- Pacman-dev - lists.archlinux.org

Re: [pacman-dev] difficult symlink "puzzle"
by ngaba＠petra.hos.u-szeged.hu 14 Jul '07

14 Jul '07

Hi! IMHO this problem is more general and serious, and this should be solved. Either you specify some strict Arch packaging standards to avoid this bug or you improve pacman. I prefer the 2nd, and that is not easy; I make some suggestions here to discuss: First I define the goal: the content of foo's .FILELIST exists on filesystem iff foo is installed. OK, I assume that the goal above holds for all installed package; we just need to keep this during package install/removal. Install: -If we install /dir/file: That exists until package foo is removed (in normal case). If dir is a symlink, we MUST record the "real" destination of file to localdb/foo/files too, because the dir symlink can be removed/modified during an other package removal (or update!!: you can easily imagine a modified perl example where current is a symlink to the latest version...) or by the user. -If we install /dir/dir_symlink: First, we treat dir2 as a regular file, and do the steps above. Then we check if /dir1/dir2 exists. If no, we simply create it pointing to the requested dir. If yes and that is not a real dir, we can stop with file conflict. If yes and dir_symlink is a real dir, we can move its content to the dir_symlink-pointed dir, then delete/create_link dir_symlink; we must also record the file movements to the localdb as above (this is slow, queryfileowner needed, but this a rare case). Remove: -If we remove /dir/file: no extra work needed, just we need to use "real" path. -If we remove /dir/dir_sl: If dir_sl_pointed is empty or doesn't exist, then simply remove dir_sl. If dir_sl_pointed is not empty then we must create a real dir named dir_sl, and moved some files (we scan localdb for files need to be installed as /dir/dir_sl/file) from dir_sl_pointed to there (and record the movements to localdb again.). Bye, ngaba PS: We can implement these without db conversion (and keep compatibility). ---------------------------------------------------------------- This message was sent using IMP, the Internet Messaging Program.

4 40

[pacman-dev] xdelta-ing in pacman 3.1
by Dale Ogilvie 13 Jul '07

13 Jul '07

I for one would like to see this fly. Someone just bumped OpenOffice up a notch, hello Mr 158M download... For me the xdeltas should just be created as far as possible "without maintainer intervention" and stored alongside the package. One could put a size saving threshold on the delta into makepkg, if the delta is over say 50% of the size of the total package, it doesn't get uploaded... Once pacman 3.1 gets into testing I'll look at the code again. I've been running pacman with wget-xdelta.sh for some time now, its sad to see messages like these flash by... Proceed with installation? [Y/n] :: Retrieving packages from extra... --23:48:45-- ftp://mirror.pacific.net.au/linux/archlinux/extra/os/i686/k3b-1.0. 1-1_to_1.0.2-1-i686.delta => `k3b-1.0.1-1_to_1.0.2-1-i686.delta' Resolving mirror.pacific.net.au... 61.8.0.17 Connecting to mirror.pacific.net.au|61.8.0.17|:21... connected. Logging in as anonymous ... Logged in! ==> SYST ... done. ==> PWD ... done. ==> TYPE I ... done. ==> CWD /linux/archlinux/extra/os/i686 ... done. ==> PASV ... done. ==> RETR k3b-1.0.1-1_to_1.0.2-1-i686.delta ... No such file `k3b-1.0.1-1_to_1.0.2-1-i686.delta'. Dale O

2 1

[pacman-dev] Updated hungarian translation
by ngaba＠petra.hos.u-szeged.hu 13 Jul '07

13 Jul '07

Hi! The po files are attached. Some notes: -src/pacman/remove.c:115, add.c:122 "error: failed to add target '%s' (%s)" <- this should be "error: failed to add target '%s' to the transaction (%s)" as earlier (why was this change reverted?) -lib/libalpm/add.c:815: "could not update provision '%s' from '%s'" <- this is not very informative (and needless imho) Bye, ngaba ---------------------------------------------------------------- This message was sent using IMP, the Internet Messaging Program.

2 2

[pacman-dev] multiple extraction from package
by Xavier 13 Jul '07

13 Jul '07

The last commit made me remind of a little issue I wanted to bring a few hours ago, thanks Dan :) grep "\.INSTALL" lib/libalpm/*.c lib/libalpm/add.c: } else if(strcmp(entryname, ".INSTALL") == 0) { lib/libalpm/package.c: } else if(strcmp(entry_name, ".INSTALL") == 0) { lib/libalpm/trans.c: _alpm_unpack(installfn,tmpdir, ".INSTALL"); lib/libalpm/trans.c: snprintf(scriptfn, PATH_MAX, "%s/.INSTALL", tmpdir); The archive is read 3 times, in 3 different places. In particular, the .INSTALL file is checked for existence in package.c , for knowing the package has a scriptlet. In trans.c , the .INSTALL file is extracted to /tmp/ for running the preinstall scriptlet. And in add.c , the .INSTALL file is extracted to $dbpath/local , which is by the way causing some issues : http://www.archlinux.org/pipermail/pacman-dev/2007-July/008693.html I wonder if it wouldn't be possible to read the archive just in one place, and extracting files just once. Or well, I don't know what is the best way, extracting files to /tmp/ and then moving them around (like for backup files), or extracting them several times from the archive (like .INSTALL) , or maybe keeping the special dot files from the archive in memory or something. Well, it's not that important, but it's related to the issue in the link above, that I would like to fix, so I just wanted to know if everyone was happy with the current way pacman reads/extracts files from the archive.

