* Add a check for direct use of sysctl(8). [1]
* Add a check for use of the deprecated macro INSTALLS_SHLIB, and point people
to USE_LDCONFIG instead. [2]
* Add a check for plists with less than 6 items, and point people to
PLIST_FILES instead. [3]
PR: 100472 [1]
91933 [2]
100190 [3]
Submitted by: Babak Farrokhi <babak@farrokhi.net> [1]
This commit should largele be a NOOP as it only adds support
for DESTDIR undefined. This does allow us to start testing
ports with DESTDIR set, but this is as of yet not supported.
Although this has been extensively tested on pointyhat, this
is a very intrusive change and some cases may have been
overlooked. Please contact Gabor and me if you find any.
PR: 100555
Submitted by: gabor
Sponsored by: Google Summer of Code 2006
- from Changes
- Following some useful e-mail discussions with Mark Linimon
and Dan Langille, switch to using the MASTER_PORT variable
to detect master/slave relationships between ports.
MASTER_PORT will be set in every slave port whereas
MASTERDIR doesn't have to be. Here a slave port is one
defined as using port metadata from some other port, and
doesn't have to go as far as requiring both master and slave
to use the same Makefile and other ports infrastructure.
- Keep track of the version of the software used to create the
cache, and prompt for a cache rebuild if the software has
been upgraded since. A cache rebuild is definitely required
for the 1.6 -> 1.7 update.
- Don't quit on errors parsing make variables at the cache-init
or cache-update stage. Print copious error messages, but carry
on processing. Entries for ports that fail at this stage will
not be created in the cache.
- Don't quit on errors parsing make variables at the cache-init
or cache-update stage. Print copious error messages, but carry
on processing. Entries for ports that fail at this stage will
not be created in the cache.
PR: 100759
Submitted by: maintainer (Matthew Seaman)
- move pkg-message to files/pkg-message.in and use SUB_FILES
- move post-extract to post-patch and use REINPLACE_CMD
- replace ${PREFIX} in man-page with %%PREFIX%% and then REINPLACE_CMD
over it.
- Make replacement in former post-extract actually work due to
changed layout of the file.
PR: ports/100584
Submitted by: Edwin Groothuis <edwin@mavetju.org>
Approved by: cperciva@#bsdports
OpenBsd-Ports, PkgSrc and maybe others.
You can browse and search through your Portstree and perform actions
like Upgrades, Installs, Uninstall a.s.o. More advanced features like
setting Options and Vulnerability checking are also available. For a
complete list, check the Homepage or the Changelog.
Functionality is KPorts' main goal, not simplification by reducing
options; however KPorts should be rather easy to use.
WWW: http://kports.sf.net
PR: ports/99653
Submitted by: Hannes Hauswedell <hannes.hauswedell@gmail.com>
(MFD - Merge From Developing)
* Fix a bug with a time delta. When build took over 24 hour.
* Allow origins in ALT_PKGDEP.
* Add -e (--emit-summaries) option to show summary messages.
* Fix portupgrade working when stdin is closed. [1]
- Get back sf.net to MASTER_SITES
- Add CONFLICTS with portupgrade-devel
PR: ports/99179 [1]
Submitted by: Hannes Hauswedell <hannes.hauswedell_at_gmail.com>
It's exactly like 2.1.3 but without summary messages and
fixed bug when BDB4 was defined but tried only bdb1 driver and
fall back to dbd driver.
- Add an option WITH_BDB1. Both BDB1 and BDB4 options can be off.
Ruby dbd driver will use in the case.
- 97406 (ghostscript refactoring) was removed from the -exp run due to errors
- 95841 (remove legacy code) was backed out after commit, due to its still
being needed for kdelibs3 among others
This is a bug fix and general improvement release. From the change log:
- Incorporate local ports/categories into the INDEX by scanning Makefile.local
for additions to $SUBDIR. Mentioned on freebsd-ports@freebsd.org by
Rong-En Fan <rafan (at) infor.org>
- Clean up error messages, and factor out common progress counter code.
- Add a new Category object type, and cope better with changes to categories
(ie. ports being hooked up / unhooked from the tree). Problem pointed out by
Colin Percival <cperciva (at) freebsd.org>
- Instead of calling perl to compile exactly the same bit of perl code from
/usr/ports/Mk/bsd.ports.mk about 15,000 times (ie once per port) as a
consequence of actually running 'make describe', just call perl once, use make
to print out the basic variables from the port and replicate the processing
done ourselves.
