| Age | Commit message (Collapse) | Author | Files | Lines | |
|---|---|---|---|---|---|
| 2008-03-11 | Use std::string instead of PkgIterators (which don't persist over cache ↵ | Daniel Burrows | 1 | -11/+15 | |
| updates, d'oh) to store the set of obsolete packages. | |||||
| 2008-03-11 | Fix the display of the number of newly obsolete packages. | Daniel Burrows | 1 | -8/+8 | |
| The problem is that %d in fragf does not, in fact, substitute an integer argument; instead it disables the DIM attribute. | |||||
| 2008-03-11 | Add checkboxes in the UI for some of the new dependency-handling options. | Daniel Burrows | 1 | -0/+23 | |
| 2008-03-11 | Print a list of newly obsolete packages after a command-line update. | Daniel Burrows | 1 | -8/+45 | |
| 2008-03-11 | When it looks like the user mistyped a package name, use a direct search for ↵ | Daniel Burrows | 1 | -24/+15 | |
| the package instead of using a hand-generated matcher. This avoids having to consider weird cases where the invalid package name doesn't parse as a matcher and we would generate a confusing error message (e.g.: "aptitude install +5"). | |||||
| 2008-03-11 | Add support for installing build-dependencies at the command-line. (Closes: ↵ | Daniel Burrows | 14 | -37/+371 | |
| #243317) | |||||
| 2008-03-11 | Tweak the code to select source packages so that it gives precedence to ↵ | Daniel Burrows | 1 | -10/+46 | |
| actual source packages, and handle source package names that differ from the binary package name properly. | |||||
| 2008-03-09 | Add a ?source-version matcher that matches against the source version number. | Daniel Burrows | 1 | -0/+77 | |
| 2008-03-09 | Add a ?source-package matcher to pick out the binaries of a particular ↵ | Daniel Burrows | 1 | -0/+77 | |
| source package. | |||||
| 2008-03-09 | Factor out the select-a-source-package-and-version logic from ↵ | Daniel Burrows | 3 | -171/+305 | |
| cmdline_changelog.cc. This was a bit fiddly, because the "changelog" action doesn't actually need a source package reference. When it finds one it uses it, but it can also operate in a string-based manner (by relying on the fact that you can look up a changelog just from its name and version). | |||||
| 2008-03-08 | Add support for my new auto-inst-filtering code to make holds actually stick. | Daniel Burrows | 2 | -0/+29 | |
| 2008-03-06 | Allow the user to manually invoke the 'reject hold-breaking actions' logic. | Daniel Burrows | 6 | -6/+86 | |
| 2008-03-06 | When the resolver starts up, automatically reject forbidden versions and ↵ | Daniel Burrows | 3 | -9/+21 | |
| actions that break package holds. The user can override these rejects individually in the usual manner, or globally via Aptitude::ProblemResolver::Allow-Break-Holds. This should make the resolver's behavior a lot more intuitive for users, but it doesn't erase all the bugs on this subject due to the fact that apt's auto-install logic will break holds. This is one of many places (e.g., conflicts handling, reliable choices in safe-upgrade) where I think that apt's auto-install handling is dragging aptitude down. It may be time to replace it for good. | |||||
| 2008-03-05 | Add missing trailing \ns to several error messages so that they're actually ↵ | Daniel Burrows | 1 | -6/+6 | |
| readable. Sorry about that, bubulle :-(. | |||||
| 2008-03-05 | Allow the user to enable the "safe" resolver for other command-line actions. | Daniel Burrows | 3 | -20/+101 | |
| The documentation desperately needs to be improved, but the code works. | |||||
| 2008-03-04 | Ignore more errors that happen while loading the cache, instead of treating ↵ | Daniel Burrows | 2 | -5/+12 | |
| them as failures later. (Closes: #468751) To fix that bug, you also need the last few error-handling changes I made. | |||||
| 2008-03-04 | Don't abort an update if we failed to parse the package lists, since we want ↵ | Daniel Burrows | 1 | -5/+5 | |
| to download new ones! | |||||
| 2008-03-04 | Don't segfault in "update" et al when there's an error parsing the package ↵ | Daniel Burrows | 1 | -0/+3 | |
| lists. | |||||
| 2008-03-04 | Add untested code to support ignoring upgrades in a 'safe' resolution. | Daniel Burrows | 3 | -8/+25 | |
| 2008-03-04 | Split the safe-upgrade logic into the general command-line layer. | Daniel Burrows | 3 | -155/+179 | |
| 2008-03-04 | Add support for adding user tags to the packages modified by a command. | Daniel Burrows | 7 | -2/+172 | |
| I've gone back and forth a few times on the ideal semantics here. It might be nice if the program could automagically guess the packages that the user "meant" to target and only tag up those, but this requires some smarts and I suspect it would be unpredictable (bad!). The current compromise is that implicit tagging (with no pattern saying which packages should be tagged up) will just tag up anything that's being installed, upgraded, or removed; that covers several common cases and is straightforward to implement and understand. Perhaps once this is in the wild I'll have more information about how people use it on which to base an informed decision. | |||||
