September 2009 - Pacman-dev - lists.archlinux.org

[pacman-dev] use of /usr/local with pacman
by Jeff 06 Sep '09

06 Sep '09

I just updated the installed pacman and noticed makepkg.conf's use of /usr/local in strip, doc, and man dirs. Should pacman allow any files to be installed into /usr/local? I've been an administrator on various UNIX clones for more than a decade and packages that were installed by the package manager, whether precompiled or a build recipe, always avoided /usr/local and any "local" modifications were put exclusively in /usr/local with PATH and library search path having /usr/local first so the user's modifications would take precedence. Just curious if pacman makes such a policy or if it is up to the PKGBUILD author. -- Jeff My other computer is an abacus.

3 4

[pacman-dev] Yum-presto like
by Matias Hernandez Arellano 05 Sep '09

05 Sep '09

It's posible to use or create something like yum-presto for pacman? -- Matias Hernández (Msdark) www.msdark.archlinux.cl ArchLinux-CL - WikiAdmin MSN: matiasfh(a)gmail.com Twitter : www.twitter.com/msdark Plurk : www.plurk.com/msdark Linux User: #445065 http://keys.gnupg.net/pks/lookup?op=get&search=0xEA3EEC672189419C

4 9

[pacman-dev] makechrootpkg and $PKGDEST
by bardo 05 Sep '09

05 Sep '09

Is it my impression or there are some problems with $SRCDEST and $PKGDEST? I didn't define any of these variables in both my makepkg.conf, the guest one and the chroot one. In makechrootpkg, lines 155-157, I see: --------- source $uniondir/etc/makepkg.conf [ -d "$uniondir/pkgdest" ] || mkdir "$uniondir/pkgdest" if ! grep "PKGDEST=/pkgdest" "$uniondir/etc/makepkg.conf" >/dev/null 2>&1; then echo "Setting PKGDEST in makepkg.conf" echo "PKGDEST=/pkgdest" >> "$uniondir/etc/makepkg.conf" fi [ -d "$uniondir/srcdest" ] || mkdir "$uniondir/srcdest" if ! grep "SRCDEST=/srcdest" "$uniondir/etc/makepkg.conf" >/dev/null 2>&1; then echo "Setting SRCDEST in makepkg.conf" echo "SRCDEST=/srcdest" >> "$uniondir/etc/makepkg.conf" fi --------- So here the chroot's makepkg.conf gets sourced, and it shouldn't inherit any of these variables, but they get forced (in the chroot) to /pkgdest and /srcdest. Problem is, later (lines 240-248) the files don't get moved to $WORKDIR as before, but they go to $PKGDEST, which *should* be empty (because it's the guest system's pkgdest, which got sourced before adding the variables there): --------- if [ -e $pkgfile ]; then if [ -n "$PKGDEST" ]; then echo "Moving completed ${_pkgname} package file to ${PKGDEST}" mv $pkgfile "${PKGDEST}" else echo "Moving completed ${_pkgname} package file to ${WORKDIR}" mv $pkgfile "${WORKDIR}" fi fi --------- I don't understand because they go the /{src,pkg}dest, but it looks like they shouldn't. Even more, there's no directory check on the guest system, so guess what... if the dirs don't exist, the source and dest files get renamed to /srcdest and /pkgdest respectively... This looks like a bug to me, but given how many times I am wrong (especially at 2 a.m.) I'd prefer to wait for some feedback before opening one. Corrado

3 3

Re: [pacman-dev] [PATCH] '-Spp and -Supp' options were added to generate uri list without the downloaded package uris
by Xavier 04 Sep '09