PR: 99454
Submitted by: author/maintianer
=============
1. Add -b option to make a backup package of the installed port
2. Add -g option to make a package of the new port
3. Add -o option to replace an installed port with a new port from a
different location (e.g., portmaster -o emulators/linux_base-fc4 linux_base)
The second argument can also be specified as the name of the installed port
(e.g., linux_base-8-8.0_15)
4. Add -s option to detect no-longer-needed dependencies, and offer the user
a choice to delete them.
General improvements:
=====================
1. Update usage() to reflect reality, and organize it better
2. Switch to using DEPORIGIN to detect missing dependency tracking, which
provides dramatically fewer false positives, and more useful tools to fix
actual problems
3. Factor out more common code into functions
4. Only pay attention to +REQUIRED_BY files for dependency tracking if they
are > 0 bytes
5. Simplify option parsing
6. Make whitespace more consistent in -a message
7. Only grep for dependencies if we're upgrading an installed port
8. If using -o, or a port moved, try to find the names of the old distfiles,
and offer the user the choice to remove all of the, or prune the stale ones.
Bug fixes:
==========
1. Fix the PATH if a user is using ccache
2. Handle the case where there is no ORIGIN in a +CONTENTS file
3. Ignore errors in the tail of fetchlog, since if the file is gone we'll
exit the loop anyway
* Add -e (--emit-summaries) option and show summary messages only when
the option defined or verbose mode is on. [1]
* Allow origins in ALT_PKGDEP. This announced in pkgtools.conf
but did not work really. [2]
* Run 'make config' before all operations unless -j (--jet-mode)
option specified. [3]
* Add lock on operations with pkgdb.db. Now you can safe run a few
portupgrade(1).
* Incorrectly calculated upgrade time when it is more 24 hours. [4]
* Other fixes and impovements.
PR: ports/99086 [4]
Reported by: Doug Poland <doug at polands.org> [3,4]
Requested by: many [1]
Submitted by: skv [2]
- allow to depend on csup (default) or cvsup [2]
- default database is now mysql [2]
This port still needs a lot of work and it would be great if one of the wizards
that know tindy's insides would take maintainership (current maintainer seem
to be gone, no feedback rom him for > 1 month).
PR: ports/97433 [1]
Submitted by: Shaun Amott <shaun@inerd.com> [1]
itetrcu (me)
Approved by: lawrance (mentor)
o Add ALT_INDEX array to pkgtools.conf. The array holds additional
INDEX files. Useful for local categories.
o Add to pkgtools.conf a new dirrective: include_hash('glob').
It downloads keys and values (splitted by '=>') from files
coincided with 'glob' and returns a filled hash. The glob is
related to PREFIX. [1]
o Add a summary messages on each upgrade/install transaction:
how many tasks and how many task done. [2]
o Make portupgrade(1) with -D option really remove a broken distfile.
But it works only if FETCH_CMD set to use fetch(1) (it's default). [3]
o Bugs fixes
Requested by: skv [1]
netchild [2]
Andriy Gapon <avg at icyb.net.ua> [3]
port descriptions given the port's name or a glob. Basically to view
the descriptions of GCC (related) ports you type
portless gcc\*
and portless will display the descriptions of all gcc* ports.
There are options to display the postinstall messages or Makefiles instead.
- Martin Kammerhofer
mkamm@gmx.net
PR: ports/98739
Submitted by: Martin Kammerhofer
o Really fix *_rc_scripts. The last fix was uncomplited.
o Set both make argument and environment variable to PORT_UPGRADE=yes.
It makes possible a port or a package (via install/deinstall scripts)
to detect if it builds/installs/deinstalls under portupgrade(1) [1]
o Add to pkgtools.conf a new directive: include_eval('file')
The file will included and evaluated in the place where encountered.
The file path looking inside of PREFIX. [2]
o Add to pkgtools.conf ALT_MOVED. The array holds alternate MOVED file.
E.g. for files in EXTRA_CATEGORIES. [3]
o Allow install port by origin. E.g. portinstall lang/ruby18
o Make pkgdb offer install a stale dependency before selecting it from
installed. [4]
- Add own mirror to prevent delay while distfile appears on MASTER_SITE_LOCAL.
Requested by: skv [1,2,3], kris [4]
o Fix -O that should force ignoring dependency errors,
but really did not work [1]
o Add PKG_BACKUP_DIR environment variable where specified
a directory where will keep saved packages (-b).
Default: PKG_PATH (PORTSDIR/packages/All) [2]
o Add MAKE_ENV in pkgtools.conf that is like MAKE_ARGS but
allows to set environment variables for ports [3]
o Fix pkg_fetch(1) to guess a right source directory for
download (All or Latest). It works in much more cases now [4]
And make a notice for using of '@' in package name more clean.
o *_rc_scripts internal procs don't recognize RC-scripts without .sh.
