- Pacman-dev - lists.archlinux.org

[pacman-dev] pacman -U doesn't run ldconfig!
by Nagy Gabor 17 Apr '07

17 Apr '07

Hi! This is a serious bug I think, since -U is the same as -A if the package wasn't installed earlier. I wrote about this in my previous e-mail: http://archlinux.org/pipermail/pacman-dev/2007-April/002610.html (the correct patch can be found here: http://archlinux.org/pipermail/pacman-dev/2007-April/002632.html) I repeat this bugreport, because you may have not noticed in my "big" patch and I think this is easy (and needed) to fix: just replace 'if((trans->type != PM_TRANS_TYPE_UPGRADE) && (handle->trans->state != STATE_INTERRUPTED)) {' to 'if(handle->trans->state != STATE_INTERRUPTED) {' in libalpm/add.c Bye, Nagy Gabor

2 1

[pacman-dev] CVS update of pacman-lib (3 files)
by Dan McGee 17 Apr '07

17 Apr '07

Date: Monday, April 16, 2007 @ 21:49:06 Author: dan Path: /home/cvs-pacman/pacman-lib Modified: etc/pacman.d/mirrorlist.in (1.9 -> 1.10) scripts/makepkg (1.68 -> 1.69) src/pacman/deptest.c (1.19 -> 1.20) Bugfixes in prep for a 3.0.2 release Fix #6905- makepkg dependency checking error Fix #6915- add SUU mirror to mirrorlist.in ----------------------------+ etc/pacman.d/mirrorlist.in | 1 + scripts/makepkg | 3 +-- src/pacman/deptest.c | 2 +- 3 files changed, 3 insertions(+), 3 deletions(-) Index: pacman-lib/etc/pacman.d/mirrorlist.in diff -u pacman-lib/etc/pacman.d/mirrorlist.in:1.9 pacman-lib/etc/pacman.d/mirrorlist.in:1.10 --- pacman-lib/etc/pacman.d/mirrorlist.in:1.9 Mon Apr 9 16:13:31 2007 +++ pacman-lib/etc/pacman.d/mirrorlist.in Mon Apr 16 21:49:06 2007 @@ -5,6 +5,7 @@ # United States Server = ftp://ftp.archlinux.org/@@REPO@@/os/@CARCH@ Server = ftp://ftp.nethat.com/pub/linux/archlinux/@@REPO@@/os/@CARCH@ +Server = ftp://locke.suu.edu/linux/dist/archlinux/@@REPO@@/os/@CARCH@ Server = ftp://ftp-linux.cc.gatech.edu/pub/linux/distributions/archlinux/@@REPO@@/os… Server = ftp://mirror.cs.vt.edu/pub/ArchLinux/@@REPO@@/os/@CARCH@ Server = http://mirrors.easynews.com/linux/archlinux/@@REPO@@/os/@CARCH@ Index: pacman-lib/scripts/makepkg diff -u pacman-lib/scripts/makepkg:1.68 pacman-lib/scripts/makepkg:1.69 --- pacman-lib/scripts/makepkg:1.68 Wed Apr 11 12:41:29 2007 +++ pacman-lib/scripts/makepkg Mon Apr 16 21:49:06 2007 @@ -182,8 +182,7 @@ pmout=$(pacman $PACMAN_OPTS -T $*) ret=$? if [ $ret -eq 1 ]; then #unresolved deps - #strip out the pacman prefix from "requires: xyz" - echo $pmout | sed 's|requires:||g' + echo $pmout elif [ $ret -ne 0 ]; then error "pacman returned a fatal error ($ret): $pmout" exit 1 Index: pacman-lib/src/pacman/deptest.c diff -u pacman-lib/src/pacman/deptest.c:1.19 pacman-lib/src/pacman/deptest.c:1.20 --- pacman-lib/src/pacman/deptest.c:1.19 Fri Mar 9 00:33:06 2007 +++ pacman-lib/src/pacman/deptest.c Mon Apr 16 21:49:06 2007 @@ -81,7 +81,7 @@ } if(!found) { - MSG(NL, _("requires: %s"), saved_target); + MSG(NL, "%s", saved_target); retval = 1; } free(saved_target);

