summaryrefslogtreecommitdiff
AgeCommit message (Collapse)AuthorFilesLines
2008-06-21Added tag 0.4.11.5 for changeset 40d8efe8ff72Daniel Burrows0-0/+0
2008-06-21Bump the version number to 0.4.11.5.Daniel Burrows3-2/+32
2008-06-21Refactor the code to strip columnification out of cmdline_search and into a ↵Daniel Burrows3-36/+58
function in cmdline_util.
2008-06-18Automated merge with http://hg.debian.org/hg/aptitude/headDaniel Burrows1-2/+6
2008-06-18Add a default rejection of solutions that would remove Essential packages. ↵Daniel Burrows1-1/+4
(Closes: #486748) This is slightly unideal in two ways that should be fixed eventually. First, the UI is less than desirable if this leads to a situation where nothing can be resolved: the user just gets the message that dependencies can't be resolved, with no indication why; perhaps there should be some explanation that some versions were rejected. Second, the removal of Essential packages will appear in the UI as an alternative resolution, just one that's rejected by default; I'm not sure if this is good or not. Third, the UI should probably continue to do a last-minute check for Essential packages before running an installation; I'm not fixing this right now because it's more complicated to do so (to do it right you should remember which Essential packages the user said it was OK to remove and only remove those, and maybe also remove them from that list if the user reinstates them).
2008-06-18Treat packages with the "Important" flag (meaning apt) like Essential ↵Daniel Burrows1-1/+2
packages when assigning version scores (i.e., give the removal of apt a large negative score).
2008-06-18Dutch translation update. Closes: #486858Jens Seidel2-804/+872
2008-06-10Handle target types in target::description().Daniel Burrows1-0/+2
2008-06-10Add a missing return value for an operator=.Daniel Burrows1-0/+1
2008-06-10Fix signedness errors and handle cmdline_version_curr_or_cand in ↵Daniel Burrows1-2/+17
find_source_package (even though it might not actually be used).
2008-06-10Return 0 if a user-tag operation succeeds.Daniel Burrows1-0/+2
2008-06-10Fix a signedness error.Daniel Burrows1-1/+1
2008-06-10Since the return value of do_cmdline_changelog is never used and isn't ↵Daniel Burrows2-2/+2
meaningful, don't return anything at all.
2008-06-10Fix signedness errors -- they're harmless since none of the numbers involved ↵Daniel Burrows1-2/+2
will ever be large at all.
2008-06-10Explicitly handle all possible enum values.Daniel Burrows1-0/+6
2008-06-10Remove an unused variable.Daniel Burrows1-1/+0
2008-06-10Move the addition of -Wall and -Werror to CXXFLAGS into the toplevel ↵Daniel Burrows7-7/+8
Makefile to ensure they actually get used everywhere.
2008-06-07Added tag 0.4.11.4 for changeset 91f4cc3cf470Daniel Burrows0-0/+0
2008-06-07Update version number to 0.4.11.4.Daniel Burrows3-2/+40
2008-06-07Updates from "make dist".Daniel Burrows3-1589/+1868
2008-06-07Updates from "make dist".Daniel Burrows35-700/+700
2008-06-07Added tag 0.4.11.4 for changeset fbfe1aa99fbfDaniel Burrows0-0/+0
2008-06-07Fix the XSL template that makes literal text bold. (Closes: #473580)Daniel Burrows1-1/+4
In addition to making literal text bold again, this also causes the literal occurrences of ".deb" to not be interpreted as groff escapes.
2008-06-07Automated merge with http://hg.debian.org/hg/aptitude/headDaniel Burrows9-24/+296
2008-06-07If -v is passed to safe-upgrade and there are broken deps, show the set of ↵Daniel Burrows3-7/+25
actions prior to dependency resolution in addition to the usual list of broken dependencies.
2008-06-07Implement the "disable column formatting" feature via a new command-line ↵Daniel Burrows5-15/+125
option, "--disable-columns". (Closes: #141929) This should make 'aptitude search' more useful in scripting contexts or other situations where its output is being used noninteractively.
