summaryrefslogtreecommitdiffstats
path: root/urpmi
Commit message (Collapse)AuthorAgeFilesLines
* update copyrightThierry Vignaud2009-11-241-1/+1
|
* make sure we generate dudf reports for transaction errorsChristophe Fergeau2009-10-161-0/+4
| | | | Fixes #53690
* add --download-all options to enable downloading of all packages before installChristophe Fergeau2009-10-131-0/+1
|
* more dudf cleanupsChristophe Fergeau2009-07-301-1/+1
|
* small clean-ups in text messagesChristophe Fergeau2009-07-301-2/+0
|
* dudf patch from Olivier Rosello for the Mancoosi projectChristophe Fergeau2009-07-301-1/+27
|
* Change error code when some packages couldn't be installedChristophe Fergeau2009-03-241-2/+2
|
* Return error code when the install is aborted during an upgradeChristophe Fergeau2009-03-241-2/+2
|
* Revert "urpmi doesn't have a --auto-orphans parameter, fixes bug #48946"Christophe Fergeau2009-03-231-2/+13
| | | | | The parameter was actually implemented and working, dunno how I overlooked that.
* Remove remnants of --auto-orphans codeChristophe Fergeau2009-03-201-12/+2
| | | | | | | | | I recently removed references to urpmi --auto-orphans from urpmi man pages, but in fact, the bug was that the option wasn't parsed, the rest of the code was here. However, given that many people are surprised by which packages get auto-orphaned, it's better if it's not possible for now to remove auto-orphaned packages at install time. This commit removes the auto-orphans code from urpmi
* urpmi doesn't have a --auto-orphans parameter, fixes bug #48946Christophe Fergeau2009-03-201-1/+0
|
* Mark packages as "manually installed"Christophe Fergeau2009-01-131-0/+9
| | | | | | | | With the orphan mechanism, there is no way to manually mark a package that was installed as a dep of another package as not being a potential orphan. Mimic apt and if an already package is passed as an argument to urpmi, remove it from the potential orphans list. Fixes bug #45054.
* add some docPascal Rigaux2008-12-181-0/+3
|
* - urpmiPascal Rigaux2008-11-141-0/+1
| | | | | | o in --auto, do not allow to install a package substring match (you can use -a to force it)
* - urpmiPascal Rigaux2008-11-031-0/+1
| | | | | o ensure download when updating media (with --auto-update) is quiet
* create urpm::msg::ask_yes_or_no() to remove duplicationPascal Rigaux2008-10-211-3/+3
|
* do not call {ask_yes_or_no} callback with explicit "(y/N)" otherwise it isPascal Rigaux2008-10-211-1/+1
| | | | | displayed in GUIs (gurpmi/rpmdrake/...)
* - urpmiPascal Rigaux2008-10-201-14/+1
| | | | | o fix --auto-update ignoring --media and default-media (#45097)
* - urpmi.update, urpmi.addmedia, urpmi.removemedia:Pascal Rigaux2008-10-141-1/+1
| | | | | | | o do not check wether the media are valid, it allows "mv /etc/urpmi/urpmi.cfg.{backup,} ; urpmi.update -a" to work (nb: not equivalent with urpmi.addmedia, since pubkey will not be imported)
* - urpmi, library:Pascal Rigaux2008-10-061-0/+1
| | | | | | o handle --replacefiles, will be used by mdkonline (require perl-URPM 3.19)
* - urpmi, rpmdrake:Pascal Rigaux2008-10-061-2/+2
| | | | | | o nice exit code for "bad signature" fatal error. Fixes rpmdrake continuing on bad signature (#44575)
* when using --bug, copy /root/.rpmdrake tooThierry Vignaud2008-09-171-1/+1
|
* - library:Pascal Rigaux2008-09-111-8/+5
| | | | | o create urpm::select::conflicting_packages_msg() for rpmdrake
* (set_env) extract it so that rpmdrake can support the --env optionThierry Vignaud2008-09-091-14/+1
|
* add local metalink generationPer Øyvind Karlsen2008-07-091-0/+1
|
* - urpmiPascal Rigaux2008-07-081-2/+0
| | | | | o handle --debug-librpm
* o handle "unrequested orphans" (similar to "deborphan")Pascal Rigaux2008-07-071-1/+20
|
* remove unused varsPascal Rigaux2008-07-031-8/+4
|
* nicer var namesPascal Rigaux2008-07-031-3/+3
|
* create db_open_or_die_() and use itPascal Rigaux2008-07-011-1/+1
|
* simplify rpmdb_to_synthesis() prototypePascal Rigaux2008-07-011-1/+1
|
* - urpmiPascal Rigaux2008-06-261-0/+5
| | | | | | | o display a message "Package foo is already installed" when asking "urpmi foo bar" and only installing bar (#41593) (requires perl-URPM 3.16)
