| Age | Commit message (Collapse) | Author | Files | Lines | |
|---|---|---|---|---|---|
| 2010-04-24 | Fix several places where gui.h is included as a nonlocal header. | Daniel Burrows | 2 | -4/+3 | |
| Using <> means the source directory isn't searched, which causes trouble with the scons build, which doesn't put the source directory into the include path. Most of aptitude follows this convention, but these two files have trouble with it. | |||||
| 2010-04-21 | Add SConscript files for the rest of the source. | Daniel Burrows | 2 | -0/+62 | |
| gtk still isn't built since I need to set up its build environment etc. | |||||
| 2010-04-21 | Fix dpkg_terminal to not directly reference log4cxx. | Daniel Burrows | 1 | -2/+2 | |
| 2010-04-21 | Rename the log4cxx namespace to logging. | Daniel Burrows | 8 | -40/+40 | |
| There are various problems with log4cxx, especially that it seems to be unreliable and leads to random crashes. It might be a good idea to eventually move to a different framework or roll our own. Decreasing the number of explicit dependencies on log4cxx is a good way to do that. | |||||
| 2010-03-05 | Overhaul the resolver to distinguish between tiers and changes to tiers; ↵ | Daniel Burrows | 1 | -5/+3 | |
| this is the first step towards being able to minimize the number of changes meeting a criterion. The biggest change for this is that the new cost objects (tier operations) are not totally ordered, so various places that used to take a maximum now take a least-upper-bound instead; similarly for minimum and greatest-lower-bound. Still to do: find a sound way to allow the resolver to add together costs instead of upper-bounding them: "this change will force us to remove 10 packages". The new code seems to be a little too slow -- probably it lost some optimizations by accident. It does seem to be correct, though. | |||||
| 2010-01-17 | Fix a startup crash on some architectures caused by not properly protecting ↵ | Daniel Burrows | 1 | -2/+6 | |
| the call to Glib::thread_init(). (Closes: #555120) | |||||
| 2010-01-17 | Revert some exploratory changes to the Glade file that accidentally got ↵ | Daniel Burrows | 1 | -27/+170 | |
| checked in. | |||||
| 2010-01-13 | Consistently use a signal_ prefix on signals emitted by the area code. | Daniel Burrows | 2 | -12/+12 | |
| 2010-01-13 | Write the standard implementation of the notification interface. | Daniel Burrows | 1 | -0/+40 | |
| 2010-01-13 | Write a prototype for a create_notification() function that creates a ↵ | Daniel Burrows | 1 | -0/+5 | |
| notification (like create_area et al). | |||||
| 2010-01-13 | Explicitly state that set_progress() emits a signal. | Daniel Burrows | 1 | -1/+6 | |
| 2010-01-13 | Rename notification to notification_info. | Daniel Burrows | 1 | -2/+3 | |
| 2010-01-13 | Implement the backing class for tab_info. | Daniel Burrows | 1 | -0/+92 | |
| 2010-01-13 | Make set_tooltip() take a non-const Window pointer, to match the ↵ | Daniel Burrows | 1 | -1/+1 | |
| corresponding Gtk::Widget routine. | |||||
| 2010-01-13 | Make the interesting top classes exported by area.h trackable, to avoid any ↵ | Daniel Burrows | 1 | -4/+4 | |
| nasty surprises down the road. | |||||
| 2010-01-13 | Implement create_area(). | Daniel Burrows | 2 | -0/+8 | |
| 2010-01-13 | Make my terminology more consistent in the doccomments in area.h. | Daniel Burrows | 1 | -3/+6 | |
| 2010-01-13 | Fix the hey-this-is-C++ note in area.h. | Daniel Burrows | 1 | -1/+1 | |
| 2010-01-13 | Make active tab management the responsibility of the view. | Daniel Burrows | 1 | -16/+12 | |
| 2010-01-13 | Implement notification management for areas. | Daniel Burrows | 1 | -0/+36 | |
| This was a copy-and-paste job, which suggests maybe making an abstraction that covers both tabs and notifications when I get a chance... | |||||
| 2010-01-13 | Reimplement the code for managing tabs, based on multi_index with hashing to ↵ | Daniel Burrows | 1 | -9/+58 | |
| avoid any scalability surprises. | |||||
| 2010-01-10 | Switch from html2text to elinks for generating the flat text documentation. | Daniel Burrows | 1 | -170/+27 | |
| This should resolve a lot of complaints from users about encoding issues in the text manual. | |||||
| 2010-01-10 | Finish up the interface for the top-level go-between code and start ↵ | Daniel Burrows | 3 | -11/+349 | |
| implementing it. | |||||
| 2009-12-01 | Start working on data structures to represent the top-level UI structure. | Daniel Burrows | 1 | -0/+202 | |
| The idea here is to give it a model-view-controller flavor, so that the individual views don't get so tightly tied into the main UI's code. That hopefully will make it a bit easier to experiment with radically different UI designs. | |||||
| 2009-11-29 | Fix all VBox, VSeparator and VPaned widgets to have an explicit orientation. | Daniel Burrows | 1 | -0/+12 | |
