December 2006 - Pacman-dev - lists.archlinux.org

[pacman-dev] rankmirrors script
by James Rosten 17 Jan '07

17 Jan '07

I have merged testing a single mirror into rankmirrors. Testing a mirror file is still the same usage. To test a mirror just do: rankmirrors -u ftp://archlinux.puzzle.ch/current/os/i686 or rankmirrors --url ftp://archlinux.puzzle.ch/current/os/i686 I also added a test to make sure the mirror filename provided actually exists. The patch is below. ~ Jamie / yankees26

2 1

[pacman-dev] pacman3 logging issues
by Dan McGee 12 Jan '07

12 Jan '07

Just thought I'd send this email before I forgot about it. Here is a log from my pacman3 -Syu yesterday: http://pastebin.archlinux.org/548 The aspell upgrade alone was logged 9 times...so I'm thinking there is a bit of an issue here with the way log entries are being written to the file. I haven't looked at the logging code enough yet to know what is going on, but I thought I'd at least present a concrete example to the list. Happy holidays to everyone! -Dan

3 5

[pacman-dev] pacman-optimize updates
by James Rosten 11 Jan '07

11 Jan '07

Ok, after testing it, some changes needed to be made and so they were made. Now it works (at least for me it worked). This is just an idea, but wouldn't it be interesting if we used the stat_busy, stat_done, and stat_fail stuff from /etc/rc.d/functions for the messages in pacman-optimize? Anyway, here is the patch for FS #5009 (woops, just caught another bug.........it wouldn't be able to work after a successful one but it would work after the failure ('cause pacman.new was still haning around.........fixed). It is below as usual. ~ Jamie / yankees26

3 9

[pacman-dev] makepkg fix (FS #5923)
by James Rosten 06 Jan '07

06 Jan '07

This is a fix for Flyspray bug #5923 http://bugs.archlinux.org/task/5923. Patch is below. ~ Jamie / yankees26

3 2

[pacman-dev] Long functions and other minor fixes
by James Rosten 03 Jan '07

03 Jan '07

This is my massacre of _alpm_load_pkg in package.c (just joking about the massacre). I moved the extremely large for loop into its own function (and made it a while loop). Feel free to change the name of _alpm_pkg_checker (first thing that came to mind). The patches for package.c and package.h are below. ~ Jamie / yankees26

3 4

[pacman-dev] My $0.02 on the Frugalware v.s. Arch bit...
by Scott Horowitz 02 Jan '07

02 Jan '07

I hate to wade into this, but let me just say... there's no doubt in my mind that this will never be resolved while the (inflammatory) sentiment in the pacman-g2 README exists. For starters, the whole point being made about the API not being stable, which seems to be points 1 and 2 in the README, are plain silly. No project should take a project's cvs development code and expect the API to remain stable; Frugalware knew what it was getting into when it grabbed this code to put into their repositories. There's a reason that pacman3 hasn't been released yet on Archlinux, and that there haven't been Archlinux frontends developed. Pacman3 development shouldn't be hindered because someone decided to release the code into the wild, and it's clearly going under some large refactoring. As for point 3, Aaron's supposed disinterest in contributor's patches - I think it has been quite clear from the mailing list that this isn't the case. The real justification behind this sentiment, which was brought up in the original email and I agree needs improvement, is that there could use more openness (such as in response to a WONTFIX patch) between all the parties. Then there's the stuff about how development has fallen behind schedule (what project hasn't?) and Aaron asked Frugalware to fork things (I'm pretty sure this was more tongue-in-cheeck about how Frugalware is doing things than anything). These seem pretty petty to me. In my opinion, two things need to happen, one from each side. On the Archlinux side, I think Aaron should make a better attempt to be more transparent in his responses, which would hopefully encourage debate. On the Frugalware side, I think the whining about API stability needs to be dropped and, rather than take the "fork" talk as a challenge, they should be asking what needs to be done to make the two paths more compatible. Happy Holidays, Scott