* - urpmiPascal Rigaux2008-06-021-5/+1
| | | | | | o fix --replacepkgs when a package appears more than once in urpmi db (#40893) (need perl-URPM 3.14)
* - urpmi:Pascal Rigaux2008-05-191-18/+13
| | | | | | o --auto-update should behave like urpmi.update when mirrorlist is outdated (cf http://forum.mandriva.com/viewtopic.php?t=86837)
* - urpmq:Pascal Rigaux2008-04-031-0/+1
| | | | | | o --suggests is currently misleading, introduce --allow-suggests and explain the user that --suggests really means --allow-suggests (#39726)
* - urpmi:Pascal Rigaux2008-03-171-0/+4
| | | | | o tell the user to "restart system" when it is needed
* - urpmi:Pascal Rigaux2008-03-141-1/+11
| | | | | | o never suggest --install-src for spec file (#38876) o do not allow "urpmi --install-src foo.spec"
* bump copyright yearThierry Vignaud2008-03-131-1/+1
|
* enhance cdrom fallback without perl-Hal-CdromsPascal Rigaux2008-03-111-5/+8
|
* o use hal to wait-for/mount cdroms: Pascal Rigaux2008-02-291-2/+5
| | | | | | | | | | | | | | | | you can now use more than one cdrom drive big commit, no time to split it... some details: - split urpm::removable into urpm::removable and urpm::cdrom - fix setting $urpm->{removable_mounted} - in try_mounting_non_cdroms(), check for local url, simpler than checking is_cdrom_url and then checking file_from_local_medium - fix _try_mounting_medium() not mounting non-cdrom anymore (regression introduced during devel) - even if urpmi says it wants cdroms in a given order, if you make a mistake, it will use the one you give (as a side-effect, it simplifies the sort on "already mounted" cdrom (cf _sort_media disappearing))
* do not pass the device around (we don't want to force a device anymore)Pascal Rigaux2008-02-291-1/+1
|
* - urpmi:Pascal Rigaux2008-02-251-2/+3
| | | | | | o display "%s of packages will be retrieved." (need perl-URPM 3.10 and synthesis built with @filesize@)
* don't use global configs_dir when using --envPascal Rigaux2008-02-251-0/+1
|
* deffine 'message' callbackThierry Vignaud2008-02-251-0/+4
|
* add --debug in usage and man pagePascal Rigaux2008-02-151-0/+1
|
* nocheck_access is quite wrong nowadays afaikPascal Rigaux2008-02-051-1/+0
|
* add --no-suggests to urpmi usagePascal Rigaux2008-01-251-0/+1
|
* add --aria2-options to --help outputPer Øyvind Karlsen2008-01-191-0/+1
|
* - urpmi:Pascal Rigaux2008-01-181-16/+22
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | o --buildrequires deprecate --src, --src is kept for compatibility but its behaviour is changed a little (it doesn't download src.rpm anymore) ie implement what i suggested on cooker mailing list: Date: Thu, 15 Nov 2007 18:21:32 +0100 Subject: urpmi --src issues, suggesting replacing --src with --buildrequires current situation: % urpmi /xxx/foo.src.rpm #=> installs foo.src.rpm buildrequires % urpmi --install-src foo bar #=> downloads foo.src.rpm bar.src.rpm and "rpm -i" % urpmi --help | grep -- --src --src, -s - next package is a source package. % urpmi --src foo #=> urpmi says it will install foo buildrequires + foo.src.rpm #=> in fact, it only installs foo buildrequires weird cases in current urpmi: % urpmi --src foo bar #=> equiv to "urpmi --src foo ; urpmi bar" % urpmi foo --src bar #=> equiv to "urpmi --src foo bar" above :-( % urpmi --src foo --src bar #=> equiv to "urpmi --src foo bar" :-( suggestions: 1) we could make --src a global option: % urpmi --src foo bar #=> would install buildrequires of foo.src.rpm and bar.src.rpm cons: it breaks compatibility (though i'm quite sure noone uses the previous behaviour) 2) deprecate --src, introduce with --buildrequires: % urpmi --buildrequires foo bar #=> would install buildrequires of foo.src.rpm and bar.src.rpm 3) when doing "urpmi /xxx/foo.src.rpm", warn the user that it should use either "--install-src" or "--buildrequires", and that urpmi is defaulting to "--buildrequires" 2) and 3) are implemented now
* fix using urpmi --env where the media are virtual (by removing "virtual" ↵Pascal Rigaux2008-01-171-1/+5
| | | | from urpmi.cfg)