It's wrong for 6.0+ [5]
o Fix pkgdb -s [6]
o Add firefox in a browser list in portcvsweb(1)
o Revive NEWS file
PR: ports/56301 [1]
ports/81454 [2] (based on)
ports/92896 [3]
ports/94054 [4]
ports/95775 [5]
ports/96832 [6]
Reported by: Roman Neuhauser <neuhauser at sigpipe.cz> [1]
Fernan Aguero <fernan at iib.unsam.edu.ar> [3]
Andrew Webster <awebster at connectalk.com> [4]
Rong-En Fan <rafan at infor.org> [5]
Richard Brooksby <rb at ravenbrook.com> [6]
Submitted by: KOMATSU Shinichiro <koma2 at lovepeers.org> [2]
- Update to 2.1.0
o rewrite version checking. Now it's compliant with pkg_version(1) algorythm
o fix shebangs in test/*
o add test in test_pkgversion.rb from test-pkg_version.sh
o fix test_pkgdb.rb
o fix pointyhat URL in pkgtools.conf and pkgtools.rb
o make comparisons when updates from packages. It compared package names
instead versions. It gives a weird results when port was renamed
(eg. mozilla-firebird -> firefox) [1]
o don't show a warning about origin changing when origins are the same. [2]
It rids us from messages like:
Package origin of 'pan' has been changed: 'news/pan' -> 'news/pan'
PR: ports/62990 [1], ports/98153 [2]
Submitted by: Leland Wang <llwang at infor.org> [1]
tobez [2]
Approved by: maintainer (silence for a month, no activity seen for 4 monthes)
daichi (blessed)
ports (silence)
a number open PRs from 2003 year
Inspired by: az
* The check for quoted COMMENTs had a typo. However, this check has been
consolidated for other macros that should not be quoted. Merge the COMMENT
check into the BROKEN, et al check. [1]
* Properly pluralize the message which counts the number of fatal errors
and warnings. [2]
* Add distinfo checking support. [3]
* Huge refactoring of the error handling code. If the -g argument is specified,
errors will be grouped together to avoid spamming the console with
repeated lines.
* Add a check for icon installation, and recommend INSTALLS_ICONS.
PR: 97207 [1]
96089 [2]
96923 [3]
Submitted by: lawrance [1]
ehaupt [2]
sem [3]
Tested by: mezz
1. Update DEPORIGIN in addition to @pkgdep entries in +CONTENTS
2. Give more details about why a port moved
3. grep for the old port name in existing +CONTENTS files, since the
new name will not appear at all
Also, stretch the window to wait for a background 'make checksum' to
complete to 5 seconds instead of 3
easier to leave it running unattended for large updates. (This can be
toggled off.)
2. Implemented a completely unattended mode which accepts the defaults for
all portmaster dialogues. (Note, this does not help with ports that have
interaction other than 'make config', such as ghostscript, etc.)
3. Added -u to sort(1) calls to try and address a bug report about duplicate
lines in +REQUIRED_BY files. I was not able to reproduce this bug, but
better safe than sorry.
4. Verbose mode is now slightly more informative about what's happening.
5. More common code factored out into functions.
6. The -n flag now cancels the build of the port called on the command line too.
7. Improved (?) the stale distfile deletion routine to better handle ports
that have multiple distfiles.
8. We now launch 'make checksum' (which performs a fetch if needed) in the
background while recursing through dependencies.
9. Various code cleanups, such as adding -t to a mktemp invocation that was
missing it, more local variables in functions, more consistent flag variable
values, etc.
Don't restrict Apache version to 1.3
The csup change won't work now, and the port should be updated - as promised by
the maintainer 3 weeks ago.
PR: ports/94288
Submitted by: itetcu (me), maintainer
Reviewed by: maintianer, jmelo, sem
Approved by: maintaner, lawrance (mentor)
* bugfix: do not ignore option for case-insensitivity when searching
for packing list files
* added "-b something" option to simulate -f '/something$', since I got
tired of typing the later
There are two patches in the PR mentioned bye pkg-message, but
this tool depends on the first one and doesn't work with the
second one. This should be clearly mentioned.
PR: ports/96100
Submitted by: Gabor Kovesdan <gabor.kovesdan@t-hosting.hu>
package is installed or not using a precalculated regex.
This speeds up "portaudit -a" with around a factor of 10.
The change is slightly modified from the one from the PR by using
pkg_info -aE instead of ls /var/db/pkg for determining installed
packages.
Submitted by: Kuang-che Wu <kcwu@csie.org>
PR: ports/92942