7 11

[pacman-dev] pacman-optimize updates
by James Rosten 31 Dec '06

31 Dec '06

I came up with a patch for FS #5009 (the error when /var/lib/pacman is a loopback device or other mounted partition. I have not been able to test it yet, as I am on my mac (so sending this with Mail, so let us hope it doesn't mess this up :P), but in a few hours I'll be on my Arch computer and I'll test it then and obviously report back. Also, fixed the step number in the comments ('cause I am that nit pickish), because there were two step 1's. ~ Jamie / yankees26

1 0

[pacman-dev] Long functions and other minor fixes
by James Rosten 30 Dec '06

30 Dec '06

I've refactored _alpm_checkconflicts, but there appears to be something slightly off. With a normal cvs build of pacman3, trying to install sonata-svn with normal sonata produces this: loading package data... done.checking dependencies... :: sonata-svn conflicts with sonata. Remove sonata? [Y/n] y error: failed to prepare transaction (conflicting dependencies) :: sonata-svn: conflicts with sonata But with my patch: loading package data... done.checking dependencies... done. checking for file conflicts (1/1) [###############################################################] 100% error: failed to prepare transaction (conflicting files) sonata-svn: /usr/bin/sonata exists in filesystemsonata-svn: /usr/lib/python2.4/site-packages/mmkeys.so exists in filesystemsonata-svn: /usr/lib/python2.4/site-packages/mpdclient3.py exists in filesystemsonata-svn: /usr/lib/python2.4/site-packages/mpdclient3.pyc exists in filesystemsonata-svn: /usr/lib/python2.4/site-packages/sonata.py exists in filesystemsonata-svn: /usr/lib/python2.4/site-packages/sonata.pyc exists in filesystemsonata-svn: /usr/share/applications/sonata.desktop exists in filesystemsonata-svn: /usr/share/locale/de/LC_MESSAGES/sonata.mo exists in filesystemsonata-svn: /usr/share/locale/fr/LC_MESSAGES/sonata.mo exists in filesystemsonata-svn: /usr/share/locale/pl/LC_MESSAGES/sonata.mo exists in filesystemsonata-svn: /usr/share/locale/ru/LC_MESSAGES/sonata.mo exists in filesystemsonata-svn: /usr/share/pixmaps/sonata.png exists in filesystemsonata-svn: /usr/share/pixmaps/sonatacd.png exists in filesystemsonata-svn: /usr/share/pixmaps/sonatacd_large.png exists in filesystemsonata-svn: /usr/share/sonata/CHANGELOG exists in filesystemsonata-svn: /usr/share/sonata/README exists in filesystemsonata-svn: /usr/share/sonata/TODO exists in filesystemsonata-svn: /usr/share/sonata/TRANSLATORS exists in filesystem errors occurred, no packages were upgraded. As you can see it flipped out, but even with the normal pacman3 it still fails to install. Both were executed with: sudo ./pacman.static -A /var/abs/local/sonata-svn/sonata-svn-246-1.pkg.tar.gz Using -U instead of -A yields the same results. So below is the patch for you guys to look over and see what might be screwing this up (cause I have no idea). ~ Jamie / yankees26

1 0

[pacman-dev] Long functions and other minor fixes
by James Rosten 30 Dec '06

30 Dec '06

After trying to build pacman3 so I could test something for stonecrest, I found some errors in my patch and Dan's (very small, you forgot to put the label error before return(-1);) while compiling. So below is the better patch than the other patch (which should be ignored). It patches package.c, package.h, and sync.c of libalpm. ~ Jamie / yankees26