04 Sep '09

On Fri, Sep 4, 2009 at 3:49 AM, Xyne<xyne(a)archlinux.ca> wrote: > On Thu, 3 Sep 2009 15:20:03 +0200 > Xavier <shiningxc(a)gmail.com> wrote: > >> I have a preliminary patch for this. I am not really happy with the C >> code I wrote for doing the substitutions, but it works. >> >> Does it make any sense to have this for -R and -U operations ? I >> implemented that but I don't really see any utility so I wanted to >> drop it. > > I think it makes sense for -R at the moment. It would also make sense > for -U if -U could handle dependency resolution. If you have working > code then I think you should just leave it in for now as it will be > good to have later. > >> Final concern : do you think this feature is useful when pacman asks >> questions and prints a lot of stuff before the actual output ? >> For many operations, the output is quite crap, and I don't know how to fix this. >> >> $ pacman -Rc --print gtk2 >> >> Remove (14): tango-icon-theme-0.8.1-1 midori-0.1.9-1 firefox-3.5.2-1 >> xulrunner-1.9.1.2-1 wireshark-1.2.1-1 mtr-0.75-4 >> libwebkit-1.1.10-2 libsoup-2.26.3-1 libunique-1.0.8-1 >> gconf-2.26.2-2 emacs-23.1-3 librsvg-2.26.0-1 audacious-2.1-1 >> gtk2-2.16.5-1 >> >> Total Removed Size: 279.73 MB >> >> Do you want to remove these packages? [Y/n] >> tango-icon-theme 0.8.1-1 >> midori 0.1.9-1 >> firefox 3.5.2-1 >> xulrunner 1.9.1.2-1 >> >> $ pacman -S --print base >> base package not found, searching for group... >> :: group base (including ignored packages): >> acl attr bash binutils bzip2 coreutils cpio cracklib >> cryptsetup dash db dbus-core dcron device-mapper dhcpcd dialog >> diffutils e2fsprogs file filesystem findutils gawk gcc-libs >> gdbm gen-init-cpio gettext glibc grep groff grub gzip hdparm >> initscripts iputils jfsutils kbd kernel-headers kernel26 >> kernel26-firmware klibc klibc-extras klibc-kbd >> klibc-module-init-tools klibc-udev less libarchive libfetch >> libgcrypt libgpg-error libpcap libusb licenses logrotate lvm2 >> lzo2 mailx man-db man-pages mdadm mkinitcpio mlocate >> module-init-tools nano ncurses net-tools pacman pacman-mirrorlist >> pam pciutils pcmciautils pcre perl popt ppp procinfo procps >> psmisc readline reiserfsprogs rp-pppoe sdparm sed shadow >> sysfsutils syslog-ng sysvinit tar tcp_wrappers texinfo tzdata >> udev usbutils util-linux-ng vi wget which wpa_supplicant >> xfsprogs zlib >> :: Install whole content? [Y/n] >> kernel-headers 2.6.30.5-1 >> tzdata 2009l-1 >> glibc 2.10.1-4 >> attr 2.4.43-1 >> acl 2.2.47-2 >> ncurses 5.7-2 > > > Hmmmm... that's a good point. It might seem superfluous but I think > it's good to keep that. It would enable scripts to handle the dialogue > through a pipe and then easily parse the output so I see a utility in > it. It would probably be less useful for the user directly though, but > even then it would be interesting for some dialogues to see which > dependencies some of the group members will pull in (or cascadingly > remove). Consider the DE groups for example. They should pull in other > packages that are not part of the group dialogue. > > That does not go well with my idea that print operations should imply noconfirm though (see attached patch and commit log). I guess this requires further thinking and discussion. I also forgot to mention that I don't like the group interactivity , but that is another problem.

1 0

[pacman-dev] [PATCH] Feedback from installed package while searching
by djszapi2＠gmail.com 04 Sep '09

04 Sep '09

From: Laszlo Papp <djszapi2(a)gmail.com> To get feedback while searching instead of using another utility for this purpose, whether the desired packages are installed. You can see example for it in case of yaourt. Signed-off-by: Laszlo Papp <djszapi2(a)gmail.com> --- src/pacman/sync.c | 4 ++++ 1 files changed, 4 insertions(+), 0 deletions(-) diff --git a/src/pacman/sync.c b/src/pacman/sync.c index dc93621..9935d6d 100644 --- a/src/pacman/sync.c +++ b/src/pacman/sync.c @@ -330,6 +330,10 @@ static int sync_search(alpm_list_t *syncs, alpm_list_t *targets) printf(")"); } + if (alpm_db_get_pkg(db_local, alpm_pkg_get_name(pkg))) { + printf(" [%s]", _("installed")); + } + /* we need a newline and initial indent first */ printf("\n "); indentprint(alpm_pkg_get_desc(pkg), 4); -- 1.6.4.1