| 2008-03-04 | Hey, execve works a lot better if you actually NULL-terminate the list of ↵ | Daniel Burrows | 1 | -1/+2 | |
| strings. | |||||
| 2008-03-04 | If updating the debtags database fails, report the actual command that we ↵ | Daniel Burrows | 1 | -4/+4 | |
| ran (including the "update"). | |||||
| 2008-03-01 | Fix the style attribute applied to user-tags. | Daniel Burrows | 1 | -1/+2 | |
| 2008-03-01 | Display user tags next to debtags tags. | Daniel Burrows | 2 | -5/+27 | |
| 2008-03-01 | Add command-line operations to add and remove user tags to packages. | Daniel Burrows | 4 | -0/+188 | |
| 2008-03-01 | Fix bugs that turned up in the user-tags code and do a rename of usertags -> ↵ | Daniel Burrows | 2 | -22/+41 | |
| user_tags. | |||||
| 2008-03-01 | Treat package names containing question marks as search patterns, just like ↵ | Daniel Burrows | 6 | -4/+22 | |
| names containing tildes. | |||||
| 2008-03-01 | Fix the user-tags parser to eliminate an infinite loop. | Daniel Burrows | 1 | -1/+4 | |
| 2008-03-01 | Actually hook up the ?user-tag matcher. | Daniel Burrows | 1 | -5/+7 | |
| 2008-03-01 | Use '?user-tag', not '?usertag', for the user-tag matcher. | Daniel Burrows | 1 | -1/+1 | |
| 2008-03-01 | Add untested backend support for letting the user set tags (arbitrary ↵ | Daniel Burrows | 3 | -26/+401 | |
| strings) on packages, and a matcher to search using these tags. This code is currently inaccessible from the interface; I'm just checkpointing it at the moment. | |||||
| 2008-02-28 | Fix the new safe-upgrade code to actually look past the first solution and ↵ | Daniel Burrows | 1 | -10/+19 | |
| to actually generate upgrades. Previously, not only were we never looking at more than one solution, but that one was being thrown away because the main driver routine would always claim that dependency resolution failed (whoops). | |||||
| 2008-02-28 | Improve the internal error generated when duplicate solutions are encountered. | Daniel Burrows | 1 | -1/+8 | |
| 2008-02-28 | Explicitly warn the user when the resolver produces a solution to ↵ | Daniel Burrows | 1 | -1/+4 | |
| 'safe-upgrade' and then runs out of time trying to find more. | |||||
| 2008-02-28 | Try to maximize the set of installed packages when running "safe-upgrade". | Daniel Burrows | 1 | -7/+51 | |
| This works by mandating all of aptitude's selections at each step, then asking for the next solution; when the resolver converges or runs out of solutions, the last solution that was produced is used. | |||||
| 2008-02-27 | Rename ?configfiles to ?config-files. | Daniel Burrows | 1 | -3/+3 | |
| 2008-02-26 | Add a ?priority matcher. | Daniel Burrows | 1 | -31/+33 | |
| 2008-02-22 | Explain slightly more about what a failed lock means, to hopefully make ↵ | Daniel Burrows | 1 | -1/+1 | |
| things a bit more clear to users. | |||||
| 2008-02-22 | Make ?any-version produce an any-matcher, not an all-matcher. | Daniel Burrows | 1 | -2/+1 | |
| 2008-02-21 | Add a ?bind matcher that's the same as the special binding syntax, but works ↵ | Daniel Burrows | 1 | -0/+27 | |
| for any match syntax, not just ?-matchers. | |||||
| 2008-02-21 | Don't require translators to know that UNINST is to be translated literally. | Daniel Burrows | 1 | -1/+1 | |
| This also would make it easier to translate UNINST in the future. | |||||
| 2008-02-21 | Remove the legacy download_bar and download_screen modules. | Daniel Burrows | 7 | -689/+1 | |
| 2008-02-21 | Don't try to support translated search strings. | Daniel Burrows | 1 | -76/+31 | |
| 2008-02-20 | Change the syntax of explicitly bound matchers from ??VARIABLE: to ?for ↵ | Daniel Burrows | 1 | -49/+69 | |
| VARIABLE: Hopefully this will be a good sight more readable than the previous version. | |||||
| 2008-02-20 | Eliminate another literal ?widen and add another one of Jens's translator ↵ | Daniel Burrows | 1 | -4/+7 | |
| comments, since it's still important that they keep the question marks. | |||||
| 2008-02-20 | Apply string improvements and translator comments from Jens Seidel to the ↵ | Daniel Burrows | 1 | -12/+36 | |
| matcher code. In some cases I edited a string to remove the need for translators to be aware of issues raised by his translator comments instead of adding them, or edited the translator comments to make them more accurate. | |||||
| 2008-02-20 | Fix compile errors in the unused setset class, so that its test passes. | Daniel Burrows | 1 | -18/+2 | |
| 2008-02-20 | Fix a compile error introduced by ee514d358e53. | Daniel Burrows | 1 | -2/+2 | |
| 2008-02-20 | src/cmdline/cmdline_prompt.cc: Add a required trailing newline to a message. | Noritada Kobayashi | 1 | -1/+1 | |
 |
index : aptitude | |
| [no description] | Dyson |
| summaryrefslogtreecommitdiff |