2 1

[pacman-dev] Oh Em Gee, I'm back
by Aaron Griffin 16 Apr '07

16 Apr '07

"Wait, you were gone?" Yeah yeah, I've been kinda AWOL for around a week. Some times real-life hits you like a ton of bricks... I'd go into it, but it's not real interesting, and nothing major happened. It's just the fact that I've been working quite a lot recently, and have suffered from computer burn-out. But I realized something. If I let this get me down, I'll pretty much never get anything done. It's kinda like lifting weights. When you start off, you do a small amount, and get sore, then repeat, and eventually the original weight is easy. That's what I need to force myself into. The more you do, the more you *can* do. But I'm digressing. Anyway, I have a huge amount of emails to go through (~30 or so, not counting flyspray emails) tonight, so if there's anything important that I should be looking at, feel free to ping me so I can get it first and get up to speed. -- Aaron

3 2

Re: [pacman-dev] Fwd: Pacman translation into Polish
by Mateusz Jędrasik 15 Apr '07

15 Apr '07

Hi, It seems the pacman translation itself is all done and finished; There seems to be a minor glitch tho (at least from the ones I noticed), depicted on a screenshot which is available for viewing at http://szklo.eu.org/~imachine/tmp/pacman.png due to its size. Basically, the [###-] progress bar doesn't align properly next to the 'checking file conflicts' line. It's two characters to the left. I've tried simply adding two spaces to the translated text msgstr, but despite being a nasty 'hack' it doesn't fix the issue anyway. It would be great to resolve it. Cheers, //m. -- Mateusz Jędrasik <m.jedrasik(a)gmail.com> tel. +48(51)69-xxx-xx, +44(772)664-2342 http://imachine.szklo.eu.org

2 2

[pacman-dev] Southern Utah University Arch Linux Repository
by Thomas S Hatch 14 Apr '07

14 Apr '07

I am pleased to announce the SUU Arch Linux repository, this server is not only mirroring Arch, but it is running it as well, please add us to the repository and mirror lists lists for pacman. We have set up a full mirror so there are iso downloads as well. [testing] ftp://locke.suu.edu/linux/dist/archlinux/current/os/i686 [current] Server = ftp://locke.suu.edu/linux/dist/archlinux/current/os/i686 [extra] Server = ftp://locke.suu.edu/linux/dist/archlinux/extra/os/i686 [unstable] Server = ftp://locke.suu.edu/linux/dist/archlinux/unstable/os/i686 [community] Server = ftp://locke.suu.edu/linux/dist/archlinux/community/os/i686 and for x86_64 of course! -Thomas S Hatch Southern Utah University

2 2

[pacman-dev] Duplicated requiredby conclusion
by Nagy Gabor 12 Apr '07

12 Apr '07

Hi! Yes, we fixed this problem: We didn't check the existency of the entry before inserting it to the requiredby field. But how could duplicated entries appear? Update is a remove-then-install process, so we simply remove the package first (and removing it from requiredby fields), then install it (which is not installed now, so that is not listed in any requiredby fields). The answer: That remove process is not a real remove transaction (a pseudo transaction is created); look into add.c(389): ==== _alpm_trans_init(tr, PM_TRANS_TYPE_UPGRADE, ... ==== Wow, the transaction type is UPGRADE (this is mainly for logging imho). But look into trans.c/_alpm_trans_update_depends(342): ==== if(trans->type == PM_TRANS_TYPE_REMOVE) { void *data = NULL; rqdby = alpm_list_remove(rqdby, pkgname, _alpm_str_cmp, &data); FREE(data); deppkg->requiredby = rqdby; } else { if(!alpm_list_find_str(rqdby, pkgname)) { rqdby = alpm_list_add(rqdby, strdup(pkgname)); deppkg->requiredby = rqdby; } } ==== So, we want to add it, instead of removing. After adding "if(!alpm_list_find_str(rqdby, pkgname))", this is not so dangerous (but needless), but this a _very odd_ method. However, this is absolutely not safe yet: If the remove was successful, but we cannot install the package, we get false requiredby entries. Bye, Nagy Gabor