5 8

[pacman-dev] [PATCH] Feedback from installed package while searching
by Laszlo Papp 03 Sep '09

03 Sep '09

To get feedback while searching instead of using another utility for this purpose, whether the desired packages are installed. You can see example for it in case of yaourt. Signed-off-by: Laszlo Papp <djszapi2(a)gmail.com> --- src/pacman/sync.c | 5 +++++ 1 files changed, 5 insertions(+), 0 deletions(-) diff --git a/src/pacman/sync.c b/src/pacman/sync.c index dc93621..3ae6927 100644 --- a/src/pacman/sync.c +++ b/src/pacman/sync.c @@ -330,6 +330,11 @@ static int sync_search(alpm_list_t *syncs, alpm_list_t *targets) printf(")"); } + pmpkg_t *ipkg = alpm_db_get_pkg(db_local, alpm_pkg_get_name(pkg)); + if (ipkg) { + printf(_(" [installed: %s]"), alpm_pkg_get_version(ipkg)); + } + /* we need a newline and initial indent first */ printf("\n "); indentprint(alpm_pkg_get_desc(pkg), 4); -- 1.6.4.1

1 0

[pacman-dev] [PATCH] Feedback from installed packages while searching
by Laszlo Papp 03 Sep '09

03 Sep '09

From: root <root(a)djszapi.localdomain> To get feedback while searching instead of using another utility for this purpose, whether the desired packages are installed. You can see example for it in case of yaourt. Signed-off-by: root <root(a)djszapi.localdomain> --- src/pacman/sync.c | 4 ++++ 1 files changed, 4 insertions(+), 0 deletions(-) diff --git a/src/pacman/sync.c b/src/pacman/sync.c index dc93621..9935d6d 100644 --- a/src/pacman/sync.c +++ b/src/pacman/sync.c @@ -330,6 +330,10 @@ static int sync_search(alpm_list_t *syncs, alpm_list_t *targets) printf(")"); } + if (alpm_db_get_pkg(db_local, alpm_pkg_get_name(pkg))) { + printf(" [%s]", _("installed")); + } + /* we need a newline and initial indent first */ printf("\n "); indentprint(alpm_pkg_get_desc(pkg), 4); -- 1.6.4.1

3 2

[pacman-dev] http://archlinux.org/pacman doesn't work
by Laszlo Papp 03 Sep '09

03 Sep '09

Hello! When will http://archlinux.org/pacman be available ? 404 - Page not found. Best Regards, Laszlo Papp

4 6

[pacman-dev] My repo has moved
by Allan McRae 03 Sep '09

03 Sep '09

With the reinstall of the Arch server, my pacman working git repo has now moved to git://projects.archlinux.org/users/allan/pacman.git Allan

3 3

Re: [pacman-dev] makepkg -i should fail when packages cannot be installed
by Allan McRae 03 Sep '09

03 Sep '09

Pierre Schmitz wrote: > Hi, > > I would file a bug report but somehow... > > It seems makepkg -i just throws a warning if it cannot install the package > after build. I would prefer if it would at least return something different > from 0. > > My scripts assume that everything is fine if makepkg retuns 0 and an inter > package file conflict wasn't noticed due to the behaviour that makepkg did > return 0 even if something went wrong. > Hmmm... there was a bug report that was the opposite. The reasoning here is that makepkg did not fail, rather pacman failed to install the successfully built package. The same occurs when pacman fails to remove deps as asked by makepkg. Allan _________________________________________________________________ Use Hotmail to send and receive mail from your different email accounts. http://clk.atdmt.com/UKM/go/167688463/direct/01/

5 5