4 5

[pacman-dev] Doxygen Caller Graphs (was: Some doxygen comments.)
by Dan McGee 13 Jul '07

13 Jul '07

On 7/12/07, Andrew Fyfe <andrew(a)neptune-one.net> wrote: > I've been using doxygen to create caller graphs... > http://neptune-one.homeip.net/~andrew/pacman/doxygen/d3/dd7/lib_2libalpm_2a… I can't let you one-up me. :) http://code.toofishes.net/pacman/doc/ That is pretty extensive documentation, with some pretty sizable call/caller graphs. Some of it might have to be cut back at some point (preferably in the code, patches welcome). I generated it using Doxygen and making some changes to our 'stock' Doxyfile. These can be found in my htmldocs branch in its usual place at code.toofishes.net. If you want a local copy of this, I put the tarball up as well (71 MB): http://code.toofishes.net/pacman/doc/pacmandoc.tar.gz -Dan

3 5

[pacman-dev] localized log ?
by Xavier 13 Jul '07

13 Jul '07

Is it a good idea to localize pacman.log also ? Do other apps do this as well ? pacman.log can be handy sometimes for restore/repair purpose, if it was always in english, it would be very easy to parse, but having to take several languages in consideration make it much much harder. I've at least 2 examples from the forums of little scripts that require pacman.log in english : http://bbs.archlinux.org/viewtopic.php?pid=260452#p260452 http://bbs.archlinux.org/viewtopic.php?pid=260961#p260961

4 3

[pacman-dev] [GIT] pacman branch, master now at v3.0.0-300-g1c9f30b
by Dan McGee 13 Jul '07

13 Jul '07

Hello, This is an automated email from the git hooks/update script, it was generated because a ref change was pushed to the repository. Updating branch, master, via 1c9f30b9fab998e2d89dd307a90122618d746cb6 (commit) via 8d62835ba18a7fb662c2f3a23b86a16fe1386388 (commit) via 5ed501565ef8caf4fe2043066750ed7564a2c52d (commit) via 723478560179bc541e89627c43b1cea512fab2a1 (commit) from c0a7d9d82d03d9831f53dbb30357ddb9f10e29ec (commit) - Log ----------------------------------------------------------------- commit 1c9f30b9fab998e2d89dd307a90122618d746cb6 Author: Chantry Xavier <shiningxc(a)gmail.com> Date: Fri Jul 13 15:30:37 2007 +0200 libalpm/deps.c : fix for remove044 pactest. Patch from Nagy that makes removedeps use alpm_depcmp. I also renamed removedeps to recursedeps, as it can have a more general usage, and added an include_explicit argument, so we can control if packages explictly installed are added or not. Note: Small changes made by me (Dan) as well- mostly some English grammar correction and a few other cleanups. Signed-off-by: Chantry Xavier <shiningxc(a)gmail.com> Signed-off-by: Dan McGee <dan(a)archlinux.org> commit 8d62835ba18a7fb662c2f3a23b86a16fe1386388 Author: Chantry Xavier <shiningxc(a)gmail.com> Date: Fri Jul 13 10:41:40 2007 +0200 Remove conflict message for non-existent files (reworked) Signed-off-by: Chantry Xavier <shiningxc(a)gmail.com> Signed-off-by: Dan McGee <dan(a)archlinux.org> commit 5ed501565ef8caf4fe2043066750ed7564a2c52d Author: Chantry Xavier <shiningxc(a)gmail.com> Date: Fri Jul 13 11:06:53 2007 +0200 remove sync022 and sync023 pactest. Currently the replaces flag is only used for forcing the replacement of package (eg because of a renaming) on sysupgrade operation. If the replaces flag should also be used for resolving conflicts, then it should be discussed and specified correctly first. References: http://bugs.archlinux.org/task/7415#comment17207 http://www.archlinux.org/pipermail/pacman-dev/2007-June/008553.html commit which introduced the replace flag : 4fcec8f03ff790bfb23b826eee169bc9c0b00142 Signed-off-by: Chantry Xavier <shiningxc(a)gmail.com> commit 723478560179bc541e89627c43b1cea512fab2a1 Author: Chantry Xavier <shiningxc(a)gmail.com> Date: Thu Jul 12 13:49:23 2007 +0200 libalpm/remove.c : Rsd combination. Currently the d (nodeps) option skips the s (recursive) part, rendering the Rsd combination totally useless. This patch makes a recursive removal still possible using the nodeps option, as Romashka asked there : http://bugs.archlinux.org/task/6057#comment17784 Signed-off-by: Chantry Xavier <shiningxc(a)gmail.com> ----------------------------------------------------------------------- Diffstat: lib/libalpm/conflict.c | 12 +++-- lib/libalpm/deps.c | 104 ++++++++++++++++++++++------------------------ lib/libalpm/deps.h | 2 +- lib/libalpm/remove.c | 32 ++++++++------ pactest/tests/sync022.py | 22 ---------- pactest/tests/sync023.py | 22 ---------- 6 files changed, 76 insertions(+), 118 deletions(-) hooks/update --- Git Source Code Management System hooks/update refs/heads/master \ c0a7d9d82d03d9831f53dbb30357ddb9f10e29ec \ 1c9f30b9fab998e2d89dd307a90122618d746cb6

