add - cd to the directory
containing a binary package before running pkg_add, to ensure it can
pick up dependencies.
use buildlink2 and set USE_X11 will use x11-links, or if they will simply
symlink the X11 files directly into ${BUILDLINK_DIR}. If USE_X11_LINKS is
"YES" (the default), then x11-links is used. If USE_X11_LINKS is "NO",
then the X11 files are symlinked directly.
There is a choice here between speed and flexibility. Using x11-links is
faster since you don't need to spend time symlinking all of the X11 files
into ${BUILDLINK_DIR}. However, if you update your X11 installation, then
the x11-links package will likely be out-of-date. Conversely, directly
symlinking the X11 files will protect you from changes in your X11
installation when you update, but there is a (small) time penalty to have
to do it for every package that uses X11.
pkgsrc, require the new version instead of the one from pkgsrc.
This means there will be no more messages that look like
Version mismatch: 'pkg_install' 20020827 vs 20020723
Change reviewed by the wiz. :)
branch. WARNS=2 turns on all kinds of stuff that isn't really appropriate,
and isn't turned on in current. "-Wwrite-strings", in particular, triggers
an error on a bit of harmless code in "file.c". Closes PR pkg/17840.
Use _REQUIRE_BUILTIN_* to specify that the headers/libraries of some parts
of X11 that are available from 3rd parties must be part of the standard X11
distribution or else their symlinks aren't installed as part of x11-links.
This is to allow x11-links to contain all of the parts of the standard X11
distribution.
The problem that prompted this is that newer releases of XFree86-4.x
contain not just new libraries but new inter-library dependencies. Some
of the X11 libraries present in XF86-4.x that were also in XF86-3.x now
depend on libXpm.so or libfreetype.so, which, from a buildlink
point-of-view, must also be found during the link phase when building
software.
Check the return code from the second invocation of installprereq(), to
make sure that a failed installation of a rpe-requisite package kills
the attempt to add the package. This reverts pkg_add(1) to traditional
behaviour.
This update adds a prominant message at the end of the run explicitly stating
which class of NO_BIN_ON_{FTP,CDROM} pkgs have been included or excluded
from the set. This should help prevent accidental license violations.
Suggested by Hubert Feyrer in private email.
- by default cdpack now excludes NO_BIN_ON_CDROM packages. This is the most
important change. Previously all pkgs were included.
- added flag to allow NO_BIN_ON_CDROM packages
- added flag to exclude NO_BIN_ON_FTP packages
- increased the verbosity resulting from the verbose flag
- added a debug flag to preserve tmp files
* when creating BUILDLINK_CONFIG_WRAPPER_SED we also need to create
a corresponding REPLACE_BUILDLINK_SED to reverse the effects
* account for packages that set USE_X11BASE to an empty string.
* create BUILDLINK_CONFIG_WRAPPER_SED patterns if required
* interpret "Makefile.common" as well as "Makefile" for packages
that share common information
issues pointed out in private email by Thomas Klausner:
* Include $PKGVERSION instead of RCS revision in script output
* Add a few line breaks to make the output more readable
* -buildlink-config-wrapper is a hardcoded suffix in
bsd.buildlink.mk. Therefore, make sure the created config wapper
targets actually have this suffix.
* Prevent pkgconfig logic from twice including
pkgconfig/buildlink.mk
* Add missing \" in created sed rules
* Include createbuildlink revision in created files
a buildlink.mk from a package's Makefile and PLIST.
XXX: this is an attempt to make the creation of buildlink.mk files more
XXX: systematic and less error-prone than copying existing buildlink.mk
XXX: files and modifying them for a new package. Any help in making this
XXX: script less simplistic and more versatile is gladly welcome!