3 5

[pacman-dev] Crypto algorithms
by Dan McGee 12 Apr '07

12 Apr '07

We have a rather ragtag set of cryptography functions in use right now, and weird files being used (md5driver? wtf). Looking around I came across these LGPLed hash libraries that we can probably use in pacman. BeeCrypt- http://directory.fsf.org/security/crypt/BeeCrypt.html Documentation- http://beecrypt.sourceforge.net/doxygen/c/index.html Any reason not to go for this? I haven't looked super closely but I think it would make sense to clean this up a bit in order to allow for any future extensibility we may want. These seem like cleanly written algorithms that we can drop in to our code base as they are LGPL. Thoughts appreciated. Have to start cleaning somewhere. -Dan

8 15

[pacman-dev] Fwd: Pacman translation into Polish
by Mateusz Jędrasik 12 Apr '07

12 Apr '07

Hi, I have begun work on translating pacman 3 into the Polish language, supposing there is no ongoing work yet. The files I attach here are alpha at best (done halfway through pl_PL.po, without testing in actual pacman itself since its source build seems to have odd issues with linking pacman.static against libacl, which is, in fact, installed and likely working on my system), the reason why I announce this work is mostly so that other people do not choose to follow my footsteps redoing what is already done by me. I shall post the rest shortly; stay tuned. The attached file goes into src/pacman/po in pacman-3.0.1.tgz source tarfile. Work on the lib/libalpm/po/pl_PL.po file is underway. I have also sent this e-mail earlier however I see it has not yet reached the mailing lists (perhaphs because I wasn't subscribed to it at the time) and, willing to rather avoid somebody starting a Polish translation without knowing about my current efforts, I'm willing to risk being responsible for a duplicate copy on the mailing lists. Regards, //m. -- Mateusz Jędrasik <m.jedrasik(a)gmail.com> tel. +48(51)69-444-90, +44(772)664-2342 http://imachine.szklo.eu.org

3 5

[pacman-dev] Git tree update
by Andrew Fyfe 11 Apr '07

11 Apr '07

Just a quick update on what I'm currently working on... fakeroot - http://repo.or.cz/w/pacman.git?a=shortlog;h=fakeroot Rediffed with recent changes to makepkg (nudge nudge Dan ;P) autoconf_scripts - http://repo.or.cz/w/pacman.git?a=shortlog;h=autoconf_scripts The first stage to removing hard-coding of version numbers and paths. changes so far are: * add autoconf substitution for version number * add -V|--version options to scripts ToDo: * remove hard-coding of config files and paths * merge in gettext changes Andrew

1 0

[pacman-dev] CVS update of pacman-lib/scripts (makepkg)
by Dan McGee 11 Apr '07

11 Apr '07

Date: Wednesday, April 11, 2007 @ 12:41:29 Author: dan Path: /home/cvs-pacman/pacman-lib/scripts Modified: makepkg (1.67 -> 1.68) Fix FS #6872- CFLAGS were not being exported ---------+ makepkg | 4 ++++ 1 file changed, 4 insertions(+) Index: pacman-lib/scripts/makepkg diff -u pacman-lib/scripts/makepkg:1.67 pacman-lib/scripts/makepkg:1.68 --- pacman-lib/scripts/makepkg:1.67 Tue Apr 3 22:03:44 2007 +++ pacman-lib/scripts/makepkg Wed Apr 11 12:41:29 2007 @@ -885,6 +885,10 @@ unset LC_ALL LANG umask 0022 + # ensure CFLAGS and CXXFLAGS are used + export CFLAGS + export CXXFLAGS + #check for "exit on syntax error" shell option echo $SHELLOPTS | grep errexit 2>&1 >/dev/null set_e=$?

1 0