1 0

[pacman-dev] Long functions and other minor fixes
by Dan McGee 29 Dec '06

29 Dec '06

Some more functions that are worth taking a look at on the libalpm side: add.c: _alpm_add_commit be_files.c: _alpm_db_read _alpm_db_write conflict.c: _alpm_checkconflicts _alpm_db_find_conflicts package.c: _alpm_pkg_load (getting a bit long) remove.c: _alpm_remove_commit server.c: _alpm_downloadfiles_forreal sync.c: _alpm_sync_sysupgrade (getting there) _alpm_sync_prepare _alpm_sync_commit And some minor things I found in the code- cleaned up an extern enum, changed 'md5' variable to 'hash' as it can be either md5 or sha1, resorted db.h for clarity, and fixed some English grammar issues. Here is a patch: Index: lib/libalpm/alpm.h =================================================================== RCS file: /home/cvs-pacman/pacman-lib/lib/libalpm/alpm.h,v retrieving revision 1.64 diff -u -u -r1.64 alpm.h --- lib/libalpm/alpm.h 22 Nov 2006 09:03:41 -0000 1.64 +++ lib/libalpm/alpm.h 22 Dec 2006 18:34:01 -0000 @@ -399,7 +399,7 @@ /* * Errors */ -extern enum __pmerrno_t { +enum __pmerrno_t { PM_ERR_MEMORY = 1, PM_ERR_SYSTEM, PM_ERR_BADPERMS, @@ -469,7 +469,9 @@ /* Downloading */ PM_ERR_CONNECT_FAILED, PM_ERR_FORK_FAILED -} pm_errno; +}; + +extern enum __pmerrno_t pm_errno; char *alpm_strerror(int err); Index: lib/libalpm/backup.c =================================================================== RCS file: /home/cvs-pacman/pacman-lib/lib/libalpm/backup.c,v retrieving revision 1.7 diff -u -u -r1.7 backup.c --- lib/libalpm/backup.c 31 Oct 2006 06:39:59 -0000 1.7 +++ lib/libalpm/backup.c 22 Dec 2006 18:34:01 -0000 @@ -55,9 +55,9 @@ ptr++; /* now str points to the filename and ptr points to the md5 or sha1 hash */ if(!strcmp(file, str)) { - char *md5 = strdup(ptr); + char *hash = strdup(ptr); FREE(str); - return(md5); + return(hash); } FREE(str); } Index: lib/libalpm/db.h =================================================================== RCS file: /home/cvs-pacman/pacman-lib/lib/libalpm/db.h,v retrieving revision 1.20 diff -u -u -r1.20 db.h --- lib/libalpm/db.h 20 Nov 2006 09:10:24 -0000 1.20 +++ lib/libalpm/db.h 22 Dec 2006 18:34:01 -0000 @@ -44,10 +44,14 @@ pmlist_t *servers; }; +/* db.c, database general calls */ pmdb_t *_alpm_db_new(char *root, char *dbpath, char *treename); void _alpm_db_free(void *data); int _alpm_db_cmp(const void *db1, const void *db2); pmlist_t *_alpm_db_search(pmdb_t *db, pmlist_t *needles); +pmdb_t *_alpm_db_register(char *treename, alpm_cb_db_register callback); + +/* be.c, backend specific */ int _alpm_db_install(pmdb_t *db, const char *dbfile); int _alpm_db_open(pmdb_t *db); void _alpm_db_close(pmdb_t *db); @@ -58,7 +62,6 @@ int _alpm_db_remove(pmdb_t *db, pmpkg_t *info); int _alpm_db_getlastupdate(pmdb_t *db, char *ts); int _alpm_db_setlastupdate(pmdb_t *db, char *ts); -pmdb_t *_alpm_db_register(char *treename, alpm_cb_db_register callback); #endif /* _ALPM_DB_H */ Index: lib/libalpm/sync.c =================================================================== RCS file: /home/cvs-pacman/pacman-lib/lib/libalpm/sync.c,v retrieving revision 1.85 diff -u -u -r1.85 sync.c --- lib/libalpm/sync.c 1 Dec 2006 09:32:30 -0000 1.85 +++ lib/libalpm/sync.c 22 Dec 2006 18:34:03 -0000 @@ -382,7 +382,7 @@ { pmlist_t *deps = NULL; pmlist_t *list = NULL; /* list allowing checkdeps usage with data from trans->packages */ - pmlist_t *trail = NULL; /* breadcrum list to avoid running into circles */ + pmlist_t *trail = NULL; /* breadcrumb list to avoid running in circles */ pmlist_t *asked = NULL; pmlist_t *i, *j, *k, *l; int ret = 0;

2 4