1 0

[pacman-dev] FS7228: Update {PKGBUILD,install}.proto
by Andrew Fyfe 13 Jul '07

13 Jul '07

Attached is a patch to update {PKGBUILD,install}.proto - Add group and options field - Quote $startdir - Add vim modeline - Add check to see if function exists in install.proto before trying to call it, avoids the annoying bash error if the function is missing. Andrew --- PKGBUILD.proto.orig 2007-03-02 20:42:07.000000000 +0000 +++ PKGBUILD.proto 2007-06-05 14:11:56.000000000 +0100 @@ -11,6 +11,7 @@ arch=() url="" license=('GPL') +groups=() depends=() makedepends=() provides=() @@ -18,13 +19,17 @@ replaces=() backup=() install= +options=() source=($pkgname-$pkgver.tar.gz) noextract=() -md5sums=(generate with makepkg -g) +md5sums=() # generate with 'makepkg -g' build() { - cd $startdir/src/$pkgname-$pkgver + cd "$startdir"/src/$pkgname-$pkgver + ./configure --prefix=/usr make || return 1 - make DESTDIR=$startdir/pkg install + make DESTDIR="$startdir"/pkg install } + +# vim:set ts=2 sw=2 et: --- install.proto.orig 2004-11-03 01:14:37.000000000 +0000 +++ install.proto 2007-06-05 14:11:47.000000000 +0100 @@ -35,4 +35,6 @@ op=$1 shift -$op $* +[ "$(type -t "$op")" = "function" ] && "$op" "$@" + +# vim:set ts=2 sw=2 et:

9 17

[pacman-dev] [PATCH] Cleanup code handling a sysupgrade of pacman itself
by Dan McGee 13 Jul '07

13 Jul '07

Anyone have a problem with this change (Aaron)? -Dan --- The old code attempted to upgrade pacman by itself and automate the process by creating a new transaction and a bunch of other complicated stuff. We should just leave such actions to the user to decide instead of trying to do it for them, and it caused problems when pulling in dependencies. For now, simply request that the user run a 'pacman -S pacman' operation. If they want, they can also answer 'N' at the prompt and the upgrade proceeds as normal. Signed-off-by: Dan McGee <dan(a)archlinux.org> --- src/pacman/sync.c | 19 +------------------ 1 files changed, 1 insertions(+), 18 deletions(-) diff --git a/src/pacman/sync.c b/src/pacman/sync.c index 8982b95..25563e3 100644 --- a/src/pacman/sync.c +++ b/src/pacman/sync.c @@ -544,24 +544,7 @@ int pacman_sync(alpm_list_t *targets) ":: operation. If you wish to continue the operation and\n" ":: not upgrade pacman seperately, answer no.\n")); if(yesno(_(":: Cancel current operation? [Y/n] "))) { - if(alpm_trans_release() == -1) { - fprintf(stderr, _("error: failed to release transaction (%s)\n"), - alpm_strerror(pm_errno)); - retval = 1; - goto cleanup; - } - if(alpm_trans_init(PM_TRANS_TYPE_SYNC, config->flags, - cb_trans_evt, cb_trans_conv, cb_trans_progress) == -1) { - fprintf(stderr, _("error: failed to init transaction (%s)\n"), - alpm_strerror(pm_errno)); - return(1); - } - if(alpm_trans_addtarget("pacman") == -1) { - fprintf(stderr, _("error: pacman: %s\n"), alpm_strerror(pm_errno)); - retval = 1; - goto cleanup; - } - break; + goto cleanup; } } } -- 1.5.2.1

3 3

[pacman-dev] Remove CHOST from makepkg.conf
by Andrew Fyfe 13 Jul '07

13 Jul '07

Are there any objections to removing CHOST from makepkg.conf? It's never seen by package Makefiles as it's not exported, and as a general rule it's a BAD idea to mess with CHOST. If there's no objections I'll do up a patch. Andrew

4 6