summaryrefslogtreecommitdiffstats
Commit message (Expand)AuthorAgeFilesLines
...
* urlparse.urljoin() only works correctly with recognized protocols andZac Medico2008-12-121-1/+4
* When processing output from the "depend" phase, use the number of lines asZac Medico2008-12-122-5/+20
* When attempting to close fd 9 at the end of the depend phase, use exec so thatZac Medico2008-12-121-1/+1
* Bug #250148 - Prevent ebuild.sh subprocess from inheriting file descriptorZac Medico2008-12-121-9/+19
* Bug #250166 - To avoid accidental regeneration of digests with the incorrectZac Medico2008-12-121-21/+29
* When in --debug mode, make the depgraph call self.digraph.debug_print() toZac Medico2008-12-122-5/+13
* For py3k compat, avoid using list.sort(cmp). (trunk r12184)Zac Medico2008-12-121-7/+11
* For py3k compat, remove Package.__cmp__() and related cmp() function call.Zac Medico2008-12-121-8/+1
* Fix deprecated_profile_check() to account for PORTAGE_CONFIGROOT. Thanks toZac Medico2008-12-122-4/+9
* Avoid TypeError in BlockerCache.__iter__() that's triggered withZac Medico2008-12-121-0/+3
* Bug #250212 - Add a new 'upstream.workaround' qa category, and use it forZac Medico2008-12-124-3/+9
* Remove the part about 'link level dependency awareness' which is not supportedZac Medico2008-12-111-2/+0
* Bug #249586 - Use writemsg() to send output to stderr instead of stdout.v2.1.6Zac Medico2008-12-071-10/+15
* If pkg_nofetch needs to be spawned inside fetch() and it happens thatZac Medico2008-12-071-0/+29
* Allow 'system' and 'world' arguments simultaneously since it doesn't hurt.Zac Medico2008-12-061-4/+0
* Add docs for PORTAGE_GPG_DIR and PORTAGE_GPG_KEY that are used by repoman forZac Medico2008-12-061-0/+10
* Validate $ROOT inside the --resume merge list in order to avoid a pottentialZac Medico2008-12-061-0/+12
* Make fixpackages use a new "#" symbol for /var/db updates (previously '*'v2.1.6_rc3Zac Medico2008-12-051-2/+11
* Make fixpackages less noisy by only generting '*' characters for packagesZac Medico2008-12-042-5/+14
* Enable FEATURES=fixpackages by default since the performance is prettyZac Medico2008-12-042-1/+3
* Make sure depgraph._slot_conflict_explanation() doesn't display the sameZac Medico2008-12-041-1/+1
* Bug #249185 - For common cases in which USE deps trigger SLOT conflicts, giveZac Medico2008-12-041-18/+98
* Break references between the depgraph and the dropped tasks insideZac Medico2008-12-031-0/+1
* When --keep-going drops packages due to unsatisfied dependencies, includeZac Medico2008-12-031-0/+1
* It's not safe to use the git commit -a option since there mightZac Medico2008-12-031-0/+23
* When calling `git ls-files -m` to list modified files, use --with-tree=HEADZac Medico2008-12-031-2/+2
* Fix depgraph._dep_check_composite_db._visible() to reject installed packagesZac Medico2008-12-021-7/+7
* Skip sourcing of profile.bashrc during the depend phase. (trunk r12136)Zac Medico2008-12-021-8/+11
* Source bashrc files as late as possible, giving them the opportunity toZac Medico2008-12-023-5/+8
* Fix inconsistencies between the "clean" and "cleanrm" phases. (trunk r12126)Zac Medico2008-12-011-3/+3
* Bug #249185 - For more useful output in cases when one or more USE depsZac Medico2008-12-011-50/+122
* Change depgraph._slot_collision_info() from a set to a dict that containsZac Medico2008-12-011-11/+16
* Add a sanity check inside depgraph._add_pkg() to ensure that the dependenciesZac Medico2008-11-291-1/+5
* Add 'automatically resolved' blockers to the --pretend docs. (trunk r12120)Zac Medico2008-11-291-1/+2
* Use stat st_dev attributes instead of the older approach. (trunk r12117)v2.1.6_rc2Zac Medico2008-11-271-2/+1
* Don't call prepare_build_dirs() inside doebuild() when called for parallelZac Medico2008-11-271-1/+4
* For bug #235642, include the stat mode, uid, and gid bits in the hardlink key.Zac Medico2008-11-261-1/+2
* In movefile(), ignore the hardlink_candidates parameter when it's an empty list.Zac Medico2008-11-261-1/+1
* Bug #235642 - Create hardlinks when merging identical files. This works by us...Zac Medico2008-11-262-7/+58
* Don't direct to /dev/null when creating $EBUILD_EXIT_STATUS_FILE inside die.Zac Medico2008-11-261-2/+1
* Bug #205044 - When creating $EBUILD_EXIT_STATUS_FILE, don't direct output to ...Zac Medico2008-11-261-3/+9
* Make Scheduler.merge() bail out early if PORTAGE_TMPDIR is not properly set.Zac Medico2008-11-261-0/+16
* Add some more clarification to the 'ebuild phase exited unexpectedly' message.Zac Medico2008-11-261-3/+14
* Bug #248603 - Disable warnings about missing repo_name entries for now, since...Zac Medico2008-11-261-1/+1
* Eliminate redundant Manifest separation code by combining mychanged + mynew s...Zac Medico2008-11-261-13/+4
* Remove redundant reference to 'mynew' since 'myupdates' already contains thos...Zac Medico2008-11-261-2/+2
* Bug #248464 - With git, there's never any keyword expansion, so there'sZac Medico2008-11-261-6/+20
* Bug #248782 - Handle permission error in EbuildFetchonly.execute() if PORTAGE...Zac Medico2008-11-261-1/+6
* Note that file names in /etc/portage/package.* directories are sorted ascendingZac Medico2008-11-261-2/+3
* Bug #248599 - Remove docs for the --list-sets option.Zac Medico2008-11-251-3/+0