| Workaround for bug #526030 in glade. | |||||
| 2009-11-29 | Spontaneous changes from running glade. | Daniel Burrows | 1 | -3/+3 | |
| 2009-11-28 | Proof-of-concept support for displaying the status of each changelog ↵ | Daniel Burrows | 4 | -79/+199 | |
| download in the space where the changelog will appear. Ideally this would provide a bit more feedback (download status and so on), but that will require more information to be sent from the download queue to the foreground thread. | |||||
| 2009-11-27 | Move the changelog parsing background thread into ↵ | Daniel Burrows | 1 | -144/+14 | |
| src/generic/apt/changelog_parse. | |||||
| 2009-11-27 | Rewrite the changelog code to be based on aptitude's download queue, not the ↵ | Daniel Burrows | 3 | -360/+259 | |
| download_manager system. This should fix the crashes on shutdown (untested so far), by allowing the background worker threads to be shut down cleanly. It also makes the changelog downloading code a lot cleaner and simpler. | |||||
| 2009-11-20 | Be paranoid about the fact that some architectures think file offsets are ↵ | Daniel Burrows | 1 | -7/+10 | |
| signed. (Closes: #557212) | |||||
| 2009-11-18 | Overhaul how temporary file creation is handled to use a single directory in ↵ | Daniel Burrows | 1 | -2/+1 | |
| /tmp and to try harder to remove everything when the program exits. This is something I've meant to do for a while but only got around to now. It fixes several annoying problems that have only gotten more annoying in the GTK+ version: for instance, you no longer get a ton of entries in /tmp if the program exits unexpectedly (instead, you get only one -- I could try harder to make it so you don't get any, but that could be tricky). It also ensures that on a normal exit, everything is really cleaned up. | |||||
| 2009-11-14 | Eliminate accesses to freed memory when a callback destroyed the object ↵ | Daniel Burrows | 1 | -0/+28 | |
| containing the callback, by taking a strong reference to "this" at stack scope. | |||||
| 2009-11-13 | More informative logging when a weak screenshot cache entry is revived. | Daniel Burrows | 1 | -1/+2 | |
| 2009-11-13 | Modify all the places where weak pointers are changed to strong pointers to ↵ | Daniel Burrows | 1 | -3/+3 | |
| use lock(). lock() is the routine that returns a NULL pointer on failure instead of throwing a fatal exception. | |||||
| 2009-11-12 | Rename the screenshot cache size configuration option. | Daniel Burrows | 1 | -1/+1 | |
| 2009-11-12 | Log when a screenshot cache entry is destroyed, to help debug whether ↵ | Daniel Burrows | 1 | -0/+3 | |
| entries are being leaked. | |||||
| 2009-11-11 | Subject-verb agreement in a comment. | Daniel Burrows | 1 | -1/+1 | |
| 2009-11-11 | Downgrade a log message for a situation that turns out to be OK from WARN to ↵ | Daniel Burrows | 1 | -4/+4 | |
| DEBUG. | |||||
| 2009-11-11 | Remove some trailing whitespace. | Daniel Burrows | 1 | -1/+1 | |
| 2009-11-11 | Don't zero out the size of screenshots that are evicted from the cache. | Daniel Burrows | 1 | -3/+0 | |
| This may have made sense at one time ... now it doesn't. | |||||
| 2009-11-11 | When returning entries from the weak cache, don't connect up their signals ↵ | Daniel Burrows | 1 | -0/+5 | |
| again. Note: it looks like screenshots are never being deleted. May need to look for reference cycles here. | |||||
| 2009-11-11 | Log a message when we can't find a weak cache entry. | Daniel Burrows | 1 | -1/+5 | |
| 2009-11-11 | Don't crash when logging a message about dropping a weak entry from the cache. | Daniel Burrows | 1 | -1/+1 | |
| 2009-11-11 | Log useful information about the computed size of a screenshot. | Daniel Burrows | 1 | -1/+11 | |
| 2009-11-11 | Ensure that we catch changes to the amount of cached pixbuf data caused by ↵ | Daniel Burrows | 1 | -0/+4 | |
| non-incremental loads. Those loads don't invoke area_prepared() before they invoke ready(). | |||||
| 2009-11-11 | Instead of connecting area_prepared to both a cache entry's area_prepared() ↵ | Daniel Burrows | 1 | -2/+7 | |
| method and its signal, have area_prepared() log a useful message and invoke the signal. | |||||
| 2009-11-11 | Merge with changes from laptop. | Daniel Burrows | 2 | -39/+140 | |
| 2009-11-11 | If a screenshot is already in the cache of loaded images, actually use it ↵ | Daniel Burrows | 1 | -1/+9 | |
| instead of waiting for an event that will never happen. | |||||
| 2009-11-11 | When an entry is returned from screenshot_cache::find(), move it to the end ↵ | Daniel Burrows | 1 | -1/+12 | |
| of the LRU list. | |||||
| 2009-11-11 | Fix update_entry_size() to only update the entry's size if it's still in the ↵ | Daniel Burrows | 1 | -11/+11 | |
| cache. Entries that were evicted and replaced shouldn't be updated. | |||||
 |
index : aptitude | |
| [no description] | Dyson |
| summaryrefslogtreecommitdiff |