2008-06-05Update of help-ru.txt by Yuri Kozlov <kozlov.y@gmail.com> (see #483943)Jens Seidel1-3/+47
2008-06-04In the internal resolver self-test, check that the set of packages and ↵Daniel Burrows1-1/+138
versions known to the resolver is consistent with reality.
2008-06-04Automated merge with http://hg.debian.org/hg/aptitude/headDaniel Burrows1-1/+8
2008-06-04Russian translation update. Closes: #483943Jens Seidel2-443/+380
2008-06-02Add a note to the manpage documentation of "aptitude download" indicating ↵Daniel Burrows1-1/+8
that patterns are allowed.
2008-06-02Don't treat self-conflicts as "replacements" when assigning full replacement ↵Daniel Burrows1-0/+15
bonuses. (Closes: #483920) Previously, aptitude would pick treat some arbitrary versions of a package as "full replacements" of that package. This meant they got the rather large "full replacement" bonus and were heavily favored by the solver, leading to surprising outcomes.
2008-05-31Allow patterns as arguments to "aptitude download", and correctly handle ↵Daniel Burrows1-31/+63
non-pattern package names with version specifiers that contain tildes. Thanks to Vincent Labrecque for pointing this out.
2008-05-31Make the definition of \satisifies a bit clearer.Daniel Burrows1-2/+3
2008-05-29Install aptitude-defaults.??_??, not just aptitude-defaults.??. (Closes: ↵Daniel Burrows1-1/+1
#483620)
2008-05-28Fix a link to the Install-Recommends configuration option.Daniel Burrows1-2/+2
2008-05-26Added tag 0.4.11.3 for changeset b5c583a812adDaniel Burrows0-0/+0
2008-05-26Update the version number to 0.4.11.3.Daniel Burrows6-8/+81
2008-05-26Ran update-po.Daniel Burrows3-6356/+6583
2008-05-26Ran update-poDaniel Burrows35-7100/+6834
2008-05-26Add a space between a package name and the Provides indicator with ↵Daniel Burrows1-0/+2
--show-why and -v. When printing out the reasons, aptitude would print things like "FooP<- Bar" to mean that "Foo is provided by Bar". I think that "Foo P<- Bar" is much more readable and fits well with "Foo D: Bar".
2008-05-26Don't die with an assertion error when --show-why is used and an impure ↵Daniel Burrows1-1/+5
virtual package exists. I had the idea that there would never be a chain from a manually installed package to the prospective package-to-be-installed that started with a Provides. Of course this is wrong: the prospective package itself can directly provide a name that's also a manually installed package.
2008-05-17Fix getting changelogs of bin-nmued packages from the curses interface. ↵Daniel Burrows1-2/+5
(Closes: #333468) Apparently I overlooked this when I added SourceVer support at the command-line.
2008-05-17Fix the manpage to talk about Recommends-Important instead of ↵Daniel Burrows1-2/+2
Install-Recommends (Closes: #480533).
2008-05-17Un-reverse the logic in "aptitude changelog" that decides whether to use a ↵Daniel Burrows1-1/+3
source package to find a package's changelog. (Closes: #481458) Previously it was using the source package if and only if there wasn't one available, which is not exactly what is desired and also produces weird error messages.
2008-05-15Norwegian Bokmal translation update, added new help-nb.txt fileJens Seidel3-7/+144
2008-05-16po/ja.po: Update from "1120+10f+15u+106o" to "1145t+106o".Noritada Kobayashi1-33/+29
2008-05-15po/ja.po, help-ja.txt: Make many improvements, including terminological ↵Noritada Kobayashi2-30/+34
unifications.
2008-05-15po/ja.po: Correct a mistake.Noritada Kobayashi1-1/+1
2008-05-15help-ja.txt: Synchronize with the original English version, help.txt.Noritada Kobayashi1-4/+52
generated by cgit v1.2.3 (git 2.39.1) at 2024-07-07 12:45:15 +0000